Adiposity, major plasma biomarkers and the risk of cardiovascular diseases and mortality in Chinese adults

Background: Uncertainty remains on the associations of adiposity with stroke types and subtypes. Moreover, the role of adiposity on cardiovascular disease (CVD) and non-CVD mortality remains to be clarified among certain population subgroups (e.g., those with diabetes). Reliable assessment of these associations is particularly important in China, where mean levels of adiposity and rates of CVD, especially stroke, differ importantly from those in Western populations. Methods: In 2004-08 the China Kadoorie Biobank recruited >512,000 adults (41% men, mean age 52.0 years) from 10 diverse areas in China, recording, during 10- years’ follow-up, 16,898 CVD deaths, 46,712 incident cases of ischaemic stroke (IS) and 10,775 intracerebral haemorrhage (ICH) events. A nested case-control study of ~12,000 incident strokes and ~6000 controls measured 18 plasma biomarkers. Multiple linear regression was used to examine cross-sectional associations of adiposity with blood pressure and plasma biomarkers. In prospective analyses, Cox regression was used to estimate adjusted (for confounders, i.e., age, sex, study area, smoking, alcohol, education and physical activity) hazard ratios (HRs) for each disease associated with body mass index (BMI) and other adiposity measures (e.g., waist circumference) and, among a subset of participants, with plasma biomarkers. The percentage of excess risk attenuated after adjustment for mediators was estimated using HRs. Results: The overall mean (SD) BMI at baseline was 23.6 (3.2) kg/m2 . Adiposity was positively associated with blood pressure and adverse plasma biomarker profiles. Each 1 SD higher BMI was associated with 5.1 mmHg higher SBP, 0.2 mmol/L higher random plasma glucose (RPG), 0.1 mmol/L higher LDL-cholesterol, 0.1 mmol/L lower HDL-cholesterol, 0.2 log mg/L higher log high-sensitivity C-reactive protein and 3.0 u/L higher alanine aminotransferase. There was a log-linear positive association of BMI with IS (HR=1.19 [95% CI 1.18-1.20] per 1 SD higher BMI, adjusted for confounders), and main IS subtypes (lacunar: 1.21 [1.19- 1.23]; non-lacunar: 1.24 [1.22-1.28]). Approximately half of the BMI-IS association was explained through baseline SBP (HR reduced from 1.19 to 1.10). Accounting for intra-individual variation in SBP suggested that about three quarters of the BMI-IS association is explained through usual SBP (reduced the HR to 1.05). Among a subset of individuals, major plasma biomarkers attenuated the HR of the association of BMI with IS to 1.04. For ICH, BMI showed no association at levels below 25.0 kg/m2 (n=4485; 0.97 [0.93-1.02] per 1 SD higher) but a positive association at levels ≥25 kg/m2 (n=2003; 1.20 [1.12-1.28]). Adjustments for baseline SBP and RPG reversed these associations, with an adjusted HR of 0.92 (0.89-0.94) per 1 SD higher BMI throughout the range examined. Further adjustment for selected plasma biomarkers among a subset of participants ii attenuated the inverse association with ICH (n=3952; 0.95 [0.92-0.99]). Although there was a positive association of BMI with CVD incidence, there was a U-shaped association with CVD mortality, with the lowest risk at BMI 22.5 to <25.0 kg/m2, similarly in those with (n=23,843) and without (n=422,871) diabetes at baseline. For other measures of adiposity, the associations with plasma biomarkers and disease risks were similar to those for BMI. Conclusions: Among relatively lean Chinese adults, adiposity was positively associated with IS and its main subtypes, mainly through its effect on blood pressure. Given blood pressure, there was a strong inverse association of adiposity with ICH, which was attenuated towards the null after allowing for plasma biomarkers. The contrasting associations of adiposity with CVD incidence and mortality indicated poor survival following disease onset among those with lower BMI, in both individuals with and without diabetes.