Abstract
OBJECTIVE: To examine interprovincial variations of arthritis prevalence focusing on comparisons between Quebec and the rest of Canada. METHODS: Data were derived from the 1991 General Social Survey (GSS), the 1991 Health and Activity Limitation Survey (HALS), and the 1994 and 1996 National Population Health Surveys (NPHS). Arthritis was variously ascertained through self-report of people aged 15 years or older. Prevalence in Quebec was compared with other provinces using extremal quotients (EQ) and correlation analysis. Multiple logistic regression analysis (1996 NPHS) was used to determine whether the low prevalence in Quebec persisted after controlling for confounding factors including age, sex, education, marital status, occupation, body mass index (BMI), comorbidity, and smoking. RESULTS: Quebec consistently had the lowest provincial prevalence of arthritis, with age-sex adjusted prevalences of 18.4% (GSS), 1.9% (HALS), 8.8%, and 10.1% (1994 and 1996 NPHS), which were significantly lower than the corresponding national estimates: 21.2%, 3.1%, 12.9%, and 13.3%. EQ from different surveys varied from 1.5 to 3.0 (significantly > 1). Correlation analyses reveal that relative rankings for provinces were consistent in all surveys. Logistic regression analyses showed a low risk of arthritis for Quebecois: odds ratio 0.75 (95% confidence interval 0.65, 0.87) after controlling for potential confounding factors. CONCLUSION: The low prevalence of arthritis observed in Quebec cannot be explained by potential confounding factors included in the NPHS and warrants further epidemiological studies.
