Abstract
A large number of observational studies have found a J-shaped relationship between alcohol intake and ischemic heart disease (IHD) risk. However, some studies suggest that the alleged cardio-protective effect may be an artifact in the way that the elevated risk for abstainers is due to self-selection on risk factors for IHD. The aim of this paper is to estimate the association between alcohol and IHD-mortality on the basis of aggregate time-series data, where the problem with selection effects is not present. In addition, we will analyze SES-specific mortality to investigate whether there is any socio-economic gradient in the relationship at issue. SES was measured by educational level. We used IHD-mortality in three educational groups as outcome. Per capita alcohol consumption was proxied by Systembolaget's alcohol sales (litres of alcohol 100% per capita 15+). Swedish quarterly data on mortality and alcohol consumption spanned the period 1991Q1-2020Q4. We applied SARIMA time-series analysis. Survey data were used to construct an indicator of heavy SES-specific episodic drinking. The estimated association between per capita consumption and IHD-mortality was positive and statistically significant in the two groups with primary and secondary education, but not in the group with postsecondary education. The association was significantly stronger the lower the educational group. Although the associations were generally stronger for males than for females, these differences were not statistically significant (P > 0.05). Our findings suggest that the detrimental impact of per capita consumption on IHD-mortality was stronger the lower the educational group.