Objectives: Epidemiological studies evaluating the association of nut consumption with coronary heart disease (CHD) risk have produced inconsistent results. The current study aimed to assess the CHD risk for the highest versus the lowest categories of nut consumption, the dose-response association of CHD for every 1 serving/week increment in nut consumption, and the heterogeneity among studies and publication bias.
Methods: Pertinent studies were identified by a search in PubMed and Web of Knowledge up to January 2014. A random-effect model was used to combine the results. The dose-response relationship was assessed by restricted cubic spline and variance-weighted least squares regression analysis. Publication bias was estimated using Egger's regression asymmetry test.
Results: Ten articles with 14 studies including 6302 CHD cases were included in this meta-analysis. Pooled results suggested that highest nut consumption amount versus lowest amount was associated significantly with a reduced risk of CHD [summary relative risk (RR)=0.681, 95% confidence interval (CI)=0.592-0.783, I(2)=62.7%], especially among USA (summary RR=0.671, 95% CI=0.591-0.761) and prospective studies (summary RR=0.660, 95% CI=0.581-0.748). A linear dose-response relationship was found, and the risk of CHD decreased by 10% for every 1 serving/week increase intake of nut (summary RR=0.90, 95% CI=0.87-0.94) No publication bias was found.
Conclusion: Our analysis suggested that higher nut consumption might have a protective effect on CHD risk, especially in the USA, which needs to be confirmed.