Cannabis Use in Germany Increased Dramatically Across Both Time Periods and Birth Cohorts | Research

The takeaway

Analysis of nearly 79,000 German adults over 26 years found cannabis use odds increased roughly 9-fold from 1995 to 2021, with younger birth cohorts having dramatically higher use, and use declining steeply with age.

Epidemiologists, drug policy researchers, and anyone tracking cannabis use trends in Europe.

What the researchers found

Period effects showed cannabis use odds rising from OR 0.33 in 1995 to OR 2.96 in 2021. Cohort effects were even more dramatic, rising from OR 0.02 for those born in 1936 to OR 16.69 for those born in 2002. Age effects showed steep decline from OR 9.44 at age 18 to OR 0.13 at age 59.

Why it matters

Germany partially legalized recreational cannabis in 2024. This comprehensive pre-legalization baseline shows cannabis use was already rising dramatically through both secular trends and generational shifts, providing essential context for evaluating legalization's impact.

The numbers in context

N=78,678 across 10 survey waves (1995-2021). Age 18 OR: 9.44 (95% CI: 8.17-10.9). Age 59 OR: 0.13 (95% CI: 0.10-0.17). Period 1995 OR: 0.33, Period 2021 OR: 2.96. Cohort 1936 OR: 0.02, Cohort 2002 OR: 16.69. Cannabis users more likely to also drink alcohol and smoke tobacco.

How the study worked

Binary logistic regression using generalized additive models analyzed data from 10 waves (1995-2021) of the German Epidemiological Survey of Substance Abuse (ESA), a nationally representative household survey. N=78,678 adults aged 18-59. Age, period, and cohort effects estimated simultaneously.

What this study cannot tell us

Self-reported data subject to social desirability bias. Survey methodology changes over 26 years may affect comparability. Age-period-cohort models have inherent identification problems. German context may not generalize. Household surveys may miss heavy users and institutionalized populations.

How to read the evidence

Very large sample across 10 survey waves with sophisticated statistical modeling, though self-reported data and inherent APC model limitations apply.

When this study was published

Published 2025, analyzing 1995-2021 survey data.

The bigger picture

The steep age-related decline in use suggests most people who try cannabis do not continue long-term. But the massive cohort effect (800-fold difference between 1936 and 2002 birth cohorts) means each successive generation is starting from a higher baseline, which could eventually shift the age-prevalence curve upward.

Questions still open

Whether the steep age-related decline will persist in cohorts that grew up with legal cannabis
How Germany's post-legalization prevalence compares to this pre-legalization trend

Common questions

Does this mean most young Germans use cannabis?

The cohort effects are relative, not absolute. Having much higher odds than older cohorts does not mean most young people use cannabis, just that the proportion is much higher than it was for previous generations at the same age.

Do most people stop using cannabis as they age?

This data strongly suggests yes. The odds of past-year cannabis use dropped more than 70-fold from age 18 to age 59, indicating that the vast majority of users discontinue or dramatically reduce their use over the life course.

Read the original research

Age-period-cohort analysis of past 12-month cannabis use in Germany (1995-2021).

Addiction (Abingdon, England), 120(9), 1759-1769

View on PubMed ↗Publisher’s paper ↗

Citation

Schöllner, Natalie S; Glos, Adrian; Möckl, Justin; Krowartz, Eva-Maria; Hoch, Eva; Olderbak, Sally. (2025). Age-period-cohort analysis of past 12-month cannabis use in Germany (1995-2021).. Addiction (Abingdon, England), 120(9), 1759-1769. https://doi.org/10.1111/add.70087