Chocolate is a rich dietary source of various bioactive flavonoid compounds. Despite being one of the most popular foods worldwide, the association between chocolate consumption and long-term mortality remains unclear. The objective of this study is to determine the associations between chocolate consumption and long-term overall and cause-specific mortality, to evaluate dose–response and potential mediators, and to conduct an updated meta-analysis based on prospective cohort studies. We performed a prospective analysis in the Alpha-Tocopherol, Beta-Carotene cancer prevention (ATBC) Study with a total of 27,111 men who were recruited between 1985 and 1988 and followed through 2015. Exposure data of daily chocolate consumption was obtained from validated baseline food frequency questionnaire. Hazard ratios (HRs) and 30-year absolute risk differences (ARDs) including 95% confidence intervals (CI) for overall and cause-specific mortality were estimated using multivariable-adjusted Cox proportional hazards regression models. An updated meta-analysis of cohort studies was also conducted. During 482,807 person-years of follow-up, a total of 22,064 men died. The multivariable analyses showed a statistically significant inverse association between chocolate consumption and risk of overall mortality, with HRs of 0.91, 0.89, 0.89, and 0.88 for the increasing categories 2–5 as compared with those in the lowest category (P_trend < 0.0001, and P for nonlinearity < 0.0001). We observed significantly lower mortality from cardiovascular disease (CVD), heart disease and cancer, representing 13%, 16% and 12% risk reductions for the highest compared to lowest chocolate category, respectively (all P_trend ≤ 0.002; all P for nonlinearity < 0.0001). The inverse associations of chocolate consumption with risk of overall, CVD and heart disease mortality were generally consistent across cohort subgroups (e.g., body mass index and serum cholesterol). Mediation analysis showed that 4.3% of the inverse association of chocolate and overall mortality was mediated through reducing blood pressure. Within the updated meta-analysis of cohort studies (21 risk estimates, 908,390 participants and 65,407 events), greater consumption of chocolate (per 5 g/day) was associated with a lower risk of CVD incidence and mortality (pooled relative risk = 0.98, P value < 0.001; P for nonlinearity < 0.001). The predefined subgroup analyses generally revealed consistent inverse chocolate-CVD risk associations. In this prospective study, calorie-balanced greater consumption of chocolate was inversely associated with lower overall, CVD, heart disease and cancer mortality. The systematic review and meta-analysis provide support for the inverse chocolate-CVD association. Our findings may provide evidence to partially allay concerns regarding adverse health outcomes from low-to-moderate chocolate consumption.

巧克力是各种生物活性类黄酮化合物的丰富膳食来源。尽管巧克力是世界上最受欢迎的食品之一，但巧克力消费与长期死亡率之间的关系仍不清楚。本研究的目的是确定巧克力消费与长期总体死亡率和特定原因死亡率之间的关联，评估剂量反应和潜在介质，并根据前瞻性队列研究进行更新的荟萃分析。我们在 α-生育酚、β-胡萝卜素癌症预防 (ATBC) 研究中进行了前瞻性分析，该研究在 1985 年至 1988 年间招募了 27,111 名男性，并跟踪至 2015 年。每日巧克力摄入量的暴露数据是从经过验证的基线食物中获得的频率问卷。使用多变量调整的 Cox 比例风险回归模型估算总体死亡率和特定原因死亡率的风险比 (HR) 和 30 年绝对风险差异 (ARD)，包括 95% 置信区间 (CI)。还对队列研究进行了更新的荟萃分析。在 482,807 人年的随访中，共有 22,064 名男性死亡。多变量分析显示，巧克力消费量与总体死亡风险之间存在统计学上显着的负相关关系，与最低类别相比，增加类别 2-5 的 HR 分别为 0.91、0.89、0.89 和 0.88（P 趋势< _0.0001  ，P表示非线性 < 0.0001)。我们观察到心血管疾病 (CVD)、心脏病和癌症的死亡率显着降低，与最低巧克力类别相比，最高巧克力类别的风险分别降低了 13%、16% 和 12%（所有 P 趋势 ≤ 0.002 ；_所有 非线性P < 0.0001）。巧克力消费量与整体风险、CVD 和心脏病死亡率的负相关在队列亚组（例如体重指数和血清胆固醇）中通常是一致的。中介分析表明，巧克力与总体死亡率之间的负相关有 4.3% 是通过降低血压介导的。在最新的队列研究荟萃分析中（21 项风险估计、908,390 名参与者和 65,407 项事件），巧克力摄入量增加（每 5 克/天）与 CVD 发病率和死亡率风险降低相关（汇总相对风险 = 0.98，P值<0.001；P对于非线性 < 0.001）。预定义的亚组分析通常揭示出一致的巧克力-CVD 风险反向关联。在这项前瞻性研究中，在热量平衡的情况下，摄入更多的巧克力与降低总体心血管疾病、心脏病和癌症死亡率呈负相关。系统评价和荟萃分析为巧克力-CVD 关联提供了支持。我们的研究结果可能提供证据，部分减轻人们对低至中等巧克力消费带来的不良健康后果的担忧。