RESUMO
OBJECTIVE: Studies investigating the impact of chocolate consumption on cardiovascular disease (CVD) have reached inconsistent conclusions. As such, a quantitative assessment of the dose-response association between chocolate consumption and incident CVD has not been reported. We performed a systematic review and meta-analysis of studies assessing the risk of CVD with chocolate consumption. METHODS: PubMed and EMBASE databases were searched for articles published up to 6 June 2018. Restricted cubic splines were used to model the dose-response association. RESULTS: Fourteen publications (23 studies including 405 304 participants and 35 093 cases of CVD) were included in the meta-analysis. The summary of relative risk (RR) per 20 g/week increase in chocolate consumption was 0.982 (95% CI 0.972 to 0.992, I2=50.4%, n=18) for CVD (heart failure: 0.995 (0.981 to 1.010, I2=36.3%, n=5); total stroke: 0.956 (0.932 to 0.980, I2=25.5%, n=7); cerebral infarction: 0.952 (0.917 to 0.988, I2=0.0%, n=4); haemorrhagic stroke: 0.931 (0.871 to 0.994, I2=0.0%, n=4); myocardial infarction: 0.981 (0.964 to 0.997, I2=0.0%, n=3); coronary heart disease: 0.986 (0.973 to 0.999, n=1)). A non-linear dose-response (pnon-linearity=0.001) indicated that the most appropriate dose of chocolate consumption for reducing risk of CVD was 45 g/week (RR 0.890;95%CI 0.849 to 0.932). CONCLUSIONS: Chocolate consumption may be associated with reduced risk of CVD at <100 g/week consumption. Higher levels may negate the health benefits and induce adverse effects associated with high sugar consumption.