rethinkTHC Search
Menu
Study breakdown

Medical cannabis dispensary openings increased young children's poison exposures by 52%, but recreational dispensaries then decreased them by 42%

ObservationalStrong evidence
The takeaway

Analysis of 36,161 pediatric cannabis exposures reported to poison centers found medical dispensary openings increased exposures in children ages 2-6 by 52.3%, but subsequent recreational dispensary openings decreased them by 42.4%.

Parents, pediatricians, policymakers designing cannabis regulations, and public health officials focused on child safety.

52% increase with medical dispensaries; 42% decrease with recreational

What the researchers found

Medical cannabis dispensary openings were associated with a 52.3% increase in cannabis exposures among children ages 2-6. However, when recreational dispensaries opened, exposures in this age group decreased by 42.4% relative to medical-only states. Children 7-11 also saw a 26.6% decrease with recreational dispensaries. No significant effects were found for adolescents or young adults.

Why it matters

The counterintuitive finding that recreational dispensaries reduced pediatric exposures compared to medical-only states suggests that maturation of cannabis markets brings better packaging, labeling, and public education that protects children.

The numbers in context

36,161 cannabis exposures ages 2-20 from 2016-2021. Ages 2-6: 96.3% unintentional exposures; 52.3% increase with medical dispensaries (CI 37.5-67.0, p<.001); 42.4% decrease with recreational (CI -62.2 to -22.6, p<.001). Ages 7-11: 82.4% unintentional; 26.6% decrease with recreational (CI -45.1 to -8.1). Ages 12-17: 79.9% intentional. Ages 18-20: 77.5% intentional.

How the study worked

Difference-in-difference analysis of 36,161 cannabis-related exposures for individuals aged 2-20 reported to the National Poison Data System from 2016 to 2021. Effects of medical and recreational cannabis dispensary openings were estimated by age group (2-6, 7-11, 12-17, 18-20).

What this study cannot tell us

Poison center data capture only reported exposures, likely underestimating true incidence. Cannot account for all state-level policy differences. Recreational dispensary openings often coincide with enhanced packaging and education requirements. Data ends 2021.

How to read the evidence

Strong: nationwide poison center data with difference-in-difference design providing robust causal inference across multiple states and years.

When this study was published

Published 2026. Data from 2016 to 2021.

The bigger picture

This study challenges the straightforward narrative that cannabis legalization increases pediatric poisonings. The temporal pattern suggests an initial risk period with medical dispensaries that is mitigated as states develop regulatory infrastructure including child-resistant packaging requirements.

Questions still open

  • What specifically about recreational legalization reduces pediatric exposures? Are child-resistant packaging requirements the primary driver? Does the initial medical dispensary spike reflect increased availability or increased product diversity (edibles)?

Common questions

Does cannabis legalization increase child poisonings?
Medical dispensary openings increased young children's exposures by 52%, but subsequent recreational legalization decreased them by 42%, likely due to improved packaging and safety regulations.
What age group is most at risk for accidental cannabis exposure?
Children ages 2-6 accounted for the most unintentional exposures (96.3% were accidental), while adolescent exposures were predominantly intentional.

Read the original research

Cannabis and pediatric cannabis exposure - evidence from America's Poison Centers.

Journal of child psychology and psychiatry, and allied disciplines, 67(3), 400-412

Citation

Steuart, Shelby R; Bethel, Victoria; Bradford, W David. (2026). Cannabis and pediatric cannabis exposure - evidence from America's Poison Centers.. Journal of child psychology and psychiatry, and allied disciplines, 67(3), 400-412. https://doi.org/10.1111/jcpp.70058

Explore the wider topic