A large new review of the evidence has concluded that consuming the equivalent of two standard drinks a day substantially increases the risk of premature death, disability and several forms of cancer compared with drinking little or no alcohol.
What the research examined
Researchers reviewed more than 7 200 scientific publications and combined disease‑specific risk estimates with national health data sets to model how lifetime alcohol consumption affects the likelihood of developing serious illness or dying early. The analysis, published in the Journal of Studies on Alcohol and Drugs, focused on how different average weekly intakes map to health outcomes across a lifetime.
The team assessed connections between alcohol and a wide range of immediate and long‑term harms, including:
- oesophageal, oral and breast cancers;
- cardiovascular disease;
- liver disease;
- injuries from accidents and violence.
Key findings
The central result reported by the authors is that people who consume an average of 14 drinks per week (about two drinks a day) have an elevated lifetime risk: roughly 1 in 25 such drinkers will die from causes attributable to alcohol. By contrast, average consumption of up to seven drinks per week was associated with only smaller increases in risk for most conditions examined.
| Average weekly intake | Reported risk |
|---|---|
| 0–7 drinks | Small increases for most conditions |
| ~14 drinks | Substantially higher lifetime risk; ~1 in 25 deaths alcohol‑related |
“Even low levels of alcohol use come with health risks ... and that risk continues to increase the more someone drinks,” said Kevin Shield, senior author of the study.
Co‑author Timothy Naimi described the study as providing a quantitative framework across the drinking spectrum, noting that while dietary guidance in some countries uses a “less‑is‑best” message, it often lacks specific numerical thresholds.
What this means for South Africa
Although the analysis used US data sets, its methods and conclusions are relevant for South Africa where alcohol consumption contributes substantially to the burden of disease, including trauma, HIV‑related harms, non‑communicable diseases and interpersonal violence. The study reinforces that risk rises progressively with consumption rather than there being a clearly safe ‘‘moderate’’ level for everyone.
Health practitioners, policymakers and public‑health communicators should note three practical implications:
- Clinical advice should reflect that even what many people consider moderate drinking carries measurable risks.
- Harm‑reduction strategies and brief interventions in primary care remain important tools to reduce population harm.
- Policy measures that reduce overall population alcohol intake — such as pricing, restricted hours, and limits on availability — have potential to lower alcohol‑related mortality and morbidity.
Limitations and caution
The authors combined many studies with varying designs and populations and modelled lifetime outcomes using national data; findings therefore reflect population‑level risks rather than precise predictions for any individual. Differences in drinking patterns, unit sizes and co‑existing health conditions mean personal risk can vary. The paper does not suggest an absolute safe level of alcohol for all people and highlights that risk accumulates with higher consumption.
For clinicians and members of the public seeking practical advice, caution and individualised counselling remain important. People who are pregnant, have a history of substance dependence, liver disease, certain cancers or are taking medications that interact with alcohol should be advised to avoid drinking entirely.
Given South Africa’s already high burden of alcohol‑related harm, the study adds weight to calls for clearer national guidance and stronger preventive measures. Communicators should combine the study’s quantitative findings with local data and context when advising the public.
Policy responses, health‑service planning and patient counselling should treat even modest drinking as a modifiable risk factor, limited through both individual behaviour change and population‑level interventions.
