Article Text

Download PDFPDF
  1. Kelley-Anne Sabarre
  1. Epidemiology and Community Medicine, Faculty of Medicine, University of Ottawa, ON, Canada


Introduction Arthritis is a chronic, progressive, inflammatory disease; it is known as the leading cause of chronic pain in North America. Researchers identified cigarette smoking as a major preventable risk factor for arthritis, and sex as a non-modifiable risk factor for arthritis.

Objective To determine the association between a smoking–sex interaction, and the prevalence of arthritis among obese Canadians.

Methods Using data from the Canadian Community Health Survey (CCHS), cycle 4.0, interaction variables were created to demonstrate the effect of smoking, the effect of sex, and the joint effect of smoking and sex in relation to the prevalence of arthritis. The association between exposures of interest, smoking and sex, and health outcome of interest, prevalence of arthritis, was examined using multiple logistic regression and log binomial regression modelling. Measures of additive interaction (relative excess risk due to interaction (RERI), attributable proportion due to interaction (AP), synergy index (S)) using both additive and multiplicative models were calculated and assessed.

Results The prevalence of the health outcome of interest was common; 29.7% of the study population reported having arthritis. The OR of arthritis was: 1.285 (1.148, 1.439) for male smokers; 1.606 (1.431, 1.802) for female non-smokers; and 2.240 (1.997, 2.513) female smokers compared with male non-smokers. All ORs were statistically significant, with the exception of individuals with some postsecondary education compared with individuals with less than secondary school education (OR=0.866 (0.747, 1.005)). Using logistic regression, measures of assumed additive interaction implied statistical significance of a biological interaction (RERI: 0.349 (0.142, 0.556), AP: 0.156 (0.065, 0.247), S 1.391 (1.102, 1.756)). The relative risk of arthritis was: 1.19263 (1.10369, 1.29105) for male smokers; 1.38933 (1.28697, 1.50282) female non-smokers; 1.61469 (1.49868, 1.74350) for female smokers compared with male non-smokers. Using log binomial regression, measures of additive interaction implied no statistical significance of a biological interaction (RERI: 0.033 (−0.072, 0.137), AP: 0.020 (−0.045, 0.085), S: 1.056 (0.879, 1.269)).

Conclusions Although it appears that a synergistic smoking–sex interaction was present when using a multiplicative model, the additive model did not confirm this synergistic effect. Given that the additive model reports more accurate risk for a common outcome such as arthritis, this leads us to believe that smoking and sex do not have a synergistic effect on the prevalence of arthritis among obese Canadians.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.