The three-question alcohol screening tool known as the AUDIT-C has long been a staple of clinical settings. The World Health Organisation originally developed it for face-to-face primary care consultations, but practitioners now use it far beyond waiting rooms and GP surgeries. A large Finnish study published in Addiction in 2026 raises a question worth sitting with: are we applying the same thresholds to people whose risk profiles are fundamentally different?

The answer, drawn from nearly 880,000 person-years of follow-up data, is yes. That matters considerably for how we understand alcohol-related harm across the lifespan.

What the AUDIT-C Actually Measures

The AUDIT-C asks three questions: how often someone drinks, how many drinks they typically have on a drinking occasion, and how often they consume six or more drinks in one sitting. Each item scores zero to four, giving a total of zero to twelve. The higher the score, the greater the apparent risk.

It is a quick tool. That is its strength and also its limitation. A single score compresses a great deal of individual variation into one number. When someone compares that number against a fixed threshold, it either flags a person as at risk or it does not. Whether that threshold suits a 30-year-old man, a 70-year-old woman or an 85-year-old taking multiple medications is a different question entirely.

A Decade of Follow-Up Data on AUDIT-C Alcohol Screening

Finnish researchers linked five nationwide population surveys, conducted between 2011 and 2017, to health registers tracking hospitalisations, deaths and prescriptions for alcohol dependence medication. They followed 95,477 adults aged 20 and over. Over the study period, 1,444 new alcohol-attributable events occurred, including 846 within the first five years.

The headline finding is clear. As AUDIT-C scores rise, the risk of a registered alcohol-attributable event rises in a roughly exponential pattern. Each additional point on the scale linked to a 49% higher hazard of such an event, after adjusting for age and sex.

What stands out further is what emerges when the data split by sex and age group.

Women and Men Show Different AUDIT-C Alcohol Screening Risk Curves

Each additional point on the AUDIT-C scale linked to a 61% increase in hazard among women, compared with 45% among men. Risk among women reached statistical significance at a score of two. Among men, meaningful elevation began at three.

The clinical literature has long pointed in this direction, recommending lower cut-off thresholds for women. These results confirm it. What makes this study stand apart is the data quality. Rather than relying on self-reported outcomes or selected clinical populations, the researchers drew on nationwide register data. That approach substantially reduces the reporting bias that tends to cloud alcohol research.

Older Adults Are Not One Group

This is where the study makes perhaps its most practically useful contribution. Adults aged 65 and over routinely appear as a single group in clinical guidelines. The data here argue that is a significant oversimplification.

Among men, the optimal AUDIT-C cut-off for predicting alcohol-attributable harm sat at six for those aged 20 to 64, fell to five for those aged 65 to 79, and dropped further to three for those aged 80 and above. Among women, the corresponding figures were five, three and two.

To put that in concrete terms: applying the current single threshold of four across all adults aged 65 and over in Finland produced a positive rate of 42% among men aged 65 to 79. The evidence suggests five is a more appropriate cut-off for that group. Meanwhile, clinicians likely missed a meaningful share of genuinely at-risk individuals aged 80 and above. Overidentification in one subgroup and underidentification in another are not minor calibration issues. They affect where attention and resources go.

Why Older Adults Show Risk at Lower AUDIT-C Scores

The researchers are careful not to overstate what their data explain. Several mechanisms are plausible. Older adults metabolise alcohol differently, carry less body water and more often take medications that interact with alcohol. At any given AUDIT-C score, they showed higher rates of alcohol-attributable events than younger adults. A high score in someone aged 80 may also partly reflect decades of accumulated drinking history rather than current intake alone, making it a marker of lifetime exposure as much as present behaviour.

Younger people with high scores, by contrast, may not yet show the chronic register-level outcomes this study captured. They face their own elevated risks, including injury and mental health consequences, but these sit outside the study’s scope.

What the Numbers Look Like in Practice

The study modelled five-year absolute probabilities of an alcohol-attributable event by age, sex and AUDIT-C score. At a score of five, that probability ranged from 0.8% among women aged 20 to 64 to 2.0% among men aged 80 and above. These are not alarming absolute figures. But the exponential shape of the risk curve means that at higher scores, risk climbs steeply. That gradient is steeper still for women and the oldest adults.

This pattern reinforces what accumulating evidence already tells us: the greatest concentration of alcohol-attributable risk sits at the higher end of the score range.

Applying AUDIT-C Alcohol Screening Results More Precisely

The study does not argue that the AUDIT-C is flawed. It argues that a single fixed threshold applied uniformly across all adults rarely serves every purpose well. In clinical screening, erring towards sensitivity and using a lower threshold makes sense. For population monitoring, too much sensitivity inflates prevalence figures and obscures genuine trends.

Applying age and sex-specific cut-offs would leave prevalence rates unchanged among adults under 65. Among men aged 65 to 79, rates would fall from 42% to 30%, while those aged 80 and above would see rates rise from 14% to 28%. For women, the picture shifts upward across both older age groups: from 16% to 26% in the 65 to 79 bracket, and from 3% to 15% among those aged 80 and above.

For those working in alcohol awareness, health promotion or prevention programmes, the core message is straightforward. Age and sex are not afterthoughts when interpreting AUDIT-C alcohol screening results. They are central to what the score actually means. The evidence now exists to act on that, and for the oldest age groups especially, doing so could make a real difference.

Source: dbrecoveryresources