Lovell · Experimental and clinical psychopharmacology 2020 · Systematic review and meta-analysis · n=30 studies (1,613 participants: 849 cannabis users, 764 controls)

Cognitive outcomes associated with long-term, regular, recreational cannabis use in adults: A meta-analysis.

Cited 103 times in the scientific literature.

Level 3 - non-randomized controlled study

Systematic review and meta-analysis of non-randomized observational comparative studies.

PubMed 31670548 · doi:10.1037/pha0000326 · record verified 2026-08-26

What was done

A systematic review and univariate meta-analysis evaluated cognitive functioning in adults with long-term (mean ≥2 years), regular (mean ≥4 days/week) recreational cannabis use tested during abstinence (mean ≥12 hours). The authors searched five databases through May 2019 following PRISMA guidelines. Cognitive performance was classified into six domains (attention, executive function, learning and memory, decision making, information processing, working memory) plus a global cognition measure. Moderator analyses examined duration of use, age of onset, and prolonged abstinence (≥25 days).

What was found

Across 30 studies (849 cannabis users, mean age 30.7 years; 764 controls, mean age 30.3 years), cannabis use was associated with statistically significant, small-magnitude deficits in executive function, learning and memory, and global cognition, alongside moderate deficits in decision making. Differences in attention, information processing, and working memory were small and nonsignificant. The abstract reports no numerical effect sizes, confidence intervals, or p-values. Duration of use and age of onset did not moderate outcomes; however, executive function deficits were nonsignificant in analyses of prolonged abstinence (≥25 days).

Why it matters

This review demonstrates that residual cognitive deficits in long-term adult cannabis users are domain-specific and predominantly small, with executive deficits potentially resolving after extended abstinence.

Limits

The abstract reports no numerical effect estimates, confidence intervals, or heterogeneity statistics. Included evidence consists of non-randomized comparative studies susceptible to baseline confounding, and average sample sizes per included study were relatively small (averaging ~54 total participants per study).

Cited by