Physical activity, body mass index and heart rate variability-based stress and recovery in 16 275 Finnish employees : a cross-sectional study

This is an elec onic ep in o he o iginal a icle. This ep in may di e om he o iginal in pagina ion and ypog aphic de ail. Au ho (s): Ti le: Yea : Ve sion: Please ci e he o iginal e sion: All ma e ial supplied ia JYX is p o ec ed by copy igh and o he in ellec ual p ope y igh s, and duplica ion o sale o all o pa o any o he eposi o y collec ions is no pe mi ed, excep ha ma e ial may be duplica ed by you o you esea ch use o educa ional pu poses in elec onic o p in o m. You mus ob ain pe mission o any o he use. Elec onic o p in copies may no be o e ed, whe he o sale o o he wise o anyone who is no an au ho ised use . Physical ac i i y, body mass index and hea a e a iabili y-based s ess and eco e y in 16 275 Finnish employees : a c oss-sec ional s udy Föh , Tiina; Pie ilä, Julia; Helande , Elina; Myllymäki, Te o; Lindholm, Ha i; Rusko, Heikki; Kujala, U ho Föh , T., Pie ilä, J., Helande , E., Myllymäki, T., Lindholm, H., Rusko, H., & Kujala, U. (2016). Physical ac i i y, body mass index and hea a e a iabili y-based s ess and eco e y in 16 275 Finnish employees : a c oss-sec ional s udy. BMC Public Heal h, 16, A icle 701. h ps://doi.o g/10.1186/s12889-016-3391-4 2016 RESEARCH ARTICLE Open Access Physical ac i i y, body mass index and hea a e a iabili y-based s ess and eco e y in 16 275 Finnish employees: a c oss-sec ional s udy Tiina Föh 1* , Julia Pie ilä 2 , Elina Helande 2 , Te o Myllymäki 3 , Ha i Lindholm 4 , Heikki Rusko 5 and U ho M. Kujala 1 Abs ac Backg ound: Physical inac i i y, o e weigh , and wo k- ela ed s ess a e majo conce ns oday. Psychological s ess causes physiological esponses such as educed hea a e a iabili y (HRV), owing o a enua ed pa asympa he ic and/o inc eased sympa he ic ac i i y in ca diac au onomic con ol. This s udy’s pu pose was o in es iga e he ela ionships be ween physical ac i i y (PA), body mass index (BMI), and HRV-based s ess and eco e y on wo kdays, among Finnish employees. Me hods: The pa icipan s in his c oss-sec ional s udy we e 16 275 indi iduals (6863 men and 9412 women; age 18–65 yea s; BMI 18.5–40.0 kg/m 2 ). Assessmen s o s ess, eco e y and PA we e based on HRV da a om bea - o-bea R-R in e al eco ding (mainly o e 3 days). The alida ed HRV-de i ed a iables ook in o accoun he dynamics and indi iduali y o HRV. S ess pe cen age ( he p opo ion o s ess eac ions, wo kday and wo king hou s), and s ess balance ( a io be ween eco e y and s ess eac ions, sleep) desc ibe he amoun o physiological s ess and eco e y, espec i ely. Va iables desc ibing he in ensi y (i.e. magni ude o ecognized eac ions) o physiological s ess and eco e y we e s ess index (wo kday) and eco e y index (sleep), espec i ely. Mode a e o igo ous PA was measu ed and pa icipan s di ided in o he ollowing g oups, based on calcula ed weekly PA: inac i e (0 min), low (0 < 150 min), medium (150–300 min), and high (>300 min). BMI was calcula ed om sel - epo ed weigh and heigh . Linea models we e employed in he main analyses. Resul s: High PA was associa ed wi h lowe s ess pe cen ages (du ing wo kdays and wo king hou s) and s ess balance. Highe BMI was associa ed wi h highe s ess index, and lowe s ess balance and eco e y index. These esul s we e simila o men and women (P< 0.001 o all). Conclusion: Independen o age and sex, high PA was associa ed wi h a lowe amoun o s ess on wo kdays. Addi ionally, lowe BMI was associa ed wi h be e eco e y du ing sleep, exp essed by a g ea e amoun and magni ude o eco e y eac ions, which sugges s ha PA in he long e m esul ing in imp o ed i ness has a posi i e e ec on eco e y, e en hough high PA may dis u b eco e y du ing he ollowing nigh . Ob iously, se e al ac o s ou side o he s udy could also a ec HRV-based s ess. Keywo ds: Body mass index, Hea a e a iabili y, Physical ac i i y, Physiological s ess, S ess, S ess assessmen * Co espondence: [email p o ec ed] 1 Depa men o Heal h Sciences, Uni e si y o Jy äskylä, P.O. Box 35FIN-40014 Jy äskylä, Finland Full lis o au ho in o ma ion is a ailable a he end o he a icle © 2016 The Au ho (s). Open Access This a icle is dis ibu ed unde he e ms o he C ea i e Commons A ibu ion 4.0 In e na ional License (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons license, and indica e i changes we e made. The C ea i e Commons Public Domain Dedica ion wai e (h p://c ea i ecommons.o g/publicdomain/ze o/1.0/) applies o he da a made a ailable in his a icle, unless o he wise s a ed. Föh e al. BMC Public Heal h (2016) 16:701 DOI 10.1186/s12889-016-3391-4 Backg ound Physical ac i i y (PA) is known o ha e posi i e e ec s on heal h [1, 2]. Rou ine PA educes s ess and enhances psychological wellbeing, which is pa icula ly impo an o he p e en ion and managemen o ca dio ascula disease, among o he ch onic diseases [3]. Regula PA is known o educe he isk o many ad e se heal h ou - comes. Some PA is be e han none; howe e , o mos heal h ou comes, addi ional bene i s a e achie ed i he amoun o PA inc eases h ough highe in ensi y, g ea e equency, and/o longe du a ion. Acco ding o he 2008 Physical Ac i i y Guidelines o Ame icans, mos heal h bene i s occu wi h a leas 150 o al minu es o mode a e in ensi y o a leas 75 min o igo ous in en- si y ae obic PA pe week. Howe e , addi ional bene i s occu wi h mo e PA [4, 5]. In addi ion o he bene icial e ec s o PA on physical heal h, hese guidelines a e also ele an o men al heal h [6]. Al hough leisu e- ime PA has inc eased among Finnish adul s [7], physical inac i - i y is a majo p oblem and isk o heal h, in all coun- ies. Fu he mo e, physical inac i i y is associa ed wi h being o e weigh [8] and he cu en a e o o e weigh adul s wo ldwide has been desc ibed as an epidemic o e en a pandemic. This si ua ion is a majo public heal h isk because being o e weigh is associa ed wi h diseases including co ona y hea disease, s oke, diabe es and cance [9]. Toge he wi h physical inac i i y and o e weigh , s ess a wo k is a majo public heal h isk. I may e en lead o ca dio ascula disease [10] wi hou comple e eco e y [11]. S ess has been shown o educe pa icipa ion in leisu e- ime PA [12, 13]. Fu he mo e, wo kplace s ess may p edic a u u e inc eased isk o insu icien PA [14]. No mal weigh is associa ed wi h good sel - epo ed sub- jec i e heal h [15], including low s ess le els [16, 17]. E i- dence sugges s ha psychosocial s ess is associa ed wi h he de elopmen o adiposi y [18]. Howe e , acco ding o p e ious s udies, he associa ion be ween subjec i e s ess and body composi ion is inconsis en , wi h e idence bo h suppo ing [16, 19] and e u ing [20, 21] he idea ha s ess is associa ed wi h adiposi y. A ecen sys ema ic e- iew epo ed ha he associa ions o psychosocial ac o s a wo k wi h weigh - ela ed ou comes we e weak and somewha inconsis en [22]. Psychological s ess causes sympa he ic esponses in he au onomic ne ous sys em (ANS), such as educed hea a e a iabili y (HRV) [23]. HRV e e s o he a i- a ion in in e als be ween hea bea s and e lec s ca diac au onomic modula ion. Physiological s ess can be de- ined as an inc eased body ac i a ion le el, when sympa- he ic ac i i y domina es he ANS and pa asympa he ic ac i a ion is low. S ess is associa ed wi h educed HRV, owing o a enua ed pa asympa he ic and/o inc eased sympa he ic ac i i y in ca diac au onomic con ol. Reco e y e e s o a educed body ac i a ion le el, when pa asympa- he ic ac i a ion domina es he ANS o e sympa he ic ac i i y [24–26]. HRV analysis can be used as a comple- men a y ool o assess gene al heal h [27]. HRV analysis du ing sleep has he po en ial o explo e he sleeping b ain, wi h possible implica ions o men al heal h [28]. P e ious HRV-s udies ha e mainly used adi ional ime-domain and equency-domain measu es o HRV, such as oo mean squa e o successi e R-R in e als (RMSSD) and he a io o low equency powe o high equency powe (LF/ HF a io). The adi ional measu es o HRV ep esen he a e age le el o he au onomic ac i i y o e a pe iod o he ime. Ca diac au onomic ac i i y is e y dynamic and a ies du ing he day depending on s ess, eco e y and PA. The e o e, he usabili y o he adi ional measu es o HRV is limi ed in eal-li e condi ions. Addi ionally, hese mea- su es a e e y indi idual which u he limi s hei usabili y in s ess assessmen and clinical wo k. Howe e , i is also possible o p o ide applied hea a e (HR) and HRV- de i ed s ess and eco e y a iables ha ake in o accoun he dynamic changes in au onomic ac i i y and indi idual- i y o HRV including in o ma ion ha is di icul o ob ain om adi ional measu es o HRV. The majo i y o p e ious s udies on he associa ion o PA wi h s ess ha e used subjec i e assessmen me hods o adi ional measu es o HRV in he assessmen o s ess. The p e ious s udies suppo he associa ion o PA wi h inc eased HRV [29, 30]. Howe e , accu a e and objec i e me hods a e needed o eliably assess PA, as well as o assess HRV-based s ess and eco e y in eal- li e. By u ilizing a me hod ha acknowledges he dynam- ics and indi iduali y in HRV in eal-li e, he aim o his s udy was o in es iga e he ex en o which PA and BMI a e associa ed wi h HRV-based indica o s o s ess and eco e y on wo kdays. The s udy was conduc ed among 16 275 Finnish employees who had pa icipa ed in bea - o-bea R-R in e al eco ding as a pa o li es yle counseling be ween 2007 and 2015. Mo e speci ically, ac- coun ing o age and sex, we in es iga ed he p e alence o s ess and eco e y acco ding o he pa icipan s’ob- jec i ely measu ed PA le el and sel - epo ed body mass index (BMI). Uniqueness o he p esen s udy is in he in- di idual and dynamic me hod used in he assessmen o physiological s ess and eco e y. Me hods S udy design and pa icipan s This c oss-sec ional s udy in es iga ed he amoun and in- ensi y o objec i e HRV-based s ess and eco e y on wo kdays in a eal-li e sample o 16 275 Finnish em- ployees (6863 men and 9412 women; age 18–65 yea s; BMI 18.5–40.0 kg/m 2 ). The pa icipan s nonselec i ely ep esen a c oss-sec ion o ypical Finnish employees in- cluding bo h manual and non-manual labou employees. Föh e al. BMC Public Heal h (2016) 16:701 Page 2 o 13 The majo i y o he pa icipan s we e appa en ly heal hy wi hou ch onic diseases. The exclusion c i e ia o pa icipa ion in he R-R in e al eco dings included se- e e ca diac disease, e y high blood p essu e (≥180/ 100 mmHg), ype 1 o 2 diabe es wi h au onomic neu - opa hy, se e e neu ological disease, e e o o he acu e disease, and BMI >40 kg/m 2 . These exclusion c i e ia ep- esen ed by he analysis so wa e manu ac u e a e p e- sen ed in de ail p e iously [31]. The cha ac e is ics o he pa icipan s a e p esen ed in Table 1. Da a collec ion The no el me hodology used o de e mine he pa ici- pan s’s ess, eco e y and le el o weekly PA, was based on HRV da a om bea - o-bea R-R in e al eco dings. These eco dings we e olun a ily pe o med on em- ployees as a pa o he p e en i e occupa ional heal h ca e p og ams p o ided by hei employe s be ween 2007 and 2015. The clinical pu pose o hese measu e- men s is p esen ed comp ehensi ely in a p e ious pape by Mu ikainen e al. [31]. The da a eco dings used in he p e ious s udy we e ga he ed be ween 2007 and 2013, wi h a s udy popula ion size o 9554. These da a we e used in he p esen s udy, supplemen ed wi h e- co dings om 2014 o 2015. The s udy had a u he in- clusion c i e ion o a minimum o 4.5 h bea - o-bea R-R in e al eco ding du ing sleep a e a wo kday. Ano he inclusion c i e ion was he a ailabili y o R-R in e al da a, including a leas one wo kday (≥4 h o wo k) and one day o , wi h a measu emen pe iod o 16–30 h/day ( om wake-up o wake-up). Pa icipan s who had con- sumed alcohol on he moni o ing days we e excluded. In o ma ion abou wo kdays, wo king hou s, days o and sleep pe iods was ob ained om dia ies ha he pa - icipan s we e eques ed o keep du ing he measu e- men pe iod. The analyzed da a consis ed o success ully eco ded (measu emen e o <15 % and <30-min e- co ding b eak) days. The low o he pa icipan s in- cluded in he analysis is p esen ed in Fig. 1. HRV-based assessmen o PA, s ess and eco e y Ambula o y bea - o-bea R-R in e al da a we e used o de e mine he amoun and in ensi y o PA, s ess and e- co e y. Using he Fi s bea Bodygua d de ice (Fi s bea Technologies L d., Jy äskylä, Finland), eal-li e R-R in e al da a we e eco ded, usually o e 3 days ( ypic- ally wo wo kdays and one day o ) and analyzed using Fi s bea Analysis Se e so wa e ( e sion 6.3, Fi s bea Technologies L d.), which included a powe ul a i ac de ec ion and co ec ion ea u e o i egula ec opic bea s and signal noise. The so wa e calcula es HRV in- dices second-by-second using he sho - ime Fou ie ans o m me hod, and calcula es HR- and HRV-de i ed a iables o espi a ion a e, oxygen consump ion, on-o kine ics (inc easing o dec easing HR), and pa ame e s desc ibing excess pos -exe cise oxygen consump ion using neu al ne wo ks. The ea e , he so wa e di ides he measu emen da a in o cohe en da a segmen s and ca ego izes hese segmen s in o di e en physiological s a es, such as PA o di e en in ensi ies, s ess and e- co e y [32–34], by aking in o accoun indi idual cha - ac e is ics (e.g. indi idual le els and scales o HR and HRV, and he indi idual ela ionships be ween HRV and au onomic con ol) [35]. The ca ego iza ion o he da a is desc ibed in Addi ional ile 1: Table S1. Mo e in o ma- ion abou his analysis me hod is a ailable in a pape by Fi s bea Technologies L d. [36]. De ec ion o s ess and eco e y a iables A e da a ca ego iza ion, he HRV-based a iables de- sc ibing he amoun and in ensi y o s ess and eco e y on wo kdays we e de ec ed. S ess pe cen ages (i.e. p o- po ions o s ess eac ions, du ing he day and du ing wo king hou s) and s ess balance ( a ios be ween eco - e y and s ess eac ions du ing sleep) desc ibe he amoun o s ess and eco e y, espec i ely. The a iables desc ib- ing he in ensi y (i.e. magni ude o ecognized eac ions) o s ess and eco e y we e s ess index (du ing he day) and eco e y index (du ing sleep), espec i ely. These a i- ables and hei calcula ions a e p esen ed in Addi ional ile 1: Table S1. The co ela ion coe icien be ween wo con- secu i e wo kdays a ied om 0.74 o 0.88 o he ad- i ional HRV a iables, om 0.64 o 0.93 o HRV-de i ed a iables o s ess, and om 0.42 o 0.49 o HRV-de i ed a iables o eco e y du ing sleep. Calcula ion o weekly PA Backg ound in o ma ion abou age, sex, sel - epo ed heigh and weigh , and sel - epo ed PA class [37] modi- ied om Ross and Jackson [38], was collec ed in Table 1 Cha ac e is ics o he pa icipan s Va iable All (n= 16275) Men (n= 6863) Women (n= 9412) Mean ± SD Min Max Mean ± SD Min Max Mean ± SD Min Max Age 44.8 ± 9.9 18.0 65.0 44.5 ± 9.9 18.0 65.0 45.0 ± 9.9 18.0 65.0 Body mass index (kg/m 2 ) 26.0 ± 4.1 18.5 40.0 26.6 ± 3.5 18.6 40.0 25.5 ± 4.4 18.5 40.0 Sel - epo ed ac i i y class 0–10 4.8 ± 1.8 0.0 10.0 4.9 ± 1.9 0.0 10.0 4.8 ± 1.8 0.0 10.0 Physical ac i i y (mins/week) 186 ± 227 0 2629 246 ± 258 0 2629 142 ± 189 0 1865 Föh e al. BMC Public Heal h (2016) 16:701 Page 3 o 13 conjunc ion wi h R-R in e al eco dings using ques ion- nai es. Backg ound in o ma ion was used o es ima e maximal HR [39] and maximal VO 2 [40] which we e hen used in he es ima ion o VO 2 . The maximal HR used o u he calcula ions was co ec ed acco dingly i a pe iod wi h HR highe han he es ima ed maximal was ound om he eco ding. F om he second-by- second VO 2 es ima ions, each pa icipan ’s mean VO 2 o each minu e o he measu emen day was calcula ed. The minu e-by-minu e VO 2 es ima ions we e hen con- e ed o mul iples o he es ing me abolic a e (MET) by di iding he VO 2 alues by 3.5. The o al numbe o 1-min segmen s wi hin he ollowing h esholds: mode - a e PA 3 o <6 METs and igo ous PA ≥6 METs, du ing each measu emen day (including days o ), we e calcu- la ed. Con inuous bou s o PA las ing o ≥10 min we e included in he es ima ion o weekly PA. These con inu- ous bou s o PA we e calcula ed sepa a ely o wo kdays and days o , and, i he measu emen pe iod included wo o mo e wo kdays o days o , an a e age was cal- cula ed. The ac i i y minu es sco e o each day (mod- e a e PA minu es + igo ous PA minu es × 2) was calcula ed. The ea e , he amoun o PA was ex apo- la ed using he ollowing o mula: PA minu es pe week = (5 × mean wo kday ac i i y sco e) + (2 × mean day-o ac i i y sco e). These calcula ions ha e been p e iously desc ibed in mo e de ail [31]. Based on he weekly PA minu es, he pa icipan s we e di ided in o he ollowing PA g oups: inac i e (0 min), low (0 < 150 min), medium (150–300 min) and high (>300 min). Assessmen o body composi ion BMI was calcula ed om he sel - epo ed weigh and heigh (kg/m 2 ). The pa icipan s we e hen di ided in o he ollowing g oups: no mal weigh (18.5 o <25 kg/m 2 ), o e weigh (25 o <30 kg/m 2 ) and obese (30–40 kg/m 2 ). Analysis Da a p ocessing and s a is ical analysis we e pe o med using R 3.2.2 e sion (R Founda ion o S a is ical Com- pu ing). P- alues we e wo-sided and a p- alue o <0.05 was conside ed s a is ically signi ican . Because o he size o he da a, 99 % con idence in e als (CIs) we e de- e mined (Fig. 2) ins ead o con en ional 95 % CIs. The main ou come a iables o he s udy we e s ess pe cen age and s ess index, calcula ed o he whole day, s ess pe cen age calcula ed o wo king hou s, and s ess balance and eco e y index calcula ed o sleep. These a iables we e de i ed om he bea - o-bea R-R in e al eco dings on wo kdays. Fo a mo e de ailed de- sc ip ion o he a iables see Addi ional ile 1: Table S1. In addi ion, HR and adi ional HRV pa ame e s, includ- ing RMSSD and he LF/HF a io, we e calcula ed om he bea - o-bea R-R in e al eco dings on wo kdays. These a iables we e calcula ed sepa a ely o waking hou s and sleep, and RMSSD was calcula ed using a 5- min window. I he measu emen pe iod o a subjec in- cluded wo o mo e wo kdays, an a e age was calcula ed and he mean alues o he ou come a iables we e used in he analysis. Fig. 1 Flow o he pa icipan s and measu emen days included in he analysis (BMI, body mass index) Föh e al. BMC Public Heal h (2016) 16:701 Page 4 o 13 Fo he desc ip i e s a is ics, he means and s anda d de ia ions o he ou come a iables we e calcula ed sep- a a ely o men and women, and s a i ied based on PA, BMI and age. Di e ences in he ou come a iables be- ween PA, BMI and age g oups we e es ed using he K uskal-Wallis es . The esul s a e shown in Addi ional ile 2: Tables S2-S4. To show he e ec s o BMI and PA g oup on he HRV-based s ess and eco e y a iables, he age-con olled mean alues and 99 % CIs o he HRV-based s ess and eco e y a iables, by BMI and PA g oup, a e p esen ed in Fig. 2. Linea models we e employed o s udy he e ec s o PA g oup, BMI and age on HRV-based s ess and eco - e y a iables. In he models, age and BMI we e inco po- a ed as con inuous p edic o a iables and objec i ely measu ed PA g oup was inco po a ed as a ca ego ical p edic o a iable. The models we e gene a ed sepa a ely o men and women. The e e ence alue o age was se o 18 yea s and o BMI o 18.5 kg/m 2 . A simple linea leas squa es eg ession model (p ocedu e lm in R) was applied o p edic he s ess pe cen age du ing he day. As con i med by isual inspec ion, he assump ion o Fig. 2 S ess and eco e y by physical ac i i y and body mass index g oups wi h age-con olled mean alues and 99 % CIs o ou pu a iables. Physical ac i i y g oups: inac i e (0 min/week), ed ●; low (0 < 150 min/week), black ▲; medium (150–300 min/week), blue ■; high (>300 min/ week), g een ♦ Föh e al. BMC Public Heal h (2016) 16:701 Page 5 o 13 linea eg ession conside ing he no mal dis ibu ion o he esiduals was no ul illed o s ess pe cen age du - ing wo king hou s, s ess index and eco e y index. Thus, a Box-Cox ans o ma ion was applied on hese dependen a iables [41]. The Box-Cox coe icien was de e mined by maximizing he log-likelihood unc ion and was ounded o wo decimal places be o e ans o m- a ion. Tobi eg ession model (p ocedu e glm using i e a- i ely eweigh ed leas squa es in R) was applied o modeling s ess balance wi h a ixed lowe and uppe limi o −1 and 1, espec i ely. The in e ac ions o he p edic- o s we e no included in he inal eg ession models be- cause he coe icien o de e mina ion o he in e ac ion models was only a ew pe cen age poin s g ea e han o he simple models. Resul s The o al numbe o wo kdays included in he analysis was 28 314, wi h measu emen s ob ained om 16 275 pa icipan s (men 6863; women 9412). The pa icipan s’ cha ac e is ics a e shown in Table 1. The pa icipan s’ mean age was 44.8 yea s (44.5 yea s o men; 45.0 yea s o women) and he mean BMI was 26.0 kg/m 2 (26.6 kg/ m 2 o men; 25.5 kg/m 2 o women). The pa icipan s’ mean sel - epo ed ac i i y class was 4.8 (4.9 o men; 4.8 o women) indica ing ha , on a e age, he pa ici- pan s we e in ol ed in PA 2–3 imes pe week and hei o al weekly PA was abou 2 h. The mean weekly mi- nu es o objec i e moni o ing-based PA was 186 (246 o men; 142 o women). The mean weekly minu es o PA in he g oup o low PA was 78 o men and 74 o women, in he g oup o medium PA 222 o men and 215 o women, and in he g oup o high PA 545 o men and 496 o women. The numbe o pa icipan s in he PA, BMI, and age g oups is p esen ed in Addi ional ile 3: Table S5. Di e ences in ou come a iables be ween PA g oups (Addi ional ile 2: Table S2) we e s a is ically signi ican excep o LF/HF a io du ing waking hou s and sleep, s ess balance in men, and HR and s ess balance in women. Fo bo h men and women, he high PA g oup had he highes RMSSD (du ing waking hou s and du - ing sleep) and eco e y index, and he lowes s ess pe - cen age (du ing he day and du ing wo king hou s) and s ess index. Di e ences in ou come a iables be ween BMI g oups we e s a is ically signi ican o bo h men and women (Addi ional ile 2: Table S2). No mal-weigh indi iduals had he highes RMSSD (bo h du ing waking hou s and du ing sleep), s ess balance and eco e y index, he lowes s ess pe cen age du ing he day and he lowes s ess index. S ess pe cen age du ing wo king hou s was lowes in obese indi iduals. In bo h sexes, di e ences in ou come a iables we e s a is ically signi ican be ween age g oups, excep o HR du ing sleep in women (Addi ional ile 2: Table S3). The younges age g oup (18–30 yea s) had he highes RMSSD (bo h du ing waking hou s and du ing sleep), s ess balance and eco e y index, and he lowes s ess index. S ess pe cen ages du ing he day we e lowes in he younges age g oup in women, and in he oldes age g oup (51–65 yea s) in men. S ess pe cen ages du ing wo king hou s we e lowes in he oldes age g oup in bo h sexes. Figu e 2 shows he e ec o PA and BMI g oup on he s ess and eco e y a iables wi h he e ec o age con- olled. The high PA g oup had he lowes mean s ess pe cen age du ing he day and du ing wo king hou s in all h ee BMI g oups, a e adjus men o age (Fig. 2). Mean s ess index alues inc eased as he BMI g oup changed om no mal weigh o o e weigh and o e - weigh o obese, ega dless o he PA g oup. In addi ion, ega dless o he PA g oup, obese indi iduals had he lowes s ess balance and eco e y index. The linea model esul s a e shown in Table 2. Medium (P<0.05)andhigh(P< 0.001) PA g oups, lowe BMI (P< 0.001), and olde age (P< 0.001) we e associa ed wi h lowe s ess pe cen ages du ing he day. Medium (P< 0.05) and high (P< 0.001) PA le el, highe BMI (P< 0.001), and olde age (P< 0.001) we e associa ed wi h lowe s ess pe cen ages du ing wo king hou s. S ess pe - cen age esul s du ing he day and du ing wo king hou s we e simila o men and women. Highe BMI (P<0.001) and olde age (P< 0.001) we e associa ed wi h highe s ess index, bo h in men and in women. In addi ion, medium (P<0.01) and high(P< 0.01) PA we e associa ed wi h lowe s ess index in women. Medium (P< 0.01) and high (P< 0.001) PA, and highe BMI (P< 0.001) we e associa ed wi h lowe s ess bal- ance, bo h in men and in women. Mo eo e , olde age was associa ed wi h lowe s ess balance in men (P< 0.001). Highe BMI and olde age we e associa ed wi h lowe eco e y index, in men and in women (P< 0.001). BMI explained he highes p opo ion o a iance in s ess balance (2.2 % o men and 3.1 % o women) compa ed wi h PA and age. Discussion The pu pose o his s udy was o in es iga e he amoun and in ensi y o objec i e HRV-based s ess and eco e y on wo kdays. The sample g oup comp ised 16 275 Finnish employees, who had pa icipa ed in bea - o-bea R-R in e al eco ding as a pa o li es yle counseling in he cou se o hei e e yday li es be ween 2007 and 2015. Mo e speci ically, he ela ionships be ween PA, BMI, and HRV-based s ess and eco e y we e in es i- ga ed. Fo bo h sexes, a high le el o PA and lowe BMI Föh e al. BMC Public Heal h (2016) 16:701 Page 6 o 13 Table 2 Resul s o he linea models Men Women Pa ame e Es ima e 95 % Cl Lowe Uppe P alue Va iance explained (%) Pa ame e Es ima e 95 % Cl Lowe Uppe P alue Va iance explained (%) S ess (%), 24 h a 2.356 h 1.476 h In e cep 52.3153 50.7730 53.8575 <0.001 50.0658 48.9790 51.1527 <0.001 Age (18 yea s = 0) −0.1263 −0.1600 −0.0926 <0.001 0.780 i −0.0869 −0.1151 −0.0587 <0.001 0.387 i BMI (18.5 kg/m 2 = 0) 0.1541 0.0580 0.2502 0.002 0.144 i 0.0909 0.0264 0.1553 0.006 0.081 i Physical ac i i y le el (inac i e = 0) 1.586 i 1.050 i Low physical ac i i y class 0.3235 −0.6903 1.3372 0.53 −0.8892 −1.5471 −0.2313 0.008 Medium physical ac i i y class −1.1405 −2.2144 −0.0667 0.04 −1.9539 −2.7608 −1.1470 <0.001 High physical ac i i y class −3.9695 −5.0062 −2.9328 <0.001 −4.3772 −5.2620 −3.4923 <0.001 S ess (%), wo king hou s b 4.152 h 1.721 h In e cep 807.3480 777.0476 837.6483 <0.001 677.8091 656.0738 699.5443 <0.001 Age (18 yea s = 0) −4.8729 −5.5355 −4.2103 <0.001 2.942 i −2.66482 −3.22838 −2.10127 <0.001 0.905 i BMI (18.5 kg/m 2 =0) −4.6640 −6.5515 −2.7766 <0.001 0.341 i −4.54067 −5.82984 −3.25149 <0.001 0.504 i Physical ac i i y le el (inac i e = 0) 1.590 i 0.584 i Low physical ac i i y class −5.5094 −25.4261 14.4073 0.59 −3.40975 −16.5664 9.746938 0.61 Medium physical ac i i y class −24.7847 −45.8819 −3.6874 0.021 −17.9339 −34.0695 −1.79829 0.030 High physical ac i i y class −85.2896 −105.6576 −64.9215 <0.001 −61.7878 −79.4833 −44.0923 <0.001 S ess index, 24 h c 27.113 h 27.448 h In e cep 1.2406 1.2403 1.2409 <0.001 1.2414 1.2412 1.2417 <0.001 Age (18 yea s = 0) 0.0001 0.0001 0.0001 <0.001 21.992 i 0.0001 0.0001 0.0001 <0.001 19.655 i BMI (18.5 kg/m 2 = 0) 0.0002 0.0001 0.0002 <0.001 3.819 i 0.0001 0.0001 0.0001 <0.001 3.031 i Physical ac i i y le el (inac i e = 0) 0.115 i 0.145 i Low physical ac i i y class 0.0002 0.0000 0.0004 0.024 −0.0001 −0.0003 0.0000 0.031 Medium physical ac i i y class 0.0000 −0.0002 0.0002 0.91 −0.0002 −0.0004 −0.0001 0.002 High physical ac i i y class 0.0001 −0.0001 0.0003 0.47 −0.0003 −0.0004 −0.0001 0.002 S ess balance, sleep e 3.669 h 3.244 h In e cep 1 0.9975 0.9287 1.0663 <0.001 0.6166 0.5672 0.6661 <0.001 In e cep 2 −0.5292 −0.5503 −0.5082 <0.001 −0.5537 −0.5705 −0.5368 <0.001 Age (18 yea s = 0) −0.0066 −0.0081 −0.0052 <0.001 1.153 i −0.0007 −0.0019 0.0006 0.31 0.006 i BMI (18.5 kg/m 2 =0) −0.0273 −0.0315 −0.0231 <0.001 2.179 i −0.0253 −0.0282 −0.0224 <0.001 3.080 i Physical ac i i y le el (inac i e = 0) 0.445 i 0.233 i Low physical ac i i y class −0.0715 −0.1162 −0.0269 0.002 −0.0126 −0.0423 0.0172 0.41 Föh e al. BMC Public Heal h (2016) 16:701 Page 7 o 13 Table 2 Resul s o he linea models (Con inued) Medium physical ac i i y class −0.0799 −0.1272 −0.0326 <0.001 −0.0572 −0.0937 −0.0207 0.002 High physical ac i i y class −0.1327 −0.1784 −0.0870 <0.001 −0.0822 −0.1222 −0.0421 <0.001 Reco e y index, sleep 9.685 h 3.297 h In e cep 592375.4298 573358.1497 611392.7100 <0.001 457913.8301 444648.7930 471178.8673 <0.001 Age (18 yea s = 0) −5092.7330 −5508.5947 −4676.8712 <0.001 7.753 i −2100.7061 −2444.6430 −1756.7692 <0.001 1.501 i BMI (18.5 kg/m 2 =0) −4825.0722 −6009.6783 −3640.4662 <0.001 0.921 i −4038.5841 −4825.3668 −3251.8014 <0.001 1.065 i Physical ac i i y le el (inac i e = 0) 0.154 i 0.026 i Low physical ac i i y class −19214.5686 −31714.7934 −6714.3437 0.003 1051.0555 −6978.4656 9080.5766 0.80 Medium physical ac i i y class −6153.1124 −19394.2693 7088.0445 0.36 −4452.2347 −14299.7786 5395.3093 0.38 High physical ac i i y class −9653.5430 −22437.0465 3129.9606 0.14 −5732.9247 −16532.4796 5066.6302 0.30 BMI body mass index a Linea eg ession b Box-Cox linea eg ession using ans o ma ion coe icien o 1.62 c Box-Cox linea eg ession using ans o ma ion coe icien o −0.79 e Tobi eg ession Box-Cox linea eg ession using ans o ma ion coe icien o 3.18 h The p opo ion o a iance explained by he whole model i The p opo ion o a iance explained by he p edic o a iable. Calcula ed as he di e ence be ween he p opo ion o a iance explained by he whole model and he p opo ion o a iance explained by a model including all he p edic o a iables, excep o he p edic o in ques ion Föh e al. BMC Public Heal h (2016) 16:701 Page 8 o 13