HABITUAL PHYSICAL ACTIVITY AND THE ASSOCIATION

WITH DISEASE SEVERI'N AND EXERCISE CAPACITY IN

CYSTIC FIBROSIS: A PILOT STUDY

by

Susan L. Pollock

A thesis submittcd in confonnity with the requhments for the degrce of Master of Science

Graduate Department of Community Health University of Toronto

The aushor has granted a non- exchisive licence allowing the National L h u y of Canada to npioduce, loan, distnite or sel1 copies of this thesis in microform, p a p or electronic formats.

The author retains ownership of the copyright in this thesis. Neither the

may be printed or othawise teproduceci without the author's permission.

L'autnn a accordé me licence non exchisive permettant A la BibIiotMque nationale du Canada de np.oduke, p*, dimiiuer ou vendre des copies de cette thbe sous la fonne de microficbeI/nhn, de reproduction sur papier ou sin format 61ectronique.

L'auteur conserve la propnCtC du droit d ' a m qui protège cette thèse. Ni la thèse ni &s exbraits substantiels de celie4 ne doivent être imprimés ou auûement reproduits sans son autorisation.

Master of Science - 2000

Swan L Pollock

Graduate Department of Community Health

University of Toronto, Canada

Objectives: To detennine rccnlltment and compliance rates for the collection of habitual physicd

activity data, to profile trends in habitual physical activity patterns, and to detennine if a positive

relationship exists bctween lung function, habituai physical activity, nutritional status and

exercise capacity in cystic fibrosis (CF) patients.

Methods: For this cross-sectional pilot study, 40 CF patients were recdted from the CF clinic

at the Hospital for Sick Children over a three month period. Patients providecl a self-report of

their habitual physical activity patterns using two suwey instruments: the Habitua1 Activity

Estimation Scale (HAES), and the 7-Day Activity Diary. Rilmonary function and nutritional

status data were coliected from clinic records. Data on exercise capacity were available for a

subset of patients who completed their annual exercise test on or near to 4 3 months) the study

recrui tment date.

Resulis: High necmitment (89%) and compliance ( M E S , 100%; Acrivify Diary. 82%) rates

indicateâ the feasibility of collecting habinial physicai activity data in this study population.

Positive and sipficant correlation coefficients were demonstraîed bctween habituai physical

activity (HAES), lung hinction, nuiritional status and e x d s e capacity. Patients reportcd

significantly higher Total Activity scores (mild to vigorous physicai activity) for a typical

weekend day (HAES, 8.0 f 3.0 hodday; Activity Diaty, 4.8 f 2.3) than weekàay (HAES, 5 7 k

2.8; Actiwity Diary, 339 f 1.2). The Total Activt'iy score for a typicd weekday, deriveci fiom the

HAES, was demonstcated to be a sigeificant predïctor of fmed expiratory volume in one second

(Fm), the most important vaiabk in desctibhg lmg discase.

Co11cIiwio119: These pilot study rcsults will sme as the prceursor for a longitudùial follow-up

study that will begin to aâ&ess the direction of the causal ralaîiortship beniveen habitual physical

activity and FEVI in CF.

ACKNOWLEDGEMENTS

To Mary Corey, my supervisor, who provideci me with both opporiunity and inspiration. T h a h

for always reminding me of the "big pictm" and for allowing me to find my own way.

To Jane Schneiderman-Walker, my fnend and committce member who shand my vision and

who taught me about the importance of "pruning." Thanks for naduig and n-reading my thesis

and for never king shy to voice your opinions!

To Jacek Kopec and Joe Reisman, members of my cornmittee who provided me with helpfd and

invaluable feedback throughout the various phases of my thesis.

To the CF patients and their families who generously àonated their time to this study. Thanks for

compkting your Activity Diaries with such enthusiasm!

To the staff of the CF clinic who offereâ their assistance to me whenever called upon. Thank-you

for a i i of your invaluable help.

To Petet Katzmarzyk, Susan Jagial and Mary Chipman for nading and conûibuting to my thesis.

To Janey, whose emails and conversation entertained me at the worst and best of times, and

whose honesty and friendahip contnbuted to the successfur completion of this project.

To Lenny, for her tales of Ted and Ca~per, and for her contribution of dog pictures to the office

wds. Thanks for rnaking sure thaî I wasn't foqetting to eat, and for aii your advice.

To the nst of th team in Rwm 5406-thanks f a your support.

To aîi my fncnds who never gave up on m. Thenks for un&rstanding my absence durlng the

last few months.

To my Epidemiology coileagues, especiaIly Ali. Sue, Micheiie and Laurie, it's been fun! I

couldn't have asked for better classrnates.

To mom, dad, Heather, Marc, Ali, Trev, Grandpa, Maddie and Natasha, for believing in me.

Thank-you for instilling in me a love of learning and for teaching me the value of perseverance.

To Darren, for his love, support, and sense of humor, and for providing me with welcome

diversions fmm my work. n i a t h for helping me to open my eyes and see Life beyond my thesii.

CONTENTS

2 BACKGROUND mea.ea-..-....a ..ma..nama.ma..anna.nwM*a*w a . 4

2.1 CYSTIC FIBROSIS ........... .+ ... .. ......... ......... ...... * ........ *.* .............. ............................... O ................... 4

2 3 PHYSICAL A C ï M T Y IN CHILDREN AND YOUTH .................................................................. 8

.................................................................................................................... 2.2.1 HEALTH BENEFWS 8

........................................................................ 2 3 3 PATTERNS IN CHROMCAUY IW, PATIENTS 9

............................................................... 2.23 EXERUSE CONSIDERATIONS IN CF PATIENTS 9

................................................................................. 2 3 EXERQSE TRAINING PRûûRAMS IN CF 12

23.1 CHARACERiSTICS OF EXERUSE PRûGRAMS .................................................................. 12

.................................................... ............... 23.2 SUPERVISED PROGRAMS ....................*................ 13

2A3 UNSUPERVISED PROGRAMS ................................................................................................... 16

23.4 CHALLENGES OF TRAINING PRûGRAMS .. ............................ .............................................. 17

2.4 RELATIONSHIP BETWEEN HABïïüAL PHYSICAL ACMWTY. LUNG FUNCTION.

NUTRPfIONAL STATUS. AND EXERCISE CAPACITY IN CF ....................................................... 18

2 3 . OF PHYSICAL ACTMTY ................. ........ .................................................... 20

25.1 DOUBLY LABELED WATER .................................................................................................... 20

2.5.2 OBSERVATION ........................O ..................................... .............................................. 21

2.53 MOVEMENT ASSESSMENT DEVICES .... ........C..tt.....t-..t... .... .....,,. 21

......................... 23.4 HEART RATE MOmORING........ ..*e*.*+.o.*+.**.C...*.*...***.*H.**t*... ..... 22

3 M O W - ..m...mm.m~.m*.*o.*no~~omeom~.em~~~~mm.m~.m.muH~em...mm.a.~o.*~ ......m...... 32

3.1 OVERVIEW OF THE STUDY DESIGN o.................. ....................... ...........................O............... 32

3.2 PRACITCAL PLANNING AND IMPLEMENTATIOEJ .......................................................... 32

... .................................................................................... 33 SUBJECT SELECTION....... ......r.....r+....r.. 34

33.1 ELlGlBILSrY CRï ïERk ............................................................................................................ 34

33.2 SUBECT RECRUITMENT ...................................................................................................... 34

3.4 DATA COLLECïION ..................................................................................................................... 35

3.4.1 SUBJECT DATA .................... ............*......., .................................................................... 3 5

3.42 PULMONARY FUNCTION ......................................................................................................... 36

3.43 MEASUREMENT OF HA8lTUAL PHYSICAL A C T M T Y ..................................................... 37

3.43.1 THE HABlTüAL ACTMTY ESTIMATION SCALE (HAES) .................................... . . . . 37

3.432 THE 7-DAY ACI'NITY D M ................................................. t,........ ..................... 38

3.4.4 ADMINISTRATION OF THE HAES AND THE A C T M W DIARY ............................O................ 38

3.43 NUTRIMONAL STATUS .. .................*..*..m.........-.. ............................................................... 39

....... 3.4.6 EXERCISE CAPAQTY ..r..........r.i..r~.~ttt~..~..~..*~~~*~.......~~....*~......*-*......*-..**..~....***... 40

3.4.7 COMPLIANCE WïïH PHYSIOTHERAPY r ~ ~ ~ ~ ~ . H ~ . ~ ~ ~ ~ ~ ~ ~ . ~ ~ o ~ o H ~ o ~ ~ ~ t . ~ o o C ~ .... ..H.*o..w.oH*.**-.*-*. 42

3.43 POTENTIAL CORR5ATES OF PHYSICAL ACTMTY PARTIQPATION** ...c.**..t,Ce**.. 42

................................................................................................................................. 3.5 S m SIZE 44

3.6 DATA ANALYSIS ........... ...... ................* ....*..*...........*.*..................*...........-.. .......*...*............. 44

306.1 DATA SCORING .......................................................................................................................... 44

3.6.1.1 THE HAES ................................................................................... ....................... O.... 4 5

3.6.l.2 THE ACMWTY DïARY ................... , ..................................................................................... 45

3.6.13 ACïiVïîYTYPE ......................*............................................................................................... 46

3.6.2 DATA QUALiTY CONTROL ...*.......................... ,., ......*............................................................. -46

3.63 DESCRPTNE STATISTICS ..........*......... ....., ............................................................................. 48

... 3.6.4 PEARSONS CORRELAïïON ..................*...*............* ..,. 48

3.6.5 REGRESSION ANALYSES .............................................................. .................................... 4 9

3.6.5.1 MULTIPLE CORRELATION ................ .....,..*..*..*.....*....... ..................................................... 49

3.6.5.2 MODEIJNG FEVI AS OUTCOME VARIABLE ...................*......................................... 50

4.1.1 PATIENT CHARACi'ENSTICS ..........................~............................................................... 5 1

4Al.l Si'üDY PNUiciPANïS .*...*+....*....*...*..o......... W.* .....l..t..**...t.........* ..C...*.t....l..*..*l.*.*..H..*. .*..*...O 51

4.1.1.2 CF POPULATION VALUES ........... .................................................... ......... 5 1

..... .... ...... 4.1.13 STüDY REFüSALS ...... ...................... ..... ».. .......................................................... 52

4.l.2 MEASUREMENT OF HABiTüALPKYSICAL ACTIVITY .......*...... .........U....*................ 53

............................... 4.l.2.l THE HAES .....................*..*... ............ ..... ............. ................ 53

................................. 4.l.2.2 THE 7-DAY ACTIVfTY DfARY ... .....HH.......H...........C......e.H..C.....).. 55

4.133 A COMPARISON OF DATA FROM TEE HAES AND THE ACTfVITY DIARY .........-.... 58

4.13 PULMONARY FüNCTION .......................... ....... ...............O........... 58

4.1.4 ANNUAL EXERCISE TESTIN0 ................................................................................................. 59

................................................................................ 4.1A.i A SUB4 AMPLE OF STUDY PATIENTS 59

4.1.43 ASSESSMENT OF EXERCISE CAPACITY ....................................................................... 5 9

4.1.S COWUANCE W ï ï H PHYSlOTHERAfY ................................................................................ 6û

4.1.6 PARENT QUESMONNAIRE ....................................................................................................... 6û

............................................................................................ 4.1.7 A C T N ï ï Y CLASSiFiCATION 6 2

4.1.73 PERSONAL CARE ................................................... ... .......... .. ............................................... 63

4.1.73 DOMESTIC WORK .+................................................................................................................. 63

4.1.73 SCHOOL.RELATED. ............................................................................................................... 6 3

4.l.7.4 EMPLOYMENT .......... r.c ............................................................................................... 6 3

4.1.73 UNSTRUCX'URED PLAY ..................... .. ............................................................................. ...... 64

4.1.7.6 WALKING .......... .........,.+..*.+.... .................................................................................................. 64

4.1.7.7 SPORT ........................................................................................................................................ 65

4.1.7.8 UND- ................................................. ... ......................................................................... 65

4 3 CORRELATION ANALYSES .......................................O..... t., ........................................................ 72

4.2.1 THE HAES AND THE ACrrVPrY DiARY ...........* ....... .... .......... ... 72

43.6 A SUMMARY OF ReSULTS FROM CORRELATION ANALYSES .......................* ................ 80

4 3 REGRESSION ANALYSES ...... ................... .................................................................................. 81

43.1 MUTLIPLE CORRELATION ...........*o.H.... .... te ... .............. 82

4A1.1 SüMMARY OF MULTIPLE CORRELATION ....................................... ...... 84

4 3 2 MODELJNG FEVi AS THE OUTCOME VARIABLE ............................... ......... 93

4 . 1 A G ............................................................................................................................................ 94

432.2 PERCENT OF DEAL WEIGHT ............................................................................................... 94

43.23 HAES ACTMTY SCORES ...................................................................................................... 95

432.4 MULTIPLE VARIABLE M O D U .......................................................................................... 95

........................................... 43.25 SUMMARY OF FINAL MULTIPLE REGRESSION 97

S D I S ~ ~ I O N w m m m m ~ m m ~ m m m m m ~ w m ~ * m m w m m m n m - m m ~ m n ~ m m m m w w m m m ~ * m m ~ * ~ ~ m + n m ~ m ~ ~ * * m w * m w m m m ~ m m m m m w ~ m m w w w 99

5.1 INTERPRETATION OF RESULTS ............................... ............................................................... . 99

5.1.1 FEASIBILlW ............................................................................................................................... 99

5.13 HABïïüAL PHYSICAL ACMWTY ......................................................................................... Iûû

5.l.2.l TRENDS FOR THE HAES AND THE A C T M T Y DIARY .................................................. 1ûû

5.13.2 COMPARISON OF THE HAES AND THE A C î M T Y DIARY ........................... ...,... ..... 102

S.13 EXERCISE CAPACl'iY ........................... ..*.... ........................................................................... 103

5.1d C O M P W C E ~PHYSIOTHERAPY ......... ................ .. ........., ..................................... 104

5.l.S PARENT QUESTlONNAIRE.. ................................................................................................... 104

5.1.6 ACMViTY QIPLSStmCATION ................. .....r.*.r.+................ ...................................... ....... *.o.... 1M

5.1.7 SIMPLE CORRELATION OF ACI'MTY SCORE WlTH FEVt .. ..............n...-.....H.......t.. -- 106

SA8 REGRESSION ANALYSES ~ ~ ~ * ~ ~ ~ ~ n ~ ~ ~ ~ ~ ~ ~ b ~ o ~ ~ ~ o ~ ~ ~ ~ ~ * * ~ ~ ~ ~ ~ o ~ * * ~ ~ ~ o * ~ ~ ~ b ~ ~ ~ ~ ~ ~ * ~ ~ I ~ * ~ ~ ~ ~ ~ ~ - ~ ~ ~ ~ o ~ ~ C C ~ ~ ~ ~ ~ ~ m m ~ Io8

................................................................... .............................. 5.2 METIIODOLOGICAL ISSUES W.* 108

53.1 STUDY LIMPTATIONS .................... ......... .......................................................................... 108

.................... S.Zl.1 COMPLIANCE WITH, AND RETURN OF. THE 7-DAY ACTIWI'Y D m 109

52.1.2 RANDOM AND SYSTEMATIC ERROR IN SELF-REPORT OF A C T M T Y ..................... 109

........................................................................... 5.2.13 GENERALEABIlSrY OF THE RESUtTS 110

53.1.4 DIRECTION OF CAUSALSI"Y ......................................................................................... ..... . 111

5.23 SUGGESTED MODIFICATIONS OF THE CZTRRENT PROTOCOL. ................................... 111

5 3 FUTURE DIRECTIONS ......................................,........ ................................................................. 112

....................................................................... 53.1 PROPOS AL FOR A LONGITUDINAL STUDY 112

5.4 CONCLUSIONS ............................................................................................................................ 114

REFERENCE LIST ............................................................................................................................ 116

LIST OF TABLES

.......................... TABLE 3.1 SUMMARY OF THE SURVEY INSTRUMENTS ..,,,...................... 43

TABLE 3.2 ACTMTY SCORES COMPJLED FROM THE HAES AND TEIE ACTMTY

D M Y ...............-.........*........*.......*...*.......... .......................................................................................... 47

.........*................... .......*.......................** TABLE 4.1 STUDY PATIENT cf fCTERiST ICS ...... 5 2

TABLE 4.2 CF PATIENT CHARACTERISTICS FOR ALL ELIGlBLE PATIENTS ATTENDMG

...................................................................................... CLINIC FROM MARCH 9-JUNE 4. 1999 5 2

TABLE 4.3 A C T M T Y HOURS COMPILETl FROM THE HAES .................................................... 55

TABLE 4.4 ACTlVITY HOURS COMPiLED FROM THE ACTMTY DMRY ............................... 56

TABLE 4.5 PERCENT OF PREDICTED PWONARY FUNCTION .............................................. 58

TABLE 4.6 EXERCISE PARAMETERS.* ....... , ....................... ............................................. 6 0

TABLE 4.7 CORRELATES OF PHYSICAL ACTMTY PARTICIPATION .............. 62

TABLE 4.8 CORRELATION BETWEEN THE TOTAL AND HXGH A C T M T Y SCORES OF

THE KAES AND THE ACTMTY DIARY ......................................................................................... 72

TABLE 4 3 CORRELATION FOR WEEKDAY A C T M T Y SCORES OF THE HAES AND THE

A m DIARY ...... ............t.......t.....t..- 74

TABLE 4.10 CORRELATION FOR WEEKEND ACMVïW SCORES OF THE HAES AND THE

ACT'MTY DLARY ...m. ...*..****.+.****".*.*+.*...*.......+.*.*....*.*.*..**.....*..r.......~t...*.......~.....t*........*...~.*..**....**.. ...O 75

TABLE 4.11 CORRELATION FOR WEICiHTED A m SCORES OF THE HAES AND

THE ACTMTY DIARY .........................O................ ...o.............w.......w..........w...................~......*.t........ 76

... TABLE 4.12 CORRELATION BETWEEN WEEKDAY AND WEEKEND A~IVITY SCORES n

TABLE 413 CORRELATION ANALYSIS FOR FEVl .........,...,,,...........u.~e.~~..e..n..~.....CH . . 78

TABLE 4.14 CORRELATION BETWEEN ACTMTY SCORES AND POTENTIAt

CORRELATES OF PEZYSICAL ACTMTY PARTiCJPAmON ..................................................... .... 79

.................................... TABLE 4J5 MULTIPLe CORRELATION FOR THE ACTMTY SCORES 87

TABLE 4J6 MULTIPLE CORRELATION FOR FEV,. WITH DIFFERENT ACTMTY SCORES

INCLUDED AMONG THE CANDIDATE PREDICTORS ............................... ..,.. ..., ... 88

TABLE 4J7 MODEUNG FENc AS THE OUTCOME VARIABLE .................. .+ ............................. 98

LIST OF FIGURES

FIGURE 4.1 AVERAGE ACTMTY LJZvELS FOR THE HAES ........ -.. ...........~....C.....CI....... . 5 4

............. FIGURE 4 3 AVERAûE ACMVRY LEWU FOR THE ACTMTY DIARY .................. .. 57

........................................................ FIGURE 4.3 A C T I W N CLASSWICATION (CATEGORY O).. 66

FIGURE 4.4 ACXïVïR CLASSIFICATION (CATEGORY S).. ................... .......H...*.........*...**..... 67

.......................................................... FIGURE 4.5 A C T M T Y CLASSIFICATION (CATEûûRY 6) 68

FIGURE 4.6 A C T N R Y CLASSIFICATION (CATEGORY 7) ................. ...... ................................... 69

FIGURE 4.7 ACMWTY CLASSIFICATION (CATEGORY 8) ...................................... . 70

FIGURE 4.8 ACTNiTY CLASSIFICATION (CATEGORY 9) ................* ............. .........,... 71

FIGURE 4 9 PLOT OF FEVl BY AGE ................................................................................................ 89

FIGURE 4.10 PLOT OF FEVl BY PERCENT OF IDEAL WEIGHT ..................... ....... .......... ........... W

FIGURE 4.11 PLOT OF FEV, BY HAES WEEKDAY TOTAI, ACMVITY ..................................... 91

FIGURE 4.12 PLOT OF FEVl BY HAES WGHTED TOTAL ACTMTY .................................... 92

LIST OF APPENDICES

A ASUMMARY OFEXERCISE~VENTTONSTUDfES INCE5 ........................ .., ............. 124

B CONSEM: AND ASSENT FORMS ............... - ......... ..................................................................... 130

C GLOSSARY OF TERMS AND ~BREVIATIONS ............................................ 138

D THE HABïI'UAL, ACTMTY ESTIMATION SCGLE (HAES) ................ .. .................................. 140

E THE 7-DAY ACTNLTY DIARY .................................................................................................... 147

F PARENTQUESTIONNAiRE ....................... ..... ............................................................................ 153

G PûWER CALCULATION .............................................................................................................. 158

H SUMMARY OF ACTMTY TYPES .............................................................................................. 159

INTRODUCTION

1.1 Sinnificance

Cystic fibrosis (CF) is a chnic , inheriteà disorder chamcterized by progressive lung

disease in the majority of patients. Physiotherapy is preacn'bed as a conventioaal treabnent of

pulmonary impairment caused by lung infection and mucus obstruction. The goal of chest

physiotherapy is to improve or maintain lung function. Exercise may also ùenefit pulmonary

function, and has ken proposad as a supplementary therapy in CF.' An exercise training

program of an appropriate frequency, intensity, type and duration may psitively affect the

pulmonary status of CF patients through the facilitation of sputum expectoration,' and an

increase in respiratory muscle endurance? Patient cornpliance with a long-term urercise training

program, however, is difficult to maintain? Romoting healthy lifestyle changes in CF patients

may be a more practical means of incnasing physical activity levels. A longitudinal study is

needed to &termine whether the habituai physical activity level of CF patients is a significant

p d c t o r of the rate of change in pulmonary function.

Prior to the onset of a long-term investigation, an appropriate measunment of habitual

physicai activity is necessary for the CF population. The cumnt pilot study was conducted to

memm habitual physicd activity in a sample of CF patients at The Hospital for Sick Children

('SC). Two instMnents were used to capture this infinmation: the Habituai Activity Estimation

Scde @AES) and the 7-Day Activity Diary. Both insm~aeats provide data on the fkequcncy.

intensity and duration of activity. The Acîivity Diary, however, provides additional information

on the type of activity king paformod.

CF patients at HSC are ftbsuently overwhelmed by requests for their participation in

research studies. One important goal of this pilot study was therefore to calculate the recniitment

and compliance rates for the collection of habitual physical activity data in order to predict the

likeiihood of success in patient recnlltment and compliance for a long-term study. Tmds in the

relationship of pulmonary func tion with habitual ph ysical activity level, nutri tional status, and

exercir capacity were also examined These results were intendecl to encourage integration of

the regular collection of habitual physicai data into clinic management, generate further

hypotheses, and guide the design of hitue clinical trials of exercise as therap y.

For this pilot project, we hypothesized that a positbe relationship would be &monstrated

between pulmonary jiuacîion (forced expiratory volme in one second, Fm/), habitual physical

activity (Activiîy score), num'tional statu (percent of ideal weight) and excrcise capacity

(nrcu~*ml work capaciiiy or onygen consumption). We pmposed that CF patients with high values

for FEVl would have high activity scons, in addition to high percent of idcal weight and

maximai work capacity or oxygen consumption values. Although it is plausible that a causal

relationship ktween lung function and habitual physical activity could work in either direction.

the issue of causality could not be addresseci in this cross-sectional pilot study.

1.2 Obiectives

The objectives of the cumnt study were:

Rimsry

I . To detennine recruifment and conipliimce rates for the coilection of habituai physical

OCtMity dotajbm the CFpopulation,

2. To qvm~lfy tkfie~uetlcy, intensity, type Mddtumfa of hubitual pIrysical activity in CF

chiIriren.

3. To compare the datofrom two &ffermt rneasumnents Qhabincat physicd llctivity: the

7-Day Activity Dimy and the Habirual Activity Estimation Scde (HAES).

4 To detennine if a positive relationship exists between pulmnary fkncrion. habitua1

physical activity, nutnional stm and exercise capacity.

-adUr

5. To examine potential correlates of physical activiîy partz*cpati*on in CF children.

BACKGROUND

2.1 Cvstic Fibrosis

Cystic fibrosis is a chronic disease which is usually manifested in early childhood, and is

characterized by an abnormal transport of chloride ions across the epitheiial üning of multiple

surfaces inclwüng the nasal cavities, lower aimays, exocrine pancreatic ducts, and sweat ducts?

Clinical manifestations of CF include chronic obstructive pulmonary disease. pancreatic

insufficiency, and elevated sweat electrolyte levels? Although CF is the rnost common life-

shortening, autosomal recessive genetic disease in Caucasian populations~ this incurable disease

is much less common in Afncan and Oriental races? 1n Caucasian populations, the carrier rate is

appmximately 1 in 20. and the incidence is estitnated at 1 in 2500 live bas.'

CF was fmt recognized as a disease in the 1930's, at which time many patients âied in

their fmt year of life? Subsequent developments in CF research and treatment have dramatically

improved the iife expectancy of these patients h m 11 yem in 1960 to 3 1 years in 19977

hproved prognosis can be attributed to extensive international research efforts. At HSC in

Toronto. the gene nsponsible for CF was located in 1989.8 This discovery has enabled scientists

to consider gene therapy as an eventual tnatment for CF. Rennements in therapy, the aâvent of

anti-bacterial antibiotics, and more muent recognition of mild cases have also contributed to

improvements in pgnosis? Despite this progress, some patients stiiî do not survive put

childhood Resemh has demonstratecl that factors associated with good prognosis include care

in a CF chic, mak gcn&rp good nutrition, normal intesthai fat absorption! and hi&

~ a r d i o ~ c i l m o ~ ~ ~ fimess.'O

Lung fiuiction declïne is a primary featun of CF patients. An estimaîed 7596 of patients

suffet h m recumnt pulmonary infections: and 90% of ail deaths fiam CF cm be attniuted to

irnversible lung damage." Rilmonary dysfiuiction usually begins in inf'cy with an abnormally

fiequent cough, and eventually inhibits the cfiild's ability to clear smalI aimay secmions.

Bacterial infections arise when mucus becornes trapped in the aimays. As the cilia are unable to

clear inhaled bactena b m the lungs. chroNc irritation and inflammation of the aimays nsuit.

The bacterial agents most likely to cause infection are Pseudomonas aeruginosa, Staphyiococcus

aureus, and Burkholderia c e p ~ i a . ' ~ Virai agents also cause infection in CF patients. Originating

in the upper respiratory tract, these infections can descend to cause bronchitis. In CF, complete

obstruction of the bronchus by mucus will lead to a collapse of portions of the lung. As air

cannot move into and out of the lungs properly, less oxygen will mach the bloodstmam. Duriag

late childhood and early adulthcmd, many CF patients show a subsequent increased sensitivity of

the bronchi to a variety of challenges, ranging h m exercise to inhaled proteins?

Pancreatic insufficiency, an inability of the pancreas to secrete enough enzymes for food

digestion, is another common clinical feature of CF. Appmxiniately 85% of patients are unable

to absorb adequate amounts of nutrients h m their diets, and consequently quire enzyme

supplementation to help prevent growth delays and weight loss. Chest infections, in addition to

panmatic insufficiency, may also lead to weight loss through a reâuction in appetite or

inmascd energy expendit~n!~

Many factoni influence the p w t h pattern of CF patients, includhg lung disease

severity, respiratory infiection frequency, the extent of nutrient mdabsorption, and the &grec to

which tceatment affects these conditions. The mean birth weight of CF infants is appximately

5% lower than that of healthy babies." hi spite of delays in pubaty and skeletai mturity, C F

children usuaüy experieace steady incrases in height and weight during the first 10 years of Mee

These increases, however, arc below the averages for n o d chiidren. Final height and weight

values d l k much greater for CF childm who do not have sigmficant pulmonary involvement

than for children with severe lung impairment?

Beginning with their diagnosis, CF patients at HSC integrate regular clinic visits iuto

their lives. Approximetely once every ihrre months, patients and their families consult with a

s pecializeâ team of heal th professionals, which includes doctors, nwses, ph ysiotherapists,

dieticians, and social workers. Measurements of height, weight, and temperature are taken in

clinic, and sputum samples are collected. In chiIdren older than six years, partial puimonary

function testing (spirometry) is performed at every clinic visit. Exercise testing accompanies full

puhonary function testing once per year. Chest x-rays and blood tests are conducted on a

biannual basis. Xmys show changes within the lungs, helping the doctor to assess the patient's

clinical status. Blood tests demonstrate b e r functioning. nutritional status, hemoglobin levels,

the presence or absence of infection, and the degree to which the patient can fight infcctioneL2

Another component in the management of CF is in-hospital treatment. Penods of

hospitalization usually range fkom 7-10 Qys, and are necessary if an acute chest infection d œ s

not respond to at-home treatment This intensive trcabnent is compriseci of intravenous

antibiotics, physiotherapy, and adquate r e d 2 T ~ t m e n t ngimens routinely conducted at home

can resume once hospitaîization is complete.

The aim of CF matment is to aileviate symptoms and correct organ dysfunction. Patient

contentment and cornpliance with trratm#it an important considerations in deciding on the best

regimen to pscribe, and becom evm more important to consider in adolescent patients.

Panmatic enzyme replacement, the pmcription of antibiotics, and the clearance of lower airway

semtions comprise the plimpsr f o m of treatment.

To combat weight loss, the clinic dietician prescribes a high-energy meal plan and caldc

supplementation. CF patients need an energy intnLc of appximately 120-150% of n o d

requirements. Enzyme and vitamin supplements are ncommended for panmatic insuftcient

individuals. l2 Treatments for pulmonary impairment include bronchodilator therapy, antibiotics,

and chest physiotherapy. These therspeutic regimens have becn very successful in modifying the

natural history of lung disease to the extent that serious lung damage can now be delayeù, in

some patients, for many years.

Inhaled bronchodilator medications relax the branchial tubes in the lungs. The subsequent

increases in both the size of the air passages and in expiratory flow rates make it easier for

patients to breathe. Bronchodilators are useful in combating occasionai bouts of wheezing.

shortness of breath. and persistent coughing. Antibiotic therapy is prescribed for chest

exacerbations and for regular disease management. The patient's response to antibiotic therapy

may include a decrease in the fkquency of chest exacerbations, improvements in lung fimction, a

&mase in the bacterial density of sputum, and enhanced well-kingbL2

Chest physiotherapy mobilizes sputum from smallet to larger ainvays so that it can be

removed by cough. In addition to increasing sputum expectoration. physiotherapy may also

minimize lung tissue damage. Two common physiotherapy techniques nquire the assistance of a

caregiver. manuai percussion of the chat and postural drainage. Two independent techniques are

forceci expiration by the patient, and expiration against a positive pressure r n ~ k ! ~ Physiotherapy

should be ieguiarly administered so that infcctiou cm bs âelayed or prcvtntcd by the nduction

in obstructive mucus. Patients are recommended to perfonn two to thne sessions of

physiotherapy pet day, with cach session lasting 20 to 30 minutes?

Effective chest physiotherapy may requin special equipment and a caregiver to assist in

administration. and adequate compliance with treatment is essential. Compliance with

physiotherapy is often ciifficuit to encourage. however, especiaily if the physiotherapy sessions

interfere with dail y activities or set CF childmi apart h m theu peers. A supplemental treatment

for CF patients is ngular physical activity, which encoumges patients to cou&, and may leaâ to

inmased sputum expectoration4 and maintenance of pulmonary function?

2.2 Phvsical Activitv in Children and Youth

2.2.1 Heaith Benefits

The health effects of physical activity participation in chilâren and adolescents have not

been well documenteci. Based on the available research, the potential health benefïts of regular

physical activity likely incluâe: improved peak bone mass, immune status, catdiorespiratory

fitness, and psychologicai weii-king, teduced adiposity, and a decreased risk of musculoskeletd

injury. As a result of these reported benefits. physicai activity guidelines for adolescents have

been developed. Experts recommend that hdthy adolescents participate in a minimum of 3

weekly moderate to vigorous activity sessions, with each session lasting at least 20 to 30

minute^.'^

The health benefits of physicai activity in children and youth are difficult to determine.

Activity patterns in chiIbn and adolescents Vary as a function of gender, age, and season. Males

are g e n d y more active than fernales, and overaii participation in physical activity tends to

dccrease with age. Time spent in vigorous activity, however, may actuaily i n m u e with age.

Children appear to be more active when outdoors, and seasonal variation in activity patterns has

been rep~rted."*'~

2.23 Patterns in Cbronidy 11l Patients

in light of evidence supporting physical activity participation in healthy populations,

nsearchers surmise that regular physical activity may also be beneficial for individuals with

chronic disordem. Encouraging physical activity participation in childhood wiii help to establish

lifestyle habits that may modify the course of some diseases, and research suggests that physical

activity shouid becorne a regular component of CF management.''

Studies indicate that chronically ill children are more sedentary than their healthy

counterparts. Reduced physical activity in children with c h n i c disorciers may be amibutcd to

parental fear of h m , school worries about injury, physician concems about disease exacerbation

due to physical stress, timeconsuming matment ngimens, or fewer oppominities to exercise.

Exercise-related studies of children with asthma, Dom's syndrome, cerebral palsy, juvenile

rheumatoid arthritis, and CF suggest that exercise training programs are not deûimental. On the

contrary, these pmgrams may result in improved selfconcept, enhanced medical trcatment

benefits, and increases in physical fitness." In CF patients, extensive nsearch has been

conducted to determine the effects of exercise training on general weU-king, lung function and

exercise capacity.

2.2.3 Exerck Consiàerations in CF Patients

The htaith beaefits of lecnational exercirre for CF patients were fint recognized in 1969.

Enmise hbas since km proposcd as an adjunct to chest physiotherepy,'9 and is now widely

accepteci as a therapeutic intervention in CF? Studits have indicatcd that exacise training can

inmatje spuhim expectoration i u g h heightened coughing. impmvt respiratory muscle

endurance, and contribute to the maintenance of puimonary fuaction status.' In addition, exercise

may impmve clinical statu, fostcr the child's participation in nomial activities of childhmd, and

improve theù morale." As a supplementmy -ment, exercise is appeaiing because children cm

perform theu therapy independently or with their pers in activities of their own choosing. While

there is no evidence to suggest whether reguIar participation in physicai activity cm improve

prognosis in CF patients or be as effective as chest physiotherapy, CF patients who do not

exercise may experience nductions in muscle tone, strength, endurance, and aerobic ~apacity.'~

Although exercise in CF patients is considered to be generally safe, there may be

additionai risks for some individuals. Possible problems inch& a reduction in blood oxygen

resulting in tissue death. dehydration, the inducement of violent coughing, and weight loss.' As a

nsult of these potential risks, extra precautions may be necessary. If oxygen &saturation occurs

during exercise, it can k deviated with oxygen supplemmtation. Training at a target hem rate

of 10 beats per minute klow the point where desaturation occurs should also be prescritmi. A

gradua1 increase in activity levels should precede the onset of any training program if the patient

exhi bits ph ysical decondi tioning. ' Exercise testing is a usehl tooi in the assessrnent of disease pmgression." CF patients

suffer prognssively impairecl exetcise capacity as theû lung disease wor~ens.~' Although heart

rate and blood pressure responses to exerçisc arc normal, exercise testing demonstrates that

maximal values for work capacity, oxygen consumption and h e m rate are Iess than expccted

when compareâ to the hcaithy population. In CF, exen5se is limitcd by lung mechuiics, whereas

in heaithy individuals, exerciae i i Iimited by the cardiovascdar system. ExCrcise training

pro- of sufficient intcnsity in CF may nsuit in bcnefits to excrcise capacity, including

imptonments in maximai work capacity, maximal onygen consumption, maximal minute

ventilation, respiratory muscle endurance, and anaerobic ~ireshold.'~

Guidelines have been suggested for extrcise recommendations in CF patients. The

activity choice will depend partially on the patient's access to appropriate equipment and

f d t i e s . Aerobic activities that will pmmote sputum expectoration, such as swimming, running.

skating and dancing, should be encouraged h m an early age." Participation should entail th=

to five sessions per week at an intensity of 70 to 85% of the maximal heart rate. CF patients are

generally not able to attain an age-prcdicted maximal heart nite due to ventilatory limitations.

Therefore, maximum heart rate should be individually detemiined. Patients should be ailowed

significant input into the choice of activities in order to ensm gooâ compiiance with the

pmgram.'g

Cornpliance with exexcise programs cm be diKcult to foster in CF patients due to the

existence of time consuming matment regimes, forgedùlness, or a negative attitude towaràs

exercise. Although many CF centers do not offer organized exercise programs, such programs

may encourage cornpliance. Summer camps, in-patient hospital programs. and home-based

program are several possibilities. While summer camps offer brief periods of intense activity in

a positive social setting, in-patient hospitai programs cm teach patients how to incorporate

regular exercise into their Lives. At-home exercise participation requires patients to pprtelue

independentiy in unstnictured programs. In North America, however, summer camp and gmup

exercise programs arc no longer tt~ommcndcd due to the risk of cross-inféction of Burkholdaia

cepria." niercfore. home-bascd exercise progcams am the most viable option for CP patients.

As cornpliana with long-term uracise training pgrams is di&cult to maintain, a more

fa ib le alternative for CF patients may be a change in habituai physical activity psttcms.

Whereas exercise is characterizcd by an imposed and ngimented training program. habitual

physical activity refm to the level of activity one has incorporated into their daily Me. Changes

in habitual physical activity would represent a liftstyle modification. and may promote long-term

hedth benefits more easily than exercise programs. Until now, the quantification of and foliow-

up of habituai physical activity in CF patients have not been adequately addnssed. Research hm,

however, been conducted to examine the d e of exercise training programs in positively

affecting the health of CF patients. These studies an important to examine because they provide

evidence of the benefits of physical activity in CF.

2.3 Exercise Training Ronrams in CF

2.3.1 Characterisüa of Exercise Rognm~

Exercise-related studies in the CF population have been conducted to cietennine the

effects of exercise training programs on pulmonary function, exercise capacity, sputum

production, respiratory muscle endurance and psychological well-being. The various studies

have implemented training prognuns of differing types. duration, fnsuency, and intensity.

The degree of supervision is an important consideration in exercise presaiption.

Supe~sed pmgx'ams require regular inmention and program monitoring by an exercise

specialist such that the safety of the individual can be @odicPlly checked and individual

tailoring of exeiçise programs cm occur. These factors likely foster high patient compiiance with

the training program. On the contrary, the success of unsuperviscd training pmgrams depends on

the participmt's sociai support and seKmotivation. Aithough cornpliance with the program may

k less than optimal. unsupmised programs rsquire mlliimal monitoring. and thus have the

The desireci duration of the exercise program should be decideci on @or to the onset of

training. Whereas long- training pmgmms have been shown to benefit lung fuaction in

the effects of short-term exercise pgrams are still equivocal. Although positive results fiom

shortened training pend have ken demonsttgted, these effects are often short lived. 233

Training program of varying intensitiea en also pnsented in the CF literanire. High intensity,

exhaustive pmgrams have demonstrateci positive m~ults,~ but these pgrams may be difficult to

incorporate into daily life. Exercise pmgrams of lesser intensity have been considered as a more

feasible dternative.' Specific study examples of the various training program are presented

below.

2.3.2 Supervid Pmgrams

Supe~sed training progcams can occur in a formalized hospital setting, as demonstrated

by utc ch^ and AmericanL6 based studies. The conflicting results h m these investigations

demonstrated either benefit to pulmonary function,= or no harm to patients with training." In the

Amencan s t ~ d ~ : ~ there was no reported difference in pulmonary f'unction for subjects

mdornized to two groups: either two sessions of exercise plus one session of physiothmpy

(n=9), or three sessions of physiotherspy (n=8). Patients (mean age=ISA years) stayed on the

program for an average of 13 days. and perfonned exercise on a cycle ergorneter. Over this short

study pend the investigatom ensured hi@ cornpliance with trcatment, and concluded that

substiaiting exerdee sessions for regular physiotherapy was not harmful to patients.

The dutation of hospitalization paiods for CF patients are generally short. and are

brefore not conducive to the irnplernentation of an extended, superviseci training program.

Rccreational settings arc m a e suited to long-tenn pgrams. Eight supervised programs have

ban conducteci in recrcationai settings under the supe~sion of a traincd profcssional. The

activity types included swimming. n ~ i n i n ~ , ~ trampoline ex t rc i s~ ,~~ intensive and muiti-

activity training in a camp ~ e t t i n ~ , ~ ~ and aembic txercise directed by an instmctor. 3132

Two studies have evaluated the effets of a 12-week? d a seven wœk swim pm8f8m24

on the heplth of CF patients n$pectivcly. Research bas suggested the vaiue of swimming as a

training mode in CF because the welî-moisturizcd environment may inmase sputum

expectoration. Submersion-induced subtk changes in volume and presswe within the respiratory

system may also contribute to knefits in CF patients with smmming programs.u The seven

week s w i m program was coupled with repeated evaluation of lung function and sputum

production in 1 1 patients (median ag-10 years).u Children maintained their usual schedules of

therapy ihroughout the study. Sputum volumes were found to be significantly higher on

swimming days than on non-swimming days. In addition to enhanced mucus clearance. results

also demonstrated significant improvements in pulmonary function following the training

program. Although these patients w m recommended to remain physicaiiy active after

completion of the swim training, pulmonary huiction values eventuaily renimed to baseline

velues. The positive changes in pulmonary huiction may have resuîted fiam the sswimming itself,

or h m the supmised nahm of the program. Results h m the 12-week swimming program,n

on the other hand, demonstrateci no benefit to pulmonary function.

Trampoline jumping has been suggested as an appealing alternative to more typical and

arduous forms of exercise, and may mimic the actions of chest physiotherapy. Eigbt children

(mean agesll.5 years) participateci in an eight-week exmise program on a trampoline. for a

niaxinium of 109 minutts l~eck.~~ The @en& participatcd in a program that included rocking,

jogging, and jumping on a mini-trampoline. This program nsuiteâ in significant impvement in

faccd vitai capacity and maximal oxygcn coiuumpion.

Patients (age rpngc=IO-30 yeare) vol~~~ttered for a three-month d n g pmgram in

another supervised s t ~ d ~ . ~ Twenty-one CF patients wen ailocated to the exercise group. whüe

the remaining 10 patients formed the control gmup. The exercise group participated in an aerobic

program three times per week that included wm-up, jogging, and cool-down components. The

control group was askeâ not to change dieV usuai activity. The exercise group demonstrateci

significant improvements in exercise capacity and in respiratory muscle endurance, but no

change in pulmonary function. In the control p u p , fomd exphtory volume in one second

showed significant decline.

Superviseci pmgrams have also been implemented in a pediatric rehabilitation centna

and at a s m e r ~ a r n ~ . ~ ~ h e s e types of settings provide optimal conditions for children to

participate in physical activity for brief time periods. Seven CF patients (mem age13.4 years)

participated in four weeks of summer camp, unâergoing 1.5 daily hours of intensive swimming

and canoeing aimed at upper body endurance? Respiratory muscle endurance, which has been

comlated with fatigue nsistance, impmved by 56.7% in these patients. No conclusions could be

drawn from this study, however. about the potential benefits of lower body endurance exercise.

At a pediaüic rchabilitation centre? patients partjcipated in a 17-day intensive training period,

replacing usual physiotherapy with physical activity. Although pulmonary function did show

signifiant improvement, less intensive emrcise may not have demonstratecl benefit to patients.

Results h m th- studies iUustraie diffmntial eff- of supmised training on lung

hmction, ex&r capacity, and respiratory muscle endurance in CF. A multitude of foctors mut

be considered @or to the onset of diis type of program, incltuhg the type, duration, and setting

of the program. A summary of the superviscd pfograms are depicteci in AppendiXA as Table A.I.

Seven unsupeivised training pmgrams have been implemented in the CF population. 333-38

h s u p e ~ s e d propus are usually canieâ out in the home setting, and an generally longer in

duration than supervised program due to minimal monitoring requirements.

The two longest studies to date have been 30 months3' and three years3 in duration. The

effects of physical training on work capacity, lung fhction and clinical condition were examined

in seven patients (mean age=10.4 yeam) participating in daily exercise over 30-months?' The

exercise program was compriscd of skipping, jumping, jogging end swimming. Chest

physiotherapy was withdnwn aftu the initial 12 months. As puhonary function parameters

were unchanged over the course of the study, the investigators concludeci that physical training

could replace conventional chest physiotherapy except where strenuous exercise was not

feasible.

In a three-year exercise inmention study in 65 CF patients (age range=7-19 years), the

effects of long-term training on lung hmction were evaluated? Subjects were randomized to

eitha an exercise or a control group. The exetcise group was instmcted to exercise a minimum

of t k times per week for 20-minute sessions, while the conml group was asked to maintain

their usual activity. Patients h m both groups continued their usual schedules of chest

physiotherapy. Subjects participaieci in aerobic activities of their own choosing, with a guideline

target hem rate pvidcd by the exercise physiologist. Results indicated a pater annual mean

rate of decline in forced vital capacity for the contml gmup cornparrd with the exercise p u p .

No Merences in exercise capacity were demonstrated.

An Australian studf invcstigatcd the effits of a 12-week home exercise program,

comprised of sprints and jogging, on 41 txerche and 45 conbol patients (age mp8-14 years).

The exercise group demonstrateci significant impmvements in exercise capacity, but no change

in puimonary function. The investigators noted declining compli~nce in the latter half of the

program which may have contributed to these non-significant muits.

The literature indicates a neeâ for m e r nsearch in the area of unsupe~sed programs.

Whereas one study suggested that exercise could have equivalent benefit to patients as

conventional physiotherapy~s another study nported no benefit to pulnionary function with

unsupervised training?' The unsupervised program am presented in Appendix A as Table A.2.

Another possibility for exercise training is to combine supervisecl and unsupervised

components into one exercise program. A study in the Netherlands showed improved pulmonary

function (forced expiratory volume in one second) following a short period of hospitalization,

and an increase in exercise capacity after one-year of at-home training? This type of program is

useful if patients begin exercise training during a hospital stay. and want to maintain a similar

level of physical activity once they rem home. This study is presented in Appendk A as Table

A.3.

23.4 Chaifenge~~ of Trsiiiing Program

The effects of e x d s e craining on CF patients are difficult to interpret due to small

sample &es and potential confounding variables. hvestigatom find it difficult to remit enough

patients for their studies because an extensive tims cornmitment is requifed, and some

individuals are not intmstcd in pmuing excrcise training. Las to follow-up is aiso a problem in

CF patient studies, m e r emphasizing the n d for a large sample sim. Chest exacerbations, for

example. may intafkre with the training protocol. Codounding variables mut be adequately

controIIed for. Potcntial confounders to consider inch&: the hetcrogcneity of the disease and its

dive= naturai course, unprsdictabk occumnas of infection, complexity of the thaipeutic

approsch, and variable degrees of cornpliance with inatment.40 These fwtors must al1 be

accounted for if conclusions are to be made on the relationship between exercise training and

health in CF,

Compüance issues and the individuai needs of the patient must also be given cmfui

consideration. Rogrtms that are too long or too intense* may deter patients h m paiticipating.

Individual tailoring of training pmgrams, such that each prograrn meets the training needs and

interests of the patient, is essential if heaith effects are to be demonstrated. Finaily, the benefits

in lung function and exercise capacity seen with training may not last beyond the training

program itself? Given the challenges involved in implemnting successful exercise training

programs in CF, M e r evaluation of habitual physicai activity and its relationship to important

health outcomes is necessary.

2.4 Relationshi~ between Habitual Ph~sical Activity. Lune Function. Nutritional

Status. and Exercise Ca~acitv in CF

Several studies have begun to explore the relationship between habitual physical activity

and lung function in CF. In one investigation, the physicd activity levels of 30 CF patients and

3 1 control subjects (ages 7-18 years) wen ~ o r n p a n d ~ ~ The investigators aiso evaluated whether

habituai physical activity was nlated to exercise performance and pulmonary function in these

CF patients. Kriska's Modinable Activity Questionnaire was used to assess habituai physical

activity. Total hours of physicai activity per wœk did not differ ktween the C F (7.3 hodweek)

and the control(7.1 hounlwcek) groups. The total hom engaged in vigorws activity, however,

were significantly lower in the C F patients (1.3 hourslwak) comparecl with the healthy controls

(3.2 hourdwœk). The rcsults indicated that activïty levd was unrelateci b both txercise

toleranct and lung hmction.

A Quebcc study ueed the Habituai Activity Estimation Scale w) as an estimate of

habitua1 physical activity in enamining the nlationship between activity levels, lung fuaction and

nutritional statu in 36 CF patients (aga 6-16 years)!2 Consistent with pnvious usage of the

HAES, patients classified theu actinty into 4 differcnt categories, and their activity level was

represented by the total percentage of time speat in the somewhat active (i.e., waücing) and

active (i .e., biking) categories. Boys and girls speni 8.1 and 7.5 hours respectively being at lest

somewhat active. In children with significant lung disease, activity level was positively

associated with nutritionai status, but unrelated to lung function. h m these nsults, the study

investigators concluded that activity levels may be rsstncted by nutritional status in more

severel y affected patients.

Disease progression in CF is characterized by malnutrition and lung hinction decline. As the

disease advances, exercise capacity is also limited? Researchers have h ypothesized that

pulmonary function may affect exercise capacity by limiting ventilatory capacity, or through

skeletal muscle deconditioning. Nutritional status may be influentid through reâuctions in

muscle mass or changes to muscle quality? The inter-relationship of nutritionai status,

pulmonary function and exercise capacity has ban p~viously investigated. a,r3,rs

In 50 CF patients, exercise testing was performed to determine limiting factors of

exercise cap~acity?~ in this sample of patients, exercise capocity was demonstrated to be

correlateci with both lung function and nutriti011al status. Another study was conducted in 22 CF

patients with advanced lung disease to evduate the effects of nutntional status on exercise

capacitya Results indicated that exercise capacity w u relatcd to the degree of respiratory

impairment. and independently infiuend by nutritionai status. In an investigation of 14 CF

patients, lung diseare sewrity and paipheral muscle hmction were both dctermined to be

lunithg factors of exercisc ~ a ~ a c i t ~ . ~ At HSC, long-term folîow-up of CF patients demonstrated

that the institution of a high fat diet resulted in improved nutritional status, and a subsequent

increase in survival! These saidies have demonstrated an association between nutritionai status.

exercise capacity and lung function in the CF population. The potentiel contribution of habituai

physical activity to this nlationship needs Mer investigation, as does the quantification of

habitua1 physical activity in CF.

2.5 Measwement of Phvsical Activit~

Pediaüic chronic disease nsearch has advocated the importance of incorporating a

quantifiable rneasure of physicai activity into reguiar disease management." A consensus has not

ken reacheâ, however, on the best mcthod of measuriag physical activity. The difficulty lies in

the many possible choices for physical activity quantification. ranging h m isotope consumption

to heart retc monitoring and self-report. Racticalîty, cost and measurement accuracy are

important considerations in chwsiag the most appropriate method.

2.5.1 Doubly LaMd Watet

Doubly-labeled water @LW) is a relatively new technique that is being considered as the

gold standard of energy expenditure estimation. Although the expense of this procedure deems it

impractical in largescale epidemiological snidies, the accuracy of DLW malw it an appealing

choice for field and laôoratory studies!' One limitation of this procedure, however, is that it

canot be ueed to &termine the proportion of energy expenditure associatecl specifically with

physical activity?

For this proceâun, patients ingest a quantity of water with a known concentration of

hydrogen and oxygen isotopes. ûnce the isotopes have distriibuted themselves m equilibrium

with body water, labeled hydrogcn exits the body as water, and labelcd oxygen leaves the body

as water and carbon dioxide. By knowing the difference in the elimination rates of the 2 isotopes,

the pduction of carbon dioxide cm be calcula& Total energy expenditure is calculateci from

standard respiratory gas exchange tquations?

2.52 O b ~ e ~ t t i o n

In contrast to DLW, the observation method is a niatively simple means of assessing

physical activity. Observation is usefd in studying pnschool and school-aged children. where

alternate methode may be neither feasible nor appropriate. An observer monitors the activities of

a subject for a predetermined amount of time, and records data on the type. fkequency, duration

and intensity of each activity. To calculate total energy enpaiditure, the estimated energy cost of

each activity is multiplied by the time spent in that activity?'

The observation method haî not been previously validateci, although it has been used as a

standard by which other techniques are compareci. Due to restrictions on time and cost, this

method is best suited for small studies. The= are severai other disadvantages to this technique.

Observations can only be made over short tirne periods, and are not refiective of habitua1

physical activity. Additionaîly, subjects may alter their usud activity if they know that they are

being observed!'

23.3 Movement Asmsmetlt Devices

Movement assessrnent devices arc wom on the tNnk or extremities, and are designed to

k t l y monitor movemcnt and store this data for subsequent analysis. Two examples of thcse

motion xnsors are the palorneter and the accelcrometer. Although the pedometer depends on

accelemtion and deceleration. it is not an a~celemrneter!~ An accelemmetcr mcasures

a c e l d o n of the body.

nie pedometer counts the nurnber of steps taken or estimates the distance waIked. It is

usehil in estimating physicai activity associateci with walking or m g , pmviding a quick

assessrnent of activity Ievels. ïndividuals tend to walk with a pa te r impact on one fwt than the

other, and pedometer readings will Vary depending on which side of the body the pedometu is

wom!' The pedometer wil l also underestimate activity Ievels for sports such as cycling, skating,

and rowing, where activity is not baseci on vertical displacement.

The most cornman single-plane acceleromeer is the Caltrac. This device clips close to

the body at the non-dominant hip to ensure an accurate estimation of acceleration and

deceleration. The Caltrac is designed to provide an activity score that nflects the intensity and

quantity of the movement, and allows for an estimation of energy expendiw for activities

within the vertical plane? Three-cümensional accelerometers reflect movement outside the

vertical plane, and can estimate energy expenditure for many different types of activities. These

triaxial accelemmeters measun motion in the vertical, horizontal and diagonal planes. There are,

however, disadvantages to use of the accelerometer. Energy expended as a result of lifting heavy

objects cannot be nflected in the data For water sports such as swimming, waterpmfing of the

accelerometer is necessary. For cycling, the &vice muet be wom on the leg instead of the

trunkt4'

2.5.4 Heart Raîe Monitoring

A demonstrated positive relationship behvem heart rate monitoring and cnergy

expenditure has led to the use of h m rate monitoring to estimate energy expendiiun. This

technique is appealing in populations where it is difficuit to assess physical activity kvels by

other means, such as in young children. Relatively inexpensive heart rate monitors are available

to assess daüy activity, complete wîth M-day storage capacity for minute by minute data

collection of h m rate? To obtain an individual's heart rate p d e , a minimum of four to five

recording days is nccessary in order to account for the variability seen in heart rate

measurements. Many factors e t heart rate and could infiuence the nlationship between heart

rate and energy expenditure, including the active muscle group, anxiety, training, ambient

temperatureT and the type of muscle contraction?' Using the hem rate monitor in conjunction

with a self-report instrument could better reflect physical activity than if the monitor is used

alone.

25.5 Seif-Report Instniments

In epidemiological saidies. activity diaries and questionnaires are the most common

techniques used for assessing habitua1 physical activity. These instruments can either be self-

administered or completed with the help of an interviewer. Parents or teachers may serve as

proxy respondents in estimstng the physical activity levels of young children. Habitual physical

activity patterns or physicd activity levels in the recent past may be of most interest to a

researcher. Activity type, intensity, duration and fkequency are the most common variables that

are meesund, and in some cases. an estimation of energy expenditure can be computod. 47.48

Advantages of using disries and questionnaires to mess physical activity include low

cost, quick administration tirne, the abiüty to gatha information over an extended time period,

and minimal staff* However, the vaüdity and nliability of these instnunents must be continually

evaluateâ. Disadvmtagts include potentiai recd bias, and subjective respoases to the questions.

In addition. studies indicate that self-report instruments an not useful in children unâer 10 years

01d?~*" Roxy response is, however, an acceptable alternative for completion of these

instruments.

253.1 The Aetnlty D m

Activity diaries requin subjects to &tail their daüy activity at intervals ranghg h m 1

minute to 4 hours. Individuais may bmadly classiry UicK activities, or express their activity

pattems in more specific temu, using prepared forms. Although this type of extensive reporthg

may be tedious, subjects can employ shorthand methods to efficiently report their activity levels,

and rnay discover patterns in their weekly activity, making recording easier.

A 3-day activity diary was developed by Bouchard et al to estimate the amount and

pattern of daily energy exPenditudg Every 15 minutes over a 3day period, 150 adults and

chilâren (ages 10-50 years) categorized their activities on an energy cost sale h m 1 to 9. Mean

energy expenditurc was positively comlated with physical working capacity and negatively

comlated with body fat. Reliability testing indicated high repducibility for this measurement.

This 3-day diary was modified as a 7-day activity diary in a Swedish investigation?' In

two groups of randomly selected adolescents (aged 15 years) h m two regions of Sweden

(n=374), activity pattmis were recordecl over a 7 day period. Each subject was provided with a

detailed explanation and demonstnition of the diary. Total energy expenditure and physical

activity level w m computed from the diary entries. Resuits dernonstrsted high and concordant

physical activity levels and energy expenditure in the two subject groups.

Validation of this activity diary was p g f o d in a sub-sample (n=SO) of these Swedish

adolescents? The 7&y activity diary was cornparcd to the doubly labeled water methoci, and

was found to pproviâe close es- of physicai activity level and total e n q expenditure.

These results suggestcd the suitability of the diaq for estimating energy txpnditurt in groups of

adolescents.

Although not previously used in a CF population. a Mar version of this activity diary

has been used in children with chioaic A sample of 23 cMdren (ages 5-1 1 yem) with

juvenile rheumatoid arthntis (JRA) were compared to an equd sample of healthy childnn on

àaily physical activity levels using a 3-day activity record. The investigators hypothesized that

childm with JRA would be las active than their healthy counteqsits. The 3day activity record

was completed by parents. and pmvided data on the type and intensity of the activities at 15-

minute intervals. Mean physical activity values w m significantiy lower in JRA patients than for

controls.

In a study of Quebec chilâren and youthy the 3-&y activity diary was used to examine

the relationship between puimonary function and habitua1 physical activity in 424 boys and 366

girls (ages 9-18 yem). Estimates of daily energy expenditure and time spent in moderate to

vigorous physical activity were computed h m the diary. R d t s demonstrateci that habitua1

physical activity was not associated with puhonary function in this sample of childmi and

adolescents. The investigators speculated that this lack of association could have been due to

inadquate stimulus, age, growth status, and type of activity.

The diary offers several advantages. Data cm be collected for many subjects

simultaneously, and the type, dmtion and hrquency of the activities can be recorded. This

technique incm a low cost to the investigator, and an estimation of energy expendihin is

possible. Due to a possible reduction in sccuracy over long periods of data collection$' however,

phone cals and careful description of the instniment rnay be necessary to encourage accuracy

and complcteness.

2.5.5.2 QUestionnPirwP

Many questionnaires have been designed to quantify physical activity for various periods,

including: the previous 24 hours, one week, three months, and one year. Attempts have dso been

made to quanti0 Metime and occupational physicai activity, and to estimate usual or habitml

physicai activity. These questionnains are either self-administereâ, or administercd ui-puson or

over the phone by an interviewer?' Several of the more commonl y cited questionnaires are

described below . The Revious Day Physical Activity Recall (PDPAR) requires subjects to recall after-

school activities h m the prwious day. Activity is recordai in 30-minute intervals with a rating

of relative intensity provided by the subject. A listing of common youth activities is intendeci to

aid in subject recall. One studJ4 concludeci that PDPAR provided an acceptable estimate of

relative energy expenditure over an 8-hour recording period and accurately identified bouts of

moderate to vigorous physical activity. Although this instrument provides a summq of type,

fnsuency, intensity, and dmtion of the activities, then are iimitations. Results depend on the

accuracy of a one-day mail over an 8hour period, which may not be representative of usual

physicai activity participation. A validity study was perfonned in ol&r children only, and the

appropriateness of PDPAR for younger children is ~nascertained?~

The Physical Activity Questionnaire for Older Children (PAQC) is a self-atiministered

7day mal1 measun designcd to evduate moderate to vigorous habitual physicai activity during

a specific season. Physical activîty is describecl as sports or games that make you bnathe or

sweat hard, or maice your legs tirrd An activity score is calculateci h m a compilation of 9

questions. Although the PAQC was show to ôe a cost-efficient methocl of assessing children's

physicai activity during the s c h d it c m o t bs used to esbcmstt caloac expenditurc, is not

designed to UUKSS summer or holiday levets of physicd activity, and Qes not discriminate

between moderate and vigorws physical activity levels.

The Godin-Shephard Physical Activity Survey is a simple, quantitative physical activity

self-report masure in which subjects report the number of times in an average week that they

spent more than 15 minutes in activities classifieci as mild, modemte or strenuous. Examples of

activities in each of the thne categories is pmvided to subjects? This instrument was found to

be diable and positively comlated with PAR in one study, and investigators concluded that it

could be a usefiil self-report measure in young childrai? However, this insrniment may be

subject to recall bias as individuals must ncail theû participation in physical activity over seven

days.

The Habitua1 Activity Estimation Scale (HAES) is a self-report instrument âesigned to

estimate habitua1 physical activity. Completion of the HAES entails estimating the proportion of

time spent at each of four different activity categories for a typical day?* The HAES hm ôeen

used in both hedthy and ctuonically ill pediatric populations. In 36 childnn with acute

lymphoblastic leukemia, bone mass, minerai metabolism and physical activity were assessed.

The HAEs was completed for the most recent weckday and weekend day at 6-month intavals

over the course of the hempy? In a group of 156 grade nine students in Southem Ontario, the

relationship between babitual physicd activity, as estimateci thtough the HAES, and self-

perceptions of physical activity among Native and non-Native adolescents was e ~ a l u w d ~ ~ In an

investigation whkh examined normative values for lumbar spine bone mass in childnn in

relation to age, gender, dietary intake. and physical activity, the KAES was also employed to

assess physical activityPL Although îhis instrument hm wt ban used to estimate cnergy

expenditure, these stuciies have &monstrated successful use of the HAES as a seEreport

instrument in different populations

Investigators have attempted to determine the accuracy of the HAES in the following

thne studies. The sensitivity of the HAES to &tect change in activity patterns was evaluated in

24 hypoactive children (ages 5-16 years) with chronic conditions. Parents completed the HAES

for their childnn prior to and following an 8-week outpatient program &signecl to incrase

physical activity. Significant increases in physical activity were demonstrated in comparing the

two parental reports of activity.'* In a study of 52 girls (6-12 years), the relationship between

percent body fat, body mass index, and inactivity was evaluated. Parents reporteci the activity

levels of their children using the HAES for a typical weekday and typical weekend day.

Significant, positive correlations were demonstrated for inac tivity with both percent body fat and

body mass index?' In a sample of 17 schwl children, a positive and significant conelation was

found between activity rneasured by a Caltrac accelerometer and the active category of the

HAESPZ Although finther vaüdation of the HAES is desirable, results from these investigations

demonstrate bat the HAES is a practical and convenient tool that provides an acceptable level of

accuracy in measuring habitua1 physical activityPO

2.553 Summ~ry of Sell-Report Instruments

Questionnahs and diaries are practical and inexpensive tools used to quantify physical

activity in epi&miological studies. Research is stiii necded. however, to determine the validity

and nliability of these measms. Validation is challenging due to inadequate criteria by which to

judge these self-report masures. An instniment such as the HAES gives subjects a bill range of

activity options, h m inactivity to high activity. lessenhg the likelihood of b i d estimations

which might occur if only high activity Ievels w m npoireds8 nie bcst techniques will captuse

ail types of physical activity, including occupational, leisure, and household In addition, the

instrument should be able to quantiry the type, duration, frequency and intensity of the activity. It

may be necessary to integrate the use of two different activity measures into one study in order to

capture al l dimensions of activity.

25.6 Summary of Physicai Activity Measurememt

Many methods are avdable to measurc habitual physical activity. The quantification and

pmfiling of habituai physical activity is necessary in order to follow these pattems over the, and

to examine the association of physical activity with other important variables, such as lung

function in cystic fibrosis. An additional importance of the quantification of physical activity is

to explore potential cornlates of physical activity participation. This knowledge could M e r

explain differences in activity pattans betweea groups. and promote the adoption and

maintenance of regular physical activity participation.

2.6 Potential Correlates of Phvsical Activitv Partichation

As the health benefits of reguiar physical activity are well estabüshed in adults, the

habitual physical activity patterns of children and youth are also kginning to be investigated

more closely. Studies have suggested that an active childhood may pndispose individuals to

similar activity patterns in adultho0dp3 and may subsequently d u c e the risk of c h n i c

diseaseeU The quantification of habituai physical activity in childhood is an important,

preliminary step in understanding this association,

There is also a necd to study the potential comlatea of exercise behavior in childhood

and adolescence. These fxtors may help to explain the differences in self-motivation between

active and inactive individuais, and dlow for the establishment of guidelines to promote regular

participation in physical activity* Agc and gcada have alrieady kai recogaized as important

cietednants of exercï8e behavior, but modifiable fiwtors such as psychosocial and

environmental variables are of spccial interest, and rcquire hir<her studYew

The potential comlates of physicai activity behavior cm be summarized into three

categories. Persod chot~~teristics include age, seif-motivation, weight, and personal attitudes.

Envirolunental ckructerisrks cornpise access to facilities, family or peer influence, and

climate. Activity chamcteristics encompass activity intensity and type." In the CF population,

understanding the potential detemiinants of physical actinty couid have invaluable implications

for health c m professionais who maice recommendations on exercise participation to patients.

Studies in the healthy population have aiready proposeà potential comlates of physical activity

participation.

Modifying physical activity patterns through changes in the environment appears to k

more practical than implementing expensive public intervention pmgrams. A study was

conducted in 110 college students to examine variabks that may influence physical activity

behavior. The physical environment was the pximary variable of interest. In this investigation,

self-report items were uscd to assas environmental variables at home, in the neighborhood, and

on fnquently traveled routes. Home exercice equipment and accessibility of exercise facilities

were significant comlates of physical activity participation. After adjustrnent for neighbohood

sociocconomic status. however, home exercise equipment was the only significant cornlate."

In a p u p of 528 chiloiren (mdian age=ll yem), demographic, psychosocid, and

environmeneal dctcnninants of physicai activity were assesseci. Age, gender, television watching,

and la& of exercisa quipment in the home w a e found to be significant comlates of low

activity status aftaebination of non-8igDiticant variables. Rcsdts suggestcd that p m t s can

produce a home nivitonnient conducive to physical activity involvement by monitoring the time

that their childnn spend watchiag television, aud by providing wess to exercise equipmentP6

Robinson et al sought to evaluate the relationship between hours of television viewing,

adiposity, and physical activity in 671 f e d e adolescents (mean ag-124 years). Wours of after-

school television viewing, physical activity level, and stage of sexual maturation were assessed

with self-report instruments. In this sample, resuls suggested a weak association of television

viewing time with adiposity, physicai activity, or change in either over tirne? More nsemh is

needed to further assess the relationship between physical activity and television viewing.

For 30 Young children (ages 5-9 yem) and their parents, investigators sought to

determine the relationship between parent and child activity levels. The children exhibited

physical activity pattems that were similar to theu parents. These results suggesteâ that

children's physical activity participation may be incnased through the promotion of ngular

physical activity in their

Physical activity patterns are cletennineci by the complex interaction of biological,

psychological. social, cultural and environmental factors. No single variable can account for the

variance in childrni's physical activity participation. Further studies in both healthy and diseased

populations are needed to stimulate hypotheses and provide information for designing and testing

longitudind intervention prograrn~.~ Wective activity promotion interventions hold promise for

improving the hdth of both childm and adults h u g h the encouragement of exercise

involvement .@

METHODS

3.1 Overview of the St& Des@

This was a cross-sectional pilot study. The first component of the study was descriptive:

subject characteristics, nutritionai statu, pulmonary function measuranents, and habitual

physical activity levels of CF patients were dedbed. The second component was analyticai: the

relationship of pulmonary fùnction with habinial physicai activity, nutritional status, and exercise

capacity was explored.

3.2 Practical Planning and Irn~lementation

The study coordinator (graduate stuclent) conceptualized this pmject in September of

1998. The six-month planning period that foiiowed was necessary for the development of an

appropriate study pmtocol, and for securing project and ethics approval from HSC.

A thomugh literature search, conducted by the study coordinator, confirmed that nsearch

in the pmposed area was insufficient. This search entailed exploration of the PUBMED database

using the following key words: qsftcfibrosis, Mitml ppltycal activiîy, ppulmu~~~ryfunction~

and exetcise cqacity. While the quantification and evaiuation of habituai physical activity in CF

patients had not k e n adequately addrcssed, numemus studies had previously examined the

relationship between exerciee training programs and lung huiction. Thenfore, the study

coordinator concluded that the association ktwcen habitual physical activity and pulmonary

function needed m e r investigation in CF.

A second litcrature scarch was conducteci to assess the femiility and appmpriatcness of

diBertnt techniques useci to quantify phpical activity. Of particth intmst w a e ~ t ~ r c p o r t

instniments that could be s e E a d m i n i i e ~ by CF patients. Two instruments, which together

provideci a comprehensive profile of habituai physical activity, were chosen for this stuày.

Patients in the CF cliaic at HSC expenence frequent requests to participate in both

invasive and time-consuming tescarch snidies. A study coordinator must therefore infonn the CF

clhic staff of a new pmject, and gain thek approval prior to patient contact. For the cunait

pmject, support h m clinic staff was also intendeci to encourage maximal participation from

patients. In December of 1998, the study coordinator presented her thesis proposal at a weekiy

meeting attended by the CF c l i ~ c team, which included: one social worker, physiotherapist,

dietician, exercise ph ysiologist, biostatistician, five ph y sicians, two nurses. and three study

coordinators. The study coordinator invited the ch ic team to critique the proposal, offer

feedback and ask questions. The pmject received unconditional support, and the CF ch ic team

acknowledged its clinical relevana.

The study coordinator also arrangeci to meet individually with members of the cliaic

team. The physiotherapist offered a summary of her exercise recommendations for patients, and

discussed common problems in encouraging compiiance with therapy regimens. The exercise

physiologist contributed her observations on the wi& specrnim of physical fitness levels in CF

patients. Infonnai discussions with the pulmonary hction laboratory technicians proved to be

invaluable during subject rrcnlltment. The technicians helpeâ the snidy coordinator to identify

both potentiaî participants and those patients who were ineiigible due to chest exacerbations.

At HSC, all newly pfoposeâ nsc~rch pjects must undergo a comprehensive review

process wîth the Research Ethics Boani of the Rescatch Institute. The sndy coordinator prcpared

the pmject proposai pccording to established guidelines, which includcd designing the

appropriate consent and assent fomis. The 10-pagc proposai and appendices w a e submitted for

intanai review to cIinicians mthin the Division of Respiratory Medicine. Two intemai rtviewers

careflllly scn~tinized the project, and provided written and oral suggestions for improvemnt. The

study coordinator discussed the necessity for revisions with the reviewers and h a superviaor

prior to milMg changes. The revised proposal was nsubmitted to the R e m h Ethics Board,

and due to the non-invasive natm of this project, it nceived expedited appmval. The study

coordinator initiated subject m i t m e n t and data collection procedures in March of 1999.

3.3 Subiect Selection

33.1 EUgibility Criteii

Study participants were selected from patients (7-18 years) under the regular c m of the

CF chic at HSC (n=188). The number of eligible patients was determineci by extncting data

h m the Toronto CF database to determine how many patients had visiteû the HSC clinic over a

one-year penod (January of 1998 to January of 1999). Patients younger than seven years cannot

accurately perform the pulmonary function and exercise tests necessary for inclusion in this

study, and were therefore not approached to participate. Al1 other patients within the specified

age range who attendcd their three-month routine clinic visit within the designated study period

(March 4-June 6,1999) were deemed eiigible for participation (n=138). Patients w m excluded

if they presented a their c h i c visit with a chest exacerbation or were hospitaiîzed at this tirne.

Under these circumstances, the study coordinator decideci that their selfieport of habinial

physical activity would not k fepresentative of their usual physicai activity pattms.

3.3.2 Subject Recruitment

In oidct to f d t a t c efficient subject rtctuitmcnt, the study coo~nator nviewed cIinic

lists of acheduid patients prior to u c h of t b weekîy CF clinics. EIigible patients were

idencified bascd on tbeit age. Communication with chic staff allowed the study COOfdinator to

detenaine the most appropriate tim to a p p a c h patients. The thme other CF chic study

coordinators were consultai to ensure no overlap in subject recniitment.

Eligible patients were identified and approsçhed by the shidy coordinator while they

waited in clinic. These patients and their families were provideci with both a written and an orai

description of the study protoc01. and were given an oppommity to ask questions and to assess

the time requirements of their involvement. Benefits and rislrs of the study were outüned

Patients gave their informeci consait in W n g upon their agreement to participate. For children

younger than 16 yem, a parent provideci informed consent, and the child gave their o m

informed assent. The consent and assent forms are show in &pend& B.

During each Uuee-hour chic, the study coordinator made contact with as many patients

as possible. In an average clinic, a maximum of two patients could k successfully recniited

Generally, there was not enough time to recnait more than two patients. Depending prtidy on

the patient's understanding of the study nquirernents, the study coordinator spent between 20

minutes and one hour with patients. In addition, patients were often calleci away for routine

testing and consultation with clinic staff. In these cases, data collection was delayed. The study

coordinator had to wait for the patient to ntum to the waiting m m , pnventing the remitment

of other patients.

3.4 Data Collection

3.4.1 SubJect Data

For each patient rrcmitcd, the study coordinator miewed updated chic records to

obtain necessary subject d.ia The patient's age was calcuiateù h m th& d t m e n t and birth

dates. The clinic nurse mutinely collectô measurrs for height and weight and documents them in

the patient's fiie. These values were morded by the study coordinator, and uscd to compute a

masure of nurritionnl statu, percent of ideal weight.

Patients who refused to participate in the study wem documented In order to compare

these non-participants to the study sarnple, relevant data was dso coiiected h m their clinic

records. Subsequently, a cornparison of these hvo groups could be made.

3.4.2 Puimo~~~vy Fundion

Puimonary function tests are routinely conducted in CF patients in the pulmonary

function laboratory at HSC. These tests are designeci to identify and quantify the degree of lung

impairment." For the cumnt investigation, al1 study patients Wonaed pulmonary fuaction

tests either through full pulmonary function testing or spinwetry.

Spirometry is the procedure used to measure the rate at which the lung changes volume

during forced breathing maneuvers. The most commonly performed test procediin uses the

forced expiratory vital capacity maneuver. This procedure, in which the subject inhales

maximalîy and then exhales as rapidly and completely as possible, pmvided the pulmonary

function measuns necded for the culTent studY?O

At HSC, spirometry data is collecteci on the Vmax system (Sensor Medics Corp. Yoba

Linda, Califomia) according to accepted standards?' Pulmonary function variables are expressed

as percent of pmdicted value for height and sex? The four most common measmments are:

forced vital capsCity WC), forced expiratory volume in one second ml), peak expiratory

flow 0, and average forced expiratory flow over the middle 50% of WC

Complete data was available for all study patients on these four measUrCs. These tems are

further defined in 4pendUr C.

PEVi is the most reptoducible70 and sensitive indicator of disease severity? It has

previously been estabiished as the best pndictot of sW7nvai in fl For these nasons. FEVi

was the variable used to r e p e n t pulmonary function for ail data analyses.

3.4.3 Measurement of H;iibituai Pbysicai Activity

Two instruments were used in the cumnt study to quantifj~ habitual physical activity: the

Habitual Activity Estimation Scak and the 7Day Activity Diary. These instruments w e n chosen

because together they pmvided a comprehensive profile of habitual physical activity (frrquency,

intensity, type and duration), and could be seif-administered by the study patients. These

instruments are summarized in Table 3.1.

3.4.3.1 The Habituai Activity Estimation Serk (BAES)

The HAEs was initially designed as a ciinically feasible tool to assess habitual physical

activity levels in children with chmnic disorciers? In the HAES, as outlined in Appendix D, a

typical waking day is delineated by five periods: wake-up time, the thne main me&, and

bedtime. Activity intensity is partitioned into four categones: inactive fie., sleeping). somewhat

idve (i. e. , siîting), somewhcrt actiw (i. e.. waiking) and active (i.e., running). The HAES

captures data on the intensity, fkequency and duration of activity. The somewk active and

&ve categories were used to compiie the Acrivity scores* Patients were asked to recall one

d weekday and one ~ G M I Q I weekend day h m the past two weeks. They were then able to

complete the HAES. by estimating the proportion of timc spent in each of the four activity

categories for the specioed weekday and wetkcnd day nspectively. Data wcre coiiected for a

weekday and weekend day separatcIy to acwunt for any difft~tllccs in habituai activity patterns

for these two periods.

3.4.33 Tba 7 - h y Adlvitg Msrg

In addition to the intensity, hquency and duration of activity measured with the HAES.

the 7Day Activity ~ i a r y : ~ ~ as illustnted in Apperufa E, also captures data on the rype of

activity being perfonned For sevea consecutive days, study paîients useà the Activity Diery to

classify the intensity and type of their activity in 15-minute increments acconling to aine

pTedete&ed categories. Study patients listed the entin m g e of their activities, ngatdess of

intensity level. Category 1 repnsented inactivity (sleeping), Categories 2 and 3 npresented

sitting and standing respectively, Categories 4 and 5 represmted walking, Categories 6,7, and 8

represented physical activities of mil& moderate and hard intensity respectively, and Category 9

repnsented compctitive sport.

The Acriviry Diary Activity scores were compiled from cotcgories 4-9. For each diruy

entry, the patients were also asked to record specific details on the types of activitia they were

involved in. These entries allowed the study coordinator to constnict a profile of physical

activity, and smed as confirmation that the appropriate intensity level had beni documented. As

categories 1-3 represented varying leveh of inactivity, they were not incorporated into the

Activity score. These categories were important, however, as they encouraged patients to report

on the entire range of activity, decrrasing the chances for ovet-estimation of physical activity.

3.4.4 Administration of the LIAES rad the Activity Diary

The HAEs was self4dminisMed by ai l snidy patients in the clinic waiting area.

CompIetion of the two scaks (weekday and weekmd) typically nquind l a s than 15 minutes,

although a few patients nquir#l as long as 25 minutes to cornpletc it. Even though wrïtten

instructions and a sample entry wczc providtd to cach patient, the study coordinator offered

additional instruction and assistance as needed The completcd HAES waa ceninieci to the study

coordinator in clinic, who immediately examined it for completeness and accurscy. Missing

values or apparent m m were âiscwsed with the patient and comcted if necasary.

In clinic, patients were also introduced to the Activity Diary and given extensive

instruction on its p p e r completion. The study coordinator showed patients excerpts h m a

sample diary and reviewed the diffaent ceiegory definitions with them. Patients were sent home

with the diaries, and asked to prospectively record their activity for the following seven days.

During the completion period, the study coordinator telephoned each patient at leapt once to

encourage cornpliance and respond to any concems. The completed diaries were mailed to the

study cwrdinator for scoring and analysis.

The literature suggests that younger children are unable to provide an accurate

description of their own physical activity." Therefon, parents of children younger than 13 years

in the cumnt study were carefully bnefed on completion of the HAES and the Activity Diary,

and asked to serve as proxy nspondents. The study coordinator recommendcd that parents probe

their chilâren about their schwkelated activities, and where necessary, consult with teachers

and schwl staff. The childnn pmvided their parents with valuable input on their activities,

enabling them to cornpletc thcse measures.

3.4s Nutritionai Status

Measuns of height and weight wen used to calculate both height and weight percentiles

and percent of ideal weight, accordiag to standards of Tanner et al?* Although these standards

were publishd in 1965, the methods of meastument are considered superior, and oie mutinely

used h clinical rcscarch at HSC. nie hdght and weight pacentües illustraied whetha the study

patients d i f f ' fnrm the^ heahy pars in the sam agc-sex g r ~ u ~ ? ~ Percent of ided weight

was calculatcd as an hdicatot of nutritional statu.

To compute percent of ideal weight, the patient's height is f h t plotted on a growth cbsit

and the height centile is detelipuled. The ideai weight for height is detennined by finding the

weight on the same centile as that for height. age and sex. Percent of ideal weight is a moasure of

actual weight diviàed by ideal weight for height, multiplied by 1007 Pezcent of ideal weight

was previously found to be a diable indicator of nutritional statw in both healthy subjects, and

in childnn and aduits with CF?

3.4.6 Exedse Caprity

Exercise testing was pesfomed in those study patients who were scheduled for their

annuai exercise test on or near to their study rrcniitment date. The annual scheduling of the

exercise test and the three-month recruitment period meant that only a sample of the total study

patients underwent exercise testing on the same &y as study ncniitment. Exercise data was

accepted hwn the previous or subsequent chic visit (& 3 months) if the patient's ciinical status

had not changeci significantly beiwcm visits. based on the stability of their pulmonary function

test results ktween clinic visits. It has ban previously established that a change of less than

12% predicted in FEVI is within the normal range of ~ariabilit~?~

Patients perforrned a maximal, inmmentd cycling t d 9 0 n an electrically braked cycle

ergorneter (Rodby Electronik AB, Enhoma, Sweden). Similar to previous HSC

investigations?&" incnments were chosen b& on the patient's sex and height, so that the

patient typically terminateci the test in 8 to 10 minutes due to exha~stion?~ The last completed

minute of the exercise test was docmc~~ted as Uu patient's maximai workioad. niroughout the

test, patients b t h e d tluough a mouthpicce connected to a two-way Y-valve of low resistance

and dead space. Aeut rate (Lead II, ECG), oxygen saturation (Ohmeda, Biox 3700 pulse

oximetcr, Louisnlle, KY). inspirrd minute ventilation ~atiQnson-Cowan dry gas meter,

Manchester, England), mixed expired oxygen (Applied Eîectrochernistry oxygen aaalyzer,

Sumyvale CA), carbon dioxiâe (C02-44A rapid nsponse carbon dioxide analyzer, Fitness

Instrument Technologies Corp., Quogue, NY) and respiratory rate (themister) were monitored

continuously on an eight-channel recorder. Ox ygen consumption, carbon dioxide production,

tidal volume and respiratory exchange ratio were measund continuously on-iine through an

automated exercise testing pmgmm dtveloped in the exemise lab at HSC. Wood pnssun was

measured at rest and immediately after peak exercise by a mercury sphygmommometer.

The exercise test was conducted by an experienced exercise physiologist who evaluated

the validity of the data. In several study patients, a maximal effort was not achieved on the cycle

ergorneter due to maturity level and uncontroiIable coughing, and these data were not used The

exercise test provided data on the following mairimal exercise panuneters: heart rate, minute

ventilation, oxygen saniration, oxygen consumption, work capacity, and ventüatory resewe.

These terrns are defined in Appendù C. Values for percent body fat and lean body mass were

caiculated by the exercise physiologist prior to the exercise test. Percent body fat was computed

h m the sum of four skinfolds @iceps. triceps, subscapularis. and suprailliac)?' Lean body m a s

was calculated by subtracting the patient's fat mass (calculateâ h m percent body fat) h m their

total weight.

For the ppurposes of the current study, maximai oxygen consumption and d m d work

capacity were the two most relevant exercise parameters. In tht CF population, maximal oxygen

consumption has k e n suggested as a prcdictor of ~urvival,'~ while maximal work capacity has

been documented as an indicator of overaiî clinicai striais."

3AS7 CampUuiee with Physbthempy

Cornpliance with physiotherapy was an important variable to evaiuate in the cumat

study. Most CF patients are prescribed daily chest physiotherapy to meintain or improve their

pulmonary hction, yet compliance is difficult to encourage. Some patients deliberately replace

their physiotherapy with reguiar physical activity under their own assumption that the benefits

will be quivalnit. Patients who do comply with physiotherapy, on the other hand, may have less

time for participation in regular physical activity. A measun of compliance with physiotherapy

was essential considering its potential to confound the relationship between habitual physical

activity and puimonary function.

Patients were asked to rank their compliance with physiothmpy h m three possible

nsponses. This scoring system was pnviously developed for use in the CF population." The

choices, zero, one and ~ Y O represented poor, pamhl andfil1 cornplimice nspectively. The clinic

physiotherapist was able to verify or dispute the patient's response, if necessary.

3.4.8 Potential Corniates of Wysicai Activity Participation

Potential comlates of physicai activity participation were evaluated through a series of

questions answered by parents. The Pamt Questionnaire (PQ), as shown in Appendix F, was

intendeâ to provide data on environmental f ~ r s that could be nlated to the study patient's

participation in physical activity. The factors of interest included: parental activity level. the # of

organizcû activities that the child was involved in, pfoximity of the household to sporting

facilities, safety of the f d y ' s neighborhood, and the number of exercise-related items in the

home. The shidy coordinator compilcd the PQ fiom relevant questions pubiished in other

A description of the PQ is prmidcd in Table 3.1. surveys.

Table 3.1: Sumrnary of the survey instruments

s w e y nt Rvpasc Type of Approxûnate Recordhg Variables Administration Time to Period Asses&

Adniiaister

Habitual Activity Estimation Scaie wA=)

7-Day Activity Diary

To assess Seif- 15-25 Usual #of hm spent habituai physical administered in minutes week&zy at each of 4 activity levels by chic, patient and uctiviïy levets patient ncall and parentt weekend

day activity

To prospectively Self- 15-20 1-week # of hrs spent assess habituai administered at minutes / day at each of 9 physical activity home, patient activity levels, levels and parentt type of activity

P m t Questionnaire To examine Self- 10-15 Once @QI potential administered in minutes

correlates of chic, parent physical activity participation

parent's activity level, # of organized activities, pronimity to facilities, safety of neighborbod, home exercise equipment

3.5 S m l e Size

As determineci h m the CF database, 188 patients (7-18 years) made at least one mutine

visit to the CF c h i c between January of 1998 and January of 1999. This estimated numkr of

patients saved as the sampling h e for the c m n t pilot study. For the purpose of sample size

estimation, the study coordinator approximated that 50% of the eligible number of patients

would be schedulcd to visit the CF chic during the thnc-month rtcniitmenî period From these

94 potentiai participants, a sample size of 40-50 patients was p d c t e d The final sample size for

this study was 40 patients.

Logistics and estimates of feasibility were used to generate the estimated sarnple size

pnor to the onset of subject recruitment. Time constraints of CF patients in clinic, personnel

limitations, and refusals were the main limitations of sample size for this pilot study.

In order to conduct a longitudinal study based on the demonstrated trends of this pilot

study, a power calculation wül be needed. This power calculation is depicted in AppendLr G, and

will be the basis for a longitudinal study examining the relationship of FEVl with habitua1

physical activity , nutritional status and exercise capacity.

3.6 Data Andysis

The Statisticai Analysis S ystem (SAS), version 6.12, was the sofhvm package used for

statistical analysis. A permanent SAS dataset was c ~ t e d for storage of the snidy data.

3 1 Datasforhg

To compare the habituai physical activity data capnued by the HAES and the Activity

Diary, comparable activity scons werc needed B a d on the activity category definitions for the

HAES and the Activiiy Diary, cotegories 4-5 of the Activiîy Diary repmmted the somewhat

uctnte category of the HAES, and categoriw 6-9 rcprcscntcd the OCtiVe category of the HAES.

3.6.1.1 The HAES

The complete HAES âaîa was entercd into a Microsoft Exce1497 program tha it was

previously developeâ to compile activity data hwn die HAES. The program computed the totai

proportion of tirne (96) that the subject spent in each of the four activity categories (inactive,

somewhat inactive, sornewhat active. and active) for the reported weekday and w e e h d day

nspectively. These percentage values w e n subsequentîy converteci into hours/&y using a SAS

PwP='*

The final HAES Activiry scores, as depicted in Table 3.2, were compiled from the two

categories cepresenting the highest levels of physical activity: somewhat actiw and acrive. Based

on a pnvious study, habituai physical activity can be represented by a combination of the

sornewhat active and active categories (Total Activity), or by the active category (High Activity)

oniy?

For the curnnt study, separate scores were calculated for the HAES data to reflect

diffmnces between both weekday and weekend day activity, and total and high activity levels.

A Weighted Actbity score was ais0 compiled for each of the Total and Hi@ Activity levels to

npresent a combination of weekday and weekend day activity. The Weighted Aaivify score was

created by assigning a weight of five to the weekday activity and a weight of two to the weekend

activity, and then divided by seven.

3.613 T& Activity Dhry

The stuây coordinator developed score sheets for this study to compile activity data h m

the Activity Diary. As the subjects nported theh activity in Srninute increments, each 15-

minute t h e pcriod was &signateci as one unit. For each of the nine activity categories (1-9). a

summry n u m k was computed basai on the totai amber of uni& in that category. The total

number of hours spent in each category was subsequently calculated (total number of units 1 4).

The Activity Diary Active scores. as depicteci in Table 3.2, were based on activity

categories 4-9. In a previous study, a compilation of cutegories 6-9 repsented moderate to

vigorous physical a ~ t i v i t ~ . ~ ~ which was tenned High Activity for this investigation. Subsequently,

a summation of nitegories 4-9 repnsented mild to vigoroue physical activity. which was tenned

Total Actiwity. Similar to scoring for the HAES, separate Activity Diary Actr'vity scores were alao

cornputeci to represent weekday and weekend activity, and Weighred Activity scores were

calculated to combine weekday and weekend day activity.

3.6.13 Activity Type

The activities listed in the Activity Diary were classified individually by their Vpe. A

listing of the different types of activities (categories 4-9 on1 y) was first compilecî from all the

Activity Diaries. The activity types and the time spent perfonning each type were then dlocated

to an activity classification within their spedic activity category. For each activity category

(categones 4-9). the eight possible activity classifications were: porsonol care, domestic wo&

school-related, employment, unstructund play, wwalng, p o r t and tmd$ned. Histogramo were

matcd to compare the amount of time spent in each activity classification.

3.6.2 Data Qwuty Contn,l

The raw dita was enterai into the SAS dataset by the study coordinator, and data checks

were conducted to ensun the accuracy of the data Maximal and minimal values for al1 variables

were examine& and any discnpant values were checkcd against the raw drta In addition, a

netarch assistant u n f d a r with the study entered the data a second time into a t e m m

dataset. A complnson program written in SAS connrmd the accuracy of the data entry.

Table 3.2: Activity scons compilecl h m the BAES and the Activity Diary

High ~criviiy'

Weighted Total Actbity

Weighted High Activity

Catenories: Carnories :

Somewhat 4+5+6+7+8+9 active* + active'

Cate~ories: Catenories :

Active only 6+7+8+9

[(S * Weekday [(Weekday Total Total Activity) Activity) + + (2 * Weekend (Weekend Total Total Activity)]/7 Actinty)]n

[(5 * Weekday [ ( ' k d a y High High Activity) Activity) + + (2 * Weekend (Wcchd High High Activity)]!7 ActiVity)]n

t scores for both weekday and wcekend &y rtspcctjvcly tt somewhat active: e.g., wallting

active: e.g, biking

3.63 Dcreelpîîve StatWct~

Univariate analysis was conducted to compute the range, fnquency. man and standard

deviation (SD) of each study variable. SAS m a m pmgrams were used to compute weight and

height percentiles, percent of i&ai weight. and m e n t pndicted values for pulmonary function.

The study dataset was merged with data files in the Toronto CF database to obtain additional

subject data, including biah date and gender. A dummy variable was created for gender, a

categorical variable.

PROC UNVARIATE was used to compare weekday and weekend Activity scores by

means of a p a i d t-test. PROC 'LTEST was used to evaluate whether any statistically significant

differences wne present ktween males and fernales.

3.6.4 Pearson y s Correlation

Pearson's comlation was used to assess the strength of the association between pairs of

variables through the creation of comlation d c e s . The significance of a comlation

coefficient is a function of both the magnitude of the conelation coefficient (R), and the sampie

size. Therefore, the importance and sangth of each celationship was assessed in addition to the

significance of the conelation coefficient R-square values were assesseci as an approximation of

the amount of variabiiity in one variable that was explained by association with anothet variable.

Scattet plots were also createâ for each of the pairs of interest to evaiuate whether outlying data

points w e n causing the correlation coefficient to be larger or smailer than the overail trend?

The comlations of the different HAES and Activity DÙwy Activity scores with FEVL w m

the primmy focus. As part of mode1 building, the comlation coefficients and pmbability vatues

(p-vaiues) were examined to determine which of the diffenat activity scores &monstrPted the

strongtst relatioaship with FE&. The HAES and Activity Diary ActMty scores that Qmionstratcd

both significant pvalues and the highest comlation coefficients when snaiyzed with FWl wcre

of most interest for regression andysîs.

The correlation coefficients for the association of the ActMry scores and FEVl with age.

gender, peEent of ideal weight, meXimal oxygen consumption. maximal work capacity, and

compiiance with physiotherapy were of secondary interest. The süength of these comlation

coefficients were imponant in esdmating which variables might be important in the relationship

between FEVi and the Activity scores, and thus practical for multiple regression analysis. In

addition, the comlation of the HAES Activiv scores with the Activity Diary ActrYity scores w m

examined to estimate the degree of consistency between these two measunments.

3.6.5 Regmaaion Anaîyics

In sepiuate multiple correlation analyses, either FEVi or one of the Activity scons was

considerad as the outcome variable as the direction of causation between lung h c t i o n and

habitua1 physicai activity could not be established h m this pilot study. The ActWity scores used

in the final ngnssion models were chosen based on the strength and significance of the

association dernonstrateci in multiple comlation.

The expianotory variables were:

i) FEVl or an Activity DiaryMMS Activity score (whichewr wa9 not the outcome), ii) percent

of ideal weight. iii) one of maximai oxygen comumption, or manimal work capacity, iv)

compiiance with physiotherapy, v) age, and vi) gender.

3.6.5.1 Multiple Condation

Muitipk comlations between variables were examined by means of the ail possible

regression mbsets technique. AU possible regession anaiysis rcquires the fitting of every

possible regnasion cquation thrt involves the outcorne variable plus the explanatory variables.

As m e r model building for the final multiple regnssion analysis, the technique of 1

possible regession analysis was used to find the best fitting model. It is pferable to stepwise

regnssion, where important nlationships among variables may be misscd. This technique was

well-suited for this study because the number of explanatory variables was rellveiy mail. The

adjusted R-square values and the Cp Mallow statistic were evaluated to detemine the best

models. PROC REG in SAS produces efficient output of these statistics.

The model that was most pSVSUIlonious (i.e., the fewest explanatory variables explaining

the most variability in the outcorne variable as determincd by the adjusted R-square values), and

for which the Cp Maiiow statistic most closely matched the number of parameters in the model

was considered the be the strongest candidate for final multiple regression analysis.

3.6.5.2 Modehg FEVl as the Outeome Variable

Multiple linear regression involves expnssing the continuously distributed outcome

variable as a linear function of a certain number of explanatory variables." PROC REG in SAS

was used to generate mon detailed results of the best multiple ngression models selected h m

the results of multiple comlation. The adjusted R-square values, parameter estimates and

standard errors were examined to evaluate changes in the amount of variability explained with

additional variables in the model. An interaction term was also included.

RESULTS

4.1 Descri~tive Statistics

The study variables are presented as mean f standard deviation (SD) for males. females.

and the total study population. It was important to look for any ciifferences between males and

femaîes in this sampie because of a nported poorer sumival in fernale CF patients compand io

males?8

4.1.1 Patient Characteristics

4.1.1.1 Study Participants

Twenty-three males (57.5%) and 17 females (42.5%) were recruited nom the CF c h i c at

HSC over a three-month period (March 9-June 4, 1999). Subject characteristics for these 40

study patients are detailed in Table 4.1. Mean height and weight values below the 5 0 ~ percentile

in this group of patients indicateâ subnormal gr0wth.l3 Percent of ideai weight, however, did not

reveal any &gne of maln~trition.'~

4.1.1.2 CF Population Values

Table 4.2 displays the characteristics for 77 males (55.8%) and 61 females (44.2%) who

met the study eligibility criterion (7-18 yem) and also attended the CF clinic during the three-

month study recruinnent peW. The study sample was similar to ali eligibk CF clinic patients.

Table 4.1: Study patient characteristics (mean i SD)

Variable M'aie (a = 23) Feude ( n = 17) Total (a = 40)

Age (ye-) 12.6 * 2.9 12.2 + 3.9 12.4 4: 3.3

HeiQt (cm) 146.9 î 15.8 144.3 + 16.7 145.8 + 16.0 wei%t (irg) 37.9 î 12.1 39,l i 14,2 38.4 î 12.9

Height percentile 34.6 i 27.7 45.8 î 26.2 39.4 r 27.3

Weight paantile 32.2 I 24.0 41.8 î 23.7 36.3 î 24.0

% of ideal weight 99.5 I: 9.9 96.4 + 8.7 98.2 i 9.5

Table 4.2: CF patient diatacteristics for al1 eligible patients attending dinic fiom March 9-June 4,1999 (mean ,: SD)

Variable Male (n=77) Femik (n = 61) Total (n = 138)

Age (Y=@ 11.9 i 2.9 12.6 3.1 12.2 I: 3.0

Height (cm) 144.6 i 16.5 147.3 * 13.5 145.8 + 15.3 @g) 36.9 * 12.8 39.4 & 11.0 38.0 * 12.0

Height percentile 38.7 î 28.4 44.8 î 26.3 41.4 i 27.6

Weight percentile 35.3 î 28.0 35.8 26.5 35.6 î 27.2

8 of ideal weight 101,s î 17.9 95.2 & 12.4 98.7 i 16.0

4.1.13 Study R e k d s

The study coordinator apploachcd 45 patients, 40 of whom agned to participate in the

study. There was an 89% rccruitment rate. Five patients refused to participate: one wos moving

h m the province, one was invoIved in another rescarch sndy, one w u hospitaiizcd at the tim

of recruitment, and two wen UIUIItcrested in the study pmtocol. Relevant clinic daîa for these

two boys and thres girls w m coilected (man f SD: age, 13.3 f 2.7 ycrw; height, 146.8 f 10.6

cm; weight. 36.7 i 7.5 kg; height perccntile, 29.2 i 20.5; weight percentile, 26.3 f 239; percent

ofi&d weight, 97.9 i 17.8). This p u p of study nfusals, who also had subnonnal growth and

nomel nutritional statue, did not differ h m the study patiente.

4.1.2 Measmement of Eabitual Phpicai Activity

4.1.2.1 The HAES

AU study patients completed the HAES either in clinic, 97.5% (39/40) or at home, 2.5%

(1/40). Results from the HAES indicated sigdficantly higher scons for the Total and High

Activity categoties of a weekend day than a weekday, suggestiag that the study patients were

mon physicaüy active on a weekend day than a weekday (sec Table 3.2 for clarification of the

Activity scores). Figure 4.1 depicts the average activity levels for each category of the HAES.

The reported time spent in the Total Activity category for males and females was slightly lower

than that reported in another CF population?2

Although not statistically significant, maies consistently nported more time than females

in the High Activity category, while females reported more tim than males in the Total Activity

category. These results are shown in Table 4.3.

Figuie 4.1 : Average Activity Levels for the HAES

w-Y w- we-Y weekend W-kday wee(cend Weeûday Weekend lCYACTlVE lNACTlVE SOMEWHAT SOMEWHAT SOMEWHAT SOMEWHAT ACTIVE ACTIVE

INACTIVE INACTIVE ACTIVE ACTIVE

Tabk 4.3: Activity hours (hodday) compiled b r n the HAES (mean t SD)

Maies (n = 23) Femaies (a = 1'7) Totai (n = 40) Activity Category

Weekday Score

Total Activityt 5.5 + 2.8 6.0 î 2 3 5.7 î 2.8

High ~ctivity'l 3.1 r 1.7 2.6 1.3 2.9 î 1.5

Weekend Day Score

Total ActivitYt 7 3 + 3.5 8.2 I 2.4

High Activitye 4,s I 3.3 3.1 2.5

Weighted Score

Total ~c t iv i t~ ' 6.2 f 2.7 6.6 f 2.5

High ~c t i v i t~* 3.5 + 2.0 2.7 f 1.3

t somtwhat active + active tt active

4.1.2.2 The 7-Day Activity D h y

Thirty-nine patients (97.5%) a m to complete the 7day Activity Diary at home. One

patient did not agree because he was moving residences, and did not thînk that the diary wouid

be representative of his usual activity. From the 39 patients who were sent home with the

Activity Diary, 33 (84.6%) retumed th& diaries by mail or at a subsequent clinic visit to the

stucly coordinator. Thirty-two of the 33 retumed dimies (97%) were filled out according to

instructions. One àiary was too iacomplete for any score to be estirnated. The study patients who

completed the Activity Diary did not differ frnn those who did not complete it.

The Activity scores reportai in Table 4.4 were based on an average of the 5 nported

weekdays or 2 reported weekend days such that a cornparison with the HAES could be made (see

Tabk 3 2 for clarification of the Activity scores). Average activity levels are iiiustrated in

Figure 4.2. The Activity scons showed similarands to those of the HAES. Aithough not

statisticaily significant, d e s consistently spent mon time in the High Acti~ty category, while

females spent more time in the Total Activity category. Significantl y more tirne was reporteci for

the High and Total Activity categories of a weekend &y than a weekday.

Table 4.4: Activity hours (hodday) compiled from the 7-Day Activity Diary (mean + SD)

Activity Category Male8 (a = 17) Femaies (n = 15) T o u (n = 32)

Weekday Score

Total ~ctivit~' 3.5 * 1.1 4.3 * 1.2 3.9 I 1.2

High ~ c t i v i t ~ ~ 2.3 î 1.0 1.9 * 1.2 2.1 * 1.1 Weekend Day Score

Total Activityt 4.7 î 1.9 5.1 î 2.8 4.8 * 2.3 High ~ c t i v i t ~ ~ 3.1 i= 1.7 2.2 & 2.0 2.7 î 1.9

Weighted Score

Total ActivitYt 3.8 I 1.3 4.5 + 1.4 4.1 f 1.4

High ~ctivity't 2.5 I 1.1 2.0 I 1.3 2.3 f 1.2

t catcgory 4+5+6+7+8+9 tt catcgory 6+7+8+9

Figura 4.2: Average Activity Levels for the Activity Diary

4.1.2.3 A Compuiion of Data h m the HAES and the Activity DtPry

In comparing self-report of activity h m the HAES and the Activity Diary for patients

who completed both instruments, significantly more thne was reported in both the Total and

High Activity categories (weekday and wakend) as reported with the HAES. This suggests that

the HAES and the Diary may have been capniring different dimensions of activity.

4.1.3 Rilmonary Funclion

Pulmonary function data, illuscrated in Table 4.5, were coUected for al1 patients.

Depending on the statu of their lung disease ml), patients were classifiecl as sevcrely (do%

predicted), moderately (50-8046 pndicted) or rnildly (> 80% predicted) impaireci, Twenty-six

patients (65%) had mild pulmonary impallment, 1 1 patients (27.5 9) haâ moderate impairment,

and thne patients (7.5%) had severe impairment, demonstrating that this sample was comprised

rnainly of patients with mild pulmonary disease. This high prevalence of mild lung impairment

has been previously documenteci for pre-adolescent and adolescent CF patients?

Table 4.5: Percent of precücted pulmonary function (mean i SD)

Variable mes (a = 23) Fe- (n = 17) ToW (a = 40)

ml 86.7 + 18.4 84.2 I 24.5 85.7 i 21.0

FVC 89.8 i 16.0 90.9 A 18.0 90.3 I: 16.7

FEF25-75 74.8 î 28.8 66.8 +31.6 71.4 * 29.9

PEF 83.5 î 15.2 80.8 î 21.8 82.3 r 18.1

Definition of abbreviations, sec Appendix C

4.1A.l A sub-sample of study patients

Only patients scheduled for their annual exercise test on or near to C+ 3 months) their

nmiitment date had data on exercise capacity Although exercise data was available fm 26

patients (65.0%), the exercise tests fiam 2 patients (1 M, 1 F) were considend invaiid by the

exercise physiologist. The female patient was too immature to perform the test, and was

uncooperative with the testing protocol. The male patient was unable to complete the test due to

uncontrollable coughing. These results were not included in the final analysis. Twenty-two of the

26 patients (85%) completed their exercise test on the same day as their study ncruitment date.

The remaining 4 patients (15%) completed their exercise test prior to, or foilowing their

rrcniitment date & 3 months). As a nsult of the two invalid exercise tests, final exercise

capacity data was avaüable for 24 patients.

4.1.4.2 Amessrnent of exercise capacity

Values for exercise parameters are docurnented in Table 4.6. Study patients (n=24) had

mean maximal values for workîoad and oxygen consumption slightly lower than those ceported

in a similarly aged group of hdthy ~hildren,~ suggesting that these CF patients had reàuced

enmise capacity compared to normal, school age ~ h i l d r e n ? ~ ~ Fernales had a statisticaily

significant higher percent body fat than males. These values are consistent with trends presented

elsewhere." reflecting mual clifferences between heaithy boys and girls?'

Table 4.6: Exercise parameters (mean t SD)

Variable Mlaiea n Fernales n Total n

(beatslmin) 179.0 k6.1 14 175.3 + 14.4 10 177.5 * 10.3 24

V-(~min) 70.6 f 24.1 14 60.6 I 19.9 10 66.5 i 22.5 24

Sa02 at max (96) 9S.Oî1.4 12 93.6k3.6 9 94.4î2.6 21

Vo2,max min) ~.s*o.s 14 1.2î0.4 10 1,4*0.5 24

v 02max (ml.kg%nin-L) 39,7 i 5.8 14 34.5 î 7.9 10 37.5 i7.1 24

Wmax (46 predicted) 82.2 i 9 .8 14 86.6 a 19.1 10 84.0 î 142 24

MVV 72.2 î 26.1 15 64.4 t26.9 12 68.7 f 26.3 27

Q M W 97.0 I: 18.1 14 103.4 I 35.4 10 99.7 î 26.2 24

% body fat 12.5 î 2.9 15 22.3 *3.2 11 16.7 î 5.8 26

Lean body mass (kg) 32.3 * 11.2 15 27.6 * 9.6 11 30.3 î 10.6 26

Max, parameter at maximal woriûoad; Definition of abbreviations, sec Appeadix C

4.1.5 Cornpliance with Pbgsiothecapy

Patients were asked to rank their compiiance with physiotherapy h m choices of O, 1 or

2. The mean compliance value was 1.6 î 0.6 (males, 1.7 t 0.6; fernales, 1.5 t O.7), which

indicated a hi& reporteci mean level of cornpliance with physiotherapy.

Parent questionnaires, comprised of six questions, were successfully completed by 97.5%

of parents. One parent took the questiomsire home to complete because of the constmints in

chic, but did not retum it to the study coordinator. The data are presented in Table 4.7

In Question 1, parents ranked their own kvel of physical activity participation, based on a

scak of 1-5 (1, inactive; 5, vigomus physical activity). In Question 2. patents reportcd the

number and type of organized physical activities in which their childrcn were i~ivoIved. The most

commonly listed organizations were a sport's team or league (33.3%). and public park or

m a t i o n department activities (28.2%). In Question 3, the child's accessibüity to exercise

related facilities in their community was assessed. The most commonly cited facilitics were a

public park (949%). playing field (82.1%). skating rink (74.4%). and public recteation center

(71.8%). In Question 4, parent's evaluated the safety of theK neighôorimd (1. very unsafe; 5.

very safe). For Question 5, parents chose from a list of chmcte~9tics that b a t describeci their

own neighbohood. This question was intended to estimate the number of neighborhood-nlated

features that were either conducive or non-conducive to exercise. The most commonly cited

feanues were a high fresuency of people walliing or exercising (84.6%). enjoyable scenery

(82.1%). sidewaiks (79.5%). and street Lights (69.2%). Findly, in Question 6. parents estimateci

the number of exercise-nlated items in their home. The most commonly cited items were

running shoes (100%). bicycle (97.4%), sports equipment (Le., b a h and racquets) (97.496). and

skates (89.7%).

Table 4.7: Potential cornlates of physical activity pdcipsiion (man t SD)

Variable Males (n = 22) Femaim (n = 17) Totai (n = 39)

Parent's activity level (range: 1 to 5)

Numbu of organizeâ physical activities (range: O to 8) 1.5 * 1.4 1.2 I 1,O 1.3 I 1.3

Accessibility of facilities (range: O to 18) 9.3 * 3.8 10.4 * 4.4 9.8 k4.1

Neighborhood spfety (range: 1 to 5)

Neighborhood feaaires conducive to exercise 2.4 * 1.3 2.3 I 1,3 2.4 î 1.3 (range: -4 to +4)

Exercise-rdated items in the home (range: O to15) 7.0 î 1.7 6.8 I 1.9 6.9 I 1.8

4.1.7 Activity Classification

The activity types reported by patients for Categories 4 to 9 of the Activity Diary

represented activities of mild to vigorous intensity. The study coordinator compiled a list of

approximately 200 terms used by patients to describe their activity. This list is found in Appendix

H. Many activity types were listai wider mon than one category of intensity. Soccer, for

example, was listed by the same individual in both Categories 8 and 9. The 2 0 t e m were

subsequently aüocated to one of 8 classifications: personal cure, domestii work, schwl-related,

entploynient, unstmcturedplay. walkhg, sport, and undefine& These classifications were chosen

to encompass and describe al1 the diffcrent activity types in the most specific manner possible.

The iesults of this activity classification are prcsentcd in Figures 4 3 to 4.8 for each of the six

categories (4 to 9). The results are presented in average ddly minutes. The following section

outlines the trends in the data according to activity classification.

4.1.7.1 Pemnal Care

Personut cure encompassed activities such as getting ready for school or bed, and CF

seIf-treatment. These types of activities were only reported for the two lowest intensity

categones, categories 4 and 5. In Category 4, an average of 53 daily minutes was reporteci, with

this number decreasing to 0.4 daily minutes in Category S.

4.1.7.2 Domestic Work

Domestic work (household and outside chores) incluàeû cleaning, cooking, dusting,

mowing the lawn, shopping, watering the plants, taking out the garbage. and raking. Activities

included in this classification were nported in Categories 4,s. and 6. The average time spent in

this classification deciined with increasing intensity of the category. In Category 4, 13.0 average

daily minutes were reported, with this number decreasing to 4.0 and 2.3 minutes for Categories 5

and 6 respective1 y.

4.1.73 School-Related

Childm also reported their participation in specific school-reluted ach'w'ties such as

classmom activity, homework, acting, helping the teacher, and school field trips. Schwlielated

activities were reported for Categories 4,s and 6 on1 y, indicating that these activities were not of

a high intensity. The highest report for these activities is 2.0 daily minutes for Category 4,

followed by 0.9 minutes for Category 5 and 0.13 minutes for Category 6.

4mle7m4 Empbyment

As this study sample was compriscd of chilârcn as old as 18 years. an employment

classincation was inclpded Employment included Ifter-school and weekend job, in addition to

delivering papem. These Pctivitie~ were Listed in Categories 4 to 7, indicating the range of

intensity requhements in these jobs. For Category 4.62 average daüy minutes were spent in

employment, with this number dtcreasing dramaticaiîy to 0.13 minutes in Category 5. h m

Category 5 to Category 6 there was a large inmase to 7.4 minutes, foilowed by a demase to 0.8

miriutes in Category 7.

4.1.73 Unstmchucd Play

Both childm and adolescents engage in fkee play. Although this type of activity does not

meet the structured definition of sport, it cm nevertheless be quite ckmanâing, and yet erratic in

nature. Uhstmctured play was reported for al1 six categonui, Categones 4 to 9. The activities

that were labtled in this classification included group games (hide and seek, hopscotch), recess,

playground activity. swinging, and bail playing. In Category 4,4.6 average daüy minutes were

nported. A predominance of unstruchind play was documenteci for Categories 5 (12.4 daily

minutes) and 6 (13.1 daily minutes). Values for Categories 7 (4.6 minutes), 8 (0.8 minutes) and 9

(0.07 minutes), however, demonstrated that less time was spent in utlstructured p l q at these

higher intensities.

4.1.7.6. Wallring

The definition used to describe the inteusity levels of Categories 4 and 5 were walking

indoors and waîking outdoors respcctively. Correspondingly, study patients reportcd the majonty

of walking time in these catcgories. Activities in this classification includcd going to lunch,

sociaiizing with fnenâs, wslking the h g , Wring work breaks, and walking 'harâ'. A total of 20.2

daiiy minutes were reportcd for Category 4 and 37.2 minutes for Category 5. Values for

categories 6 (0*6 minutes) and 7 (0.7 minutes) indicated that cMdren spmt minimal time

participating in high intetlgity waiking.

4.1.7.7 Sport

Sport is a structmd, definable activity that cm be p e r f o d at many different intensity

levels. For this study sample, sports were npoited in a l l six categories and there were a wide

range of activity types: mnning, b h g , track and field, trampoline jumping, dancing, fis-

bowiing, fighting, swimming, and lacrosse. For Categories 4 and 5, only 0.27 and 0.87 average

daily minutes were reported, demonsûating that childnn did not rank their sporting activities in a

low intensity category. Categories 6.7, and 8 on the other han& wcre comprised of 27.5,35.9

and 18.4 daily minutes of activity nspectively. By cornparison, the total hours spent in sporiing

activities of higher intensity, category 9 (4.2 minutes) were relatively small. This suggested that

this group of CF chifdren were spending relatively little time in competitive sports compared

with sporiing activities of lower intensity.

4.1.7.8 Undeîined

Some study patients did not complete the Activity Diary according to instructions, and

subsequently, recorded the category intensity of the activity without listing the activity type. So

as not to exclude this data fiom the activity profiles, the total number of hours at the specific

intensity was classifieci in a separate classification, undejlned.

Figure 4.3: Activity Classification (category 4)

Figum 4.4: Activity Classification (categoy 5)

Figure 4.5: Activity Classffication (category 6)

Figun, 4.2 Activity Classification (category 8)

4.2 Comlation Analyses

4.2.1 The BAES and the Adivity DLuy

Pearson's comlation coefficients were calcdated to evaluate the relationship between the

Activity scores of the HAES and the Activity Diary. A separate comlation was conducted

between each Activity score of the HAES and the comparable Activity score of the Activity

Diary. Table 4.8 illustrates the correlation between the Weekday, Weekend, and Weighted Total

and High Activity scores of the HAES and the Activity Diary respectively.

Positive associations were confirmed for each pair of activity scores. A significant

comlation was demonstrated for the Weekday High Activity scons of the HAES and the Diary,

and for the Weekend Total Activity scores. For the Weighted Activity score, significant

comlations were demonstrated between both the Total and High Activity scores of the two

instruments. Although the strength of the comlation coefficients are similar to those reported

between other pairs of self-report instnrment~,~ the relatively moderate values (R < 0.5) imply

that the HAES and the Activity Diary were only captwing some of the same dimensions of

habitua1 physical activity, or suggest the presence of different reporting biases for the two

instruments.

Table 4.8: Comlation between the Total and High Activity scons of the HAES and the Activity Diary (n = 32)

Weekday Weekend Weightcd

R P R Q R P

Total Activity .33 .O7 A4 .O1 37 .O4

High Activity Al .O2 -31 .O9 .O9 ,005

4.23 Habituai Paysicaî Activïty

Comlation analyses were conducted for both the HAeS and the Activity Diary to

evaluate the association of each of the Total and High Activity scores (weekday, weekend and

weighted) with M l , 96 of ideal weight, age, gender, cornpliance with physiotherapy, Wmax a

and V%mx. Table 4.9 &tails Weekday Activity, Table 4.10 illustrates Weekend Activity, and

Table 4.1 1 depicts the Weighted Activity score, a combination of weekday and weekend activity.

Table 4.9 demonstrates several important nlationships. Positive, significant cornlation

coefficients were demonstrated for the associations of the Total and High Activity scons of the

HAES with FEVi, 96 of ideal weight, and Wmax. Pattems for the Activity Diary were very

different. A negative, significant relationship was demonstrated between the High Activity score

and age, and a rnarginally significant positive cornlation was seen between Total Activity and

gender, suggesting higher Total Activity for fernales.

As Qcumented in Table 4.10 for weekmd activity, then was less overall correlation of

Activity scores with candidate comlates. A positive, significant relationship was dcmonsmted

between the HAES Total Activity score and FEVI, and between the HAES High Activity score t

and V02max. For the Activity Diary. a negative, significant relationship was demonstrated fot

the Total Activity score with age, suggesting a decnase in Total Activity with age.

The comlation andysis for the Weighted Activity Scores of the Activity Diary and the

HAES are shown in Table 4.1 1. No significant correlations were âemonstrated for the Activity

Diary. For the HAES, however, positive and significant nlationships were demonstrated for the

Total Activity category with FEVI, % of idepl weight, and Wmax. These results are similar to

those seen for the Wœkday Activity scons. Positive and significant nlationships were dm

d e r n o m for the High Activity categocy with FEV,, and %rum.

Table 4.9: Correlation for Wee* Activity scores of the HAES and the Activity Diary

Totai Activity High Activity Totai Activity High Activity

Variable R P R P R P R P

Physiotherap y -, 14 .37 -.O7 .65 -. 19 -30 .19 .29 cornpliance

t n = 24 for HAES, n = 18 for Diary

Table 4.10: Correlation for WeeRend Activiîy scores of the HAES and the Actinty Diary

Totai Activity High Aeüvity Totil Advity High Activity

Variable R P R P R P R P

Ph ysiotherap y -.22 .18 0.22 .18 .18 .32 .26 -16 cornpliance

t n = 24 for HAES, n = 18 for Diary

Table 4.1 1: Correlation for Weigkd Activity scores of the HAES and the Activity Diary

BAES Activity (n=40) . Diary (n = 32)

Totai Activity m@ ~ftivitg ToW Actîvity Hi* Activity

Variable R P

mv1 .44 .O05

% of ideal wt 38 .O2

Age -16 .32

Gerider .O9 -58

Ph y siotherap y -. 18 .26 cornpliance

Wmaxt .41 .O5

Y~2maxt .21 .32

t n = 24 for HAES, n = 18 for Diary

4.23 WcdrQy and Weekend Activlty

As presented in Table 4.12, cornlation anaiysis demonstrated positive and significant

comlations between Weekday and Wakend High and Total Activity scores for each of the

HAES and the Activity Diary. These results &monstrateci that children who reported king

physicaliy active during a typical weekday also reported hi@ levels of physical activity for a

typicai weekend day.

Table 4.12: Comlation behveen Weekday and Weekend Activity scons

Totai Activity High Activity

4.2.4 FE&

Table 4.13 demonstrates the comlation analyses for Fmt with 46 of ideal weight. age,

gender, cornpliance with physiotherapy, V b a x and ~ 0 ~ ~ . These analyses were impomt in

beginning to understand the relationships between FEVl and those variables which may be

important in the multiple regression models.

These data indicateâ a significant conelation for FEVi with % of ideal weight and

Wmax. Although there are no other significant comlation coefficients, the direction of the

cornlation coefficient for age is consistent with previous studies that have shown FEV1 to

decluie with age?*

Table 4.13: Comlation Aiialysis for FEVl (n 340)

4.2.5 Potentid Cornlates of Phpical Activity farticipation

The degree of association between each of the Activity Scons and the potential correlates

of physical activity participation was evaluated by means of a cornlation coefficient. The nsults

are demonstrated in Table 4.14. The HAES Activity Scores demonstrated the strongest, positive

relationships with two variables: parent's physical activity level and orgmized physicd uctivity

involvement. The Weekda y, Weekend, and Weighted Total Ac tivity scores demonstrated

significant relationships with parent's physical rtivity level. The Weekend Total, Weekend

High and Weighted High Activity scores demonstrated significant associations with organized

physical activity. The Weekday and Weighted Total Activity scons of the HAES also showed a

si gni ficant nlationshi p with accessibility &xercise-related facitities. For the Activi ty Diary, on

the other han& there was only one positive. significant comlation: the Weighted Total Activity

score with organized physical activity. The nsults suggested a positive nlationship of parent's

physical activity level, involvement in organizcd physical activity, and accessibility of exercist

relateci faciiities with habituai physical activity.

Table 4.14: Correlation bctwecn Activity sams and potentid correlates of physicai activity participation

WeekendT .43 ,007

H -10 .5s

Activity (n=32)

WeekendT .O2 -91

Wetkend H .O9 .61

4.2.6 A Siimmsrg of Resuîts h m Correlation Analyses

The correlation analyses were performed to probe the strength and direction of the

nlationships behveen the Activity scores of the Activity Diary and the HAES, and between the

Activity scons and the other study variables, most importantly FWI. The intent of this process

was to decide which Activity Scores were of most interest for further inclusion in multiple

regression analysis.

Positive and significant associations were demonstraad between the HAES Activity

scores and FEVi. The Weekday High Activity score demonstrated the strongest, significant

comlation with FEVl, followed by the Weighted Total Actinty score, the Weekday Total

Activity score, the Weighted High Activity score, and finali y, the Weekend Total Activity score.

Only the Weekend High Activity score was not strongl y correlated with FEVl. Scatter plots

demonstrated that these correlations were not artifactual; then were no outlying variables that

greatiy enhanced or obscured an o v e d pattern. Results h m comlation analysis between the

Activity Diary Activity scores and FEVI, b y cornparison, àemonsaatcd weak associations.

Although further analysis of the Activity Diary scores took place, mon emphasis was placed on

the inclusion of the HAES Activity scores in multiple regression models, specifically those

scores that àemonstrated the highest correlation coefficients with FEVi.

In determinhg which other study variables couid be of importance in multiple regession.

the comlation coefficients and pvaiues ftom Table 4.13 (Correlation Analysis for FEV1) were

assessed. Pacmt of i&al weight demonstrateci the shwigest correlation with FEVl, foilowed by

Wmax and y. Percent of ideai weight and age were expected to be important in m e r

modeling of Fm in this study. Duc to the small sample of patients with &a on exercise

ca@t~ (n = 24). Wmax w a e not condaed for incIusion in hrther anaiyses.

Their potnuial importance in a friture study of habituai physical activity with a larger sample,

however, shouid not discounteci. Signincant, positive correlations were also demonstrated for

these exercise parameters with the HAES Activity scores.

Two of the study variables that were included in Pearson's correlation analysis, gen&r

and compliance with physiotherapy, were not continuous. For gender, a binary variable. pvalues

for the test of no comlation between gender and masures of FEV, or activity reflected identicai

p-values for t-tests of no difference in mean values khueen males and females. The ordinal

compüance score was developed in an attempt to quanti@ an underlying continuous

measunment of physiotherapy compliance, and was therefore treated as a continuous variable

for this study. Whether or not this score does in fact represent an underlying linear variable will

be explond in hi- long-term study of a larger sample.

Although the Pearson's comlation coefficient may not be a valid estimate of the true

population comlation coefficient for non-normal variables, R is still appropriate for the test of

the nul1 hypothesis of no comlation, provided that one of the variables is normal? The literature

further suggests that a Pearson's comlation coefficient is appropriate for a categorical variable

that is assigneci meaningfd numericd values?

Although the conelation of FEVl with geader and compüance with physiotherapy was

c ' neither stmng nor statisticaily signifiant, both variables were ntained for regression analysis

due to an established rclationship betwœn age and FEV1. and because of the potential

importance of compliance with physiotherapy to both FEVl and habituai physical activity.

4.3 m s s i o n A n d m

Although complete a was avaiiabie for 18 of the 40 study patients, regression anaiyses

wen not conducted on this d sample of the total study population. Analyses werc conducted

on eitha the total numbcr of study patients (n = 40). or on the sample thaî completed the Activïty

Diaries (n = 32).

4.3.1 Multiple Correlation

Multiple comIation analysa w a e conducted by means of the al1 possible remsion

subsets technique. In separate analyses. eitha PEVl or one of the Activity scores was considmd

as the outcome variable. The other variable (FEVl or one of the Activity scores) was then

included as an explanatory variable in addition to age, gender, cornpliance with physiotherapy

and % of i&ai weight The two best models for each Activity score were chosen and have been

presented in Tables 4.15 and 4.16. Tk best models were chosen with consideration to both the

adjusted R-square value and the Cp Mallow statistic.

The R-square is the fraction of the variance in the dependent variable that the regmision

mode1 explains. It dways inmases as more variables are added to the ngnssion equation, even

if these new variables have very Little preâictive value. The adjusted R-square (R2adj), on the

other han& makes you pay a penalty for using additional degnes of fnedorn?' Thmfore, the

adjusted R-square allowed for a cornparison of models with difierent numbers of parameters in

this analysis. The best regression models were g e n e d y considend to be those with the largest

adjusted R - ~ q ~ a r r .

Although the sdjusted R-square was the primary criterion in chaosing the k s t models,

the Cp Mdow statistic was also used to screen for the best regrtssion models. The vdue for Cp

had to be within an acceptable range, near to p @ = the # of parameters = # of preâictor

variables + inteme~t)?~ Thacforc, a combination of fwtors was considercd in selccting the best

and m a t parsimonious models for nnal regrwsioa anrlysis.

Table 4.15 demonstrates the inter-relationships between the Activity scores and the

explanatory variables that best accounted for the variabïiity in these scores. For the HAES

Activity scores, the best individual predictors of these scores were FEVl, followed by percent of

ideal weight. For the Activity Diary Activity scons. on the other hand, either age or gender was

the best individual predictor variable, followed by compiiance with physiotherapy. The highest

adjusted R-squares were demonstrated for the HAES Activity scores. The explanatory variables

accounted for less variability in the Activity Diary Activity scores, suggesting that other

unknown variables were contributing to the variability seen for these scores. This effect was not

due to the reduced sample size (n=32) because when analysis was conducted for this smaller

subset, the R-square values were very similar. The stmngest relationships were demonstrated for

the HAES Weekâay Totai Activity score with age and FEVl (R2adj = 0.27) and for the HAES

Weighted Total Activity score with age and FEV ad adj = 0.23).

Table 4.16 contains the results for multiple comlation analyses where FEVi was the

outcome variable. The results are listed according to the Activity score that was included in the

list of potential predicton. With inclusion of a HAES Activity score as a potential explanatory

variable, percent of ideal weight was the ks t individual predictor of FEV,, followed by the

HAES Activity score. When an Activity Diary score was included as the activity variable,

percent of ideai weight was again the single ks t individual predictor of FEVl, foilowed by either

the Activity Diary scom or cornpliance with physiotherapy. The two best models containeci one

of two HAES Activity scores as a candidate pndictor: the Weekday Totai Activity score. or the

Weighted Totai Activity score. The multiple correlation of Weekday Total Activity, age, and

percent of ideal weight with FEVi pioduccd an adjustcd R-square of 0.41. The comlation of the

Weightad Total Activity score, age and percent of ideal weight with FEVi produccd an adjusted

R-square of 0.40. A separate analysis conducted in the subset of patients who completed the

Activity Diiuy (n =32) also iadicated that the inclusion of these two HAES Activity scores in

separate comlation analyses accounted for the most variability in ml. These analyses

demonstrateci that these two HAES Activity scores ahould be included as potentiai explanatory

variables in fiuther multiple repssion analyses on FEVI.

4.3.1.1 Summ~ry of Muiüple C o r d t b n

n i e primary intent of this multiple comlation malysis was to examine the potential

predictors of FEVi in this study sample. Specificaily, it was important to evaiuate which Activity

scores explained the most variability in mi, the most important variable in desmibing lung

disease in CF. FEVi has already been indicated as the best predictor of survival in CF

and is the primary outcome variable of interest in many CF clinical studies.

Multiple comlation anaiysea were conducted with either one of the Activity scores or

with FEVl as the outcome variable. The results h m these separate analyses confiied a sirnilar

pattern: a positive relationship was demonstrateci between FEVl and the Activity scons, and this

relationship was strongest for the HAES Activity scores. Specificaily. two HAES Activity scores

were demonstnrted to explain the most variability in W1. Candidate explanatory variables also

explained the most variability in these same two HAES Activity scons.

When FEVl was the outcorne variable, the adjusteci R-square values for the best multiple

variable models were higher than the adjusted R-square values for the k s t models whm the

Activity scores wem the outcome variable. #en the HAES Activity scores w a e the outcome

variable, the adjusted Rsquan values ranged h m 0.1 1 to 0.27, meaning that 11% to 27% of the

variabiiity in these HAES Activity scores were explained by the candiAatc explanatory variables.

Whm the Activity Diary scores w m entend as the outcome variable, 5% to 17% of the

variabüity in these scons was explained by the explanatory variables. Pinaliy, w h m FEVl was

entered as the outcome variable, 33% to 4196 of the variabiîity was explainai by the set of

explanatory variables that iacluded the HAES Activity scons, whereas 24% to 27% of the

variability was explained by the set of explanatory variables that included the Activity Diary

scores.

These ~ u l t s confirrned the importance of FEVi as the primary outcome variable for

M e r multiple ngression analysis, and clarified the significant role of the HAES Activity

scores, as compared with the Activity Diary scons, as the candidate explanatory variables to

represent habituai physical activity. In particular, two RAES Activity scores wen indicated for

inclusion as explanatory variables in the final ngression models (Weekday Total and Weighted

Total Activity). In addition to these two Activity scores, age and % of ideal weight were also

indicated as potential explanatory variables based on their consistent pnsence as expianatory

variables of FEVl in multiple comlation. Plots of FEVl by age, percent of ideal weight,

Weekday Total Activity and Weighted Total Activity have been included in Figures 4.9-4.12 to

illustrate the nlationships ktween these variables.

Gender is an important variable in pndicting prognosis of CF patients, and is therefore an

important variable to include in any suMval model. For this analysis, however, genàer did not

enter into any of the best models where FEVi was the outcome variable, even though it did

consistently tum up in the modcls where the Activity scons were the outcome variable. When

both the Activity score and gen&r wcre e n t d as cuididpte explanatory variables, the Activity

score was always selected as the more important explanatory variable. This suggcsted that both

veriabks were explainhg simüar variability in FEVt. Widiin the nauow age nnp of this study

sample, gender did not have any important pradictive vdue in the model. This may nflect the

gender effect being to some extent explaincd by the Activity scores. These nsults do not

undermine the importance of gender in predicting CF survival. Nonetheless, gender was not

includeâ as a candidate predictor variable in final ngnssion analysis, which focused on the

parameters for predicting FEVi.

Table 4.15: Multiple comIation for the Activity scons

Activity Seare ss Best Modeis the Outcorne

WDTHAES Age. ML 0.27 1.9 Age, Gender, FEVl 0 2 7 2.8

WDHHAES FINl, 96 ideal wt Age, FEV1, 96 ideai wt

WTHHAES Genâer, FEVl 0,15 2,4 Age, Genâer, mi, Compiiance 0.14 4.8

WDTAD Gender, FEVl Genâer, FEVl, 96 i&al wt

WETHAES mi

FEVI, Compliance

WDHAD Age Age, Gender

WTT AD Age, Gender Age, Genâer, 96 ideal wt

WEHHAES Age, Gender, Cornpliance 0.10 2.3 Age, Gender, ml, Compüance 0.08 4.0

WTHAD Age. Gender Age. Gender, Compüance

WEHAD FEVl, Cornpliance Gcnder, FEVl, Compiiance

AD, Activity Diary; cornplionce, complianee with physiotberrpy; wt, weight; WDH, Wdcday High; WDT,

Table 4.16: Multiple comlation for ml, with diffemt Activity scores included among the candidate predictols

A~tivity Seor~ Bcet Md& AdJ. R- C@) included as a square Mlailow Candidate Predic tor

wDmAES Age, WDTHAES, 9b ideal wt 3.8 ~ g e , WDTHAES, Compliance, % ideal wt

Age, WTIKAES, % ideal wt Age, WTTHAES, Compliance, % ideal wt

Age, WDHHAES, % i&al wt Age, WDHHAES, Compliance, % ideai wt

Age, WETHAES, 9b ideai wt Age, W ' , Compüance, % ideal wt

Age, WTHHAES, % ideal wt Age, WTHHAES, Compliance, % i&d wt

Age, % ideal wt Age, Compliance, 96 ideal wt

Age, Compliance, % ideai wt Age, WEHAD, Compliance, 9b ideal wt

Age, Compüance, % ideai wt Age, WTHAD, Cornpliance, % ideal wt

Age, Cornpliance, 96 ideal wt Age, WETAD, Compliance, 46 ideai wt

Age, Cornpliance, % i&d wt Age, WDT'AD, Cornpliatm, % ideal wt

Age, Compüance, % idcai wt Age, WDHAD, Compii~na, 9b ideal wt

Age, Compi.îance, 9b ideai wt Age, WITAD, Compiianœ, 96 ideal wt

AD, Activity Diary; wt, weight; cornpliince, compiinnct with physiotherapy; WDIE, Weekday High; WDT,

Wakday Total; WET, Weekend T e WEK Wabend HÏgh; WTT, Weighted Totai; WTH, Weighted High

\

Figure 4.9: Pbt of FEVf by Age

\

Figure 4.12: Plot of FEVl by HAES Wdghbed Tiptel AdMLy

Multiple linear regression was pafonned to determine the nature of the individual and

combined effects of t h . explanatory variables on PEVi. The explanatory variables were: i) age.

ü) percent of ideal weight. and iii) either the HAES Weekday Total Activity score or the HAES

Weighted Total Activity score. As there were two Activity scons to consider, the same steps

were followed to include each of these scons into the different potential madels.

Aithough both FEVl and the Activity scores were outcome variables in the al l possible

regcession analysis, only FEVl was an outcome variable in the final multiple ngnssion because

the.primary nlationship of interest was the explanation of variability in FEVl by habitua1

physical activity. In addition. only the HAESActivity score was used as an explanatory variable

for activity in the final multiple regression andysis because the Activity Diary Activi~ scores

demonstrated lirnited association with FEVl in both Pearson's correlation and al1 possible

regression.

Every possible combination of variables was examined. For each model, the regression

coefficients and theu standard emrs were evaluated to detemine the relative influence of each

variable on The regression coefficient signifies the amount of change in the outcome

variable per one unit change in the explanatory variable, while controlling for al1 other variables

in the model. If the ngression coefficient was large relative to its standard emr, then that

variable was considend to be a sipificant pndictor of FE&. If the value of the regression

coefficient nmained stable with ctdditional vuiabies in the model, then the vaiable was b m e d

an independent prcdictor. If the tegression coefficient did not remain stable wiüi the addition of

more variables, then confounding of the varirrble's effit was indicatcd. The results an pnseated

in Table 4.17. .

43.2.1 Age

The negative association between age and FEVl bas been previously established,** and is

M e r confirmed by results h m simple comlation between these two variables (R= -0.26,

pcO. 11). Aithough not statistically sigdficant, the negative sign of the correlation coefficient

suggested that FEVi decreaseg with age. From multiple comlation, we see that age is the ihird

best individual predictor of the varhibility in FEVi, following behind percent of ideal weight and

the HAES Activity score. When age was entend into a model as the oaly predictor variable of

FE&, 4.0% of the variability in PEVl was explaineci, as indicated by the adjusteci R-square

value. The regression coefficient demonstrated that for a one-year increase in age, FEVl would

demase by 1.63% predicted. This npss ion coefficient, however, was not significant (p=0.11).

4.3.2.2 Percent of Ideai Weight

In addition to FWi, nutritional status is also an indicator of disease severity in CF. From

simple comlation, we see a comlation coefficient of 0.47 (p=û.O2) between FEVl and percent

of ideal weight. demonstrating a positive and significant nlationship between these variables.

From the nsults of multiple comlation. percent of ideal weight was consistently the single best

individual predictor of FEVI.

With percent of ideal weight entaed individuaiiy into a model of FINl, the adjwted R-

square was much higher than the value seai when age was mtered singly into a model. However,

this was expccteâ because a significant nlationship was demonstrated between FEVi and percent

of ideal weight in simple comlation, but not betwem FEVL and age. The adjutcd R-square

indicated that 20.0% of the variability in FEVi was explpiaed by percent of ideal weight for this

study population. The regrasion coefficient, for which the pvalue was sippincmt @ = 0.002).

indicated that for an increase of 1% in p e n t of ideai weight. there was en incnase of 1.05 %

predicted in FEVi.

43.2.3 HAES Activity Seores

Although a causal nlationship between FEVl and habitua1 physicd activity has not been

estabLished this snidy hypothesized that a positive nlationship would be demonstrateci between

these two variables. Results h m comlation analyses (simple and multiple) indicated that the

HAES Activity scores, as compareci with the Activity Diary Activity scores, had the strongest

comlation with FEVl and thus explaineci more variability in FEVl than the Activity Diary

scores. Fmm the six possible HAES Activity scores, two were selected from multiple comlation

analyses as the best potential preàictors of FEVl: Weekday Total Activity and Weighted Total

Activity. These scores were factored into separate regression models.

By itself, the Weighted Activity score explained more of the variability in FEVl (17.0%)

than did the Weekday Activity score (15.0%). The regmision coefficients for the Weekday and

Weighted Activity scores indicated that for a one-howlday increase in Total Activity, FEVl

would increase by 3.1% predicted (p = 0.008) and 3.5% predicted (pc0.05) respectively.

43.2.4 Multiple Variable Moàeis

A second and third variable wu, added to each of the models containing only one

explanatory variable, and the effects on the adjusted R-square. the regnssion coefficients and the

standard emrs were evaluated,

As a single prodictor of FEVi, age was not significant. Whm entaed into a mode1 with

the Activity score or with percent of iâed weight, however, age became a significant predictor of

FEVIS in this two vari-able modeI, the effcçts of both explmatory variables (age + % ideal

weight, or ags + Activity score) on wcrc augmentcd For exampIe, the effect of percent of

ideal weight on PEVl was enhancecl when age differences were explaineà. While the estimates

of standard emr remained relatively stable, the parameter estimates increased. Consequently, the

significance of these variables as predictors of FEVl incteased, and confounding was indicated

Confounding is an association between two variables that is caused by a third variable (the

confounâer), which is conrlated with the f h t two va~iables?~ This enhancement effect

demonstrated the need to consider controlling for age, for exarnple stratifjing by age group, in a

similar analysis of a larger sample size.

When percent of ideal weight and the Activity score were entered into the same two

variable model, the effects of both variables were reduced which also indicated confounding,

This was not surprising kcause then was a moderate, yet significant comlation of the Activity

score with percent of iàeai weight. Therefore, these two variables were explaining sorne of the

same variability in ml. While the standard error estimates remaimd stable, the parameter

estimates and the significance of these variables also decrraseâ.

The best model incluâed al1 thne explanatory variables, age, the HAES Activity score

(Weekday Total or Weighted Total), and percent of ideal weight. With the Weighted Activity

score in the three vanable model, the adjusted R-square indicateâ that 40.0 % of the variability in

FEVl was explained by the combined effects of age, percent of i&ai weight, and the Weighted

Activity score. This was an increase of 9.0% h m the best two variable model, which included

age and percent of ideal weight. nie three variable madel that included the Weekday Activity

score demonstrated an adjusted R-square of 41.046, which was an incnase of 10.0% over the

best two variable mode1 (age and pcrccnt of i&d weight).

While Table 4.17 iUustrPt#, the totai vuiability explaineci by the best multiple variable

model @ment of ided weight, agc, and Weekday Totai Activity scon), the contribution of cach

predictor variable to the total variability explained was examineci by means of partial comlation.

In this diree variable model, percent of ideal weight, age and Weekday Total Activity

respectively explained 20.096, 11.0% and 10.0% of the variability in FEVI.

Although gender did not contribute to any muitivariable models. the possibility of gender

intexaction was considmd Age and gendet were combined wïth each of the HAES Activity

scores to form interaction terms. These terms were entered into different rnodels with the other

explanatory variables. including gender. The interaction tem. however, were not statisticdy

significant, and did not effect change to the adjustecl R-square value. A larger sample size may

be necessary to detect a significant interaction between these variables.

4.3.2.5 S u a m t q of Finai Multiple Regmaion

The final mode1 was chosen b d on the principles of parsimony and common sense.

The thne variable mode1 explained enough additional variability in m l , as compared with the

best two variable model. to justify inclusion of a t h i d variable. the HAES Activity score. The

relationship between FEVI and each of the thne explanatory variables dso makes biologicai

sense, as h a ôeen previously discussed.

These results h m multiple regnssion indicated a negative relationship between age and

FEVl in this study s~mple, and a positive relationship for both percent of idcal weight and the

HAES Activity scores with FEVi. The choice of the HAES Activity score (Weekday Total or

Weighted Total) did not affect the amount of variability explained in FEVl by the three variable

model. The thrce explanatory variables wen statistically significant predictors of FEVt in the

final model. These muits indicated thaî eitha of these two HAES Acaivity scores should be

considaed for inclusion in a modcl to explain vaziabiiity in FEVi in CF patients.

Table 4.17: Modeling PEVl a, the outcome variable

Mode1 Au. hterœpt hameter Stsadud p-value R2 Estirnate Error

One Varidle:

AGE 0.04

96 IDEAL WT 0.20

wwmAEs O* 15

W'ITHAES 0-17

Two Variables,.

AGE + % IDEAL WT 0.3 1 Age % Ideal Wt

96 IDEAL WT + WDTaAES 0.24 96 Ideal Wt WDTHAES

%IDMLWT+WTTHAES 0.26 96 Ideal wt WTTHAES

AGE + WDTEiAES 0.27 Age WDTHAES

AGE + WTTHAES 0.27 Age WlTHAES

Three Variables:

AGE + 9b IDEAL WT + WDTEAES 0.41 Age % Ideal Wt WlYImEs

Am+% Ideal Wt+WTTBAES 0.40 Age % Ideal Wt wTmAEs

WDTHAES, HAES Wcekday Total Acîivity score; WTTHAES, HAES Weightcd Tocal Activity soore; wt, weight

CHAPTER 5

DISCUSSION

5.1 Interpretation of the Resuits

Results from this cross-sectional pilot study codirmed the hypothesis that a positive

relationship would be demonsrnited between FEVi. habitual physical activity scores. percent of

ideal weight, and exercise capacity. In addition, ngression analyses demonstrated that the two

best Activity scores in pndicting FEVl were the HAES Weekday Total and HAES Weighted

Total Activity scores. A combination of either of these scores with two other explanatory

variables, age and percent of ideal weight, explained approximately 40% of the variability in

FEVl, the most important variable in âescribing CF lung disease. These positive results, in

addition to the physiological and psychological benefits that CF childrni may &rive M m

ngular physical activity, confirm the need for continued nsearch to investigate the causal

nlationship between habitual physical activity and lung hction.

5.1.1 Feasibuity

A primary objective of this study was to determine the ncnritment and compliance rates

for the collection of habituai physicai activity data using the HAES and the 7-Day Activity

Diary. The non-invasive and relatively short nature of this study k l y contnbuted to the high

recruitxncnt rate (89%) demonstratecl for this CF cünic population. Separate compliance rates for

the HAES (100%) and the Activity Diary (82%) dtmonstrated that most patients were retuming

the instruments, and f011owing study protoc01 with rcgarûs to their accmte cornpkioa.

The compliance rate with the HAES was bigbct than with the Diary, WIsly as a n u i t of

the telitive case of adminisartioa for the HAES. The EUES was completcd in approximately 15

minutes by ma t patients in cünic, whaeas completion of the Diary requhd IS to 20 minutes

each day over a consecutive period of seven days. These ciifferences, demonstrateci in this pilot

sample, suggest that the Diary would best serve specifically as a study tool to colIect data on the

rypc of activity being performed, information that cannot be captureci b y the HAES. The HAES,

as has been pnviously s ~ ~ ~ e s t e d , ~ ~ c o u l d more easily be integcated into routine chic follow-up.

In addition. the HAES is sensitive to detect change in actinty patterns in populations?8 and

therefore is suited to longitudinal usage intended to chart activity pattems.

5.1.2 Habituai Physid Activity

5.1.2.1 Trends for the HAES and the AcLlvity Diary

A cornparison of Activity scores between this study sample, the healthy population and

other groups of CF patients is advantageow in evaluating whether the habitual physical activity

pattems of the study patients represented expected levels. Direct cornparisons of Activity scores

can be made between the cumnt study and those investigations that also utiiized the HAES and

the Activity Diary in their study population. For investigations that have uscd altemate self-

report methods to collect physical activity data, general trends in the data can k compared,

including male versus female and weekday vereus weekend activity patterns.

The Total Activity scores for the curent study were consistentiy lower than those

reported for other study populations. In a similady aged group of 36 CF children? patients uscd

the HAES to report their perceived average of wual weekday and weekend day activity. Males

reported 8.1 hours/day and females reportcd 7.5 houfflday. The cumnt study population ais0

fcported lower Total Activity scores thm those documenteci for a group of healthy children

(Maies, 9.1 î 0.4 hours; Fernales, 8.9 f 0.4 hours),1° implying that the CF patients of the currcnt

study may have bem Iess physically active than thcir healthy pe9s. Anotha study of habituai

physical activity in CF sccm to support this suggestion. CF patients npo~W fewer hours

engaged in vigorous physical activity c o m p ~ with a healthy control group?' The h o u

engaged in Total Activiîy, however, did not differ ktwcen groups.

The iiterature suggests that healthy maies participate more frrquently in vigonwls

activities than females.16 Although not statisticaüy significant, maies in the cumnt study

population consistently reported more t h e in High Activity than fernales, confirming the trend

suggested by the literature. In a study of heaithy youth,n moderate to vigorous activity (High

Activity) was assessed by means of the 3-day Activity Diary. Pattems in the data reveal that

males reported consistently higher values for High Activity cornparrd with femaies.

III contrast with High Activity scores, Total Activity scores wen consistently higher in

fernales compared with males for the current investigation. These results suggested that femaie

patients spent more time in mild to moderate ph ysical activities. Two other studies involving CF

patients42 and healthy individu al^^^ nspcctively repurted the opposite trend: higher Total

Activity for males versus femaies. This indicates the need for further investigation of trends in

physical activity participation for males and femaies in a larger sample of CF patients.

The pilot study nsults confimed the importance ofcollecting weekday and weekend day

activity separately. For ail Activity scores. the reports of weekend day activity were significantly

larger than reports for weekday activity. In general, the evidence is equivocai as to whether

children an typicaily more active on a weekday or a weekend The literanue does,

however. agne that diffemces l k l y eexist. Che study suggested chat children may be more

active on the weekend due to the seâentary nature of school. Use of the 3day Activity Diary in

heaithy stuây p ~ ~ u l a t i o n s ~ ~ ~ ~ has e n s d coUcction of habituai physical activity data for both

weekday and weekaid days, confîtming the need to mapure thcse activity intervals sepoiratcly.

Positive, significant correlation coefficients were dernonstrattd between weekday and

weekend Total and High Activity scores for the HAES and the Diary. These results suggested

that children exhibit consistent pattems of physical ectivity participation throughout the week

These trends may nsult h m infîuences in the home environment, such as parental

encouragement of physical activity, or may be related to extemal environmental cues, such as

accessibility of exercise-related facilities.

This study supportcd some of the trends in physical activity participation previously

documented in the literature. Recautions must be taken, however, in àrawing conclusions From a

small sample of patients. Differences in Activity scores between males and females. for example.

were not statistically signifcant. and a larger sample is needed to assess the tnic effects of

gender on habituai physical activity participation. The higher values for Total Activity in females

versus males have not been nported elsewhere. and may reflect a nonnpresentative sample of

study patients. Altematively, these differences could be attributed to unique activity behaviors in

this CF population. Research is still needed to m e r understand and profile the activity patterns

of male and female CF patients in a larger sample.

5.132 Cornparison of the HAES and the Activity Diary

A correlation coefficient was cdculated between the HAES and the Activity Diary in this

pilot study to evaluate the constnict validity of the instniments. Constnict validity can be

confirmeci if a hypothesizcü association ktween the survey measure and a measun of the same

concept actualiy exists? The stmigth of the positive correlation between the HAES and the

Diary was only moderate. In Ilddition. mdy patients nported signifîcantfy more t h e for the

High and Total Activity scores as reportcd with the HAES comparai with the Diary. This

suggested some Merences ktwccn the two instruments.

Other snidies have investigateâ the association between M i n t seKreport measuns of

activity. A comparison study of four habituai physical activity in fernales

âemonstrated agreement between these instruments ranging h m r = 0.17 to d . 4 0 . A second

study demonstrated comlation coefficients behvan survey instruments ranging h m d . 2 2 to

d.39. These comlations are similar to the range of moderate conelation coefficients

documentecl for the comparison of the HAES and the Diary (range h m r = 031 for Weekend

High Activity to r = 0.49 for Weighted High Activity).

The results h m correlation analysis between the Activity Diary and the HAES are

similar to those of other studies. However, the moderate smngth of the correlations suggests that

there may have been na1 differences between the report of sctivity using the Activity Diary and

the mal1 of activity for HAES completion. Recall limitations, ramiom error, or ümitations due to

proxy nsponse may have affected the vaiidity and reliability of the results? Further

investigation in a larger sample is warranted to determine whether real differences exist between

the Activity Diary and the HAES.

5.1.3 Exerciae Capacity

A positive. significant comlation was dmionstrated for the HAES Activity scores with

maximal work capacity and maillmel oxygen consumption. Two other studies have demonstrated

similar resuits in healthy ~hildrcn.%~ although the HAES was not used to quantify habitua1

physical activity. The rclationship betwem physical actinty Ievel and e x d s e capacity is still

equivocal. In a 3-year exercise intervention trial in CF patients? a minimum of thne weekly

m b i c sessions was enough to e f f i t change in pulmonary hction. however, insuffident to

effect any change in wercise capacity. A training efféct may not have been demonstrateci

because of the patients rclatively high fitncss levels at baseline. Establishing thresholds of

physical activity participation for a training effect ha9 many challenges in CF: the start and stop

nature of young children's physical activity patterns. the demanding daily treatment routines of a

CF family, and the potential apathy toward exercising in the adolescent group. These factors

must be considered if fiinirt activity intewtntion studies are to be conducteci.

In m e r investigating the association between physical activity and exercise capacity,

several factors muet be considemi, including the methods useà to evaluate habituai physical

activity patterns and exercise capacity status, and the contribution of gemtic endowment to both

physical activity and physical fitness. In CF, the longitudinal follow-up of exercise capacity and

habitua1 physical activity in a large number of patients will allow for a heightened understanding

of this nlationship.

5.1.4 Cornpliance with Physiotherspr

The compliance with physiotherapy scores did not contribute as pndictors of FEVl in the

current study. Although compliance with physiotherapy was the third best individual predictor of

the Activity Diary scons. it did not contribute to the explanation of variability in multiple

correlation for FEVl. In the CF population, the accuracy of compliance with physiotherapy

scores, and their actuai condation with lung hction status are challenging issues. Patients who

comply with physiotherapy are iikely heaithier than those who do not. Ch the other han& some

patients who fa1 healthy do not comply with their ptescribed physiotherapy because they

personally dam it to k unncccssary. This extreme variaôility in rtual compüance with

physiouiaapy may have rended it unimpomt in this cro~s~sectionai regression analysis.

5.1s Psrwt Q u e s t i o ~

Analysis of data h m the parent questionnaire nvealed sevaal potcntiaily important

relatiowhips for physicat activity participation with acccssibüity of facilitics, partllt's physicai

activity level and organized phyaical activity involvement. The mean score for neighborhood

safety suggested that the majority of parents felt they lived in extmnely spfe neighborhoods. As

then was minimal variability in the scores. thae wen ais0 no important comlatioas with the

Activity scores. In a larger study sample, there may have ken more variability in ihis score. and

therefore a larger effect.

These results suggested associations common to other studies. Childnn of active and less

active parents exhibited similar physical activity patterns as their parents in one such study?' An

examination of the relationship between a mile walwrun test of physical fitness and cornlates of

activity participation demonstrated that involvement in organized physical activity was

correlated with a fitness test?' Physical activity level. however. was not assessed in this

investigation.

Cornmon relationships seen between this study and the literature suggested that physical

activity participation in CF patients rnay be affected by factors common to the healthy

population. The significant relationships for the Activity Diary and the HAES with previously

documented correlates of physical activity participaiion implied M e r relevance for the use of

these two instruments in diis population.

5.1.6 Activity Cladicatioa

The classification system developcd for the activity types listeci by this CF study

population couid eventually allow for cornparisons m s s populations, and could help in the

detennination of which activities may k usehil for effecting change in health siatus in

intavation trials. Tm& in the classilïcation of activity type accordhg to category intensities

h m the Activity Diary suggcsced that the study population was assigning the suitable intensity

to theif activity types. Common sensc suggeated thaî persona1 care activitits, f a example, would

not be listed in caiegorics higher than intensity 4 or 5, the two categories of most miId intensity.

This was the case. Similarly, then were nlatively few activities in the sport classification listed

at lower intaisities, or for the most physically demanding category, indicating that most childrm

reporteci sport participation at a moderate hteasity. These results were also expecteà, as it is

unkly, for example, that many childnn would be involved in cornpetitive sport. The need for

an undefined category. bowevet. suggested that children do nquirc more explicit instructions on

detailing their activity so that they will be more specific in their activity descriptions.

5.1.7 Simple Comiation of Activity Score with FEVi

Trends in the data suggested a stronger association for FEVi with the HAES Activity

scores than with the Activity Diary Activity scores. Som of the potential diffennces between

the instruments that could have contributecl to these discrepmcies have a h d y been suggested.

Then was also stronger comlation between FEVl and the Weekday Activity scores than with

the Weekend Activity scores. These results may have been attributed to more variability in

weekend activity compand with weekday activity. Patterns of habitua1 physicai activity rnay be

more consistent during the school week as school routines are such that most children have an

opportunity to participate in at least some physicai activity. On the weekend, children may range

h m totally inactive to highly active.

Although these results may insinuate that a record of weekend activity is of littie value, it

is stiil important to measun physical activity for both wakdays and weekend âays. In a larger

sample, weeltaid activity may be indicated as important, especialiy if some chilârea perform the

majority of their physicai actinty on Saatrday and Sunday. This amount of physical activity may

be enough to demonstrate a positive und important nlationship with lung function.

For the cumnt study, a positive and significant correlation was demonsaated for the

HAES Activity score with lung hction and nutritional status, suggesting that those with

decreased lung function or inadequate kidyweight are less active. Similarly, in the healthy

population? a significant negative relationship ha9 ken demonstrated between physical activity

level and fatness. This association suggests that childnn who sn nutritionaily sound, neither too

fat nor too thin, are more k l y to be engaged in cegular physical activity. For the CF population

where many patients are nutritionaily compromised, these results suggest the importance of

longitudinal follow-up of this nlationship to determine whether changes in habitual physical

activity are related to change in nutritional status. Oniy pmperly designed intervention saidies

will be able to evaluate whether ppcticipation in habituel physical activity can effect change in

nutritional status.

Only two other studies have examinai the specific nlationship of lung huiction with

habituai physical activity in CF childnn and adolescents. The results from these two studies

diffend h m those of the cumnt investigation. In 36 CF patients:2 a significant relationship

was demonstrated between the Total Activity score of the HAES and nutritionai status in

children with clinicall y significant lung disease. However, thm was no relations hip between

activity level and lung function in these patients. In a study by Nixon et al:' activity level was

unreleted to both exercise tolerance and pulrnonary function in 30 CF patients.

Results of these two studies likely diffed from those of the cumnt study due to the use

of differcnt masurement tools, the cross-sectionai nature of the study designs, varying disease

statw, small numbers of patients, seasonal ciifferences in the periods of activity chta collection,

and differtnt activity intervals measured (wœkdPy versus wœkend). Rather than suggesting a

lack of consistent association betwan important sady vasiables in CF, these dimpant rwuits

further clarify the need for a longitudinal stuây in the CF population with repeated measunment

and follow-up of habitua1 physical activity, lung function, nutritional status, and exercise

capaci ty.

The results h m simple comlation suggested that luag functioa and the HAES Activity

score were nlated in a positive direction. Mthough more research is needed to determine the

direction of causality for these hvo variables, they may act upon one another within a positive

fcedback loop. For enample, relatively stable lung huictioning may promote participation in

ngular physical activity. In turn, physical activity participation may act to M e r maintain or

even improve lung function. Establishing causalit y, however, will nquire long-term follow-up of

CF patients into adulthood, and studies of specific activity interventions to alter physical activity

patterns.

5.18 Regretmion Anaïyses

These results h m multipIe regnssioa indicated a negative relationship between age and

FEVi in this study sample, and a positive nlationship for both percent of ideal weight and the

HAES Activity scores with FEVi. The choice of the HAES Activity score (Weekday Total or

Weighted Total) did not affect the amount of variability cxplained in FEVi by the three variable

model. These reaults indicated that either of these two HAES Activity scores could be considered

for inclusion in a model to explain variability in FEVl in CF patients.

5.2 Methodolonical Issues

5.2.1 swdy Limitriao~ls

This study indicated the importance of continuing to evaluate the nlationship between

habituai physical activity and lung fiinction in CF, such that the direction of cawality betwan

these two variables can be established Rior to fiirthcr investigation, however, it is important to

identiw study limitations, and to provide raommtndation~ for friture resexucb in this area. The

following sections outline some of the challenges and recommendations i n c d in conducting

this pilot study.

5.2.1.1 CompUPirc with, anà Return of, the 7-Day Acüvity D m

Experience h m this pilot study indicated that q p l a r phone contact during the week of

data collection was necessary to ensure pmpet completion and retwn of the Activity Diary. If the

Activity Diary was not retumed by mail, patients were asked to bring in their completed Activity

Diary to their next clinic visit. Clear instructions, detailed examples of a sarnple day, and

encouragement to tül out the Activity Di*ary as close as possible to the tirne of the chic visit

were also important.

Although the relatively high cornpliance rate with the Activity Diary suggested that a 7-

day Diary was a feasible methoci of collecting habituai physical activity data, a 3-day Activity

Diary may ensure even higher compüance. A 3-day ~ i a r ~ ' ~ would include s d f - ~ ~ p o ~ of two

weekdays and one weekend day, and could k especially useful as a study tool in longitudinal

followiip of W patients where more than one completion of the Activity Diary is necessary.

5.2.1.2 Ranàom and Systematic Emr in Sdl-Report of Activity

In any self-report of physical activity, thae is a risk of ranQm or systematic (bias)

emr? Although the assessrnent of habitual physical activity h m the HAES and the Activity

Diary was subject to this risk, specific design feanaes w a e intended to minimize the potential

for these types of mrs.

The potmtiai for ranâom emr dunng self-administration of the sumy instniments was

addressed in this sady. For the Activity Diary, thae was a risk for unintentional emrs during

the recording of activity inteasity. h part to minimizt this risk, subjects wae sslrcd to

specificaiiy describe the activity type next to their entry for activity inteasity. This enabled the

study coordinator to check one entry against the other to ensun that the intensity matched the

type. For the HAES, potential emrs in ncording were immdiately assessed upon the patient's

completion of it. If the assigned percentages for each time p e n d as provided by the patient, did

not add up to 10096, the study coordinator documented this discrepancy, and consuited with the

patient to nctify the inconsistency. If patients perceived dificulties in comctly assigning

percentage values to each activity level, the stuây coordinator completed a sampie entry with

hem, or invoked the assistance of the puent.

Both the Activity Diary and the HAES nquind self-report of activity for al1 levels of

intensity. This was intended to =duce a patient's tendency to over-estimate physical activity

participation or to engage in non-repnsentative activity patterns during the week of recording.

When patients were first introduced to the study pmtocol in clinic, the study coordinator

emphasized the importance of recording USMI activity patterns. Examples of many Merent

types and intensities of daily activities wen pfovided to patients, mging h m sleep and

television viewing to involvement in competitive sport* These guidelines were foiiowed with

each newly recmited patient io demase the chance of obtaining a biased estimate of habinial

physical activity.

5.2.1.3 Generriupbiiity of the Results

Due to the short-tem nature of this study, habituai physical activity was coliected for

only the s p ~ g season (March-June). Research has suggested that there is variation in the

physical activity participation according to the time of yeu. Whüe one study documented a

marked inmase in activity during the s u m m r months,I6 awtber study suggestd a ckciine in

activity kvds dwing the sumrn~r.~~ Therefort, as sctivity patterns poteatiaily vrry by season, the

nsults fiom thie pilot study cannot be generaüzed to otha seasonal periods duriag the year. In

future long-term stuàies. the collection of habituai physical activity data shodd be appmpriately

timd to capture physicai activity data during al1 four semons so that a seasonai effect can be

eshateci.

5.2.1.4 Direction of Causaüty

The mss-sectional design and smail sample size of this study limi ted an y efforts to

evaiuate the direction of causality between habitual physical activity and pulmonary function.

Instead, the presence of a positive, statistical association between Activity scores &rived h m

the HAES and FEVl suggested the need for longitudinal follow-up of habitual physical activity

patterns in CF patients. Routine assessrnent of habitual physical activity, in addition to the

collection of pulmonary huiction. nutritionai status and exercise data already available through

reguiar clinical foiiow-up, would be instrumental in beginning to addrrss the issue of causality.

5.2.2 Suggested Modifications of the Cumnt Rotocol

Results h m the cumnt study an equivocal regardng the use of the 7-Day Activity

Diary in quantifying habituai physical in CF. The Activity Diary did not demonsmte a

significant or smng relationship with the primsry outcome variable. FEVI, and this finding

warrants m e r investigation. The Activity Diary and the HAES may have been capturing

different dimensions of habituai physical activity in the CF patients. Altemately, the iim

consuming nature of accurate completion of the 7&y Activity Diary may mder it impractical

for longitudinal usage. Future reseacch rnay indicate the importance of quantifjing habituai

physical activity by mwis of a single instniment that woulà fimction to collect infomraton on

ail dimensions of activity at one time period This type of instnmient muid potentially combine

featuns fiom both the HAES and ihe Activity Disry.

A s w e y instrument of this design rnay entaiI ha* patients complete the HAES in

clinic, while comspondingly listing the usual activity types. In addition to providing data on

activity type, patients would also be accountable for the percentage estimatar that they ncorded

This option, however. needs M e r consideration with regards to its feasibiiity and accmy.

5.3 Future Directions

53.1 Proposai lor a bngitudinai Study

Survival regression studies in CF have demonstratecl how pulmonary function and

nutritional status affect sumival. It has been pnviously established that FEVi is the best

predictor of s~rvival?~ and consequently, this variable is widely accepted as a surrogate for

mortality risk in CF. Research has examinecl faftors that &termine the severity of lung disease in

order to clarify the survival model: which is essentid for the timing of procedures such as lung

transplant. In addition to FEVi and nuhitional statua, other important predictors of sumival are

age, gender and specific gene mutations.

A masure of habitua1 physical activity may further explain some of the variance in

Wi, and thus enhance the CF survival model. Habituai physical activity, however, has neither

been followed longitudinaliy in the CF population, nor considend for inclusion in the survival

model. Rior to the inception of a longitudinal study, a quantifiable measun of habituai physical

activity that could be incorporated into the clinical follow-up of patients was neeâeà. The cumnt

pilot study was designeci to invcstigate this issue.

This pilot study examincd the fcasibility and cornpliance rates for the collection of

quantifiable data on habituai physical activity. This shidy also explod the â h t i o n and strength

of the relationship betwem Mitual physicai activity and =VI, while taking into collsideration

.otha variables that may influence thïs nlationship. Resdts implied a positive nlationship

between FEVl and habituai physicd activity. On the b&s of these encoumghg nsults, a

longitudinal study has been proposed, and the project protoc01 submitted for fimding in October

1999 to the Canadian Cystic Fibmsis Foundation.

The purpose of the pmposeâ pmject is to estubiish ifhabiturrl pkysical uctivity is a

significmt predictor of the change in m. Patients will be recniited from the CF clinics at two

study centers, HSC and Montreal Children's Hospital, between which there is a long history of

collaboration. The inclusion of two study centers will inmase the sample size and ailow for

smnger conclusions to be Qawn h m the &ta. Data collected over a two-year W o d h m both

the HAES and a 3day Activity Diary wili pennit the creation of a comprehensive profile of

habitua1 physical activity in a large study population. The 3day Diary, a modification of the 7-

day Activity Diary used in the pilot study, has been proposeci to encourage cornpliance over the

two-year study pend. Exercise capacity, nueitional status, and puhonary function data will

also be mutinely collected h m clinic records in keeping with the protocol of the pilot study.

However, this longitudinal study will aliow for more complete collection of data on exercise

capacity, and pmnit the inclusion of this data in regnssion analyses. A power calculation,

conducted for the grant application, has b a n included in Appmdtr G.

The short-term goals of this pmposed project include the integration of the HAES into

regular clinic proceâures for aU CF patients, and inclusion of this measun into the Toronto CF

database. F m the combined mdts of the 3-day Activity Diary and the HAES. a

compnhmsive CF activity profile wili be crcakâ, which wi l l enable bctter understandi11g of

activity behaviors in the C F population. A quantification of activity level wil i aüow us to

evaluate the rclaticmship of pulmonary function with habituai physicd activity, nutritional status

mi exefcisc capacity in a large study population. Capturing activity on a rcguiar bads win d o w

for the assessrnent of the effecis of seasonal variability on activity patterns, and subscquently on

lung huiction.

There are no hown longitudinal sadies that have examincd and monitored habitual

activity patterns and the relationship to diseass progression in CF. This pmposed project will

encourage the foîiow-up of CF activity pattems into aduithood, such that the causal relationship

between activity level and disease progression cm be eddressed. As demonstrateci by the pilot

study, the activity measures employed in Uiis investigation are well suited for clhic use, and the

HAES may allow for longitudinal tracking of activity patterns. Long-term goals include the

development of recommendations for exetcise prescription, and future coilaboration with the

Clinical Studies Netwotk (CSN) to &velop a plan for longitudinal, multi-center intervention

studies throughout Canadian CF clinics. An understanding of the inter-nlationship of pulmonary

function with habitual physical activity level, nutritional status, and exercise capacity, will allow

for improvements to our mode1 of survival in CE

5.4 Conclusions

This pilot study met its primary objectives. The results suggested a positive association of

weekday activity level (HAES) with both nutritional statw and pulmonary function, and

supportecl the hypothesis that CF patients would be receptive to providing data on theu activity

patterns. While this study has begun to demonstrate the practicaiity of using the HAES and the

Activity Diary in CF patients to quantify habituai physicai activity, preliminuy cross-sectional

trends indiate the need for longitudinal followiip.

Exercise training pmgrams are difncult to foster in the CF population wherc &mancihg

trament regimens can consume much of the CF patient's spare time. The promotion of

increcwd habituai physicai activity, howevcr, is hypothaized to effect change in the clinid

status of these patients, and is iikely casier to foster than cornpliance with an exercise training

program. The cuennt study was intendeci to gather supportive evidence necessary to warrant a

longitudinal study of the nlationship berneen FEVl and habitual physical activity. A

longitudinal study, bas& on the nsults of this pilot study, may eventually establish whether

habituai physical activity can help to explain variability in the suniival of CF patients. If the

answer is yes, the importance of encouraging regular ph ysical activi ty participation in CF

patients wili be confirmed.

REFERENCE LIST

Thompson K. Cystic fibrosis-update on exercise. Phys Sports Med 1990; 18:103-106.

Keens TG, Krastius IRB, WannamaLa EM, Levison H, Crozier DN, Bryan C. Ventilatory muscle endurance training in normal subjects and patients with cystic fibrosis. Am Rev Respir Dis 1977; 116:853-860.

Schneideman-Waiker J, Pollock SL, Corey ML, Wiikes DD, Cmny GJ, Pedder L, et al. A mdomized cotrolied trial of a thne year home exercise p m g m in cystic fibrosis. In Press-Jn of Pediatr, 1999.

Jackson ADM, Dodge JA, Cuthbert AW, Van Heyningen V, Porteous DJ. Cystic Fibrosis. Oxford: Oxford University Press, 1989.

Meams M. Cystic Fibrosis: the First 50 Years-A Review of the CLinical Roblems and their Management. In: Dodge JA. Bmck DJH, Widdicombe IH, editors. Cystic Fibrosis-Cumnt Topics. Chichester: John Wiley & Sons, 1993:217-250.

Corey M, Farewell V. Detenninants of mortality h m cystic fibrosis in Canada, 1970- 1989. Am J Epiâemiol 19%; 143:10(n-1017.

Cystic Fibrosis Foundation. Patient registry annuai report 1996.1997; Bethesda, MD.

Kerem B, Rommens J, Buchanan J, Markiewicz D, Cox T, Chakravarti A, et al. Identification of the cystic fibrosis gene: genetic analysis. Science 1989; 245:1073- 1080.

Kerem E, Reisman JJ, Corey ML, Canny GJ, Levison H. Rediction of mortality in patients with cystic fibrosis. N Engl J Med 1992; 326: 1 187-1 191.

Nixon PA, Onnstcin DM, Kelsey SF, Doerschuk CF. The prognostic value of exercise testing in patient with cystic fibrosis. N Engl J Med 1992; 327: 1785-1788.

ûrenstein DM, Henke KG, Cemy FJ. Exmise and cystic fibrosis. The Physician and Sportsrnedicine 1983; 9 (2 Suppl):6-18.

Cystic Fibrosis Team: The Hospital for Sick Childnn. Cystic fibmsis: a teaching manual for parents and patients. 1992; 3rd ed Toronto, ON.

Lai H-C, Cony M, FtzSimmons S, Kosorok MR. Fameil P. Cornparison of growth status of patients with cystic fibmsis betwen the United States and Canada, Am J Clin Nutr 1999; 69:S 1-538,

Sallis J, Patrick K Physical activity guidelines for adolescents: consensus statemtnt. Pecl Ex Sci 1994; 6:3O%3 14,

Sallis J. Epiûemiology of physical ectivity and fitness in childrcn and adolescents. Crit Rev Food Sci Nutr 1993; 33:40348.

Pate RR, Long BJ, Heath O. Descriptive epidemiology of physical activity in adolescents. Ped Ex Sci 1994; 6:434-447.

hutzenhiser JK, Clark R. Physical activity and exercise in children with cystic fihsis. Jomal of Pedatric Nming 1993; 8: 1 12-1 19.

Singsen B. Physical fitness in children with juvenilie cheumatoid arthntis and other chronic pediatnc illnesses. Pediatric Rheumatology 1995; 42:1035-1050.

Boas SR. Exercise ncommendations for individuals with cystic fibmsis. Sports Medicine 1997; 24: 17-37.

Coates AL. Exercise in CF: Does in help? Does it tell us anything? Pediatr Puhonol 1987; 3 Suppl:7 1-72.

Orenstein DM. Exercise testing in cystic fibmsis. Pediatr Pulmonol 1998; 25:223-225.

Pegues DA, Carson LA, Tablan OC, FitzSimrnons S. Roman SB, Miller JM, et al. Acquisition of Pseudomonas cepacia at summer camps for patients with cystic fibrosis. Summer Camp Study Group. J Pediatr 1994; 124:694-702.

Zach M, Oberwaldner B. Hausler F. Cystic fibrosis: physical exercise versus chest physiotherapy. Arch Dis Child 1982; 57:587-589.

Zach MS, Purnr B, Oberwaldner B. Effect of swimming on forced expiration and sputum clearance in cystic fibrosis. Lancet 198 1; 2: 1201- 1203.

Heijerman HG, Bakker W, Sterk PJ, Diflanan H. Oxygen-assisted exercise training in adult cystic fibrosis patients with pulmonary limitation to exercise. Int J Rehab Res 1991; 14: 101-1 15.

Cemy FI. Relative effects of branchial drainage and exercise for in-hospital c a n of patients with cystic fibrosis. PhysicaI Therapy 1989; 69:633639.

Edlund LD, French RW, Herbst JJ, Ruttenbag HD, Ruhling RO, Adams TD. Eff- of a swimming pgcam on children with cystic fibrosis. Sports Medicine 1986; 140:80- 83.

Orenstcin DM. Exercise conditioning and cudiopulmonary fitness in cystic fibmsis: the effects of a 3 month supervisai e d n g program. Chest 1981; 80392.

Stanghelle TI(, Hjeltncs N, Bangmi H. Michalsen EL Effécts of daily short bouts of trampoline wdse àuring 8 w& on the puimonary fiinction and the mPximal oxygen uptake of cbildiai with cystic fibmsis, ht J Sparts hkd 1988; 9 Suppl:32-36.

30. Kaens TG. Ventilatory muscle endurance training in normal subjects and patients with cystic fibmsis. Am Rev Respir Dis 1977; 1 l6:853.

3 1. Braggion C, Coniacchia M, Miano A, Schena F, Verlato G, Masteiia GD Exercise tolerance and effects of training in young patients with cystic fibrosis and mild aiiway obstruction. Pediatr Pulmonol1989; 7: 145-152.

32. Stanghelle JK, Winnem M, Roaldsen K. de Wit S. Notgewitch JH, Nilsen BR. Young patients with cystic fibrosis: attituâe toward phyeicai activity and influence on physical fitness and spimmetric values of a 2-week training course. Int J Spom Med 1988; 9 Suppl:25-3 1.

33. & Jong W, Grevink R, Roda RJ, Kaptein A, van der Schans CP. Enect of a home exercise training program in patients with cystic nbroais. Chest 1994; 105:463-468.

34. O'Neill PA, Dodd MA, Phillips B. Poole J, Webb AK. Regular exercise and reduction of breathiessness in patients with cystic fibmsis. Br J Dis Chest 1987; 8162-69.

35. Ancireasson B. Jonson B. Komfalt R, Nordmark E, Sanâstrom S. Long-term effects of physical exercise on working capacity and pulmonary function in cystic fibrosis. Acta Paediatr Scand 1987; 76:70-75.

36. Blomquist M. Freyschuss U, Wiman CO, Strandvik B. Physical activity and self treatment in cystic fibrosis. Arch Dis Child 1986; 61:362-367.

37. Hoizer FJ, Schnall R, Landau U The effect of a home exercise porgramme in childscn with cystic fibrosis aud asthma. Aust Psediah J 1984; 20:297-302.

38. Salh W. Bilton D, Dodd M, Webb AK. Enect of exercise and physiothecapy in aiding spuaim expectoration in ad& with cystic fibrosis. Thorax 1989; 44:1006-1008.

39. Heijerman HG, B a k r W, Sterk PJ, Dijkman Et lang-term effccts of exercise training and hyperaümentation in adult cystic fibrosis patients with severe pulmonary dysfunction. Int J Rehab Res 1992; 15:252-257.

40. KoIIberg H. Cystic fibmsis and physical activiw an introduction. Int J Sports Med 1988; 9:2-5.

41. Nixon PA, bnstein DM, Keisey SF. Habitua1 physical activity in childrcn and ado1cscents with cystic fibrosis. Peû Ex Sci 1999; 11: 18 1 (Abstract).

42. ouc cher GP, Lands LC, Hay JA, Homby L. Activity Ievels and the relationship to lung function and nutritional status in childrcn wîth cystic fibrosis. Am J Phys Med Rehabil 1997; 7633 1 1-3 15.

43. Marcotte JE, Gfisble R Lenson H, Coaes AL, Canny GJ* Multiple factors M t txcrcise capacity in cystic fi'brosis. Pediatr PiiImonoI 1986; 23274-28 1.

44. Lands LC, Heigenhauser GIP, Jones NL. Analysis of factors limiting maximal exeinse performance in cystic ûbmsis. Clinicd Science 1992; 83:391-397.

45. Marcotte JE, Canny GJ, Grisdale R, DuMond KJ, Corey ML, Zinman R. et al. Effects of nutritional status on exacise performance in advanced cystic fibrosis. Chest 1986; 90:375-379.

46. Hay J. Development of a 24 hour ncall measure. Canadian Association of Sport Sciences Annual Confertnce 199 1.

47. Montoye HJ, Kemper HCG, Saris WHM, Washburn RA. Measuring ph ysical activity and energy expenditure. 1st cd Windsor, ON: Human Kinetics, 19%.

48. Freedson P, Melanson E. Measuring physicd activity. In: Docherty D, editor. Measment in Pediaüic Exercise Science. 1st ed. Victoria, B.C.: Hman Kinetics, 1996:261-283.

49. Bouchard Cl Tremblay A, Leblanc C. Lortie G, Savard R, Thenault G. A method to assess energy expenditun in chilnn and adults. Am J Clin Nutr 1983; 37:461-467.

50. Bratteby L-E. Sandhagen B. Lotborn M, Samuelson O. Daily energy expenditure and physical activity assessed by an activity diary in 374 randomly selectd 15-year-old adolescents. Eur J Clin Nutr 1997; 51:592-600.

51. Bratteby L-E, Sandhagen B. Fan Hs Samuelson O. A 7day activity diary for the assessrnent of daily energy expenditure validateci by the doubly labeiled water methocl in adolescents. Eur I Clin Nutr 1997; 5 1:58S-S9l.

52. Hendemn CJ, Love11 DJ, Specker BL. Campaigne BN. Physical activity in children with juvenille rheumatoid arthritis: quantification and evduation. Arthritis Care and Research 1995; 8: 114-1 19.

53. Eisenmann JC, Katzmarzyk PT, Theriault G, Song TMK, Malina RM, Bouchard C. Physical activity and pulmonary huiction in youth: the Quebec family study. Ped Ex Sci 1999; 11:208-217.

54. Weston AT, Petosa R, Patc RR. Validation of an instrument for rneasurement of physical activity in youth. Med Sci Spom Exerc 1997; 29:138-143.

55. Crocker PRE, Baiky DA. Faulkner RA. Kowallcsi KC, McOrath R. Measuring generai levels of physical activity: pnliminary evidenct for the Physical Activity Qucsti0mai.e for Oldcr Children. Mcd Sci Sports Exerc 1997; 29: 1344-1349.

56. Oodin 0. Shephard RI. A simpIe mcthod to assess exercisc behavior in the community. Can J Appl Sport Sci 198% 10141-146.

Sallis JI& Buono MJ, Roby JI, Micale FG, Nelson JA. Seveaday recall and other physical activity self-teports in childm and adolescents. Msd Sci Sports Exerc 1993; 2S:gg- 108.

Hay JA. Development and testing of the habituai aclinty estimation scde. Proceedings of the XIXth International Symposium of the European Orwp of Pediatnc Work Physiology 1997; 2: 125-129.

Halton JM, Atkinson SA, Hay J, Webber CE. Mntin RF, Barr RD. Ostopenia in children with acute lyrnphoblastic leukernia &hop Tms 1995; 18:684 (Abstract).

Hay J, Shephard RI. Perceptions and patterns of physicd activity: a cornparison of MohawkKayuga and non-native adolescents. American Journal of Human Biology 1998; 10:629-635.

Atkinson SA, Halton J, Hay J, Gordon C, Webber C. Nonnative values for lumbar spine bone mass in children in relation to age, gender, and physical activity. Jn Bone Min Res 1991; 6:S132 (Abstract).

Moelgaard C. Bone mineraiization during growth; the influence of growth, puberty, and calcium intake. 19%; University of Copmhagen. Denmark.

DiLorenzo TM, Stucky-Ropp RC, Vander Wal JS, Ootham W. Detenninants of exescise among chiidren. II. a longitudinal anaysis. Reventive Medicine 1998; 27:470-477.

Sallis JK, Simons-Morton BG, Stone EJ, Corbin CB, Epstein LH, Faucette N, et al. Detemiinants of physical activity and interventions in youdi. Med Sci Sports Exerc 1992; 24:S248-S257.

Sallis JK, Johnson MF, Calfas KJ, Caparosa S, Nichols JF. Assessing perceived physical environmental variables that may influence physical activity. Reseaich Quarterly for Exercise and Sport 1997; 68:345-351.

Pate RR, Tmst SG, Felton GM, Ward DS, Dowda M, Saunders Re Cornlates of physical activity behavior in nual youth. Rescarch QuarterIy for Exerciae and Sport 1997; 68:241-248.

Robinson TN, Hammer LD, Külen JD. Kraemtr HC, Wilson DM, Hayward C, et al. Does television viewîng increasc obesity and rcâuce physical pctivity? Cross-sectional and longitudinal analyses among aàolescmt girls. Pediatrics 1993; 9 1:273-280.

F ' n P, Evenson S. Familiai aggregation in physicd activity. Reseprch Quarterly for Exercise and Sport 1991; 62:384-389.

Dishman RK, Sallis JK, ûrcnstein DR The detaminuits of physicd activity and exCrci=. Public Hialth R w 1985; 100:158-171.

70. Anonymous Inteptation of pulmonq fiinction tests: a practical guide. Philadelphia: Lippincott-Raven, 1997.

7 1. American 'ïhoracic Society. Standardization of sphmetry - 1994 Update. Am J Respir Crit Care Med 1995; lS2:llO7-1136.

72. Corey M, Levison H, Crozier DN. Eve- to seven-year course of pulmonary function in cystic fibmsis. Am Rev Respir Dis 1976; 1 14: lO8SlO92.

73. Rosenfeld M. Davis R, FiaSinunoas S, Pepe M, Ramsey B. Gender gap in cystic fibmsis moaality. Am J Epidemiol 1997; 145794-803.

74. Tanner JM, Whitehouse RH, Takashima S. Standards h m birth to manuity for height, weight, height velacity and weight velocity. British childmi, 1%5. Arch Dis Child 1966; 41:613-635.

FitzSimmons SC. Changing epidemiology of cystic fibrosis. J Pediatr 1993; 122:l-9.

Moore BJ, Durie PR, Forstner 00, Penchan P. The assessment of nutritional status in chiidren. Nutr Res 1985; 57:9749.

Szeinberg A, England S. Mindorff C, Fraser IM, Levison H. Maximal inspiratory and expiratory pressures are nduced in hyperinfiated, malnourished young adult male cystic fibrosis patients. Am Rev Rmpù Dis 1985; 132:766-769.

American Thoracic Society. Lung function testing: Selection of reference values and interpretative strategies. Am Rev Respu Dis 1991; 144: 1202-1218.

Godfny S, Davies CM, Wozniak ER, Barnes C. Cardio-respiratory nsponse to exercise in normai children. Clin Sc Mol Bi01 1971; 40:419-442.

Reisman JJ, Rivington-Law B. Corey ML, Marcotte JE, Wannamaker E, Harcourt D, et al. Role of conventional physiothcrapy in cystic fibrosis. J Pediatr 1988; 1 13:632- 636.

Dumin M A , Rahaman MM. The assessment of the amount of fat in the humsn h m meastuement of skinfold thickness. Br J Nutr 1967; 21:681-689.

Coates AL, Boyce P, Shaw DG, Godftzy S, Meams M. The relationship b e ~ e c n the chest radiograph, regional lung act ion studies asing 13N2, exacise tolerance and clinicai condition in children with cystic naOeis. Arch Dis Child 1981; 56:106-1ll.

Passero M. Remor B, Salomon I. Patient nportcd cornpliance with cystic fibrosis therapy. Clin M m 1981; 20:264-266.

McArdle WD, Katch FI. Kaîch VL. Exercisc Physiology. 3rd cd. Philaddphi$: Lea & Ftbiger, 1991.

Pate RR, Dow& M. Rose JG. Associations between physical activity and physicai fi mess in American children. AJDC 1990; 144: 1 123- 1 129.

Cody RP, Smith JK. Applied Statistics and the SAS Programming Language. 4th ed. New Jersey: Pnntice-Hall, Inc., 1997.

Kelsey JL, Whittemore AS, Evans AS, Thompson WD. Methods in Observational Epidemiology. 2nd ed. New York: Oxford University Press, 1999.

Cony ML, Edwards L, Lenson H, Knowles MR. Longitudinal analysis of pulmonary function decline in patients with c ystic fibrosis. J Pediatr 1997; 13 1 :809-8 14.

Reisman, I. J., Schneidennan-Waiker, Je, Canny, O. J., Wilkes, D. D., Mappa, L., Levison, He, Lu, W-Y., and Cony, P. Normal responses to progressive and steady state exedse in chilâren 11 to 18 years. Unpublished Work, 19%.

Washington RL, van Gundy JC, Cohen C, Sondheimer HM, Wolfe RR. Normal aerobic and anaerobic exercise data for North American schwl-age children. J Pediatr 1988; 1 12:223-233.

Cale LA, Almond L. Children's activity levels: a review of studies conducted on Bxitish chiidren. Phys Ed Rev 1992; 15: 1 1 1-1 18.

Koo MM, Rohan TE. Cornparison of four habituai physical activity questionnaires in girls aged 7-15 yr. Med Sci Sports Exerc 1999; 3 1:421-427.

Snedecor GW, Cochrane WG. Statistical Methods. 6th ed. Iowa: The Iowa State University Press, 1967.

Selvin S. Ptactical Biostatistical Methoàs. Caiifomia: Wadsworih Publishing Company, 1995.

Glantz SA. Slinka BK. Rimer of Applied Regmision and Analysis of Variance. 1st ed. New York: McGraw-Hill, hc., 1990.

Rowlandp AV, Eston RG. Ingledcw DK Relationship ktween activity levels. aembic fitncss, and body fat in 8- to 10-yr-old chilhn. J Appl Physiol1999; 86: 1428-1435.

Saüis JF, McKenzie TL, Alcaraz JE. Habituai physical activity and healthielateà physicai fitness in fourth-grede cMdren. AJDC 1993; 147:8%.

Pate RR, Ross JG. Factors associateci with health-rtlated fitness. JOPERD 1987; Novemk-Decemk.93-95.

Dotson CO, Ross JG. Relationship betwan activity patterns and fimess. JOPERD 1985; January:86-89.

100. Hyan RE, S c d m PD. Nalumm M. hterpretation of Pulmonary Function Tests. Philadelphia. Lippincott-Raven, 1997.

APPENDIX A

A SUMMARY OF EXERCISE INTERVENTION STUDIES IN CF

- a

go: œ

Table A.2: Unsurxrvised el

Patient motivation, inmesî, clinicai stability and panllnity to hospital

kontinued) Study Duradon

Phase A: 6 rnonths

Phssc:B:6months

Basdine vermis PbriseAvasus Pbasc B

Coatrol patients, askdiom -ge u ! d sctivity .

Conclusions

APPENDIX B

CONSENT AND ASSENT FORMS

Consent Fonn

HABITUAI, PHYSICAL ACTIVrrY AND THE ASSOCIATION WiTH DISEASE SEVERXTY AND EXERCISE CAPACEY IN CYSTIC FIBROSE: A PILOT STUDY

Sue PoUock, BPE Master of Science Candidate in Epidemiology, Department of Public Health Sciences, The University of Toronto (supe~sed by Dr. Mary Corey) Research Assistant, Division of Gmtmenterology and Nutrition, The Hospital for Sick Childnn Phone: (416) 8 13-5751

Dr. Mary Corey, PhD E~iderniolos!ist and Scientist. Division of Gastroeniaology and Numon, The Hospital for Sick Childrcn Associate Professor, Departments of Paediaüics and niblic Health Sciences, The University of Toronto Phone: (416) 8 13-5750

Jane Schneidennan-Waiker, MSc Exercise Physiologist, Division of Respiratory Medicine, The Hospital for Sick Childm Phone: (4L6) 813-7654 (extension: 4407)

Dr. Ja Reisman, MD, FRCP(C), MBA irologi&Division of Respiratory Medicine,

The Hospital fot Sick Children clsonate Rof«sot, The Facuity of Medicine,

The University of Toronto Phone: (416) 8 13-6167

Furpœe of the Researcb

We are conducting this study to document the M y activity patterns of cystic fibrosis patients. One of our aims is to evaluate the rclationship of different amounts, intensities, and types of activity with puimonary function and exercise performance. Knowleâge gained h m this study may help us benet understand the unique role of other therapies, such as exercise, in the treatment of cystic fibrosis. Additionally, we plan to invcstigate household and family-related fsctors that may pmmote participation in specific activities. This will help us make recommendations towards improving cornpliance to specific therapy pmgrams.

Description of the Reseoreh

If you choose to participate, the tim we nqUe h m you, as outlined below, will be minimal.

i)

ii)

iii)

iv)

During your c h i c visit, one of the study investigaton, will be avaiiable to answer any questions you may have, and to m e r explain this nsearch project. Your participation in this study mil not interfm with your standard treatment.

If you agree to s i p the consent form, you will be givm a questionnaire to complete while at the chic. This questionnaire asks you about your usual activity, and should take approximately 15 minutes to complete. There are also a few questions for parents to answer nlated to their own activity levels and other household factors.

Next, we will provide you with an activity diary to take home. We wiU ask you to fil1 out this diary for the 7 days folîowing your clinic visit, & M n g al1 your activities (sleep includeâ!). Upon completion of your activity record, you will need to send it back to us at the hospital. We may telephone you during the study period to offer support.

For children l a s than 12 years of age, parents will need to help with both the questionnaire and the activity diary. Investigato~ wiU need access to patient health mords to obtain information on pulmonary hnction. exercise testing, and chical statu. We may rtquire furthet input, by phone, to clPrify intonnation a h the saidy.

Then are no known harms associateci with participation in this study. However, you may find completion of the 7&y activity diary to be a minor time inconvenicnce.

Although you (or your child) wi i i not ducctiy benefit h m this study, we h o p tbat the uifonnation we collcct on your activity Ievels wiIi heIp us better undenitaad the d e that otha therapics, such as exacise therspy, couid play in the management of cystic l%msis. At the end of the study, we wiU m d you a summary of your o v d activity Ievel.

Confidentiality WU be respectai, and no information that discloses the identity of the subject will be nlcased or published without consent unless required by law. For your information, the nsearch consent form will be insemi in the patient heaith mord.

Participation in nseacch is voluntsry. If you choose not to participate, you and your family will continue to have access to quaiity c m at HSC. If you choose on behalf of your child to participate in this stuây, you can withdraw your child fiom the study at any t h e . Again, you and your family wiiî continue to have access to quality can at HSC.

For participants 16 y- otage a d o k r :

"1 acknowledge that the research procedm dehbed &ove have ban explained to me and that any questions that 1 have asked have been answed to my satisfhction. 1 have ban informeci of the alternatives to participation in this study, including the right not to participate and the right to withdraw without compmising the quality of medical c m at The Hospital for Sick Children for me and for other members of my family. As well, the potential harms and discornforts have been explained to me and I also understand the benefits (if any) of participating in the research study. 1 how that 1 may ask now, or in the future, any questions I have about the study or the restarch pmceduns. 1 have ban assured that recorcîs relating to me and my care wili be kept confidential and that no intorrnation will be released or printed that would disclose personal identity without my pemission unless required by law."

1 hereby consent to participate.

The Person who may be contacteci Nam of Patient and Age about the research is:

who may be reached at telephone #k Signature (if 16 p. or over)

Name of person who obtained consent

Date

For parene of participants leai tb.n 16 y t u r of age:

'1 acknowledge that the nsearch procedures describeci above have been explained to me and that any questions that I have askcd have k e n answmd to my satisfaction. 1 have been idormeci of the aitematives to participation in this study, includhg the right not to participate and the right to withdraw without compromising the quality of mcdicai can at The Hospital for Sick Chiidren for my child and for other members of my famiiy. As well, the potential hanns and discornforts have been explained to me and I also understand the benefits (if any) of participatiag in the nsearch stuây. 1 know that 1 may ask now, or in the future, any questions 1 have about the study or the research procedures. 1 have k e n assured that records rclating to my child and my child's care wiiî be kept confidentid and that no information will be nleased or printed that would disclose personal identity without my permission unless nquind by law."

1 hemby consent for my child to participate.

The Person who may be contacted Name of Patent about the research is:

who may be reached at telephone #: Signature

Name of person who obtained consent.

Date

Signature

Titie of Sbdy

HABITVAL PHYSICAL ACï IV ï ïY AND THE ASSOCIATION WITH DISEASE SEVERlTY AND EXERCISE CAPACIT'Y IN CYSTIC FiBROSIS: A PILXlT STUDY

1) Sue Pollock, BPE -Research Assistant, The Hospital for Sick Children -Graduate Student, The University of Toronto

2) Dr. Mary C o q PhD Scientist and Epidcmiologist, The Hospital for Sick Children -Associate Professor, The University of Toronto

3) Jane Schneidennan-Waiker, MSc -Exercise Ph ysiologist, The Hospiital for Sick Children

4) Dr. Joe Reisman, MD, FRCP(C), MBA -Pediatric Respirologist, The Hospital for Sick Children -Associate Rofessor, The University of Toronto

Questions vou mav have about the stumt:

Why are ne doing tbis study?

At the cystic fibrosis chic, doctors, nurses and physiotherapists ns intmsted in your activity patterns. Remarchers at the hoapital also want to know how you spend your time, including how much TV you watch, how often you go shopping with Ken&, end how many sports you play. This study wiii aiïow us to describe your activity levels, and compare them with other childm and other CF ciinics. Wc would aiso Lik to know why you choosc the activities that you do. For example, why Q some kids s p d thcir summers at the local pool. while others prefer to play on the cornputsr? This information will help us encourage aiî CF lads to make good choices about thcir chosen activities.

what rrUi happen d* thir study?

At your clinic visit, you wil l meet one of the study investigators. She wii i explain the study to you. If you and your parents agnt to participate. we wiii give you a short questio~maire to n11 out, esking you about the amount of time that you spend each day at high and low activity levels. If you are las than 13 years old, your parents will help you to fiii out this questiomain. Parents WU also have their own questions to answer about thcmselves and your neighbo&d We will also give you an activity diary to tak home. This is a record of your daily activities over a one week t h period After 7 days of filling out the diary, you will need to smd it back to the hospitai. We wiil provide you with a stamped, addresseci envelope to send it in.

A n the= good things and baà things about the ihdy?

We don't think that there is anything bad about this study. In fact, we think it will be fin for you to describe yow activity panmis, and to find out how you spend your time each day. At the end of the study, we wiil send you a summary of your o v e d activity level.

Who rvill know about what 1 did la the study?

Any information that we collect about you during this study wiil only be shared with the other study investigatonr. If we need to i&ntiQ you. we wiii only use your hospital identification number.

Yes. If you do not want to join ihis study, or decide halfway through that you no longer want to participate, notify the hospitai or your parents. We wiil respect any decision you make about your own participation.

For participants lem than 16 years of age:

"1 was pment when read this fom and gave hislher verbal assent."

Name of person who obtained assent

Date

GLOSSARY OF TERMS and ABBREVIATIONS

Forced Ex~Vstory Fiow Rate (FEF3-7s): The average forced expiratory flow rate over the middle b

50% of the forced vital capacity test.lm

Forced Ex~iratory Volume in one second (FEVI): The volume of air exhaleci in the first second

of the forced vital capacity test.Im

Forced Vital Ca~acitv WC): A maneuver in which the subjects exhales as rapidly and

completely as possible following a maximal in~~iration. '~

Maximal heart rate (HRmax): A measun of the number of times the heart bats per minute at

maximal exercise."

Maximal minute ventilation (vEmax): The amount of air that a subject inspires or expins in one

minute at maximal e~ercise.~

Maximal oxygen saturation (Sa02 at mm): An assessrnent of pulmonary p exchange at

maximal exercise."

Peak ex~iratorv flow rate (PEF): The maximal expiratory fiow rate that occurs shortly afkr the

onset of expiration in the WC test?"

Maximal oxvgen conswn~tion ( 9 0 ~ ~ ) : The maximal rate et which oxygen can be consumed

per minute. It is considmd to be the kst physiologie indicator of an individuai's aality to

transport and utiiize oxygen, and is a masure of an individuai's aerobic capacity."

1mai work caywity (wmax): The maximal workld a subject achi~ves during exercise

testingU

Ventüatorv ceserve (vE/MVV): The ratio of maximal ventiiation to maximum voluntary

ventüation. The test involveci fm mpxunal volrmtary ventilation requires the subjoct ta bmthe

138

hard and fat for 15 seconds. These nsults are extrapolated to 60 seconds, and repoaed in

minute?'

APPENDIX D

THE HABITUAL ACTIVITY ESTIMATION SCALE (HAES)

THE HABITVAL ACTMTY ESTIMAION SCALE (HAES)

Date :

This questionnaire contains specific questions about your activities of daily living. Try to answer each question as best as you possibly can.

Please nad the al i the directions canfully before completing the questiomaire. This wili make the task easier for you.

Instructions

This fom asks you to estimate your level of activity during different time pefiods throughout an average weekday (Monday to Fhday) and weekend (Saturday or Sunday). Please mal1 the activities of one typical dry within the past several weeks. For each t h e period, please estimate the percentage of time that you spent in each of 4 different activity levels. The different activity levels are deecribeâ below. For each of the time paiods, the totai time spent in all activity levels must add up to 100 96. Following is a sample question and answer senes. When completing the fom, only give the percentage of time in each category. You do not need to write in what you were actually doing!

Sample:

From when you finished breakfiut until you started lunch, please estimate the percentage of time

that you spent in each of the following activity levels:

a) inactive (i.e.. having a nap)

b) sornewhat inactive 60aD (i.e., watching TV)

C) somcwhat active 25% (i.e., shopping)

d) active JO% (i.e., riding a bicycle)

e) totai

Thene an ais0 a series of questions asking about the timing and length of meals and sleeping

habits. Plcase answer ail of these qutstims completely.

AC- LE= DESCRIPTIONS

These descriptions give you examples of activities that are hlpical of each activity level. You

should refer back to these descriptions as often as you need when completing your estimates.

a) inactive - sleeping, lying Qwn, resting, napping

b) somewhat inactive - Sitting, reading, watching television, playing vide0 gams, time in front

of the cornputet, playing games or activities which an mostly done sitthg down

c) sonwwhat active - walking, shopping, light household chores

d) active - ninning, jumping, skipping, bicycling, skating, swirnming, games that require lots of

movernent and make you breathdsweat hard

WeeWv Acn'vin,: For an average weekday (Le., Monday, Tuesday, Wednesday, Thursday or M y ) , please

estimate the percentage of tirne that you spent in each activity level.

1) After getting out of bed until starting breakfast:

a) inactive

b) somewhat inactive

C) somewhat active

d) active

e) totai

2) After finishing breakfiut until staxting lunch:

a) inactive

b) somewhat inactive

c) somewhat active

d) active

e) total

3) Afkr finishing lunch until starting supper:

a) inactive

b) somewhat inactive

c) somewhat active

d) active

e) total

4) A f h finishing supper until kâtime:

a) inactive

b) somewhat inactive

C) somewhat active

d) active

e) totai

For the average weekday that you are refedng to, please answer the following questions as

accurately as possible in the spaces pmvided.

s) At what time did you get out of bed in the moming?

6) At what time did you start eaîing breakfast?

7) How long did you spend eating breakfast? minutes

8) At what time did you start eating lunch?

9) How long did you spend eating lunch? minutes

10) At what time did you star& eating supper?

11) How long did you spend eating supper? minutes

12) At what time did you go to bed that evening?

13) For the average weekday that this questionnaire has asked you about, would

you rate your ovedl level of activity as being: (please chle)

a) very inactive

b) inactive

C) somewhat inactive

d) somewhat active

e) active

f) veryactive

14) Compand to other ment weekâays, would you rate yowlf on this

chosen &y as being: (please circle)

a) much less active than usud

b) a Little less active than usual

c) about the same as usual

d) a littie more active

e) much more active than usual

. ActtviLv:

For an average weekend day (i.e., Saturday or Sunday), please estin.uk the percentage of time

that you spent in each activity level.

15) After getting out of bed untii staiting breakfast:

a) inactive

b) somewhat inactive

C) somewhat active

d) active

e) totai

16) Mer finishing breakfast until starting lunch:

a) inactive

b) somewhat inactive

c) somewhat active

d) active

e) total

17) After finishing lunch until starting supper:

a) inactive

b) somewhat inactive

C) somewhat active

d) active

e) totai

18) After finishing suppei until bcdtimt:

a) inactive

b) somewhat inactive

C) soxnewhat active

d) active

e) totai

For the average weekend day that you are refenhg to, please answer the foilowing questions as

accuraîely as possible in the spaces providcd

19) At what time did you get out of bed in the moming?

20) At what fime did you start eating breakfast?

21) How long did you spend eating breakfast? minutes

22) ~t what time did you start eating lunch?

23) How long did you spend eating lunch? minutes

24) At what time did you start eating supper?

25) How long did you spend eating supper? minutes

26) At what time did you go to bed that evening?

27) For the average weekend day that this questionnaire has asked you about,

wouid you rate your ovedl level of activity as being: (please circle)

a) very inactive

b) inactive

c) somewhat inactive

d) somewhat active

e) active

f ) very active

2û) Compared to other recent weekend days, wouid you rate yourseif on this

chosni day as being: (please circle)

a) much less activethan wual

b) a iittie less active than usual

c) about the saute as usual

d) a iittie more active

e) much more active than usual

29) if you have any commcnts about your activity patterns that you tbink are

important, pl- mention them in the s p m below.

APPENDME

THE 7-DAY ACTnrlTY DIARY b

APPENDIX F

PARENT QUESTIONNAIRE

PARENT OUESTIONNAIRE

Dear pamts,

In order to complete our study on pediaûic activity patterns in cystic fibmsis, we also need to ask you a few questions about your own habits, household, and your child's involvement in activities. PIease answer as accurately as possible.

Child's name

Name of person completing this questiomain

Relationship to child

Date

Place an K next to the orie section which b a t describes YOUR current Ievel of daüy physicai activity:

You have a sit-down job and no regular physical activity.

Tlme to four hours of accumulated waking or standing in your working day are usual. You have no ngular organized physical activity dhng kisure time.

You are sporadically involved in mreational activities, such as weekend golf or tennis, occasional jogging, swimming, or cycling.

Usuai job activities might include lifting or srair climbing, or you participa& ngularly in ~ationaüfitncss activities such as jogging, swimming or cycling at least three buncs per week for 30 to 60 minutes each time.

You participate in txtcnsivt physical activity for60 minutes or mon at least four days pa

2. Now we WIN ask you abut YûUR CHILDsS participation in mmmudty baseâ phydcai acüvitime Please &le either YES or NO.

At the present tim, is YOUR CBILD involved in:

a public park or recreation department activities YES

b. physical activities within a church group YES

c. sports team or league YES

d. YMCA, YWCA or similar organization YES

e. health club, or pnvate sports lessons YES

f. Brownies, Cub Scouts, or other scouting group YES

g. 4-H or other farm club YES

h. another organization that promotes physical activity YES

If "Yes", please specify type/narne organization(s)

3. For each of these places where you an execcîse, p h ladicate CIit is on a h q d y trPvcleü route, andhot easiîy accessible to you (i.e., withtn a 5-minute àrîve h m your chiid's scbool or home).

a, aerobic dance studio

b. basketball court

6 bike lane or trails

e. golf course

f. health spdgym

g. mareialartsstudio

h. playing field (soccer, football, softbail)

i. public park

j. pubüc recreation center

k. racquetbalVsquash court

m. skating rink

n, sports store

O. swimming pool

p. walkinghüring trail

q. tennis court

r dance studio

A How d e do you tcd wPlWng la your neighborhood durjbg the &y? Pleacc chle ontv ong numbet fiom 1 - 5, whae 1= very unsefe and 5 = very safe.

5. Piara indicate which of the foiiowing apply Q your ndghborhoad

sidewatks

heavy trafnc

hilis

stnet lights

dogs that are unattended

enjoyable scenery

fnquently see people walkinglexercising

high crime

Herne indlfste which items you have in your home, yarà, or spartment complex

stationary aerobic equipment

bicycle

h g trampoline

running shoes

swirnming pool

weightlifting equipment, i.e., f ke weights, Universal gym, Nautilus equipment

toning devices, i.e., ankle weights, Dynabands, Thighmaster

aerobic workout vide0 or audiotapes

step aerobics, siide aembics

skates ( d e r blades. ice skates)

sports quipment (baiis, racquets)

surf board, boogie board, windsurf board

Cam, row boat, kayak

skis (water skis, domihiil skis, crosssountry skis)

APPENDM G

POWER CALCULATION

Given 188 eligible patients and an estimated 75% nsponse rate, we expect 140 patients

to participate in a longitudinal study. An estimate (bmed on objective monitoring) of the overall

percentage of adolescents in the healthy population engaged in regular moderate to vigorous

activity is 50%,16 therefore, we pmject approximattly 70 patients in each of the active and

inactive groups. The primary outcorne variable for our study, percent predicted FEVl. was found

to have a standard deviation of 18.6 in a similerly aged sample of cystic fibrosis childmi? At a

5% significance level. this sample of 140 patients has 89% power to detect a clinically

significant difference in mean FEVi of 10% pdcted betwem high and low activity groups of

qual numbers.

za

where: q3 =unbiown d = 10% predicted (the magnitude of diffmnce we an interested

in detecting between mean FEVl percent predicted for the active vems inactive groups

a = 18.6 n = the estimated numbcr of exposecl individuals=70 r = the ratio of umxposed to exposeci individuals=l/l za12 =1.%

APPENDIX H

SUMMARY OF ACTMTY TYPES

CF treatmen t crying Getting Dressed Getting Redy for Bed Gening Ready for Schaol Getting Ready for Sports Searching for Something

Domestic Work

Cleaning Cleaning (mm) Cooking Dusting Taking out the Garbage Making Bed Mowing Lam Packing up Supplies Putthg Away Omccnes Raking Lami Shopping Shopping for Grocenes Watering PlanWGrass

Acting Classes Drama Classes cornputer work Hdping Teacher Homework

A€tcr-schwl I wakend job

Hicle and Seek Playing Playing with F d l y Singing and Clapping Recess

Going to Lunch Socializing WaMng Around Walking h g Walking Indoors Work Break

Gym Class

CATEGORY S

Drying Self Off Searching for Something

Domestic Work

Outside Chorts Cleaning B k Mowing the L a w Packing up Stuff Ralsng the Lawn Shopping S weeping the S tain Washing the Car Watmog the GrasdPlants

CIssses School Field Trip Homcwodc

Ball Playing Sidewak Chaik h * g Party Games Amusement Park Carnival Oioup Gams Hide and Seek Hopscotc h Outside Play Playing Video Gama Play gmund Recess Swinging

Socializing Walking Dog Walking Outdoors Walking to Bus Walk To/Fmrn Schwl Walking to Work Work Break

Bike Hocda Hoop Hopping Running Yoga

CATEGORY 6

CrIrying Stuff aeaning Cicaning (mm) Hou8twork Wmhing Car wateriag Plmt!3 Yard Work

Drama Classes

Emlo~ment

AfierSchoo1/FYeekend Work

Hopscotch Game: 4-square Piaying in the Gym Playground Playing Playing in the School Yard Playing with Sister Recess S winging

Baii Playing Baseball Basketball Bike Bowling Playing Catch PE Class Cool-Down, Post-Soccer Dancing Dht Biking Dodgeball Fighting Fishing Tag Golf Gym GymnPstcs Hockey Roilerblade Rdg/JopgUig SkatcboBtding SItipping soccei: S=-B=M

RunningStairs Swim T-Bal1 Traick & Field Tmpoline

After-School and Weekend Work

Ultstructured Play

Chasing Someone Playing Recess

Walking

Badminton Bal1 Playing Baseball Basketball Bike Dancing DodgebaU Football Tag Gym Class Gymnastics Hockey Hado Hoop Lactosse Running Squash Running Stairs Soccer Soccef-Baseball Swim Tat-Bo TnickdField Trampoline Warsn-Up, Pre-Soccc~

CATEGORY 8

Rayground Playing Recess

Aerobics Badminton Bail Playing Bike Dancing DodgebaU Tag Gym Class Gym Class (Capture the Hag,) Gym Class (Scooters) Hockey Lacrosse Ksrate Rollerblading Roller-skating Running So~~er-Drills SoccerScnmmage Running S tain Track and Field

CATEGORY 9

red Pla\!

Reccss

Basketbal1 Bike Dance RoUctblding Soccer Track andField