Approved Research

Target trial emulation to assess the health effects of alcohol consumption

Harvard School of Public Health, Boston

Lay summary

Understanding the causal relationship between alcohol consumption and its impact on health is paramount due to the widespread use of alcohol globally. Alcohol is known to be a significant risk factor for various chronic diseases like heart disease, cancer, liver disease, and premature death. However, the precise nature of these relationships remains uncertain, partly due to the limitations of study designs employed so far.

Our research project seeks to address this critical gap by using data from the UK Biobank - one of the world's largest and most comprehensive prospective cohort studies - to emulate a hypothetical randomized trial of the long-term health effects of alcohol. We aim to assess how the risk of disease changes when 'social drinkers', which make up the majority of alcohol users globally, start to drink in moderation, heavily, only monthly, or stop drinking by using a sophisticated study design framework and statistical methods. The emulation will be conducted according to a pre-specified protocol with eligibility criteria, treatment strategies and allocation, follow-up, outcomes, causal contrasts of interest, and statistical analysis.

The project is expected to be completed within 36 months and the results are expected to be published in a high-impact journal.

Given the high prevalence of alcohol consumption worldwide, it is critical to know what diseases can be caused by alcohol consumption to inform action at the population level and also individual behavior. To date, numerous country recommendations and guidelines label low to moderate levels of alcohol consumption as "safe," although these are based on quantified associations from observational studies that are potentially biased and therefore unreliable. This research project aims to provide more clarity, as far as the underlying data allow, on the question of which diseases are related to alcohol and whether low-to-moderate consumption is really 'safe'.