Swedish Data From 250,000 Teenagers Shows Cannabis Use Shifts as a Whole Population, Not Just Among Heavy Users

Public health approaches to alcohol have been shaped for decades by a statistical observation: when average drinking in a population rises, heavy drinking rises too. The relationship is tight enough that reducing average consumption across the whole population - not just targeting problem drinkers - is considered the most effective prevention strategy. Now a study drawing on 33 years of Swedish school surveys asks whether the same principle applies to cannabis.

The answer, according to researchers from Stockholm University and the Karolinska Institutet, is yes. Published in the journal Addiction, the analysis of more than 250,000 students aged 15 to 18 finds that cannabis use among Swedish adolescents follows a stable distributional pattern: when average use goes up or down, all user groups shift in the same direction and in proportion to each other.

The data and what was measured

The study draws on national school surveys conducted by the Swedish Council for Information on Alcohol and Other Drugs (CAN), covering grade 9 students and second-year upper secondary school students from 1990 to 2023. The sample spans a period that includes substantial swings in adolescent cannabis use in Sweden - rises during the 1990s, declines in the 2000s, and renewed increases in more recent years.

The researchers did not simply track whether students had ever used cannabis; they examined the frequency distribution of use across time, looking at how the spread from non-users through occasional users through frequent users changed as average consumption fluctuated. Their central finding: the shape of that distribution remained highly stable. When average use increased, it increased across all user groups proportionally, not by expanding a small tail of heavy users while the rest of the population remained unchanged.

"Increases in average use are not driven solely by a small group of heavy users, but by broader changes in behavior among users in general," said co-author Thor Norstrom, Professor Emeritus at the Swedish Institute for Social Research at Stockholm University.

The consequence: higher average use means more at-risk young people

The proportional shift has an important public health implication. Periods of higher average cannabis use in the Swedish school data coincided with marked increases in the proportion of adolescents who used cannabis very frequently - the group most likely to experience cannabis-related problems including dependence, academic impairment, and mental health effects.

This is not a trivial finding. A common assumption in drug policy discussions is that social norms primarily influence light or experimental use, while heavy or problematic use is driven by individual vulnerability factors that policies cannot reach. If that were true, broad-population prevention efforts would have limited effect on the users who matter most. The Swedish data suggest otherwise: when norms, availability, and attitudes shift across a society, the effects appear across the full spectrum of users.

"Our results suggest that adolescent cannabis use is characterized by collective changes, in which social networks, norms, and the broader societal climate play an important role," said co-author Hakan Leifman, researcher at the Department of Clinical Neuroscience at Karolinska Institutet.

What the total consumption model implies for prevention

The pattern the researchers describe matches what alcohol researchers call the total consumption model, developed in the 1970s by researcher Ole-Jorgen Skog and refined since. The model holds that drinking behaviors in a population are correlated: people influence each other's consumption through social interaction, and as a result, changes in average consumption produce changes across the entire distribution, not just at the extremes.

If cannabis use follows the same logic, prevention strategies that target only high-risk adolescents - those already using heavily, or those with family histories of substance problems - address only part of the problem. Policies that affect the whole population's exposure to cannabis, including pricing, legal status, marketing, and social norms, should also affect the proportion of young people who end up as heavy users.

The Swedish context and its limits

Sweden is an unusual case study for cannabis research. The drug has remained illegal throughout the study period, with relatively restrictive enforcement compared with many European countries. Cannabis use rates among Swedish adolescents are considerably lower than in countries like France or the Netherlands. Whether the same distributional patterns hold in populations with higher baseline use or in jurisdictions that have legalized cannabis for adult use is an open empirical question the study cannot answer.

The researchers note that their findings carry particular relevance given international trends toward permissive cannabis policy. Even in countries where cannabis remains illegal, changing international norms - visible through media, tourism, and peer networks - may influence adolescent attitudes. "This underscores the importance of a broad public health perspective in preventive efforts targeting cannabis use among young people," Norstrom said.

The study was funded by Stiftelsen Ansvar for Framtiden (SAFF), grant 2024-0063. No competing interests were declared.