Background: The exact pathogenic mechanism of knee osteoarthritis (OA) is still unknown. With the exception of clinical treatment to alleviate symptoms, or total knee replacement, there is currently no effective treatment method. Consequently, an in-depth etiological and epidemiological study of knee OA can provide clues for diagnosis, treatment and scientific research, and will ultimately have a beneficial effect on public health.

Methods: A cross-sectional community study in the rural village of Gaoyou was conducted in 3428 Chinese adults (aged ≥ 40 years). Subjects completed an interviewer-administered questionnaire, evaluating knee pain and associated disability, analgesia, use of health services, past medical history, walking, income, smoking, and use of oral contraceptives, and standardized weight-bearing knee radiographs were obtained. Patient demographic characteristics and biochemical parameters were recorded.

Results: Single-factor regression analysis indicated that age, overweight, central adiposity, high low-density lipoprotein cholesterol (LDLC), high total cholesterol (TC), high triglycerides (TG), dyslipidemia, hypertension and low income were the associated factors for knee OA in females; age, high LDLC, hypertension, low income and frequent walking were the associated factors for knee OA in males. Interestingly, male heavy smokers were less likely to develop severe knee OA compared with non-smokers. Stepwise logistic regression analysis indicated that age and overweight were the associated factors for knee OA for all individuals. Although central adiposity, high LDLC, high TC, high TG, dyslipidemia, hypertension and low income appeared to be related to knee OA in females according to univariate analysis, these factors were not identified in stepwise logistic regression analysis. In addition although age, high LDLC, hypertension and frequent walking were also the associated factors for knee OA in males by stepwise logistic regression analysis, smoking as a protective factor was not identified in this analysis.

Conclusions: In this study, aging, obesity, frequent walking, low income and relevant multiple metabolic disorders were the associated factors for knee OA. Smoking might be associated with a lower prevalence of OA in male smokers according to univariate analysis. A retrospective association of smoking with OA may constitute an important etiologic clue, but further well-designed, large-scale prospective controlled trials are required to confirm these findings.