Am J Clin Nutr 1980 Acheson 1147 54

  • Upload
    ek

  • View
    232

  • Download
    0

Embed Size (px)

Citation preview

  • 7/30/2019 Am J Clin Nutr 1980 Acheson 1147 54

    1/8

    m ethods in nu trition

    Th e Am e ric an Jou rn a l o fC lin ica l N u trition 33: M AY 1980, p p . 11 47 -1 154 . P rin te d in U .S .A . 1 1 47

    The m easu rem en t o f fo od and ene rg y in take inm an -an eva lua tio n o f som e techn iques 3K . J . A cheson ,4 I. T . C am pbe ll,5 0 . G . E dho lm ,6 D . S . M ille r , and M . J . S tock7

    ABSTRACT A dieta ry surv ey w as carr ied ou t on an A nta rct ic b ase o ver a pe riod of 6 to 1 2m onth s on tw e lve ind iv idua ls. Th ree m eth od s of d ete rm in in g ind iv idua l food and ene rgy in takewere u sed -w e igh ing and re cord ing of food as ea ten in com b in ation w ith fo od tab les , d ie ta ry re cal l,a nd b om b calorim etry o f du plic ate m eals . U se of w eighed foo d in take s an d fo od com po sition tab lesu nd ere stim ated energy in ta ke by 7% w hen com pared to ana lys is of d up lica te m ea ls by b om bcalorim etry . O ne w eek w as fo und to be th e m os t pra ctic al p eriod o ver w h ich in take sho uld b em ea sured to d eterm ine hab itual fo od in tak e. Am . J . C lin . N u tr. 33 : 1 147 -1 154 , 198 0 .

    The m easu rem en t o f h ab itua l food andene rg y in take in m an h as been sa id to b e o neof the h ardest ta sks a phy sio lo g ist can und e r-take . Th e tw o basic p rob lem s a re th e accu ratede te rm ina tion ofthe su b jects cu stom ary fo odin take , an d the co nve rsion o fth is in fo rm a tionto en ergy and nu tr ien t in takes. A ny techn iq ueh as to be accu ra te b u t how eve r, shou ld no tb e so in ten sely app lied as to in te rfe re w ithth e sub jec ts d ie ta ry hab its and thus alte r thepa ram ete r b ein g m easu red . A no the r p rob lemis fo r how long food in take sho u ld be m ea-su red be fo re the in fo rm atio n g leaned can b esa id to be a true re flectio n of hab itu a l foo din take , i.e ., the fo od an ind iv idu a l no rm a llyconsum es to p rov id e th e ene rgy an d nu trien trequ irem en ts fo r h is regu la r ev eryd ay ac tiv i-tie s . T he stand ard m e tho ds o f d ete rm in in gfood in tak e h ave b een d iscu ssed e lsew here(1 -4 ) bu t can b e summ arized brie fly a s fo l-l ows . 1 ) In d irec t de term in atio n from in fo r-m a tio n on group co nsum ption , in spectio n o ffam ily bu dge ts , la rd er inv en to ries , f igu re s fo rag ricu ltu ra l p ro duc tion , e tc. 2 ) E stim a tion b yrecall o f foo d con sum ed ove r th e la st day ,w eek , o r m on th . 3 ) M easu rem en t and reco rd -in g of food in take as eaten .

    C onv e rsio n of fo od in tak e in fo rm atio n toene rgy and nu tr ien ts is do ne b y chem ica lan aly sis o r bom b ca lo rim e try o f du p licatesamp le s (5 ) or th e u se o f foo d tab le s such asth ose o f M cC ance an d W iddow son (6 ) .

    H ow accurate a re these techn iq ues? Fo rhow long shou ld food in take b e m easu red tode te rm in e hab itu al in tak e? T o try and an-sw er these question s a p ro lo nged d ie ta ry su r-vey u sing th ree d iffe ren t tech n iqu es fo r m ea -su ring foo d and ene rg y in take w as carr iedou t on 12 ind iv idua ls spend ing 1 yea r on anA nta rctic base (H alley B ay ). T he su rv ey w asca rr ied ou t a s pa rt o f an ene rg y ba lan cesu rvey an d the pe rio d of inv estiga tion ond iffe ren t ind iv idua ls va ried be tw een 6 an d 12mon th s .

    E xp erim en ta l desig nT he w e igh in g and reco rd ing of food in take

    and co llectin g dup licate fo od sam ples p re -I From th e D ep artm ents of N u trit ion an d P hy sio log y ,

    Q ueen E liz ab eth C olle ge, L ondon W 8 7AH and th eD iv is ion of H um an Physio lo gy , N ation al In stitu te fo rM edic al R esearch , L ondon NW 3 , Eng lan d .

    2 K .J.A . and I.T .C . sup por ted by the B rit ish A ntarct icSu r v ey .

    3 A dd res s rep rin t re que sts to : D r . M . J . S tock D ep art-m en t o f P hys io logy , S t G eo rge s H o sp ita l M edic alS chool, L ondon , SW 17 ORE , England .

    4 Presen t address : Ins titu te of P hys io logy , U niv ersi tyof L au san ne, Sw itz erla nd . P re sen t address : U niv er-sity D epar tm ent of A naesth esia , S t. Jam e s H o sp ita l,L eed s, England . P resen t address : Schoo l of E nv i-ronm en tal S tud ies, U niversity C ollege , L ondon ,E ng land . P resen t add res s: D epa rtm ent of P hys io l-o gy , S t G eo rge s H osp ita l M edic al S chool, L ondon ,SW I7 O RE , Eng lan d .

  • 7/30/2019 Am J Clin Nutr 1980 Acheson 1147 54

    2/8

    1 1 48 A CHESON ET A L .

    sen ts a v arie ty o f organ iz ation al prob lem s ifcarried ou t in a norm al c iv iliz ed env ironm en tand the sub ject carries on w ith h is ev ery daylif e . I t is a ted io us job th at requ ires a h ighdegree o fm otiv ation on th e part o f the sub jec tand o f ten a great d eal o f encouragem ent f romthe inv estigato r. In the A ntarctic these task sare easier in th at the sub jec ts are f ree rang ingbu t w ith in a re lativ e ly con f ined area.

    H alley B ay (75#{176}1 S , 2 6# {1 76 }2W) is a B ritishA n tarc tic S urv ey base on the C aird C oas t o fth e W edde ll S ea. C lim atic and en v ironm en talcond ition s hav e b een d escribed p rev iou sly(7 ). A t the tim e o f stu dy 22 m en w ere resid en tthere, en gaged p rinc ipally in the stud y o fupper atm osphere phy s ics, m e teoro log y , andm agne tism . A lth ough som e trav e ling w asdon e, u sually f or recreatio n and holid ay s, thesc ien tif ic program s w ere essen tially static andth e m en spen t the y ear m ore or less con f inedto the base area. A ll p ersonne l had the ir ow nsp ec if ic task s to perf orm -trac tor m echan ic ,cook , g eophy sic is t, e tc . In add ition , all h elp edw ith the dom estic cho res and rou tin e basejo bs-scru bb in g f loo rs , lay in g tab les, trans-p ortin g f ood bo x es and fue l d rum s, and rais-ing bu ried stores .

    T he sub jec ts and observ ers liv ed in c lo secon tac t and de tailed observ atio ns that w ouldhav e presen ted con siderab le prac tical p rob -lem s e lsew here w ere reasonab ly easy to carryou t .

    A ll f o od at H alley B ay w as transportedf rom the U n ited K ingdom and brough t byre lie f sh ip on ce a y ear. T h e d iet co nsisted o ftinned and dehy d rated food , su pp lem en tedtw o o r three tim es a w eek by f ro z en m eat.M eals w ere eaten at regu lar tim es in a com -m unal d in ing hall and w ere prepared by tw op ro f essional cook s.

    B reak f ast con sisted o f cereal and a coo k edd ish , lun ch o f b read and soup , and d inn er o fa m ain d ish and a d essert. B read w as f resh lybak ed . C ak es w ere u sually m ade f or the af te r-n oon tea break . Food w as also f ree ly av ail-ab le at any o ther tim e o f day o r n igh t f orany one w an tin g to prepare it f o r h im se lf .M aterials and m eth ods

    D etails o f all 12 sub jec ts are g iv en in T ab le 1 . M eanage w as 24 y ears , m ean w eight o v er th e y ear w as 71 .56k g , and m ean height w as 179 cm . Du ring the pe rio d o fstudy th e f ood in tak e o f each sub jec t w as determ in ed f oran average of 1 w eek o f each m on th . O ne subjec ts f o od

    in tak e w as m easured alm os t co ntinuous ly f o r 50 w eek sand ano the rs f or 29 w eek s. T hree m ethod s o f determ in-in g f ood and en erg y in tak e w ere used .

    Weigh ed in ta ke a s e a tenE ach w eek o f study th e subje cts w eighed and record ed

    all f o od con sum ed . Ene rgy in tak e w as d eriv ed f rom thefood com pos itio n tables o f M cCance and W iddow son(6) u sin g the A tw ater n utrien t co nv e rsion facto rs o f 4 , 9 ,and 4 k cal/g of carboh y drate, f at , and prote in , respec -tiv e ly . T h e calculatio ns w ere perf orm ed b y com pu ter (8) .

    Dur ing each w eek of study intake w as m easu red indiscrete periods of 24 hr startin g at a tim e of day con-v enient to the subje ct. S ubjects w eighed and reco rdedth eir ow n fo od intake ( f oo d item and w e igh t) in note-boo k s w ith w h ich they w ere prov ided . D ietary scalesw ere placed at strategic points around the base (din ingroom , k itchen , bar, lab oratories, etc). T w o sorts o f scalesw ere used-a C hatillon dietary scale w eighin g up to 50 0g in 2 g div isio ns and a S alte r d ie tary scale w e igh ing upto I k g in 5 g div isio ns . T h e accu racy of the scales w asch eck ed and ad jus ted by th e deale rs be fore leav ing theUn ited K ingdom .

    A ll item s w ere w eig hed and reco rded separately c x -cept f or d iscre te standard item s such as can s of bee r orblo ck s of cho colate f or w h ich an av erage w eig ht w astak en (th e sub jec t m ere ly record ed the natu re o f th enum ber consum ed). Subjects w ere issued plastic con -taine rs f or b utte r and sugar th at w e re w e igh ed at th estart and fm ish of each 24-h r pe rio d . T ea and co f f eew ere not w eighed but added m ilk and sugar w ere.

    D ie ta r y r e ca llO nce during each w eek of study th e subjects w ere

    asked to w rite dow n everything they could rem em bereatin g during th e prev ious 24 hr. T h is w as usually doneon a prin ted dietary ques tio nnaire (9 ) b ut on a num berof occas ion s sub jec ts w ere ask ed to p rov ide th is in f o r-mation on a blank sh eet o f pap er. N o attem pt w as m adeto in te rv iew th em and th ey w ere le f t to f ill in th e que s-t ion naire in their ow n w o rds.

    T he record s o f d ietary recall w ere analy sed ind epen-

    T A B L E 1Phy s ical characteristic s ando ccupation s of sub jec ts

    S u b jec t H e igh t W e igh t A ge O ccu pationcm kg yr

    1 185 92.11 26 G eneral ass istant2 174 63.07 23 T racto r m echanic3 165 64.89 26 M ed ical o f f icer4 178 71.90 24 N utrition ist5 179 6 4 . 8 8 2 1 Cook6 174 67.8 6 26 E lec tric ian7 182 7 1 .96 29 M eteoro log ist8 193 78 .40 24 G eophy s icist9 177 68.26 24 Ionosphericis t

    10 183 78.36 21 M eteorolo gis tII 183 6 6 . 4 7 22 C ook12 176 70.53 22 lo no sph ericist

    M e an 179 71.5 6 24

  • 7/30/2019 Am J Clin Nutr 1980 Acheson 1147 54

    3/8

    FOOD AND EN ERGY IN TAKE 1 1 49

    dently by workers experienced in the analysis of suchinformation using their own standard techniques. T henutrient conversion factors used for fat, carbohydrateand protein were the A twater factors of 9, 4, and 4 kcal/g, respect ively.

    Ana ly sis o f 24 -hr d up lica te

    For one period of 24 hr during each week of study aduplicate sample was taken of every item consumedincluding alcoholic and nonalcoholic drinks other thanwater, and placed in a plastic container. A t the end of 24hr the entire 24-hr duplicate was homogenized and thehomogenate weighed to the nearest 2 g. A 300- to 400-gsample of the homogenate was taken, weighed, andfrozen to await analysis in the U nited K ingdom.

    Eventually the sample was thawed then dried toconstant weight in an oven at 105 C for a period of48 to72 hr. The dried sample was ground to a f ine powderinitially with a pestel and mortar, then in a coffee grinder,and finally placed in the oven for a further 24 hr toevaporate moisture absorbed during the grinding proc-ess. The resulting powder was analyzed for gross energyin the ballistic bomb calorimeter (5) . A nalyses wereperformed in triplicate or until the standard error of themean was less than 1%. Pulvenised dried sucrose wasused as the standard. A correction for alcohol lost in thedrying process was made using the alcoholic content ofdrinks given in the food tables (6) and taking 7 kcal/g asthe energy value of alcohol.

    T he metabolizable energy ofthe entire duplicate sam-plc was calculated using the equation of M iller andPayne (5); metabolizable energy = 0.95 gross energy(kilocalorie) - 7.5 nitrogen (gram). Since the nitrogencorrection is small, it was calculated from the total weightof ingested nitrogen provided by food tables.

    ResultsM ean 24 hr energy intakes (SD ) com-

    puted from the weighed intake using the A t-water factors are given in T able 2 along withthe mean body weight and the total numberof individual days each subject was studied.M ean intake was 3210 kcal w ith a range of2430 to 3970 kcal.

    Table 3 shows the coefficient of variationofthe mean 24-hr intake ofeach subject. T hisgives a measure of the variability of day today intake and in order to provide a measureof the variability of intake from week to weekand longer periods, the follow ing procedurewas adopted. For each subject mean dailyintakes were calculated for periods of oneweek, and the coefficient of variation of thesemeans is given in Table 3 and indicates thevariation in intake from week to week. Thisprocedure was then used to calculate coeffi-cients of variation for means determined overperiods of 2, 3, and 4 weeks.

    TABLE 2M e an 24-hr energy intake, mean bodyweight, and number ofday s studi ed

    . M ean 24-hr energy in. . Total no. of daysSubject M ean w ei ghttake (SE) studiedkcal k g

    I 2 4 3 0 1 3 0 9 2 . 1 1 4 22 3 2 7 0 1 2 0 6 3 . 0 7 5 23 2730 5 0 6 4 . 8 9 1734 3370 40 7 1.90 3135 2580 1 30 6 4 . 8 8 696 3 1 8 0 9 0 6 7 . 8 6 7 07 3470 120 71.96 428 3730 160 78.40 359 3 3 3 0 9 0 6 8 . 2 6 1 0 1

    10 3970 90 78.36 63I I 3 4 1 0 1 4 0 6 6 . 4 7 4 21 2 3 0 5 0 90 70.53 83

    Mea n 3 2 1 0 7 1 . 5 6 9 0a Energy intakes derived from weighed food intake

    and A twater nutrient conversion factors.

    TA BLE 3Coefficient of variation of energy intakesover periods of 1 d ay and I , 2, 3, and 4 weeks

    Subject I D ay I W k 2 W k 3 W k 4 W kI 34.6 17.32 25.7 14.63 24.9 13.6 12.1 11.7 9. 54 18.2 10.5 8.1 7.3 6.9S 42.1 28.9 28.1 25.16 23.8 8.1 3.1 5.47 21.9 13.18 25.5 6.69 26.2 10.1 8.2 6.5

    10 18.2 8.8 7.3 9.111 27 20.512 26.7 5.9 5.6 6.5

    a SD/mean, %.

    Enough data were available to completethis analysis for 4 weeks on only two subjects,for 3 weeks on five subjects, and for only 1week on the remaining subject. I n all casesth e coefficient ofvariation dropped markedlywhen going from daily intakes to weeklyintakes, i.e., the variability of energy intakewas much less from week to week than it wasfrom day to day. W hen intake was consideredin periods of 2, 3, and 4 weeks the subjectson whom enough data were available showedonly a very small decrease in the coefficientof variation as the period of observation in-creased.

    V alues for energy intake derived from 24-hr dietary recall are given in T able 4 along

  • 7/30/2019 Am J Clin Nutr 1980 Acheson 1147 54

    4/8

    1150 A CH ESON ET A L .TABLE 4Comparison of energy intakes derived from weighed food intakeand dietary recall

    E n er gy i nt ak es Errornu b j e c t

    Weighed intake (W )k c a l

    R ecall (R )k c a l

    R-Wk c a l

    R/Wr a n g e

    1 25502650

    2370 (Q )1875 (P)

    -180-775

    0.66-1.320.61-0.81

    22

    2 30503230

    2300 (Q )1825 (P)

    -750-1405

    0.65-0.700.53-0.66

    34

    3 32102224

    21 15 (Q )1276 (P)

    -1095-948

    0.56-0.710.33-0.76

    25

    4 3372 2627 (Q ) -745 0.62-1.06 155 2296 2014(Q) -282 0.75-1.26 56 3021 2111 (Q) -910 0.51-0.90 97 2895

    33902425 (Q )2505 (P)

    -470-885

    0.72-1.060.63-0.94

    42

    8 38903787

    3600 (Q )2327 (P)

    -290-1460

    0.980.45-0.85

    13

    9 3196 2048(Q) -1148 0.49-0.84 1010 3797 3424(Q ) -373 0.68-1.03 911 3755

    26653140(Q)2305 (P)

    -615-360

    0.63-0.960.69-1.10

    42

    12 2468 2153 (Q ) -315 0.77-0.98 4a D ietary recall methods involved either a printed questionnaire (Q ) or a blank sheet of paper (P) to record the

    previous days food intake. bNo. of 2 4- hr observ ati ons.

    with the 24-hr energy values computed fromthe weighed intake for the same days. T hesame nutrient conversion factors, 4, 9, and 4kcal!g for protein, fat, and carbohydrate, re-spectively, were used in the calculation ofenergy content in both cases, so any differ-ence between the two groups of values ispurely a function of the subject s ability toremember what he ate in the previous 24 hr.I t should be remembered that the subject hadalso weighed and recorded each item con-sumed. Table 4 gives the dietary recall valueswhen the printed questionnaire was used andalso when some subjects were asked to writedown on a blank sheet of paper what theycould remember eating over the preceding 24hr.

    I n terms ofenergy, subjects underestimatedtheir food intake by 21% using the printeddietary questionnaire. On only six out of the68 occasions did recall overestimate food con-sumption. Some subjects showed a greaterfacility for remembering what they had eatenthan others, but the performance of most wasrather erratic. I t would appear, however, thatthe questionnaire produced the best results-an underestimate of mean intake by 2 1% asagainst 33.6% when the blank sheet of paperwas used-a difference that was statistically

    significant (P < 0.05). Inspection of the recallrecords to determine the cause of the discrep-ancy between recall and records of weighedintake showed that the subjects usually leftout one complete food item and tended tounderestimate the size of their portions.

    In T able 5 values are given for daily me-tabolizable energy intake calculated using theenergy factors of A twater (4, 9, and 4 kcal, g)and those used by M cCance and W iddowson(3 .75 , 9 .3 , and 4.1 kcal!g ofcarbohydrate, fat,and protein, respectively) compared with me-tabolizable energy derived from bomb cab-rimetry measurements. M eans are presentedfor a total of 96 comparisons. Use of A twaterfactors overestimated metabolizable energycontent by a mean of 6.7% compared withdirect analysis and the M cCance and W id-dowson factors overestimated energy by4.6% . Of the results based on A twater factors,52% were within 10% ofthe bomb figure, 31%within 10 to 20% , 15% between 20 and 30% .Using a paired t test the difference betweenthe analytical f igure and that derived usingA twater factors were signif icant (P