Abstract

Previous studies have suggested that smoking might be protective against the development of osteoarthritis (OA) of the knee. A group of 1003 women aged 45-64 years (mean 54.2 years) from the Chingford general population survey were studied to examine the effect of cigarette smoking on the prevalence of radiologically confirmed OA at different sites. Standard anteroposterior radiographs of the hand and knee were available in 985 women. Disease classification was made on the basis of radiological OA and symptomatic radiological OA. Odds ratios were calculated and adjusted for age and body mass index. A total of 463 (46.2%) women were ever smokers compared with 540 (53.8%) non-smokers. Ever smokers had consumed an average of 14.9 cigarettes a day for a mean of 25.7 years. For radiological OA of the distal interphalangeal joint (DIP) (140 women), proximal interphalangeal joint (40 women), carpometacarpal joint (160 women), and knee joint (118, women) there was no reduced risk of OA in ever smokers. In the small number of subjects with generalised OA (22 women) there was a non-significant 40% reduction of radiological OA in ever smokers (odds ratio 0.63; 95% confidence interval 0.24 to 1.68). Results were similar for subjects with radiographic clinical OA, except the DIP joint which showed a positive association between smoking and Heberden's nodes (odds ratio 2.02, 95% confidence interval 1.89 to 3.42). Results were similar when analysed using current smokers against never smokers. These results do not support an inverse association between cigarette smoking and OA in women. A possible inverse relation with the small subgroup of women with generalised OA and an effect of cigarettes on disease severity cannot, however, be discounted.