Urine THC Levels Dropped About 50% in People Who Self-Reported Reducing Cannabis Use | Research

The takeaway

Across 7 CUD treatment trials, participants who reported reducing cannabis use had urinary THC levels averaging 391 ng/ml lower than non-reducers, with about a 50% average decrease from baseline.

CUD treatment researchers, harm reduction advocates, and clinicians monitoring patient cannabis reduction.

50% average decrease in urinary THC among self-reported reducers

What the researchers found

Participants self-reporting cannabis use reduction had significantly lower urinary cannabinoids compared to non-reducers (difference of 391 ng/ml, 95% CI 231-551, p < 0.001). Average urinary cannabinoids decreased by up to 50% from baseline in the reduction group. However, no reliable individual-level cut-point could be established for creatinine-normalized THC.

Why it matters

Cannabis harm reduction research has lacked a biochemical verification method for use reduction (unlike abstinence, which can be verified). This study takes the first step toward establishing that self-reported reduction correlates with measurable decreases in urinary THC.

The numbers in context

N = 471 from 7 trials. 220 reducers, 251 non-reducers. Difference: 391 ng/ml (95% CI: 231-551, p < 0.001). ~50% average decrease in urinary cannabinoids for reducers. Individual CN:THC cut-point: -39.9 (95% CI: -70.3 to 2.9, not significantly different from zero).

How the study worked

Aggregated analysis of 7 CUD treatment trials (n=471 non-abstinent participants). Reduction defined as 50% decrease in frequency and/or 75% decrease in amount. Longitudinal models compared urinary cannabinoids between self-reported reducers (n=220) and non-reducers (n=251), controlling for parent study, study week, baseline levels, and covariates.

What this study cannot tell us

Aggregated data from different trials with varying designs and populations. Cannabis reduction was self-reported with potential bias. Creatinine normalization may not adequately account for hydration variation. 51% of original participants met criteria for analysis.

How to read the evidence

Aggregated analysis of 7 trials with biological verification. Moderate evidence for group-level but not individual-level biochemical verification.

When this study was published

Published in 2025.

The bigger picture

The inability to biochemically verify cannabis reduction at the individual level is a major gap in harm reduction research. While this study shows group-level verification works, individual verification will require further development, possibly with more frequent or standardized sampling.

Questions still open

Could more frequent urine sampling improve individual-level detection? Would hair or blood testing be more reliable for verifying reduction?

Common questions

Can a urine test prove someone reduced their cannabis use?

At the group level, yes: people who reported reducing use had about 50% lower urinary THC. But no reliable individual-level cut-point was established, so a single urine test cannot definitively confirm reduction for an individual.

Why does this matter for treatment?

Many cannabis users aim to reduce rather than quit entirely. Without biochemical verification of reduction, treatment trials and clinical monitoring rely solely on self-report, which has known biases.

Read the original research

Detecting cannabis use reduction through biochemical verification of urinary cannabinoids: An aggregated analysis of cannabis use disorder treatment trials examining average reductions and individual cut-points.

Addiction (Abingdon, England)

View on PubMed ↗Publisher’s paper ↗

Citation

McClure, Erin A; Neelon, Brian; Tomko, Rachel L; Dowd, Ashley; Gray, Kevin M; McRae-Clark, Aimee L; Baker, Nathaniel. (2025). Detecting cannabis use reduction through biochemical verification of urinary cannabinoids: An aggregated analysis of cannabis use disorder treatment trials examining average reductions and individual cut-points.. Addiction (Abingdon, England). https://doi.org/10.1111/add.70291