Alcohol unlikely to cause prostate cancer, but may shape how it progresses

Prostate cancer is the most common cancer in men in developed countries, which makes it an obvious target for risk reduction. Yet whether alcohol raises the risk of getting it, or of dying from it, has never been settled. Part of the problem is that studies asking men about their drinking are easily muddled: drinkers tend to differ from non-drinkers in many other ways, and men who feel unwell often change their habits. This study tried a different route, using inherited genetic differences in how the body processes alcohol, which are fixed at birth and cannot be caused by lifestyle or illness.

Researchers analysed data from 25 studies within the international PRACTICAL Consortium, covering 23,868 men diagnosed with prostate cancer and 23,051 men without it. They examined 68 genetic variants across 8 alcohol-metabolising genes, the alcohol dehydrogenase and aldehyde dehydrogenase families, which govern how quickly alcohol and its by-products are broken down. Each variant was tested for links with prostate cancer diagnosis and with death from prostate cancer, separated by tumour grade. Results from the 25 studies were then pooled to give an overall picture.

For getting prostate cancer in the first place, there was little evidence that any of the alcohol-metabolising variants made a difference. The story changed for deaths from the disease: four variants in two genes passed the study's strict statistical threshold. Three variants in the ALDH1A2 gene were linked with lower prostate cancer mortality, with hazard ratios of 0.78 and 0.76, though these three travel together genetically and are effectively one signal. One variant in ALDH1B1 was linked with higher mortality in men with low-grade prostate cancer, with a hazard ratio of 1.43.

Observational studies of alcohol and prostate cancer have produced unclear answers, which the authors attribute partly to the methodological limits of that kind of research. This genetic approach sidesteps some of those limits, and its conclusion splits the question in two: alcohol looks unlikely to affect whether prostate cancer develops, but may influence what happens after it does. That distinction is new and useful, though open questions remain. The mortality findings rest on a small number of variants, and gene variants are a proxy for drinking rather than a direct measure of it, so replication is needed before the progression link is firm.

If you have wondered whether drinking raises your odds of developing prostate cancer, this analysis is a reason to think it probably does not, and it comes from a design that is harder to fool than surveys of drinking habits. The suggestion that alcohol might influence the course of the disease once diagnosed is more tentative, resting on a few gene variants rather than measured drinking, so it is a thread for researchers to pull rather than a settled fact. Neither finding says anything about alcohol's well-known effects elsewhere in the body, so this is one piece of a much larger picture, not a verdict on drinking overall.