Alcohol consumption is prevalent around the world and is increasing in many middle-income countries including China. While the harmful effects of heavy drinking for certain major diseases (e.g., liver cirrhosis, stroke and several types of cancer) are well established, the broader impact of alcohol use on a wide range of diseases remains largely unexplored. Moreover, the causal relationships of alcohol, particularly moderate consumption (i.e., one to two drinks per day), with disease risks (such as ischemic heart disease [IHD]) are not well understood.
In our recent research, we conducted a comprehensive investigation into the relationships between alcohol use and risks of over 200 different diseases in Chinese adults.
What did we do?
We analysed data from the China Kadoorie Biobank (CKB), which followed over 512,000 adults recruited during 2004-08 from ten urban and rural areas across China.
Since alcohol consumption patterns differed hugely between men and women (in CKB one-third of men vs. only 2% of women drank regularly, i.e., at least once a week), we focused our analyses primarily on men. We examined the associations of self-reported drinking patterns with the risks of 207 different diseases identified through linkage to electronic hospital records over a 12-year period.
What did we find?
Our findings revealed that alcohol consumption was associated with increased risks of 61 diseases in men. Among these, 28 diseases were already established by the World Health Organization (WHO) to be alcohol-related (such as liver cirrhosis, stroke, and several gastrointestinal cancers).
Importantly, we also found associations between alcohol and 33 diseases not previously established by the WHO to be alcohol-related, including gout, cataract, certain fractures, and gastric ulcer.
Furthermore, certain drinking patterns, such as drinking daily, drinking in heavy “binge” episodes, or drinking outside mealtimes, increased the risks of certain diseases, particularly liver cirrhosis.
How did we assess causality?
Establishing a causal relationship between alcohol consumption and disease (e.g., to confirm that alcohol causes the disease, rather than being correlated with the disease because of other lifestyle or health factors) is challenging in traditional epidemiology.
To address this, we conducted a genetic analysis using genetic variants (ALDH2-rs671, ADH1B-rs1229984) that are common in East Asians. These genetic variants greatly reduce alcohol tolerability and strongly determine alcohol consumption levels (by causing unpleasant reactions upon drinking, i.e., Asian flushing response). As these genetic variants are unrelated to other lifestyle factors (e.g., smoking or socioeconomic status), they can be used to assess the potential causal relevance of alcohol for disease risks through a method called Mendelian randomisation.
What did our genetic findings show?
Our genetic findings supported a dose-dependent causal effect of increasing alcohol intake on the identified alcohol-related diseases as a whole. Among men, for every four drinks per day, there was a 14% higher risk of established alcohol-related diseases and 6% higher risk of diseases not previously established to be alcohol-related.
Some diseases, such as liver cirrhosis and gout, exhibited an even stronger dose-response effect (risk doubling for every four drinks per day). Additionally, stroke showed a 38% higher risk.
While there was no increased risk of IHD with higher alcohol intake, there was also no evidence for any causal protective effects of moderate drinking against IHD. Our findings for IHD and other diseases do not support the belief that there are health benefits of moderate drinking.
Caution when interpreting this study
As few women drink in China (coupled with potentially more under-reporting of drinking by women than men for cultural reasons), we were unable to assess the effects of alcohol on disease risks in Chinese women. However, women in our study served as a useful control group in the genetic analyses, indicating that the excess disease risks in men were likely to be caused by drinking alcohol rather than by other mechanisms related to the genetic variants.
Additionally, we were unable to examine the relationships between alcohol consumption and specific diseases with low hospitalisation rates (e.g., dementia or depression).
Lastly, since our study participants predominantly consumed spirits, we could not evaluate the effects of specific alcohol types such as red wine.
Future direction and implications
We plan to conduct similar comprehensive investigations to evaluate the causal effects of alcohol use on many diseases in populations from other countries, which may have different drinking patterns, beverage types, and patterns of hospitalization. This will contribute to the growing body of causal evidence about the scale of alcohol-related harms in different populations, and will inform preventive measures in different countries.
Our study has shed light on the expanded harms of alcohol consumption, going beyond what was previously established. Our findings suggest that the global burden of alcohol use may be underestimated, and emphasize the need to strengthen alcohol control policies in China and globally.
Written by Dr Pek Kei Im, Research Fellow, University of Oxford, and Dr Iona Y Millwood, Associate Professor, University of Oxford.
All IAS Blogposts are published with the permission of the author. The views expressed are solely the author’s own and do not necessarily represent the views of the Institute of Alcohol Studies.