BMI, Occupational Activity, and Leisure-Time Physical Activity: An Exploration of Risk Factors and Modifiers for Knee Osteoarthritis.

Kathryn Remmes Martin, Diana Kuh, Tamara B. Harris, Jack M. Guralnik, David Coggon, Andrew K. Wills

Research output: Contribution to journalAbstractpeer-review


Background/Purpose: Risk of knee osteoarthritis (OA) is increased by obesity, and also by physical activities which mechanically stress the joint. The few studies which have examined the interaction between body mass index (BMI) and activity have yielded inconsistent findings. We examined whether the association of BMI across adult life on midlife knee OA was modified by occupational activity and/or leisure-time physical activity (PA).
Methods: Data came from a nationally representative British birth cohort in which 2597 participants underwent clinical examination and assessment of knee OA at 53y. Main exposures at 36, 43 and 53 years: BMI (kg/m2), self-reported leisure-time PA (inactive, less-active, and most-active) and occupational activity – kneeling/squatting; lifting; climbing; or sitting, with likelihood of exposure (unlikely, somewhat-likely and most-likely) assigned using a job-exposure matrix. Odds ratios (OR) for knee OA were estimated at each time-point using logistic regression. Interactions between BMI and occupational or leisure-time PA were tested using a likelihood ratio test (LRT); where p was<0.05, we identified elevated risk by examining BMI in each activity-level, and activity in each BMI-stratum based on standard deviations (-1SD, 0SD, and 1SD). Analyses were stratified by sex, and adjusted models included socioeconomic position and health status variables.
Results: For men, evidence of an interaction between BMI and lifting (LRT p=0.01) occurred only at 43y. BMI increased knee OA risk for men who were most-likely to lift at work (OR per z-score of BMI: 3.55, 95%CI: 1.72-7.33). Interestingly, among those men most-likely to lift at work, those with lower BMI (≤0SD) were at lower risk of knee OA than those somewhat-likely to lift (-1SD – OR: 0.14, 95%CI: 0.03-0.60; 0SD – OR: 0.30, 95%CI: 0.11-0.84). For women, evidence of an interaction between BMI and leisure-time PA (LRT: p=0.005) occurred only at 43y. BMI increased knee OA risk at higher levels of PA (OR per z-score of BMI: 1.59, 95%CI: 1.26-2.00 in inactive; 1.70, 95%CI: 1.14-2.55 in less-active; and 4.44; 95%CI: 2.26-8.36 in most-active). Interestingly, among those women who were most-active, those with lower BMI (≤0SD) were at lower risk of knee OA than those less-active (-1SD – OR: 0.14, 95%CI: 0.04-0.48; 0SD – OR: 0.36, 95%CI: 0.18-0.73).
Conclusion: Our results suggest that high BMI may be more detrimental to joint health among individuals exposed to greater levels of activity. A better understanding of these relationships may be required to optimize the potential benefits of PA and identify high risk groups within particular occupations.
Original languageEnglish
Article number800
Pages (from-to)S348-S348
Number of pages1
JournalArthritis & Rheumatism
Issue number10
Publication statusPublished - Oct 2012
Event76th Annual Scientific Meeting of the American-College-of-Rheumatology (ACR) and Association-of-Rheumatology-Health-Professionals (ARHP) 2012 - Washington, Washington, United Kingdom
Duration: 9 Nov 201214 Nov 2012

Cite this