The effects of alcohol on health are a major topic of discussion in both scientific literature and international media, with rival claims about how beneficial or detrimental moderate drinking might be. Recently, two US reports published within weeks of each other by eminent medical experts (NASEM, 2025; Shield et al., 2025) came to polar opposite conclusions. Two years earlier, Canada’s Guidance on Alcohol and Health (CGAH; Paradis et al., 2023) characterised “low-risk drinking” as no more than two drinks per week, down from earlier guidelines of two or three drinks per day. We try to explain here some of the underlying methodological differences between these influential reports and argue that all three have, to different degrees, underestimated the net impacts of alcohol on health (see also Stockwell, Priore and Im, 2025a and 2025b).
The US Academies of Science, Engineering and Medicine (NASEM, 2025) concluded that while moderate drinking was associated with a 10% increased risk of breast cancer, it was also associated with a 22% reduction in risk of heart disease and of death from all causes by 16%. It also concluded there was insufficient evidence on other cancers at lower levels of consumption. “Moderate drinking” was defined as no more than one US standard drink of alcohol (=14g ethanol) per day for women and two for men (Dietary Guidelines Advisory Committee, 2020).
Soon after, the US Department of Health and Human Services (Shield et al., 2025) released an analysis of the current US Dietary Guidelines on Alcohol (Dietary Guidelines Advisory Committee, 2020) by an independent group titled the Interagency Coordinating Committee on the Prevention of Underage Drinking (ICCPUD). Like NASEM, they concluded that low-volume alcohol use increased the risk of breast cancer. However, ICCPUD also identified low-volume alcohol as a cause of multiple cancers of the digestive system. They additionally concluded that drinking within the current US guidelines still significantly increased risks of death, injury and illness.
Comparing the three expert reports
What led to these discrepant conclusions? The short answer is that NASEM decided to review only a highly restricted range of recent articles. Critically, in relation to cancer, NASEM only considered 23 articles in total while both the ICCPUD and CGAH reviews each included several hundred (see Table 1). NASEM’s stated reason for excluding hundreds of cancer articles was a little perverse. It was to exclude studies that may have included “sick quitters” in their abstainer comparison group, lest this gave the false appearance that low-volume alcohol use protected against cancer. (“Sick quitters” refers to people who quit drinking due to health reasons, and their inclusion in studies can lead to conclusions that moderate drinking is better for health than abstaining, due to the poorer health of these sick quitters.) However, nearly all studies report increased, not reduced, cancer risk from low-volume use regardless of whether they suffered from a sick quitter bias, so this drastic step was not strictly necessary. Sick quitter bias may, however, still lead to under-estimation of level of cancer risk. For example, in a meta-analysis of alcohol and prostate cancer, Zhao et al. (2016) found that after excluding studies with sick quitter bias, low-volume alcohol was associated with a 23% increase in prostate cancer risk, up from 8% when such studies were included.
Another key difference was that while ICCPUD and CGAH focused on causes of death known to be alcohol-caused (e.g., liver cirrhosis, alcohol poisoning), NASEM considered deaths from all causes. The problem here is that only about 5% of deaths in North America are wholly or partially alcohol-attributable (Griswold et al., 2018). This opens the door for considerable bias and confounding from uncontrolled lifestyle, health, and socioeconomic factors from the other 95% of non–alcohol-attributable deaths. For example, decades of research show that “moderate drinkers” are systematically different from abstainers in ways unrelated to their alcohol use e.g. they tend to have higher incomes and better access to healthcare (Naimi et al, 2017). This means that their moderate alcohol use can be a sign rather than a cause of good health. The previous (and substantially higher) Canadian low risk drinking guidelines had relied on all-cause mortality studies (Stockwell et al, 2012).
The NASEM report also downplayed critical new evidence from randomised controlled trials (RCTs) and Mendelian Randomization (MR) studies (e.g., Van de Luitgaarden et al., 2022; Millwood et al, 2023) which do not support health benefits from low-volume alcohol. While we need more RCTs and MR studies to allow for triangulation of evidence to finally settle this continuing controversy, we do note one RCT has already been published (Voskoboinik et al., 2020) showing that patients with history of a cardiovascular event (atrial fibrillation) halved their relapse rate if randomised to abstinence rather than continued drinking. Furthermore, RCT studies suggest that providing low-volume alcohol significantly increases blood pressure (Roerecke et al., 2017), may worsen arterial function (Del Giorno et al., 2022) and has no effect on blood glucose control (Gepner et al., 2015). These are all key risk factors for cardiovascular disease.
The critical role of ischaemic heart disease (IHD) in estimates of alcohol’s disease burden
The 2020 Global Burden of Disease (GBD) group (Bryazka et al., 2022) estimated 1.8 million deaths attributable to alcohol globally, down from 2.8 million in 2016 (Griswold et al., 2018). One reason for this sudden loss of a million deaths was the 2020 assumption that up to 9 drinks/day protects against IHD (the most prevalent cardiovascular condition) whereas the 2016 GBD Group assumed this up to 5 drinks/day. Canada’s 2023 Guidance identified a meta-analysis by Zhao et al. (2017) as the best available for estimating alcohols impact on IHD. That review estimated nonsignificant reductions in IHD at up to only one drink per day using a model recommended by the authors as having best dealt with systematic bias. The ICCPUD group also used the Zhao et al. (2017) meta-analysis for their estimates. However, ICCPUD chose a different model from this paper (not the one recommended by the authors), resulting in their estimating “low risk alcohol use” to be up to 6 drinks per week i.e., three times higher than in the recent Canadian exercise.
Understandably, most commentators and media reports on alcohol dwell on the headline conclusions from these expert reports. Diving into the weeds of the specific methods used in each exercise reveals how different assumptions about how alcohol impacts risk of IHD, globally a leading cause of death, can swing estimates up, down or sideways. Furthermore, there are several conditions potentially caused by alcohol that are not included in these estimation exercises such as prostate cancer (Zhao et al., 2016) and gout (Im et al, 2023). Some potentially alcohol-caused conditions (e.g., dementia, pancreatic cancer) have been excluded because of the spurious appearance of protective effects from alcohol at low doses (e.g. Lu et al., 2017). Even for conditions with well-established causal associations with alcohol (e.g., cancers of digestive system and breast) levels of risk are underestimated across the board because about 80% of mortality studies do not account for sick quitters and related biases (Zhao et al., 2023). Not even the Canadian Guidance made allowance for these downward biases in their estimates of alcohol’s risk to health (Paradis et al., 2023).
Conclusions
The current evidence on alcohol and health suggests strong grounds for both scepticism about health benefits from low-volume alcohol use, and also for supposing alcohol’s contribution to the burden of disease and injury has been systematically underestimated.
Written by Professor Tim Stockwell and Isabella Priore, Canadian Institute for Substance Use Research, and Dr Pek Kei Im, Nuffield Department of Population Health, 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.
Funding
This work was funded by the Canadian Institute for Substance Use Research (CISUR), University of Victoria, BC, Canada. Pek Kei Im is funded by a Wellcome Career Development Award (302990/Z/23/Z). CISUR had no role in the preparation, review, or approval of the manuscript or the decision to submit the manuscript for publication.
References
Bryazka, D., Reitsma, M. B., Griswold, M. G., et al. (2022). Population-level risks of alcohol consumption by amount, geography, age, sex, and year: a systematic analysis for the Global Burden of Disease Study 2020. Lancet, 400(10347), 185-235. https://doi.org/10.1016/s0140-6736(22)00847-9
Del Giorno, R., Maddalena, A., Bassetti, S., & Gabutti, L. (2022). Association between alcohol intake and arterial stiffness in healthy adults: A systematic review. Nutrients, 14(6), 1207. https://doi.org/10.3390/nu14061207
Dietary Guidelines Advisory Committee. (2020). Scientific Report of the 2020 Dietary Guidelines Advisory Committee: Advisory Report to the Secretary of Agriculture and Secretary of Health and Human Services. https://doi.org/10.52570/dgac2020
Gepner, Y., Golan, R., Harman-Boehm, et al. (2015). Effects of initiating moderate alcohol intake on cardiometabolic risk in adults with type 2 diabetes: A 2-year randomized, controlled trial. Annals of Internal Medicine, 163(8), 569–579. https://doi.org/10.7326/M14-1650.
Griswold, M., Fullman, N., Hawley, C., et al. (2018). Alcohol use and burden for 195 countries and territories, 1990–2016: A systematic analysis for the Global Burden of Disease Study 2016. The Lancet, 392(10152), 1015–1035. https://doi.org/10.1016/s0140-6736(18)31310-2
Im, P.K., Wright, N., Yang, L. et al. Alcohol consumption and risks of more than 200 diseases in Chinese men. Nature Medicine 29, 1476–1486 (2023). https://doi.org/10.1038/s41591-023-02383-8
Lu, P.Y., Shu, L., Shen, S.S., et al. (2017). Dietary Patterns and Pancreatic Cancer Risk: A Meta-Analysis. Nutrients, 9(1).
Millwood, I.Y., Im, P.K., Bennett, D., et al. (2023). Alcohol intake and cause-specific mortality: conventional and genetic evidence in a prospective cohort study of 512 000 adults in China. Lancet Public Health, 8(12):e956- e967. doi: 10.1016/S2468-2667(23)00217-7.
Naimi, T. S., Stockwell, T., Zhao, J., Xuan, Z., Dangardt, F., Saitz, R., Liang, W., & Chikritzhs, T. (2017). Selection biases in observational studies affect associations between ‘moderate’ alcohol consumption and mortality. Addiction, 112(2), 207-214. https://doi.org/10.1111/add.13451
National Academies of Sciences, Engineering, and Medicine. (2025). Review of evidence on alcohol and health. National Academies Press. https://doi.org/10.17226/28582
Paradis, C., Butt, P., Shield, K., et al. (2023). Canada’s guidance on alcohol and health: Final report. https://ccsa.ca/sites/default/files/2023-01/CCSA_Canadas_Guidance_on_Alcohol_and_Health_Final_Report_en.pdf
Roerecke, M., Kaczorowski, J., Tobe, S. W., et al. (2017). The effect of a reduction in alcohol consumption on blood pressure: A systematic review and meta-analysis. The Lancet Public Health, 2(2), e108–e120. https://doi.org/10.1016/S2468-2667(17)30003-8
Shield, K., Keyes, K. M., Martinez, P., et al. (2025). Draft report: Scientific findings of the Alcohol Intake & Health Study for public comment. https://www.stopalcoholabuse.gov/media/pdf/Report-on-Alcohol-Intake-and-Health.pdf
Stockwell, T., Priore, I. and Im, P. (2025a). The US Academies of Science, Engineering and Medicine were economical with the truth about alcohol and health. Journal of Studies on Alcohol and Drugs, 86(4), 651–656 https://doi.org/10.15288/jsad.25-00075
Stockwell, T., Priore, I. and Im, P. (2025b). Fast Takes: Getting whiplash? A short guide to recent conflicting reports on alcohol and health. Journal of Studies on Alcohol and Drugs. https://www.jsad.com/do/10.15288/jsad-FT.06.30.2025-41/full/
Van De Luitgaarden, I. A. T., Van Oort, S., Bouman, E. J., et al. (2022). Alcohol consumption in relation to cardiovascular diseases and mortality: A systematic review of Mendelian randomization studies. European Journal of Epidemiology, 37(7), 655–669. https://doi.org/10.1007/s10654-021-00799-5
Voskoboinik, A., Kalman, J. M., De Silva, A., Nicholls, T., Costello, B., Nanayakkara, S., Prabhu, S., Stub, D., Azzopardi, S., Vizi, D., Wong, G., Nalliah, C., Sugumar, H., Wong, M., Kotschet, E., Kaye, D., Taylor, A. J., & Kistler, P. M. (2020). Alcohol abstinence in drinkers with atrial fibrillation. New England Journal of Medicine, 382(1), 20–28.
Stockwell, T. & Butt P, Beirness D, Gliksman L, & Paradis C. (2012). The basis for Canada’s new low-risk drinking guidelines: A relative risk approach to estimating hazardous levels and patterns of alcohol use. Drug and Alcohol Review. 31(2), 126-134. Article first published online: 29 SEP http://onlinelibrary.wiley.com/doi/10.1111/j.1465-3362.2011.00342.x/epdf
Zhao J, Stockwell T, Naimi T, Churchill S, Clay J, Sherk A. (2023). Association Between Daily Alcohol Intake and Risk of All-Cause Mortality: A Systematic Review and Meta-analyses. JAMA Netw Open;6(3):e236185. doi:10.1001/jamanetworkopen.2023.6185
Zhao, J., Stockwell, T., Roemer, A., & Chikritzhs, T. (2016). Is alcohol consumption a risk factor for prostate cancer? A systematic review and meta-analysis. BMC Cancer, 16(1), 845–845. https://doi.org/10.1186/s12885-016-2891-z
Zhao, J., Stockwell, T., Roemer, A., Naimi, T., & Chikritzhs, T. (2017). Alcohol consumption and mortality from coronary heart disease: An updated meta-analysis of cohort studies. Journal of Studies on Alcohol and Drugs, 78(3), 375–386. doi:10.15288/jsad.2017.78.375
Table 1: Risk estimates for moderate drinking and mortality outcomes provided by NASEM (2025) and other published systematic reviewsd
| US National Academies Reviews | Canada’s Guidance and other Reviews | ||||||
| Outcome | N studies | Alcohol level/day | RR (95% CIs) | Source | N studies | Alcohol level/day | RR (95% CIs) |
| All-cause Mortality | 8 | 1-14g/day (women) 1-28g/day (men) | 0.84, 0.81–0.87 | Zhao et al., 2023* Stockwell et al., 2024* | 107 6 | Up to 24g/day “ | 0.93 (0.85-1.01) 1.10 (0.97-1.24) |
| Cancers: | “ | ||||||
| Breast | 4 | “ | 1.10 (1.02-1.19) | Sun et al. (2020) | 22 | Each 20g/day increase | 1.20 (1.17-1.42) |
| Colorectal | 3 | “ | 1.09 (0.98-1.23) | Vieira et al. (2017) | 11 | Each 10g/day increase | 1.08 (1.07-1.11) |
| Esophagus | 2 | “ | Not assessed | Bagnardi et al. (2015) | 54 | Each 20g/day Increase | 1.30 (1.28−1.32) |
| Larynx | <4 | “ | Not assessed | Bagnardi et al. (2015) | 41 | Each 20g/day Increase | 1.33 (1.24−1.42) |
| Mouth | <4 | “ | Not assessed | Bagnardi et al. (2015) (mouth +pharynx) | 52 | Each 20g/day Increase | 1.61 (1.52−1.72) |
| Pharynx | <4 | “ | Not assessed | ||||
| Prostate | <4 | “ | Not assessed | Bagnardi et al. (2015) Zhao et al. (2016) ** | 43 27 | >12.5<50g/day Up to 24g/day | 1.06 (1.01−1.11) 1.08 (1.04-1.11) |
| Liver | <4 | “ | Not assessed | World Cancer Research Fund (2018) | 14 | Each 10g/day increase | 1.04 (1.02-1.06) |
| Heart disease (IHD) | 2 | “ | 0.78 (0.68, 0.90) | Zhao et al. (2017) | 23 | Up to 24g/day | (0.75, 1.21) |
d Reproduced with permission of editors of the Journal of Studies on Alcohol and Drugs; *All-cause mortality studies were not used directly in the development of Canada’s Guidance on Alcohol and Health (CGAH) but these reviews were conducted alongside in order to inform the process. **CGAH did not include prostate cancer as an outcome. Bagnardi et al. (2015) did not report an equivalent level of “moderate alcohol use” so the more recent Zhao et al. (2016) estimates for this outcome are also presented.