Work characteristics, personal resources, and employee well-being

Work characteristics, personal resources, and employee well-being: A longitudinal study among Finnish firefighters

People and WorkResearch Reports 109

Auli Airila

Work characteristics, personal resources, and em

ployee well-being: A

longitudinal study among Finnish firefighters

Airila A

uli

In today’s dynamic work life, employees are expected not only to be healthy and able to work efficiently, but also to be highly engaged and motivated. However, at the same time, the prevalence of mental health disorders and diseases of the musculoskeletal systems is also high. Therefore, understanding and enhancing employee well-being as a whole is essential. The present study examined the longitudinal effects of work characteristics (i.e., job demands and resources), personal resources and lifestyle on employee well-being, using data on Finnish firefighters.

This study showed that job demands, job resources, personal resources, and lifestyle all affect employee well-being over time. It offers new information for researchers, policy-makers, HRM at workplaces, and occupational health services about possibilities to enhance employee well-being, particularly work engagement, and consequently prevent ill-health and promote long working careers.

109

ISBN 978-952-261-525-1

Orders:Finnish Institute of Occupational HealthArinatie 3FI-00370 HelsinkiFinland

E-mail [email protected]

ISBN 978-952-261-525-1 (paperback)ISBN 978-952-261-526-8 (PDF)

ISSN-L 1237-6183ISSN 1237-6183

Cover picture: Saimi Airila

People and Work

Editor in chief Harri Vainio

Scientific editors Irja Kandolin Timo Kauppinen Kari Kurppa Anneli Leppänen Hannu Rintamäki Riitta Sauni

Editor Virve Mertanen

Address Finnish Institute of Occupational Health Arinatie 3 FI-00370 Helsinki Tel. +358-30 4741 www.ttl.fi

Cover design Mari Pakarinen Cover picture Saimi Airila, 6 years ISBN 978-952-261-525-1 (paperback) ISBN 978-952-261-526-8 (PDF) ISSN-L 1237-6183 ISSN 1237-6183

Press Suomen Yliopistopaino Oy – Juvenes Print, Tampere 2015


Auli Airila

People and Work Research Reports 109

Finnish Institute of Occupational Health Helsinki 2015

AcAdemic dissertAtion

to be presented, with the permission of the Faculty of social sciences of the University of Helsinki, for public examination in Auditorium XiV, University main Building (Unioninkatu 34, 3rd floor),on 20th march 2015, at 12 noon.

Author´s Auli Airilaaddress: centre of expertise for Work and Work organizations Finnish institute of occupational Health Helsinki, Finland

Supervisors: research director research Professor Jari Hakanen Helsinki collegium for Advanced studies University of Helsinki, Finland centre of expertise for Work and Work organizations Finnish institute of occupational Health Helsinki, Finland

Professor Anna-maija Pirttilä-Backman department of social research social Psychology University of Helsinki, Finland

Reviewers: Professor siw tone innstrand department of social Work and Health science the norwegian University of science and technology (ntnU) trondheim, norway

Professor saija mauno school of social sciences and Humanities University of tampere, Finland

Opponent: Professor Ulla Kinnunen school of social sciences and Humanities University of tampere, Finland

3

Contents

ABBreViAtions............................................................ 5

List oF originAL ArticLes .......................................... 6

AcKnoWLedgements ................................................... 7

ABstrAct .................................................................... 9

tiiVisteLmä ................................................................ 12

1 introdUction ......................................................... 151.1 conceptualization of employee well-being ................ 17

1.1.1 Affective well-being ...................................... 191.1.2 Health-related well-being .............................. 241.1.3 Life satisfaction ........................................... 29

1.2 theoretical frameworks for explaining employee well-being ............................................................ 30

1.2.1 Job demands-resources model ...................... 301.2.2 conservation of resources theory ................... 331.2.3 Broaden-and-Build theory .............................. 351.2.4 comparison of the theoretical frameworks ....... 36

1.3 study context: rescue services ............................... 371.4 Aims of the study .................................................. 40

2 metHods ................................................................. 432.1 Participants .......................................................... 432.2 measures ............................................................. 46

2.2.1 employee well-being ..................................... 462.2.2 Job demands ............................................... 492.2.3 Job resources .............................................. 502.2.4 Personal resources ....................................... 512.2.5 Lifestyle factors ............................................ 51

2.3 statistical Analyses ............................................... 53

3. resULts ................................................................... 563.1 relationship between work engagement and work ability (study i) .................................................... 563.2 Work engagement as a mediator between job resources, personal resources, and work ability (study ii) ......... 58

4

contents

3.3 mood trajectories and their relationship to work ability, self-rated health, and life satisfaction (study iii) ...... 613.4 Pain-depression trajectories, work characteristics, and individual factors (study iV) ............................. 63

4 discUssion ............................................................. 674.1 motivational state of work engagement associated with work ability ................................................... 674.2 motivational process and its health-related properties .. 70

4.2.1 Longitudinal evidence of the importance of job and personal resources ................................. 704.2.2 Health resources as drivers and outcomes of the motivational process in the Jd-r model ...... 71

4.3 mood trajectories were differentially related to well-being ............................................................ 734.4 developmental paths in musculoskeletal pain and depressive symptoms, and health-enhancing effects of job resources and individual factors ..................... 754.5 methodological considerations ................................. 794.6 implications for future research .............................. 824.7 Practical implications ............................................. 854.8 conclusions .......................................................... 87

reFerences ................................................................ 89

originAL ArticLes ..................................................... 109

5

AbbReviAtions

AIC Akeike’sinformationcriteriaBaB Broaden-and-BuildtheoryBIC BayesianinformationcriterionBMI BodymassindexCFA ConfirmatoryfactoranalysisCFI ComparativefitindexCI ConfidenceintervalCOR ConservationofResourcestheoryDPS DepressivesymptomsEFA ExploratoryfactoranalysisERI Effort-RewardImbalancemodelJDC JobDemand-ControlmodelJD-R JobDemands-ResourcesmodelMLE Maximum-likelihoodestimationMSP MultisitemusculoskeletalpainLCGM LatentclassgrowthmodelingOR OddsratioPOMS ProfileofmoodstatesRMSEA RootmeansquareerrorofapproximationSDT Self-determinationtheorySEM StructuralequationmodelingSRH Self-ratedhealthTLI Tucker-LewisindexT1 Time1(1996)T2 Time2(1999)T3 Time3(2009)UWES UtrechtworkengagementscaleWAI WorkabilityindexWHO WorldHealthOrganization

6

List oF oRiginAL ARtiCLes

I Airila,A.,Hakanen,J.,Punakallio,A.,Lusa,S.,&Luukkonen,R.(2012).Isworkengagementrelatedtoworkabilitybeyondworkingconditionsandlifestylefactors?International Archives of Occupational and Environmental Health,85,915–925.

II Airila,A.,Hakanen,J.J.,Schaufeli,W.B.,Luukkonen,R.,Punakallio,A.,&Lusa,S.(2014).Arejobandpersonalresourcesassociatedwithworkability10yearslater?Themediatingroleofworkengagement.Work & Stress,28,87–105.

III Airila,A.,Hakanen,J.J.,Luukkonen,R.,Lusa,S.,&Punakallio,A.(2013).Positiveandnegativemoodtrajectoriesandtheirrelationshiptoworkability,self-ratedhealthandlifesatisfaction:A13-yearfollow-upstudy.Journal of Occupational and Environmental Medicine,55,779–785.

IV Airila,A.,Hakanen,J.J.,Luukkonen,R.,Lusa,S.,Punakallio,A.,&Leino-Arjas,P.(2014).Developmentaltrajectoriesofmultisitemusculoskeletalpainanddepressivesymptoms–theeffectsofjobdemandsandresourcesandindividualfactors.Psychology & Health,29,1421–1441.

Theoriginalarticlesarereprintedinthisreportwiththekindpermissionofthecopyrightholders,SpringerScience+BusinessMedia(I),Taylor&Francis(II,IV),andWoltersKluwerHealth,LippincottWilliams&Wilkins(III).

7

Acknowledgements

There are many people who have helped and supported me during my PhD project.

I am most grateful to my supervisors, Professor Anna-Maija Pirttilä-Backman and Research Professor Jari Hakanen. Thank you Anna-Maija for your support, thorough guidance and help throughout the study and writing process. It has been a huge privilege to work with Jari, an outstanding expert in positive work psychology and employee engage-ment. I would never have been able to complete my thesis without your endless encouragement, positive and constructive feedback, wise advice, and support. Thank you so much, Jari!

I also wish to warmly thank the reviewers of my thesis, Professor Saija Mauno from the University of Tampere and Professor Siw Tone Innstrand from The Norwegian University of Science and Technology, for their valuable comments on my manuscript.

I am also indebted to the co-authors of the original publications of my thesis. Dr. Sirpa Lusa and Dr. Anne Punakallio introduced me to the Finnish firefighters’ world of work and offered me the opportunity to use longitudinal research data on firefighters. Thank you both! Dr. Ritva Luukkonen has been a highly valuable guide in the world of statistics, thank you for all your help and supervision in the statistical analyses! I also want to thank Professor Wilmar Schaufeli from the Utrecht Uni-versity and Dr. Päivi Leino-Arjas from FIOH for their collaboration, guidance, and advice.

I have also learned a great deal from the doctoral seminars held by the discipline of Social Psychology at the University of Helsinki. I wish to thank Professor Anna-Maija Pirttilä-Backman and Professor Inga Jasinskaja-Lahti for their stimulating and insightful seminars. I also thank my peer PhD students from the seminars for commenting on my papers.

8

AcKnoWLedgements

Iextendmywarmestthankstoallmyco-workersatFIOH.IamgratefultoDr.AnneliLeppänen,DirectoroftheCentreoftheExper-tiseforDevelopmentofWorkandOrganizations,forallowingmetouseFIOH’sfacilitiesandresourcestocarryoutthiswork.IalsothankmyformerteamleaderDr.AriVäänänenforhissupport,forgrantingmestudyleaves,andforhisconstructivecommentsonmymanuscript.IamgratefultoallmycolleaguesatFIOHwithwhomIhavehadtheopportunitytowork,includingMinnaToivanen,BarbaraBergbom,PiiaSeppälä,MinnaJanhonen,JenniErvasti,KirsiYli-Kaitala,andmanyoth-ers.Thankyouforyourpeer-support,practicalcommentsonmywork,andinspiring,criticalandexcitingdiscussionsoverlunchandcoffeebreaksthroughouttheyears!IalsothankVirveMertanen-HalinenforprofessionaleditingofmythesisandAliceLehtinenforlanguageediting.

Financialsupportduringthisresearchprocesshasalsobeenimpor-tant.ForthisIthankTheEmergencyServicesCollegeandTheFinnishDoctoralProgramonLabourandWelfareStudies–LabourNet.Inaddi-tion,IthanktheUniversityofHelsinki,TheFinnishWorkEnvironmentFund,andtheFinnishConcordiaFundforofferingmescholarshipstointernationalconferences.

Mostofall,Iamgratefultomyfamilyandfriends.IowemywarmestthankstomyparentsTuireandSeppoforencouragingandsupportingmeovertheyears.Inparticular,IamgratefultomymotherandherhusbandAntti,andmyparents-in-lawEevaandMauriforalltheirsup-port,andfortheiressentialhelpwithchild-care.Combiningwork,PhDstudies,andfamilyhasbeenmucheasierwithyourhelp.IalsowanttothankmydearfriendsRiikkaandTiina,foryourfriendshipandsupport,forlistening,andforournot-always-so-scientific-but-highly-importantdiscussions!Alovingthankyougoestomydearchildren,Laura,Oskari,andSaimi.Thanksforinspiringme,andforourlovely,oftenamusingdiscussionsrelatedtomythesis.Youhavetaughtmemorethanyouknow!AndspecialthankstoSaimiforthepositivecoverillustration!Last,butdefinitelynotleast,IwanttothankmydearhusbandMarkus,forallyourencouragement,continuoussupport,helpinvariousways,andlove.

Kerava,March2015

Auli Airila

9

AbstRACt

Intoday’sdynamicworklife,employeesareexpectednotonlytobehealthyandabletoworkefficiently,butalsotobehighlyengagedandmotivated.However,atthesametime,theprevalenceofmentalhealthdisordersanddiseasesofthemusculoskeletalsystemsisalsohigh.There-fore,understandingandenhancingemployeewell-beingasawholeisessential.

Theaimofthisthesiswastoexaminethelongitudinaleffectsofworkcharacteristics(i.e.,jobdemandsandresources),personalresources,andlifestyleonemployeewell-beingbyapplyingthreetheoreticalframeworks–JobDemands-Resources(JD-R)model,ConservationofResources(COR)theory,andBroaden-and-Build(BaB)theory.Thesetheoreticalframeworkswerechosenastheyoffertheoreticalexplanationsfortheunderlyingpsychologicalprocessesbetweentheantecedentsandwell-beingoutcomes.Inthisthesis,employeewell-beingwasunderstoodasamultidimensionalconstructcoveringaffective,cognitive,andhealth-relatedcomponents.Throughthisconceptualization,bothpositiveandnegativeaspectsaswellasjob-relatedandcontext-freeindicatorsofwell-beingwereacknowledgedandstudied.

Morespecifically,thisstudyaimedtoextendcurrentknowledge,firstly,byexaminingtherelationshipbetweentheaffective-emotionalstateofworkengagementandworkabilityanditssub-dimensions.Assuch,thisstudycontributedtocurrentoccupationalhealthpsychologybyexaminingtheroleofworkengagementinrelationtoworkabilitybeyondlifestylefactorsandworkcharacteristics.Secondly,thisstudyaddednewcontributionbyinvestigatingtheroleofworkabilityinthemotivationalprocessoftheJD-Rmodel.Inaddition,thedevelopmentalpathsinpositive(i.e.,vigor)andnegative(i.e.,fatigue)affectivestatesandtheir

10

ABstrAct

relationtowell-being,aswellasintra-individualchangetrajectoriesinmultisitemusculoskeletalpainanddepressivesymptomsandtherelatedantecedentswereinvestigated.Byapplyingaperson-centeredapproach,thisstudywasabletoidentifyindividualdevelopmentalpathsovertime.

ThedatawascollectedwithquestionnairesamongFinnishfirefightersduringa13-yearperiodwiththreemeasurementpoints(i.e.,1996,1999,and2009).Twoslightlydifferentdatasetswereusedinthefoursub-studies.Dataset1(StudiesI–II)consistedoffirefighterswhorespondedtothequestionnairesinboth1999(T2)andin2009(T3),andwerestillemployedintheirprofessionin2009(n=403).Dataset2(StudiesIII–IV)includedthosefirefighterswhorespondedtothequestionnairesateachmeasurementpoint,andwerestillemployedintheirprofessionin2009(n=360).Severalvalidatedmeasuresofworkcharacteristics,personalresources,lifestyleandwell-beingwereapplied.

Thestudyshowedthatapositivestateofworkengagementconsistingofvigour,dedication,andabsorptionwassignificantlyassociatedwithworkabilityevenafteradjustingforvariousindividualandworkchar-acteristics.Furthermore,lifestylefactors,workcharacteristicsandworkengagementweremorestronglyrelatedtothesubjectiveworkabilitydimensionsthanthetwomoreobjectivesub-dimensions(i.e.,numberofdiseasesandsickleave).ThesecondmajorfindingandatheoreticalcontributionofthestudywasthatthemotivationalprocessoftheJD-Rmodelalsohashealth-relatedproperties.Moreprecisely,jobandpersonalresourceshadlong-termeffectsonworkengagementandconsequentlyonworkability,thusexpandingthepotentialpositiveoutcomesofthemotivationalprocessincludedintheJD-Rmodel.Thus,thedualroleofworkabilitybothasapredictor(i.e.,health-relatedresource)thatmayfosterengagement,andasanoutcomeofthemotivationalprocesswasfound.

Inaddition,thisstudyshowedthatdifferentdevelopmentalpathsinpositive(i.e.,vigor)andnegative(i.e.,fatigue)affectivestatesarepos-sible.Typically,aratherhighlevelofvigorandlowleveloffatiguewasobserved.However,othercombinationsofpositiveandnegativeaffectivestatesalsoexisted,forexample,atrajectorywithhighbutdecreasinglevelofvigorandhighleveloffatigue.Theseresultssuggestthatthepositiveandnegativeaffectmaynotalwaysbedirectoppositesorindependenteither.Inaddition,althoughbothvigorandfatiguewereratherstableover

11

ABstrAct

time,somechangesoccurredoverthe13-yearfollow-upperiod.Moreprecisely,adeclineinpositivemoodandanincreaseinnegativemoodwereapparentbetween1996and1999whereasthechangeswereratherminimalduringthefollowingten-yearperiod.Moreover,developmentalpathsofaffectivestatesweredifferentiallyrelatedtowell-being.There-fore,theresultssuggestboththepossibilityandneedtofosterpositivedevelopmentalpathsinaffectivestates,i.e.,toincreasevigorandpreventfatigueatwork,inordertomaintainandimproveemployeewell-being.Similarly,differentdevelopmentalpathsinmultisitemusculoskeletalpainanddepressivesymptomswereobserved,andchangesovertimeprovedtobepossible.Theresultsalsoindicatedthatjobdemands,jobandpersonalresourcesaswellaslifestylewerepartlydifferentlyrelatedtopain-depressiontrajectories.Thisneedstobeacknowledgedwhenplanninginterventionsforpreventingmusculoskeletalpainanddepres-sivesymptoms.

Toconclude,thebeneficialroleofjobresourcesandpersonalre-sources,andsimilarlythedetrimentaleffectofjobdemandsinexplainingemployeewell-beingwereshowninthislongitudinalstudy,andthus,provingfurthersupportfortheassumptionsoftheJD-Rmodel,CORtheory,andBaBtheory.Evidently,thefindingssuggestthatamotivatedandenergeticworker,whohasresourcefuljobandnottoohighmentalandphysicaldemandshasbetterwell-beingthanaco-workerwhoislessengagedandexperiencesamoreunfavourablesituationintermsofdemandsandresources.Moreover,agoodlevelofself-esteem,anoptimisticviewofthefutureandhealthylifestylehabitsmayincreasethelikelihoodofbeinghealthyandhappy.Overall,theseresultssuggestthatworkcharacteristicsmayplayamorecrucialrolethanlifestyleinenhancingemployeewell-being.Assuch,workorganizationshavegoodopportunitiestoreinforceandmaintainwell-beingoftheiremployees,andconsequentlypreventill-healthandpromotelongworkcareers.

12

tiivisteLmä

Jatkuvastimuuttuvatyöelämäasettaatyöntekijöilleyhäenemmänvaatimuksia.Hyväterveysjatyökykyeivätyksinäänriitä,vaantyön-tekijöidenoletetaanolevanmyösentistäsitoutuneempiajamotivoitu-neempiatyöhönsä.Samanaikaisestimielenterveysongelmatsekätuki-jaliikuntaelinvaivatovatyhäsuurempiaongelmiatyöikäistenkeskuudessa.Siksityöntekijöidenhyvinvoinninkokonaisvaltainenymmärtäminenjaedistäminenovatentistätärkeämpiä.

Tämäntutkimuksentavoitteenaoliselvittäätyöhönliittyvienteki-jöiden,yksilöllistenvoimavarojenjaelintapojenyhteyttätyöntekijöidenhyvinvointiin.Työhönliittyvinätekijöinätutkittiintyönvaatimuksiajatyönvoimavaroja(esimiestuki,työntekijöidenvälisetsuhteet,työ-tehtäväänliittyvätvoimavarat).Tutkimuksessahyödynnettiinkolmeateoreettistaviitekehystä,jotkaovattyönvaatimusten-työnvoimavaro-jen(TV-TV)malli,voimavarojensäilyttämisenteoriajapositiivistentunteidenlaajentumis-jarakentumisteoria.Nämäkolmeteoreettistaviitekehystävalittiin,koskaniidenavullaolimahdollistatarkastellavoimavarojen,elintapojenjahyvinvoinninvälisiäyhteyksiä.Työnteki-jöidenhyvinvointiontässätutkimuksessaymmärrettymoniulotteisenakokonaisuutena,jossaotetaanhuomioonhyvinvoinninaffektiiviset,terveyteensekäkognitiiviset(kutenelämääntyytyväisyyteen)liittyvätosa-alueet.Työntekijöidenhyvinvointiaontarkasteltumyönteisestäjakielteisestänäkökulmasta.Siksionselvitettysekätyönimuajatyökykyäettätuki-jaliikuntaelintenkipuoireitajamasennusoireita.

Tutkimuslaajensinykyistätietämystäselvittämällätyönimunyhteyttätyökykyynjaseneriosa-alueisiin.SamallaseselkeyttityökyvynrooliaTV-TV-mallinmotivaatiopolulla,jonkaonaiemminajateltusisältävänainoastaanmyönteisiäorganisatorisiaseurauksia,kutenhyväntyöstäsuo-

13

tiiVisteLmä

riutumisen.Lisäksitutkimuksessaselvitettiinmyönteisten(tarmokkuus)jakielteisten(väsymys)mielialojenkehityspolkujajaniidenyhteyttähyvinvointiinsekätuki-jaliikuntaelintenkivunjamasennusoireidenkehityspolkujajaniitäennustaviatekijöitä.Yksilökeskeinenlähestymis-tapamahdollistiyksilöllistenkehityskulkujenpaikallistamisen.

Tutkimusaineistonaolisuomalaisillepalomiehilletehtykyselytut-kimus.Aineistokoottiin13vuodenaikanakolmellakyselyllävuosina1996,1999ja2009.Neljässätutkimuksenosatutkimuksessakäytettiinkahtahiemanerilaistaaineistoa.Aineisto1(osatutkimuksetI–II)koostuiniidenpalomiestenvastauksista,jotkaosallistuivatkyselyynvuosina1999(T2)ja2009(T3)jatyöskentelivätedelleenammatissaanvuonna2009(n =403).Aineisto2(osatutkimuksetIII–IV)kattoiniidenpalomiestenvastaukset,jotkaosallistuivatkyselyynjokaisellakolmellamittauskerrallajajotkatyöskentelivätedelleenammatissaanvuonna2009(n=360).Tutkimuksessakäytettiinuseitavalidejamittareita,joillatutkittiintyönvaatimuksiajavoimavarojasekäyksilö-jaelintapatekijöitä.

Tutkimusosoitti,ettätyönimuelimyönteinentunne-jamotivaatio-tila,jokamuodostuutarmokkuudesta,omistautumisestajauppoutumi-sesta,olimerkitsevästiyhteydessätyökykyynsenkinjälkeenkunuseatyksilöönjatyöhönliittyvättekijätolivakioitu.Lisäksitutkimusosoitti,ettäelintapoihinjatyöhönliittyvättekijätsekätyönimuolivatvahvem-minyhteydessätyökyvynsubjektiivisiinulottuvuuksiinkuinsenobjek-tiivisiinulottuvuuksiin,kutentautienmääräänjasairauspoissaoloihin.

Toinentutkimuksenkeskeinenhavaintojateoriaakehittävätulosolise,ettäTV-TV-mallinmotivaatiopolkuunsisältyyterveyteenliittyviäominaisuuksia.Toisinsanoentyöhönjayksilöönliittyvillävoimavara-tekijöilläolipitkänaikavälinvaikutuksiatyönimuun,jokaedelleenoliyhteydessätyökykyyn.Näinollentutkimusosoitti,ettäTV-TV-mallinmotivaatiopolkuvoijohtaamyösparempaanhyvinvointiinjaterveyteen.Samallatyökyvynkaksoisroolisaitukea:työkykyonmotivaatiopolullasekäselittäjä(eliterveyteenliittyväyksilönvoimavara),jokaedistäätyönimua,ettämotivaatiopolunmyönteinenterveyteenliittyvälopputulema.

Tutkimusosoittilisäksi,ettäerilaisettarmokkuudenjaväsymyksenkehityspolutovatmahdollisia.Useimmitenpalomiehilläolikorkeatarmokkuudenjamatalaväsymyksentila,muttamuitakinmyönteistenjakielteistenmielialojenyhdistelmiäesiintyi.Tuloksetosoittivat,ettäkielteinenjamyönteinenmielialaeivätoleselkeästivastakkaisiataipääl-

14

tiiVisteLmä

lekkäisiä.Vaikkamyönteisetjakielteisetmielialatosoittautuivatvarsinpysyviksi,joitakinmuutoksiakintapahtui13vuodessa.Esimerkiksivuosien1996ja1999välillätarmokkuuslaskivastaavastiväsymyksenkasvaessa,kuntaasseuraavankymmenenvuodenaikanamuutoksetmielialoissaolivatvarsinvähäisiä.Tutkimusosoittimyös,ettätarmok-kuudenjaväsymyksenerilaisetkehityspolutolivateritavallayhteydessähyvinvointiin.Tutkimusosoittikin,ettämyönteistenmielialojenkehit-täminenonmahdollista,sillätarmokkuudenedistäminenjaväsymyksentorjuminenmyötävaikuttavattyöntekijöidenhyvinvointiin.

Lisäksitutkimusosoitti,ettätuki-jaliikuntaelintenkipuoireetjamasennusoireetvoivatkehittyäeritavoin.Myöstyönvaatimukset,työ-hönjayksilöönliittyvätvoimavaratsekäelintapatekijätovateritavallayhteydessäerilaisiinkipu-masennus-kehityspolkuihin.Tämäonsyytähuomioida,kunkipu-jamasennusoireitapyritäänvähentämään.

Tutkimusosoitti,ettätyönvoimavarat,yksilönvoimavaratjatyönvaatimuksetvaikuttavattyöntekijöidenhyvinvointiinhyvinpitkälläkinaikavälillä.NäinollentutkimusvahvistaaTV-TV-mallin,voimavarojensäilyttämisenteoriansekäpositiivistentunteidenlaajentumis-jaraken-tumisteorianolettamuksia.Tutkimuksentuloksetosoittavat,ettämoti-voitunutjaenerginentyöntekijä,jollaonuseitavoimavarojasisältävätyösekäalhaisettyönhenkisetjafyysisetvaatimukset,voiparemminjaonterveempikuintyöntekijä,jokaonvähemmänsitoutunuttyöhönsäjajollaonepäsuotuisammatolosuhteettyönvaatimustenjavoimavarojensuhteen.Myöshyväitsetunto,optimistinensuhtautuminentulevai-suuteensekäterveellisetelintavatvoivatmyötävaikuttaatyöntekijänhyvinvointiin.

Tuloksetviittaavatsiihen,ettätyöolotekijätvoivatollaelintapate-kijöitätärkeämpiätyöntekijöidenhyvinvoinninedistämisessä.Näihintekijöihinmyöstyöpaikatvoivatvaikuttaa,jasitenedistäätyöntekijöidenhyvinvointiajamyötävaikuttaamyöstyöurienpidentymiseen.

15

1 intRoduCtion

Promotingemployeehealthandwell-beingisessentialinourever-changingworklifeinwhich,forexample,agrowthofinformationtechnologyisproducingnewrequirementsforemployees.Therefore,employeesareexpectedtobehealthy,productive,proactive,energized,highlycommitted,andhaveagoodworkabilityinordertoperformefficiently(e.g.,Leiter&Bakker,2010).Thus,bothhealth-relatedandaffectivequalificationsofemployeesarerequired:beinghealthyandabletoworkisnotsufficient.Inaddition,highlevelsofmotivationanden-gagementareneeded.However,simultaneously,everyyearover20 000peopleinFinlandretireondisabilitypensions,mostoftenbecauseofmentaldisordersanddiseasesofthemusculoskeletalsystems(FinnishCentreforPensions,2013).Thus,evidently,understandingandenhanc-ingemployeewell-beinginitsentiretyisessential:howtobuildahealthyworklifethatallowsemployeestousetheirfullcapabilitiesatworkandsimultaneouslyenjoygoodhealthandwell-being?

Traditionally,employeewell-beinghasbeenviewedfromanega-tiveperspective,thatis,asoccupationalstress,disease,mentalillness,weakness,andsymptoms.Thus,understandingthecausesandconse-quencesofill-healthwasthemainfocusofoccupationalhealthstudiesthroughoutthedecades(e.g.,Cooper,1998;Cooper&Marshall,1976;Dewe,O’Driscoll,&Cooper,2010;Karasek&Theorell,1990;Lazarus&Folkman,1984).Fromthebeginningofthemillennium,theriseofpositivepsychology(Seligman&Csikszentmihalyi,2000;Snyder&Lopez,2002)haspromptedinterestinpositiveaspectsofwell-being.Thefocushasthereforegraduallyshiftedfromstressandillnesstosocialpsychologicalprocessesthatcontributetoandenhancewell-being.Inparticular,thischangehasresultedinarapidincreaseinstudiesofpositive

16

1 introdUction

organizationalbehaviour,suchasworkengagement(Demerouti,Bakker,Nachreiner,&Schaufeli,2001;Schaufeli,Salanova,González-Romá,&Bakker,2002),andpsychologicalcapital(Luthans,Youssef,&Avolio,2007).Theincreasingamountofresearchonworkengagementandotherpositiveaffectivewell-beingstatesisnotsurprisingasthebenefitsofsuchaffectivestatesonbothemployeesandorganizationsarewidelyacknowledged.Indeed,previousstudiesindicatethatengagedemployeesperformbetter,areproactiveandmorecommittedtotheirjobs,havefewersicknessabsences,andaremorewillingtocontinuetheirjob(forareview,seeHalbesleben,2010).Thus,inordertobenefitfromthefavourableconsequences,itiscrucialtounderstandthedrivers(e.g.,jobresourcesandpersonalresources)ofsuchpositiveaffectivework-relatedstates.

Theideaofpositivepsychologywasnottoreplacetraditionalviewofpsychologyfocusingonill-healthandhealthproblems,rather,itwastocomplementandenlargethefield.Therefore,currentresearchonso-cialpsychologicalhealthstudiestriestofindabalancebetweenpositiveandnegativeaspectsofemployeewell-being,andaimstounderstandtheentirerangeofwell-being,notjustthepositiveornegativesideofit(e.g.,Schaufelietal.,2002;Schaufeli&Bakker,2004).Overall,inhealthpsychology,healthisunderstoodasaproductofnotonlybiologicalprocesses,butalsoofpsychological,behavioral,organizational,andsocialprocesses(e.g.,Engel,1977;Houdmont&Leka,2010;Odgen,2012).Similarly,thisstudyaimstoframeacomprehensiveviewofemployeewell-being,intakinganindividualandorganizationalperspective,andinexaminingbothpositivewell-being(e.g.,workengagement,workabilityandlifesatisfaction)aswellasnegativeaspectsofhealth,suchassymptomsofdepressionandmusculoskeletalpain.Inparticular,thefocusofthisstudyisonthepsychosocialprocessesrelatedtoemployeewell-being,specially,onthemechanismslinkingmotivational(i.e.,workengagement),situational(i.e.,jobresourcesandjobdemands),andindi-vidual(i.e.,personalresourcesandlifestyle)characteristicsonemployeewell-being.Thesemechanismslinkingmotivational,situationalandindividualcharacteristicsonwell-beingwereexaminedusingdatafroma13-yearfollow-upperiodonasampleofFinnishfirefighters.

Apparently,firefightersexperiencebothpositiveandnegativeaspectsofwell-being.Workasafirefightercanbehighlymotivatingasthejob

17

1 introdUction

includesawidevarietyoftaskswhichenablestheuseofone’sknowledgeandcompetence,aswellassupportiveteamwork.However,atthesametime,firefightersarefoundtohaveseveralhealthproblems,suchasmus-culoskeletaldisorders,sleepingproblems,post-traumaticstressdisorder,depressionandevenworkdisability(e.g.,Fullerton,Ursano,&Wang,2004;Haslam&Mallon,2003;Punakallio,Lusa,Luukkonen,Airila,&Leino-Arjas,2014;Regehr,Hill,Knott,&Sault,2003).Itisclearthathealth-protectingjobresourcesandpersonalresources,aswellasjobdemandsthatmayreducewell-being,formpartofthejoboffirefighters.However,sofar,previousstudiesonfirefighters’well-beinghavemainlyfocusedonthenegativeside,suchasriskfactorsofthejob,riskbehavior,andhealthproblems,insteadofpositiveaspectsofwell-being.

Thus,thisstudyaimstofillinthegapsincurrentliteraturebypro-vidingacomprehensiveperspectiveoffirefighters’well-beingthroughinvestigatingboththepositiveandnegativeaspectsofwell-beingandtherelateddemandsandresourcesbyusingalongitudinaldata.Moreover,byapplyingaperson-centeredapproachinexaminingemployeewell-beinginthelongtermthisstudybringsnovelinformationaboutthedevelopmentalpathsatindividuallevelaboutpositiveandnegativeaf-fectivestates,aswellaswithregardtomultisitemusculoskeletalpainanddepressivesymptoms.Theoretically,thisstudyaimstoexpandonprevi-ousresearchbyfocusingonhowmotivationalcharacteristics(i.e.,workengagement)maypredicthealth-relatedwell-being(i.e.,workability),andhowworkabilityitselfmaybeconsideredahealth-relatedresourcethatmayhaveabeneficialeffectonwell-beinginthelongterm.Despiteusingadataofasingleoccupation,thisstudyhasastrongtheoreticalbasis,anditisthereforeassumedthatthemodelsandresultscanlargelybegeneralizedtootheroccupationsaswell.

1.1 Conceptualization of employee well-being

Awidevarietyofconstructshavebeenusedtodescribewell-being(seeforexample,Ryan&Deci,2001).Asoutlined,inoccupationalhealthpsychology,well-beinghasfrequentlybeendistinguishedbypositiveandnegativeperspectives(e.g.,Mäkikangas,Schaufeli,Tolvanen,&Feldt,2013;Schaufelietal.,2002).Inaddition,conceptssuchasaffective

18

1 introdUction

well-being(Warr,1990,2007),subjectivewell-being(Diener,1984,2000),andpsychologicalwell-being(Ryff&Keyes,1995)havebeenapplied.Inaddition,adistinctionbetweenjob-relatedandnon-jobre-latedwell-beinghasbeenmade(Warr,1990).Comparedtothemodelofaffectivewell-being(Warr,1990)–whichisbasedonthedimensionsofpleasureandarousal–subjectivewell-being(Diener,1984,2000)andpsychologicalwell-being(Ryff&Keyes,1995)offerasomewhatmorecomprehensiveviewofwell-being.First,subjectivewell-beingreferstheextenttowhichpeoplefeelandthinkthattheirlifeisgoingwell.Assuch,subjectivewell-beingincludesbothaffectivedimensions(i.e.,moodsandemotions)aswellasacognitiveevaluationofone’ssatisfactioningeneral(i.e.,lifesatisfaction)(e.g.,Diener,2000;Diener,Suh,Lucas,&Smith,1999).Psychologicalwell-being,ontheotherhand,includesnotonlyaffectiveandcognitivecomponents,butalsobehaviouralandmotivationalaspectsofwell-being(Ryff&Keyes,1995).Tosumup,severalconstructions–partlyoverlapping–ofwell-beingexist,andatthemomentnosingleagreed-upondefinitionofemployeewell-beingexists(Salanova,DelLíbano,Llorens,&Schaufeli,2014).

Inthisstudy,employeewell-beingisunderstoodasamultidimen-sionalconstructcoveringaffective,cognitive,andhealth-relatedaspectsofwell-being.ThisstudycombineselementsfrommodelsbyWarr(1990),Diener(1984,2000),andRyffandKeyes(1995),withhealth-relatedviewsaboutwell-being.Assuch,threedimensionsofwell-beingarestudied:affective well-being(i.e.,workengagement,andpositiveandnegativeaffectivestates),health-related well-being(i.e.,workabil-ity,self-ratedhealth,andmultisitemusculoskeletalpain),andcognitivecomponentofsubjectivewell-being(i.e.,lifesatisfaction).Underthisclassification,bothpositiveandnegativesidesaswellasjob-relatedandcontext-freeindicatorsofwell-beingareacknowledged.Inthatsense,thisstudyisfollowingthelineofoccupationalhealthpsychologyresearchthatemphasizestheholisticviewofemployeewell-being(e.g.,Schaufelietal.,2002;Schaufeli&Bakker,2004;vanHorn,Taris,Schaufeli,&Schreurs,2004).Next,thesethreedimensions–affective,health-related,andcognitive–arepresentedindetail.

19

1 introdUction

1.1.1 Affective well-being

Asatheoreticalapproachtoaffectivewell-being,atypologyofaffectivewell-beingbyPeterWarr(1990)hasbeenapplied.InWarr’s(1990)typology,affectivewell-beingiscomprisedoftwoorthogonaldimen-sions:pleasureandarousal(seealsoRussell,1980).Inhismodel,anylevelofpleasuremaybeaccompaniedbyhighorlowlevelsofarousal,andsimilarly,aparticularlevelofarousalmaybeeitherpleasurableorunpleasurable.Tomeasureaffectivewell-being,threeaxesareused:discontented-contented,anxiety-comfort,anddepression-enthusiasm.Thus,themodelacknowledgesbothpositiveandnegativeaspectsofwell-beingwhichbothareofinterestinthecurrentstudy.FromWarr’smodel,thefocusofthisstudyisinthethirdaxis,i.e.,depression-enthusiasm,inwhichpositivefeelingssuchasenthusiasmandpositivemotivationaresituatedoppositefeelingsofdepressionandsadness.Moreprecisely,fouraffectivewell-beingindicatorsareexaminedthatcanbeincludedinthedepression-enthusiasmaxisofWarr’stypology(1990):workengagement,vigor,fatigue,anddepressivesymptoms,respectively.IntermsofWarr’smodel,work engagementandvigorrefertoenergeticandactivatedstates,andpositivepleasurefilledwithenthusiasm,andthus,theycanbelocatedatthepositiveendofthedepression-enthusiasmaxis.Whereas,feelingsoffatigueanddepressive symptomsarelocatedatthenegativepoleofthesameaxis.Hereafter,theconceptualizationoffouraffectivewell-beingindicatorsofthisstudywillbepresentedinmoredetail.

Work engagement

Workengagementhasemergedasanaffectivewell-beingconstructtomeasurethepositivework-relatedstateofmind(e.g.,Bakker,Schaufeli,Leiter,&Taris,2008).Researchonengagementatworkhasgrownconsiderablyduringthelast20years;however,acomprehensiveandcommonunderstandingofworkengagementasanacademicconceptdoesnotyetexit.Instead,therearediverse,andpartlycontradictory,conceptualizationsofthework-relatedengagement.

Personal engagement.ThefirstconceptualizationofengagementatworkwasmadebyKahn(1990)whodefinedpersonalengagementas“harnessingoforganizationmembers’selvestotheirworkroles:inen-

20

1 introdUction

gagement,peopleemployandexpressthemselvesphysically,cognitively,emotionallyandmentallyduringroleperformances(p. 694)”.Assuch,engagementissimultaneousself-expressionandself-employmentbyemployeesinroleperformances,thatis,employeesbothdrivepersonalenergiesintoworkroles(i.e.,self-employment),anddisplayselfwithintherole(i.e.,self-expression).Inotherwords,engagedemployeesbringtheirpersonalselvesintotheirworkastheyidentifywithit.Despitepresentingacomprehensivetheoreticalmodelofengagement,Kahnhasnotproposedanoperationalizationofthemodel(Schaufelietal.,2002).Inaddition,onlyafewempiricalstudiesbasedontheKahn’smodelexist(e.g.,May,Gilson,&Harter,2004;Rothbard,2001;Saks,2006).

Job engagement.AccordingtoMaslachandLeiter(1997)–thepio-neersofburnoutresearch–jobengagementcanbecharacterizedasapositiveantithesisofburnout.Tothem,engagementischaracterizedbyenergy,involvementandefficacy,whicharethedirectoppositesofthethreeburnoutdimensions(i.e.,exhaustion,cynicism,andreducedpro-fessionalefficacy).Therefore,engagementandburnoutareconsideredtoexistonthesameunderlyingcontinuum,withjobengagementatoneendandburnoutattheother.Assuch,thisconceptualizationlimitstheopportunitytoexamineengagementinitsownright(seeSeligman,2002;Seligman&Csikszentmihalyi,2000)asanindependent,distinctconceptofburnout.Inaddition,definingengagementandburnoutasdirectoppositesdisablestheinvestigationoftherelationshipbetweenthesetwoconcepts(Schaufelietal.,2002).

Work engagement.Athird–andthemostcitedandstudied–defini-tionofengagementconsidersworkengagementasindependentposi-tiveconstruct,althoughnegativelyrelatedtoburnout(Schaufelietal.,2002).Consequently,workengagementisdefinedasapositive,fulfilling,affective-motivationalstateofwork-relatedwell-being,characterisedbyvigor,dedication,andabsorption(Schaufelietal.,2002).Vigorreferstohighlevelsofenergyandmentalresiliencewhileworkingandthewillingnesstoinvesteffortinone’swork,andpersistenceeveninthefaceofdifficulties.Dedicationischaracterizedbybeingstronglyinvolvedinone’swork,andexperiencingasenseofsignificance,enthusiasm,inspira-tion,pride,andchallenge.Absorptionreferstobeingfullyconcentratedandhappilyengrossedinone’swork,wherebyhavingasenseoftimepassingquicklyanddifficultiesdetachingoneselffromwork.Assuch,

21

1 introdUction

workengagementisamultidimensionalconstructincludingbehavioural-energetic(vigor),emotional(dedication),andcognitive(absorption)components.Mostscholarsagreethatengagementincludesdimensionsofvigoranddedication,however,thesignificanceofthethirdindicatorofworkengagement(i.e.,absorption)islessunanimous(Bakkeretal.,2008).Inaddition,thedistinctionofworkengagementfromburnoutissupportedbyavastamountofempiricalevidence(forameta-analysis,seeHalbesleben,2010).Duetotheindependenceofworkengagement,itispossible,tosomeextent,tobeengagedandburn-outsimultane-ously,thatis,tohavebothpositiveandnegativeaffectivestatesatthesametime.Similarly,researchhasshownthatworkengagementisin-dependentofseveralrelatedconstructs,suchasflow(Csikszentmihalyi,1990;Schaufeli&Bakker,2010),jobinvolvement(Lodahl&Kejner,1965;Salanova,Agut,&Peiró,2005),jobsatisfaction(e.g.,Hackman&Oldman,1976;Hertzberg,1959;Schaufeli&Bakker,2010),andworkaholism(e.g.,Mäkikangasetal.,2013;Schaufeli,Taris,&Bakker,2008).Thus,despiteapartialoverlap,theconceptofworkengagementhasanaddedvalueovertheserelatedconceptsandmustthereforebedistinguishedfromthem(seealsoSchaufeli&Bakker,2010).

Previousstudiesonworkengagementhaveconsistentlyshownthatjobresources(e.g.,autonomy,socialsupport,skillvariety,andperform-ancefeedback)andpersonalresources(e.g.,self-efficacy,optimism,andself-esteem)facilitateengagement(e.g.,Christian,Garza,&Slaughter,2011;Crawford,LePine,&Rich,2010;Halbesleben,2010;Simpson,2009;Xanthopoulou,Bakker,Demerouti,&Schaufeli,2007),andconsequently,engagementisfoundtoberelatedtopositiveoutcomesatwork,includingorganizationalcommitment(e.g.,Hakanen,Bakker,&Schaufeli,2006),jobperformance(Salanovaetal.,2005),andwork-unitinnovativeness(Hakanen,Perhoniemi,&Toppinen-Tanner,2008a).However,evidenceoftherelationshipbetweenengagementandhealthisstillscarce(Bakkeretal.,2008;Bakker&Leiter,2010),andtherefore,thisstudyaimstoenlargecurrentknowledgebyexaminingtherelation-shipbetweenengagementandhealth-relatedwell-being,i.e.,workability.

22

1 introdUction

Positive and negative affective states

Asoutlined,workengagementreferstoapositivework-relatedaffectivestate.However,affectivestatescanalsobeconceptualizedascontext-freedispositionalconstructs,suchaspositiveandnegativeaffect(e.g.,Watson,Clark,&Tellegen,1988;Weiss&Cropanzano,1996).Bydefinition,affectcanbedistinguishedbymoodsandemotions.Moodsareregardedasmorestableaffectivestatesthanemotions,whichareconsideredmoreintenseandshort-lived(Brief&Weiss,2002;Cropanzano,Weiss,Hale,&Reb,2003).

Positiveaffectreferstofeelingslikeenergetic,active,alert,inspired,andenthusiastic,whichare,infact,closelyrelatedtothevigordimensionofworkengagement.Incontrast,negativeaffectreferstoaffectivestatessuchasbeingdistressed,upset,nervous,orafraid.(Watsonetal.,1988.)Bothpositiveandnegativeaffectscanbeconceptualizedasstateortrait:withstateaffectreferringtoone’sfeelingsatagivenpointintime,whereastraitrepresentsmorestableindividualdifferencesinthelevelofaffect(e.g.,Pressman&Cohen,2005).Severalscalestomeasurepositiveandnegativeaffectivestatesexist;thePositiveandNegativeAffectSchedule(PANAS;Watsonetal.,1988),andtheProfileofMoodStates(POMS;McNair,Lorr,&Droppelman,1971)amongothers.

Thereisnoconsistentviewastowhetherpositiveandnegativeaffectarebipolaropposites(e.g.,Russell&Carroll,1999)orindependentofeachother(e.g.,Watson&Tellegen,1999).Thediscussionhassimi-laritieswiththeaforementioneddiscussionbetweenworkengagementandburnout(e.g.,Maslach&Leiter,1997;Schaufelietal.,2002).Theindependenceofpositiveandnegativeaffectwouldmeanthatanincreaseinonedoesnotnecessitateadecreaseintheother(Cropanzanoetal.,2003),andthus,peoplecouldhavebothpositiveandnegativefeelingsatthesametime(e.g.,Cacioppo&Berntson,1994;Larsen&McGraw,2011;Larsen,McGraw,&Cacioppo,2001).Incontrast,thebipolarviewassumesthatpositiveandnegativeaffectareoppositeendsofabipolarcontinuum,andpeoplecannothavepositiveandnegativefeelingssimultaneously(e.g.,Green,Goldman,&Salovey,1993;Rus-sell&Carroll,1999).Inthisstudy,thefocusisonthreeaffectivestateswhichrepresentbothpositive(i.e.,vigor)andnegative(i.e.,fatigueanddepression)polesofaffect.Moreover,boththeindependenceandbipolarviewoftheaffectivestateswillbeexamined.

23

1 introdUction

Mostoftheconceptualizationsofvigorsharetheideathatvigorisapositiveaffectivestatethatischaracterizedbyhighlevelsofenergyandvitality.Forexample,RyanandFrederick(1997)havedefinedvigorasanindividual’ssubjectivefeelingofenergyandliveliness.Intheworkcontext,vigorhasbeendefinedasasetofinterrelatedaffectivestatesexperiencedatwork,andassuch,ischaracterizedbyfeelingsofphysicalstrength,emotionalenergy,andcognitiveliveliness(Shirom,2010).Asalreadynoted,vigorhasalsobeendefinedasacorecomponentofworkengagement(Schaufeli&Bakker,2010).Hence,vigorisdefinedashighlevelsofenergy,mentalresilience,andwillingnesstoinvesteffortinone’swork.Inthisstudy,vigorisviewedasapositiveaffectivestatecharacter-izedbye.g.,activity,energy,cheerfulness,andliveliness.Hence,vigorcanbesituatedatthepositivepoleofdepression-enthusiasmaxisinWarr’s(1990)affectivewell-beingmodel.Previousstudies,albeitthusfarmainlybasedoncross-sectionaldata,suggestthatvigorispositivelyrelatedtowell-being.Forexample,vitality–onecomponentofvigor–wasfoundtoberelatedtophysicalandpsychologicalhealth(Ryan&Frederick,1997).Moreover,Shirom,Toker,Berliner,Shapira,andMelamed(2006)foundthathighlevelsofvigorwerenegativelycorrelatedwithseveralinflammationbiomarkers.However,thereisapaucityoflongitudinalresearchinvestigatingtheeffectofvigoronemployeewell-being,andinparticular,thepossibleco-occurrencewithfatigue,thus,thisstudyaimstoanswertheseshortcomingsidentifiedinpreviousresearch.

Fatigue−anaspectofnegativeaffect–isrelatedtoaperson’senergylevelinasimilarwaytovigor.Intheworkcontext,fatiguecanbecon-ceptualizedasanexhaustioncomponentofburnout(Maslach&Leiter,1997).Exhaustionisdefinedasthedrainingofemotionalresourcesandfeelingsofchronicfatigue(Maslach&Leiter,1997),andassuch,itdi-rectlyreferstoaperson’sreducedenergylevel(seealsoSchaufeli&Buunk,2003).FollowingWarr’s(1990)typology,fatiguecanbesituatedatthenegativepoleofthedepression-enthusiasmaxis.Inthisstudy,fatigueisviewedasanegativeaffectivestatethatischaracterizedbyreducedemotionalresources.Previousevidenceshowsthatfatigueisrelatedtowork-relatedstressreactions,psychosocialoverload,subjectivehealthcomplaints(Sluiter,deCroon,Meijman,&Frings-Dresen,2003),andevendecreasedworkability(deCroonetal.,2005).Similarly,exhaustion–asevereformoffatigue–hasbeenfoundtoberelatedtolifesatisfac-

24

1 introdUction

tion(e.g.,Demerouti,Bakker,&Schaufeli,2005),andself-ratedhealth(Hakanenetal.,2006).However,thejointeffectsofaffectivestates(i.e.,fatigueandvigor)arelargelyunknown,aswellastheirrelationshiptoemployeewell-beinginthelongterm.Therefore,thisstudycontributestocurrentknowledgebyexaminingthelinksbetweentheseaffectivestatesandtheirrelationshiptowell-beingduringa13-yearperiod.

Similarly,depressioncanbedefinedasanegativeaffectivestate.Assuch,depressionischaracterizede.g.,bysadness,irritability,hopeless-ness,feelingsoffailure,andsocialwithdrawal(e.g.,Beck,Steer,&Gar-bin,1988).Hence,depressionreferstoawayofreactingtochallengesthatareperceivedasimpossible,amelancholicmoodortoaclinicalmentaldisorder(Gruenberg&Goldstein,2003).FollowingWarr’s(1990)typology,depressioncanbesituatedatthenegativepoleofthedepression-enthusiasmaxis.Inthisstudy,depressivesymptomssuchasfeelingmiserable,sad,hopeless,andlonelywereexamined.Consist-entevidenceindicatesthathighjobdemandsarerelatedtodepression(Bonde,2008;Netterstrømetal.,2008),andsimilarly,theprotectiveeffectsofwork-relatedpsychosocialfactors,suchassocialsupport,arelargelyevident(forreviews,seeNetterstrømetal.,2008;Stansfeld&Candy,2006).Inaddition,lifestylefactorssuchasalcoholconsumption,exercise,andsleepingproblemshavebeenfoundtobeassociatedwithdepression(e.g.,Lopresti,Hood,&Drummond,2013).However,therelationshipbetweendepressivesymptomsandmusculoskeletalpainislargelyunknowndespitethefactthatbothcauseanenormousamountofhealth-relatedproblems.Hence,thisstudyaimstobroadencurrentknowledgebyexaminingthelongitudinaldevelopmentpathsofbothdepressivesymptomsandmusculoskeletalpain,andtheantecedentsrelatedtothesepaths.

1.1.2 Health-related well-being

Inadditiontoaffectivewell-being,thisstudyaimstoexaminehealth-relatedaspectsofemployeewell-being.Apparently,healthasaconstructiscomplex,andseveralconceptualizationsofhealthexist.Forexample,healthcanhavemedical,physical,mental,andsocialcomponents(e.g.,seeWordHealthOrganization,1986).Traditionally,biomedicalmodelsofhealthdefinehealthastheabsenceofdiseaseordisability,andfocuses

25

1 introdUction

onthecausesofdiseases,preventionandcure(seee.g.,Engel,1977;Larson,1999).Inaddition,healthmaybedefinedasanabsenceofphysicaldisability,psychologicaldisabilityorpain(Brown,Kazi,Spitz,Gertman,Fries,&Meehan,1984).Inthisstudy,healthisviewedfromabroadperspective,thatis,coveringbothill-healthandsymptomsofpain,physicalandmental(dis)ability,andsubject’sownevaluationsofhis/hermentalandphysicalstate,andthus,ashealth-relatedwell-beingindicators,work ability,self-rated health,andmultisite musculoskeletal painareinvestigated.

Work ability

InFinnishworklife,theconceptofworkabilityhasbeenusedforseveraldecades.However,thereisnoclearconsensusonwhatworkabilityisorhowitshouldbedefined.Differentpractitioners,suchashealthcareprofessionals,pensionandrehabilitationauthorities,andresearchersmayhavedifferentperceptionsanddefinitionsofworkability.Inaddition,theconceptualizationofworkabilityhaschangedasworklife,andthedemandsrelatedtoit,haschanged.(Gould,Ilmarinen,Järvisalo,&Kos-kinen,2008.)Previously,workabilitywassimplydefinedasabalancebetweenthedemandsofworkandtheresourcesoftheindividual,fol-lowingtheideaofastress-strainmodel(e.g.,Rohmert,1986).Therefore,Ilmarinen,TuomiandKlockars(1997)havepostulatedthatworkabilityreferstoemployee’sabilitytocarryouthis/herworkinrelationtothedemandsoftheworkandhis/herhealthandmentalresources.Assuch,impairedworkabilityisbelievedtoresultfromanimbalancebetweenjobdemandsandindividualresources.Later,amultidimensionalviewofworkabilityhasbeenadoptedinwhichjob-relatedresourcesandmicroandmacroenvironmentsoutsideworklifehavebeenincorporatedinadditiontothatofjobdemandsandindividualresources(e.g.,Ilmarinen,Tuomi,&Seitsamo,2005;Järvikoski,Härkäpää,&Mannila,2001).Twodifferentworkabilityconceptsapplyingamultidimensionalviewexist:amultidimensionalworkabilitymodel(Järvikoskietal.,2001),andaholisticworkabilitymodel(Ilmarinenetal.,2005).Inthisstudy,theholisticworkabilitymodelhasbeenappliedasitisthemostemployedframeworkforexaminingworkability(e.g.,vandenBerg,Elders,Zwart,&Burdorf,2009).

26

1 introdUction

Intheholisticworkabilitymodel,developedbyIlmarinenetal.(2005),workabilityisdescribedasamultidimensionalconstructconsist-ingofboththeresourcesoftheindividualandfactorsrelatedtoworkandworking,andtheenvironmentoutsideofwork.Thedimensionsofworkabilityareoftendepictedintheformofa“Workabilityhouse”(Ilmarinen,Gould,Järvikoski,&Järvisalo,2008).Thefirstthree“floors”ofthehouseareformedbyindividualresources:healthandfunctionalcapacity;knowledgeandskills;andvalues,attitudes,andmotivation.Thefourth“floor”isthatofwork,consistingofworkingconditions;workcontentanddemands;workcommunityandorganization;andsupervi-soryworkandmanagement.Inaddition,family,closecommunity,andsocietyareincludedaspartofthemacroenvironmentofworkability.Basically,all“floors”playaroleincreatingsignificantprerequisitesforworkability,although,thebalancebetweenindividualresourcesandworkisthecoreoftheholisticmodel.

Awiderangeofpredictorsandconsequencesofworkabilityhasbeenexamined.Inparticular,individualcharacteristicsandwork-relatedriskfactorshavebeenextensivelyexaminedasantecedentsofworkability(forareview,seevandenBergetal.,2009).Forexample,individuallifestylefactors1suchasalcoholconsumption,exercise,andsleep(e.g.,Ilmarinenetal.,1997;Pohjonen,2001),andwork-relatedfactorssuchasmentalandphysicalworkdemands,andmanagement(e.g.,Pohjonen,2001;Tuomietal.,1991)havebeenfoundtobeassociatedwithworkability.However,sofar,themotivationalaspectsofhumanresourceshavenotbeenexaminedwiththesameintensityasbiographical,lifestyle,orwork-relatedfactors,despitethefactthataffective-motivationalfactors−suchasworkengagement−areconsideredessentialfactorsrelatedtoworkability(e.g.,Ilmarinen,2009).Moreover,health-relatedindicators,suchasworkability,maythemselvesbeimportantresourcesthatboostposi-tiveaffectivestates,suchasworkengagement,andconsequentlyfurtherimprovehealthandwell-being.Infact,TheWorldHealthOrganization(WHO)defineshealthasapositiveconstructincludingphysical,mentalandsocialwell-being,thatis,“aresourceforeverydaylife”ratherthan

1Aspoorhealthbehaviors(e.g.,smoking,alcoholconsumption,andlackofexercise)mayhaveserioushealthconsequences(e.g.,Odgen,2012),lifestylefactorswereincludedinthepresentstudy.However,theemphasisofthisstudywasonthepsychosocialprocessesthatlinkjobresources,personalresourcesandemployeewell-being.

27

1 introdUction

theobjectiveofliving(WHO,1986).Thus,healthcanbeconceptualizedasakindofcapitalinwhichindividualsmayinvestinordertoachievepositivefuturehealthoutcomes(Williamson&Carr,2009).Inasimilarvein,itcanbearguedthatworkabilityisahealth-relatedresourcethatislikelytoberelatedtofuturewell-being.Therefore,thisstudyaimstobroadencurrentknowledge,byfirstly,examiningtherelationshipbetweenworkengagementandworkability,andsecondly,investigatingworkabilityasahealth-relatedresource.

self-rated health

Similartothatofworkability,self-ratedhealthcanbeconceptualizedasanindicatorofhealth-relatedwell-being.Conceptually,self-ratedhealth(SRH)isdefinedasaperson’sglobalassessmentofhis/hergeneralstateofhealth(DeSalvo,Bloser,Reynolds,He,&Muntner,2006).Althoughreflectingindividual’sownperceptionofhis/hergeneralhealth,SRHcanberegardedasavalidindicatorofone’soverallhealthstatusbecauseseveralmeta-analyseshaveshownthatSRHispredictiveofmorbidityandvariousobjectivehealthindicators,evenmortality(e.g.,DeSalvoetal.,2006;Idler&Benyamini,1997).Assuch,self-ratedhealthisanimportantindicatorofwell-beingthatmayinfluenceindividual’squalityoflife,aswellasmotivationtowardwork,andtherefore,relevantfactortobetakenintoaccountinthecurrentworklife.

Infact,antecedentsofSRHhavebeenwidelyexamined.Forexample,evidenceonthedetrimentaleffectofnegativeaffect,suchasdepres-sivesymptoms,onself-ratedhealthexists(e.g.,Barger,2006;Idler&Benyamini,1997).Inasimilarway,psychosocialwork-relatedfactorssuchassocialsupport(Barger,2006;Idler&Benyamini,1997),aswellasjobvariety,jobcontrol,anddecisionauthority(Laaksonen,Rahkonen,Martikainen,&Lahelma,2006;Niedhammer&Chea,2003;Pikhartetal.,2001)havebeenfoundtobeassociatedwithgoodSRH.Neverthe-less,forexample,anunderstandingtherelationshipbetweenpositiveaffectandSRHisstilldeficient(e.g.,Shirom,Toker,Berliner,Shapira,&Melamed,2008).Moreover,thelongitudinalrelationshipbetweenthejoint-effectofaffectivestatesandSRHstillremainslargelyunknown.Therefore,thisstudyaimstofillinthesegapsintheliterature.

28

1 introdUction

musculoskeletal pain

Amongworkingpopulation,musculoskeletalpainisacommonhealthproblem,andinfact,musculoskeletaldisordersconstitutethemostfre-quentlyreportedhealthcomplaintinEurope(Hauke,Flintrop,Brun,&Rugulies,2011).Painintheback,shoulder,andneckarefoundtobethemostcommonmusculoskeletalpainsymptomsamongemployees(e.g.,Weevers,vanderBeek,Anema,vanderWal,&vanMechelen,2005),andcauseavastnumberofsicknessabsences(Munce,Stansfeld,Blackmore,&Stewart,2007),andmoredrastically,workdisability(Kamaleri,Natvig,Ihlebaek,&Bruusgaard,2009;Mirandaetal.,2010).Moreover,paininmultiplesiteshasbeenfoundtobeevenmoreharmfulthanpaininasinglesite(e.g.,Haukkaetal.,2013;Øverland,Harvey,Knudsen,Mykletun,&Hotopf,2012),andtherefore,understandingthemechanismscausingmultisitemusculoskeletalpainisessential.

Recently,thestudyonmultisitemusculoskeletalpain(MSP)anditsdeterminantshasgainedmoreattention,andseveralworkcharacteris-ticsandlifestylefactorshavebeenfoundtoberelatedtoMSP(forareviewseeHaukeetal.,2011;Shiri,Karppinen,Leino-Arjas,Solovieva,&Viikari-Juntura,2010).Forexample,lowsocialsupport,highjobdemands,lowjobcontrol,lowdecision-makingauthority,andlowskilldiscretionwererelatedtomusculoskeletaldisordersinameta-analysisincluding54longitudinalstudies(Haukeetal.,2011).Indeed,inseveralexplanatorymodelsofmusculoskeletalpainitisassumedthathighjobdemandsandlowsocialsupportleadtopsychosocialstresswhichinitsturnisrelatedtomusculoskeletaldisorders(e.g.,Bongers,deWinter,Kompier,&Hildebrandt,1993;Haukeetal.,2011;Melin&Lundberg,1997).Someevidenceindicatesthatmultisitemusculoskeletalpainanddepressivesymptomsmaydeveloptogether(e.g.,Bair,Robinson,Katon,&Kroenke,2003;Demyttenaereetal.,2007).However,painanddepres-siondonotalwaysco-exist(e.g.,Magni,Moreschi,Rigatti-Luchini,&Merskey,1994),andthus,theymightalsohavedifferentdevelopmentalpathsovertime.Nevertheless,thelongitudinalevidenceisstillscarce,andthus,inthisstudy,thelongitudinalrelationshipbetweenmultisitemusculoskeletalpainanddepressivesymptoms,andtherelateddemandsandresources,areexamined.

29

1 introdUction

1.1.3 Life satisfaction

Lifesatisfactionisdefinedasakeyindicatorofpositivesubjectivewell-being(Erdogan,Bauer,Truxillo,&Mansfield,2012;Pavot&Diener,2008)thatreferstoacognitive,globalevaluationofthequalityofone’slifeasawhole(McDowell,2010;Pavot&Diener,1993).Hence,itas-sessanindividual’sconsciousglobaljudgmentofhis/herlifebyusingtheperson’sowncriteria,thatis,individualsmayevaluateandemphasizethedifferentdomainsoflivedifferentially.Therefore,asacognitivecon-struct,lifesatisfactioncanbediscriminatedfromaffectivecomponentsofwell-being,suchaspleasantaffect.Moreover,affectivereactionsoftenrespondtoimmediatefactorsandthoseofashortduration,whereaslifesatisfactionmayreflectalong-termperspective.(Pavot&Diener,1993.)Nevertheless,bothcognitiveandaffectivecomponentsofwell-beingareneededtocaptureacomprehensivepictureofemployeewell-being.Therefore,besideaffectiveandhealth-relatedindicatorsofemployeewell-being,acognitiveindicatorofsubjectivewell-being(i.e.,lifesatisfaction)wasalsoincludedinthecurrentstudy.

Asacontext-freeindicatorofwell-being,lifesatisfactioncande-velopinanydomainoflife.Apparently,personalitytraits,culture,andlifecircumstancesarerelatedtolifesatisfaction,butalsopsychologicalprocessessuchasgoalsandcopingstrategiesmayinfluencelifesatisfac-tion(Dieneretal.,1999;Steel,Schmidt,&Shultz,2008).Inaddition,severalwork-relatedantecedentsoflifesatisfactionhavebeenidentified(forareview,seeErdoganetal.,2012).Moreover,ithasbeenshownrecentlythatpeoplereportmorelifesatisfactionincountrieswherepositiveemotionsarehighlyvalued(Bastian,Kuppens,DeRoover,&Diener,2014).Inaddition,lifesatisfactionisfoundtoberelatedtoseveralhealthrelatedoutcomes,suchasalowlevelofsleepcomplaints(Brand,Beck,Hatzinger,Harbaugh,Ruch,&Holsboer-Trachsler,2010),andreducedcardiovascularmortality(Chida&Steptoe,2008)whichalsoillustratestheimportanceofexamininglifesatisfactioninworklifecontext.However,suchevidenceismoreorlesslacking,asisalsothecaseforlongitudinalrelationshipsbetweenaffectivestatesandlifesatisfaction,andthus,thisstudyaimstoprovideanswersfortheseshortcomings.Morespecifically,thisstudycontributestocurrentliteraturebyexaminingtherelationshipsbetweendevelopmentalpathsofpositive(i.e.,vigor)andnegative(i.e.,fatigue)affectandlifesatisfactionovertime.

30

1 introdUction

1.2 theoretical frameworks for explaining employee well-being

Awidevarietyofwork-relatedfactors,psychosocialprocesses,sociocul-turalfactorsaswellasphysiologicalfactorsarefoundtoberelatedtoemployeewell-being.Theimportanceofworkingconditionsonemployeewell-beingisemphasizedintheleadingworkstressmodels,thatis,JobDemands-Control(JDC)model(Karasek,1979),andtheJobDemands-Resources(JD-R)model(Demeroutietal.,2001;Schaufeli&Bakker,2004).Inthisstudy,theJD-Rmodelisusedasthemainconceptualframeworkforexaminingemployeewell-beinganditspredictors.Assuch,theJD-Rmodelisadescriptivemodelthatspecifiestherelation-shipsbetweenworkcharacteristics(i.e.,jobdemandsandjobresources),personalresources,andemployeewell-being.However,itisnotsufficienttoexplaintheunderlyingpsychologicalprocessesbetweentheanteced-entsandwell-beingoutcomes.Thus,additionaltheoreticalframeworksareneeded.(Schaufeli&Taris,2014.)Inparticular,theConservationofResources(COR)theory(Hobfoll,1989,2001),andBroaden-and-Build(BaB)theoryofpositiveemotions(Fredrickson,2001)thathavefrequentlybeenappliedasexplanatorytheoreticalframeworksinthefieldofoccupationalhealthpsychologyoffermoreexactpredictionsforthelinksbetweenresourcesandwell-being.Inthisstudy,alongwiththeJD-Rmodel,CORandBaBtheorieshavealsobeenusedasmaintheoreticalframeworkstoguidetheoreticalreasoninginexplainingtherelationshipbetweenworkcharacteristics,personalresources,andemployeewell-being.Theseframeworksarepresentedinmoredetailinthefollowingsections.

1.2.1 Job demands-Resources model

JobDemands-Resources(JD-R)model(Demeroutietal.,2001)isaconceptualframeworkthathasoftenbeenusedtoinvestigatetheante-cedentsofemployeewell-being,suchasworkengagementandburnout.TheJD-Rmodelcanbecharacterizedasacomprehensivemodelthatsimultaneouslyaimstopredictbothill-healthandmotivationatwork.Rootsofthemodelcanbefoundinthebalancemodelsofjobstress,suchasJDCmodel(Karasek,1979)andEffort-RewardImbalance(ERI)model(Siegrist,1996).

31

1 introdUction

TheJD-Rmodelproposesthatdiverseworkcharacteristics,suchasjobdemandsandjobresources,haveeitherpositiveornegativeeffectsonwell-being.IntheJD-Rmodel,job demandsaredefinedas“thosephysical,social,ororganizationalaspectsofthejobthatrequiresus-tainedphysicalormentaleffortandarethereforeassociatedwithcertainphysiologicalandpsychologicalcosts”(Demeroutietal.,2001,501).Examplesofjobdemandsarephysicaldemands,timepressure,workoverload,androleconflict.Inthisstudy,physicalandmentaldemandsofthejobwereexamined.Inturn,job resourcesreferto“thosephysi-cal,psychological,social,ororganizationalaspectsofthejobthatmaydoanyofthefollowing:(a)befunctionalinachievingworkgoals;(b)reducejobdemandsandtheassociatedphysiologicalandpsychologicalcosts;(c)stimulatepersonalgrowthanddevelopment”(Demeroutietal.,2001,501).Examplesofjobresourcesareautonomy,socialsupportfromcolleaguesandsupervisor,skillutilization,andparticipatingindecisionmaking.Inthisstudy,supervisoryrelations,interpersonalrelations,andtaskresourceswereexaminedasindicatorsofjobresources.

ThebasicassumptionoftheJD-Rmodelisthattwodistinctpsycho-logicalprocesses−thehealth-impairmentprocessandthemotivationalprocess−aredifferentlyrelatedtowell-being.Firstly,thehealth-impair-ment processassumesthathighjobdemandsandlowjobresourcesleadtoburnout,andconsequentlytoill-health(e.g.,depression,cardiovasculardisease,andpsychosomaticcomplaints).Thus,burnoutactsasamediatorbetweenjobdemandsandill-health.Secondly,themotivational processassumesthatjobresourcesleadtoworkengagement,which,inturn,hasapositiveeffectonorganizationaloutcomes,suchascommitmentandperformance.Inthisprocess,workengagementactsasamediatorbetweenjobresourcesandpositiveorganizationaloutcomes.(e.g.,Bakker&Demerouti,2007;Bakkeretal.,2008.)

AccordingtothelaterformulationsoftheJD-Rmodel,personalresourceshavebeenaddedtothemodel,astheymayhavesimilarmotivationalpotentialtothatofjobresourcesandmaybepositivelyrelatedtoworkengagement,andconsequentlytopositivework-relatedoutcomes(Xanthopoulouetal.,2007).Bydefinition,personalresourcesarepositiveself-evaluationsthatarelinkedtoresiliency,andrefertoanindividual’ssenseofabilitytosuccessfullycontrolandimpactonhis/herenvironment(Hobfoll,Johnson,Ennis,&Jackson,2003).Inthisstudy,

32

1 introdUction

self-esteem(seeJanssen,Schaufeli,&Houkes,1999;Rosenberg,1965;Rosenberg,Schooler,Schoenbach,&Rosenberg,1995)andoptimism(seeScheier&Carver,1985,1992)havebeeninvestigatedasindicatorsofpersonalresources.

EmpiricalevidencelargelysupportstheassumptionsofthetworelatedprocessessuggestedintheJD-Rmodel(foroverviews,seeBakker&Demerouti,2007;Hakanen&Roodt,2010;Schaufeli&Taris,2014).Mostofthestudiesarebasedonacross-sectionaldesign;however,somelongitudinalevidenceoftheJD-Rmodelexists,aswell.Forexample,a3-yearfollow-upamongFinnishdentists(Hakanen,Schaufeli,&Ahola,2008b)supportedboththehealth-impairmentandthemotivationalprocess.Moreprecisely,jobdemandspredictedburnoutovertime,whichinturnpredictedfuturedepression;andjobresourceswererelatedtofutureworkengagementwhichinturn,predictedorganizationalcom-mitment.Inaddition,a1-yearfollow-upstudyamongDutchmanagersconfirmedbothprocessesproposedbythemodel(Schaufeli,Bakker,&VanRhenen,2009).

TheadvantageoftheJD-Rmodelistheoptiontosimultaneouslyinvestigatebothnegative(i.e.,health-impairment)andpositive(i.e.,motivational)processes.Inaddition,thewidevarietyofjobdemands,resources,andoutcomesincludedinthemodeladdsthemodel’sflex-ibility,thatis,themodeldoesnotrestrictitselftoaspecificjobdemandorjobresource,insteadanydemandandanyresourcemayaffectem-ployeewell-being.Consequently,thisincreasesthemodel’savailabilityandusabilityindiverseworkcontextsandorganizations.(Schaufeli&Taris,2014.)Hence,theJD-Rmodelcanbeappliedinvariousoccupa-tionalfieldsasthemodelconsidersthespecificdemandsandresourcesofeachjob.Indeed,themodelhasbeentestedamongawidevarietyofoccupationsindiversecountries.Forexample,evidenceexistsamongindustrialworkers(e.g.,Bakker,Demerouti,deBoer,&Schaufeli,2003;Schaufeli&Bakker,2004),whitecollarworkers(e.g.,Hakanenetal.,2006,2008a,2008b;Korunka,Kubicek,Schaufeli,&Hoonaaker,2009;Schaufeli&Bakker,2004),bluecollarworkers(e.g.,Hansez&Chmiel,2010;Hu,Schaufeli,&Taris,2011),andevenamongvolunteers(e.g.,Lewig,Xanthopoulou,Bakker,Dollard,&Metzer,2007),andfamily-ownedbusinessworkers(e.g.,Hu&Schaufeli,2011).Theevidenceindicatesthatthemodelcanbereliablyusedindiverseworkenviron-

33

1 introdUction

ments.However,studieshavemainlybeenconductedamongwhitecollarworkersinsteadofmorephysicallydemandingjobs,suchasfirefighting.Nevertheless,demandsandresourcesandtheirconsequencesalsoneedtobeexaminedinphysicallydemandingjobsinordertoexaminethemodel’susefulnessinsuchacontext.

OnelimitationoftheJD-Rmodelhasbeenitsneglecttoelucidatetherelationshipbetweenjobandpersonalresourcesandhealth-relatedoutcomes.Researchhasemphasizedtheeffectofworkengagementonorganizationalcommitmentandjobperformancealthoughworken-gagementcouldalsoberelatedtohealth-relatedoutcomes.Forexample,organizationaloutcomessuchascustomerloyalty(Salanovaetal.,2005),organizationalcommitment(Hakanenetal.,2008b),innovativeness(Hakanenetal.,2008a),andwork-familyenrichment(Hakanen,Peeters,&Perhoniemi,2011)havebeenexamined.Inotherwords,intheJD-Rmodel,health-relatedindicatorsareconsideredoutcomesofthehealth-impairmentprocessresultingfromhighlevelsofjobdemands,ratherthananoutcomeofthemotivationalprocess.However,someevidenceexistsoftherelationshipbetweenworkengagementandhealthoutcomes(Hakanen&Schaufeli2012,Seppäläetal.,2012).Nevertheless,thelon-gitudinalpathfromjobresourcesviaworkengagementtoworkabilityhasnotpreviouslybeeninvestigated.Moreover,health-relatedindicatorsmayalsobeimportantresourcesthatboostworkengagementandconsequentlyfurtherimproveemployeewell-being.Therefore,thisstudyexaminestheroleofworkabilityinthemotivationalprocessoftheJD-Rmodel.

1.2.2 Conservation of Resources theory

Hobfoll’s(2001)ConservationofResources(COR)theoryisastressandmotivationaltheorythatconsiderstheacquisitionandaccumulationofresourcesapivotaldrivetoinitiateandmaintainpeople’sbehaviour,andconsequently,enhancewell-being.ThebasictenetofCORtheoryisthatpeoplestrivetoobtain,retain,foster,protect,andbuildresourcesthattheyvalue.Resourcesaredefinedas“thoseentitiesthateitherarecentrallyvaluedintheirownright,oractasmeanstoobtaincentrallyvaluedends”(Hobfoll,2002,p.307).Fourbroadcategoriesofresourcesaredistinguished:objects(e.g.,home,food,tools,andsocioeconomicstatus);conditions(e.g.,jobresources,suchassocialsupportandjob

34

1 introdUction

control);personalcharacteristics(e.g.,personalresources,suchasskillsandself-efficacy);andenergies(e.g.,time,money,andknowledge).Practically,CORtheoryassumesthatpeoplestrivetomaintainandaccu-mulatediverseresources,andconsequently,theseresourcesmaypreventthemfromhealthproblems.Accordingly,stressislikelytooccurwhenresourcesarethreatenedorlost,orwhenindividualsinvestinresourceswithoutreceivingtheexpectedlevelofbenefits.Inotherwords,againinresourcesresultsinwell-being.(Hobfoll,2002,seealsoSalanova,Schaufeli,Xanthopoulou,&Bakker,2010.)

CORtheoryhastwokeyprincipals.Firstly,peopleneedtoinvesttheirresourcesinordertodealwithstressfulconditionsandavoidnegativeconsequences.Forexample,socialsupportfromcolleaguesasaresourcemaypreventstress,andconsequently,alsopreventnegativehealthout-comes.Therefore,thosewithgreaterresourcesarelessvulnerabletostressthanthosewithfewerresources.Secondly,peoplemustinvestresourcesinordertoprotectagainstfutureresourceloss,andgainnewresources.Accordingly,thosewithgreaterresourcesaremorecapableoffuturere-sourcegain,andthus,constituting“gainspirals”.Consequently,inthelongterm,theaccumulationofresourcesresultsinpositiveoutcomes,suchasbettercoping,adaptation,andwell-being.Onthecontrary,thosewholackresourcesaremorelikelytohavefurtherresourceloss,andconsequently,sufferfromhealthproblems.Thus,CORtheoryproposesthatresourcesandwell-beingaffecteachotherreciprocally(i.e.,gainorlossspirals).(Hobfoll,2002,seealsoSalanovaetal.,2010.)

Recently,CORtheoryhasfrequentlybeenappliedasatheoreticalframeworkinstudiesofemployeewell-being,andempiricalevidencelargelysupportstheassumptionsofthetheory.Forexample,ina3-yearlongitudinalstudyamongFinnishdentists(Hakanenetal.,2008a)positiveandreciprocalcross-laggedassociationswerefoundbetweenjobresourcesandworkengagement(i.e.,indicatorofemployeewell-being),andbetweenworkengagementandpersonalinitiative,whichinturnhadapositiveeffectonwork-unitinnovativenessovertime.Similarly,gainspiralsbetweenpersonalandjobresourcesandwell-beingwerefoundinalongitudinalstudyamongteachers(Salanova,Bakker,&Llorens,2006).Morerecently,furtherempiricalsupportforCORtheorywasfoundinathree-yearlongitudinalstudythatobservedgainspiralsbetweenjobandhomeresourcesandwell-being(Hakanenetal.,2011).

35

1 introdUction

Tosummarize,CORtheory,alongsidetheJD-Rmodel,assumesthathighlevelsofresourcescanbebeneficialforemployeewell-beinginthelongterm.TheresourcegainofCORtheorycanberelatedtomotivationalprocessoftheJD-Rmodelasbothemphasizethepositiveeffectofresourcesonwell-being.Similarly,theresourcelossofCORtheoryhassimilaritieswiththehealth-impairmentprocessoftheJD-Rmodelasbothresultinnegativehealthoutcomes.Therefore,inthisstudy,CORtheoryisappliedasatheoreticalframeworktoexaminewhetherjobresourcesandpersonalresources(i.e.,self-esteem)predictworkabilityinthelongterm.

1.2.3 broaden-and-build theory

TheBroaden-and-Build(BaB)theory(Fredrickson,2001)emphasizestheimportanceofpositiveemotionsinthedevelopmentofotherresources,andconsequently,positivehealthoutcomes.Assuch,BaBtheoryprovidesapossibletheoreticalexplanationforthemechanismthatlinkspositiveaffectivestatesandemployeewell-being.AccordingtoBaBtheory,posi-tiveemotionssuchashappiness,joy,andlove,broadenpeople’sthought-actionrepertoires,buildtheirenduringphysical,intellectual,social,andpsychologicalresources,andconsequentlyleadtobetterwell-being.Moreprecisely,thebroaden hypothesisassumesthatpositiveemotionsbroadenpeople’sawarenessandthinkingwhichhelpsthemtodrawonawiderrangeofnewideas.Forexample,joybroadenspeople’sthinkingandactionbyencouragingtheurgetoplayandbecreative.Overtime,thisbroadeningofthoughtsandattentionleadstodiscoveringandbuildingnewpersonalresources(i.e.,build hypothesis)and,consequently,toim-provedwell-being.Ontheotherhand,negativeemotionsnarrowpeople’smomentarythoughtsandactions.Inaddition,positiveemotionsareas-sumedto“correct”or“undo”theaftereffectsofnegativeemotions.Thetheorynotonlyassumesthatpositiveemotionssignalflourishing,butmoreimportantly,thatpositiveemotionsalsoproduceflourishing,eveninthelongterm.Hence,similartoCORtheory,BaBtheoryproposesthatemotionsandwell-beingaffecteachotherreciprocally(i.e.,gainspirals).

AwealthofresearchsupportstheproposalsoftheBaBtheory.Forexample,FredricksonandJoiner(2002)showedthatpositiveaffectandbroad-mindedcopingreciprocallyenhancedoneanother,initiating

36

1 introdUction

upwardspiralsofenhancedwell-being.Similarly,inanexperimentalstudyamongcollegestudents,higheractivationstateofamusementbroadened the scopeofattentionand though-actionrepertoires(Fredrickson&Branigan,2005),thussupportingtheassumptionsofBaBtheory.Inaddition,empiricalevidenceindicatesthatpositiveemotionsareassociatedwithfewersymptoms,lessillness,lesspainandinjury,andbetterhealth,eveninthelongterm(forreviewssee,Lyubomirsky,King,&Diener,2005;Pressman&Cohen,2005).Forexample,workengagement–asanaffectivewell-beingstate–hadapositiveeffectonlifesatisfactioninaseven-yearfollow-upstudyamongdentists(Hakanenetal.,2006).Similarly,vitality–anotherindicatorofaffectivewell-being–wasfoundtoberelatedtophysicalandpsychologicalhealth(Ryan&Frederick,1997).Takentogether,thesefindings,albeitthusfarmainlybasedoncross-sectionalstudies,suggestthatpositiveaffectivestatesarepositivelyrelatedtoemployeewell-being.Similarly,evidenceexistsonthedetrimentaleffectofnegativeaffectonwell-being(e.g.,Billings,Folkman,Acree,&Moskowitz,2000).Insummary,forthepurposesofthisstudy,BaBoffersaplausibleexplanationinexaminingtherelation-shipbetweenworkengagementandworkabilityaswellastheeffectsofpositiveaffectivestatesonemployeewell-being.

1.2.4 Comparison of the theoretical frameworks

Toconclude,theJD-Rmodelinparticular,CORtheoryandtosomeextentBaBtheoryallemphasizetheimportanceofresourcesinmaintain-ingandincreasingemployeewell-being,andthus,theyareappliedinthepresentstudy.However,somedifferencesintheconceptualizationofresourcesbetweenthemodelsexistandneedtobementioned.FortheJD-Rmodel,resourcesarerelatedtothespecificjobinquestion,whereasCORdefinesresourcesonagenerallevelcoveringresourcesfromobjecttoenergies,andfromconditionstopersonalcharacteristics.Inturn,BaBfocusesonpositiveemotionsasaresource,forthisreasontherefore,theJD-RmodelandCORaremorewidelyapplicable,andhavecon-sequentlybeenusedmoreextensivelyinthisstudyaswell.SimilarlytoCORandBaB,thereisanassumptionofthegainspiralsovertime,thatis,accumulationofresourcesresultsinresourcegainorupwardspiralsinlongterm,andconsequently,betterwell-being.Likewise,JD-Rmodeldiscussesthepositivelinksbetweenjobandpersonalresourcesandem-

37

1 introdUction

ployeewell-being.Therefore,forthepurposesofthisstudy,whichaimstoexaminethelong-termrelationshipsbetweenjobandpersonalresourcesandemployeewell-being,thesetheoreticalframeworksprovideimportantperspectivesonthesubject.Asstated,avastamountofempiricalevidenceofallthreemodelsexists.However,strictlyspeakingonlylimitedevidenceoflong-termrelationshipsbetweenresourcesandwell-beingexists(seealsoSalanovaetal.,2010),andthus,furtherresearchisneeded.

1.3 study context: Rescue services

Asindicated,theJD-Rmodel(Demeroutietal.,2001)proposesthatjobdemands,andjobandpersonalresourcesarerelatedemployeewell-being.Moreimportantly,thesedemandsandresourcesmaydifferbetweenoccupations,andthus,understandingtheworkcontextisimportantinexaminingemployeewell-being.ThecontextofthisstudyisrescueservicesinFinland,andinparticular,operationalfirefighterswithintherescueservices.Therefore,inordertounderstandtheparticularjobdemands,jobresources,andpersonalresources,andtheirrelationshipstowell-being,itisessentialtohaveanoverviewoftheoperationalfire-fighters’jobandtheirworkenvironment.

InFinland,thereare22rescueserviceregionsthatcarryoutthedutiesofrescueservicesintheirrespectiveregions.MunicipalitiesarejointlyresponsibleforrescueservicesinFinland.Inaddition,theMinistryoftheInteriordirectsandsteersrescueservicesandoverseestheircoverageandquality.Thereareapproximately360full-timefirebrigadeswitharound4,300permanentfull-timeemployeesofwhichapproximately3,500workinoperationalrescuetasksandmedicalfirstaidservices(e.g.,asfirefighters,sub-officers,andfirechiefs).Inaddition,volunteer,insti-tutional,industrial,andmilitaryfirebrigades(i.e.,contractfirebrigades)participateinperformingrescueservicesdutiesbyagreementbetweenthefirebrigadesandtheregionalrescueservices.Overall,inapproximately520firebrigades,thereareabout13,400personsemployedbycontractfirebrigades.(Pelastusopisto,2014.)

Workinrescueservicesincludesdiversetasksandworkenviron-ments.AccordingtotheRescueAct(379/2011),therescuedepartmentsareresponsibleforguidance,education,andprovisionofadviceaimedatpreventingfiresandotheraccidents,preparingforthecombating

38

1 introdUction

ofaccidentsandtakingappropriateactioninaccidentsanddangeroussituations,andinthelimitingoftheconsequencesofaccidents,andsupervisorydutiesofrescueservices.Inaddition,tasksinclude,forex-ample,warningthepopulationofaccidentsanddangeroussituationsaswellastaskscarriedoutaspartofrescueoperations.In2013,Finnishrescueserviceregionsconductedapproximately105000rescuetasks,comprisingof13500fires,16000otheraccidents(e.g.,trafficaccidents,tasksrelatedtohazardoussubstances),andaround75 500otherrescuetasks(suchastestingsmokealarms,rescuingpeopleandanimals,andfirstresponse).(Pelastusopisto,2014.)

Inaddition,firebrigadesareallowedtocarryouttasksbelongingtoemergencymedicalcareservicesinjoint-operationwiththemunicipalboardofthehospitaldistrict.Emergencymedicalcareservicesinclude,forexample,urgenttreatmentofpatientswhohavesufferedaninjuryorasuddenonsetofanillnessprimarilyoutsideofhealthcaretreatmentfacilities;transportofpatientstothetreatmentunitwiththemostap-propriatemedicalservices;andreferralofpatients,relativesofpatients,andotherindividualsinvolvedinincidentstopsychosocialsupportservices,wherenecessary(HealthCareAct1326/2010).In2013,firebrigadeshadnearly500 000emergencymedicalcaretasks.Atpresent,emergencymedicalcaretasksrepresentthemajorityofalltaskswithintherescueserviceregions.

Diversetasksfromfireandrescueservicestoemergencymedicalcaretaskssethighrequirementsforfirefighters.Indeed,firefightingisaphysically,mentally,andsociallydemandingjob.Smokediving,fireandrescueoperations,workingintheroofs,andcleaninguptheincidentsites(e.g.,infireandroadaccidents)arephysicallymostdemandingtasksoffirefighting(Lusa,Louhevaara,&Kinnunen,1994).Inaddition,goodphysicalhealthisneededinemergencymedicalcaretasks,e.g.,withpatienttransportation.Ontheotherhand,beingconstantlyalert,workinglonghoursinunpredictableworkenvironments,andoccasion-allyseveretimepressureincreasethementalworkloadoffirefighters.Moreover,goodmentalhealthisneededinmanagingademandingjobthatalsoincludesrescuinghumanlives,andevenfacingdeath.(e.g.,Beaton,Murphy,Johnson,Pike,&Corneil,1998;Haslam&Mallon,2003.)Socially,firefightingrequiresgoodcommunicationskills(e.g.,Beaton,Murphy,Pike,&Corneil,1997).Moreprecisely,workrequires

39

1 introdUction

constantinteractionwithpartners,suchascolleaguesandsupervisors,customers,otherpublicauthorities,andpublicmedia.Inparticular,workingina24-hourshiftincloselyco-ordinatedteamsnecessitatesgoodrelationshipsbetweentheco-workers.Practically,firefighterseat,sleep,andexercisetogether,andoftentheirsocialactivitiestendtooccurinthecompanyofotherfirefighters(Beatonetal.,1997;Mankkinen,2011).Inaddition,indangerousworksituationsfirefightershavetobeabletofullytrusttheirworkmates,andthus,goodsocialrelation-shipsbetweenco-workersareextremelyimportant.Furthermore,goodsupervisoryrelationsareessential,andareinfactfoundtoberelatedtoadecreasedlevelofmentalworkloadamongfirefighters(Lusa,Punakal-lio,Luukkonen,&Louhevaara,2006).Inadditiontorequirementsofgoodsocialrelationshipswithintheworkingcommunity,emotionallyhighlydemandingcustomers,forexampleangry,anxious,orimpatientpatients,ordemandingandthreateningpatientsandtheirrelatives,maybechallengingforfirefighters,andincreasethementalpressureofwork.

Apparently,highjobdemands–physical,mental,andsocial–arepartofthefirefighters’dailyworktasks.However,atthesametime,firefightingmayincludeseveraljobresourcesthathelptoreducejobdemandsandtherelatedphysiologicalandpsychologicalcosts,andstimulatepersonalgrowthanddevelopment(seeDemeroutietal.,2001).Jobresourcesoffirefighters’mayinclude,forexample,goodrelationshipsbetweenco-workers(e.g.,Pillai&Williams,2004),supervisorysupport(e.g.,Mitani,Fujita,Nakata,&Shirakawa,2006),andthefairdistributionofworktasks(e.g.,Lusaetal.,2006).Inaddition,personalresourcesmayactasapositiveresourceforfirefighters.Forinstance,someevidenceexistsofthepositiveeffectofself-esteemonfirefighters’well-beingandjobsatisfaction(Saijo,Ueno,&Hashimoto,2008).

Obviously,physical,mental,andsocialdemandsofthejobmayberelatedtofirefighters’well-being,asalsosuggestedbytheJD-Rmodel.Indeed,previousstudiesindicatethatfirefighterssufferfromseveralhealthproblems,including,forexample,sleepproblems(Haslam&Mallon,2003),hazardousdrinking(Boxer&Wild,1993),muscu-loskeletalproblems(Bos,Mol,Visser,&Frings-Dresen,2004;Sluiter&Frings-Dresen,2007),anddepressivesymptoms(Carey,Al-Zaiti,Dean,Sessanna,&Finnell,2011;Fullertonetal.,2004;Plat,Frings-Dresen,&Sluiter,2012).InFinland,about40%ofthepermanentoperative

40

1 introdUction

Finnishfirefightersarenotabletoworkuntiltheirnormalretirementage.In2000–2013,musculoskeletaldisorders(44%),cardiovasculardiseases(13%)andmentaldisorders(12%)werethemostcommonreasonsforearlyretirementamongFinnishfirefighters.(Fors,Keva,personalcommunication,2014.)Therefore,inordertomaintainandimprovehealthandwell-beingofFinnishfirefighters,itisessentialtounderstandthelongitudinalrelationshipsbetweenworkcharacteristics,personalresources,lifestyle,andwell-being.

1.4 Aims of the study

Theaimofthisstudywastoextendunderstandingoftherelationshipsbetweenworkcharacteristics(i.e.,jobdemandsandjobresources),per-sonalresources,andlifestyleonemployeewell-beingusingdatafroma13-yearfollow-upperiodonFinnishfirefighters.

Firstly,theaimwastoexaminetheroleofaffective-motivationalstateofworkengagementinrelationtoworkability.Secondly,thisstudyexam-inedtheroleofworkabilityintheJD-Rmodelbothasahealth-relatedresource(i.e.,predictor)thatislikelytoinfluenceworkengagement,andasanoutcomeofthemotivationalprocess.Thirdly,thedevelopmentalpathsinpositive(i.e.,vigor)andnegative(i.e.,fatigue)affectivestatesandtheirrelationshiptowell-beingwereinvestigated,aswellasintra-individualchangetrajectoriesinmultisitemusculoskeletalpainandde-pressivesymptomsandtherelatedantecedents(i.e.,workcharacteristics,personalresources,andlifestyle).Methodologically,thisstudyaimedtoextendthecurrentknowledgebyapplyingaperson-centeredapproach(i.e.,latentclassgrowthmodeling)alongwithavariable-basedapproach(e.g.,multivariateanalysisandstructuralequationmodeling).Hence,severalpredictors,indicatorsofemployeewell-being,andmethodologieshavebeenused.Theresearchquestionsbysub-studieswereasfollows:

1. Doesworkengagementrelatetoworkabilityevenafteradjustingforage,lifestylefactors,andworkcharacteristics?Areage,lifestylefactors,workcharacteristics,andparticularlyworkengagementassociatedwithdifferentdimensionsofworkability?(StudyI)

2. Doesworkengagementmediatetherelationshipbetweenjobre-sourcesandfutureworkability?Doesworkengagementmediate

41

1 introdUction

therelationshipbetweenpersonalresources(i.e.,self-esteem)andfutureworkability?Doesworkengagementmediatetheimpactofworkabilityin1999onworkabilityin2009?(StudyII)Thespecifichypotheseswere:

H1:Jobresourcesin1999willbepositivelyrelatedtowork-abilityin2009throughworkengagementin2009.

H2:Self-esteemin1999willbepositivelyrelatedtoworkabilityin2009throughworkengagementin2009.

H3:Workabilityin1999willbepositivelyrelatedtoworkengagementin2009,whichinitsturnwillbepositivelyrelatedtosubsequentworkabilityin2009.

3. Aretheredifferentdevelopmentalpathsbasedonpositive(vigor)andnegative(fatigue)affect?Dothesedevelopmentalpathschangeorremainstableovertime?Arethesedevelopmentaltrajectoriesdiffer-entlyassociatedwithwell-being(i.e.,workability,self-ratedhealth,andlifesatisfaction)duringthe13-yearfollow-upperiod?(StudyIII)

4. Aretheredifferentdevelopmentalpathsofmultisitemusculoskeletalpainanddepressivesymptomsduringthe13-yearfollow-upperiod?Dojobdemands,jobresources,personalresources,andlifestylepredictbelongingtoaparticularpain-depressiondevelopmentalpath?(StudyIV)Thespecifichypotheseswere:

H1a:Aonelargetrajectorywithlowlevelsofbothsymptoms(Low Symptoms)willemerge.

H1b:Asmallertrajectorywithahighlevelofdepressivesymptoms(High Depression)willemerge.

H1c:Asmallertrajectorywithahighlevelofmusculoskeletalpain(High Pain)willemerge.

H1d:Atrajectorygroupwithhighlevelsofbothsymptoms(High Symptoms)willemerge.

H2: Highlevelsofjobdemandsandlowlevelsofjobre-sourcesandoptimismatbaselinepredictbelongingtothetrajectorygroupswithhighlevelsofpainanddepressivesymptoms.

TheresearchframeworkandtherelationshipsbetweenthestudyvariablesarepresentedinFigure1.

42

1 introdUction

Figu

re 1

. The

res

earc

h fr

amew

ork

of t

he s

tudy

.

!"#$%&'(

)*+'&)%,'

-./012%&'()*+'&)%,'

3#$"*4"*0)#%&'

*"&%5)

#0'

674"*810)*/'

*"&%5)

#0'

9%0+'*"

0)7*2"0'

:45

;10;'

6"&<="0$"";

'

6$7,/'3'

6$7,/'3>'33'

6$7,/'3?'

6$7,/'333'

6$7,/'333'

6$7,/'333'

6$7,/'3?'

6$7,/'3?'

6$7,/'3?'

6$7,/'3?'

6$7,/'3?'

6$7,/'3?'

6$7,/'3'

6$7,/'3>'33'

6$7,/'3>'33'

6$7,/'33'

6$7,/'33'

@)*+'

"#A%A";"#$'

@)*+''

%B1&1$/'

9*%C"2$)*1"0'

)<'!

6-'%#,'

D-6'

6$7,/'33'

6$7,/'33'

E)B',";%#,0'

-"*0)#

%&'

*"0)7*2"0'

E)B'*"0)7*2"0'

6$7,/'33'

9*%C"2$)*1"0')<'

81A)*'%#,'<%5A7"'

6"&<=*%$",''

."%&$.'

F1<"'

0%50<%25)#

'

:G9H:!I6'

:G9H:!I6'

-JID

3H9:

J6'

F1<"0$/&"'<%2$)*0'

6$7,/'3'

6$7,/'33'

43

2 metHods

2.1 Participants

ThedataofthisstudyispartofaquestionnairestudyamongFinnishfirefightersconductedbytheFinnishInstituteofOccupationalHealthin1996(T1),1999(T2),and2009(T3).ThestudywasapprovedbytheEthicsCommitteeoftheHUSHospitalDistrict,andwasperformedaccordingtotheHelsinkiDeclaration.Eachsubjectgavewritteninformedconsentbeforeparticipation.ThequestionnairesareavailableinFinnish(Lusa-Moseretal.,1996;Punakallio&Lusa-Moser,1999;Punakallio&Lusa,2011).TheapprovalletterfromtheEthicsCommitteeandtheletterofintenttotheparticipantsareavailablefromtheauthoronrequest.

In1995,astratifiedsamplingofallpermanentFinnishoperativemalefirefighters(n=3512)wasconductedbasedonregistersofthethreetradeunionsofFinnishfirefighters.Thesamplingwasstratifiedaccordingtothenumberoffirefightersintheareaandtheirage.Atbaseline,in1996,aques-tionnairewassentto1,124professionaloperativefirefightersinFinland.Thesampleconsistedofoperatingpersonnelinrescueservices,whichismainlyresponsiblefordailyfirefighting,rescueandmedicalemergencies,andmedicalemergencyserviceoperations.Inthefirstphase,76%(N=849)returnedthequestionnaire.Oftherespondents,14wereexcludedbecauseofchangeofjob,sickleave,ornotbeingprofessionalfirefighter,thusthefinalsamplesizewas835.In1999,afollow-upquestionnairewassentto1,106participantsofthesampleselectedin1996whohadorhadnotresponded.Theresponseratewas72%(N=794),and67wereexcluded,sothefinalsamplewas727.In2009,1,061questionnairesweresenttofirefightersregardlessofpreviousresponsestatus,and68%returnedthequestionnaire(N=721),and270wereexcluded(finalsampleN=451).Figure2illustratestheprocessofdatacollection.

44

2 metHods

!"#$%&'&()*+,-.(

/ &'#'01'&()*+,2.(

/ 3$&/04'(5'65'7

'*8()*+-.(

!"#$%&'&()*+,9.(

/ #:0*4'(3;(<3=()*+>.(

/ 1?#@($'0A'()*+B.(

/ 1%=168%8'C(*38(

D53;'11?3*0$()*+>.(

E8506F'&(107

D$?*4()*+,,B9.(

!"#$%&'&()*+>G.(

/ &'#'01'&()*+9-.(

/ #:0*4'(3;(5'1?&'*#'(35(*3(

0&&5'11(0A0?$0=$'()*+,H.(

I%'163**0?5'1(1'*8()*+,,B9.(

J?*0$(107

D$'()*+-GH.((

!"#$%&'&()*+B9.(

/ #:0*4'(3;(<3=()*+,9.(

/ &?10=?$?8K(D'51?3*(35(1?#@(

$'0A'()*+GH.(

/ %*'7

D$3K'&()*+>.(

/ '7

D8K(L%'163**0?5'1(

)*+,B.(

!"#$%&'&()*+BM2.(

/ 3$&/04'(5'65'7

'*8()*+,BN.(

/ &?10=?$?8K(D'51?3*(35(1?#@(

$'0A'()*+-B.(

/ #:0*4'(3;(<3=()*+BN.(

/ 3O(&%8K()*+,2.(

/ %*'7

D$3K'&()*+B.(

/ '7

D8K(L%'163**0?5'1(

)*+,-.(

P'1D3*&'&()*+-9N.(

Q063*R?&'(3D'506A'(

F5'F4:8'51()*+GH,B.(

,NNN()SB.(

,NN>()S,.

(B22N()SG.(

J?*0$(107

D$'()*+9H,.((

J?*0$(107

D$'()*+MBM.((

P'1D3*&'&()*+MB,.(

P'1D3*&'&()*+MN9.(

I%'163**0?5'1(1'*8()*+,,2>.(

I%'163**0?5'1(1'*8()*+,2>,.(

T0801'8(B()*+G>2.(

/ 0#6A'$K(R35@?*4(

F5'F4:8'51(R:3(

:0A'(5'1D3*&'&(

?*(,NN>U(,NNNU(

0*&(B22N(

T0801'8(,()*+92G.(

/ 0#6A'$K(R35@?*4(

F5'F4:8'51(R:3(

:0A'(5'1D3*&'&(

?*(,NNNU(0*&(

B22N(

Figu

re 2

. Col

lect

ion

of t

he d

ata.

45

2 metHods

Twoslightlydifferentdatasetshavebeenusedinthefoursub-studies.Dataset 1consistedoffirefighterswhorespondedtothequestionnairesinboth1999(T2)andin2009(T3),andwerestillemployedintheirprofession(n=403).ThisdatasethasbeenusedinStudiesIandII.Dataset 2includedthosefirefighterswhorespondedtothequestionnairesateachmeasurementpoint(T1,T2,andT3),andwerestillemployedintheirprofession(n=360).ThisdatasethasbeenusedinStudiesIIIandIV.AssomeofthestudyvariablesweremeasuredonlyatT2(e.g.,self-esteem),thesetwodatasetswereused.ThecharacteristicsofthesetwodatasetsarepresentedinTable1.

OfDataset1,atotalof148oftherespondentsfrom1999didnotanswerin2009.Thedrop-outswereolder,hadlowereducation,smokedmoreoften,andhadlowerworkabilitythanthosewhorespondedonbothoccasions,indicatingslightlybetterlifestyleandworkabilityamongthesampleofthisstudy.However,whentheagedifferencewastakenintoaccounttherewerenodifferencesbetweenthetwogroups.Moreover,thedrop-outsandtheparticipantsdidnotsignificantlydif-ferinrelationtoself-esteemandjobresources.OfDataset2,atotalof475respondentsfrom1996didnotparticipatein2009.Statisticaltestsrevealedthatthedrop-outsweresignificantlyolder,hadlowereduca-tion,andslightlypoorerwell-beingintermsofworkability,depressivesymptoms,andself-ratedhealth.Inaddition,thedrop-outshadhighermentalworkload,poorerlifestylehabits,andloweroptimismthanthosewhorespondedatallthreetimes.Incontrast,thedrop-outsandthepar-ticipantsdidnotsignificantlydifferinrelationtolifesatisfaction,vigor,fatigue,musculoskeletalpain,physicalworkload,jobresources,andalcoholconsumption.Tosumup,thereweresomedifferencesbetweentheparticipantsandthedrop-outs,howeverthesewereforthemostpartminor.Theattritionanalysesaredescribedindetailintheoriginalarticles.

46

2 metHods

Table 1. Descriptive characteristics of the study samples.

Characteristics Dataset 1(Studies I–II)

Dataset 2(Studies III–IV)

Sample size 403 360

Average age at T3 48.5 (range 35–62, SD = 5.4)

48.7 (range 35–62, SD = 5.4)

Basic education

Primary or elementary school education 80.9% (n = 321) 81.0% (n = 288)

Secondary school education 19.1% (n = 76) 19.2% (n = 69)

Vocational education

Firefighter 87.7% (n = 315) 87.5% (n = 281)

Sub-officer 29.4% (n = 105) 30.4% (n = 97)

Fire chief 9.8% (n = 35) 9.4% (n = 30)

Mean work experience, in years 25.3 (SD = 5.8) 25.5 (SD = 5.7)

Shift work 84.3% (n = 337) 84.1% (n = 301)

2.2 measures

2.2.1 employee well-being

Affective well-being

Work engagement(StudiesI–II)wasmeasuredbyusingtheshortversionoftheUtrechtWorkEngagementScale,theUWES-9(Hakanen2002,2009;Schaufeli,Bakker,&Salanova,2006).TheUWES-9isvalidatedandthemostwidelyusedmeasureforworkengagement(Bakkeretal.,2008;Seppäläetal.,2009),consistingofnineitems.Themeasurehasbeentranslatedinto21languagesandusedamongvariousoccupationalgroups.Theinstrumenthasthreesub-scales:vigor(e.g.,“Atmywork,Ifeelburstingwithenergy”),dedication(e.g.,“Myjobinspiresme”),andabsorption(e.g.,“Iamimmersedinmywork”).Eachofthesub-

47

2 metHods

dimensionswasassessedusingthreeitems.Theitemswereratedonaseven-pointfrequency-basedscale(0=never,6=daily).Thescalewashighlyreliable(T3Cronbach’sα=0.95;forvigorα=0.89;dedicationα=0.90;absorptionα=0.90).

Vigor(StudyIII),fatigue(StudyIII)anddepressive symptoms(StudyIV)weremeasuredusingtheFinnishversionoftheProfileofMoodStates(POMS)(Hänninen,1989;McNairetal.,1971)including38itemsthatreflectbothpositiveandnegativeaffectivestates.Respondentswereaskedtoindicatehowthespecificitemsdescribetheirstateduringthelastweekratedonafive-pointfrequency-scale(0=not at all,4=very much).Theconstructvalidityofthescaleissatisfactory(Peterson&Seligman,2004).POMShasbeenwidelyusedinempiricalstudies,forexample,insportspsychologyandstudiesoncancerpatients(seeforexampleBeedie,Terry,&Lane,2000;Cella,Tross,Orav,Holland,Silberfarb,&Rafla,1989).POMShasalsobeenusedinsomeorganiza-tionalpsychologysettings(e.g.,Sonnentag&Niessen,2008).The vigor subscaleconsistedofsixitems:active,energetic,cheerful,vigorous,lively,andalert(T1α=0.94,T2α=0.95,T3α=0.94).The fatigue subscaleconsistedofthreeitems:exhausted,fatigued,andweary(T1α=0.82,T2α=0.86,T3α=0.87).Andfinally,the depression subscaleconsistedofsevenitems:miserable,sad,depressed,hopeless,blue,lonely,anddistressed(T1α=0.87,T2α=0.92,T3α=0.89).

Health-related well-being

Work ability(StudiesI–III)wasmeasuredbytheWorkAbilityIndex(WAI)questionnaire(Tuomi,Ilmarinen,Jahkola,Katajarinne,&Tulkki,1998).Theindexconsistsofsevendimensions,namely:(i)thesubjectiveestimationofcurrentworkabilitycomparedwithlifetimebest(0–10points);(ii)subjectiveworkabilityinrelationtojobdemands(2–10points);(iii)thenumberofcurrentdiseasesdiagnosedbyaphysician(1–7points);(iv)thesubjectiveestimationofworkimpairmentduetodiseases(1–6points);(v)sickleaveduringthepastyear(1–5points);(vi)ownprognosisofworkabilitytwoyearsfromnow(1,4or7points);and(vii)psychologicalresources(1–4points).TheWAIindexrangesfrom7to49points,andahigherscoreindicatesbetterworkability.TheWAIisthemostwidelyusedquestionnaireonworkabilityandavalidated

48

2 metHods

measureofworkability(vandenBergetal.,2009).Furthermore,satis-factorytest-retestreliabilityoftheindexhasbeenobserved(deZwart,Frings-Dresen,&vanDuivenbooden,2002).

InStudiesIandIII,thetotalindexofWAIwasused.However,duetothelargeamountofmissingdata(n=174–199)inestimatedworkimpairmentduetodiseases(sub-dimension4),thisitemwasexcludedfromalltheanalyses.ThereforetheslightlymodifiedWAIindexhadarangefrom6to43points(α=0.70–0.78,formoredetailseetheoriginalpapers).ThecontinuoussumscoreoftheWAIwasusedinStudiesIandIII.Inaddition,inStudyI,thesub-dimensionsoftheWAI–i.e.,theindividualitems-wereusedasseparatedependentvariables.

InStudyII,workabilitywasmeasuredbyasingle-itemoftheindex,namelythesubjectiveestimationofcurrentworkabilitycomparedwithlifetimebestwithascalefrom0to10:“Assumethatyourworkabilityatitsbesthashadavalueof10.Howmanypointswouldyougiveyourcurrentworkability?(0meansthatcurrentlyyoucannotworkatall)”.Thesingle-itemquestionofworkabilitywasusedbecausethetotalworkabilityindexcannotberegardedasafactorinwhichitssevendimensionswouldloadsimilarlyintoasinglefactor.ThereforetheindexcouldnotbeusedintheSEMmodel.Priorstudieshaveindicatedastrongassociation(r=.87)betweenthetotalWAI-scoreandthesingle-itemindicator(e.g.,Ahlstrom,Grimby-Ekman,Hagberg,&Dellve,2010).Inaddition,boththetotalWAIandthesingle-itemquestionhaveshownsimilarpatternsofassociationswithdiversehealth-relatedoutcomes(e.g.,Ahlstrometal.,2010).Thus,asingle-itemquestionofworkabilityisagoodalternativetothetotalWAI-index,andhasbeenwidelyusedinFinnishworklifeandhealthsurveys(e.g.,Kauppinenetal.,2010).

Self-rated health(StudyIII)wasmeasuredbyasingle-itemquestionbyasking“Howdoyourateyourhealthcomparedwithyouragepeers?”withafive-pointscale(1=muchworse,5=muchbetter).ThereliabilityoftheSRHquestionisfoundtobeasgoodasorevenbetterthanthatofmostofthemorespecificquestionsonhealth(Lundberg&Manderbacka,1996).Moreover,thequestioniscloselyrelatedtoobjectivemeasuresofhealthandevenmortality(Chida&Steptoe,2008).

Multisite musculoskeletal pain(MSP)(StudyIV)wasmeasuredusingsevenitemsadaptedfromavalidatedNordicMusculoskeletalQuestion-naire(Björkstén,Boquist,Talbäck,&Edling,1999;Kuorinkaetal.,

49

2 metHods

1987).Painwasmeasuredinneck,shoulders,forearms/hands,hips,knees,andradiating,andlocalpaininlowback.Thequestionwas:“Estimatehowmanydaysaltogetheryouhavehad…(i.e.,neck)painduringpreceding12months?”Allitemswereratedonafive-pointscale(1=never,2=1–7days,3=8-30days,4=over30days,5=daily).Thevariablewascategorizedintotwocategories:‘0=nopain’(painon0–7days),‘1=pain’(painmorethan7days).Finally,allsevendichotomizedvariablesweresummedandthesumscoreofMSPatallsiteswascalcu-lated(0=nopain,7=paininsevensites).

Life satisfaction

Lifesatisfaction(StudyIII)wasasingle-itemquestion“Howsatisfiedareyouatpresentwithyourlife?”.Theitemwasratedwithafive-pointscalefrom1(veryunsatisfied)to5(verysatisfied).Single-itemmeasuresoflifesatisfactionhavepreviouslybeeneffectivelyused(Lucas&Don-nellan,2012,seealsoWanous,Reichers,&Hudy,1997).

2.2.2 Job demands

Jobdemandsweremeasuredwithphysicalandmentalworkload(Stud-iesIandIV).Physical workloadwasmeasuredusingfouritemsadaptedfromViikari-Junturaetal.(1996).Physicalworkloadwascoveredbyquestionsonthefrequencyofworkinginfourdifficultworkpostures:(i)workingonone’sknees,crouched,orcrawling;(ii)posturesinwhichthebackisbent;(iii)posturesinwhichthebackistwisted;and(iv)workingwithahandorhandsaboveneck-shoulderlevel.Allitemswereratedonafour-pointscale(forexample,1=notatall,4=overanhourduringtheshift).(Fordatasets1/2:T1α=0.73/0.73,T2α=0.74/0.72,T3α=0.79/0.78).Mental work load2(e.g.,Tuomi,Toikkanenetal.,1991)consistedofthreeitems:Excessivedemandsofthejob;responsibilityofthejob;andfearoffailureandmistakesatwork.Itemswereratedonafive-pointscale(0=notatall,4=verymuch)(forbothdatasets:T1α=0.77;T2α=0.79,T3α=0.75).

2NB.IntheoriginalpaperoftheStudyI,theterm“jobdemands”hasbeenusedinsteadofmentalworkload.

50

2 metHods

2.2.3 Job resources

Threediversejobresourceswereexamined.AccordingtotheJD-Rmodel(Demeroutietal.,2001),eachjobhasitsownjobresourcesthatarerelevantintermsofemployeewell-being.Thus,theappliedjobre-sourceswereselectedaccordingtopriorstudiesthathaveidentifiedtheseresourcestobeimportantforfirefighters:(1)supervisory support(e.g.,Haslam&Mallon,2003;Mitanietal.,2006);(2)supportive interpersonal relations(e.g.,Saijo,Ueno,&Hashimoto,2007);and(3)task resources(e.g.,Lusaetal.,2006).Thethreejobresources–supervisoryrelations,interpersonalrelations,andtaskresources–wereadaptedfromtheOc-cupationalStressQuestionnaire(OSQ)(Elo,Leppänen,Lindström,&Ropponen,1992).TheOSQmeasurestheperceivedworkenvironmentanditseffects.Itiswell-validatedinFinlandandhasbeenusedamongseveraloccupationalgroups(Eloetal.,1992).

Supervisory relations(StudiesI–II,IV)includedfiveitemscoveringsupervisorysupport,supervisorycontrol,andrelationshipsbetweenemployeesandsupervisors(fordatasets1/2:T1α=0.81/0.81,T2α=0.80/0.82,T3α=0.84/0.83).Anexampleitemis“Doyougetsupportandhelpfromyoursupervisorwhenneeded?”Interpersonal relations(StudiesIIandIV)consistedoffouritems:conflictsbetweenemploy-ees,conflictsbetweenyoungerandolderworkers,cooperationinone’swork-unit,andrelationshipsbetweenemployees(forbothdatasets:T1α=0.74;T2α=0.72,T3α=0.77).Anexampleitemis“Whatkindofaretherelationsbetweenco-workersinyourworkplace?”Task resources(StudiesI–II,andIV)includedthreeitems:decision-makingonissuesconcerningone’stasks,opportunitiestouseone’sknowledgeandskillsatwork,andfeedbackonsuccessinworktasks(fordatasets1/2:T1α=0.67/0.67,T2α=0.68/0.67,T3α=0.72/0.72).Anexampleitemis“Canyouuseyourknowledgeandskillsatwork?”Alljobresourceitemswereratedonafive-pointscale(e.g.,1=notatall/practicallynever,5=verymuch).InStudiesIandIIacontinuoussumscoresofthejobresourceswereused,whereasinStudyIV,thesumscoresweredichotomizedattheirmediantocreatehighandlowgroupsofjobresources.

51

2 metHods

2.2.4 Personal resources

Self-esteem(StudyII)wasmeasuredbytheRosenbergSelf-EsteemScale(Rosenberg,1965)consistingof10items.Rosenberg’sself-reportscaleisavalid,reliable,andthemostwidelyusedmeasureofself-esteem(Gray-Little,Williams,&Hancock,1997;Marsh,1996).Itincludesbothpositive(e.g.,“Onthewhole,Iamsatisfiedwithmyself ”),andnegative(e.g.,“AttimesIthinkIamnogoodatall”)items.Allitemswereratedonafour-pointscalerangingfrom1=stronglydisagreeto4=stronglyagree(T2α=0.81).

Optimism(StudyIV)wasasingle-itemquestionfromthevalidatedWorkAbilityIndexQuestionnaire(Tuomietal.,1998):“Haveyourecentlyfeltyourselftobefullofhopeforthefuture?”Theoriginalfive-pointscale(0=never,4=often)wascategorizedintolow(<3)andhigh(≥3)levelsofoptimism.

2.2.5 Lifestyle factors

Alcoholconsumption,bodymassindex(BMI),smoking,physicalexer-cise,andsleepproblemswerestudiedasindicativeoflifestyle(StudiesIandIV).Alcohol consumptionwasmeasuredusingasingle-itemquestiononthefrequencyofalcoholconsumptionwithaneight-pointscale(1=never,8=dailyoralmostdaily).InStudyI,acontinuousvariablewasused.InStudyIV,thevariablewascategorizedintolow(<6)andhigh(≥6)alcoholconsumption.BMI(StudyI)wascalculatedbydividingbodyweight(kilograms)bythesquareofbodyheight(metres).Smokinghabitswereelicitedusingadichotomous(yes-no)questiononcurrentsmoking.Physical exercisewasassessedthroughasingle-itemquestiononthefrequencyofleisure-timeexerciseactivity,usingathree-pointscale(1=notatall,2=occasionally,3=frequently).InStudyIV,thevariablewascategorizedintolow(<3)andhigh(3)physicalactivity.Finally,afour-itemscaleofsleep problemswasderivedfromtheBasicNordicSleepQuestionnaire(Partinen&Gislason,1995):Difficultiesinfallingasleepduringthepastthreemonths;sleepingwellduringthepastthreemonths;awakingtooearlyinthemorningandnotbeingabletofallbackasleepduringthepastthreemonths;andextremetirednessduringdaytime.Alltheitemswereratedonafive-pointscale(1=not

52

2 metHods

atall,5=daily/almostdaily)exceptsleepingwell,whichwasmeasuredwithathree-pointscale(1=well,3=moderately,5=poorly)(fordatasets1/2:T1α=0.76/0.76,T2α=0.79/0.78,T3α=0.79/0.80).InStudyIV,thesumscorewasdichotomizedusingmediantocreatelow(<8)andhigh(≥8)sleepingproblems.Table2summarizesthevariablesusedinthesub-studies.

Table 2. Variables used in the sub-studies.

Variables Study I Study II Study III Study IV

Affective well-being

Work engagement x x

Vigor x

Fatigue x

Depressive symptoms x

Health-related well-being

Work ability

Work ability index x x

Single-item x

Musculoskeletal pain x

Self-rated health x

Life satisfaction x

Job demands

Physical work load x x

Mental work load x x

Job resources

Supervisory relations x x x

Interpersonal relations x x

Task resources x x x

Personal resources

Self-esteem x

Optimism x

Lifestyle factors

Alcohol consumption x x

BMI x

Smoking x x

Physical exercise x x

Sleeping problems x x

53

2 metHods

2.3 statistical Analyses

Multiplestatisticalanalyseswereusedinthisstudy.LinearregressionanalysiswasusedinStudyI,structuralequationmodelinginStudyII,latentclassgrowthmodelinginStudiesIII–IV,mixedlinearmodelsinStudyIII,andmultinomialregressionanalysisinStudyIV.ContinuousvariableswereappliedinStudiesI,II,andIII,whereasduetothestudydesign,categorizedvariableswereappliedinStudyIV.AnalyseswereconductedusingPASWStatistics18forWindows(StudyI),AMOS18.0softwarepackage(StudyII),andSAS(version9.2)software(StudiesIII–IV).

In Study I,exploratoryfactoranalysis(EFA)wasusedtoexaminewhetherthedifferentscalesofjobdemandsandjobresourcescouldbedistinguishedfromeachother.Multicollinearitywaschecked,andfoundnocollinearityproblemsinthedata(seeMyers,1990).Toexaminewhetherlifestyle,jobdemands,jobresources,andworkengagementwererelatedtotheWAIoritssixsub-dimensionsin2009,linearregressionanalysiswasused.ThebaselineWAIandthesub-dimensionsoftheWAItenyearsearlierwerecontrolled.Thus,intotalsevendifferentmodelswereanalysed.

In Study II,structuralequationmodeling(SEM)techniqueswereusedtotesttheroleofworkengagementasamediatorbetweenjobandpersonalresourcesontheonehandandworkabilityontheother.SEMwaschosenduetoseveraladvantages.Firstly,ittakesaconfirmatory(i.e.,hypothesis-testing)ratherthanexploratoryapproachtothedataanalysis.Secondly,SEMprovidesexplicitestimatesoferrorvarianceparameters,thatis,byusinglatentvariablesitispossibletoassessandcorrectmeasurementerrors,whichallowsmorereliabletestsofthestudymodels.Thirdly,bothunobserved(i.e.,latent)andobservedvariablescanbeincorporatedinSEMprocedure.Fourthly,SEMallowsforthepossibilityoftestingmodelsagainsteachother.Finally,SEMallowsreliablemethodstotestindirecteffects.(Formoredetails,seeByrne,2010.)Inthefirstphase,themeasurementmodel(confirmatoryfactoranalyses,CFA)thatdefinesrelationsbetweentheobservedandunob-servedvariableswastested(seeMulaik&James,1995).Aftertestingthemeasurementmodel,structuralequationmodelsweretested.Modelfitwasevaluatedusinggoodness-of-fitindices(i.e.,Chi-squareχ2test,theRootMeanSquareErrorofApproximation,theComparativeFit

54

2 metHods

Index,andtheTucker-LewisIndex),andconventionalrulesofthumbfortheircut-offs(seeBrowne&Cudeck,1993;Byrne,2010).Inaddi-tion,tocomparethedifferentmodels,Akeike’sInformationCriterion(AIC)wasused.Finally,abootstrappingon2,000subsamplesfromtheoriginaldatausingtheMLestimatorwithbias-corrected95%confidenceintervalsforeachoftheparameterbootstrapestimateswasconductedinordertoexaminewhetherjobresources,self-esteem,andworkabilityatT1yieldedanindirecteffectviaworkengagementonworkabilityatT2(seee.g.,Hayes,2009).

In Study III,alatentclassgrowthmodeling(LCGM)(Jones&Nagin,2007;Nagin,1999)toidentifylatentmood(i.e.,vigorandfatigue)trajectorieswasused.LCGMisasemi-parametricstatisticaltechniqueusedtoidentifydistinctsubgroupsofindividualsfollowingasimilarpatternofchangeovertimeinavariableofinterest(Andruff,Carraro,Thompson,Gaudreau,&Louvet,2009).Theparametersofthemultitrajectorymodelswereestimatedbymaximumlikelihoodestimation,andthecensorednormalmodeldistributionwasused.Todecideontheadequatenumberoflatentclasses,differentcriteriawereused:(i)TheBayesianInformationCriterion(BIC)statistics;(ii)TheAICstatistics;(iii)theposteriorprobabilities;and(iv)theusefulnessandclarityofthelatentclassesinpractice.Inordertoinvestigatetherelationshipbetweenthelatentmoodclassesandemployeewell-being,themixedlinearmodelswereused.Moreover,thet-testswithBonferronicorrectionwerecarriedouttoinvestigatewhetherdifferencesexistedonthelevelsofworkability,self-ratedhealth,andlifesatisfactionbetweenthedifferentlatentclasses.Agewasadjustedineachmodel.

In Study IV,theeffectsofjobdemands,jobresources,personalresources,andlifestyleonpain-depressiontrajectorieswereanalysedusingtheLCGMinasimilarwaytoStudyIII.Inordertoinvestigatetherelationshipbetweenthelatentclassesandantecedentsin1996,multinomiallogisticregressionanalyseswereconducted.Theassociationswerepresentedasoddratios(OR)andtheir95%confidenceintervals(95%CI).Allmodelswereadjustedwithage,andfinally,allsignificantpredictorswereputintothesamemodelinordertoexaminetheirin-dependenteffects.Inthefinalmodel,variablesthatwerenotsignificantatthecriticallevel0.05inrelationtoeitheroftheexaminedtrajectorieswereeliminated.

55

2 metHods

Amoredetaileddescriptionofthestatisticalanalysesisprovidedintheoriginalarticles.Table3summarizesthestudytype,sample,mainvariables,anddataanalysesofeachsub-study.

Table 3. Study type, sample, main variables, and statistical analysis of the Studies I–IV.

Study Study type Sample Main variables Data analysis

Study IThe relationship between work engagement and work ability

Cross-sectional

n=403 Job demands; job resources; lifestyle factors; work engagement; work ability

Exploratory factor analysis (EFA); linear regression analysis

Study IIThe mediating role of work engagement; the role of work ability in the motivational process of the JD-R model

Longitudinal sample with two waves

n=403 Job resources; personal resources (self-esteem); work engagement; work ability

Confirmatory factor analysis (CFA); structural equation modeling (SEM); boostrapping using the MLE

Study IIIThe co-development of vigor and fatigue, and their relationship with well-being

Longitudinal sample with three waves

n=360 Vigor; fatigue; life satisfaction; self-rated health; work ability

Latent class growth modeling (LCGM); mixed linear models

Study IVThe co-development of musculoskeletal pain and depressive symptoms, and their antecedents

Longitudinal sample with three waves

n=360 Job demands; job resources; personal resources (optimism), lifestyle factors; depressive symptoms; multisite musculoskeletal pain

Latent class growth modeling (LCGM); multinominal logistic regression

56

3 ResuLts

Theresultsarepresentedbysub-studies.Firstly,therelationshipbetweenworkengagementandworkabilityisinvestigated(StudyI).Secondly,theroleofworkabilityinthemotivationalprocessoftheJD-Rmodelisexamined(StudyII).Thirdly,theco-developmentofvigorandfatigueandtheirrelationshiptoemployeewell-beingwillbepresented(StudyIII).Fourthlyandfinally,theresultsofthedevelopmentalpathsofmultisitemusculoskeletalpainanddepressivesymptoms,andtheante-cedentsrelatedtoparticulartrajectorygroupswillbepresented(StudyIV).Detailedanalysesofsub-studiesarepresentedintheoriginalpapers.

3.1 Relationship between work engagement and work ability (study i)

StudyIinvestigatedwhetherworkengagementisrelatedtoworkabilityaftercontrollingforage,lifestyle,workcharacteristics,andworkabilitytenyearsearlier.Bothworkabilityindex(WAI)anditssub-dimensionswereexamined.Table4showstheresultsofthestepwiseregressionanaly-sesrelatedtototalWAI.AgewasnegativelyrelatedtoWAI.Thelifestylefactorsincludedinstep2significantlyimprovedtheregressionmodel.Morespecifically,sleepproblemswerenegativelyandphysicalexercisepositivelyrelatedtoworkability.Incontrast,alcoholconsumption,BMIandsmokingwerenotrelatedtothetotalWAIscore.Addingworkchar-acteristics(step3)tothemodelfurthersignificantlyimprovedtheregres-sionmodel.However,onlyphysicalworkloadwassignificantlyrelatedtotheWAI.Finally,workengagementwasaddedtothemodel(step4).Theresultsshowedapositiverelationshipbetweenworkengagementandworkability.Thefinalmodelexplained53%ofthevarianceoftheWAI.

57

3 resULts

Table 4. Associations of age, lifestyle factors, work characteristics, and work engagement with work ability index (WAI) in 2009 (n = 403).

Explanatory variables Work ability index

b 95% CI ∆R2 R2

Step 1 .39*** .39

WAI 1999 .53 [.36, .70]

Age -.14 [-.26, -.03]

Step 2: Lifestyle factors .09*** .48

Alcohol consumption -.01 [-.38, .35]

BMI -.16 [-.38, .06]

Smoking -1.38 [-2.79, .04]

Physical exercise 1.74 [.43, 3.04]

Sleep problems -.28 [-.46, -.10]

Step 3: Work characteristics .04** .52

Mental workload -.73 [-1.51, .06]

Physical workload -1.32 [-2.40, -.23]

Supervisory relations -.01 [-.80, .78]

Task resources .32 [-.81, 1.44]

Step 4: Work engagement .01* .53

Work engagement .61 [.09, 1.13]

Note. b = unstandardized beta-coefficient from the final step; CI = confidence interval; ∆R2 = change in explanation rate; R2 = explanation rate.* p < .05. ** p < .01. *** p < .001.

ThesecondaimoftheStudyIwastoexaminewhetherage,lifestyle,jobdemandsandresources,andworkengagementwereassociatedwiththesub-dimensionsoftheWAI.Resultsindicatedthatallindependentvariables,exceptBMIandalcoholconsumption,wereassociatedwithatleastonesub-dimensionoftheWAIaftercontrollingthebaselineWAI.AgewasnegativelycorrelatedtothetwoWAIsub-dimensions,namelycurrentworkabilityandownprognosisofworkability.Oflife-stylefactors,frequentsleepproblemswerenegativelyrelatedtocurrentworkability,workabilityinrelationstojobdemands,andpsychologicalresources.Inturn,frequentphysicalexercisewaspositivelyassociated

58

3 resULts

withcurrentworkabilityandworkabilityinrelationtojobdemands,andnegativelywithsickleave.Smokingwasnegativelyrelatedtoworkabilityinrelationtojobdemands.Workcharacteristicswereassociatedwithfivesub-dimensionsoftheWAI.Morespecifically,highmentalworkloadwasnegativelyrelatedtoworkabilityinrelationtojobde-mandsandpsychologicalresources.Highphysicalworkloadinturnwaspositivelyassociatedwithahigherfrequencyofdiseasesandsickleave.Inaddition,taskresourceswerepositivelyrelatedtoownprognosisofworkability,whereassupervisoryrelationswerenotrelatedtoanyofthesub-dimensionsoftheWAI.Workengagementwaspositivelyrelatedtothreesub-dimensionsofWAI:goodcurrentworkability,goodworkabilityinrelationtojobdemands,andahigherlevelofpsychologicalresources.Ofthedifferentsub-dimensions,thehighestexplainedvariancesofthestudymodelsconcernedworkabilityinrelationtojobdemands(48%)andpsychologicalresources(45%).Incontrast,thelowestexplainedvariancesconcernedsickleave(20%)andnumberofdiseases(24%).MoredetailedresultsarepresentedintheoriginalpaperoftheStudyI.

3.2 Work engagement as a mediator between job resources, personal resources, and work ability (study ii)

StudyIIexaminedthemediatingroleofworkengagementbetweenjobresourcesandpersonalresourcesandworkability,andtheroleofworkabilityinthemotivationalprocessoftheJD-Rmodel.Themeasure-mentmodelproducedanacceptablefittothedata.Ofthestructuralequationmodels,thehypothesizedmediationmodelfittedwelltothedataandsignificantlybetterthananyoftheotherfourtestedmodels(forfitindicesandchi-squaredifferencetestsofthefivetestedmodels,pleaseseetheoriginalpublicationoftheStudyII).Inthebest-fittingmodel,displayedinFigure3,bothjobresourcesatT13(β=.19,p<.01)andself-esteematT1(β=.19,p<.01)werepositivelyrelatedtoworkengagementatT2.Furthermore,workengagementatT2waspositively

3NB.HereT1refersto1999andT2to2009inasimilarwayasintheoriginalpaperoftheStudyII.

59

3 resULts

relatedtoworkabilityatT2(β=.29,p<.001).WorkabilityatT1alsopredictedworkabilityatT210yearslater(β=.30,p<.001),aswellasworkengagementatT2(β=.12,p<.05).Thehypothesizedmodelexplained12%ofthevarianceinworkengagementatT2and21%ofthevarianceinworkabilityatT2.Finally,bootstrappingwasusedtotestwhetherjobresources,self-esteem,andworkabilityatT1yieldedanindirecteffectviaworkengagementonworkabilityatT2.Allindirecteffectswereconfirmed(pleaseseetheoriginalpublicationofStudyIIfordetailedbootstrappinganalyses),thussupportingthemediatingroleofworkengagementbetweenjobresourcesandworkability(Hypothesis1),andbetweenself-esteemandworkability(Hypothesis2),respectively.Inaddition,workabilityatT1hadanindirecteffectonworkabilityatT2,viaworkengagement,andthus,supportingHypothesis3.

60

3 resULts

Figu

re 3

. Fin

al m

odel

of

the

med

iatin

g ro

le o

f w

ork

enga

gem

ent

betw

een

job

reso

urce

s, s

elf-

este

em, a

nd w

ork

abili

ty

(n =

403

).

Note.

***p

< .0

01; *

*p <

.01;

*p

< .0

5.

(Airi

la, H

akan

en, S

chau

feli,

Luu

kkon

en, P

unak

allio

, & L

usa,

201

3).

Job

reso

urce

s T1

Sel

f-est

eem

T1

Wor

k ab

ility

T1

Wor

k en

gage

men

t T2

W

ork

abili

ty

T2

.19*

*

.19*

*

.12*

.29*

**

.30*

**

Sel

f-rat

ed

wor

k ab

ility

Sup

ervi

sory

rela

tions

Inte

rper

sona

l rel

atio

ns

Task

reso

urce

s

Neg

ativ

e it

ems

Pos

itive

ite

ms

.91

.52

.53

!.87

.59

.96

.98

Sel

f-rat

ed

wor

k ab

ility

Vig

our

Abs

orpt

ion

Ded

icat

ion

.94

.96

.84

.17

61

3 resULts

3.3 mood trajectories and their relationship to work ability, self-rated health, and life satisfaction (study iii)

StudyIIIinvestigatedthelatentmoodtrajectories(i.e.,vigorandfatigue)andtheirrelationshipstoemployeewell-being(i.e.,workability,self-ratedhealth,andlifesatisfaction)duringa13-yearfollow-upperiod.Altogether,trajectorymodelswith1–5trajectoriesweretested.TheBICstatisticsrevealedthatafour-trajectorymodelwasthebestfittingmodelwithac-ceptableposteriorprobabilitiesforallfourgroupsrangingfrom.78to.88(fordetailinformationaboutthefitcriteriaofthetestedtrajectories,pleaseseetheoriginalpublicationofStudyIII).Thus,fourdistincttra-jectoriesthatdifferedfromeachotherintheirmeanlevelsandchangeswereidentified:(a)highpositive(n=149,41%),(b)highpositivebutdecreasing(n=106,29%),(c)moderatelypositive(n=79,22%),and(d)highnegative(n=26,7%).Thehigh positivetrajectoryconsistedofthoseparticipantswhohadahighlevelofvigorandlowleveloffatigueacrosstime.Inthisgroup,thelevelofvigorremainedstablethroughoutthe13-yearfollow-up,whereasthefeelingsoffatigueslightlyincreasedbetweenT1andT2,andremainedstablethereafter.Thehigh positive but decreasingtrajectoryconsistedofparticipantswhohadahighbutdecreas-inglevelofvigorandhighleveloffatigue.Atbaseline,thelevelsofvigorandfatiguedifferedfromeachothersubstantially.However,thesetwoaffectivestateswereclosertoeachotherbyT2andremainedratherstablethereafter.Thelatentgrouplabelledmoderately positivewascharacterizedbyamoderateandstablelevelofvigorandamoderateleveloffatigue,whichslightlyincreasedbetweenT2andT3.Inthisgroup,feelingsofvigorwereatahigherlevelthanthefeelingsoffatiguethroughoutthestudyperiod.Thetrajectorylabelledhigh negativewascharacterizedbyahighleveloffatigueandlowlevelofvigor,andbothfeelingsremainedstableovertime.ThetrajectoriesarepresentedinFigure4.

Intermsofwell-being,membersofthehigh positivetrajectoryhadbetterwell-beingthanthemembersofotherlatentmoodtrajectories.Incontrast,membersofthehigh negativegrouphadconsiderablylowerworkability,SRHandlifesatisfactionthanthemembersoftheotherlatentgroups.Inallwell-beingoutcomes,thedifferencesbetweenhigh positive but decreasingandmoderately positivelatentgroupswerenon-significant.ThedetailedanalysesarepresentedintheoriginalpublicationoftheStudyIII.

62

3 resULts

1,5 2

2,5 3

3,5 4

T1 (1

996)

T2

(199

9)

T3 (2

009)

Vigo

r Fa

tigue

1,5 2

2,5 3

3,5 4

T1 (1

996)

T2

(199

9)

T3 (2

009)

Vigo

r Fa

tigue

1,5 2

2,5 3

3,5 4

T1 (1

996)

T2

(199

9)

T3 (2

009)

Vigo

r Fa

tigue

1,5 2

2,5 3

3,5 4

T1 (1

996)

T2

(199

9)

T3 (2

009)

Vigo

r Fa

tigue

a) H

igh

posi

tive

traje

ctor

y (n

=149

) c)

Mod

erat

ely

posi

tive

traje

ctor

y (n

=79)

b) H

igh

posi

tive

but d

ecre

asin

g tra

ject

ory

(n=1

06)

d) H

igh

nega

tive

traje

ctor

y (n

=26)

Figu

re 4

. Est

imat

ed d

evel

opm

enta

l tra

ject

orie

s fo

r vi

gor

and

fatig

ue (n

= 3

60).

(Airi

la, H

akan

en, L

uukk

onen

, Lu

sa, &

Pun

akal

lio, 2

013)

.

63

3 resULts

3.4 Pain-depression trajectories, work characteristics, and individual factors (study iv)

StudyIVinvestigatedthedevelopmentofmultisitemusculoskeletalpainanddepressivesymptomsduringa13-yearfollow-upperiod;andwhetherthebelongingtoaparticulartrajectorygroupwouldbeassociatedwithbaselineworkcharacteristics,personalresources,andlifestyle.Threedistincttrajectoriesthatdifferedfromeachotherintheirmeanlevelsandchangesinmultisitemusculoskeletalpainanddepressivesymptomswereidentified:(a)LowSymptoms(n=179,49%),(b)HighPain(n=115,32%),and(c)HighDepression(n=66,19%).ThefitcriteriaofthetrajectorieswiththedifferentnumberoflatentclassesaredisplayedintheoriginalpublicationoftheStudyIV.

TheLow Symptomstrajectoryconsistedofthoseparticipantswhohadlowlevelsofbothsymptoms.Inthisgroup,thefeelingsofdepressionremainedstablethroughoutthe13-yearfollow-upperiod,whereasthelevelofmusculoskeletalpainslightlyincreasedbetweenT2andT3.TheHigh Paintrajectoryconsistedofparticipantswhohadahighandincreas-inglevelofmusculoskeletalpainandlowlevelofdepressivesymptoms.ThelatentgrouplabelledHigh Depressionwascharacterizedbyahighlevelofdepressivesymptomsandamoderatelevelofmusculoskeletalpain.Inthisgroup,thelevelofdepressivesymptomsincreaseddrasticallybetweenT1andT2,anddecreasedthereafter,howeverstillremainingatahighlevel.(SeeFigures5and6.)Takentogether,hypotheses1a–1cweresupported,asbothtrajectorygroupswithlowsymptoms(H1a),highpain(H1b),andhighdepressivesymptoms(H1c)werefound.Instead,notrajectorywithhighlevelsofbothsymptomsemerged,andthusHypothesis1dwasnotsupported.

64

3 resULts

Figure 5. Mean profiles for musculoskeletal pain (n = 360). (Airila, Hakanen, Luukkonen, Punakallio, Lusa, & Leino-Arjas, 2014).

Figure 6. Mean profiles for depressive symptoms (n = 360). (Airila et al., 2014).

!

"!

#!

$!

%!

&!

'!

(!

)!

*!

+!

#"!

,-.!/0123-1/!456#)+7!

89:;!2<95!456##'7!

89:;!=>2?>//9-5!456((7!

@>2

?>//9A>!/0123-1

/!

B%!4$""+7!B$!4#+++7!B#!4#++(7!

!

!

!

!

!

"!

"#$!

%!

%#$!

&!

&#$!

'!

'#$!

(!

(#$!

)*+!,-./0*.,!123%456!

789:!/;82!123%%$6!

789:!<=/>=,,8*2!123??6!

@&!1%5556!@%!1%55?6! @'!1&""56!

AB,CBD*,E=D=0;D!/;82!

!

!

!

!

!

"!

"#$!

%!

%#$!

&!

&#$!

'!

'#$!

(!

(#$!

)*+!,-./0*.,!123%456!

789:!/;82!123%%$6!

789:!<=/>=,,8*2!123??6!

@&!1%5556!@%!1%55?6! @'!1&""56!

AB,CBD*,E=D=0;D!/;82!

!

!

!

!

!

"!

"#$!

%!

%#$!

&!

&#$!

'!

'#$!

(!

(#$!

)*+!,-./0*.,!123%456!

789:!/;82!123%%$6!

789:!<=/>=,,8*2!123??6!

@&!1%5556!@%!1%55?6! @'!1&""56!

AB,CBD*,E=D=0;D!/;82!

!

"!

#!

$!

%!

&!

'!

(!

)!

*!

+!

#"!

,-.!/0123-1/!456#)+7!

89:;!2<95!456##'7!

89:;!=>2?>//9-5!456((7!

@>2

?>//9A>!/0123-1

/!

B%!4$""+7!B$!4#+++7!B#!4#++(7!!

"!

#!

$!

%!

&!

'!

(!

)!

*!

+!

#"!

,-.!/0123-1/!456#)+7!

89:;!2<95!456##'7!

89:;!=>2?>//9-5!456((7!

@>2

?>//9A>!/0123-1

/!

B%!4$""+7!B$!4#+++7!B#!4#++(7!!

"!

#!

$!

%!

&!

'!

(!

)!

*!

+!

#"!

,-.!/0123-1/!456#)+7!

89:;!2<95!456##'7!

89:;!=>2?>//9-5!456((7!

@>2

?>//9A>!/0123-1

/!

B%!4$""+7!B$!4#+++7!B#!4#++(7!!

"!

#!

$!

%!

&!

'!

(!

)!

*!

+!

#"!

,-.!/0123-1/!456#)+7!

89:;!2<95!456##'7!

89:;!=>2?>//9-5!456((7!

@>2

?>//9A>!/0123-1

/!

B%!4$""+7!B$!4#+++7!B#!4#++(7!

65

3 resULts

Thisstudyexaminedalsotheeffectsofjobdemandsandresources,lifestyle,andoptimismonthepain-depressiontrajectories.Firstly,theindependenteffectofeachpredictoronthepain-depressiontrajectorieswasexamined.WhentheHighDepressionwascontrastedwiththeLowSymptomstrajectory,highmentalworkloadpredictedbelongingtotheformergroup(OR4.48,CI:2.45–8.18)whenagewasincludedinthemodel.ThefindingwassimilarwhencontrastingHighPainandLowSymptoms,however,theassociationwasminor(OR1.80,CI:1.09–2.94)comparedtothatofHighDepression.Physicalworkloadwasnotasignificantpredictorinbothcontrasts.Again,withtheLowSymptomstrajectoryasreference,poorinterpersonalrelationspredictedbelongingtoHighPain(OR1.72,CI:1.06–2.79)andHighDepression(OR3.61,CI:1.99–6.57).Similarly,poorsupervisoryrelations(OR2.76,CI:1.52–5.01)andlowtaskresources(OR2.56,CI:1.42–4.61)predictedbelongingtotheHighDepression.However,theywerenotsignificantwhenHighPainwascontrastedwithLowSymptoms.Oflifestylefactors,alcoholconsumptionpredictedbelongingtoHighPain(OR2.40,CI:1.45–4.00),whereassmokingandphysicalexercisewerenon-significantpredictorsinbothcontrasts.SleepingproblemspredictedbelongingbothtoHighPain(OR2.41,CI:1.47–3.95)andHighDepression(OR3.76,CI:2.07–6.81)whencontrastedwithLowSymptoms.Similarly,lowlevelsofoptimismpredictedgroupmembershipinHighDepression(OR2.68,CI:1.49–4.83)butnotinHighPain.

Table5showstheage-adjustedresultsofthefinalmodelinwhichallantecedentsofpain-depressiontrajectorieswereanalysedatthesametime.WiththeLowSymptomsasreference,highmentalworkloadandsleepingproblems,poorinterpersonalrelations,andlowlevelofoptimismpredictedbelongingtoHighDepression.ContrastingHighPainwithLowSymptoms,alcoholconsumptionandsleepingproblemsweresignificantpredictors.Thus,Hypothesis2waspartlysupported:highlevelsofjobdemandsandlowlevelsofjobresourcesandpersonalresourcesatbaselinepredictedbelongingtothetrajectorygroupsofHighPainandHighDepression.However,afterincludingallthesignificantpredictorvariablesinthesamemodel,HighPaintrajectorywasnotanymorepredictedbyanyoftheJDsorJRs.

66

3 resULts

Table 5. Predicting of membership in pain-depression trajectories; final multivariable model. Multinominal regression analysis; odds ratios (OR) and their 95% confidence intervals (CI), (n = 360).

High Pain vsLow Symptoms

High Depression vs Low Symptoms

OR 95% CI OR 95% CI

Age 1.02 [.98, 1.07] 1.02 [.96, 1.08]

Mental work load

Low 1.00 1.00

High 1.39 [.81, 2.38] 2.87 [1.51, 5.45]

Interpersonal relations

Good 1.00 1.00

Poor 1.40 [.83, 2.34] 2.63 [1.39, 4.97]

Alcohol consumption

Low 1.00 1.00

High 2.42 [1.43, 4.11] 1.42 [.75, 2.70]

Sleeping problems

Low 1.00 1.00

High 2.13 [1.27, 3.59] 2.72 [1.44, 5.15]

Optimism

High 1.00 1.00

Low 1.23 [.73, 2.06] 1.97 [1.04, 3.72]

67

4 disCussion

Theaimofthisstudywastoexaminetheeffectsofworkcharacteristics,personalresources,andlifestyleonemployeewell-being,aswellasde-velopmentalpathsofaffectivestates,andmultisitemusculoskeletalpainanddepressivesymptomsamongFinnishfirefightersusingalongitudinaldata.Thestudyaimedtoprovideacomprehensiveviewofemployeewell-beingbyexaminingaffective,health-related,andcognitiveaspectsofemployeewell-beingapplyingthreecomplementarytheoreticalframeworks–JD-Rmodel,CORtheory,andBaBtheory.Theoretically,thisstudyextendedpreviousknowledgebybringingnewinputbothtoworkengagement,aswellastoworkabilityliterature.Firstly,thisstudyprovidedevidenceforthemotivationalpotentialofworkengagementinexplainingworkability.Secondly,thisstudyfoundsupportforthehealth-relatedpropertiesofthemotivationalprocessoftheJD-Rmodel,andparticularly,theroleofworkabilityinit.Moreover,applyingaperson-centeredapproachinexaminingemployeewell-beingovera13-yearfollow-upperiod,thisstudyprovidednewinformationonindividualdevelopmentalpathsinpositiveandnegativeaffectivestatesaswellasinmultisitemusculoskeletalpainanddepressivesymptoms.Toconclude,longitudinallinkagesbetweenworkcharacteristics,personalresources,lifestyle,andemployeewell-beingwerefound.Next,themainfindingsofthestudyarediscussedinmoredetail.

4.1 motivational state of work engagement associated with work ability

Avastnumberofstudiesexistontheassociationsbetweenindividuallifestyle,workcharacteristics,andworkability(vandenBergetal.,2009).

68

4 discUssion

However,thusfartheroleofmotivationalaspects–hereworkengage-ment–hasbeenneglectedinexplainingworkability,despitethefactthatmotivationalfactorsareconsideredessentialfactorsrelatedtoworkability(e.g.,Ilmarinen,2009).Moreover,thecontributionofworkengagementforhealthandworkabilitybeyondlifestyleandworkcharacteristicshasrarelybeenexamined(forexception,seeRongen,Robroek,Schaufeli,&Burdorf,2014).Theresultsofthisstudyindicatedthatthepositivestateofworkengagementconsistingofvigour,dedication,andabsorptionwassignificantlyassociatedwithworkability.Infact,workengagementwasrelatedbothtothetotalscoreoftheWAIandtoitsthreesub-dimensionsevenafteradjustingforvariousindividualandworkcharacteristics.Morespecifically,thisstudyshowedthatamotivatedandenergeticworker,whostronglyidentifieswithhis/herwork,ismorelikelytohavebetterworkabilitythanhis/herlessengagedco-worker.ThisfindingisinlinewithastudyofFinnishteachers(Hakanenetal.,2006)inwhichworkengagementwaspositivelycorrelatedwithself-ratedworkabilitymeas-uredwithaone-itemindicator.Inadditiontoworkengagement,otherkeyfactorspromotingworkabilityamongfirefightersweregoodsleep,frequentexercise,nottoohighphysicalworkloadormentalworkload,andgoodtaskresources.Similarfindingsexistontheeffectsofworkcharacteristics(e.g.,Fischeretal.,2006;Pohjonen,2001;Tuomietal.,1991)andlifestyle(e.g.,Kaleta,Makowiec-Dabrowska,&Jegier,2006;Lusa,Häkkänen,Luukkonen,&Viikari-Juntura,2002;Rongenetal.,2014)onworkability.

Lifestylemayberelatedtoemployeewell-beingthroughseveralpsy-chologicalpathways.Forexample,evidenceindicatesthatexercisemayincreaseself-esteem,positiveaffect,andself-efficacy(Biddle&Mutrie,2008;Elavskyetal.,2005),whichconsequentlymayincreasehealthandwell-being.Similarly,ahealthylifestylehasbeenrelatedtolowerlevelsoffatigue,tension,depression,andanxiety(Biddle&Mutrie,2008),andconsequently,betterhealth.

Thisstudyaddednewknowledgealsobyexaminingtheantecedentsofdifferentsub-dimensionsoftheWAIseparately,insteadofatotalindex.Infact,knowledgeontherelationshipsbetweenthesub-dimensionsoftheWAIandindividualandworkcharacteristicsareessentialinordertoimproveworkabilityandapplyproperlyfocusedinterventionsinworkplaces.Indeed,thedifferentsub-dimensionsoftheWAIhadpartlydifferentantecedents.Furthermore,lifestyle,workcharacteristics,and

69

4 discUssion

workengagementweremorestronglyrelatedtothesubjectiveWAIdimensions(i.e.,currentworkability,workabilityinrelationtojobdemands,ownprognosisofworkability,andpsychologicalresources)thantothetwomoreobjective(i.e.,numberofdiseasesandsickleave)WAIsub-dimensions.Generally,moderateassociationbetweenworkengagementandphysiologicalhealthindicatorshavebeenfound(e.g.,Bakker&Leiter,2010;Langelaan,Bakker,Schaufeli,VanRhenen,&VanDoornen,2006).Intheirstudyonconstructionworkers,Alavinia,deBoer,vanDuivenbooden,Frings-Dresen,andBurdorf(2009)foundthatwork-relatedriskfactorsweremorestronglyassociatedwiththesubjectiveworkabilitysub-dimensionsthanwiththehealth-relatedsub-dimensions.Asregardssickleave,onestudyfoundthatamongDutchmanagers,workengagementpredictedabsencefrequencybutnotabsenceduration(Schaufelietal.,2009).TheWAIinstrumentdoesnotdifferentiatebetweenfrequencyanddurationofsickleave,butsimplymeasuresthenumberofabsencedays.Thismayexplainwhyinthisstudyanassociationbetweenworkengagementandsickleavewasnotfound.Conceptually,ithasrecentlybeensuggestedthattheone-factormodelofworkabilityshouldbedismissed(Martus,Jakob,Rose,Seibt,&Freude,2010),andreplacedwithatwo-dimensionalinstrumentcoveringsubjectivelyestimatedworkabilityandobjectivehealthstatus.Thisstudyfurthergavesupportforsuchadistinction.Therefore,futurestudiesshouldfurtherinvestigatethedimensionalityoftheWAIandtheimportanceofthedifferentsub-dimensionswithinthetotalscore.

Toconclude,thesefindingscontributetothepreviousresearchonworkabilitybyaddingamotivational,work-relatedwell-beingconstructofworkengagementasapotentiallyimportantantecedentofworkabilityanditssub-dimensions.Theoretically,thesefindingssupportthebasicideaoftheJD-Rmodelofthedetrimentaleffectofjobdemands,andhealth-enhancingeffectsofjobresources.Practically,theseresultssuggestthatbyimprovinglifestyle,work-relatedfactors,andworken-gagement,itmaybepossibletoimproveworkability,andespeciallyitssubjectivedimensions,i.e.,self-ratedcurrentworkability,workabilityinrelationtojobdemands,andpsychologicalresources.Thus,notonlypromotinglifestyleorworkingconditions,butalsofosteringapositiveandmotivationalstateofworkengagementislikelytobevaluableinmaintainingworkability.

70

4 discUssion

4.2 motivational process and its health-related properties

4.2.1 Longitudinal evidence of the importance of job and personal resources

Supportingallstudyhypothesis,sub-studyIIshowedthatjobresourceshadmotivationalpotentialastheywererelatedtofutureworkengage-ment,andconsequentlytoworkabilityovera10-yearstudyperiod.Thus,thebeneficialroleofjobresourcesandpersonalresourcesinthemotivationalprocessoftheJD-Rmodelgainedfurthersupportfromthislongitudinalstudy.Asindicatedbythisstudy,jobscharacterizedbysupportiveconditionssuchasautonomoustasks,positiveinteractionsbetweenco-workers,andsupportandpositivefeedbackfromone’ssuper-visor,mayeveninthelong-termfosterflourishingandengagedemployeeswhoenjoygoodworkability.Thelong-termimpactsofjobandpersonalresourcesonwell-beingsupporttheassumptionofCORtheoryofslowpositiveaccumulationprocessesresultinginlongtermresourcegains.Inaddition,theseresultssupportself-determinationtheory(SDT;Deci&Ryan,2000),whichemphasizestheimportanceofsocial-contextualconditionsthateitherenhanceorhindermotivationatwork.Empiri-cally,theresultsofthisstudyarelinewithpreviousstudiesthatsuggestpositiveassociationsbetweenjobresourcesand/orworkengagementandhealth-relatedoutcomes(e.g.,Hakanen&Schaufeli,2012;Langelaanetal.,2006;Parzefall&Hakanen,2010;Seppäläetal.,2012).

Inaddition,thepresentstudyrevealedthatself-esteemasapersonalresourceplaysasignificantroleinshapingworkengagementandalsoworkabilityinthelong-termviaengagement,evenwhentheimpactsofbaselineworkabilityandjobresourceswerecontrolled.Inotherwords,thewayinwhichpeopleevaluatethemselvesislikelytoinfluencehowengagedtheyareandhowtheyassesstheirworkability.Thisresultisinlinewithpreviousstudiesonthepositiverelationshipbetweenself-esteemandwell-being(forareview,seeBaumeister,Campbell,Krueger,&Vohs,2003,seealsoDeNeve&Cooper,1998;Mäkikangas,Kinnunen,&Feldt,2004;Schimmack&Diener,2003).Moreover,thisfindingiscon-sistentwiththebasicassumptionoftheJD-Rmodelthathighlightstherelationshipsbetweenpersonalresources,workengagement,andpositive

71

4 discUssion

outcomes.Inasimilarvein,theresultsupportsbothself-enhancementtheory(Jones,1973)andBaBtheory(Fredrickson,2001)thathighlighttheimportanceofpersonalresources(e.g.,self-esteem)inmaintainingandenhancingwell-being.Tentatively,theresultsofthepresentstudyalsolendsupporttotheCORtheory’sassumptionofresourcescaravans(Hobfoll,1989,2001),thatis,increasingresources(i.e.,jobresourcesandself-esteem)tendtogeneratenewresources(i.e.,workengagementandworkability),andthusformresourcecaravans.

Severalmechanismsmaypossiblylinkpersonalityfactors(e.g.,self-esteem)andemployeewell-being.Forexample,individualswithahighlevelofself-esteemmaybemorelikelytouseproactivecopingstrategiestohandlestressfulsituations(e.g.,Antonovsky,1987;Hobfoll,1989).Assuch,positivepersonalityfactorsactasbuffersthatprotectwell-beingagainstvariousstressors(e.g.,Kivimäkietal.,2005;Mäkikangas&Kin-nunen,2003).Personalityandwell-beingmayalsobelinkedthroughhealthbehaviororphysiologicalmechanism(Antonovsky,1987;Williams,Smith,&Cribbet,2008).Evidenceindicates,forexample,thatoptimismisrelatedtoahealthylifestyle(e.g.,Giltay,Geleijnse,Zitman,Buijsse,&Kromhout,2007)andtophysicalsymptoms(Scheier&Carver,1985).Geneticindividualdifferencesmayactasathirdvariableintherelation-shipbetweenpersonalityandhealth(e.g.,Friedman&Booth-Kewley,2003;Williamsetal.,2008),aswell.Asthefocusofthisstudywasonthesocialpsychologicaldeterminantsofemployeewell-being,itdidnotexaminephysiological,genetic,andenvironmentalfactors.

4.2.2 Health resources as drivers and outcomes of the motivational process in the Jd-R model

ThebasicassumptionoftheJD-Rmodel(Demeroutietal.,2001)isthatthemotivationalprocessinitiatedbyjobandpersonalresourcesthroughworkengagementmayhaveseveralpositiveorganizationaloutcomes,suchasorganizationalcommitment,jobperformance,andproactivebehaviour(e.g.,Hakanen&Roodt,2010).However,thislongitudinalstudyshowedthatthemotivationalprocessoftheJD-Rmodelmayalsohavehealth-relatedproperties.Moreprecisely,thisstudyshowedthatworkengagementfullymediatedtherelationshipbetweenjobresourcesandpersonalresources(i.e.,self-esteem)onworkability10yearslater,thus

72

4 discUssion

expandingthepotentialoutcomesofthemotivationalprocessincludedintheJD-Rmodel.Moreover,thesefindingscontributetotheworkabilityliterature,whichhasmostlyneglectedmotivationalaspects.Inotherwords,workengagement,supportedbyresourcefuljobandhighself-esteem,playsanimportantroleinmaintainingandpromotingworkability.Hence,apositiveaffectivework-relatedstateofmind–i.e.,workengagement–canbebeneficialintermsofhealth,notonlyintermsofmotivationalbenefits.

Inadditiontotheroleofworkabilityasahealth-relatedoutcomeofthemotivationalprocessoftheJD-Rmodel,thisstudyshowedthatworkabilitymaybeanimportanthealth-relatedresourceitselfthatboostsworkengagementandconsequentlypredictsnotonlydirectlybutalsoindirectlyfutureworkability.Thus,thedualroleofworkabilitywasobserved:bothasapredictor(i.e.,health-relatedresource)andasahealth-relatedoutcomeofthemotivationalprocess.Assuch,thisstudyisoneofthefirstonworkabilitythatfocusesnotonlyontheantecedentsofworkabilitybutalsoonthepositiveconsequencesitmayhave(seealsoFeldt,Hyvönen,Mäkikangas,Kinnunen,&Kokko,2009).Hence,workabilitycanbeconsideredahealth-relatedresourcethatmayhavebeneficialeffectsonemployeewell-beingalsointhelongterm,thussupportingthebuildhypothesisofFredrickson’s(2001)BaBtheory.Inotherwords,workabilityseemsto,viaworkengagement,buildwell-being,i.e.,workability,overtime.

Theroleofworkabilityasahealth-relatedresourcethathasben-eficialeffectsonemployeewell-beingalsointhelongtermhasbeenacknowledgedinsomepreviousstudiesaswell.Forexample,Seitsamoetal.(2011)showedthatworkabilitywasastrongpredictoroflater-lifehealthina28-yearlongitudinalstudyamongFinnishmunicipalwork-ers.Similarly,Ahlstrometal.(2010)foundthatworkabilitypredictedfuturehealthamongwomenworkinginhumanserviceorganizations,withfollow-upsat6and12months.Inturn,intheir10-yearfollow-upstudy,Feldtetal.(2009)showedthatworkabilityofFinnishmanagerswasrelatedtojobinvolvementandorganizationalcommitment–bothconstructsthatarecloselyrelatedtoworkengagement.

However,itshouldbenotedthatinthisstudyitwasnotpossibletodirectlytestthepositivegaincyclehypothesisbetweenworkengagementandworkabilityassuggestedbybothBaBtheoryandCORtheorybe-

73

4 discUssion

causeworkengagementcouldnotbemeasuredatbothtimepoints(seealsoSalanovaetal.,2010).However,theseresultssuggestthepossibilityofsuchpositivereciprocalrelationshipsevolvingovertime.

4.3 mood trajectories were differentially related to well-being

Thisstudyappliedaperson-centeredapproachtoexaminethelong-termchangesinpositiveandnegativeaffectivestates.Bydoingthis,itwaspossibletoidentifyindividualdevelopmentalpathsandtheirimportanceforemployeewell-being.Theseindividualdevelopmentaltrajectoriesdonotbecomevisiblewhenusingvariable-basedapproaches,andtherefore,theperson-centeredapproachwaschosen.Moreover,thisstudyexam-inedthejoint-effectsofaffectivestates(i.e.,vigorandfatigue),andthus,addednewknowledge.

Theresultsindicatedthatdifferentdevelopmentalpathsinvigorandfatiguearepossible.Morespecifically,fourdistinctsubgroupsofindividualsthatdifferedfromeachotherintheirlevelsofpositiveandnegativemoodwerefound.Typically,firefightershadaratherhighlevelofvigorandlowleveloffatigue.Assuch,theseresultsindicatethathav-ingapositiveaffectivestateismorecommonthanexperiencingnegativefeelings(seealsoDiener&Diener,1996).However,othercombinationsofpositiveandnegativeaffectivestatesalsoexisted.Thesefindingsgivepartialsupporttoboththeindependenceandbipolarhypothesesofaffect.Ontheonehand,vigorandfatigueseemtodevelopintandem(Caccioppo&Berntson,1994).Moreprecisely,theco-occurrenceofvigorandfatiguewasparticularlyapparentamongthemembersofthehighpositivebutdecreasingtrajectoryinwhichvigorandfatiguewerealmostatthesamelevelbetweenT2andT3.Thisisinlinewiththeindependencehypothesis,whicharguesthatanincreaseinoneaffectdoesnotnecessitateadecreaseintheother(Cropanzanoetal.,2003;Mäkikangas,Feldt,Kinnunen,&Tolvanen,2012).Ontheotherhand,thebipolarviewofaffectasoppositeendsofthesamecontinuum(e.g.,Russell&Carroll,1999)alsofoundsomesupport.Inparticular,inthemosttypicaltrajectory,i.e.,thehighpositive,vigorandfatigueapparentlyrepresentedoppositetrends;thatis,membersofthislatentgrouphad

74

4 discUssion

thehighestlevelofvigorandthelowestleveloffatigueofalltrajectories,thereforesuggestingbipolarityofpositiveandnegativeaffectivestates.Theseresultssuggestthatpositiveandnegativeaffectivestatesmaynotalwaysbedirectoppositesorindependenteither.Infact,presentknowl-edgesuggeststhatbothapproaches–independenceandbipolar–maybevalidasbothofthemfoundsomeempiricalsupport,andthustheremaybenoneedtomakean“either-or”conclusionbetweenthem(Reich,Zautra,&Davis,2003).

Accordingtotheresultsofthisstudy,affectivestatesweremoderatelystableovertime;however,somechangeswereidentified.Inthehighpositivebutdecreasingtrajectoryinparticular,moodschangedsignifi-cantlybetweenthetwofirstmeasurementpoints.Infact,inallmoodtrajectories,changes–mostlynegative–weremoreobviousbetween1996and1999.Moreprecisely,adeclineinpositivemoodandanin-creaseinnegativemoodwereapparentduringthisthree-yearperiod.Incontrast,changesbetween1999and2009wereratherminimal,asalltrajectoriesfollowedarelativelystablelineduringthisten-yearperiod.Speculatively,onereasonfortheincreaseinnegativemoodduringthefirstthreeyearsandthepersistentpatternthereaftermightbefoundintheFinnishfireandrescueservices’situationatthattime,namelythestrikesinthespringof1995andwinterof1997–1998.Thesestrikeshadaconsiderableandlong-reachingeffectonfiredepartments’workcommunities(seeMankkinen,2011),andconsequently,presumablyalsoonemployees’affectivestates.Nevertheless,theoverallchangesinmoodtrajectoriesduringthe13-yearfollow-upperiodwereratherminimal.Atleastpartly,thestabilityofmoodstatesovertimecanbeexplainedbythetrait-natureofmoods.Asatrait,moodisregardedasastableaffectthatdoesnotchangeovertime.Therefore,thesefindingssupporttheideaofmoodsasratherlastingandenduringaffectivestates(Brief&Weiss,2002;Cropanzanoetal.,2003).Theratherhighstabilityofmoodstatesmayalsobeexplainedbypersonalityfactors.Forexample,temperamenttraits(i.e.,relativelystableaspectsofpersonality,e.g.,Thompson,Winer,&Goodvin,2011)havefoundtoberelatedtoaffectivestates,suchasmoodsandemotions(Bates,Goodnight,&Fite,2008;Goldsmith,1993).Therefore,changesinmoodsmaybeexpectedtobeminimalovertime.Moreover,asemotionalregulationincreaseswithage(e.g.,Grossetal.,1997),thiscouldalsoexplaintheincreasedstabilityofmoodstates

75

4 discUssion

amongfirefighters.Moreover,theresultofrelativelylittlechangesinvigorandfatigueovera13-yearfollow-upperiodisinlinewithpreviousstudiesthathaveshownratherhighstabilityinaffectivestatesovertime(e.g.,Hakanenetal.,2008b;Seppäläetal.,2009).

Byexaminingtherelationshipsbetweenmoodtrajectoriesandwell-being,thisstudyshowedthatthosewithahighlevelofpositivemood(i.e.,vigor)andlowlevelofnegativemood(i.e.,fatigue)hadbetterwell-beingthanthosewithahighlevelofnegativeaffectivity.Theoretically,thisresultisinlinewithFredrickson’s(2001)BaBtheory,whichsuggeststhatpositiveemotionsarerelatedtoincreasedwell-being,evenovertime.Overall,theseresultshighlighttheimportanceofaffectsinpredictingworkability,lifesatisfaction,andself-ratedhealth.Empirically,thesefindingsareinlinewithpreviousstudiesthathaveshownthepositiverelationshipbetweenvigorandhealth-relatedoutcomes(e.g.,Shirometal.,2008),andthedetrimentaleffectofnegativeaffectivestatesonhealth(e.g.,Salovey,Rothman,Detweiler,&Stewart,2000).

Takentogether,itseemsthatdifferentdevelopmentalpathsandchangesinpositiveandnegativemoodstatesarepossibleovertime.Therefore,theresultssuggestboththepossibilityandneedtofosterpositivedevelopmentalpathsinaffectivestates,i.e.,toincreasevigorandpreventfatigueatwork,inordertomaintainandimproveemployeewell-being.

4.4 developmental paths in musculoskeletal pain and depressive symptoms, and health-enhancing effects of job resources and individual factors

Althoughtheantecedentsofmusculoskeletalpainanddepressivesymp-tomsarelargelyknown,theknowledgeregardingthedevelopmentalpathsofMSPandDPSoverlongtermislimited,aswellastheeffectsofjob-relateddemandsandresources,personalresources,andlifestyleontheco-developmentofthesesymptoms.Thisstudyrevealedthatdifferentdevelopmentalpathsinmultisitemusculoskeletalpainandde-pressivesymptomsarepossible.Threedistinctpain-depressivesymptomtrajectorieswereidentified,i.e.,LowSymptoms,HighPain,andHigh

76

4 discUssion

Depression,thussupportingthehypotheses.However,thehypothesisconcerningatrajectorywithhighlevelsofbothsymptomswasnotsup-ported.ItislikelythatfirefighterswithhighlevelsofMSPandDPSdroppedoutfromworkandthustheavailabledata.

BothsignsofstabilityandchangeinMSPandDPSwereobserved.Thedifferentdevelopmentalpathsinmusculoskeletalpainanddepres-sivesymptomsweremostevidentinHighDepression.Inthistrajectory,changesinDPSwereevident,whereasthelevelofMSPremainedratherstablethroughoutthefollow-up,suggestingseparatedevelopmentsofthesesymptoms.Assuch,theseresultsareinlinewithpriorassumptionsofdistinctdevelopmentofmusculoskeletalpainanddepressivesymp-toms(e.g.,Magnietal.,1994).Incontrast,intheHighPaintrajectory,increasesinbothMSPandDPSwereobviousthroughoutthestudyperiod,indicatingaparalleldevelopmentofthesesymptoms.Overall,insteadofchangesinDPSintheHighDepressiongroup,ratherhighstabilityinmusculoskeletalpainanddepressivesymptomswereobserved.

Theresultsindicatedthatjobdemands,jobandpersonalresources,andlifestylewerepartlydifferentlyrelatedtopain-depressiontrajecto-ries.Bothjobdemands(i.e.,mentalworkload)andallinvestigatedjobresources(i.e.,supervisoryrelations,interpersonalrelations,andtaskresources)aswellaspersonalresources(i.e.,optimism),andlifestyle(i.e.,sleepingproblems)wererelatedtoHighDepression,whereasonlysomeofthesefactorspredictedbelongingtotheHighPaintrajectory.However,solelyHighPainwaspredictedbyalcoholconsumption.Thesefindingsargueapartlydifferentmechanismintheetiologyofmusculoskeletalpainanddepressivesymptoms.Consistentwiththesefindings,somepreviousfindingsindicatethatpainanddepressionhaveatleastpartlydifferentriskfactors(e.g.,Linton&Bergbom,2011;Miller&Cano,2009).TheresultsalsoshowedthattheantecedentsweremorestronglyrelatedtoHighDepressionthanHighPainastheriskestimateswerenotablyhigherforHighDepression.Thisindicatesthatimprovingjobresourcesandindividualfactorsmaybemorelikelytohaveaneffectondecreasingdepressivesymptomsratherthanonmusculoskeletalpain.

ThefindingsofthisstudyareinlinewiththeJD-Rmodel(Demeroutietal.,2001)thatemphasizesthedetrimentaleffectofjobdemandsonemployeewell-being.Obviously,highjobdemands,suchasaheavymentalworkload,mayactaspsychosocialstressorsatworkthatcause

77

4 discUssion

psychosocialstressreactions,andconsequently,anincreaseindepressivesymptoms.Similarly,ahighmentalworkloadmayleadtoconstantoverloadandeventually increasemusculoskeletalpain.However,contrarytohypothesisandpriorfindings(e.g.,Christensen&Knardahl,2010;Haukka,Ojajärvi,Takala,Viikari-Juntura,&Leino-Arjas,2012;Neupane,Miranda,Virtanen,Siukola,&Nygård,2013)ahighphysicalworkloadwasnotsignificantlyrelatedtopain-depressiontrajectoriesinthissample.Thiscouldatleastbeexplainedinpartbytherathersimilarworktasksoftheparticipants:difficultworkpostureswerepartofthejobforallfirefighters.Alsohealth-basedselectionfromemployment,andhencefromthedata,mayhaveweakenedtheassociations,therebyleadingtotheunderestimationoftheactualeffects.

Similarly,thislongitudinalstudygivesmoresupporttotheassump-tionsoftheJD-Rmodel(Demeroutietal.,2001)oftheenhancingeffectsofjobresourcesandpersonalresourcesonemployeewell-being.Indeed,supportinghypothesis,jobresourcesweresignificantpredic-torsofpain-depressiontrajectoriesandthiswasparticularlyapparentinrelationtoHighDepression.Moreprecisely,ininvestigatingthepredic-torsseparately,alackofallexaminedjobresourceswasrelatedtoHighDepression,whereasonlypoorinterpersonalrelationswererelatedtoHighPain.Inthefinalmodel,however,interpersonalrelationswastheonlysignificantpredictoramongjobresources,andfurthermore,solelyinrelationtoHighDepression.Indeed,convincinghealth-promotingeffectsofsocialrelationshipsandsocialsupporthavepreviouslybeenshowninsocialpsychologicalstudiesonhealthandwell-being(forreviews,see,Cohen,1988;Uchino,2004).Similarly,severallongitu-dinalstudiessuggestthatlowsocialsupportfromco-workersmaybeanimportantantecedentofdepression(e.g.,Paterniti,Niedhammer,Lang,&Consoli,2002;Stoetzeretal.,2009,seealsoHakanenetal.,2008b;Haukkaetal.,2011).Alongwiththeseresults,thefindingsofthecurrentstudysuggestthatgoodrelationshipsbetweenco-workerscanpositivelyinfluenceemployeewell-being,eveninthelongterm.Goodinterpersonalrelationsarehighlyvaluedinfirefighting,inwhichemployeesworklonghourstogetherincloselycoordinatedteams(seealsoPillai&Williams,2004).Ontheotherhand,theimportanceofsupervisoryrelationsmayremainlessinfluentialbecauseofthehighlyvaluedgoodrelationsbetweenco-workers.

78

4 discUssion

Accordingtothisstudy,personalresourcesarealsoimportantinex-plainingemployeewell-being,asalackofoptimismpredictedbelongingtotheHighDepressiontrajectory.Similarfindingsareshowninpreviousstudies,however,oftenwithashortertimelag,ontheimportanceofpersonalfactorsinexplainingwell-being(e.g.,Cannella,Lobel,Glass,Lokshina,&Graham,2007;Scheier&Carver,1992;Xanthopoulouetal.,2007).Optimismmayenhanceemployeewell-beingviaseveralpathways.First,aperson’sviewoftheworld–here,optimistic–mayberelatedtothelikelihoodofexperiencingcertainphysiologicalresponseswhenconfrontedbyenvironmentalchallenges,andthusbedirectlyre-latedtohealth(Friedman&Booth-Kewley,2003);inthiscase,tolowlevelsofdepressivesymptoms.Second,optimismmayactasabuffer.Moreprecisely,optimistsarelikelytoexperiencepositivefeelings,andwhenconfrontedbystressfulevents,arelikelybeconfidentandpersistent,whichisprotectiveofhealth.Incontrast,pessimistsaremorelikelytobedoubtfulandhesitant,whichmayleadtohealthrisks,andconsequentlytoill-health.(Carver&Scheier,2009.)

Inadditiontotheimportanceofworkcharacteristicsandpersonalresources,theresultsofthisstudysuggestthatalsogoodsleepisessentialforemployeewell-beingintermsofpositivedevelopmentinMSPandDPS.Similarly,inpreviousstudiestheimportanceofgoodsleepinpre-ventingpain(Kamaleri,Natvig,Ihlebaek,Benth,&Bruusgaard,2008)anddepression(Baglionietal.,2011)havebeenfound.Inparticular,amongfirefighterswhoworklonghours,adequatesleepisessentialformaintaininggoodphysicalandmentalhealth.Oftheotherlifestylefactors,onlyhighalcoholconsumptionwasasignificantpredictorofahighlevelofmultisitemusculoskeletalpain.Incontrast,noneoftheexaminedlifestylefactors,exceptsleepproblems,predictedbelongingtothetrajectorywithahighlevelofdepressivesymptoms.Nevertheless,similarfindingsofnoeffectsoflifestyleonpainanddepressionhavebeenfoundpreviously(e.g.,Haukkaetal.,2012;Hölzel,Härter,Reese,&Kriston,2011;Lebouef-Yde,1999).

Toconclude,inthepresentstudy,job-relateddemandsandresourceswerefoundtobemoreimportantthanlifestyleforemployeewell-being.Thus,decreasingdemandsandincreasingresourcesisworthwhileinpreventingthedevelopmentofmultisitemusculoskeletalpainandde-pressivesymptoms.

79

4 discUssion

4.5 methodological considerations

Thisstudyhasbothstrengthsandlimitationsthatneedtobeconsidered.Themainstrengthsofthestudylieinthelongitudinaldesignandtheuseofvarioussophisticatedstatisticalanalyses.Thelongitudinaldatasetwiththreewavesanda13-yeartimelagwasused.Thelongitudinalda-tasetenabled,forexample,studyingthedevelopmentalpathsovertime,andclaimingforcausality(StudiesIII–IV).Inaddition,applyingawidevarietyofstatisticalmethods(i.e.,linearregressionanalysis,structuralequationmodeling,latentclassgrowthmodeling,mixedlinearmodels,andmultinomialregressionanalysis)canbeseenasanotherstrength.Inparticular,theuseofafairlynewmethodtostudyintra-individualchangesinemployeewell-being(i.e.,latentclassgrowthmodeling)overtimecanbeconsideredasvaluable.Byapplyingtheperson-centeredap-proach,theheterogeneityinpositiveandnegativeaffectivestates,aswellasinmultisitemusculoskeletalpainanddepressivesymptoms,couldbeidentified,andthuscapturethedevelopmentofintra-individualtrajec-toriesoverthe13-yearfollow-upperiod.Finally,thisstudyfocusedonseveralimportantworkcharacteristicsandindividualfactorsandtheirrelationshiptodiversewell-beingoutcomeswithvalidatedmeasures.

Nevertheless,thisstudyhassomelimitationsthatshouldbenoted.Firstofall,thestudywasbasedonself-reportmeasures,whichmaycausesystematicmeasurementerrors(commonmethodsvariance).However,thelongitudinaldesignusedinthecurrentstudymaydiminishtheriskforcommonmethodbias(Doty&Glick,1998).AlsoinStudyI,whichwascross-sectional,itwaspossibletocontrolforbaselineworkability,andthus,diminishtheriskofcommonmethodbias.Nevertheless,fu-tureresearchwouldbenefitfromapplyingmoreobjectiveindicatorsofworkcharacteristics,individualfactors,aswellaswell-beingoutcomes.

Secondly,thestudydesignregardingpsychosocialfactorsinthequestion-nairehadsomelimitations.Forexample,self-esteemandworkengagementweremeasuredonlyonce.Thus,nocausalrelationshipbetweenvariablescouldbedetermined.Obviously,inthecaseofworkengagement,meas-uringitearlierwasnotpossibleasthemeasurewasdevelopedonlyafterthetwofirstmeasurementpoints(Schaufeli&Bakker,2003;Schaufelietal.,2002).Nevertheless,futureresearchshouldapplyafullpaneldesignincludingalljobandpersonalresourcesmeasuredlongitudinally.

80

4 discUssion

Thirdly,theconceptualizationofsomeofthemeasuresmaybecon-sideredashortcoming.Forexample,optimismandlifesatisfactionwerebothmeasuredwithasingle-itemindicatorwhichmaydiminishtheirreliabilityandvalidity.Inpreviousstudies,however,single-itemindicatorsofbothoptimismandlifesatisfactionhavebeenusedandfoundnearlyasgoodasindicatorswithseveralitems(e.g.,Carveretal.,1994;Lucas&Donnellan,2012).Inaddition,themeasureofdepressivesymptoms(i.e.,POMSdepressionsubscale)capturedsolelytheemotionalsymp-toms,notcognitiveorphysicalsymptomsofdepression.Moreover,onlydepressivesymptoms,notdepressionassuch,wereassessed.However,depressivesymptomsmeasuredbyPOMSarestronglycorrelatedwiththewell-establishedBeckDepressionInventory(Griffithetal.,2005),andtherefore,aplausiblemeasuretouse.Inaddition,itwouldbedifficulttomeasuretrajectoriesofobjectivelymeasureddepression.

Fourthly,the3and10-yeartimelagsusedinthisstudymaynotbeoptimalforinvestigatingtherelationshipsbetweenworkcharacteristics,personalresources,lifestyle,andwell-being,asotherprocessessuchasorganizationalchangesmayhaveinfluencedtheeffectofindependentvariablesontheoutcomes.Ingeneral,suchlongtimelagsmayleadtoanunderestimationofthetruecausalrelationshipbetweenstudyvariables(Zapf,Dormann,&Frese,1996).However,despitethechangesintheorganizationalstructureinFinnishfiredepartments,theworkenviron-mentsandcolleaguesforthemostpartremainedthesame.Moreover,organizationsareinfactinseveralwaystryingtopursuethestabilityagainstthechangeswithintheorganization(Leana&Barry,2000).Inrelationtothefourthlimitation,theeffectsizeswererelativelysmall.Forexample,inStudyII,theexplainedvarianceinworkengagementwas12%andinworkability21%,respectively.Nevertheless,theseeffectsareinlinewithpriorstudies(e.g.,Mauno,Kinnunen,&Ruokolainen,2007;Xanthopoulou,Bakker,Demerouti,&Schaufeli,2009).Inaddi-tion,evenrelativelysmalleffectsizesmaybesalientinpredictinghealthandwell-beingofemployees(Ford,Woolridge,Vipanchi,Kakar,&Strahan,2014).Furthermore,workcharacteristics,personalresourcesandlifestyleareonlyoneofthemanyvariablesaccountingforemployeewell-being.Physiological,genetic,andenvironmentalfactorsarealsorelatedtohealthandwell-being.Moreover,non-workcharacteristicssuchasfamily,closecommunity,andsocietyareassumedtoberelated

81

4 discUssion

toworkability(Ilmarinen,2006).However,becauseoftherelativelyhomogenousgroupoffirefighters,forexample,socio-economicandculturalfactorswerenotexamined.Nevertheless,inordertoincreasetheexplainedvariance,morefactorsshouldbeincludedintheresearchmodels.Forexample,examiningthework-familyinteractionorhomeresourcesamongfirefighterscouldbeexamined.Ontheotherhand,thesignificantrelationshipsbetweenthestudyvariablesevenovertheratherlongfollow-upperiodare,infact,indicativeoftherobustnessofthefindings.Nevertheless,infuturestudies,ashortertimelagandafullpaneldesignshouldbeused.

Fifthly,theratherhighnumberofdrop-outsmaybeconsideredalimitation.However,thedifferencesbetweentheparticipantsandthedrop-outswereeithernon-significantorratherminimalineverysub-study.Forthemostpart,thedrop-outwasprobablyduetoretirementasaresultofthelowretirementage(55years)amongFinnishfirefightersduringthestudyperiodandearlyretirementschemesandpersonalearlyretirementarrangements(under55yearsofage)whicharestillpossibleroutesforretirement.Anincrementalincreaseinactualretirementagehasonlyrecentlyoccurred.Therefore,thedrop-outfromthesamplecanberegardedasnormalandnotcausinganyparticularbiastotheresults.

Sixth,thehealthyworkereffect(e.g.,Li&Sung,1999)mayhavebiasedtheresults,especially,asgoodmentalandphysicalhealthisre-quiredoffirefighters.Therefore,thenullassociationfoundinthisstudybetweenalcoholconsumptionandworkability,forexample,maybeatleastpartlyexplainedbythehealthyworkereffect;thatis,heavydrinkerswithlowworkabilitymayhavedroppedoutofthesample.However,theparticipantsofthepresentstudydidinfacthavelowerworkabilitythanFinnishmaleemployeesingeneral(Airila,Hakanen,Punakallio,Lusa,&Luukkonen,2012),andthusitcanbeassumedthatnoconsiderablehealthyworkereffectexisted.

Finally,thisstudyfocusedononeprofessiononly–firefighters–albeitarepresentativesampleatbaseline,whichmaypotentiallythreatenthegeneralizabilityofthefindingstootheroccupationalgroups.Moreover,alltheparticipatingfirefightersweremen,andthismaypossiblyhaveaffectedthefindings.Obviously,genderaffects,forexample,personal-itycharacteristics(e.g.,Costa,Terracciano,&McCrae,2001;Feingold,1994;Mäkikangas&Kinnunen,2003)andwell-being(e.g.,Fujita,

82

4 discUssion

Diener,&Sandvik,1991;Hakanen,2009).However,theaimofthepresentstudywastotesttheoreticalmodelsandexaminetherelation-shipsbetweendemands,resources,lifestyle,andwell-being,andinthiscase,thereisnoreasontoassumethatthemodelswouldworkdifferentlyaccordingtogender.Althoughsomecautionisneeded,Ibelievethattheresultscanalsobeextendedtootheroccupationalsectors.Infact,similarevidenceoftherelationshipsbetweenworkcharacteristics,individualfactors,andwell-beingexistsinvariousoccupationsectorsandcountries(e.g.,Hakanenetal.,2006;Salanova&Schaufeli,2008).Moreover,asjobandpersonalresourcesweresignificantlyrelatedtoemployeewell-beingeveninahighlyphysicallydemandingjob,i.e.,firefighting,itisplausiblethatthesameeffectsarealsolikelytobefoundinotheroc-cupationalsectors.Ofcourse,thisremainstobetested.

4.6 implications for future research

Theoretically,thecurrentstudylargelysupportedtheassumptionsoftheJD-Rmodel,CORtheory,andBaBtheory.Atthesametime,however,thisstudyextendedcurrentknowledgebyincludingthehealth-relatedoutcome−i.e.,workability−tothemotivationalprocessoftheJD-Rmodel,aswellasfoundsupportforthelinkbetweenworkengagementandworkability.Infuturestudies,however,otherhealth-relatedvariables,inparticularphysiologicalhealth,shouldbeexaminedasendpointsofthemotivationalprocessusingalongitudinaldata(seealsoBakker&Leiter,2010).Inaddition,objectivemeasuresofemployeewell-being,suchassicknessabsence,andhealth-relatedregisters(e.g.,NationalHospitalDischargeRegister,NationalPrescriptionRegister),couldbeusedinstead,oralongside,self-reportsofwell-being.

Furthermore,ratherthanfocusingonlyontheindividuallevel,employeewell-beingshouldbeexaminedatteamlevel,too.Indeed,teamworkhasacrucialroleinemployeehealthandwell-being(e.g.,Wilson,DeJoy,Vandenberg,Richardson,&McGrath,2004),whichalsobecameapparentinthisstudywhichfoundsupervisoryrelationsandinterpersonalrelationstoberelatedtowell-beingoffirefighters.Thereisalreadysomeevidence,althoughmostlycross-sectional,thatworkengagementactsasacollectivepsychosocialconstructthatemergesfromtheinteractionandsharedexperiencesofthemembersofawork

83

4 discUssion

team(Salanova,Llorens,Cifre,Martínez,&Schaufeli,2003),andwhichmayhavepositiveorganizationalconsequences,suchasincreasedtaskperformance(Salanovaetal.,2005),collectivepositiveaffectivestate,andcollectiveefficacybeliefs(Salanova,Llorens,&Schaufeli,2011).Byexaminingemployeewell-beingasateam-levelprocess,itispossibletoidentifyteam-basedresourcesthatareessentialintermsofhealthandwell-being.Forexample,infirefighting,whichrequiresgoodteamworkandsupportiveinterpersonalrelations,identifyingteam-levelresourcesthatmaypromotewell-beingwouldbevaluable.Therefore,theroleoftheteaminemployeewell-beingshouldbeinvestigatedfurther,particularlyusinglongitudinaldatathatallowsthepossibilitytoclaimforcausality.

Asdescribedearlier,theJD-Rmodelcanbeconsideredaflexiblemod-elthatincludesawidevarietyofjobdemands,resourcesandoutcomes,andthus,issuitableandusableindiverseworkcontexts.Nevertheless,fromamethodologicalpointofview,changesinworklifeintermsof,forexample,networking,flexibility,internationalization,andmobility,setnewrequirementsforthedefinitionsofdemandsandresources.Forexample,inastudyamongvolunteerfirefighters,camaraderie–i.e.,feelingsofbelonging,asenseofsharedidentity,reciprocaltrust,andpositivebondsthatexistwithincohesiveworkgroups–hasbeenfoundtobearesourcethatmayenhancetheirhealthandwell-being(Tuckey&Hayward,2011).Hence,inordertocapturethefeaturesofaparticularworkenvironmentintheconstantlychangingandfragmentedworldofwork,theconceptualizationsofdemandsandresourcesusedinparticularoccupationalsectorsshouldbeconsideredfurther.

Moreover,theroleofthepersonalresourcesintheJD-Rmodelneedstobestudiedfurther.Inadditiontotheroleofpersonalresourcesasdirectlyimpactingwell-being,personalresourceshavebeenintegratedintotheJD-Rmodelasmediators,moderators,“thirdvariables”,andantecedentsofjobdemandsandresources,oranothercombinationofthese(formoredetails,seeSchaufeli&Taris,2014).Inthepresentstudyonlythedirectlinksbetweenpersonalresourcesandwell-beingwereexamined(i.e.,fromself-esteemtoworkabilityviaworkengage-mentandfromoptimismtopain-depressiontrajectories).Theresultsrevealedthatthestrengthoftheassociationbetweenpersonalresources(i.e.,self-esteem)andwell-being(i.e.,workability)wasofasimilarsizetothatofjobresources,whereastheeffectofoptimismonwell-being(i.e.,

84

4 discUssion

pain-depressiontrajectories)wassomewhatweakercomparedtothatofjobresources.Generally,previousstudieshavefoundmoresupportfortheimportanceofjobresourcesinexplainingemployeewell-being(e.g.,Rodríguez-Sánchez,Hakanen,Perhoniemi,&Salanova,2013;Schaufeli&Janczur,1994),however,insomecircumstancespersonalresourcesmayplayamoresignificantrole.Forexample,HakanenandLindbohm(2008)foundthatoptimismasapersonalresourcewasmoreimportantforcancersurvivors’well-beingthanthatofjobresources,whereasfortheirreferentstheeffectofjobresourceswasmorecrucial.Obviously,theimportanceofpersonalresourcesinexplainingemployeewell-being,andonthewhole,theroleofpersonalresourcesintheJD-Rmodel,needstobeexaminedmoresystematicallyusinglongitudinaldata.Moreover,itisnoteworthythatofthewidevarietyofpersonalitytraits(e.g.,DeNeve&Cooper,1998),onlytwopersonalresources–self-esteemandoptimism–wereexaminedinthisstudy.Therefore,theimportanceofotherpersonalresources,suchasself-efficacy,onwell-beingamongfirefightersshouldbeexamined.Somepreviousfindingsindicatethathighself-efficacyisaprotectivefactorofstressanddepressivesymptomsamongfirefighters(Regehretal.,2003),however,therelationshipbetweenself-efficacyandworkabilityremainsunknown.

Inaddition,theroleofjobdemandsintheJD-Rmodelremainssomewhatunclear.Recentresearchhasdiscussedtheredefinitionofjobdemandsanddistinguishingbetweenchallengedemands(e.g.,workload,timeurgency,mentaldemands,andresponsibility)andhindrancedemands(e.g.,roleambiguity,jobinsecurity,andconstraints)(Crawfordetal.,2010;vandenBroek,DeCuyper,DeWitte,&Vansteenkiste,2010;Webster,Beehr,&Christiansen,2010).Itissuggestedthatchal-lengedemandshaveamotivationalpotentialandareactuallypromotingwell-beingratherthandiminishingit.Contrastingly,hindrancedemandsaremorelikelytothwartpersonalgrowth,learningandgoal-attainment,andthus,benegativelyrelatedtoemployeewell-being.(Lepine,LePine,&Jackson,2004;Podsakoff,LePine,&LePine,2007.)Basedonthisdistinction,theindicatorsofjobdemands(i.e.,mentalandphysicalworkload)ofthepresentstudycouldbedeterminedaschallengedemandsthatwouldbelikelytoincreaseworkengagementbutalsostressandothernegativeindicatorsofwell-being.Thus,inthefuture,thedistinc-tionbetweenchallengeandhindrancedemandsofthefirefighterscould

85

4 discUssion

beexaminedinordertofindoutwhethercertaindemandscouldhavemotivationalfunctions.Nevertheless,definingdemandsashindrancesandchallengesisnotsimpleasindividualsmayevaluatedemandsdiffer-ently:forsomephysicalormentalworkloadmaybeachallengewhereasforothersarealhindrance.Moreover,hindrancesandchallengesmayhaveadifferentmeaningindiverseworkcontexts.Nevertheless,thisremainstobestudied.

Thecurrentstudyhasviewedemployeewell-beingasamultidimen-sionalconstructcoveringaffective,health-related,andcognitive(i.e.,lifesatisfaction)aspects.Assuch,bothpositiveandnegativesidesaswellasjob-relatedandcontext-freeaspectsofwell-beingwereabletocapture,andthus,followingtheprevioustheoreticalconceptualizationsofemployeewell-being(e.g.,Dieneretal.,1999;Schaufelietal.,2002;Schaufeli&Bakker,2004;vanHornetal.,2004).However,futurestudiesshouldfurtherdeveloptheconstructofemployeewell-beingthatcomprehensivelycapturesphysical,mental,andsocialaspectsofemployeehealthandwell-being.

4.7 Practical implications

Theresultsofthisstudyindicatethatjobdemands,jobandpersonalresources,andlifestylearerelatedtoemployeewell-beingevenoverthelongterm.Therefore,workplaceinterventionsaimingatimprovingwork-ingconditions,interpersonalrelations,aswellasinterventionstargetedonindividualscouldbebeneficialinordertomaintainandimproveemployeewell-being.

Attheworkplacelevel,interventionsaimedatimprovingboththephysicalandpsychosocialworkenvironmentcouldbeapplied.Thereissomepriorevidenceindicatingtheeffectivenessofworkplaceinterven-tionsonemployeehealthandwell-being.Forexample,interventionstargetingpsychosocialworkcharacteristicsarefoundtobeeffectiveinpreventingemployeehealthproblems(e.g.,Bourbonnais,Brisson,Vinet,Vézina,Abdous,&Gaudet,2006),increasingjobsatisfactionandde-creasingabsenteeism(Parks&Steelman,2008),aswellasincreasingautonomy,strengtheningsocialties,andreducinginterpersonalconflict(McLeroy,Gottlieb,&Heaney,2002).Someevidenceoneffectivework-

86

4 discUssion

placeinterventionsinfiredepartmentsalsoexists.Forexample,team-basedinterventionsandmotivationalinterviewinghadpositiveeffectsonfirefighters’nutritionbehavior,BMI,andphysicalactivityoneyearafterinterventionsamongfirefightersintheUnitedStates(MacKinnonetal.,2010).Similarly,previousworksitehealthpromotioninterventionsamongthesamestudypopulationincreasedco-workercohesion,exercisehabits,andhealthbehaviorsamongfirefighters,andreducedtheirfeel-ingsofdepression(Elliotetal.,2004).AlsointheUnitedStates,actionresearchhasshowntobeapotentialmechanismforreducingburnoutamongfirefighters(Halbesleben,Osburn,&Mumford,2006).How-ever,arecentreviewofthehealtheffectsofwellnessprogramsatpoliceandfirefighterworkplacesindicatedmixedresultsregardinghealthandwell-beingoutcomes(Rachele,Heesch,&Washington,2014).Thus,theresearchersconcludedthatempiricalevidenceforeffectivehealthinterventionsamongfirefightersremainscarce,andtherefore,furtherresearchisneeded.

Forexample,interventionstargetinginterpersonalrelationsbyim-provingcommunicationskills,trainingsessionsonteamworkandteamsupport,increasingcommunicationbetweenemployees,orincreasingemployees’participationinthedecisionmakingprocesscouldbeim-plemented.Similarly,interventionstodecreasejobdemands,improvesupervisory/followerrelationshipsandincreasetask-relatedresourcescouldbeperformed.Obviously,activeinvolvementofbothemployeesandmanagersneedstobeensured(Nielsen&Abildgaard,2013).Inaddition,employeesmayalsothemselvesredesignandcustomizetheirworkingconditions,thatistosay,“craft”theirjobs,tobetterfittheirmo-tives,strengths,andpassions.Assuch,jobcraftingreferstoself-initiatedchangesthatemployeesmakeintheirownjobdemandsandresourcesinordertoattainand/oroptimizetheirpersonalgoals.Byjobcrafting,itbecomespossibletoseekresourcesandchallengesandreducedemands.(Tims,Bakker,&Derks,2012.)Craftingone’sjobmayincludechang-ingtask,relational,orcognitiveboundariesofthework(Wrzesniewski&Dutton,2001).Therefore,firefighterscouldaccordingtotheirownneeds,forexampletosomeextent,expandordiminishthescopeoftheirtasks,changethesocialrelationshipsatwork,alterthewaytoperceivetheirtasks,orchangehowtoperformtheirtasksincaseswhichallowsvariation.However,notnecessarilyinrescueandfiresituationswhich

87

4 discUssion

arehighlyinstructedandcontrolled.Empiricalevidenceshowsthatjobcraftingbothatindividualandteamlevelshasapositiveimpactonemployeewell-beingandperformance(Bakker,Tims,&Derks,2012;Tims,Bakker,Derks,&VanRhenen,2013),andtherefore,moreeffortonlaunchingbottom-upworkplaceinterventionsisessential.

Asfindingsfromthisstudyindicate,personalresourcesandindi-viduallifestylearealsorelatedtoemployeewell-being.Thus,promotingself-esteem,self-knowledge,optimism,andahealthylifestyleamongfirefightersarealsoimportantandpossibletargetsforinterventions.Moreover,asthisstudyidentifiedseveralindividualdevelopmentalpathsbothintermsofaffectivestatesandinmusculoskeletalpainanddepres-sivesymptoms,inordertoincreaseefficiencyininterventions,itcouldbebeneficialtoidentifygroupsofpeoplethathavesimilarsituationinquestionandtargetinterventionstothem.

Tosumup,inworkplaceinterventions,bothhealth-impairmentandmotivationalprocessesoftheJD-Rmodelshouldbeinfluenced,thatis,decreasingill-healthandincreasingwell-beingbyapplyingtoolsforbothtodecreasedemandsandincreaseresources.However,featuresofthejobneedtobeconsidered,inthiscasethecharacteristicsofrescueservices.Indeed,enhancingemployeewell-beinginrescueserviceswithmultiple(i.e.,physical,mentalandsocial)demandsandspecialfeaturesoforganizationalcultureandstructure(i.e.,workingina24-hourshiftsinahighlyhierarchicalorganization),andtheunpredictablenatureofwork,mayberatherchallenging.

4.8 Conclusions

Thepurposeofthisstudywastoexaminetheeffectsofworkcharac-teristics,personalresources,andlifestyleonemployeewell-beingusinglongitudinaldata.Thebeneficialroleofjobresourcesandpersonalre-sourcesinexplainingemployeewell-beinggainedfurthersupportfromthislongitudinalstudy,andthus,supportedtheassumptionsoftheJD-Rmodel,CORtheory,andBaBtheory.Addinganewcontribution,thisstudyexaminedtherelationshipbetweenworkengagementandworkability,andfoundthatworkengagement–asapositivework-relatedaffectivestate–isassociatedwithworkability.Moreover,workability

88

4 discUssion

hadadualroleinthemotivationalprocessoftheJD-Rmodel:first,asahealth-relatedoutcome,andsecondly,asahealth-relatedresourcethatmayboostworkengagementandconsequentlypredictnotonlydirectlybutalsoindirectlyfutureworkability.Hence,thepresentstudyfoundlong-termevidenceforthehealth-relatedpropertiesofthemotivationalprocessoftheJD-Rmodel.

Furthermore,usingtheperson-centeredapproachinexaminingem-ployeewell-beingovera13-yearfollow-upperiod,thisstudyprovidednewinformationonindividualdevelopmentalpathsinpositiveandnegativeaffectivestatesaswellasinmultisitemusculoskeletalpainanddepressivesymptoms.Firstly,differentdevelopmentalpathsinvigorandfatiguewerefound,andconsequently,thosewithahighpositiveaffectivestatehadbetterwell-beingintermsofworkability,self-ratedhealth,andlifesatisfactionthanthosewithalowerlevelofpositiveaffectivestate.Secondly,differentdevelopmentalpathsinmultisitemusculoskeletalpainanddepressivesymptomswereobserved,andasassumed,highjobdemands,lowjobandpersonalresourcesandpoorlifestylehabitswererelatedtothehighlevelsofthesesymptoms.

Finally,thisstudyaimedtoofferabalancedviewofemployeewell-being,examiningitfromseveralperspectives.Obviously,notonlypre-ventingill-health(e.g.,musculoskeletalpain,depressivesymptoms)issuf-ficient,butalsoenhancingpositivewell-being(e.g.,workengagement,lifesatisfaction)isneeded:botharedimensionsofemployeewell-beingandthusessentialwhenexaminingtheantecedentsofwell-being.Similarly,severalworkcharacteristicsandindividualfactorsneedtobeconsideredinimprovingemployeewell-being.Evidently,thefindingsofthepresentstudysuggestthatamotivatedandenergeticworker,whohasresourcefuljobandlowmentalandphysicaldemands,hasbetterwell-beingthanaco-workerwhoislessengagedandhasmoreunfavourablesituationintermsofdemandsandresources.Moreover,agoodlevelofself-esteem,anoptimisticviewofthefuture,andhealthylifestylehabitsmayincreasethelikelihoodofbeinghealthyandhappier.Hence,byfocusingon,butofcoursenotonlylimitedto,improvingworkcharacteristics,personalresources,andlifestyle,employeewell-beingmaybepromoted.Assuch,workorganizationshavegoodopportunitiestoreinforceandmaintainthehealthandwell-beingoftheiremployeesandconsequentlypreventill-healthandpromotelongworkcareers.

89

ReFeRenCesAhlstrom,L.,Grimby-Ekman,A.,Hagberg,M.,&Dellve,L.(2010).Theworkabilityindexand

single-itemquestion:associationswithsickleave,symptoms,andhealth–aprospectivestudyofwomenonlong-termsickleave.Scandinavian Journal of Work, Environment & Health, 36,404–412.

Airila,A.,Hakanen,J.J.,Luukkonen,R.,Lusa,S.,&Punakallio,A.(2013).Positiveandnegativemoodtrajectoriesandtheirrelationshiptoworkability,self-ratedhealthandlifesatisfaction:A13-yearfollow-upstudy.Journal of Occupational and Environmental Medicine,55,779–785.

Airila,A.,Hakanen,J.J.,Luukkonen,R.,Lusa,S.,Punakallio,A.,&Leino-Arjas,P.(2014).Developmentaltrajectoriesofmultisitemusculoskeletalpainanddepressivesymptoms–theeffectsofjobdemandsandresourcesandindividualfactors.Psychology & Health,29,1421–1441.

Airila,A.,Hakanen,J.,Punakallio,A.,Lusa,S.,&Luukkonen,R.(2012).Isworkengagementrelatedtoworkabilitybeyondworkingconditionsandlifestylefactors?International Archives of Occupational and Environmental Health,85,915–925.

Airila,A.,Hakanen,J.J.,Schaufeli,W.B.,Luukkonen,R.,Punakallio,A.,&Lusa,S.(2014).Arejobandpersonalresourcesassociatedwithworkability10yearslater?Themediatingroleofworkengagement.Work & Stress,28,87–105.

Alavinia,S.M.,deBoer,A.G.E.M.,vanDuivenbooden,J.C.,Frings-Dresen,M.H.W.,&Burdorf,A.(2009).Determinantsofworkabilityanditspredictivevaluefordisability.Occupational Medicine,59,32–37.

Andruff,H.,Carraro,N.,Thompson,A.,Gaudreau,P.,&Louvet,B.(2009).Latentclassgrowthmodelling:Atutorial.Tutorials in Quantitative Methods for Psychology,5,11–24.

Antonovsky,A.(1987).Unraveling the mystery of health. How people manage stress and stay well.SanFransisco:Jossey-Bass.

Baglioni,C.,Battagliese,G.,Feige,B.,Spiegelhalder,K.,Nissen,C.,Voderholzer,U.,…&Riemann,D.(2011).Insomniaasapredictorofdepression:Ameta-analyticevaluationoflongitudinalepidemiologicalstudies.Journal of Affective Disorders,135,10–19.

Bair,M.J.,Robinson,R.L.,Katon,W.,&Kroenke,K.(2003).Depressionandpaincomorbidity.Aliteraturereview.Archives of Internal Medicine,163,2433–2445.

90

reFerences

Bakker,A.B.,&Demerouti,E.(2007).Thejobdemands-resourcesmodel:stateoftheart.Journal of Managerial Psychology,22,309–328.

Bakker,A.B.,Demerouti,E.,deBoer,E.,&Schaufeli,W.B.(2003).Jobdemandsandjobresourcesaspredictorsofabsencedurationandfrequency.Journal of Vocational Behavior,62,341–356.

Bakker,A.B.,&Leiter,M.(2010).Wheretogofromhere:Integrationandfutureresearchonworkengagement.InA.B.Bakker&M.P.Leiter(Eds.),Work engagement: A handbook of essential theory and research(pp.181–196).NewYork:PsychologyPress.

Bakker,A.B.,Schaufeli,W.B.,Leiter,M.P.,&Taris,T.W.(2008).Workengagement:Anemergingconceptinoccupationalhealthpsychology.Work & Stress,22,187–200.

Bakker,A.B.,Tims,M.,&Derks,D.(2012).Proactivepersonalityandperformance:Theroleofjobcraftingandworkengagement.Human Relations,65,1359–1378.

Barger,S.D.(2006).Dopsychologicalcharacteristicsexplainsocioeconomicstratificationofself-ratedhealth?Journal of Health Psychology,11,21–35.

Bastian,B.,Kuppens,P.,DeRoover,K.,&Diener,E.(2014).Isvaluingpositiveemotionassociatedwithlifesatisfaction?Emotion,14,639–645.

Bates,J.E.,Goodnight,J.A.,&Fite,J.E.(2008).Temperamentandemotion.InM.Lewis,J.M.Haviland-Jones&L.FeldmanBarrett(Eds.),Handbook of emotions(3rded.)(pp.485–496).NewYork:GuilfordPress.

Baumeister,R.F.,Campbell,J.D.,Krueger,J.I.,&Vohs,K.D.(2003).Doeshighself-esteemcausebetterperformance,interpersonalsuccess,happiness,orhealthierlifestyles?Psycho-logical Science in the Public Interest,4,1–44.

Beaton,R.,Murphy,S.,Johnson,C.,Pike,K.,&Corneil,W.(1998).Exposuretoduty-relatedincidentstressorsinurbanfirefightersandparamedics.Journal of Traumatic Stress,11,821–828.

BeatonR.D.,Murphy,S.A.,Pike,K.C.,&Corneil,W.(1997).Socialsupportandnetworkconflictinfirefightersandparamedics.Western Journal of Nursing Research,19,297–313.

Beck,A.T.,Steer,R.A.,&Garbin,M.G.(1988).PsychometricpropertiesoftheBeckDep-ressionInventory:Twenty-fiveyearsofevaluation.Clinical Psychology Review,8,77–100.

Beedie,C.J.,Terry,P.C.,&Lane,A.M.(2000).Theprofileofmoodstatesandathleticperfor-mance:Twometa-analysis.Journal of Applied Sport Psychology,12,49–68.

Biddle,S.J.H.,&Mutrie,N.(2008).Psychology of physical activity: Determinants, well-being and interventions(2nded.).NewYork:Routledge.

Billings,D.W.,Folkman,S.,Acree,M.,&Moskowitz,J.T.(2000).Copingandphysicalhealthduringcaregiving:therolesofpositiveandnegativeaffect.Journal of Personality and Social Psychology,79,131–142.

Björkstén,M.G.,Boquist,B.,Talbäck,M.,&Edling,C.(1999).Thevalidityofreportedmus-culoskeletalproblems.Astudyofquestionnaireanswersinrelationtodiagnoseddisordersandperceptionofpain.Applied Ergonomics,30,325–330.

91

reFerences

Bonde,J.P.E.(2008).Psychosocialfactorsatworkandriskofdepression:asystematicreviewoftheepidemiologicalevidence.Occupational and Environmental Medicine,65,438–445.

Bongers,P.M.,deWinter,C.R.,Kompier,M.A.J.,&Hildebrandt,V.H.(1993).Psychosocialfactorsatworkandmusculoskeletaldisease.Scandinavian Journal of Work, Environment & Health,19,297–312.

Bos,J.,Mol,E.,Visser,B.,&Frings-Dresen,M.(2004).Riskofhealthcomplaintsanddisabi-litiesamongDutchfirefighters.International Archives of Occupational and Environmental Health,77,373–382.

Bourbonnais,R.,Brisson,C.,Vinet,A.,Vézina,M.,Abdous,B.,&Gaudet,M.(2006).Effecti-venessofaparticipativeinterventiononpsychosocialworkfactorstopreventmentalhealthproblemsinahospitalsetting.Occupational and Environmental Medicine,63,335–342.

Boxer,P.A.,&Wild,D.(1993).Psychologicaldistressandalcoholuseamongfirefighters.Scandinavian Journal of Work, Environment & Health,19,121–125.

Brand,S.,Beck,J.,Hatzinger,M.,Harbaugh,A.,Ruch,W.,&Holsboer-Trachsler,E.(2010).Associationsbetweensatisfactionwithlife,burnout-relatedemotionalandphysicalexhaus-tion,andsleepcomplaints.World Journal of Biological Psychiatry,11,744–754.

Brief,A.P.,&Weiss,H.M.(2002).Organizationalbehavior:Affectintheworkplace.AnnualReview of Psychology,53,279–307.

Brown,J.H.,Kazi,L.E.,Spitz,P.W.,Gertman,P.,Fries,J.F.,&Meehan,R.F.(1984).Thedimensionsofhealthoutcomes:across-validatedexaminationofhealthstatusmeasurement.American Journal of Public Health,74,159–161.

Browne,M.W.,&Cudeck,R.(1993).Alternativewaysofassessingmodelfit.InJ.S.Long(Ed.),Testing structural equation models(pp.136–162).NewburyPark:Sage.

Byrne,B.M.(2010).Structural equation modeling with AMOS. Basic concepts, applications, and programming(2nded.).NewYork,NY:Routledge.

Caccioppo,J.T.,&Berntson,G.G.(1994).Relationshipbetweenattitudesandevaluativespace:Acriticalreview,withemphasisontheseparabilityofpositiveandnegativesubstrates.Psychological Bulletin,115,401–423.

Cannella,D.T.,Lobel,M.,Glass,P.,Lokshina,I.,&Graham,J.E.(2007).Factorsassociatedwithdepressedmoodinchronicpainpatients:Theroleofintrapersonalcopingresources.Journal of Pain,8,256–262.

Carey,M.,Al-Zaiti,S.,Dean,G.,Sessanna,L.,&Finnell,D.S.(2011).Sleepproblems,depres-sion,substanceuse,socialbonding,andqualityoflifeinprofessionalfirefighters.Journal of Occupational & Environmental Medicine,53,928–932.

Carver,C.S.,Pozo-Kaderman,C.,Harris,S.D.,Noriega,V.,Scheier,M.F.,Robinson,D.S.,Ketcham,A.F.,Moffat,F.L.,&Clark,K.C.(1994).Optimismversuspessimismpredictsthequalityofwomen’sadjustmenttoearlystagebreastcancer.Cancer,73,1213–1220.

Carver,C.S.,&Scheier,M.F.(2009).Optimism.InM.R.Leary&R.H.Hoyle(Eds.),Hand-book of individual differences in social behavior(pp.330–342).NewYork:GuildfordPress.

92

reFerences

Cella,D.F.,Tross,S.,Orav,J.E.,Holland,J.C.,Silberfarb,P.M.,&Rafla,S.(1989).Moodstatesofpatientsafterthediagnosisofcancer.Journal of Psychosocial Oncology,7,45–54.

Chida,Y.,&Steptoe,A.(2008).Positivepsychologicalwell-beingandmortality:Aquantitativereviewofprospectiveobservationalstudies.Psychosomatic Medicine,70,741–756.

Christensen,J.O.,&Knardahl,S.(2010).Workandneckpain:Aprospectivestudyofpsycho-logical,social,andmechanicalriskfactors.Pain,151,165–173.

Christian,M.S.,Garza,A.S.,&Slaughter,J.E.(2011).Workengagement:aquantitativereviewandtestofitsrelationswithtaskandcontextualperformance.Personnel Psychology,64,89–136.

Cohen,S.(1988).Psychosocialmodelsoftheroleofsocialsupportintheetiologyofphysicaldisease.Health Psychology,7,269–297.

Cooper,C.L.(1998).(Ed.)Theories of organizational stress.Oxford:OxfordUniversityPress.

Cooper,C.L.,&Marshall,J.(1976).Occupationalsourcesofstress:areviewoftheliteraturerelatingtocoronaryheartdiseaseandmentalillhealth.Journal of Occupational Psychology,49,11–28.

Costa,P.T.,Terracciano,A.,&McCrae,R.R.(2001).Genderdifferencesinpersonalitytraitsacrosscultures:Robustandsuprisingfindings.Journal of Personality and Social Psychology,81,322–331.

Crawford,E.R.,LePine,J.A.,&RichB.L.(2010).Linkingjobdemandsandresourcestoemployeeengagementandburnout:Atheoreticalextentionandmeta-analytictest.Journal of Applied Psychology,95,834–848.

Cropanzano,R.,Weiss,H.M.,Hale,J.M.S.,&Reb,J.(2003).Thestructureofaffect:Recon-sideringtherelationshipbetweennegativeandpositiveaffectivity.Journal of Management,29,831–857.

Csikszentmihalyi,M.(1990).Flow: The Psychology of Optimal Experience.NewYork:HarperandRow.

Deci,E.L.,&Ryan,R.M.(2000).The“what”and“why”ofgoalpursuits:Humanneedsandtheself-determinationofbehavior.Psychological Inquiry,11,227–268.

deCroonE.M.,Sluiter,J.K.,Nijssen,T.F.,Kammeijer,M.,Dijkmans,B.A.,Lankhorst,G.J.,&Frings-Dresen,M.H.(2005).WorkabilityofDutchemployeeswithrheumatoidarthritis.Scandinavian Journal of Rheumatology,34,277–283.

Demerouti,E.,Bakker,A.B.,Nachreiner,F.,&Schaufeli,W.B.(2001).Thejobdemands-resourcesmodelofburnout.Journal of Applied Psychology,86,499–512.

Demerouti,E.,Bakker,A.B.,&Schaufeli,W.B.(2005).Spilloverandcrossoverofexhaustionandlifesatisfactionamongdual-earnerparents.Journal of Vocational Behavior,67,266–289.

Demyttenaere,K.,Bruffaerts,R.,Lee,S.,Posada-Villa,J.,Kovess,V.,Angermever,M.C.,…VonKorff,M.(2007).Mentaldisordersamongpersonswithchronicbackorneckpain:Resultsfromtheworldmentalhealthsurveys.Pain,129,332–342.

93

reFerences

DeNeve,K.M.,&Cooper,H.(1998).Thehappypersonality:Ameta-analysisof137personalitytraitsandsubjectivewell-being.Psychological Bulletin,124,197–229.

DeSalvo,K.B.,Bloser,N.,Reynolds,K.,He,J.,&Muntner,P.(2006).Mortalitypredictionwithasinglegeneralself-ratedquestion.Ameta-analysis.Journal of General Internal Medicine,21,267–275.

Dewe,P.J.,O’Driscoll,M.P.,&Cooper,C.L.(2010).Coping with work stress: a review and critique.Malden,MA:Wiley-Blackwell.

deZwart,B.C.H.,Frings-Dresen,M.H.W.,&vanDuivenbooden,J.C.(2002).Test-retestreliabilityoftheWorkAbilityIndexquestionnaire.Occupational Medicine,52,177–181.

Diener,E.(1984).Subjectivewell-being.Psychological Bulletin,95,542–575.

Diener,E.(2000).Subjectivewell-being:Thescienceofhappinessandaproposalforanationalindex.American Psychologist,55,34–43.

Diener,E.,&Diener,C.(1996).Mostpeoplearehappy.Psychological Science,7,181–185.

Diener,E.,Suh,E.M.,Lucas,R.E.,&Smith,H.L.(1999).Subjectivewell-being:Threedecadesofprogress.Psychological Bulletin,125,276–302.

Doty,H.D.,&Glick,W.H.(1998).Commonmethodsbias:Doescommonmethodsvariancereallybiasresults?Organizational Research Methods,1,374–406.

Elavsky,S.,McAuley,E.,Motl,R.W.,Konopack,J.F.,Marquez,D.X.,Hu,L.,Jerome,G.J.,&Diener,E.(2005).Physicalactivityenhanceslong-termqualityoflifeinolderadults:Efficacy,esteem,andaffectiveinfluences.Annals of Behavioral Medicine,30,138–145.

Elliot,D.L.,Goldberg,L., Duncan,T.E., Kuehl,K.S., Moe,E.L.,Breger,R.K.R., DeFrances-co,C.L.,Ernst,D.B.,& Stevens,V.(2004).ThePHLAMEfirefighters’study:Feasibilityandfindings.American Journal of Health Behavior,28, 13–23.

Elo,A-L.,Leppänen,A.,Lindström,K.,&Ropponen,T.(1992).OSQ Occupational stress questionnaire: User’s instructions.Helsinki:InstituteofOccupationalHealth.

Engel,G.L.(1977).Theneedforanewmedicalmodel:achallengeforbiomedicine.Science,196,129–135.

Erdogan,B.,Bauer,T.N.,Truxillo,D.M.,&Mansfield,L.R.(2012).Whistlewhileyouwork:Areviewofthelifesatisfactionliterature.Journal of Management,38,1038–1083.

Feingold,A.(1994).Genderdifferencesinpersonality:Ameta-analysis.Psychological Bulletin,116,429–456.

Feldt,T.,Hyvönen,K.,Mäkikangas,A.,Kinnunen,U.,&Kokko,K.(2009).DevelopmenttrajectoriesofFinnishmanagers’workabilityovera10-yearfollow-upperiod.Scandinavian Journal of Work, Environment & Health,35,37–47.

FinnishCentreforPensions.(2013).Finnish centre for pensions pocket statistics 2013.Helsinki:FinnishCentreforPensions.

94

reFerences

Fischer,F.M.,daSilvaBorges,F.N.,Rotenberg,L.,deOliveiraLatorre,M.R.D.,Soares,N.S.,SantaRosa,P.L.F.,…Landsbergis,P.,(2006).Workabilityofhealthcareshiftworkers:whatmatters?Chronobiology International,23,1165–1179.

Ford,M.T.,Woolridge,J.D.,Vipanchi,M.,Kakar,U.M.,&Strahan,S.R.(2014).Timelagsinoccupationalstressor-straineffects:Areviewandmeta-analysisoflongitudinalstudies.Work & Stress,28,9–30.

Fredrickson,B.L.(2001).Theroleofpositiveemotionsinpositivepsychology.Thebroaden-and-buildtheoryofpositiveemotions.American Psychologist,56,218–226.

Fredrickson,B.L.,&Branigan,C.(2005).Positiveemotionsbroadenthescopeofattentionandthought-actionrepertoires.Cognition & Emotion,19,313–332.

Fredrickson,B.L.,&Joiner,T.(2002).Positiveemotionstriggerupwardspiralstowardemotionalwell-being.Psychological Science,13,172–175.

Friedman,H.S.,&Booth-Kewley,S.(2003).The“disease-pronepersonality”:Ameta-analyticviewoftheconstruct.InP.Salovey&A.J.Rothman(Eds.),Social psychology of health: key readings in social psychology(pp.305–324).NewYork:PsychologyPress.

Fujita,F.,Diener,E.,&Sandvik,E.(1991).Genderdifferencesinnegativeaffectandwell-being:Thecaseforemotionalintensity.Journal of Personality and Social Psychology,61,427–434.

Fullerton,C.S.,Ursano,R.J.,&Wang,L.(2004).Acutestressdisorder,posttraumaticstressdisorder,anddepressionindisasterorrescueworkers.American Journal of Psychiatry,161,1370–1376.

Giltay,E.J.,Geleijnse,J.M.,Zitman,F.G.,Buijsse,B.,&Kromhout,D.(2007).Lifestyleanddietarycorrelatesofdispositionaloptimisminmen:TheZutphenelderlystudy.Journal of Psychosomatic Research,63,483–490.

Gould,R.,Ilmarinen,J.,Järvisalo,J.,&Koskinen,S.(Eds.)(2008).Dimensions of Work ability: Results of the Health 2000 Survey.Helsinki:FinnishCentreforPensions,TheSocialInsuran-ceInstitution,NationalPublicHealthInstitute,FinnishInstituteofOccupationalHealth.

Goldsmith,H.H.(1993).Temperament:Variabilityindevelopingemotionsystems.InM.Lewis&J.M.Haviland(Eds.),Handbook of emotions(pp.353–364).NewYork:GuilfordPress.

Gray-Little,B.,Williams,V.S.L.,&Hancock,T.D.(1997).AnitemresponsetheoryanalysisoftheRosenbergSelf-EsteemScale.Personality and Social Psychology Bulletin,23,443–451.

Green,D.P.,Goldman,S.L.,&Salovey,P.(1993).Measurementerrormasksbipolarityinaffectratings.Journal of Personality and Social Psychology,64,1029–1041.

Griffith,N.M.,Szaflarski,J.P.,Szarflarski,M.,Kent,G.P.,Schefft,B.K.,Howe,S.R.,&Privitera,M.D.(2005).Measuringdepressivesymptomsamongtreatment-resistantseizuredisorderpatients:POMSdepressionscaleasanalternativetotheBDI-II.Epilepsy & Health,7,266–272.

Gross,J.J.,Carstensen,L.L.,Pasupathi,M.,Tsai,J.,GötestamSkorpen,C.,&Hsu,A.Y.C.(1997).Emotionandaging:Experience,expression,andcontrol.Psychology and Aging,12,590–599.

95

reFerences

Gruenberg,A.M.,&Goldstein,R.D.(2003).Mooddisorders:depression.InA.J.Tasman,J.J.Kay&J.A.Lieberman(Eds.),Psychiatry(pp.1207–1213).NewYork:JohnWiley.

Hackman,J.R.,&Oldman,G.R.(1976).Motivationthroughthedesignofwork:Testofatheory.Organizational Behavior and Human Performance,16,250–279.

Hakanen,J.(2009).Työnimunarviointimenetelmä.[UtrechtWorkEngagementScale].Hel-sinki:Työterveyslaitos.

Hakanen,J.(2002).Työuupumuksestatyönimuun–positiivisentyöhyvinvointikäsitteenarvi-ointimenetelmänsuomalaisenversionvalidointiopetusalanorganisaatiossa.[Fromburnouttoworkengagement–validationoftheFinnishversionofaninstrumentformeasuringworkengagement(UWES)inanindustrialorganization].Työ ja ihminen,16,42–58.

Hakanen,J.,Bakker,A.B.,&Schaufeli,W.B.(2006).Burnoutandworkengagementamongteachers.Journal of School Psychology,43,495–513.

Hakanen,J.,&Lindbohm,M.(2008).Workengagementamongbreastcancersurvivorsandthereferents:Theimportanceofoptimismandsocialresourcesatwork.Journal of Cancer Survivorship,2,283–295.

Hakanen,J.J.,Peeters,M.,&Perhoniemi,R.(2011).Enrichmentprocessesandgainspiralsatworkandathome:Athree-yearcross-laggedpanelstudy.Journal of Occupational and Organizational Psychology,84,8–30.

Hakanen,J.,Perhoniemi,R.,&Toppinen-Tanner,S.(2008a).Positivegainspiralsatwork:Fromjobresourcestoworkengagement,personalinitiativeandwork-unitinnovativeness.Journal of Vocational Behavior,73,78–91.

Hakanen,J.J.,&Roodt,G.(2010).Usingthejobdemands-resourcesmodeltopredictworkengagement:Analysingtheconceptualmodel.InA.B.Bakker&M.P.Leiter(Eds.),Work engagement: A handbook of essential theory and research(pp.85–101).NewYork:PsychologyPress.

Hakanen,J.,&Schaufeli,W.B.(2012).Doburnoutandworkengagementpredictdepressivesymptomsandlifesatisfaction?Athree-waveseven-yearprospectivestudy.Journal of Affective Disorders,141,415–424.

Hakanen,J.J.,Schaufeli,W.B.,&Ahola,K.(2008b).Thejobdemands-resourcesmodel:Athree-yearcross-laggedstudyofburnout,depression,commitment,andworkengagement.Work & Stress,22,224–241.

Halbesleben,J.R.B.(2010).Ameta-analysisofworkengagement:Relationshipsbetweenburnout,demands,resources,andconsequences.InA.B.Bakker&M.P.Leiter(Eds.),Work engagement: A handbook of essential theory and research(pp.102–117).NewYork:PsychologyPress.

Halbesleben,J.R.B.,Osburn,H.K.,&Mumford,M.D.(2006).Actionresearchasaburnoutinterventionreducingburnoutinthefederalfireservice.Journal of Applied Behavioral Science,42,244–266.

Hansez,I.,&Chmiel,N.(2010).Safetybehavior:Jobdemands,resources,andperceivedma-nagementcommitmenttosafety.Journal of Occupational Health Psychology,15,267–278.

96

reFerences

Haslam,C.,&Mallon,K.(2003).Apreliminaryinvestigationofpost-traumaticstresssymptomsamongfirefighters.Work & Stress,17,277–285.

Hauke,A.,Flintrop,J.,Brun,E.,&RuguliesR.(2011).Theimpactofwork-relatedpsychosocialstressorsontheonsetofmusculoskeletaldisordersinspecificbodyregions:Areviewandmeta-analysisof54longitudinalstudies.Work & Stress,25,243–256.

Haukka,E.,Kaila-Kangas,L.,Ojajärvi,A.,Miranda,H.,Karppinen,J.,Viikari-Juntura,E.,Heliövaara,M.,&Leino-Arjas,P.(2013).Paininmultiplesitesandsicknessabsencetrajectories:aprospectivestudyamongFinns.Pain,154,306–312.

Haukka,E.,Leino-Arjas,P.,Ojajärvi,A.,Takala,E-P.,Viikari-Juntura,E.,&Riihimäki,H.(2011).Mentalstressandpsychosocialfactorsatworkinrelationtomultiple-sitemus-culoskeletalpain:Alongitudinalstudyofkitchenworkers.European Journal of Pain,15,432–438.

Haukka,E.,Ojajärvi,A.,Takala,E-P.,Viikari-Juntura,E.,&Leino-Arjas,P.(2012).Physicalworkload,leisure-timephysicalactivity,obesityandsmokingaspredictorsofmultisitemusculoskeletalpain.A2-yearprospectivestudyofkitchenworkers.Occupational and Environmental Medicine,69,485–492.

Hayes,A.F.(2009).BeyondBaronandKenny:Statisticalmediationanalysisinthenewmil-lennium.Communication Monographs,76,408–420.

Hertzberg,F.(1959).The motivation to work.NewYork:WileyInternationalEdition.

Hobfoll,S.E.(1989).Conservationofresources.Anewattemptatconceptualizingstress.American Psychologist,44,513–524.

Hobfoll,S.E.(2001).Theinfluenceofculture,community,andthenested-selfinthestressprocess:Advancingconservationofresourcestheory.Applied Psychology: An International Review,50,337–421.

Hobfoll,S.E.(2002).Socialandpsychologicalresourcesandadaptation.Review of GeneralPsychology,6,307–324.

Hobfoll,S.E.,Johnson,R.J.,Ennis,N.,&Jackson,A.P.(2003).Resourceloss,resourcegain,andemotionaloutcomesamonginnercitywomen.Journal of Personality and SocialPsychology,84,632–643.

Houdmont,J.,&Leka,S.(2010).Anintroductiontooccupationalhealthpsychology.InS.Leka&J.Houdmont(Eds.),Occupational health psychology(pp.1–30).Malden,MA:Wiley-Blackwell.

Hu,Q.,&Schaufeli,W.B.(2011).JobinsecurityandremunerationinChinesefamily-ownedbusinessworkers.Career Development International,16,6–19.

Hu,Q.,Schaufeli,W.B.,&Taris,T.(2011).TheJob-Demands-Resourcesmodel:Ananalysisofaddictiveandjointeffectsofdemandsandresources.Journal of Vocational Behavior,79,181–190.

Hänninen,H.(1989).Neurotoksisten haittojen seulonta. Oirekyselyt ja psykologiset testit.[Scree-ningofneurotoxiceffects.Symptomquestionnairesandpsychologicaltests].Helsinki:Työterveyslaitos.

97

reFerences

Hölzel,L.,Härter,M.,Reese,C.,&Kriston,L.(2011).Riskfactorsforchronicdepression–Asystematicreview.Journal of Affective Disorders,129,1–13.

Idler,E.L.,&Benyamini,Y.(1997).Self-ratedhealthandmortality:Areviewoftwenty-sevencommunitystudies.Journal of Health and Social Behavior,38,21–37.

Ilmarinen,J.(2006).Towards longer worklife! Ageing and the quality of worklife in the European Union.Helsinki:FinnishInstituteofOccupationalHealth,MinistryofSocialAffairsandHealth.

Ilmarinen,J.(2009).Workability–Acomprehensiveconceptforoccupationalhealthresearchandprevention.Scandinavian Journal of Work, Environment & Health,35,1–5.

Ilmarinen,J.,Gould,R.,Järvikoski,A.,&Järvisalo,J.(2008).Diversityofworkability.InR.Gould,J.Ilmarinen,J.Järvisalo&S.Koskinen(Eds.)Dimensions of Work ability: Results of the Health 2000 Survey(pp.13–24).Helsinki:FinnishCentreforPensions,TheSocialInsuranceInstitution,NationalPublicHealthInstitute,FinnishInstituteofOccupationalHealth.

Ilmarinen,J.,Tuomi,K.,&Klockars,M.(1997).Changesintheworkabilityofactiveemp-loyeesoveran11-yearperiod.Scandinavian Journal of Work, Environment & Health,23(suppl1),49–57.

Ilmarinen,J.,Tuomi,K.,&Seitsamo,J.(2005).Newdimensionsofworkability.InG.Costa,W.Goedhard&J.Ilmarinen(Eds.),Assessment and promotion of work ability, health and well-being of ageing workers(pp.3–7).Elsevier,InternationalCongressSeries1280.

Janssen,P.P.M.,Schaufeli,W.B.,&Houkes,I.(1999).Work-relatedandindividualdetermi-nantsofthethreeburnoutdimensions.Work & Stress,13,74–86.

Jones,S.C.(1973).Self-andinterpersonalevaluations:Esteemtheoriesversusconsistencytheories.Psychological Bulletin,79,185–199.

Jones,B.L.,&Nagin,D.S.(2007).Advancesingroup-basedtrajectorymodelingandanSASprocedureforestimatingthem.Sociological Methods & Research,35,542–571.

Järvikoski,A.,Härkäpää,K.,&Mannila,S.(2001).Moniulotteinentyökykykäsitysjatyökykyäylläpitävätoiminta.[Multidimensionalviewofworkabilityandactivitiesforthemainte-nanceofworkability].Kuntoutus,3,3–11.

Kahn,W.A.(1990).Psychologicalconditionsofpersonalengagementanddisengagementatwork.Academy of Management Journal,33,692–724.

Kaleta,D.,Makowiec-Dabrowska,T.,&Jegier,A.(2006).Lifestyleindexandworkability.International Journal of Occupational Medicine and Environmental Health,19,170–177.

Kamaleri,Y.,Natvig,B.,Ihlebaek,C.M.,BenthJ.,&Bruusgaard,D.(2008).Numberofpainsitesisassociatedwithdemographic,lifestyle,andhealth-relatedfactorsinthegeneralpopulation.European Journal of Pain,12,742–748.

Kamaleri,Y.,Natvig,B.,Ihlebaek,C.M.,&Bruusgaard,D.(2009).Doesthenumberofmusculoskeletalpainsitespredictworkdisability?A14-yearprospectivestudy.European Journal of Pain,13,426–430.

98

reFerences

Karasek,R.A.(1979).Jobdemands,jobdecisionlatitudeandmentalstrain:Implicationsforjobredesing.Administrative Science Quarterly,24,285–308.

Karasek,R.A.,&Theorell,T.(1990).Healthy work, stress, productivity and the reconstruction of working life.NewYork:BasicBooks.

Kauppinen,T.,Hanhela,R.,Kandolin,I.,Karjalainen,A.,Kasvio,A.,Perkiö-Mäkelä,M.,…&Viluksela,M.(Eds.)(2010),Työ ja terveys Suomessa.[WorkandHealthinFinland].Helsinki:Työterveyslaitos.

Kivimäki,M.,Vahtera,J.,Elovainio,M.,Helenius,H.,Singh-Manoux,A.,&Pentti,J.(2005).Optimismandpessimismaspredictorsofchangeinhealthafterdeathoronsetofsevereillnessinfamily.Health Psychology,24,413–421.

Korunka,C.,Kubicek,B.,Schaufeli,W.B.,&Hoonaaker,P.(2009).Workengagementandburnout:testingtherobustnessoftheJob-Demands-Resourcesmodel.Journal of Positive Psychology,4,243–255.

Kuorinka,I.,Jonsson,B.,Kilbom,Å.,Vinterberg,H.,Biering-Sørensen,F.,Andersson,G.,&Jørgensen,K.(1987).StandardizedNordicquestionnairesfortheanalysisofmusculoskeletalsymptoms.Applied Ergonomics,18,233–237.

Laaksonen,M.,Rahkonen,O.,Martikainen,P.,&Lahelma,E.(2006).Associationsofpsychoso-cialworkingconditionswithself-ratedgeneralhealthandmentalhealthamongmunicipalemployees.International Archives of Occupational and Environmental Health,79,205–212.

Langelaan,S.,Bakker,A.B.,Schaufeli,W.B.,VanRhenen,W.,&VanDoornen,L.J.P.(2006).Doburned-outandwork-engagedemployeesdifferinthefunctioningofthehypothalamic-pituitary-adrenalaxis.Scandinavian Journal of Work, Environment & Health,32,339–348.

Larsen,J.T.,&McGraw,A.P.(2011).Furtherevidenceformixedemotions.Journal of Personality and Social Psychology,100,1095–1110.

Larsen,J.T.,McGraw,A.P.,&Cacioppo,J.T.(2001).Canpeoplefeelhappyandsadatthesametime?Journal of Personality and Social Psychology,81,684–696.

Larson,J.S.(1999).Theconceptualizationofhealth.Medical Care Research and Review,56,123–136.

Lazarus,R.S.,&Folkman,S.(1984).Stress, appraisal, and coping.NewYork,NY:Springer.

Leana,C.R.,&Barry,B.(2000).Stabilityandchangeassimultaneousexperiencesinorgani-zationallife.Academy of Management Review,25,753–759.

Lebouef-Yde,C.(1999).Smokingandlowbackpain.Asystematicliteraturereviewof41journalarticlesreporting47epidemiologicalstudies.Spine,24,1463–1470.

Leiter,M.P.,&Bakker,A.B.(2010).Workengagement:Introduction.InA.B.Bakker&M.P.Leiter(Eds.),Work engagement: A handbook of essential theory and research (pp.1–9).NewYork:PsychologyPress.

Lepine,J.A.,LePine,M.A.,&Jackson,C.L.(2004).Challengeandhindrancestress:relation-shipswithexhaustion,motivationtolearn,andlearningperformance.Journal of Applied Psychology,89,883–891.

99

reFerences

Lewig,K.A.,Xanthopoulou,D.,Bakker,A.B.,Dollard,M.,&Metzer,J.C.(2007).BurnoutandconnectednessamongAustralianvolunteers:AtestoftheJobDemands-Resourcesmodel.Journal of Vocational Behavior,71,429–445.

Li,C.-Y.,&Sung,F.-C.(1999).Areviewofthehealthyworkereffectinoccupationalepide-miology.Occupational Medicine,49,225–229.

Linton,S.J.,&Bergbom,S.(2011).Understandingthelinkbetweendepressionandpain.Scandinavian Journal of Pain,2,47–54.

Lodahl,T.M.,&Kejner,M.(1965).Thedefinitionandmeasurementofjobinvolvement.Journal of Applied Psychology,49,24–33.

LoprestiA.L.,HoodS.D.,&DrummondP.D.(2013).Areviewoflifestylefactorsthatcont-ributetoimportantpathwaysassociatedwithmajordepression:Diet,sleepandexercise.Journal of Affective Disorders,148,12–27.

Lucas,R.E.,&Donnellan,M.B.(2012).Estimatingthereliabilityofsingle-itemlifesatisfactionmeasures:Resultsfromfournationalpanelstudies.Social Indicators Research,105,323–331.

Lundberg,O.,&Manderbacka,K.(1996).Assessingreliabilityofameasureofself-ratedhealth.Scandinavian Journal of Public Health,24,218–224.

Lusa,S.,Häkkänen,M.,Luukkonen,R.,&Viikari-Juntura,E.(2002).Perceivedphysicalworkcapacity,stress,sleepdisturbanceandoccupationalaccidentsamongfirefightersworkingduringastrike.Work & Stress,16,264–274.

Lusa,S.,Louhevaara,V.,&Kinnunen,K.(1994).Arethejobdemandsonphysicalworkcapa-cityequalforyoungandagingfirefighters?Journal of Occupational Medicine,36,70–74.

Lusa,S.,Punakallio,A.,Luukkonen,R.,&Louhevaara,V.(2006).Factorsassociatedwithchan-gesinperceivedstrainatworkamongfire-fighters:a3-yearfollow-upstudy.International Archives of Occupational and Environmental Health,79,419–426.

Lusa-Moser,S.,Punakallio,A.,Louhevaara,V.,Viikari-Juntura,E.,Ilmarinen,R.,Ollila,J.,Korhonen,O.,Lindqvist-Virkamäki,S.,&Luukkonen,R.(1996).Eri-ikäisten palomiesten terveys ja toimintakyky. Osa I: kyselytutkimus.[Health,physicalandmentalcapacityoffirefightersindifferentagegroups.PartI:Aquestionnairestudy].LoppuraporttiPalosuojelu-rahastolle.Helsinki:Työterveyslaitos.

Luthans,F.,Youssef,C.M.,&Avolio,B.J.(2007).Psychological capital: Developing the human competitive edge.Oxford:OxfordUniversityPress.

Lyubomirsky,S.,King,L.,&Diener,E.(2005).Thebenefitsoffrequentpositiveaffect:Doeshappinessleadtosuccess.Psychological Bulletin,131,803–855.

MacKinnon,D.P.,Elliot,D.L.,Thoemmes,F.,Kuehl,K.S.,Moe,E.L.,Goldberg,L.,LockhartBurrell,G.,&Ranby,K.W.(2010).Long-termeffectsofaworksitehealthpromotionprogramforfirefighters.American Journal of Health Behavior,34,695–706.

Magni,G.,Moreschi,C.,Rigatti-Luchini,S.,&Merskey,H.(1994).Prospectivestudyontherelationshipbetweendepressivesymptomsandchronicmusculoskeletalpain.Pain,56,289–297.

100

reFerences

Mankkinen,T.(2011).Palomiehen ammatti työnä ja elämäntapana.[Firefighter’scareerasajobandawayoflife]:ActaElectronicaUniversitatisTamperensis,1040.Tampere:TampereUniversityPress.

Marsh,H.W.(1996).Positiveandnegativeglobalself-esteem:Asubstantivelymeaningfuldistinctionorartifactors?Journal of Personality and Social Psychology,70,810–919.

Martus,P.,Jakob,O.,Rose,U.,Seibt,R.,&Freude,G.(2010).AcomparativeanalysisoftheWorkAbilityIndex.Occupational Medicine,60,517–524.

Maslach,C.,&Leiter,M.(1997).The truth about burnout.SanFrancisco:JosseyBass.

Mauno,S.,Kinnunen,U.,&Ruokolainen,M.(2007).Jobdemandsandresourcesasantecedentsofworkengagement:Alongitudinalstudy.Journal of Vocational Behavior,70,149–171.

May,D.R.,Gilson,R.L.,&Harter,L.M.(2004).Thepsychologicalconditionsofmeaning-fulness,safetyandavailabilityandtheengagementofthehumanspiritatwork.Journal of Occupational and Organizational Psychology,77,11–37.

McDowell,I.(2010).Measuresofself-perceivedwell-being.Journal of Psychosomatic Research,69,69–79.

McLeroy,K.R.,Gottlieb,N.H.,&Heaney,C.A.(2002).Socialhealthintheworkplace.InM.P.O’Donnell(Ed.).Health promotion in the workplace(pp.459–492).Albany,NY:Delmar.

McNair,D.M.,Lorr,M.,&Droppelman,L.F.(1971).Edits manual for the profile of mood states.SanDiego,CA:EducationalandIndustrialTestingService.

Melin,B.,&Lundberg,U.(1997).Abiopsychosocialapproachtowork-stressandmusculos-keletaldisorders.International Journal of Psychophysiology,11,238–247.

Miller,L.R.,&Cano,A.(2009).Comorbidchronicpainanddepression:Whoisatrisk?Journal of Pain,10,619–627.

Miranda,H.,Kaila-Kangas,L.,Heliövaara,M.,Leino-Arjas,P.,Haukka,E.,Liira,J.,&Viikari-Juntura,E.(2010).Musculoskeletalpainatmultiplesitesanditseffectsonworkabilityinageneralworkingpopulation.Occupational and Environmental Medicine,67,449–455.

Mitani,S.,Fujita,M.,Nakata,K.,&Shirakawa,T.(2006).Impactofpost-traumaticstressdisorderandjob-relatedstressonburnout:Astudyoffireserviceworkers.Journal of Emergency Medicine,31,7–11.

Mulaik,S.,&James,L.(1995).Objectivityandreasoninginscienceandstructuralequationmodeling.InR.Hoyle(Ed.),Structural equation modeling: Concepts, issues, and applications(pp.118–137).ThousandOaks:SagePublications,Inc.

Munce,S.E.P.,Stansfeld,S.A.,Blackmore,E.,&Stewart,D.E.(2007).Theroleofdepressionandchronicpainconditionsinabsenteeism:Resultsfromanationalepidemiologicalsurvey.Journal of Occupational and Environmental Medicine,49,1206–1211.

Myers,R.(1990).Classical and modern regression with applications(2nded.).Boston,MA:Duxbury.

Mäkikangas,A.,Feldt,T.,Kinnunen,U.,&Tolvanen,A.(2012).Dolowburnoutandhighworkengagementalwaysgohandinhand?Investigationoftheenergyandidentificationdimensionsinlongitudinaldata.Anxiety, Stress, & Coping,25,93–116.

101

reFerences

Mäkikangas,A.,&Kinnunen,U.(2003).Psychosocialworkstressorsandwell-being:Self-esteemandoptimismasmoderatorsinaone-yearlongitudinalsample.Personality and Individual Differences,35,537–557.

Mäkikangas,A.,Kinnunen,U.,&Feldt,T.(2004).Self-esteem,dispositionaloptimism,andhealth:Evidencefromcross-laggeddatainemployees.Journal of Research in Personality,38,556–575.

Mäkikangas,A.,Schaufeli,W.,Tolvanen,A.,&Feldt,T.(2013).Engagedmanagersarenotworkaholics:Evidencefromalongitudinalperson-centeredanalysis.Journal of Work and Organizational Psychology,29,135–143.

Nagin,D.S.(1999).Analyzingdevelopmentaltrajectories:Asemiparametric,group-basedapproach.Psychological Methods,4,139–157.

Netterstrøm,B.,Conrad,N.,Bech,P.,Fink,P.,Olsen,O.,Rugulies,R.,&Stansfeld,S.(2008).Therelationbetweenwork-relatedpsychosocialfactorsandthedevelopmentofdepression.Epidemiological Review,30,118–132.

Neupane,S.,Miranda,H.,Virtanen,P.,Siukola,A.,&Nygård,C-H.(2013).Dophysicalorpsychosocialfactorsatworkpredictmulti-sitemusculoskeletalpain?A4-yearfollow-upstudyinanindustrialpopulation.International Archives of Occupational and Environmental Health,86,581–589.

Niedhammer,I.,&Chea,M.(2003).Psychosocialfactorsatworkandselfreportedhealth:comparativeresultsofcrosssectionalandprospectiveanalysesoftheFrenchGAZELcohort.Occupational and Environmental Medicine,60,509–515.

Nielsen,K.,&SimonsenAbildgaard,S.(2013).Organizationalinterventions:Aresearch-basedframeworkfortheevaluationofbothprocessandeffects.Work & Stress,27,278–297.

Odgen,J.(2012).Health psychology: A textbook(5thed.).Maidenhead:OpenUniversityPress.

Parks,K.M.,&Steelman,L.A.(2008).Organizationalwellnessprograms:ameta-analysis.Journal of Occupational Health Psychology,13,58–68.

Partinen,M.,&Gislason,T.(1995).BasicNordicSleepQuestionnaire(BNSQ):aquantitatedmeasureofsubjectivesleepcomplaints.Journal of Sleep Research,4(suppl1),150–155.

Parzefall,M-R.,&Hakanen,J.(2010).Psychologicalcontractanditsmotivationalandhealth-enhancingproperties.Journal of Managerial Psychology,25,4–21.

Paterniti,S.,Niedhammer,I.,Lang,T.,&Consoli,M.(2002).Psychosocialfactorsatwork,personalitytraitsanddepressivesymptoms.LongitudinalresultsfortheGAZELStudy.British Journal of Psychiatry,181,111–117.

Pavot,W.,&Diener,E.(1993).Reviewofthesatisfactionwithlifescale.Psychological Assess-ment,5,164–172.

Pavot,W.,&Diener,E.(2008).TheSatisfactionwithLifeScaleandtheemergingconstructoflifesatisfaction.The Journal of Positive Psychology,3,137–152.

Pelastusopisto.(2014).Pelastustoimen taskutilasto 2009–2013[Pocketstatisticsofrescueservices2009–2013].Pelastusopistonjulkaisu.D-sarja:Muut1/2014.

102

reFerences

Peterson,C.,&Seligman,M.E.P.(2004).Character strengths and virtues. A handbook and classification.NewYork:OxfordUniversityPress.

Pikhart,H.,Bobak,M.,Siegrist,J.,Pajak,A.,Rywik,S.,Kyshegyi,J.,…&Marmot,M.(2001).Psychosocialworkcharacteristicsandselfratedhealthinfourpost-communistcountries.Journal of Epidemiology & Community Health,55,624–630.

Pillai,R.,&Williams,E.A.(2004).Transformationalleadership,self-efficacy,groupcohesi-veness,commitment,andperformance.Journal of Organizational Change Management,17,144–159.

Plat,M-C.J.,Frings-Dresen,M.H.W.,&Sluiter,J.K.(2012).Whichsubgroupsoffirefightersaremorepronetowork-relateddiminishedhealthrequirements?International Archives of Occupational and Environmental Health,85,775–782.

Podsakoff,N.P.,LePine,J.A.,&LePine,M.A.(2007).Differentialchallengestressor-hindrancestressorrelationshipswithjobattitudes,turnoverintentions,turnover,andwithdrawalbehavior:Ameta-analysis.Journal of Applied Psychology,92,438–454.

Pohjonen,T.(2001).Perceivedworkabilityofhomecareworkersinrelationtoindividualandwork-relatedfactorsindifferentagegroups.Occupational Medicine,51,209–217.

Pressman,S.D.,&Cohen,S.(2005).Doespositiveaffectinfluencehealth?Psychological Bul-letin,131,925–971.

Punakallio,A.,&Lusa.S.(Eds.)(2011).Eri-ikäisten palomiesten terveys ja toimintakyky: 13 vuoden seurantatutkimus.[Health,physicalandmentalcapacityoffirefightersindiffe-rentagegroups:A13-yearfollow-up].LoppuraporttiPalosuojelurahastolle.Helsinki:Työterveyslaitos.

Punakallio,A.,Lusa,S.,Luukkonen,R.,Airila,A.,&Leino-Arjas,A.(2014).MusculoskeletalpainanddepressivesymptomsaspredictorsoftrajectoriesinworkabilityamongFinnishfirefightersat13-yearfollow-up.Journal of Occupational and Environmental Medicine,56,367–375.

Punakallio,A.,&Lusa-MoserS.(1999).Eri-ikäisten palomiesten terveys ja toimintakyky: Kolmen vuoden seurantatutkimus.[Health,physicalandmentalcapacityoffirefightersindifferentagegroups:A3-yearfollow-up].LoppuraporttiPalosuojelurahastollejaSisäasiainministe-riönpelastusosastolle.Helsinki:Työterveyslaitos.

Rachele,J.N.,Heesch,K.C.,&Washington,T.L.(2014).Wellnessprogramsatfirefighterandpoliceworkplaces:Asystematicreview.Health Behavior and Policy Review,1,302–331.

Regehr,C.,Hill,J.,Knott,T.,&Sault,B.(2003).Socialsupport,self-efficacyandtraumainnewrecruitsandexperiencedfirefighters.Stress and Health,19,189–193.

Reich,J.W.,Zautra,A.J.,&Davis,M.(2003).Dimensionsofaffectrelationships:Modelsandtheirintegrativeimplications.Review of General Psychology,7,66–83.

Rodríguez-Sánchez,A.M.,Hakanen,J.J.,Perhoniemi,R.,&Salanova,M.(2013).Withalittlehelpfrommyassistant:bufferingthenegativeeffectsofemotionaldissonanceondentistperformance.Community Dentistry and Oral Epidemiology,41,415–423.

103

reFerences

Rohmert,W.(1986).Ergonomics:conceptofwork,stressandstrain.Applied Psychology,35,159–180.

Rongen,A.,Robroek,S.J.W.,Schaufeli,W.,&Burdorf,A.(2014).Thecontributionofworkengagementtoself-perceivedhealth,workability,andsicknessabsencebeyondhealthbehaviorsandwork-relatedfactors.Journal of Occupational and Environmental Medicine,56,892–897.

Rosenberg,M.(1965).Society and the adolescent self-image.Princeton,NewJersey:PrincetonUniversityPress.

Rosenberg,M.,Schooler,C.,Schoenbach,C.,&Rosenberg,F.(1995).Globalself-esteemandspecificself-esteem:Differentconcepts,differentoutcomes.American Sociological Review,60,141–156.

Rothbard,N.P.(2001).Enrichingordepleting?Thedynamicsofengagementinworkandfamilyroles.Administrative Science Quarterly,46,655–684.

Russell,J.A.(1980).Acircumplexmodelofaffect.Journal of Personality and Social Psychology,39,1161–1178.

Russell,J.A.,&Carroll,J.M.(1999).Onthebipolarityofpositiveandnegativeaffect.Psycho-logical Bulletin,125,3–30.

Ryan,R.M.,&Deci,E.L.(2001).Onhappinessandhumanpotentials:Areviewofresearchonhedonicandeudaimonicwell-being.Annual Review of Psychology,52,141–166.

Ryan,R.M.,&Frederick,C.(1997).Onenergy,personality,andhealth:Subjectivevitalityasadynamicreflectiononwell-being.Journal of Personality,65,529–565.

Ryff,C.D.,&Keyes,C.L.M.(1995).Thestructureofpsychologicalwell-being.Journal of Personality and Social Psychology,69,719–727.

Saijo,Y.,Ueno,T.,&Hashimoto,Y.(2007).JobstressanddepressivesymptomsamongJapanesefirefighters.American Journal of Industrial Medicine,50,470–480.

Saijo,Y.,Ueno,T.,&Hashimoto,Y.(2008).Twenty-four-hourshiftwork,depressivesymp-toms,andjobdissatisfactionamongJapanesefirefighters.American Journal of Industrial Medicine,51,380–391.

Saks,A.(2006).Antecedentsandconsequencesofemployeeengagement.Journal of Managerial Psychology,21,600–619.

Salanova,M.,Agut,S.,&Peiró,J.M.(2005).Linkingorganizationalresourcesandworkenga-gementtoemployeeperformanceandcustomerloyalty:Themediationofserviceclimate.Journal of Applied Psychology,90,1217–1227.

Salanova,M.,Bakker,A.B.,&Llorens,S.(2006).Flowatwork:Evidenceforanupwardspiralofpersonalandorganizationalresources.Journal of Happiness Studies,7,1–22.

Salanova,M.,DelLíbano,M.,Llorens,S.,&Schaufeli,W.B.(2014).Engaged,workaholic,burned-outorjust9-to-5?Towardatypologyofemployeewell-being.Stress & Health,30,71–81.

104

reFerences

Salanova,M.,Llorens,S.,Cifre,E.,Martínez,I.M.,&Schaufeli,W.B.(2003).Perceivedcol-lectiveefficacy,subjectivewell-beingandtaskperformanceamongelectronicworkgroups.Small Group Research,34,43–73.

Salanova,M.,Llorens,S.,&Schaufeli,W.B.(2011).Yes,Ican,Ifeelgood,andIjustdoit!Ongaincyclesandspiralsofefficacybeliefs,affect,andengagement.Applied Psychology,60,255–285.

Salanova,M.,&Schaufeli,W.B.(2008).Across-nationalstudyofworkengagementasamediatorbetweenjobresourcesandproactivebehaviour.International Journal of Human Resource Management,19,116–131.

Salanova,M.,Schaufeli,W.B.,Xanthopoulou,D.,&Bakker,A.B.(2010).Thegainspiralofresourcesandworkengagement:Sustainingapositiveworklife.InA.B.Bakker&M.P.Leiter(Eds.),Work engagement: A handbook of essential theory and research(pp.118–131).NewYork:PsychologyPress.

Salovey,P.,Rothman,A.J.,Detweiler,J.B.,&Steward,W.T.(2000).Emotionalstatesandphysicalhealth.American Psychologist,55,110–121.

Schaufeli,W.B.,&Bakker,A.B.(2004).Jobdemands,jobresources,andtheirrelationshipwithburnoutandengagement:amulti-samplestudy.Journal of Organizational Behavior,25,293–315.

Schaufeli,W.B.,&Bakker,A.B.(2010).Definingandmeasuringworkengagement:Bringingclaritytotheconcept.InA.B.Bakker&M.P.Leiter(Eds.),Work engagement: A handbook of essential theory and research(pp.10–24).NewYork:PsychologyPress.

Schaufeli,W.B.,&Bakker,A.B.(2003).Utrecht Work Engagement Scale: Preliminary Manual.TheNetherlands:DepartmentofPsychology,UtrechtUniversity(availablefromschaufeli.com).

Schaufeli,W.B.,Bakker,A.B.,&Salanova,M.(2006).Themeasurementofworkengagementwithashortquestionnaire.Across-nationalstudy.Educational and Psychological Measu-rement,66,701–716.

Schaufeli,W.B.,Bakker,A.B.,&VanRhenen,W.(2009).Howchangesinjobdemandsandresourcespredictburnout,workengagement,andsicknessabsenteeism.Journal of Orga-nizational Behavior,30,893–917.

Schaufeli,W.B.,&Buunk,B.P.(2003).Burnout:Anoverviewof25yearsofresearchandtheorizing.InM.J.Schabracq,J.A.M.Winnubst&C.L.Cooper(Eds.),Handbook of work and health psychology(pp.383–425).Chichester:Wiley.

Schaufeli,W.,&Janczur,B.(1994).Burnoutamongnurses:APolish-Dutchcomparison.Journal of Cross Cultural Psychology,25,95–113.

Schaufeli,W.,Salanova,M.,González-Romá,V.,&Bakker,A.B.(2002).Themeasurementofengagementandburnout:Atwosampleconfirmatoryfactoranalyticapproach.The Journal of Happiness Studies,3,71–92.

Schaufeli,W.B.,&Taris,T.W.(2014).AcriticalreviewoftheJobDemands-ResourcesMo-del:Implicationsforimprovingworkandhealth.InG.F.Bauer&O.Hämmig(Eds.),Bridging occupational, organizational and public health(pp.43–68).Amsterdam:Springer.

105

reFerences

Schaufeli,W.B.,Taris,T.W.,&Bakker,A.B.(2008).Ittakestwototango.Workaholismiswor-kingexcessivelyandworkingcompulsively.InR.J.Burke&C.L.Cooper(Eds.),The long work hours culture. Causes, consequences, and choices(pp.203–226).Bingley,UK:Emerald.

Scheier,M.F.,&Carver,C.S.(1992).Effectsofoptimismonpsychologicalandphysicalwell-being:theoreticaloverviewandempiricalupdate.Cognitive Therapy and Research,16,201–228.

Scheier,M.F.,&Carver,C.S.(1985).Optimism,coping,andhealth:Assessmentandimpli-cationsofgeneralizedoutcomeexpectancies.Health Psychology,4,219–247.

Schimmack,U.,&Diener,E.(2003).Predictivevalidityofexplicitandimplicitself-esteemforsubjectivewell-being.Journal of Research in Personality,37,100–106.

Seitsamo,J.,vonBonsdorff,M.E.,Ilmarinen,J.,vonBonsdorff,M.B.,Nygård,C-H.,Ranta-nen,T.,&Klockars,M.(2011).Workabilityandlater-lifehealth:A28-yearlongitudinalstudyamongFinnishmunicipalworkers.InC-H.Nygård,M.Savinainen,T.Kirsi&K.Lumme-Sandt(Eds.),Age management during the life course(pp.391–398).Proceedingsofthe4thsymposiumonworkability.Tampere:TampereUniversityPress.

Seligman,M.E.P.(2002).Authentic happiness: Using the new positive psychology to realize your potential for lasting fulfillment.NewYork:FreePress.

Seligman,M.E.P.,&Csikszentmihalyi,M.(2000).Positivepsychology:Anintroduction.American Psychologist,55,5–14.

Seppälä,P.,Mauno,S.,Feldt,T.,Hakanen,J.,Kinnunen,U.,Tolvanen,A.,&Schaufeli,W.(2009).TheconstructvalidityoftheUtrechtWorkEngagementScale:multisampleandlongitudinalevidence.Journal of Happiness Studies,10,459–481.

Seppälä,P.,Mauno,S.,Kinnunen,M-L.,Feldt,T.,Juuti,T.,Tolvanen,A.,&Rusko,H.(2012).Isworkengagementrelatedtohealtycardiacautonomicactivity?EvidencefromafieldstudyamongFinnishwomenworkers.Journal of Positive Psychology,7,95–106.

Shiri,R.,Karppinen,J.,Leino-Arjas,P.,Solivieva,S.,&Viikari-Juntura,E.(2010).Theassocia-tionbetweensmokingandlowbackpain:Ameta-analysis.American Journal of Medicine,123,87.e7–87.e35.

Shirom,A.(2010).Feelingenergeticatwork:Onvigor’santecedents.In:A.B.Bakker&M.P.Leiter(Eds.),Work engagement: a handbook of essential theory and research(pp.69–84).NewYork:PsychologyPress.

Shirom,A.,Toker,S.,Berliner,S.,Shapira,I.,&Melamed,S.(2008).Theeffectsofphysicalfitnessandfeelingvigorousonself-ratedhealth.Health Psychology,27,567–575.

Shirom,A.,Toker,S.,Berliner,S.,Shapira,I.,&Melamed,S.(2006).Work-relatedvigorandjobsatisfactionrelationshipswithinflammationbiomarkersamongemployedadults.InA.Delle(Ed.),Dimensions of well-being: Research and intervention(pp.254–274).Milan:FrancoAngeli.

Siegrist,J.(1996).Adversehealtheffectsofhigh-effort/low-rewardconditions.Journal of Occu-pational Health Psychology,1,27–41.

106

reFerences

Simpson,M.R.(2009).Engagementatwork:Areviewoftheliterature.International Journal of Nursing Studies,46,1012–1024.

SluiterJ.K.,deCroonE.M.,Meijman,T.F.,&Frings-Dresen,M.H.W.(2003).Needforrecoveryfromworkrelatedfatigueanditsroleinthedevelopmentandpredictionofsub-jectivehealthcomplaints.Occupational and Environmental Medicine,60(Suppl1),i62–i70.

Sluiter,J.K.,&Frings-Dresen,M.H.W.(2007).Whatdoweknowaboutageingatwork?Evidence-basedfitnessfordutyandhealthinfirefighters.Ergonomics,50,1897–1913.

Snyder,S.R.,&Lopez,S.J.(Eds.)(2002).Handbook of positive psychology.NewYork:OxfordUniversityPress.

Sonnentag,S.,&Niessen,C.(2008).Stayingvigorousuntilworkisover:Theroleoftraitvigour,day-specificworkexperiencesandrecovery.Journal of Occupational and Organizational Psychology,81,435–458.

Stansfeld,S.,&Candy,B.(2006).Psychosocialworkenvironmentandmentalhealth–ameta-analyticreview.Scandinavian Journal of Work, Environment & Health,32,443–462.

Steel,P.,Schmidt,J.,&Shultz,J.(2008).Refiningtherelationshipbetweenpersonalityandsubjectivewell-being.Psychological Bulletin,134,138–161.

Stoetzer,U.,Ahlberg,G.,Johansson,G.,Bergman,P.,Hallsten,L.,Forsell,Y.,&Lundberg,I.(2009).Problematicinterpersonalrelationshipsatworkanddepression:ASwedishprospectivecohortstudy.Journal of Occupational Health,51,144–151.

Thompson,R.A.,Winer,A.C.,&Goodvin,R.(2011).Theindividualchild:Temperament,emotion,self,andpersonality.InM.E.Lamb&M.H.Bornstein(Eds.),Social and per-sonality development. An advanced textbook(pp.217–258).NewYork:PsychologyPress.

Tims,M.,Bakker,A.B.,&Derks,D.(2012).Developmentandvalidationofthejobcraftingscale.Journal of Vocational Behavior,80,173–186.

Tims,M.,Bakker,A.B.,Derks,D.,&VanRhenen,W.(2013).Jobcraftingatteamandindi-viduallevel:Implicationsforworkengagementandperformance.Group & Organization Management,38,427–454.

Tuckey,M.R.,&Hayward,R.(2011).Globalandoccupation-specificemotionalresourcesasbuffersagainsttheemotionaldemandsoffire-fighting.Applied Psychology,60,1–23.

Tuomi,K.,Eskelinen,L.,Toikkanen,J.,Järvinen,E.,Ilmarinen,J.,&Klockars,M.(1991).Workloadandindividualfactorsaffectingworkabilityamongagingmunicipalemployees.Scandinavian Journal of Work, Environment & Health,17(suppl1),128–134.

Tuomi,K.,Ilmarinen,J.,Jahkola,A.,Katajarinne,L.,&Tulkki,A.(1998).Work Ability Index(2ndreviseded.).Helsinki:InstituteofOccupationalHealth.

Tuomi,K.,Toikkanen,J.,Eskelinen,L.,Backman,A-L.,Ilmarinen,J.,Järvinen,E.,&Klockars,M.(1991).Mortality,disabilityandchangesinoccupationamongagingmunicipalemployees.Scandinavian Journal of Work, Environment & Health,17(suppl1),58–66.

Uchino,B.N.(2004).Social support and physical health.Understanding the health consequences of relationships.NewHaven,CT:YaleUniversityPress.

107

reFerences

VandenBerg,T.I.J.,Elders,L.A.M.,Zwart,B.C.H.,&Burdorf,A.(2009).Theeffectsofwork-relatedandindividualfactorsontheWorkAbilityIndex:asystematicreview.Occupational and Environmental Medicine,66,211–220.

VandenBroek,A.,DeCuyper,N.,DeWitte,H.,&Vansteenkiste,M.(2010).Notalljobdemandsareequal:DifferentiatingjobhindrancesandjobchallengesintheJobDemands–Resourcesmodel.European Journal of Work and Organizational Psychology,19,735–759.

vanHorn,J.E.,Taris,T.W.,Schaufeli,W.B.,&Schreurs,P.J.G.(2004).Thestructureofoccupationalwell-being:AstudyamongDutchteachers.Journal of Occupational and Organizational Psychology,77,365–375.

Viikari-Juntura,E.,Rauas,S.,Martikainen,R.,Kuosma,E.,Riihimäki,H.,Takala,E-P.,&Saarenmaa,K.(1996).Validityofself-reportedphysicalworkloadinepidemiologicstu-diesonmusculoskeletaldisorders.Scandinavian Journal of Work, Environment & Health,22,251–259.

Wanous,J.P.,Reichers,A.E.,&Hudy,M.J.(1997).Overalljobsatisfaction:Howgoodaresingle-itemmeasures?Journal of Applied Psychology,82,247–252.

Warr,P.(1990).Themeasurementofwell-beingandotheraspectsofmentalhealth.Journal of Occupational Psychology,63,193–201.

Warr,P.B.(2007).Work, happiness, and unhappiness.NewYork:Erlbaum.

Watson,D.,Clark,L.A.,&Tellegen,A.(1988).Developmentandvalidationofbriefmeasuresofpositiveandnegativeaffect:ThePANASscales.Journal of Personality and Social Psychol-ogy,54,1063–1070.

Watson,D.,&Tellegen,A.(1999).Issuesindimensionalstructureofaffect–Effectsofdescrip-tors,measurementerror,andresponseformats:CommentonRussellandCarroll(1999).Psychological Bulletin,125,601–610.

Webster,J.R.,Beehr,T.A.,&Christiansen,N.D.(2010).Towardabetterunderstandingoftheeffectsofhindranceandchallengestressorsonworkbehaviour.Journal of Vocational Behavior,76,68–77.

Weevers,H.J.,vanderBeek,A.J.,Anema,J.R.,vanderWal,G.,&vanMechelen,W.(2005).Work-relateddiseaseingeneralpractice:asystematicreview.Family Practice,22,197–204.

Weiss,H.M.,&CropanzanoR.(1996).Affectiveeventstheory:Atheoreticaldiscussionofthestructure,causes,andconsequencesofaffectiveexperiencesatwork.Research in Organi-zational Behavior,18,1–74.

Williams,P.G.,Smith,T.W.,&Cribbet,M.R.(2008).Personalityandhealth:Currentevi-dence,potentialmechanisms,andfuturedirections.InG.J.Boyle,G.Matthews&D.H.Saklofske(Eds.),The SAGE handbook of personality theory and assessment(pp.635–658).London:SAGEPublications.

Williamson,D.L.,&Carr,J.(2009).Healthasaresourceforeverydaylife:advancingtheconceptualization.Critical Public Health,19,107–122.

108

reFerences

Wilson,M.,DeJoy,D.,Vandenberg,R.,Richardson,H.,&McGrath,A.(2004).Workcha-racteristicsandemployeehealthandwell-being:Testofamodelofhealthyorganization.Journal of Occupational and Organizational Psychology,77,565–588.

WorldHealthOrganization(WHO).(1986).Ottawa charter for health promotion.Firstinter-nationalconferenceonhealthpromotion.Ottawa:WHO.

Wrzesniewski,A.,&Dutton,J.E.(2001).Craftingajob:Revisioningemployeesasactivecraftersoftheirwork.Academy of Management Review,26,179–201.

Xanthopoulou,D.,Bakker,A.B.,Demerouti,E.,&Schaufeli,W.B.(2009).Reciprocalre-lationshipsbetweenjobresources,personalresources,andworkengagement.Journal of Vocational Behavior,74,235–244.

Xanthopoulou,D.,Bakker,A.B.,Demerouti,E.,&Schaufeli,W.B.(2007).Theroleofpersonalresourcesinthejobdemands-resourcesmodel.International Journal of Stress Management,14,121–141.

Zapf,D.,Dormann,C.,&Frese,M.(1996).Longitudinalstudiesinorganizationalstressresearch:Areviewoftheliteraturewithreferencetomethodologicalissues.Journal of Occupational Health Psychology,1,145–169.

Øverland,S.,Harvey,S.B.,Knudsen,A.K.,Mykletun,A.,&Hotopf,M.(2012).WidespreadpainandmedicallycertifieddisabilitypensionintheHordalandHealthStudy.European Journal of Pain,16,611–620.


People and WorkResearch Reports 109

Auli Airila

Work characteristics, personal resources, and em

ployee well-being: A

longitudinal study among Finnish firefighters

Airila A

uli

In today’s dynamic work life, employees are expected not only to be healthy and able to work efficiently, but also to be highly engaged and motivated. However, at the same time, the prevalence of mental health disorders and diseases of the musculoskeletal systems is also high. Therefore, understanding and enhancing employee well-being as a whole is essential. The present study examined the longitudinal effects of work characteristics (i.e., job demands and resources), personal resources and lifestyle on employee well-being, using data on Finnish firefighters.

This study showed that job demands, job resources, personal resources, and lifestyle all affect employee well-being over time. It offers new information for researchers, policy-makers, HRM at workplaces, and occupational health services about possibilities to enhance employee well-being, particularly work engagement, and consequently prevent ill-health and promote long working careers.

109

ISBN 978-952-261-525-1

Orders:Finnish Institute of Occupational HealthArinatie 3FI-00370 HelsinkiFinland

E-mail [email protected]

ISBN 978-952-261-525-1 (paperback)ISBN 978-952-261-526-8 (PDF)

ISSN-L 1237-6183ISSN 1237-6183

Cover picture: Saimi Airila

Documents

Work characteristics, personal resources, and employee well-being