Role of overweight and obesity in low back disorders among men : a longitudinal study with a life course approach

Role o o e weigh and obesi y in low back diso de s among men: a longi udinal s udy wi h a li e cou se app oach Heikki F ilande , 1,2 S e lana Solo ie a, 1,2 Pe i Mu anen, 3 Ha i Pihlajamäki, 4,5 Ma kku Heliö aa a, 6 Ei a Viika i-Jun u a 2 To ci e: F ilande H, Solo ie a S, Mu anen P, e al. Role o o e weigh and obesi y in low back diso de s among men: a longi udinal s udy wi h a li e cou se app oach. BMJ Open 2015;5:e007805. doi:10.1136/bmjopen-2015- 007805 ▸P epublica ion his o y o his pape is a ailable online. To iew hese iles please isi he jou nal online (h p://dx.doi.o g/10.1136/ bmjopen-2015-007805). Recei ed 28 Janua y 2015 Re ised 2 July 2015 Accep ed 30 July 2015 Fo numbe ed a ilia ions see end o a icle. Co espondence o Heikki F ilande ; heikki. [email p o ec ed] ABSTRACT Objec i es: To assess he associa ion be ween being o e weigh o obese wi h low back pain (LBP) and clinically de ined low back diso de s ac oss he li e cou se. Design: A longi udinal and c oss-sec ional s udy. Se ing: A na ionwide heal h su ey supplemen ed wi h da a om eco ds o p io compulso y mili a y se ice. Pa icipan s: P emili a y heal h eco ds (baseline) we e sea ched o men aged 30–50 yea s (n=1385) who pa icipa ed in a na ional heal h examina ion su ey ( ollow-up). Me hods and ou come measu es: Heigh and weigh we e measu ed a baseline and ollow-up, and wais ci cum e ence a ollow-up. Weigh a he ages o 20, 30, 40 and 50 yea s we e asce ained, when applicable. Repea ed measu es o weigh we e used o calcula e age-s anda dised mean body mass index (BMI) ac oss he li e cou se. The symp om-based ou come measu es a ollow-up included p e alence o non-speci ic and adia ing LBP du ing he p e ious 30 days. The clinically de ined ou come measu es included ch onic low back synd ome and scia ica. Resul s: Baseline BMI (20 yea s) p edic ed adia ing LBP in adul hood, wi h he p e alence a io (PR) being 1.26 (95% CI 1.08 o 1.46) o one SD (3.0 kg/m 2 ) inc ease in BMI. Li e cou se BMI was associa ed wi h adia ing LBP (PR=1.23; 95% CI 1.03 o 1.48 pe 1 uni inc emen in Z sco e, co esponding o 2.9 kg/m 2 ). The de elopmen o obesi y du ing ollow-up inc eased he isk o adia ing LBP (PR=1.91, 95% CI 1.03 o 3.53). Bo h gene al and abdominal obesi y (de ined as wais - o- heigh a io) we e associa ed wi h adia ing LBP (OR=1.64, 95% CI 1.02 o 2.65 and 1.44, 95% CI 1.02 o 2.04). No associa ions we e seen o non-speci ic LBP. Conclusions: Ou indings imply ha being o e weigh o obese in ea ly adul hood as well as du ing he li e cou se inc eases he isk o adia ing bu no non-speci ic LBP among men. Taking in o accoun he cu en global obesi y epidemic, emphasis should be placed on p e en i e measu es s a ing a you h and, also, measu es o p e en ing u he weigh gain du ing he li e cou se should be implemen ed. INTRODUCTION Low back diso de s a e he mos p e alen musculoskele al heal h conce ns in popula ions and may cause a ying deg ees o disabili y. 1 Obesi y is ano he common public heal h p oblem ha con inues o inc ease. 23 The inc ease has been especially p onounced in child en and adolescen s. 45 Two ecen me a-analyses showed ha o e - weigh and obesi y inc ease he isk o bo h low back pain (LBP) and lumba adicula pain. 67 Fo LBP, he associa ions ha e been s onge in women compa ed wi h men, 8–11 howe e , o lumba adicula pain, no gende di e ence has been ound. Nea ly all s udies ha e looked a he e ec s o gene al obesi y defined by body mass index (BMI). The use o BMI has been c i icised o i s inabili y o dis inguish he di e ence be ween a and lean mass, especially in men. 12 13 Abdominal obesi y defined by wais S eng hs and limi a ions o his s udy ▪This s udy used a longi udinal design wi h a andom popula ion sample combining da a om compulso y mili a y se ice wi h a la e heal h examina ion. ▪The low back ou comes included sel - epo ed symp oms as well as clinically de ined diso de s. ▪We used bo h gene al and abdominal obesi y as weigh - ela ed indica o s. ▪The a ailabili y o mul iple measu es o weigh ac oss he li e cou se enabled us o accoun o he empo al luc ua ion o weigh o e ime and allowed explo a ion o he associa ions o ea ly indica o s o obesi y, change in obesi y s a us o e ime, as well as li e cou se BMI, wi h he ou comes. ▪Pa o he weigh - ela ed measu es du ing he li e cou se was sel - epo ed, and he e o e ecall bias may ha e a ec ed he obse ed associa ions. The mili a y eco ds did no p o ide measu es o abdominal adiposi y a baseline, he e o e we we e no able o explo e he associa ions o abdominal obesi y ac oss he li e cou se. F ilande H, e al.BMJ Open 2015;5:e007805. doi:10.1136/bmjopen-2015-007805 1 Open Access Resea ch g oup.bmj.com on June 30, 2016 - Published by h p://bmjopen.bmj.com/Downloaded om ci cum e ence has been associa ed wi h LBP in women, bu no in men. 81415 Mos p e ious s udies on he influence o obesi y on low back symp oms ha e been c oss-sec ional and he majo i y o p ospec i e s udies ha e had a ela i ely sho ollow-up ime. Repea ed measu emen s o weigh - ela ed ac o s ha e a ely been ca ied ou , espe- cially in young popula ions. 816 To he bes o ou knowl- edge, he e a e no s udies e alua ing he li e ime bu den o o e weigh and obesi y on low back ou comes. The aim o his s udy was o assess he associa ion o o e weigh and obesi y wi h LBP and clinically defined low back diso de s ac oss he li e cou se. In a longi udinal app oach, we wan ed o find ou whe he high BMI a he age o 20 yea s o du ing he li e cou se p edic s inc eased isk o low back diso de s la e in li e. Mo eo e , we looked a whe he changes in BMI du ing ollow-up a ec he isk o LBP. Finally, we explo ed he c oss-sec ional associa ions o gene al and abdominal obesi y wi h he ou comes. MATERIALS AND METHODS S udy popula ion The cu en s udy used a longi udinal design wi h wo wa es o da a collec ion (figu e 1). A andom sample o he na ionally ep esen a i e Heal h 2000 S udy (ca ied ou du ing 2000–2001), comp ising o 1866 men aged 30–50 yea s, o med ou s udy base. The main pu poses and me hods o his s udy ha e been desc ibed in de ail elsewhe e. 17 18 Pa icipan s o he fi s wa e o ou da a collec ion (baseline: Musculoskele al Diso de s in Mili a y Se ice S udy, MSDs@Mil S udy) comp ised o he 1713 (91.8%) men eligible o mili a y se ice du ing 1967–1994 whose mili a y eco ds we e ound om he egis e o he Finnish De ence Fo ces. O he andom sample, 1586 (85.0%) men pa icipa ed in he second wa e o da a collec ion ( he Heal h 2000 S udy, ollow-up). A o al o 1536 men (81.4% o he ini ial andom sample) pa icipa ed in bo h wa es and o med he eligible sample o he cu en s udy. Men who did no pa icipa e in he Heal h 2000 S udy we e sligh ly younge , belonged mo e equen ly o he Swedish speaking mino i y and mo e o en had se e e men al p oblems han hose who pa icipa ed. Di e ences in educa ion and socioeconomic s a us we e mino . 17 Fu he mo e, he e was no di e ence in base- line BMI be ween pa icipan s and non-pa icipan s o he Heal h 2000 S udy. We excluded pa icipan s wi h missing da a on LBP a ollow-up (n=151, 9.8% o he eligible sample) esul ing in a o al o 1385 (74.2% o he ini ial andom sample) pa icipan s a ailable o he analyses. Low back ou comes a ollow-up All he men unde wen a symp om in e iew and a com- p ehensi e heal h examina ion (including a physical examina ion pe o med by a specially ained physician) o a condensed heal h examina ion a home. The symp om-based ou come measu es included p e alence o LBP (non-specific LBP) and adia ing LBP du ing he p e ious 30 days. The in o ma ion on LBP was collec ed by he ollowing ques ions: ‘Ha e you e e had low back pain? (yes/no)’. Those who answe ed yes we e asked abou ecen pain and abou he ype o pain: ‘Ha e you had low back pain du ing he p eceding 30 days? (yes/ no)’;‘I you had low back pain, did i adia e? (0=no, 1=below knee, 2=abo e knee)’. Radia ing LBP was defined as pain ha adia ed ei he below o abo e he knee. The clinically defined ou come measu es included ch onic low back synd ome and scia ica diagnosed wi h p edefined c i e ia by specially ained physicians. 19 20 The diagnoses made by he physicians we e based on p e ious and cu en symp oms, documen a ion on p e- ious low back diagnoses and clinical findings in he physical examina ion. Symp oms wi h du a ion o a leas 3 mon hs o e all we e conside ed ch onic. All diagnoses we e ca ego ised as 0=no, 1=p obable, 2=defini e. The diagnoses we e classified as p obable when he pa ici- pan did no cu en ly ha e symp oms, o when he c i- e ia o he p e ious diagnoses we e ague. In he cu en s udy, a clinically defined ou come was p esen i he pa icipan had a p obable o defini e diagnosis. Spinal p oblems a baseline Low back symp oms and signs as well as spinal symp oms and signs (symp oms o signs in he ho acic o lumba spine) we e eco ded a baseline by he examining phys- ician. We used dicho omised a iables (yes/no). Weigh - ela ed ac o s a baseline and du ing ollow-up A baseline, as pa o he ec ui heal h check-up ca ied ou du ing he fi s 2 weeks o mili a y se ice, nu ses o ained medics a he ga ison heal h clinics measu ed he ec ui s’heigh and weigh in ligh clo h- ing. Fo 114 men (8.2%), da a on weigh - ela ed ac o s we e no ound om ec ui heal h examina ion eco ds, and we used he eco ds o he p emili a y se ice examina ion pe o med app oxima ely 6 mon hs p io o mili a y se ice. A ollow-up, in addi ion o measu ed du ing heal h examina ion heigh and weigh , men we e asked o ecall hei heigh a he age o 20 yea s and weigh a he age o 20, 30, 40 and 50 yea s. BMI was calcula ed by di iding weigh (kg) wi h he squa e o heigh (m 2 ). A ollow-up, gene al o e weigh and obesi y we e defined based on BMI using he WHO ecommenda ion o BMI <25 kg/m 2 (no mal), 25–29.9 kg/m 2 (o e weigh ) and ≥30 kg/m 2 (obese). A baseline, BMI was dicho omised in o BMI <25 kg/m 2 (no mal) and ≥25 kg/m 2 (o e weigh /obese). Wais ci cum e ence was measu ed a ollow-up. Abdominal obesi y was defined based on wais ci cum e - ence: <94 cm (no mal), 94–101.9 cm (o e weigh ) and ≥102 cm (obese). Wais - o-heigh a io was dicho omised in o ≤0.5 and >0.5. 2F ilande H, e al.BMJ Open 2015;5:e007805. doi:10.1136/bmjopen-2015-007805 Open Access g oup.bmj.com on June 30, 2016 - Published by h p://bmjopen.bmj.com/Downloaded om Accu acy o sel - epo ed heigh and weigh We had measu ed da a (a baseline) and sel - epo ed da a ( ecalled a ollow-up) o heigh and weigh a he age o 20 yea s o all he men. The co ela ion (Pea son co ela ion coe ficien ) was e y high o heigh ( =0.94, p<0.0001) and mode a e o weigh ( =0.77, p<0.0005). We also examined he magni ude o ecall bias o heigh a he age o 20 yea s by he ime o ecall and ound ha he olde men ecalled hei heigh simila ly o he younge men. Fo men aged 30 (n=124), 40 (n=194) and 50 (n=121) yea s a he ime o Heal h 2000 S udy da a collec ion, we also had bo h sel - epo ed as well as measu ed heigh and weigh . All sel - epo ed and measu ed an h opome ic measu es we e highly co ela ed. Fo men aged 30 yea s, he co ela ion coe ficien s o heigh and weigh we e =0.98 and =0.94, espec i ely. Fo men aged 40 yea s, he co esponding numbe s we e =0.97 and =0.95. Fo men aged 50 yea s, he co esponding numbe s we e =0.95 and =0.98. These da a sugges ha ecall bias is ela i ely mino and sel - epo ed da a could be eliably used as a p oxy o pa icipan s’heigh and weigh . Fu he mo e, he co ela ion coe ficien s (Pea son ) be ween BMI based on measu ed and sel - epo ed weigh and heigh we e 0.80, 0.85, 0.96 and 0.98, o he s udied age g oups, espec i ely. Weigh - ela ed measu es du ing he li e cou se Baseline and ollow-up in o ma ion o o e weigh o obesi y based on measu ed BMI was used o cons uc a a iable o desc ibe o e weigh o obesi y du ing he li e cou se, consis ing o six ca ego ies. The ca ego ies we e defined as he ollowing: 1=bo h baseline and ollow-up BMI <25 kg/m 2 (no mal weigh a bo h ime poin s), Figu e 1 Flow cha o he o ma ion o he p esen s udy sample (MSDs@Mil S udy, Musculoskele al Diso de s in Mili a y Se ice S udy). F ilande H, e al.BMJ Open 2015;5:e007805. doi:10.1136/bmjopen-2015-007805 3 Open Access g oup.bmj.com on June 30, 2016 - Published by h p://bmjopen.bmj.com/Downloaded om 2=baseline BMI <25 kg/m 2 and ollow-up BMI anged be ween 25 and 29.9 kg/m 2 (became o e weigh du ing ollow-up), 3=baseline BMI <25 kg/m 2 and ollow-up BMI ≥30 kg/m 2 (became obese du ing ollow-up), 4=baseline BMI ≥25 kg/m 2 and ollow-up BMI <25 kg/m 2 (o e - weigh /obese a baseline, no mal weigh a ollow-up), 5=baseline BMI ≥25 kg/m 2 and ollow-up BMI anged be ween 25 and 29.9 kg/m 2 (o e weigh /obese a base- line and o e weigh a ollow-up), 6=baseline BMI ≥25 kg/m 2 and ollow-up BMI ≥30 kg/m 2 (o e weigh / obese a baseline and obese a ollow-up). A li e cou se measu e o BMI was based on epea ed BMI alues (measu ed a baseline, sel - epo ed a he age o 30, 40 and 50 yea s and measu ed a ollow-up). I was cons uc ed by calcula ing age-s anda dised Z sco es (mean=0, SD=1) o each ime poin measu e, which we e a e aged ac oss ime poin s. Po en ial con ounde s Age a baseline was calcula ed o he day o he mea- su emen s and a ollow-up o 1 June 2000. Leng h o educa ion (yea s) was inqui ed a ollow-up. Da a on smoking and physical ac i i y we e collec ed a baseline and ollow-up. A baseline, he consc ip s we e classified as smoke s i hey answe ed ‘yes’ o he ques ion ‘Do you smoke?’. A ollow-up, daily smoke s we e classified as smoke s. A baseline, he consc ip s we e classified as physically ac i e i hey we e pa icipa ing in spo s ac i - i ies a leas once a week. A ollow-up, pa icipan s who epo ed ha hey engaged in physical exe cise ha causes swea ing o a leas 30 min a leas once a week we e classified as physically ac i e. S a is ical analyses S a is ical analyses we e pe o med using SAS V.9.4. O 681 (28.7%) missing alues on smoking a baseline, 520 alues we e impu ed using he missing da a p opensi y sco e me hod. 21 Howe e , 161 (6.8%) alues emained missing. Since ou ou comes we e based on p e alence, we used p e alence a ios (PRs) o measu e he longi u- dinal and odds a ios (ORs) o measu e he c oss- sec ional associa ions be ween weigh - ela ed a iables and low back ou comes. The e ec s o baseline and li e cou se BMI on low back ou comes a ollow-up we e s udied wi h a Cox eg ession model, wi h he o al leng h o ime o ollow-up being assigned o each indi idual. The Cox model was used ins ead o logis ic eg ession in o de o a oid undue infla ion o es ima es o associa ions be ween weigh - ela ed measu es and common low back p oblems. 22 23 The PRs and hei 95% CIs we e adjus ed o age and smoking s a us a baseline and in o ma ion o educa ion (yea s) ob ained a ollow-up. The c oss-sec ional associa ions o weigh - ela ed mea- su es wi h low back ou comes a ollow-up we e s udied wi h bina y logis ic eg ession models. The pa icipan s’ age, educa ion (yea s) and smoking s a us a ollow-up we e included in he mul i a iable models as co a ia es. We did no adjus o physical ac i i y because i was no associa ed wi h low back ou comes in ou s udy popula ion. Fu he mo e, physical ac i i y has been shown o ha e a U-shaped associa ion wi h LBP, and was he e o e consid- e ed o be an e ec modifie and no a con ounde . 924 Sensi i i y analyses we e ca ied ou excluding pa ici- pan s wi h spinal p oblems a baseline. RESULTS Cha ac e is ics o he s udy pa icipan s and p e alence o ou comes The cha ac e is ics o he s udy pa icipan s a baseline and ollow-up a e shown in able 1. The a e age ollow-up ime was 20.5 yea s (SD±6.0 yea s), anging om 6 o 33 yea s. Du ing he ollow-up, he mean BMI inc eased by 4.1– 4.4 kg/m 2 (18–20%), co esponding o 11–12 kg. The inc ease was linea ly associa ed wi h he leng h o he ollow-up ime. Bo h a baseline and ollow-up, abou one- hi d o he pa icipan s we e egula smoke s. In gene al, he p opo ion o pa icipan s wi h egula physical ac i i y was highe a ollow-up han a baseline. The p e alence o spinal symp oms o signs was abou 15% a baseline. Non-specific LBP occu ed commonly ( able 1), whe eas adia ing ype o LBP was less p e alen . The p e alence o adia ing LBP and clinically defined ou - comes inc eased wi h age (p=0.08 o adia ing LBP, p=0.002 o ch onic low back synd ome and p=0.01 o scia ica). Associa ions o BMI wi h low back ou comes du ing he li e cou se BMI a baseline p edic ed he 1 mon h p e alence o adia ing LBP assessed a ollow-up ( able 2). A one SD inc ease in baseline BMI, co esponding o 3 kg/m 2 ,was associa ed wi h a 26% inc ease in he isk o adia ing LBP. A simila end was seen in he isk o ch onic low back synd ome and scia ica assessed by a physician, al hough i did no each s a is ical significance ( able 2). Sensi i i y analyses, excluding pa icipan s wi h baseline spinal symp oms o signs, o hose wi h low back symp- oms o signs, showed no change in he obse ed associa ions. Li e cou se BMI (BMI Z sco e) was s a is ically signifi- can ly associa ed wi h 1 mon h adia ing LBP. A 1-uni inc ease in Z sco e (co esponding o 2.92 uni s o BMI) was associa ed wi h a 23% inc ease in he isk o adia - ing LBP ( able 2). Li e cou se BMI was no associa ed wi h he o he ou comes. Those wi h no mal weigh a baseline who subs an ially gained weigh du ing he ollow-up had an inc eased isk o 1 mon h adia ing LBP pain assessed a ollow-up ( able 3). Also, hose who we e cons an ly o e weigh had a wo old inc eased isk o adia ing LBP. 4F ilande H, e al.BMJ Open 2015;5:e007805. doi:10.1136/bmjopen-2015-007805 Open Access g oup.bmj.com on June 30, 2016 - Published by h p://bmjopen.bmj.com/Downloaded om C oss-sec ional associa ions o gene al and abdominal obesi y wi h low back ou comes In c oss-sec ional analyses a ollow-up, BMI ended o be associa ed wi h adia ing LBP, bu no associa ions we e seen o he o he ou comes. Wais ci cum e ence ended o be associa ed wi h adia ing LBP as well as wi h ch onic low back synd ome and scia ica ( able 2). O e weigh and obesi y defined by BMI showed a dose– esponse ela ionship wi h adia ing LBP, wi h a 42% and 64% inc ease in he isk in o e weigh and obese men, espec i ely, compa ed o men wi h no mal weigh . Fu he mo e, abdominal obesi y defined by wais - o-heigh a io was associa ed wi h an inc eased isk o adia ing LBP (OR=1.44, 95% CI 1.02 o 2.04). No associa ions we e ound o non-specific LBP, ch onic low back synd ome o scia ica assessed by a physician ( able 2). The exclusion o pa icipan s wi h baseline spinal symp oms o signs s eng hened he obse ed associa ions be ween wais ci cum e ence and adia ing LBP (OR=1.62, 95% CI 1.05 o 1.78 o abdominal obesi y defined as wais ci cum e ence ≥102 cm). DISCUSSION In ou longi udinal popula ion-based s udy ac oss he li e cou se, we ound ha BMI a he age o 20 yea s p e- dic ed adia ing LBP la e in li e. The de elopmen o obesi y du ing ollow-up inc eased adia ing LBP. Mo eo e , BMI du ing he li e cou se was associa ed wi h adia ing LBP la e in li e. Gene al and abdominal obesi y we e bo h associa ed wi h adia ing LBP wi h mino di e ences in he obse ed es ima es. No associa ions we e seen o non-specific LBP. Weake and s a is ically non-significan associa ions we e seen o BMI (longi udinally) and wais ci cum e ence (c oss- sec ionally) wi h he clinically defined ou comes. The p esen s udy has se e al s eng hs. Fi s , we used a unique s udy popula ion based on he Heal h 2000 S udy and supplemen ed wi h mili a y se ice eco ds o all men aged 30–50 yea s. The Heal h 2000 S udy well ep esen s he Finnish adul popula ion because o he andom sample and high pa icipa ion a e o up o 85%. Mili a y medical eco ds we e ob ained o almos all he men in he sample. Second, gene al and abdom- inal obesi y we e defined based on measu ed weigh - ela ed indica o s. In addi ion, we had sel - epo ed in o ma ion on weigh and heigh o each age decennium. Thi d, he ichness o weigh - ela ed mea- su es allowed us o explo e he associa ions o ea ly indi- ca o s o obesi y, change in obesi y s a us o e ime, as well as li e cou se BMI, wi h he ou comes. Fou h, ou se o low back ou comes included symp oms as well as clinically defined diso de s. Howe e , he findings o he cu en s udy mus be in e p e ed aking in o conside a ion a ew limi a ions. Since only men we e included, he popula ion being highly ep esen a i e allows gene alising ou findings o all men. The heal h examina ion a baseline did no p o ide measu es o abdominal adiposi y, he e o e we we e no able o look a he associa ions o abdominal obesi y ac oss he li e cou se. Owing o he small numbe o o e weigh o obese indi iduals a baseline, we we e no able o de ec whe he indi iduals losing weigh du ing he ollow-up a e less likely o epo back Table 1 Cha ac e is ics o he s udy popula ion a baseline (Mili a y Heal h Examina ion), 1967–1994 and ollow-up (Heal h 2000 Su ey), 2000–2001 N Mean (95% CI) P e alence (%, 95% CI) Age (yea s) A baseline 1385 19.7 (19.6 o 19.8) A ollow-up 1385 40.2 (39.9 o 40.5) BMI (kg/m 2 ) A baseline 1384 22.3 (22.1 o 22.4) A ollow-up 1385 26.5 (26.3 o 26.7) Regula smoke s A baseline 303/880 34.4 (31.3 o 37.7) A ollow-up 454/1380 32.9 (30.4 o 35.5) Physically ac i e a leisu e ime A baseline 447/846 52.8 (49.3 o 56.2) A ollow-up 1008/1370 73.6 (71.2 o 75.9) Spinal symp oms and/o signs A baseline 208/1385 15.0 (13.2 o 17.0) Low back symp oms A baseline 113/1385 8.2 (6.8 o 9.7) Wais ci cum e ence A ollow-up 1380 95.3 (94.8 o 95.9) Hip ci cum e ence A ollow-up 1380 99.1 (98.7 o 99.5) Educa ion (yea s) A ollow-up 1366 12.8 (12.6 o 13.0) LBP, p eceding 30 days A ollow-up Non-speci ic 393/1385 28.4 (26.0 o 30.8) Radia ing 162/1385 11.7 (10.1 o 13.5) Physician-diagnosed diso de s* A ollow-up Ch onic low back synd ome 100/1350 7.4 (6.1 o 8.9) Scia ica 58/1350 4.3 (3.3 o 5.5) *In o ma ion is missing o 35 men. BMI, body mass index; LBP, low back pain. F ilande H, e al.BMJ Open 2015;5:e007805. doi:10.1136/bmjopen-2015-007805 5 Open Access g oup.bmj.com on June 30, 2016 - Published by h p://bmjopen.bmj.com/Downloaded om Table 2 Associa ions o weigh - ela ed measu es wi h LBP and physician-diagnosed low back diso de s LBP, p eceding 30 days (n=1363) Ch onic LBP synd ome (n=1350) Radia ing LBP, p eceding 30 days (n=1363) Scia ica, assessed by physician (n=1350) PR/OR 95% CI* PR/OR 95% CI* PR/OR 95% CI* PR/OR 95% CI* Longi udinal associa ions be ween BMI a baseline and ou comes a ollow-up BMI a baseline†1.07 0.96 o 1.19 1.18 0.96 o 1.44 1.26 1.08 o 1.46 1.15 0.89 o 1.50 Li e cou se BMI (Z sco e)‡0.99 0.86 o 1.14 1.09 0.87 o 1.36 1.23 1.03 o 1.48 1.10 0.83 o 1.46 C oss-sec ional associa ions be ween weigh - ela ed ac o s, and ou comes a ollow-up BMI (con inuous)§ 0.95 0.84 o 1.07 1.09 0.90 o 1.33 1.16 0.99 o 1.35 1.10 0.86 o 1.41 BMI (ca ego ical), kg/m 2 <25 1.00 1.00 1.00 1.00 25–29.9 1.03 0.80 o 1.34 0.95 0.60 o 1.51 1.42 0.97 o 2.08 0.84 0.47 o 1.51 ≥30 0.88 0.61 o 1.25 1.13 0.63 o 2.00 1.64 1.02 o 2.65 1.04 0.50 o 2.15 Wais ci cum e ence (con inuous)¶ 1.00 0.88 o 1.12 1.13 0.92 o 1.38 1.15 0.98 o 1.35 1.18 0.92 o 1.52 Wais ci cum e ence (ca ego ical), cm <94 1.00 1.00 1.00 1.00 94–101.9 1.11 0.83 o 1.47 1.04 0.63 o 1.73 1.03 0.69 o 1.53 0.91 0.47 o 1.75 ≥102 1.03 0.77 o 1.39 1.24 0.75 o 2.03 1.31 0.88 o 1.96 1.19 0.64 o 2.23 Wais - o-heigh a io No mal (≤0.5) 1.00 1.00 1.00 1.00 C i ical (>0.5) 0.96 0.76 o 1.22 1.33 0.75 o 1.72 1.44 1.02 o 2.04 0.88 0.52 o 1.49 Signi ican associa ions ha e been ma ked wi h bold. *PRs ( o he longi udinal analyses) and ORs ( o c oss-sec ional analyses) and hei 95% CIs a e adjus ed o age, smoking and educa ion. †Con inuous a iable, one SD inc ease, co esponding o 3.0 kg/m 2 o baseline BMI. ‡One uni inc emen in Z sco e, co esponding o 2.92 uni s o BMI du ing ollow-up. §Con inuous a iable, one SD inc ease, co esponding o 4.0 kg/m 2 o ollow-up BMI. ¶Con inuous a iable, one SD inc ease, co esponding o 11 cm o wais ci cum e ence a ollow-up. BMI, body mass index; LBP, low back pain; PR, p e alence a io. 6F ilande H, e al.BMJ Open 2015;5:e007805. doi:10.1136/bmjopen-2015-007805 Open Access g oup.bmj.com on June 30, 2016 - Published by h p://bmjopen.bmj.com/Downloaded om symp oms a he end o ollow-up. Pa o he weigh - ela ed measu es du ing he li e cou se was sel - epo ed, and he e o e ecall bias may ha e a ec ed he obse ed associa ions. Howe e , ou assessmen s con- fi med he high accu acy o he sel - epo ed alues used in his s udy. Ou esul s a e in ag eemen wi h a ecen me a-analysis in which o e weigh and obesi y we e bo h consis en ly, hough modes ly, associa ed wi h he isk o lumba adicula pain. 7 Mo eo e , s udies on associa- ions be ween weigh - ela ed ac o s and non-specific LBP ha e epo ed inconsis en esul s, especially in men. 6925 In ag eemen wi h p e ious li e a u e, we ound weake associa ions o he clinically defined ou comes. 72627 Mo eo e , a ecen la ge c oss-sec ional s udy among consc ip s (male and emale) aged 17 yea s showed weake associa ions o obesi y wi h clinically defined ou - comes han wi h symp oms among he men. 28 To he bes o ou knowledge, his is he fi s s udy o epo associa ions o low back diso de s wi h composi e li e cou se in o ma ion o BMI based on epea ed mea- su es wi h cons an in e als. We ound ha BMI bo h a he age o 20 yea s and du ing he li e cou se p edic ed adia ing LBP la e in li e. A B i ish bi h coho s udy ound ha high BMI a he age o 23 yea s p edic ed incidence o LBP a age 32–33 yea s in bo h gende s. 16 Ou findings, oge he wi h hose o he a o emen ioned s udy, sugges ha obesi y a ound he age o 20 yea s— and especially when pe sis en —is a causal isk ac o o adia ing LBP. The e a e wo dis inc ypes o LBP— adia ing and non-specific— ha ha e a di e en cou se du ing li e and may ha e di e en ae iologies. 6 In line wi h ou findings, BMI and wais ci cum e ence p edic ed inci- dence o adia ing bu no non-specific LBP in a popula ion-based s udy among young adul s. 24 We measu ed bo h gene al obesi y and abdominal adi- posi y. This is impo an , since BMI migh no be an app op ia e measu e o cap u e adiposi y in men. 13 In ag eemen wi h ou findings, a Du ch s udy, using a andom sample om he gene al popula ion, ound bo h gene al and abdominal obesi y o be associa ed wi h adia ing bu no non-specific LBP in men. 14 Simila ly, a ecen s udy in a ep esen a i e young adul popula ion showed bo h ypes o adiposi y being asso- cia ed wi h he incidence o adia ing LBP. 24 The e a e se e al possible explana ions o he associ- a ion be ween excess weigh and adia ing LBP. Obesi y could inc ease he mechanical load on he spine by causing a highe comp ession o ea on he lumba spine s uc u es. Obese people may also be mo e p one o inju ies. 29 In addi ion, obesi y is an impo an isk ac o o a he oscle osis, which has also been linked especially wi h scia ic pain. 30 Since abdominal obesi y showed an associa ion wi h adia ing LBP, sys emic inflamma ion and pa omechanical pa hways in ol ed in he me abolic synd ome may play a ole, as discussed in he li e a u e. 31 A ecen sys ema ic e iew o win s udies concluded ha indi iduals being o e weigh o obese a e mo e likely o ha e LBP and lumba disc gen- e a ion, bu he associa ions we e weake a e con ol- ling o amilial ac o s, sugges ing ha obesi y and LBP sha e common gene ic isk ac o s. 32 In conclusion, ou findings indica e ha being o e - weigh o obese in young adul hood as well as du ing he li e cou se inc eases he isk o adia ing bu no non- specific LBP in men. Taking in o accoun he cu en global obesi y epidemic, emphasis should be placed on p e en i e measu es s a ing a you h and, also, p e en - i e measu es o u he weigh gain du ing he li e cou se should be implemen ed. S udies on he ae iology o LBP would benefi om di e en ia ing be ween adia - ing and non-specific ypes o pain. Fu u e esea ch should p o ide li e cou se s udies on he associa ions o weigh - ela ed ac o s wi h low back diso de s among women. Au ho a ilia ions 1 Cen e o Expe ise o Heal h and Wo k Abili y, Finnish Ins i u e o Occupa ional Heal h, Helsinki, Finland 2 Disabili y P e en ion Resea ch Cen e, Finnish Ins i u e o Occupa ional Heal h, Helsinki, Finland 3 Depa men o S a is ics and Heal h Economics, Finnish Ins i u e o Occupa ional Heal h, Helsinki, Finland Table 3 Associa ions o baseline and ollow-up o e weigh and obesi y wi h LBP BMI a baseline, kg/m 2 BMI a ollow-up, kg/m 2 LBP, p eceding 30 days (n=1363) Radia ing LBP, p eceding 30 days (n=1363) n* PR†95% CI n* PR†95% CI <25 <25 104 (509) 1.00 48 (509) 1.00 <25 25–29.9 161 (560) 1.00 0.76 o 1.32 70 (560) 1.34 0.89 o 2.00 <25 ≥30 33 (106) 1.18 0.74 o 1.88 17 (106) 1.91 1.03 o 3.53 ≥25 <25 2 (8) 0.84 0.17 o 4.28 0 (8) –– ≥25 25–29.9 21 (62) 1.39 0.79 o 2.50 10 (62) 2.10 1.00 o 4.47 ≥25 ≥30 26 (118) 0.76 0.47 o 1.23 18 (118) 1.65 0.89 o 3.06 Signi ican associa ions ha e been ma ked wi h bold. *Numbe wi h ou come (numbe o exposed). †PRs and hei 95% CIs a e adjus ed o age, smoking and educa ion. BMI, body mass index; LBP, low back pain; PR, p e alence a io. F ilande H, e al.BMJ Open 2015;5:e007805. doi:10.1136/bmjopen-2015-007805 7 Open Access g oup.bmj.com on June 30, 2016 - Published by h p://bmjopen.bmj.com/Downloaded om 4 Depa men o O hopaedics and T auma Su ge y, Seinäjoki Cen al Hospi al, Seinäjoki, Finland 5 Uni e si y o Tampe e, Seinäjoki, Finland 6 Depa men o Heal h, Func ional Capaci y and Wel a e, Na ional Ins i u e o Heal h and Wel a e, Helsinki, Finland Con ibu o s EV-J, SS and HF concei ed and designed he s udy. MH and HP helped acqui e he da a. SS and PM analysed he da a. HF w o e he i s d a o he manusc ip . All he au ho s con ibu ed o he in e p e a ion o he da a, ca e ully e iewed he d a and e ised i c i ically. All he au ho s ga e inal app o al o he manusc ip . Funding This wo k was suppo ed by he Academy o Finland (129475) and he Cen e o Mili a y Medicine, Helsinki, Finland. Compe ing in e es s None decla ed. E hics app o al The Sec ion o Epidemiology and Public Heal h o he E hics Commi ee o he Hospi al Dis ic o Helsinki and Uusimaa app o ed he Heal h 2000 S udy and The E hics Commi ee o he Finnish Ins i u e o Occupa ional Heal h u he app o ed he MSDs@Mil s udy. P o enance and pee e iew No commissioned; ex e nally pee e iewed. Da a sha ing s a emen No addi ional da a a e a ailable. Open Access This is an Open Access a icle dis ibu ed in acco dance wi h he C ea i e Commons A ibu ion Non Comme cial (CC BY-NC 4.0) license, which pe mi s o he s o dis ibu e, emix, adap , build upon his wo k non- comme cially, and license hei de i a i e wo ks on di e en e ms, p o ided he o iginal wo k is p ope ly ci ed and he use is non-comme cial. See: h p:// c ea i ecommons.o g/licenses/by-nc/4.0/ REFERENCES 1. Balague F, Mannion AF, Pellise F, e al. Non-speci ic low back pain. Lance 2012;379:482–91. 2. Kelly T, Yang W, Chen CS, e al. Global bu den o obesi y in 2005 and p ojec ions o 2030. In J Obes (Lond) 2008;32:1431–7. 3. Finucane MM, S e ens GA, Cowan MJ, e al. Na ional, egional, and global ends in body-mass index since 1980: sys ema ic analysis o heal h examina ion su eys and epidemiological s udies wi h 960 coun y-yea s and 9.1 million pa icipan s. Lance 2011;377:557–67. 4. Kau iainen S, Koi is o AM, Koi usil a L, e al. Sociodemog aphic ac o s and a secula end o adolescen o e weigh in Finland. In J Pedia Obes 2009;4:360–70. 5. Wang Y, Lim H. The global childhood obesi y epidemic and he associa ion be ween socio-economic s a us and childhood obesi y. In Re Psychia y 2012;24:176–88. 6. Shi i R, Ka ppinen J, Leino-A jas P, e al. The associa ion be ween obesi y and low back pain: a me a-analysis. Am J Epidemiol 2010;171:135–54. 7. Shi i R, Lallukka T, Ka ppinen J, e al. Obesi y as a isk ac o o scia ica: a me a-analysis. Am J Epidemiol 2014;179:929–37. 8. Shi i R, Solo ie a S, Husga el-Pu siainen K, e al. The associa ion be ween obesi y and he p e alence o low back pain in young adul s: he Ca dio ascula Risk in Young Finns S udy. Am J Epidemiol 2008;167:1110–19. 9. Nilsen TI, Hol e mann A, Mo k PJ. Physical exe cise, body mass index, and isk o ch onic pain in he low back and neck/shoulde s: longi udinal da a om he No d-T ondelag Heal h S udy. Am J Epidemiol 2011;174:267–73. 10. Heuch I, Hagen K, Zwa JA. Body mass index as a isk ac o o de eloping ch onic low back pain: a ollow-up in he No d-T ondelag Heal h S udy. Spine 2013;38:133–9. 11. Mikkonen PH, Lai inen J, Remes J, e al. Associa ion be ween o e weigh and low back pain: a popula ion-based p ospec i e coho s udy o adolescen s. Spine 2013;38:1026–33. 12. Snijde MB, an Dam RM, Visse M, e al. Wha aspec s o body a a e pa icula ly haza dous and how do we measu e hem? In J Epidemiol 2006;35:83–92. 13. Ro hman KJ. BMI- ela ed e o s in he measu emen o obesi y. In J Obes (Lond) 2008;32(Suppl 3):S56–9. 14. Han TS, Schou en JS, Lean ME, e al. The p e alence o low back pain and associa ions wi h body a ness, a dis ibu ion and heigh . In J Obes Rela Me ab Diso d 1997;21:600–7. 15. Lean ME, Han TS, Seidell JC. Impai men o heal h and quali y o li e in people wi h la ge wais ci cum e ence. Lance 1998;351:853–6. 16. Powe C, F ank J, He zman C, e al. P edic o s o low back pain onse in a p ospec i e B i ish s udy. Am J Public Heal h 2001;91:1671–8. 17. Heis a o S, ed. Me hodology Repo Heal h 2000 Su ey. Publica ions o he Na ional Public Heal h Ins i u e. Helsinki: KTL-Na ional Public Heal h Ins i u e, Depa men o Heal h and Func ional Capaci y, 2008:1–248. h p://u n. i/URN:NBN: i- e201204193320 (accessed 2 Jul 2015). 18. A omaa A, Koskinen S, eds. Heal h and unc ional capaci y in Finland—baseline esul s o he Heal h 2000 Heal h Examina ion Su ey. Publica ions o he Na ional Public Heal h Ins i u e. Helsinki: KTL-Na ional Public Heal h Ins i u e, Finland, Depa men o Heal h and Func ional Capaci y, 2004:1–175. h p://u n. i/URN:NBN: i- e201204193452 (accessed 2 Jul 2015). 19. Kaila-Kangas L, Leino-A jas P, Ka ppinen J, e al. His o y o physical wo k exposu es and clinically diagnosed scia ica among wo king and nonwo king Finns aged 30 o 64. Spine 2009;34:964–9. 20. Helio aa a M, Impi aa a O, Sie e s K, e al. Lumba disc synd ome in Finland. J Epidemiol Communi y Heal h 1987;41:251–8. 21. Rosenbaum PR, Rubin DB. The cen al ole o he p opensi y sco e in obse a ional s udies o causal e ec s. Biome ika 1983;70:41–55. 22. Ba os AJ, Hi aka a VN. Al e na i es o logis ic eg ession in c oss-sec ional s udies: an empi ical compa ison o models ha di ec ly es ima e he p e alence a io. BMC Med Res Me hodol 2003;3:21. 23. Thompson ML, Mye s JE, K iebel D. P e alence odds a io o p e alence a io in he analysis o c oss sec ional da a: wha is o be done? Occup En i on Med 1998;55:272–7. 24. Shi i R, Solo ie a S, Husga el-Pu siainen K, e al. The ole o obesi y and physical ac i i y in non-speci ic and adia ing low back pain: he Young Finns s udy. Semin A h i is Rheum 2013;42:640–50. 25. Kää ia S, Leino-A jas P, Rahkonen O, e al. Risk ac o s o scia ic pain: a p ospec i e s udy among middle-aged employees. Eu J Pain 2011;15:584–90. 26. Heliö aa a M, Mäkela M, Knek P, e al. De e minan s o scia ica and low-back pain. Spine 1991;16:608–14. 27. Jhawa BS, Fuchs CS, Coldi z GA, e al. Ca dio ascula isk ac o s o physician-diagnosed lumba disc he nia ion. Spine J 2006;6:684–91. 28. He shko ich O, F iedlande A, Go don B, e al. Associa ions o body mass index and body heigh wi h low back pain in 829,791 adolescen s. Am J Epidemiol 2013;178:603–9. 29. Hu HY, Chou YJ, Chou P, e al. Associa ion be ween obesi y and inju y among Taiwanese adul s. In J Obes 2009;33:878–84. 30. Shi i R, Viika i-Jun u a E, Leino-A jas P, e al. The associa ion be ween ca o id in ima-media hickness and scia ica. Semin A h i is Rheum 2007;37:174–81. 31. Sama zis D, Ka ppinen J, Cheung JP, e al. Disk degene a ion and low back pain: a e hey a - ela ed condi ions? Global Spine J 2013;3:133–44. 32. Da io AB, Fe ei a ML, Re shauge KM, e al. The ela ionship be ween obesi y, low back pain, and lumba disc degene a ion when gene ics and he en i onmen a e conside ed: a sys ema ic e iew o win s udies. Spine J 2015;15:1106–17. 8F ilande H, e al.BMJ Open 2015;5:e007805. doi:10.1136/bmjopen-2015-007805 Open Access g oup.bmj.com on June 30, 2016 - Published by h p://bmjopen.bmj.com/Downloaded om wi h a li e cou se app oach diso de s among men: a longi udinal s udy Role o o e weigh and obesi y in low back Ma kku Heliö aa a and Ei a Viika i-Jun u a Heikki F ilande , S e lana Solo ie a, Pe i Mu anen, Ha i Pihlajamäki, doi: 10.1136/bmjopen-2015-007805 2015 5: BMJ Open h p://bmjopen.bmj.com/con en /5/8/e007805 Upda ed in o ma ion and se ices can be ound a : These include: Re e ences #BIBLh p://bmjopen.bmj.com/con en /5/8/e007805 This a icle ci es 30 a icles, 9 o which you can access o ee a : Open Access h p://c ea i ecommons.o g/licenses/by-nc/4.0/non-comme cial. See: p o ided he o iginal wo k is p ope ly ci ed and he use is non-comme cially, and license hei de i a i e wo ks on di e en e ms, pe mi s o he s o dis ibu e, emix, adap , build upon his wo k Commons A ibu ion Non Comme cial (CC BY-NC 4.0) license, which This is an Open Access a icle dis ibu ed in acco dance wi h he C ea i e se ice Email ale ing box a he op igh co ne o he online a icle. Recei e ee email ale s when new a icles ci e his a icle. Sign up in he Collec ions Topic A icles on simila opics can be ound in he ollowing collec ions (1600)Public heal h (1565)Epidemiology No es h p://g oup.bmj.com/g oup/ igh s-licensing/pe missions To eques pe missions go o: h p://jou nals.bmj.com/cgi/ ep in o m To o de ep in s go o: h p://g oup.bmj.com/subsc ibe/ To subsc ibe o BMJ go o: g oup.bmj.com on June 30, 2016 - Published by h p://bmjopen.bmj.com/Downloaded om