For twenty years the received wisdom was that a drink or two a day was good for your heart. Then the message flipped, and nobody explained why. If you are wondering whether cutting back means giving up something your cardiovascular system actually needed, current moderate drinking heart health research gives a usable answer: no, it does not. The protective effect was largely an artifact of how the old studies picked their comparison group. When researchers fixed that — and then re-tested the question using genetics instead of questionnaires — the benefit shrank toward zero or reversed.
What moderate drinking heart health research shows now
Three findings carry most of the weight:
- A 2023 systematic review of 107 cohort studies covering 4,838,825 people and 425,564 deaths, published in JAMA Network Open by Zhao and colleagues, found no significant mortality benefit at low intake once study design was accounted for: relative risk 0.96 (95% CI, 0.86–1.06; P = .41) below 1.3 g of ethanol per day, and 0.93 (95% CI, 0.85–1.01; P = .08) at 1.3–24.0 g/day. Risk rose clearly higher up — RR 1.35 (95% CI, 1.23–1.47) at 65 g/day or more.
- A 2022 JAMA Network Open study of 371,463 UK Biobank participants by Biddinger and colleagues used genetics rather than self-report: per 1-SD increase in genetically predicted intake, odds of hypertension rose 1.28-fold (95% CI, 1.18–1.39; P < .001) and of coronary artery disease 1.38-fold (95% CI, 1.10–1.74; P = .006).
- A 2024 International Journal of Epidemiology analysis of 278,093 UK Biobank participants by Kassaw and colleagues, using 94 genetic variants over a median 12.6 years and 20,834 deaths, found each extra 8 g/day associated with 1.27-fold higher all-cause mortality (95% CI, 1.16–1.39) and 1.30-fold higher cardiovascular mortality (95% CI, 1.10–1.53) — with no evidence of a curve at all (P for nonlinearity ≥ 0.21 for every outcome).
Why the J-curve looked so convincing: abstainer bias
The J-curve is the shape you get graphing mortality against drinking: non-drinkers sit slightly high, light drinkers dip to the bottom of the J, heavy drinkers climb steeply. That dip was the whole argument — and the problem was the reference group.
Zhao and colleagues reported that of the 107 cohorts they reviewed, 86 put former drinkers and/or occasional drinkers into the abstainer comparison group, and only 21 were free of both biases. People who quit drinking often quit because they got sick — the "sick quitter" problem. Bundle them with lifelong non-drinkers and your baseline group looks unusually unhealthy, making everyone else, light drinkers included, look protected by comparison. Light drinkers in these cohorts also tended to be wealthier, more active and more likely to see a doctor. The dip was partly a portrait of who drinks moderately, not of what moderate drinking does.
What Mendelian randomization alcohol studies changed
Mendelian randomization is the closest thing to a randomised trial that ethics allows here. Common genetic variants — notably in the alcohol-metabolising genes ADH1B and ALDH2 — nudge how much people drink over a lifetime, and they are allocated at conception, before income, exercise or illness can intervene. Compare outcomes across genotype groups and most lifestyle confounding drops out.
Run that way, the dip mostly disappears. Biddinger's team found risk rising only marginally across light intake but accelerating sharply beyond roughly 7 to 14 drinks per week — an exponential, not a J. Kassaw's team found a straight line. When people say the "J-curve alcohol debunked" story is settled, this is the evidence they mean, though "debunked" is a stronger word than the literature strictly earns.
What Canada alcohol guidance says now
Canada moved on this earlier and harder than most countries. The Canadian Centre on Substance Use and Addiction published Canada's Guidance on Alcohol and Health in January 2023, replacing the 2011 Low-Risk Alcohol Drinking Guidelines. Its final report states that the model "did not demonstrate a significant negative nor protective net effect at lower levels of alcohol use; that is, we did not observe a J-shape curve," and concludes that "alcohol should not be promoted or used as a product to improve health."
The resulting weekly continuum of risk: 0 drinks carries no risk; 1–2 is low risk; 3–6 is moderate risk; 7 or more is increasingly high risk. A Canadian standard drink is 17.05 mL (13.45 g) of pure alcohol — a 341 mL bottle of 5% beer, a 142 mL glass of 12% wine, or a 43 mL shot of 40% spirits.
For scale, a March 2025 CCSA report by Sherk and Levesque, using the 2019 Canadian Alcohol and Drug Survey (10,293 respondents) with a sales-based correction for under-reporting, estimated that Canadians aged 15+ who drink average 13.3 standard drinks per week. Thirty-four per cent fell in the increasingly high-risk zone — 41% of males versus 28% of females.
Caveats, and what cuts the other way
Reporting this honestly means saying where it is still contested:
- The J-curve is not universally dead. A 2024 Nature Communications "burden of proof" analysis by Carr and colleagues, pooling 95 cohort, 27 case-control and 4 Mendelian randomization studies, still found a J-shape in the observational data, with lowest ischemic heart disease risk near 23 g/day (RR 0.69; 95% UI, 0.48–1.01). But their pooled MR estimate was flatly null (RR 1.00; 95% UI, 0.62–1.61). Their conclusion: study type, not biology, determines the answer.
- Zhao's low-volume estimate of 0.93 had a confidence interval touching 1.01 at P = .08. That is "no demonstrated benefit," not "proven harm at one drink."
- Mendelian randomization has real weaknesses, all listed by Kassaw's team: variants can act through other pathways (pleiotropy), they capture lifetime average intake and cannot separate a nightly glass from weekend binges, and UK Biobank had roughly a 5% response rate in a largely white-British sample, so selection and generalisability both bite.
- Self-reported drinking is under-reported almost everywhere, biasing estimates in hard-to-predict directions.
What this research does not show
It does not show that any particular drink is safer, healthier or gentler on the heart. The ethanol in a can of our vodka water cocktails is the same molecule as the ethanol in a glass of wine, and it counts the same against your weekly total. Aura is a still, uncarbonated vodka water with no added sugar and no artificial sweeteners — a formulation choice about taste and comfort, not a health claim, and nothing here should be read as one. For how we build it, read why Aura.
What to do with this today
- Convert to standard drinks first. Multiply container volume in mL by the label's ABV percentage, then divide by 1,705. A 355 mL can at 5% ABV is about 1.04 Canadian standard drinks; a 473 mL tallboy at 5% is about 1.4.
- Set a weekly number, not a per-occasion feeling. The guidance is built on weekly totals because that is what the risk data track. Write yours down.
- Stop using "heart health" as a reason to drink. If cardiovascular risk is your concern, blood pressure control, movement and sleep have evidence behind them that alcohol no longer does.
- Cap the occasion at two. CCSA flags more than 2 standard drinks per occasion as raising the risk of injury and harm to others, separately from long-run disease risk.
- Buy a format you can count. Single-serve cans are easier to track than a bottle poured by eye.
Drink less than last week and count it in standard drinks — the one instruction this whole literature agrees on. When you do open one, Aura comes in Cucumber Lime, Coconut Pineapple and Mango Peach, 355 mL cans by the case of 24. Ordering direct or wholesale from us delivers to Ontario addresses only; in Alberta look for us at AGLC-licensed private retailers, and in British Columbia at BC Liquor Stores and licensed private retailers — start at where to buy. Must be of legal drinking age in your province. Please enjoy responsibly.
This article summarises published research for general interest and is not medical advice. Speak with your physician about your own cardiovascular risk.
Sources
- Zhao J, Stockwell T, Naimi T, Churchill S, Clay J, Sherk A. "Association Between Daily Alcohol Intake and Risk of All-Cause Mortality: A Systematic Review and Meta-analyses." JAMA Network Open. 2023;6(3):e236185. doi:10.1001/jamanetworkopen.2023.6185. Full text
- Biddinger KJ, Emdin CA, Haas ME, Wang M, Hindy G, Ellinor PT, Kathiresan S, Khera AV, Aragam KG. "Association of Habitual Alcohol Intake With Risk of Cardiovascular Disease." JAMA Network Open. 2022. doi:10.1001/jamanetworkopen.2022.3849. PMID: 35333364. Full text
- Kassaw NA, Zhou A, Mulugeta A, Lee SH, Burgess S, Hyppönen E. "Alcohol consumption and the risk of all-cause and cause-specific mortality — a linear and nonlinear Mendelian randomization study." International Journal of Epidemiology. 2024;53(2):dyae046. doi:10.1093/ije/dyae046. Full text
- Carr S, Bryazka D, McLaughlin SA, et al. "A burden of proof study on alcohol consumption and ischemic heart disease." Nature Communications. 2024;15:4082. doi:10.1038/s41467-024-47632-7. Full text
- Paradis C, et al. Canada's Guidance on Alcohol and Health: Final Report. Canadian Centre on Substance Use and Addiction, January 2023. ccsa.ca
- Sherk A, Levesque C. Alcohol Consumption and Adherence to Canada's Guidance on Alcohol and Health. Canadian Centre on Substance Use and Addiction, March 2025. PDF