Table 1 Research summary

Background

Different study designs have yielded conflicting evidence about whether low to moderate levels of alcohol consumption are associated with increased or decreased risk of ischemic heart disease (IHD), a leading cause of ill health and death worldwide. Results from cohort and case-control studies predominantly show that average consumption is associated with decreased IHD risk, although findings vary by sex, disease endpoint (morbidity versus mortality), and engagement in heavy episodic drinking (HED). Conversely, Mendelian randomization (MR) studies relying on genetic variants that predict alcohol use typically find no association or harmful relationship between alcohol and IHD.

Main findings and limitations

Cautious re-evaluation using the burden of proof meta-analytic methods – systematically applied to capture potential non-log-linear relationships, control for known sources of bias, and incorporate explained and unexplained between-study heterogeneity to generate conservative estimates of risk-outcome associations – yielded estimates of the alcohol-IHD relationship that varied by study design, consistent with previous findings. Data pooled from cohort and case-control studies showed a weak association between average levels of alcohol consumption (up to ~50 g/day) and reduced IHD risk relative to no alcohol intake, while data pooled from MR studies showed no association between genetically predicted alcohol consumption and IHD risk. A primary limitation of the analysis is that it was not possible, due to insufficient data, to differentiate the estimated alcohol-IHD relationship by alcohol use subtype – that is, by average consumption characterized by frequency and quantity, by HED, or by beverage type – and to then compare subtype-specific relationships across study designs.

Implications

Using a conservative approach to consider all evidence from cohort, case-control, and MR studies, we confirmed conflicting estimates of the relationship between alcohol use and IHD derived from self-reports of intake levels versus genetically predicted alcohol use. The discrepant findings are likely driven by biases and limitations inherent in the different study designs and highlight the need to advance methodologies to obtain more definitive answers to this critical public health question. The rapidly evolving field of MR makes it possible to apply new, sophisticated MR techniques to mitigate the effects of bias in investigations of this relationship. Long-term randomized trials can be emulated using large observational databases, avoiding some of the limitations common to conventional observational studies. New MR and trial emulation approaches should be considered as ways forward to more conclusively answer pressing questions about the potential effects of alcohol consumption on IHD.

It is anticipated that the present synthesis of evidence from cohort, case-control, and MR studies assessing the dose-response relationship between alcohol intake and IHD risk will be incorporated in upcoming iterations of the Global Burden of Diseases, Injuries, and Risk Factors Study (GBD).