One in Three British Adults Now Drinks at Risky Levels, Five-Year Study Finds

A sweeping analysis of more than 135,000 adults across Great Britain has delivered one of the most detailed portraits yet of the nation’s relationship with alcohol, and the numbers are stark. Roughly one in three adults drinks at levels classified as risky, translating to an estimated 16.2 million people, while around one in 65 drinks at levels indicative of possible dependence, equivalent to approximately 814,000 adults. The findings, drawn from the Alcohol Toolkit Study and published in BMC Medicine, cover the period from October 2020 to October 2025 and provide the most granular picture to date of how harmful drinking is distributed across the population in the years since the COVID-19 pandemic reshaped daily life. At a time when alcohol-specific mortality has been climbing, the study offers public health officials a precise map of where the burden lies and who is carrying it.

The research team, led by Sarah E. Jackson of University College London’s Department of Behavioural Science and Health, analysed monthly, nationally representative survey data collected across England, Wales and Scotland. Alcohol consumption was measured using the AUDIT-C, the consumption subset of the Alcohol Use Disorders Identification Test, a validated screening instrument widely used in epidemiological research. The AUDIT-C scores respondents on three questions covering drinking frequency, typical quantity, and frequency of heavy episodic drinking, yielding a total score from 0 to 12. In this study, a score of 5 or above was defined as risky drinking, a threshold that captures patterns of consumption associated with elevated long-term health risks including cancer, liver disease and cardiovascular damage. A score of 11 or above was classified as possible dependent drinking, indicating a pattern consistent with alcohol dependence, characterised by loss of control over intake and continued drinking despite harm.

The headline prevalence figures for 2024/25 are sobering in their scale. The study found that 31.8 percent of adults reported risky drinking, with a 95 percent confidence interval of 31.5 to 32.1 percent, while 1.49 percent reported possible dependent drinking, with a confidence interval of 1.42 to 1.57 percent. Because the Alcohol Toolkit Study samples adults aged 18 and over in a manner designed to reflect the national population, these percentages can be converted into population estimates with reasonable confidence: approximately 16.2 million risky drinkers and 814,000 possible dependent drinkers in Great Britain. Put another way, for every person whose drinking pattern suggests possible dependence, roughly twenty more are drinking at levels that quietly elevate their risk of future disease. The sheer size of the risky-drinking group is what makes the finding a population-level concern rather than a niche clinical one.

Perhaps the most valuable contribution of the analysis lies in its disaggregation, revealing that harmful drinking is far from evenly spread. Risky drinkers as a group were slightly younger than the adult population overall, with a mean age of 46.0 years compared with 48.6 years, and were markedly more likely to be men, at 61.9 percent versus 48.5 percent of the general adult population. This male skew is consistent with long-standing patterns in alcohol epidemiology, but the study shows the imbalance becomes dramatically steeper at the severe end of the spectrum. Among those drinking at levels indicative of possible dependence, 77.2 percent were men. The mean age of this group was 49.1 years, suggesting that dependence in this population is concentrated in middle age, a life stage when the cumulative physiological toll of heavy drinking begins to manifest as measurable organ damage.

Socioeconomic patterning emerged as another critical dimension of the findings. Possible dependent drinkers were disproportionately drawn from less advantaged occupational social grades, with 26.7 percent belonging to these categories compared with 23.3 percent of all adults. This overrepresentation, while numerically modest at the group level, compounds existing health inequalities because disadvantaged communities already bear a disproportionate burden of alcohol-attributable mortality and morbidity. The study also found that fewer possible dependent drinkers lived with children, at 20.7 percent compared with 28.3 percent of the general adult population. The authors note this may reflect the social disruption that accompanies severe alcohol problems, including relationship breakdown and family separation, though the cross-sectional nature of the survey data means the direction of any such association cannot be established from this study alone.

Geography matters too, and the regional contrasts revealed by the analysis are among its most striking elements. Risky drinking was modestly more prevalent in northern England than elsewhere, but the gradient sharpened considerably for possible dependent drinking. In Yorkshire and Humber, 2.35 percent of adults reported drinking at levels indicative of possible dependence, a figure more than two and a half times the 0.91 percent recorded in London. This north-south divide echoes broader patterns in British health statistics, where life expectancy, chronic disease prevalence and alcohol-specific deaths all worsen along a gradient running from the affluent southeast toward deindustrialised northern regions. The concentration of dependent drinking in specific regions suggests that interventions may need to be geographically targeted rather than uniformly applied, channelling resources toward the communities where the severe end of the drinking spectrum is most heavily represented.

The timing of the study gives it particular significance. Alcohol-specific mortality in Great Britain has risen since 2020, a trend widely linked to disruptions during the COVID-19 pandemic, when lockdowns, bereavement, economic insecurity and social isolation altered drinking habits in ways that appear to have persisted for some people. By anchoring its measurement window in October 2020, the Alcohol Toolkit Study captures the population in the immediate aftermath of the first pandemic wave and follows it through five full years of change. The study’s monthly cadence is a methodological strength: rather than relying on occasional snapshots, it smooths out seasonal fluctuations in drinking, such as the well-documented December peaks, and provides a stable rolling estimate of population behaviour. With 135,295 respondents, the sample size also gives the confidence intervals enough precision to distinguish between regions and demographic groups with statistical confidence.

The measurement approach deserves attention from readers weighing the findings. The AUDIT-C is a brief self-report instrument, and self-reported consumption is known to understate true intake, since people systematically underestimate both how much they drink and how often. This means the prevalence figures, already alarming, are likely conservative. The threshold of 11 for possible dependence is a screening cut-off, not a clinical diagnosis; a formal diagnosis of alcohol dependence requires a fuller assessment covering withdrawal symptoms, craving and functional impairment. The authors are careful in their language, describing the 1.49 percent figure as possible dependent drinking rather than confirmed dependence. Even so, screening instruments of this kind are designed to cast a wide net for clinical follow-up, and when applied to a population of tens of millions, the numbers of people flagged become a genuine public health signal rather than statistical noise.

The study was supported by Cancer Research UK and the Economic and Social Research Council, and its authors declare no financial links with the alcohol industry apart from a disclosed unrelated project on alcohol-free drinks funded by Alcohol Change UK, a charity of which less than 0.6 percent of 2024/25 funds came from a non-alcoholic drinks producer. The researchers conclude that the combination of high overall prevalence, roughly one in three adults drinking at risky levels, and pronounced sociodemographic and geographic inequalities underscores an urgent need for effective population-level interventions. Such measures, in the broader public health literature, include pricing policies such as minimum unit pricing, restrictions on alcohol marketing and availability, and improved access to treatment and brief interventions in primary care. What this study adds is the evidence base for targeting: knowing that dependent drinking clusters among middle-aged men, disadvantaged social grades, households without children, and specific northern regions allows policymakers to direct screening and treatment resources with far greater precision than a one-size-fits-all national campaign would permit.

For a nation whose drinking culture is woven into social life, the message of this research is difficult to dismiss. Sixteen million people drinking at risky levels is not a fringe phenomenon but a mainstream pattern, one that guarantees a steady downstream flow of liver disease, cancers, injuries and premature deaths unless the trajectory changes. The study’s five-year window demonstrates that these patterns have proven durable since the pandemic rather than transient. As alcohol-specific mortality continues to climb, the data argue that the window for prevention is not in the clinic, where dependence is already entrenched, but in the population at large, where a modest shift in the drinking distribution of 16 million people would avert far more harm than treating the 814,000 at the severe end alone. The numbers are now on the table; the question is what Britain chooses to do with them.

Subject of Research: Prevalence and sociodemographic and geographic patterns of risky and dependent alcohol drinking among adults in Great Britain from 2020 to 2025.

Article Title: Prevalence and patterns of risky and dependent drinking in Great Britain, 2020–2025

Article References: Jackson, S. E., Cox, S., Buss, V., Hamilton, T., Brown, J., & Oldham, M. (2026). Prevalence and patterns of risky and dependent drinking in Great Britain, 2020–2025. BMC Medicine. https://doi.org/10.1186/s12916-026-05210-2

Image Credits: AI Generated

DOI: 10.1186/s12916-026-05210-2

Keywords: alcohol consumption, risky drinking, alcohol dependence, AUDIT-C, Great Britain, public health, health inequalities, Alcohol Toolkit Study, alcohol-specific mortality, regional disparities, epidemiology, BMC Medicine