Provincial differences in cardiovascular disease, deaths, and common health determinants: insights from the Prospective Urban Rural Epidemiology (PURE)-China study

Background Standardized, prospective data comparing cardiovascular disease (CVD) incidence and mortality across different geographical regions in China remain limited. It is also unclear whether observed regional disparities in these outcomes are more strongly related to differences in individual-level risk factors or broader community-level health determinants. Purpose To quantify regional differences in the incidence of CVD and death across 12 provinces in China, and to investigate the relative contribution of individual-level risk factors versus community-level determinants to these disparities. Methods This analysis included 43,105 participants aged 35–70 years without prior CVD from the PURE-China cohort, enrolled between 2005 and 2009 from 115 urban and rural communities across 12 provinces: Beijing, Jiangsu, Liaoning, Shandong, Shanxi, Jiangxi, Shaanxi, Sichuan, Yunnan, Inner Mongolia, Qinghai, and Xinjiang. Age- and sex-standardized rates for major CVD events and death of the 12 provinces were calculated per 1000 person-years. Using hierarchical Cox frailty models with a random intercept for province, we examined the extent to which individual-level factors (age, sex, cardiometabolic risks, health-related behaviors) and community-level determinants (urban/rural location, community affluence, ambient air pollution, and use of CVD prevention therapies) accounted for the variance in survival times for CVD and for death among provinces. Results The mean age of participants was 50.6 years, and 58.4% were female. Over a median follow-up of 11.9 years (interquartile range (IQR): 9.5–12.5), the age- and sex-standardized incidence of major CVD varied 6.1-fold across provinces, ranging from 2.78 per 1000 person-years in Jiangxi to 17.05 per 1000 person-years in Qinghai. Cardiovascular death and heart failure rates were highest in Xinjiang, while myocardial infarction and stroke rates were highest in Qinghai. All-cause mortality varied 4.5-fold, from 2.65 per 1000 person-years in Shandong to 12.06 per 1000 person-years in Xinjiang. Regions with higher CVD rates generally exhibited higher mean INTERHEART risk scores(as a measure of overall individual-level risk-factor burden) and were characterized by more urban communities, greater air pollution, lower community affluence, and less frequent use of CVD prevention therapies. In hierarchical models, individual-level risk factors explained 27.7% of the regional variance in survival time for CVD, whereas community-level determinants did not reduce regional variance. For death, individual-level factors explained 9.0% of the regional variance, while community-level determinants explained 26.8%. Conclusions Substantial regional disparities in CVD incidence and death exist across Chinese provinces. While individual-level risk factors primarily explain regional variation in CVD events, community-level determinants play a more prominent role in explaining disparities in mortality.