Back

Estimating thresholds for risk of cannabis use disorder using standard THC units

Lees Thorne, R.; Lawn, W.; Petrilli, K.; Trinci, K.; Borissova, A.; Ofori, S.; Mokrysz, C.; Curran, H. V.; Hines, L.; Freeman, T. P.

2025-05-21 addiction medicine
10.1101/2025.05.21.25328059 medRxiv
Show abstract

Background & AimsLower risk guidelines for safer levels of cannabis use could help to reduce the health burden posed by cannabis use disorder (CUD). We aimed to estimate risk thresholds for CUD based on delta-9-tetrahydrocannabinol (THC) consumption using standard THC units (1 unit = 5mg THC). DesignData from the CannTeen study, a longitudinal observational study consisting of five assessments over a 12-month period. SettingLondon, UK. ParticipantsParticipants were n = 65 adults aged 26-29 (46% female, 70.77% white ethnicity) and n = 85 adolescents aged 16-17 (56% female, 65.48% white ethnicity). All participants reported at least one use of cannabis during the 12-month study period and completed an assessment of DSM-5 CUD symptoms at the final follow up. MeasurementsMean weekly standard THC units were estimated using the Enhanced Cannabis Timeline Followback, a comprehensive assessment of quantity, frequency and potency of cannabis consumed. This was administered at 3-month intervals and averaged over a 12-month period. Past 12-month diagnosis of DSM-5 CUD was assessed at the final follow-up. Receiver operating characteristic curve models estimated the extent to which weekly standard THC unit consumption could discriminate no CUD from any CUD (mild, moderate, or severe), and no CUD from moderate/severe CUD, in adults and adolescents separately. Risk thresholds were selected based on cut-offs that maximised sensitivity and specificity. FindingsDiscrimination accuracy of weekly standard THC units on CUD was good, with area under the curve > 0.70 for all models. Optimal cut-offs for risk of any CUD (versus no CUD) were 8.26 units per week for adults and 6.04 units per week for adolescents. For risk of moderate/severe CUD (versus no CUD) optimal cut-offs were 13.44 units per week for adults and 6.45 units per week for adolescents. ConclusionsStandard THC units show good discrimination accuracy of CUD at different severities and in different age groups. Safer levels of cannabis use, defined by low weekly standard THC unit consumption, could be recommended in lower risk cannabis use guidelines.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.