A majority of the adult population in many developed and developing countries is overweight or obese.1,2 Obese individuals have higher rates of death from cardiovascular diseases, as well as significantly elevated risks for diabetes and many specific cancers. Overweight and obesity are also associated with higher risk for lumbar disc disease, hospitalization for low back pain, several types of osteoarthritis, sciatica, and pain.3–11 Clinical and experimental studies have shown not only mechanical but also metabolic obesity-specific pathways for the development of facet joint osteoarthritis, disc degeneration, and hypertrophy of spinal ligaments.8,12–17 These spondylotic changes narrow the spinal canal, which can progress to lumbar spinal stenosis (LSS),18 a condition associated with a health burden and impaired quality of life comparable with stroke, cardiovascular diseases, and diabetes.19 Common symptoms include leg pain, especially during walking, associated with numbness and paresthesia, and sometimes loss of motor control and bladder disturbances.
An estimated 136 of 100,000 persons older than 65 years in the United States undergo surgery for LSS every year, which is a 4-fold increase compared with 1985. We have also observed a similar trend in Europe, where LSS has become the most common indication for spine surgery in many European countries.20–23 Increasing age in the population, increased use of magnetic resonance imaging, and systematic differences in health care are major factors to account for this trend, although the impact of biological factors might also be of importance.13,17,20,24–27 With the current increased rate in LSS, the global rates of overweight and obesity have concomitantly increased dramatically since 1980.1,11,28–30
Because obese individuals are at increased risk to develop osteoarthritis in both loaded and unloaded joints,17,25,31 we hypothesized that an elevated body mass index (BMI) will also increase the rate of clinically manifested LSS. Our principal objective was to assess the relation between BMI and clinically overt LSS in a large cohort of Swedish construction workers.
MATERIALS AND METHODS
The cohort consists of Swedish construction workers who participated in a nationwide occupational health surveillance program (Bygghälsan), which was initiated through a trade agreement between employers and unions.32 The workers were invited to participate in health examinations and the participation rate was at least 80%. At inclusion, weight was measured with a scale and height was measured with a stadiometer. The workers' job title (22 categories) and smoking habits were also registered, and from 1971, the data were computerized. The examinations ended in early 1993, and the computerized register contains 389,132 persons. The proportion of females was 5%, of which 42% were office workers.
Nearly all (99.9%) Swedish residents have a personal identification number, an important tool for complete linkage with Sweden's national registries. The Swedish National Patient Register (NPR) started in 1964 and covered 83% of the Swedish population in 1972 and all inpatient care since 1987. The completeness of ascertainment and the accuracy of classification of diagnoses in the NPR are both high; the completeness for spinal diagnosis and surgical procedures in the NPR was, on average, 86% in 2001 to 2012.33,34
LSS was defined by the International Classification of Diseases (ICD) with International Classification of Diseases, Ninth Revision (ICD-9) codes 724, 724.0, or 724.00, and International Classification of Diseases, Tenth Revision (ICD-10) codes M48.0 or M48.0K. The occurrence of hospitalization that is due to LSS was collected through linkage to the NPR. In addition, a sensitivity analysis was performed using as outcome a diagnosis of LSS combined with a surgical procedure of the lumbar spine. Individuals with a prevalent LSS at cohort entry were excluded (Figure 1). The observation period was from cohort entry until December 31, 2011, death, emigration, or the occurrence of first diagnosis due to LSS, whichever occurred first. Date of death was identified by individual linkage to the national Cause of Death Register, and emigration by linkage to the national population register.
BMI, calculated as weight (kg)/height2 (m2), was used both as a categorical and as a continuous variable. Four categories were used to classify the participants into underweight (BMI <18.5 kg/m2), normal weight (BMI 18.5–24.99 kg/m2), overweight (BMI 25–29.99 kg/m2), and obese (BMI ≥30 kg/m2). Workers with extreme BMI values (<15 kg/m2 or >50 kg/m2) were not eligible for this study (Figure 1).
A Directed Acyclic Graph was constructed to identify a causal model for obesity and LSS, as well as to identify suitable covariates.35,36 Poisson regression was used to estimate age- and multivariable-adjusted incidence rate ratios (IRRs), with 95% confidence intervals (CIs). The latter model included ‘age (continuous), sex, smoking status (never, former, moderate current, heavy current, unknown), and occupation (22 categories). Nonlinear trends of risk were displayed using restricted cubic spline curves estimated by a multivariable-adjusted Poisson regression model. We used 4 “knots” placed at the 5th, 35th, 65th, and 95th percentiles of the cumulative BMI distribution. The reference level was set to 20 kg/m2. Finally, we performed stratified analysis by sex. The statistical analyses were performed with SAS, version 9.4, Stata version 11 and SPSS, version 20.
The study was approved by the regional ethical review board.
Characteristics of the participants by BMI category are presented in Table 1. Two-thirds (65%, n = 235,247) of the participants had normal weight, 29% (n = 106,730) were overweight, 5% (n = 16,695) were obese, and 2% (n = 5795) were underweight. A vast majority of the construction workers were males (95%, n = 345,921). The average age at baseline was 34 (SD, 13) years. As expected, there was a positive association between age and BMI.
During an average follow-up of 30.7 years (11,190,944 person-years), 2381 participants were diagnosed with LSS. More than half (51%) of these cases were overweight or obese at baseline. We observed an almost linear positive association between BMI and LSS (Figure 2), that is, the lowest rates of LSS were found in lean workers and the highest in obese workers. Assuming a linear relation, the incidence of LSS showed an increase of 10% (95% CI, 9%–11%) per BMI unit.
Accordingly, the IRRs of LSS varied across BMI categories, with the highest values found in obese individuals (Table 2). Compared with normal weight, obesity was associated with a multivariable-adjusted IRR of 2.18 (95% CI, 1.87 to 2.53) for LSS, and overweight was associated with an IRR of 1.68 (95% CI, 1.54–1.83). Underweight workers halved their future risk of LSS (IRR 0.52, 95% CI: 0.30–0.90).
As expected, the number of outcomes was lower (n = 1816) in patients who were diagnosed with LSS and who underwent surgery. The proportion of patients who underwent surgery out of all eligible patients with LSS was similar in the 4 BMI groups: 77% (10/13) in the underweight group, 75% (855/1144) in the normal weight group, 79% (793/1019) in the overweight group, and 77% (158/205) in the obese group (Tables 2 and 3). Nevertheless, when we used surgically treated LSS as an outcome, the results remained comparable with the original analysis (Table 3 and Figure 3). Restricted cubic spline curves stratified by sex indicated a higher IRR with increasing BMI, irrespective of sex (see Supplemental Digital Contents Figures 1 and 2, available at: https://links.lww.com/BRS/B11, for males and females, respectively). Not surprisingly, there was no statistical interaction between sex and BMI (P = 0.57).
To our knowledge, this is the first cohort study on the relation between BMI and clinical LSS. Our major finding is that higher BMI increases the risk of clinical LSS, with similar estimates in both males and females.
The findings are consistent with those from a cross-sectional subgroup analysis of the Framingham cohort,25 in which an association between osteoarthritis of the lumbar spine and a tendency for radiological LSS associated with increased BMI was shown.25 The analysis included 187 participants, of whom 13 had radiological findings of LSS. Another recent cross-sectional study of participants (n = 938), in whom 78% were considered to have more than moderate radiographical central spinal stenosis, also displayed a positive association between radiological LSS and BMI.31
Although studies specific for BMI and LSS are scarce, previous research has shown overweight and obesity to be risk factors for sciatica and hospitalization due to LBP or lumbar disc disease.3,4,9 Moreover, a BMI higher than 25 kg/m2 was found to be a risk factor for hospitalization due to LBP among metal industry workers, and overweight and obesity was, in addition, a risk factor for sciatica in municipal employees.3,4
The main strengths of our study are the prospective design, the large sample size, and the high validity of both exposure and outcome. Date of diagnosis, date of surgery, occurrence of death, and emigration data were collected through national registers known to have high accuracy.33 Complete linkage between the registers is rendered by the individual personal identification number provided to all Swedish residents.37 Furthermore, the sensitivity analysis using only cases with codes for diagnosis connected to a surgical procedure revealed similar results as the original analysis.
Several potential limitations of the study need to be discussed. Although BMI is an established measure of overweight and obesity in both the clinic and research, the measure has admittedly a major limitation in its inability to differentiate lean mass from adipose tissue. Furthermore, weight and height were assessed only once and we had, on average, a long follow-up, but given our design constraints, changes in BMI during follow-up should produce only conservatively biased estimates. Moreover, although our selected population might be considered a conceivable limitation, the design does not jeopardize the validity of the relation between BMI and LSS.38 Other possible limitations are residual confounding that is due to level of physical activity, comorbidities, and demographics. Although we adjusted for 22 occupations with diverse levels of physical activity at work, we have no data on leisure time physical activity. Somatic comorbidities are common among patients with LSS, and obesity is a known cause of cardiovascular disease, diabetes, and specific cancers. Similar to LSS, these comorbidities have been characterized as a disease process with a long induction time and are associated with age. The mean age at inclusion was 34 years and a model adjusting for comorbidities at inclusion is unlikely to affect our conclusion. Furthermore, except for emigration, we have no data on regional demographic changes, which is a possible limitation in that the surgical rates for LSS can differ in and between counties.27 These changes and differences, however, are likely to only marginally affect our point estimates.
The increase in the IRR for LSS found in obese and overweight patients is likely to be multicausal with several conceivable obese-specific pathogenetic pathways.8,39 An increase in body mass leads to an abnormal and altered load on the spine, and obese individuals have, in addition, a lower relative muscle mass than normal weight individuals, which further increases strain on the lumbar spine.8,16,40 Furthermore, besides direct biomechanical effect on cartilage and skeleton, indirect effects by changes in body mass can be mediated by mechanoreceptors, cytokines, and growth factors. These factors have the potential to alter the properties of bone matrix, ligamentum flavum, synovium, and cartilage, all of which could promote the development of osteoarthritis, hypertrophy of the ligamentum flavum, and disc degeneration.5,12–15,25,26,39 Decreased muscle mass is also associated with insulin resistance, which further weakens the skeletal muscles and promotes systemic inflammation.40,41 Adiponectin and leptin, hormones secreted by adipocytes, regulate low-grade inflammation caused by obesity, and an increase in the levels of C-reactive protein, interleukins, and tumor necrosis factors is related to the progression of spondylosis.8,39,42,43 In addition, a high serum concentration of free fatty acids is known to increase systemic inflammation and development of osteoarthritis.41,44,45 Moreover, hyperlipidemia-induced atherosclerosis is proposed as a cause of disc degeneration and ischemic pain.46,47 Finally, obesity is related to reduced walking capacity and kinesophobia, events known to also increase muscle loss and pain.48–50
Obesity and overweight are associated with an increased risk to develop LSS. Our findings indicate that obesity is one plausible explanation for the increased number of patients with clinical LSS. Whether weight loss reduces symptoms and progression of LSS remains to be established.
- In recent decades, a concurrent increase in the prevalence of obesity and spinal stenosis has occurred in many countries, and previous research has displayed several conceivable obese-specific pathogenetic pathways for the development of clinically manifested lumbar spinal stenosis.
- In our study, high body mass index was found to be a novel risk factor for the future development of lumbar spinal stenosis.
The work was supported by Swedish Council for Health Working Life and Welfare (2011–426), and the Department of Research and Development Västernorrland County Council.
Supplemental digital content is available for this article. Direct URL citation appearing in the printed text is provided in the HTML and PDF version of this article on the journal's Web site (www.spinejournal.com).
1. Finucane MM, Stevens GA, Cowan MJ, et al. National, regional, and global trends in body-mass index since 1980: systematic analysis of health examination surveys and epidemiological studies with 960 country-years and 9.1 million participants. Lancet 2011;377:557–67.
2. Swinburn BA, Sacks G, Hall KD, et al. The global obesity pandemic: shaped by global drivers and local environments. Lancet 2011;378:804–14.
3. Kaaria S, Kaila-Kangas L, Kirjonen J, et al. Low back pain, work absenteeism, chronic back disorders, and clinical findings in the low back as predictors of hospitalization due to low back disorders: a 28-year follow-up of industrial employees. Spine (Phila Pa 1976) 2005;30:1211–8.
4. Kaaria S, Leino-Arjas P, Rahkonen O, et al. Risk factors of sciatic pain: a prospective study among middle-aged employees. Eur J Pain 2011;15:584–90.
5. Liuke M, Solovieva S, Lamminen A, et al. Disc degeneration of the lumbar spine in relation to overweight. Int J Obes (Lond) 2005;29:903–8.
6. Janke EA, Collins A, Kozak AT. Overview of the relationship between pain and obesity: What do we know? Where do we go next? J Rehabil Res Dev 2007;44:245–62.
7. Stone AA, Broderick JE. Obesity and pain are associated in the United States. Obesity (Silver Spring) 2012;20:1491–5.
8. Vincent HK, Heywood K, Connelly J, et al. Obesity and weight loss in the treatment and prevention of osteoarthritis. PM R 2012;4:S59–67.
9. Wahlstrom J, Burstrom L, Nilsson T, et al. Risk factors for hospitalization due to lumbar disc disease. Spine (Phila Pa 1976) 2012;37:1334–9.
10. Berenbaum F, Eymard F, Houard X. Osteoarthritis, inflammation and obesity. Curr Opin Rheumatol 2013;25:114–8.
11. Ng M, Fleming T, Robinson M, et al. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet 2014;384:766–81.
12. Piscoya JL, Fermor B, Kraus VB, et al. The influence of mechanical compression on the induction of osteoarthritis-related biomarkers in articular cartilage explants. Osteoarthritis Cartilage 2005;13:1092–9.
13. Sairyo K, Biyani A, Goel V, et al. Pathomechanism of ligamentum flavum hypertrophy: a multidisciplinary investigation based on clinical, biomechanical, histologic, and biologic assessments. Spine (Phila Pa 1976) 2005;30:2649–56.
14. Sairyo K, Biyani A, Goel VK, et al. Lumbar ligamentum flavum hypertrophy is due to accumulation of inflammation-related scar tissue. Spine (Phila Pa 1976) 2007;32:E340–7.
15. Kosaka H, Sairyo K, Biyani A, et al. Pathomechanism of loss of elasticity and hypertrophy of lumbar ligamentum flavum in elderly patients with lumbar spinal canal stenosis. Spine (Phila Pa 1976) 2007;32:2805–11.
16. Han KS, Rohlmann A, Zander T, et al. Lumbar spinal loads vary with body height and weight. Med Eng Phys 2013;35:969–77.
17. Gandhi R, Woo KM, Zywiel MG, et al. Metabolic syndrome increases the prevalence of spine osteoarthritis. Orthop Surg 2014;6:23–7.
18. Genevay S, Atlas SJ. Lumbar spinal stenosis. Best Pract Res Clin Rheumatol 2010;24:253–65.
19. Battie MC, Jones CA, Schopflocher DP, et al. Health-related quality of life and comorbidities associated with lumbar spinal stenosis. Spine J 2012;12:189–95.
20. Ciol MA, Deyo RA, Howell E, et al. An assessment of surgery for spinal stenosis: time trends, geographic variations, complications, and reoperations. J Am Geriatr Soc 1996;44:285–90.
21. Jansson KA, Blomqvist P, Granath F, et al. Spinal stenosis surgery in Sweden 1987–1999. Eur Spine J 2003;12:535–41.
22. Deyo RA, Mirza SK, Martin BI, et al. Trends, major medical complications, and charges associated with surgery for lumbar spinal stenosis in older adults. JAMA 2010;303:1259–65.
23. Du Bois M, Szpalski M, Donceel P. A decade's experience in lumbar spine surgery
in Belgium: sickness fund beneficiaries, 2000–2009. Eur Spine J 2012;21:2693–703.
24. Perruccio AV, Power JD, Badley EM. Revisiting arthritis prevalence projections—it's more than just the aging of the population. J Rheumatol 2006;33:1856–62.
25. Kalichman L, Guermazi A, Li L, et al. Association between age, sex, BMI and CT-evaluated spinal degeneration features. J Back Musculoskelet Rehabil 2009;22:189–95.
26. Guilak F. Biomechanical factors in osteoarthritis. Best Pract Res Clin Rheumatol 2011;25:815–23.
27. Birkmeyer JD, Reames BN, McCulloch P, et al. Understanding of regional variation in the use of surgery. Lancet 2013;382:1121–9.
28. Flegal KM, Carroll MD, Kuczmarski RJ, et al. Overweight and obesity in the United States: prevalence and trends, 1960–1994. Int J Obes Relat Metab Disord 1998;22:39–47.
29. Flegal KM, Carroll MD, Ogden CL, et al. Prevalence and trends in obesity among US adults, 1999–2008. JAMA 2010;303:235–41.
30. Wang YC, McPherson K, Marsh T, et al. Health and economic burden of the projected obesity trends in the USA and the UK. Lancet 2011;378:815–25.
31. Ishimoto Y, Yoshimura N, Muraki S, et al. Associations between radiographic lumbar spinal stenosis and clinical symptoms in the general population: the Wakayama Spine Study. Osteoarthritis Cartilage 2013;21:783–8.
32. Bygghälsokohorten [Umeå Universitet Web site]. Available at: http://www.phmed.umu.se/enheter/envmed/forskning/forskningsomraden/bygghalsokohorten/
. Accessed April 20, 2015.
33. Ludvigsson JF, Andersson E, Ekbom A, et al. External review and validation of the Swedish national inpatient register. BMC Public Health 2011;11:450.
34. Strömqvist B FP, Hägg O, Knutsson B, et al. Annual register report 2014 [Swedish Society of Spinal Surgeons Web site]. Available at: http://www.4s.nu/pdf/Report_2014_Swespine_Engl_ver_141204.pdf
. Accessed April 20, 2015.
35. Shrier I, Platt RW. Reducing bias through directed acyclic graphs. BMC Med Res Methodol 2008;8:70.
36. VanderWeele TJ, Hernan MA, Robins JM. Causal directed acyclic graphs and the direction of unmeasured confounding bias. Epidemiology 2008;19:720–8.
37. Ludvigsson JF, Otterblad-Olausson P, Pettersson BU, et al. The Swedish personal identity number: possibilities and pitfalls in healthcare and medical research. Eur J Epidemiol 2009;24:659–67.
38. Rothman KJ, Gallacher JE, Hatch EE. Why representativeness should be avoided. Int J Epidemiol 2013;42:1012–4.
39. Berenbaum F, Sellam J. Obesity and osteoarthritis: what are the links? Joint Bone Spine 2008;75:667–8.
40. Vincent HK, Raiser SN, Vincent KR. The aging musculoskeletal system and obesity-related considerations with exercise. Ageing Res Rev 2012;11:361–73.
41. Dandona P, Aljada A, Chaudhuri A, et al. Metabolic syndrome: a comprehensive perspective based on interactions between obesity, diabetes, and inflammation. Circulation 2005;111:1448–54.
42. Dumond H, Presle N, Terlain B, et al. Evidence for a key role of leptin in osteoarthritis. Arthritis Rheum 2003;48:3118–29.
43. Griffin TM, Huebner JL, Kraus VB, et al. Extreme obesity due to impaired leptin signaling in mice does not cause knee osteoarthritis. Arthritis Rheum 2009;60:2935–44.
44. Brunner AM, Henn CM, Drewniak EI, et al. High dietary fat and the development of osteoarthritis in a rabbit model. Osteoarthritis Cartilage 2012;20:584–92.
45. Griffin TM, Huebner JL, Kraus VB, et al. Induction of osteoarthritis and metabolic inflammation by a very high-fat diet in mice: effects of short-term exercise. Arthritis Rheum 2012;64:443–53.
46. Leino-Arjas P, Kaila-Kangas L, Solovieva S, et al. Serum lipids and low back pain: an association? A follow-up study of a working population sample. Spine (Phila Pa 1976) 2006;31:1032–7.
47. Leino-Arjas P, Kauppila L, Kaila-Kangas L, et al. Serum lipids in relation to sciatica among Finns. Atherosclerosis 2008;197:43–9.
48. Hemmingsson E, Ekelund U. Is the association between physical activity and body mass index obesity dependent? Int J Obes (Lond) 2007;31:663–8.
49. Vincent HK, Lamb KM, Day TI, et al. Morbid obesity is associated with fear of movement and lower quality of life in patients with knee pain-related diagnoses. PM R 2010;2:713–22.
50. Vincent HK, Omli MR, Day T, et al. Fear of movement, quality of life, and self-reported disability in obese patients with chronic lumbar pain. Pain Med 2011;12:154–64.