Drinkers outlived abstainers over 30 years — with a beer-shaped catch

Whether moderate drinking protects the heart or quietly shortens life is one of the longest-running arguments in public health, and the relationship between alcohol and heart attack, stroke and death appears to vary significantly from region to region. Long-term data from a single well-defined population can help, because it follows the same people from middle age through to old age rather than comparing snapshots. This Swedish study offered exactly that: three decades of follow-up in a workforce cohort recruited in the early 1990s.

The Coeur study, conducted between 1993 and 1995, enrolled 973 middle-aged male employees of a Swedish automotive company, drawn from a randomly selected sample of 1,144 individuals, all aged 45-50 at baseline. Participants were then followed for 30 years using national health registers. The researchers calculated the accumulated incidence of first-time stroke, first-time myocardial infarction and all-cause mortality, adjusted for cardiovascular risk factors, using time-to-event methods including Kaplan-Meier estimates and hazard ratios alongside odds ratios and correlation analysis.

Alcohol consumption showed no statistically significant association with stroke. It was, however, associated with lower odds of myocardial infarction (OR 0.57) and a slight increase in survival time (HR 0.997), with consumers holding a median survival advantage of 1.3 years over non-consumers. All beverage types showed a negative association with all-cause mortality — total alcohol OR 0.57, beer OR 0.68, wine OR 0.55. The twist was beer: despite its favourable mortality association, beer consumption was linked to higher odds of heart attack (OR 1.70).

The findings land on the more favourable side of a genuinely contested literature, in keeping with the regional variation the authors note in how alcohol relates to cardiovascular outcomes. The internal contradiction is the most interesting part: alcohol overall tracked with fewer heart attacks, while beer specifically tracked with more, even as beer drinkers still showed lower all-cause mortality. Why beverage types diverge like this is an open question, and as an observational study of one male occupational cohort, it cannot separate the drink itself from everything else that differs between drinkers and abstainers.

This is a data point, not a green light. In this cohort of working Swedish men, drinkers fared better on survival and heart-attack odds than non-drinkers over 30 years, but observational associations like these cannot show that alcohol caused the advantage — and the beer finding cuts in the opposite direction within the same dataset. It is a reason to treat sweeping claims in either direction about alcohol and heart health with scepticism, particularly claims that treat all drinks as interchangeable.