Combined consideration of body mass index and waist circumference identifies obesity patterns associated with risk of stroke in a Chinese prospective cohort study

Background In China, few studies have examined the relationship between the combination of body mass index and waist circumference and the risk of stroke. Moreover, the relationship may also be different in different genders. Thus, we investigated the association between the combination of body mass index and waist circumference and the risk of stroke in Chinese. Methods This prospective cohort study included 36 632 participants aged 18 to 90 years. Participants were recruited from 60 surveillance sites (25 urban sites and 35 rural sites) across China in 2010 China Chronic Disease Risk Factor Surveillance, and followed up in 2016-2017. Incident cases of stroke were identified through questionnaires (including the basis of clinical diagnosis, imaging tests, time of diagnosis, diagnosis unit) and Cardiovascular Event Report System. Risk factors for stroke were collected at baseline using questionnaire, physical measurements and laboratory tests. Cox proportional hazards regression models were used to generate adjusted hazard ratios and 95%CI. All analyses were duplicated by gender stratification. Results During 6.42 ± 0.50 years of follow-up, 1 333 (597 males, 736 females) stroke events were observed among the 27 112 participants who did not have cardiovascular diseases at baseline. Compared with the general population who have normal weight or underweight with normal WC, those who have normal weight or underweight with abdominal obesity (adjusted hazard ratios 1.45, 95%CI 1.07-1.97 in males; 0.98, 95%CI 0.78-1.24 in females), overweight with abdominal obesity (1.41, 95%CI 1.14-1.75 in males; 1.33, 95%CI 1.10-1.61 in females), obesity with abdominal obesity (1.46, 95%CI 1.11-1.91 in males; 1.46, 95%CI 1.17-1.81 in females). Overweight with normal WC was found to be not statistically significant for both males and females (all P >0.05). Subgroup analysis found a multiplicative interaction between age and anthropometric group in females ( P for interaction <0.05). Sensitivity analysis results did not change. In the subjects with CVD risk factors, we found a similar relationship as in the general population . Conclusions Combined assessment of body mass index and waist circumference identifies obesity patterns associated with stroke risk.