1st BATCH OF FILES AVAILABLE NOW!

Last week, the NY Times and slews of other news outlets focused on the rewritten and rereleased Alcohol Intake & Health (AIH) study that originated at ICCPUD. At the same time, most all but ignored a trio of other recently published papers that asked essentially the same questions. And notably, not one of those papers appeared with the same dire warnings about health harms at even "low levels" of consumption. Two of the three studies confirmed that light-to-moderate drinking can benefit heart health but raise cancer risks. Yet a third upended that argument, finding that only heavy drinking raises cancer mortality risk. Taken together, they raise questions about both the press' fixation on the ICCPUD study and alcohol research writ large, undercutting the confidence some advocates use to argue for sweeping policy changes.Two Separate Sizable Studies Confirm Cardioprotective Effects Like the authors tapped by ICCPUD to conduct and write the AIH study, many alcohol-focused members of the public health community downplay the idea that a little bit of drinking can benefit heart health. But two new reviews didn't side step that persistent finding. And both were major undertakings, each part of long-standing efforts to understand alcohol's impact on health.One, published early this month by Nature Health, was a large re-analysis of existing research on the relationship between alcohol consumption and 20 separate chronic health outcomes. Conducted by the same team charged with producing the Global Burden of Disease project (more on that below), it broadly confirmed existing understandings after analyzing well over 800 studies dating back to the 1960s. Among them, they "found J- or U-shaped relationships" for a group of mostly cardiovascular conditions, suggesting protective effects. "Our study has shown that low-to-moderate alcohol consumption is associated with a lower risk of cardiovascular diseases," authors wrote. Ref 1Another new study, published in Addiction last month, actually uses data from ICCPUD's AIH study and shares 2 of the 6 AIH authors, Drs Kevin Shield and J&uuml;rgen Rehm. But this paper reframes some of the findings, stating more plainly that "relationships are J-shaped" for ischemic heart disease & stroke as well as type 2 diabetes. "This J-shape is observed in all major meta-analyses of cohort studies, but the lower risk at low consumption levels becomes less pronounced as more potential biases are controlled," authors cautioned, reminding that "heavy episodic" (binge) drinking appears to offset the benefits. Ref 2Recall, the AIH did not incorporate the findings of genetic analyses (or Mendelian randomization, MR studies), some of which contradict the usual "J." So this paper attempted to bridge that gap, examining MR studies of cardiovascular outcomes. Results varied widely and, ultimately, researchers determined that "the MR evidence base is too weak to rule out the cardioprotective effect at low levels of consumption and the harmful effect at high levels established in cohort studies and supported by biological pathways."That said, this paper's use of the AIH data for underlying risk curves subjects it to the same criticisms of replicability and bias. Though, given the connection, the author's clear acknowledgment of the J-curve stands in stark contrast to the rhetoric of AIH authors on the topic. These authors also make one more notable admission, writing that "most fully alcohol-attributable diseases require heavy drinking, either occasionally (e.g. alcohol poisoning) or chronically." Said differently: in isolation, alcohol causes harm almost exclusively when consumed in large quantities. At lower levels of consumption, it can only act in tandem with other risk factors and "may affect" other negative health outcomes, including major noncommunicable diseases like cancer and cardiovascular disease.Cancer Risk Confirmed Here, Questioned There The new analysis in Nature Health emphasizes cancer risk. "Even low consumption (<10 g per day) was associated with elevated risk" for 9 different cancers, authors wrote, claiming "moderate-to-strong harmful associations." Note this study's use of 10 grams of ethanol to define a drink, aligned with British standards rather than the 14 g measure more commonly used in the US. At 10 g per day, risk jumps 15% or more for 4 cancers of the digestive tract, the study found. At 20 g per day it increases by at least that much for breast cancer. It takes 30 g per day for risk to increase at least 15% for lip/mouth cancer, according to the paper.Yet another new study strongly questions the kind research reviewed for that study. Published in the Journal of General Internal Medicine last month, the analysis of data on almost 27K adults, determined that only heavy drinking carried a clear increase in risk. Defined as over 7 drinks (of about 14 g ethanol) per week for women, 14 for men, "heavy alcohol use is consistently linked with higher cancer mortality," authors wrote. Ref 3But that wasn't the case for "light," up to 3 drinks per week, or "moderate" drinking, 4-7 or 4-14 drinks for women and men, respectively. After various adjustments trying to correct for demographics, socioeconomic impacts, preexisting medical conditions, behavior and more, results confirmed that heavy drinkers were 21% more likely to die of any cancer, but light drinkers were 13% less likely. Authors found a non-significant reduction in risk among moderate drinkers, or "no associations between moderate alcohol consumption and cancer mortality." Counter to their expectation, they also saw no significant differences by sex.These researchers went about their analysis in a different way, they point out, noting that the kinds of studies used elsewhere (including in the Surgeon General's advisory on the topic in early 2025), carry "substantial limitations." Most cancer-alcohol research analyzes "public data" on average consumption and cancer incidence to make connections between the two. But these researchers looked at outcomes actually experienced by individuals. That said, authors could only analyze risk of any cancer, not cancers specifically linked to alcohol use, which "may have diluted" the risks.Implications for Guidance: Embrace Complexity Despite their finding that light drinking reduces overall cancer risk, the authors of the cancer-focused study "caution against interpreting alcohol consumption as beneficial for cancer outcomes." Instead, they believe their work "reinforces public health recommendations to limit alcohol intake." National drinking guidelines were a clear focus of the Nature Health study, which took a global view and acknowledged that guidelines "differ considerably." But such variances "reflect the complex and differential relationships between alcohol and health outcomes."The authors embrace "complexity." Guidelines must "move beyond arbitrary thresholds, and prioritize clear, evidence-based messaging that promotes population-level understanding of alcohol-related risks," they wrote. "Guidelines should emphasize reducing overall alcohol intake and discouraging heavy episodic drinking," they argue, seeing an "urgent need for clear, evidence-based communication about alcohol's carcinogenic effects."In a notable shift, these authors also discourage the use of a "universal threshold." Instead of embracing the "no safe level" rhetoric, they argue for "tailored public health guidance rather than one-size-fits-all recommendations, stressing the potential harms associated with high intake levels." That recommendation aligns rather closely with the recent update of the US Dietary Guidelines, which emphasized a simpler "drink less" message, recall.Review: Global Burden of Disease Project The team of researchers that came to those conclusions about alcohol guidance is mostly based at the University of Washington's Institute for Health Metrics and Evaluation (IHME). And that's also the homebase for the WHO-backed Global Burden of Disease study, a massive undertaking updated every few years. This paper is part of the work done for the 2023 GBD update, published late last year, based on research conducted through the end of 2023.Recall, it was the 2018 release of the GBD's alcohol-focused work that kicked into high gear the last 8 years of advocacy claiming "no safe level" of alcohol consumption. At the time, an editorial accompanying the GBD update emphasized that the "safest level of drinking is none." It was largely based on clear evidence of a cancer risk paired with questions about research confirming cardiovascular benefits of moderate consumption. The more moderate (though somewhat confounding) determinations of the 2021 update to the GBD, which found at least some safe levels of consumption, remains largely ignored as most advocates continue to hammer on the older findings.Most Alcohol & Health Research Flawed, "Weak"? Rather than embrace "no safe level" rhetoric, the IHME's new paper leans into "complexity." It kicks off by summarizing that "the relationship between alcohol and health is complex, and the evidence&hellip;is evolving," as we closed our last issue. And these researchers weren't alone. The paper in Addiction based on the data from ICCPUD's AIH also undermines the argument that the science is "settled." It agrees with two core summations of the IMHE paper: the science "continues to evolve" and much of the existing research is "prone to major limitations.""Every systematic review is only as valid as the validity of the underlying research," authors wrote for Addiction, reminding that "alcohol epidemiological research&hellip;is susceptible to risk of severe biases." And the research community should do better, they argue: "Many of our studies on alcohol consumption and health are still plagued by the same problems that were pointed out 20 years ago&mdash;even though they are, in principle, avoidable." Those include familiar issues linked to self-reporting, categorizing drinkers and controlling for confounders. "Methodological constraints impede a comprehensive and accurate assessment of alcohol's health risks," the IHME group concluded.In an attempt to address these constraints, the folks at IHME employed a novel 5-star rating system to compare the evidence base for the relationship between drinking and various health outcomes. And after careful review, most associations were rated as "weak," getting just 2 stars or fewer. The only link that got 5 stars was alcohol's impact on certain cancers of the pharynx. Researchers found no 4-star associations and a handful of "moderate," 3-star associations, including for pancreatitis and a few more cancers, plus cirrhosis and other liver diseases. Research supporting links to 9 additional health outcomes received a 2-star rating indicating "weak but consistent" links, including the increased risk of breast cancer and decreased risk of type 2 diabetes, Alzheimer's and dementia. The remaining 5 health outcomes received just 1 star, including positive associations with cardiovascular diseases.Inclusion of Injury and Harm to Others All of the above analyses focus on chronic health outcomes, but recall that the AIH study from ICCPUD controversially incorporated more acute risks as well. So does the other new paper based on that data. But even the risk of injury or overdose do not spring solely from drinking, authors acknowledge. "The risk of injury is influenced not only by the amount of alcohol consumed, but also by the person involved, the vehicle or agent causing the injury, the environment in which the injury occurs and co-use of other psychoactive substances," they wrote.The study also hints at one of the advocacy reasons for more heavily incorporating risk of injury in these types of analyses: "alcohol-attributable health harm is not restricted to the drinker," authors wrote. The paper uses the newest version of the WHO's International Statistical Classification of Diseases and Related Health Problems (ICD-11), they explain, noting that the document "now includes harm to the health of others" as the basis for a diagnosis of "harmful use of alcohol." In other words: the ICD-11 establishes a way to track the diagnosis of an individual who experiences harm from someone else's drinking. Time will tell if practitioners actually use it, these authors remind, acknowledging that "its validity when assessed by self-report is questionable."Tracking "harm to others" also opens the door to arguments that proved particularly successful in tobacco policy debates. That's why some advocates are so keen to group this data with other measures of chronic health harms. Yet clear differences remain between the kinds of harm noted here, driven almost entirely by heavy drinking and/or intoxication, and the health harms caused by second-hand smoke.Press Fixation Continues But most Americans didn't hear word one about any of this research. They were far more likely to see headlines about the rewrite and republication of the AIH study. And that attention came from more than just papers that have proven particularly sympathetic to the "no safe level" rhetoric endorsed by some of the loudest voices, like the NY Times, Washington Post and Reuters. Yet the science desks of such widely read (and often respected) news organizations continue to cover the topic of alcohol and health largely by listening to a relatively small number of researcher-advocates without seeking additional context, even in the form of other large, contemporaneous reviews of the research.Cheers,References1 Dai, X, Gakidou, E, et al, "Health effects associated with alcohol consumption: a Burden of Proof study," Nature Health, June 20262 Carr, S, et al, "A review of the relationship between dimensions of alcohol consumption and the burden of disease: 2026 update including Mendelian randomisation studies," Addiction, May 20263 Pinheiro, LC, et al, "Reevaluating the Alcohol-Cancer Link: Long-Term Cancer Mortality Outcomes in the REGARDS Study," Journal of General Internal Medicine, May 2026

New Analyses Confirm J-Curves, But Emphasize Complexity, Question Confidence

Publishing Info