Early adulthood BMI and cardiovascular disease: a prospective cohort study from the China Kadoorie Biobank

Background: The associations of early adulthood BMI with cardiovascular diseases have yet to be completely delineated. There is little reliable evidence about these associations among east Asian populations, that differ in fat distribution, disease patterns, and lifestyle factors from other populations. We aimed to study the associations between early adulthood BMI and cardiovascular diseases in a Chinese population, and the effect of midlife lifestyle factors on outcomes. Methods: In this prospective analysis, we used data from the China Kadoorie Biobank, a large and long-term cohort from five urban areas and five rural areas, using participants aged 35-70 years. The primary outcome was the incidence of cardiovascular diseases as a group, ischaemic heart disease, haemorrhagic stroke, and ischaemic stroke, which were obtained mainly through linkage to disease registries and the national database for health insurance claims. Early adulthood BMI was assessed through self-report at baseline survey. We used Cox proportional hazards regression models to examine the prospective associations. We also undertook multiplicative and additive interaction analyses to investigate the potential modification effect of midlife healthy lifestyle factors (a combined score covering smoking, drinking, physical activity, and diet). Findings: Participants were recruited for baseline survey between June, 2004, and July, 2008. During a median follow-up of 12<middle dot>0 years (IQR 11<middle dot>3-13<middle dot>1), we documented 57 203 (15<middle dot>9%) of incident cardiovascular diseases in 360 855 participants. After adjustment for potential confounders, monotonic dose-response associations were observed between higher early adulthood BMI and increased risks of incident cardiovascular diseases. Compared with an early adulthood BMI of 20<middle dot>5-22<middle dot>4 kg/m(2) (the reference group), the hazard ratios for a BMI of less than 18<middle dot>5 kg/m(2) was 0<middle dot>97 (95% CI 0<middle dot>94-1<middle dot>00), 18<middle dot>5-20<middle dot>4 kg/m(2) was 0<middle dot>97 (0<middle dot>95-0<middle dot>99), 22<middle dot>5-23<middle dot>9 kg/m(2) was 1<middle dot>04 (1<middle dot>02-1<middle dot>07), 24<middle dot>0-25<middle dot>9 kg/m(2) was 1<middle dot>12 (1<middle dot>09-1<middle dot>15), 26<middle dot>0-27<middle dot>9 kg/m(2) was 1<middle dot>19 (1<middle dot>14-1<middle dot>24), 28<middle dot>0-29<middle dot>9 kg/m(2) was 1<middle dot>34 (1<middle dot>25-1<middle dot>44), and >= 30<middle dot>0 kg/m(2) was 1<middle dot>58 (1<middle dot>42-1<middle dot>75). Except for haemorrhagic stroke, lower early adulthood BMI (<20<middle dot>5 kg/m(2)) was associated with decreased incident cardiovascular disease risks. No significant interaction was found between midlife healthy lifestyle factors and early adulthood BMI on cardiovascular disease risks. Interpretation: Increased risks of cardiovascular disease incidence were found among participants with high early adulthood adiposity, including ischaemic heart disease, haemorrhagic stroke, and ischaemic stroke. Our findings suggest early adulthood as an important time to focus on weight management and obesity prevention for cardiovascular health later in life.