Cannabis, mental health and addiction: methodological and epidemiological advances

: (Alternative Format Thesis)

Abstract

Cannabis use carries a risk of addiction, and its use has been associated with poor mental health outcomes. Over the years, cannabis potency, characterised by concentrations of Δ-9-tetrahydrocannabinol (THC), has increased in cannabis products and newer products with higher concentrations of THC have appeared in new legal markets. These changes could put people who use cannabis at higher risk of adverse outcomes, thus, understanding the association of THC with mental health and addiction is timely.

The aims of this thesis were 1) to systematically review the evidence on the association between cannabis potency and mental health and addiction; 2) to validate the Enhanced Cannabis Timeline Followback (EC-TLFB), a measurement of quantity and type of cannabis used designed to address inconsistencies in the existing literature on the association between cannabis use and adverse outcomes and 3) to investigate the association between THC and mental health and addiction across two studies: a cross-sectional study using methodological triangulation and a longitudinal study exploring directionality.

Firstly, I conducted a systematic review of observational studies on the association between cannabis potency and mental health and addiction. I found that use of higher potency cannabis, compared to lower potency cannabis, was associated with increased risk of cannabis use disorder (CUD) and psychosis, while the evidence for anxiety and depression was mixed. This review also identified limitations in the field, including on inadequate measures of cannabis potency.

Secondly, I further investigated the association between cannabis potency, mental health and addiction using cross-sectional data of participants who donated a sample of their cannabis for analysis of THC concentration and self-reported their cannabis potency preference. I investigated the association of these two measures of exposure with cannabis dependence, depression, anxiety, and psychosis-like symptoms in separate linear/logistic regression models, with the aim to triangulate the evidence. The results show high potency cannabis preference was associated with a slight increase in the risk of cannabis dependence, compared to low potency cannabis preference. However, there was no association between THC concentration in cannabis samples and cannabis dependence. I also found weak evidence of a small association between cannabis potency and increased risk of depression and anxiety and no association with psychosis-like symptoms. These results pointed to the need of standardised methods to assess the effects of cannabis use, such as the standard THC unit (5mg THC for all cannabis products and methods of administration).

To address the limitations in the measure of exposure reported in these analyses, I presented the EC-TLFB, a comprehensive assessment of cannabis use measures, including standard THC units, and validated these against objectively indexed urinary 11-Nor-9-carboxy-Δ9-tetrahydrocannabinol (THC-COOH) concentrations. Frequency of cannabis use, amount of cannabis used (measured with unaided self-report and using pictorial aided self-report), as well as total weekly standard THC units were calculated using data collected with the EC-TLFB as well as concentrations of THC in cannabis samples donated by participants and average concentrations of THC in cannabis products in the UK. All measures of cannabis exposure included were positively correlated with levels of THC-COOH in urine. Standard THC units, calculated with average concentrations of THC in cannabis in the UK and unaided self-report measures of amount of cannabis used in grams, showed the strongest correlation with THC-COOH in urine. The results show the EC-TLFB can provide a valid assessment of cannabis use measures, including standard THC units as a measure of THC dose.

These findings informed the final study which used longitudinal data to explore the direction of the association between standard THC units and CUD, anxiety and psychosis-like symptoms. This analysis included data on 5 testing sessions over a period of 12 months. I used random-intercept cross-lagged panel model (RI-CLPM) to examine the within-person reciprocal relationship between THC units and (a) CUD (b) anxiety symptoms and (c) psychosis-like symptoms at subsequent follow-ups. In the fully adjusted models, standard THC units at baseline were significantly positively associated with symptoms of CUD and symptoms of anxiety at the 3 month follow up. All other cross-lagged associations were not significant. These results showed some evidence for an association between cannabis use and subsequent CUD and anxiety symptoms. However, the evidence was not consistent across time points.

The studies presented describe limitations in the use of self-report cannabis potency as a measure of THC exposure and investigate the standard THC unit as a measure of THC dose, providing some evidence for its use as a comprehensive and valid measure of cannabis use. In addition, the results suggest higher THC exposure is associated with an increase in the risk of CUD. This thesis also reports weak evidence of an association between THC exposure with anxiety and depression and no evidence of an association with psychosis-like symptoms. However, the results should be interpreted in line with its limitations in measure of outcome for psychosis-like symptoms and study designs.

Altogether, this work highlights the significance of considering cannabis use profiles and direction of associations when examining the link between cannabis use, mental health and addiction. This thesis employs both methodological and epidemiological approaches to deepen our understanding of the relationship between cannabis, mental health, and addiction, as well as to explore new strategies for understanding the effects of THC on health outcomes.

Date of Award	7 May 2025
Original language	English
Awarding Institution	University of Bath
Supervisor	Tom Freeman (Supervisor), Lindsey Hines (Supervisor) & Sally Adams (Supervisor)