Smoking., heavy physical work and low back pain: A four-year prospective study

Transcription

1 Occup. Med. Vol. 49, No. 3, pp , 1999 Copyright O 1999 Upplncott Williams & WIMns for SOM Prirrted in Great Britain. All rights reserved /99 Smoking., heavy physical work and low back pain: A four-year prospective study W. Eriksen, B. Natvig and D. Bruusgaard Institute of General Practice and Community Medicine, University of Osb, PO Box 1130 Blindern, 0317 Oslo, Norway Data from a community-based four-year prospective study were used to test the hypothesis that heavy physical work is a stronger predictor of tow back pain in smokers than in non-smokers. Of 708 working responders without low back pain during the entire year prior to 1990, 562 (79%) completed a questionnaire four years later in A job involving heavy lifting and much standing in 1990 was a strong predictor of low back pain in smokers four years later [odds ratio (OR) = 5.53, 95% confidence interval (CI) = , p < 0.01) after having adjusted for other job characteristics, demographic factors, emotional symptoms, physical exercise and musculoskeletal pain elsewhere. In non-smokers, having a job with heavy lifting and much standing was not associated with low back pain. One explanation may be that smoking leads to reduced perfusion and malnutrition of tissues in or around the spine and causes these tissues to respond inefficiently to mechanical stress. Key words: Aetiology; job characteristics; low back pain; risk factors; smoking. Occup. Med. vol. 49, , 1999 Received 10 August 1998; accepted in fined form 8 October 1998 This study was supported by a grant from Kanazawa Medical University (C95-10) and by the Research Council of Norway, the University of Oslo and the NHO work environment fund. INTRODUCTION An association between smoking and low back disorders has been found in several epidemiological studies. 1 ' 12 However, it is still a matter of debate whether these associations really represent causal associations. 12 ' 13 One of the problems that make the conclusion uncertain is the lack of a clear causal mechanism. Some studies indicate that smoking leads to reduced perfusion and malnutrition of tissues in the spine " 16 If this is true, one would also expect that repetitive mechanical stress on these tissues (e.g., in the form of heavy physical work) might be more harmful than it would be to healthy tissues, and more inclined to evoke symptoms of strain in smokers than in non-smokers. 17 In two studies of predictors of back pain, indications of an interaction between smoking and occupations associated with heavy physical work have been found. 8 ' 11 The aim of the present study was to test the hypothesis that heavy physical work is a stronger predictor of low back pain in smokers than in non-smokers. Correspondence and reprint requests to: W. Eriksan, Institute of General Practice and Community Medicine, University of Oslo, PO Box 1130 Blindern, 0317 Oslo, Norway. Tel: (+47) ; Fax: (+47) ; w.b.eriksenosamfunnsmed.ulo.no MATERIALS AND METHODS Materials Data from a community-based four-year prospective study were used to test the hypothesis. Ullensaker is a municipality with 18,000 inhabitants, situated 40 kilometres northeast of Oslo, Norway. In 1990, all inhabitants belonging to six age cohorts years, years, years, years, years and years were sent a questionnaire. After one reminder, 2,726 (67%) of 4,050 questionnaires were returned. The responders were sent a second questionnaire four years later (1994). A total of 1,791 responders were working in The sample used in the present study comprised the 708 working responders who had not been bothered by low back pain during the previous year. Of these, 562 (79%) returned the 1994 questionnaire. Dependent variable A standardized Nordic Questionnaire 18 ' 19 was used to measure musculoskeletal pain. Both in 1990 and 1994, the responders were presented a checklist of body regions (head, neck, shoulder, elbow, hand, upper back,

2 156 Occup. Med. vol. 49, 1999 lower back, hip, knee, ankle/foot) and were asked in which part of the body they had had pain during the past year and the previous week. The words were supplemented by 'region of interest-drawings'. To describe low back pain, the 'mannequin' was marked in the region below the 12th rib and above the gluteal folds. Independent variables Gender, year of birth, civil state, body height, body weight, job characteristics, physical exercise, emotional symptoms and daily smoking habits of the subjects were recorded in The responders were asked: 'Does your job involve a large amount of: work with hands over the shoulderlevel; work in the same position over a long time; monotonous movements; heavy lifting; sitting; standing; high work-pace'. The responders were asked to mark their responses on a checklist. This registration of job characteristics gave seven dichotomous variables, among which two variables reflected positively physical exertion in the job: 'a large amount of heavy lifting' and 'a large amount of standing'. As these two variables were strongly intercorrelated, and as they were dichotomous and consequently did not give the opportunity to explore dose-response effects, we combined the two into one variable with three categories (neither; either; both). We called this variable 'heavy physical work'. The responders were asked: 'Do you smoke daily?'. The responders could reply yes or no. The daily consumption of cigarettes was not registered. The number of sessions per week of leisure-time physical exercise (each session defined as at least 20 minutes, to the level of sweating or becoming short of breath) was categorized as 'fewer than 1 per week', '1-2 per week', '3 4 per week' or 'more than 4 per week'. This variable was dichotomized ('fewer than 1 per week' and 'at least 1 per week') when entered into the logistic regression analyses, as this gave similar results to using all categories. Civil state was dichotomized into married/cohabiting or not. Body height, weight, and body mass index were divided into tertiles. Emotional symptoms (anxiety and depression) were measured with the General Health Questionnaire We scored the information according to the likert scoring procedure where the score on each question ranged from 0 to 3. The mean of the twenty scores was divided into tertiles. Statistical analyses The analysis was based on the 562 responders who completed both questionnaires. The number of responders included in the analyses was somewhat lower than 562 as there was missing information on pain, smoking, physical exercise, job factors and emotional symptoms. Chi-square test and logistic regression analysis were used. A significance level of 0.05 was chosen. The interaction between smoking and heavy physical work was investigated by (1) univariate analyses of the relationship between heavy physical work and low back pain in smokers and non-smokers; (2) logistic regression analysis of low back pain in the total material with the interaction term 'smoking x heavy physical work' as co-variate and (3) logistic regression analyses (three models) in smokers and non-smokers. In Model 1, heavy physical work was entered as the only independent variable. In Model 2, heavy physical work, age and gender were entered as co-variates. In Model 3, the co-variates were: heavy physical work, age, gender, civil state, emotional symptoms, physical exercise, musculoskeletal pain other than low back pain and monotonous movements in the job. Logistic regression analyses were conducted with all co-variates entered simultaneously. RESULTS The responders to the 1994 questionnaire (hereafter referred to as the participants) were significantly older, more often women, more often non-smokers and less inclined to have a job of high standing than the nonresponders (data not shown). The demographic and life style characteristics of the participants at time 0 (1990) are presented in Table 1, by the one-year prevalence of low back pain four years later. The overall one-year prevalence of low back pain in Table 1. Demographic and life-style characteristics in 1990, by the one-year prevalence of low back pain in 1994; numbers and row percentages Characteristics in 1990 Gender Female Male Age (yrs) Civil status Single Married/cohabiting Smoker No Physical exercise (sessions per week) < >4 Musculoskeletal pain last year (except low back pain) No Total No ' Low back pain mw% (n) (95) (100) (41) (40) (62) (33) (19) (59) (136) (131) (63) (107) (56) (21) (6) (48) (147) (195) * A total of 562 respondents completed both questionnaires, but there was missing Information on pain In 27.

3 W. Erftsen et a/.: SmoWng, heavy physical work and low back pain was 36.4%. There was no association between smoking at time 0 and low back pain four years later. Heavy physical work as a predictor of low back pain in smokers and non-smokers Among smokers, having a job with heavy lifting and much standing was a predictor of low back pain four years later (Table 2). This was not seen among nonsmokers. In a logistic regression analysis (Table 3), low back pain was predicted by the interaction term 'smoke x heavy lifting and much standing in the job' [odds ratio (OR) = 4.04, 95% confidence interval (CI) = , p < 0.05]. In the same logistic regression model, low back pain was also predicted by having a job with heavy lifting and much standing (OR = 2.30, CI = , p < 0.05), exercising less than once a week (OR = 1.55, CI = , p < 0.05), and having had musculoskeletal pain other than low back pain during the previous year (OR = 1.61, CI = , p < 0.05), but not by age, gender, civil state, smoking, emotional symptoms, or monotonous movements in the job. The variables 'high work-pace', 'much sitting in the job', 'work with hands over the shoulder-level', 'work in the same position over a long time', body height, body weight and body mass index were included as co-variates in preliminary logistic regression analyses. As they were not related to low back pain and did not influence the model in any important way, they were not included in the final model. Among smokers, having a job with heavy lifting and much standing was a strong predictor of low back pain (OR = 5.53, CI = , p < 0.01) when adjusting for age, gender, civil status, emotional symptoms, physical exercise, having a job with monotonous movements and having had musculoskeletal pain other than low back pain during the previous year (Table 4). Among non-smokers, low back pain was not predicted by having a job with heavy lifting and much standing (OR= 1.12, CI = ). Table 2. The one-year prevalence of low back pain In 1994 by heavy physical work and smoking habits In 1990; row percentages (numbers) Smoking habits Type of work Smokers Neither heavy lifting nor much standing (n = 97) Either heavy lifting or much standing (n = 44) Both heavy lifting and much standing (n = 30) Non-smokers Neither heavy lifting nor much standing (n = 222) Either heavy lifting or much standing (n = 97) Both heavy lifting and much standing (o = 36) Low back pain row % O (n) (26) P (19) <0.01 (18) (77) (38) 0.46 (16) Table 3. Factors In 1990 which predicted low back pain in Results of one logistic regression analysis with all factors entered simultaneously; odds ratios (95% confidence Intervals; n = 466 Factors in 1990 Heavy physical work x smoke Interaction 1 a Interaction 2 b Heavy physical work Either heavy lifting or much standing Both heavy lifting and much standing Smoke Gender Female (ref. cat) Male Age (yrs) (ref. cat.) Civil state Married or cohabiting (ref. cat.) Single Emotional symptoms 0 Little (ref. cat.) Medium Much Musculoskeletal pain last year (except low back pain) Physical exercise 1 + time per week (ref. cat) < 1 time per week Monotonous movements in the job No (ref. cat) OR (95% CI) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) 'Bther lifting or standing 1 x V83'- b 'Both lifting and standing' x ' yes'. 0 General Health Questionnaire 20 Ukert scores divided Into tertlles. P Table 4. Heavy physical work In 1990 as a predictor of low back pain in 1994 in smokers and non-smokers. Results of six logistic regression analyses; odds ratios (95% confidence Intervals) Smokers Non-smokers Factors in 1990 OR (95% CO OR (95% CI) Model 1 : Heavy physical work No (ref. cat) Either lifting or standing Both lifting and standing Model 2: b Heavy physical work No (ref. cat) Either lifting or standing Both lifting and standing Model 3: c Heavy physical work Either lifting or standing Both lifting and standing 2.08 ( ) 4.10 ( ) 1.95 ( ) 3.85 ( ) 1.90 ( ) 5.53 ( ) 1.21 ( ) 1.51 ( ) ( ) 1.45 ( ) 1.07 ( ) 1.12 ( ) ' Odds ratios unadjusted. b Odds ratios adjusted for age and gender. c Odds ratios adjusted for age, gender, civil state, emotional symptoms, pain other than tow back pain, physical exercise, monotonous movements In the ob.

4 158 Occup. Med. Vol. 49, 1999 DISCUSSION Among smokers, having a job with heavy lifting and much standing was a strong predictor of low back pain four years later. Among non-smokers, having a job with heavy lifting and much standing was not a predictor of low back pain. Methodological considerations The relationship between working conditions and low back pain is complex. 21 ' 22 On the one hand, certain job factors are potential causes of back pain. On the other hand, back pain may influence job factors, as persons with chronic or recurrent back problems may not to be hired into specific high-exposure jobs, or may change to lower exposure jobs. However, the influence of this 'healthy worker effect' on the results of this study was reduced by the prospective design. The participants differed significantly from the nonresponders with respect to age, gender, smoking habits and type of work. This may have influenced the results, but can hardly explain the large difference between smokers and non-smokers. A problem in all epidemiological studies of low back pain is the nonspecificity of this symptom. Even so, selfreporting of low back pain has been found to have a good reliability in test-retest analysis, and there seems to be good agreement between information given in selfadministered questionnaires and information given in interviews and clinical examinations. 19 ' 23 Episodes of low back pain between 1990 and 1994 were not registered. This may have 'diluted' and thereby weakened, statistical associations, such as the one between low back pain and heavy physical work. However, it is unlikely that a dilution of the association between low back pain and heavy physical work could have taken place only among non-smokers. Self-reports of daily smoking have a good validity in surveys such as the present one. 24 Our confidence in the registration of job factors, crudely dived into 'high exposure' or 'not high exposure', is based on face validity and the fact that intermethod surveys have shown that questionnaires have an acceptable reliability for gross activities in the job. 23 ' 26 Pain-ridden persons are probably more likely than other persons to perceive aspects of their work as strenuous. In our study, we have taken this problem into account by not including in the cohort persons with low back pain, and by adjusting for musculoskeletal pain in other regions of the body. Smoking, heavy physical work, and low back pain Many studies have shown that heavy physical work, whether it be heavy lifting, prolonged standing, pushing and pulling or heavy physical work in general is associated with low back pain. 1 " 3 * 7 ' 8 ' 13 ' 27 Heavy physical work involves long-term repetitive mechanical stress on structures in and around the spine, such as vertebral end-plates, intervertebral discs, muscles, tendons and ligaments. This may result in injuries or evoke symptoms from already present weaknesses. In our study there was also an association between heavy physical work and low back pain. However, the association was exclusively confined to the smokers. This suggests that heavy physical work is more harmful to the lower back, or more inclined to evoke symptoms from pathology in this region, in smokers than non-smokers. Indications of an interaction between smoking and occupations associated with heavy physical work have been found in other studies. 8 ' 11 In the cross-sectional study of Boshuizen et al., the association between smoking and back pain was only seen among construction workers. l! Riihimaki et al. found that smoking was a predictor of sciatic pain in blue-collar workers, but not in white-collar workers. 8 In a 14-year prospective study of middle-aged farmers, Manninen et al. found that smoking was associated with as much as a 10-fold increase in the risk of developing sciatica. 9 The association between smoking and low back pain has been reported to be stronger among persons who suffer from respiratory diseases; 1 ' 10 similar to heavy physical work, chronic cough and expiratory obstruction may also involve mechanical stress on spinal structures. One possible explanation of our findings is that smoking leads to reduced perfusion and malnutrition of tissues in or around the spine. This may weaken the power of the spine's resistance to stress. It may also interfere with the healing of injuries. To date, the nutrition of vertebrae and intervertebral discs have been found to be affected by smoking 2 ' 14 " 16 and 'disc pathology' is, indeed, the cause of many long term cases of low back pain, whether it be in the form of herniated discs or internal disc disruptions. 28 Fractures of the vertebral end-plates seem to play an important role in the pathogenesis of disc disruptions. 29 ' 30 Many episodes of low back pain, however, are probably not due disc pathology, but to pathological processes in other structures, such as muscles, tendons and ligaments. Hence, an important question is whether smoking may also affect the nutrition of these structures, or whether smoking may influence the perception of pain, for instance through nicotine-effects on the central nervous system. If so, it could also explain the association found between smoking and musculoskeletal pain in other regions of the body. 11 * 31 ' 32 In the present study, there was no association found between smoking in 1990 and low back pain in Such a trend was seen, though, in the subgroup of responders who were exposed to both heavy lifting and much standing in their jobs, but it did not reach statistical significance as this group of respondents comprised only 66 persons. The trend was of borderline significance (p = 0.06) when responders who had been bothered by low back pain in 1990 were also included in the sample (data not shown). Confounding The statistical relationship between heavy physical work and low back pain may be influenced by confounding

5 W. Eriksen et al.: Smoking, heavy physical work and low back pain 159 factors, such as socio-economic factors, psychological factors not reflected in GHQ-20 or life-style factors other than smoking and physical exercise. If nonsmokers have a more careful way of lifting than smokers it could partly explain the observed interaction between smoking and heavy work. Other predictors of low back pain Musculoskeletal pain in other regions of the body was a predictor of low back pain. Symmons et al. found that pain in other regions of the body was a predictor of back pain in middle-aged women. 33 One explanation may be that some persons have a general vulnerability to, and are prone to develop several types of musculoskeletal problems. Another explanation may be that certain types of musculoskeletal disorders tend to generalize or spread from one region of the body to another. The responders who were exercising at least once a week were less inclined to develop low back pain. Other prospective studies have also shown a negative association between sports and physical activity in the leisure-time and low back pain. 34>35 CONCLUSIONS The study indicates that heavy physical work is a stronger predictor of low back pain in smokers than in non-smokers. One explanation could be that smoking leads to reduced perfusion and malnutrition of tissues in or around the spine, and thereby makes spinal structures more vulnerable to mechanical stress. REFERENCES 1. Frymoyer JW, Pope MH, Costarica MC, Rosen JC, Goggin JE, Wilder DG. Epidemiological studies of lowback pain. Spine 1980; 5: Frymoyer JW, Pope MH, Clements JH, Wilder DG, Mac- Pherson B, Ashikaga T. Risk factors in low-back pain. J Bone Joint Surg 1983; 65A: Svensson H, Vedin A, Wilhelmsson C, Andersson GBJ. Low-back pain in relation to other diseases and cardiovascular risk factors. Spine 1983; 8: Biering-S0rensen F, Thomsen C. Medical, social and occupational history as risk indicators for low-back trouble in a general population. Spine 1986; 11: Deyo RA, Bass JE. Lifestyle and low-back pain. The influence of smoking and obesity. Spine 1989; 14: Dewey ME, Dickinson CE, Foreman TK, Troup JDG. Back pain, ventilatory function, chest symptoms, and smoking. J Spinal Disord 1989; 2: Heliovaara M, Makela M, Knekt P, Impivaara O, Aromaa A. Determinants of sciatica and low-back pain. Spine 1991; 16: Riihimaki H, Viikari-Juntura E, Moneta G, Kuha J, VidemanT,Tola S. Incidence of sciatic pain among men in machine operating, dynamic physical work, and sedentary work. Spine 1994; 19: Manninen P, Rihimak H, Heliovaara M. Incidence and risk factors of low back pain in middle-aged farmers. Occup Med 1995; 45: Leboeuf-Yde C, Yashin A, Lauritzen T. Does smoking cause low back pain? Results from a population-based study. J Manip Physiolog Therap 1996; 19: Boshuizen HC, Verbeek JHAM, Broersen JPJ, Weel ANH. Do smokers get more back pain? Spine 1993; 18: Leboeuf-Yde C, Yashin A. Smoking and low back pain: Is the association real? J Manipul Physiol Ther 1995; 18: 457^ Burdorf A, Sorock G. Positive and negative evidence of risk factors for back disorders. Scand JWbrk Environ Health 1997; 23: Holm S, Nachemson A. Nutrition of the intervertebral disc: acute effects of cigarette smoking. Uppsala J Med Sci 1988; 93: Kauppila LI. Can low back pain be due to lumbar-artery disease? Lancet 1995; 346: Battie MC, Videman T, Gill K, el al. Smoking and lumbar intervertebral disc degeneration: an MRI study of identical twins. Spine 1991; 16: Ernst E. Smoking, a cause of back trouble? BrJ Rheumatol 1993; 32: Kuorinka I, Jonsson B, Kilbom A, et al. Standardised Nordic questionnaires for the analysis of musculoskeletal symptoms. App Ergonom 1987; 18: Holmstrom E, Moritz U. Low back pain correspondence between questionnaire, interview and clinical examination. Scand J Rehabil Med 1991; 23: Goldberg D, Williams P. A User's Guide to the General Health Questionnaire. Windsor: NFER-Nelson, Burdorf A, Rossignol M, Fathalla FA, Snook SH, Herrick RF. Challenges in assessing risk factors in epidcmiologic studies on back disorders. Am J Ind Med 1997; 32: Kraus JF, Gardner L, Collins J, Sorock G, Volinn E. Design factors in epidemiologic cohort studies of work-related low back injury or pain. Am J Ind Med 1997; 32: Walsh K, Coggon D. Reproducibility of histories of low back pain obtained by self-administered questionnaire. Spine 1991; 16: US Department of Health and Human Services. Reducing the health consequences of smoking: 25 Years of Progress. A Report of the Surgeon General. DHHS Publication no. (CDC) Rockville, MD (USA): US DHHS, Viikari-Juntura E, Rauas S, Martikainen R, Kuosma E, Riihimaki H, Takala EP, Saarenmaa K. Validity of selfreported physical work load in epidemiologic studies on musculoskeletal disorders. Scand J Vftrrk Environ Health 1996; 22: Burdorf A. Reducing random measurement error in assessing postural load on the back in epidemiologic surveys. Scand JWbrk Environ Health 1995; 21: Garg A, Moore JS. Epidemiology of low-back pain in industry. Occup Med (US) 1992; 7: Schwarzer AC, Aprill CN, Derby R, Fortin J, Kine G, Bogduk N. The prevalence and clinical features of internal disc disruption in patients with chronic low back pain. Spine 1995; 20: Farfan HE A reorientation in the surgical approach to degenerative intervertebral joint disease. Orthop Clin North Am 1977; 8: Bogduk N, Twomey LT. Clinical Anatomy of the Lumbar Spine, Second Edition. London, UK: Churchill Livingstone, 1991.

6 160 Occup. Med. Vol. 49, Eriksen W, Brage S, Bruusgaard D. Does smoking aggravate musculoskeletal pain? Scand J Rheumatol 1997; 26: Brage S, Bjerkedal T. Musculoskeletal pain and smoking in Norway. J Epidemiol Community Health 1996; 50: Symmons DPM, van Hemert AM, Vandenbroucke JP, Valkenburg HA. A longitudinal study of back pain and radiological changes in the lumbar spines of middle aged women. I. Clinical findings. Ann Rheum Dis 1991; 50: Gyntelberg E One-year incidence of low back pain among male residents of Copenhagen, aged Dan Med Bull 1974; 21: Leino P. Does leisure time physical activity prevent low back disorders? Spine 1993; 18: