OBJECTIVETo verify the optimal cut-off point for overweight and obesity in Chinese adults based on the relationship of body mass index (BMI) to all-cause mortality and incidence of coronary heart disease (CHD) and stroke from pooled data of Chinese cohorts.

METHODSPooling the data of four Chinese cohorts, to analyze the age-adjusted all-cause mortality by strata of BMI, and to repeat the analyses after excluding deaths within the first three years of follow-up and after excluding the smokers. Analyzing the age-adjusted incidence of CHD and stroke by strata of BMI. Pooling the Cox regression coefficients of BMI to incidence of CHD or stroke of different cohorts using the method of weighting by inverse of variance.

RESULTSAll together 76,227 people from 4 cohorts were enrolled, with 745,346 person-year of follow-up. All-cause mortality by strata of BMI showed a U shaped curve, even after excluding deaths from the first three years and the smokers. Age-adjusted all-cause mortality increased when BMI was lower than 18.5 and higher than 28. Incidence rates of CHD and stroke increased when the level of BMI was increasing. Results of Cox regression showed that BMI was an independent risk factor for both CHD and stroke. Each amount of 2 kg/m(2) increase in baseline BMI might cause 15.4%, 6.1% and 18.8% in incidence of CHD, stroke and ischaemic stroke. Reduction of increased BMI to the level of under 24 might prevent incidence of CHD by 11% and that of stroke by 15% for men and 22% CHD and stroke for women in the population.

CONCLUSIONAttempt to define the levels of BMI < 18.5 for underweight, 24 to 27.9 for overweight and > or = 28 for obesity seemed to be appropriate cut-off points for Chinese adults.