rethinkTHC Search
Menu
Study breakdown

Blood THC Metabolite Levels — Not Self-Report — Predicted Depression Symptoms in Young Cannabis Users

Cross SectionalModerate evidence
The takeaway

Among 94 young adults, only plasma THCCOOH concentration (not self-reported use or other biological samples) uniquely predicted depression symptoms, suggesting blood-based cannabinoid testing may reveal clinical risks that self-report misses.

Psychiatrists, addiction medicine specialists, clinical researchers, and young adult cannabis users concerned about mental health effects.

What the researchers found

Plasma THCCOOH concentration uniquely predicted depression symptoms (beta = 4.43, p < 0.001), while self-reported use days, oral fluid, urine, and hair concentrations did not. All toxicological matrices except oral fluid showed adequate sensitivity (63–74%) for detecting cannabis use, and concentrations did not differ by route of administration.

Why it matters

This finding suggests that objective blood measurements capture something about cannabis exposure that self-report cannot — potentially reflecting cumulative body burden or metabolic differences that are directly relevant to mental health outcomes.

The numbers in context

70 cannabis users, 24 non-users. Ages 18–21, 64% female. Oral fluid sensitivity only 12% (vs 63–74% other matrices). Self-report and toxicological metrics correlated (r = 0.41–0.97). Plasma THCCOOH predicted depression (β = 4.43, p < 0.001). No difference by route of administration.

How the study worked

Cross-sectional study of 70 cannabis-using and 24 non-using young adults (ages 18–21, 64% female). Oral fluid, plasma, urine, and hair tested for THCCOOH. Self-reported use, route of administration (smoked flower vs. vaped concentrate), and depression symptoms assessed. ANOVAs and regression models examined relationships.

What this study cannot tell us

Cross-sectional — depression could cause heavier use rather than the reverse. Small sample. Oral fluid had very poor sensitivity (12%). Cannot determine if plasma levels reflect recent heavy use or slower metabolism. Depression measured by self-report.

How to read the evidence

Multi-matrix toxicological assessment with clinical correlation, but cross-sectional design and small sample limit causal conclusions.

When this study was published

Published 2025.

The bigger picture

If plasma cannabinoid levels predict clinical outcomes better than self-report, clinical practice might benefit from incorporating biological testing when assessing cannabis-related mental health risks, especially in young adults.

Questions still open

  • Why did only plasma THCCOOH predict depression? Does plasma level reflect cumulative exposure, metabolic differences, or recent dose? Would longitudinal studies confirm this biomarker-outcome relationship?

Common questions

Does smoking vs. vaping cannabis matter for health effects?
In this study, THCCOOH concentrations did not differ between people who smoked flower versus vaped concentrates, and self-reported use patterns were also similar. Route of administration didn't affect the depression relationship.
Why was oral fluid testing so poor at detecting cannabis?
Oral fluid had only 12% sensitivity, meaning it missed most cannabis users. This may be because oral fluid THC levels drop quickly after use and depend heavily on time since last use, making it unreliable for detecting regular users.

Read the original research

Cannabinoid profiling across toxicology samples in adolescents and young adults by route of administration and in relation to depression symptoms.

Journal of analytical toxicology, 49(8), 567-575

Citation

Wade, Natasha E; Wallace, Alexander L; Baca, Rachel; Andrade, Gianna; Happer, Joseph P; Courtney, Kelly E; Christians, Uwe; Sempio, Cristina; Klawitter, Jost; Huestis, Marilyn A; Jacobus, Joanna. (2025). Cannabinoid profiling across toxicology samples in adolescents and young adults by route of administration and in relation to depression symptoms.. Journal of analytical toxicology, 49(8), 567-575. https://doi.org/10.1093/jat/bkaf051

Explore the wider topic