High potency cannabis use, mental health symptoms and cannabis dependence: Triangulating the evidence

Highlights

• Preferring high potency cannabis was associated with risk of cannabis dependence.Further triangulation with THC concentration did not support these results.

  Evidence for an association between high potency cannabis use and other mental health symptoms is mixed.

  Improved methodology is needed to assess cannabis use in research.

Abstract

Cannabis potency (concentration of Δ-9-Tetrahydrocannabinol) has been associated with risks of adverse mental health outcomes and addiction but no studies have triangulated evidence from self-report and objective measures of cannabis potency. We hypothesised that users of high potency cannabis would have higher levels of (a) anxiety, (b) depression and (c) psychosis-like symptoms (d) cannabis dependence than users of lower potency cannabis.

Methods

A cross-sectional study of 410 participants donated a sample of cannabis for analysis of THC concentration and reported their cannabis potency preference. These two exposure measures were investigated for their association with cannabis dependence, depression, anxiety, and psychosis-like symptoms in separate linear/logistic regression models.

Results

High potency cannabis preference was associated with a slight increased risk of cannabis dependence after adjusting for confounding, with the exception of cannabis use frequency (OR = 1.16, 95% CI 1.04–1.28). No association was found between THC concentration in cannabis and cannabis dependence. There was weak evidence of a small association between cannabis potency and depression and anxiety. There was no association between high potency cannabis preference or THC concentration in cannabis and psychosis-like symptoms.

Conclusions

Users of cannabis who preferred high potency types might be at increased risk of problematic cannabis use. This should be considered with caution as we were not able to triangulate these results with an objective measure of cannabis potency. More research is needed to understand the association between high potency cannabis use and depression and anxiety.

1. Introduction

Cannabis is the most commonly used drug globally, after alcohol and tobacco (UNODC, 2021). Cannabis potency is commonly defined by the concentration of Δ9-Tetrahydrocannabinol (THC), which is its primary psychoactive component. Experimental studies have shown THC causes acute psychosis symptoms in a dose-dependent manner (D’Souza et al., 2004). In addition, pre-clinical studies have suggested that THC could also be responsible for the reinforcing effects of cannabis in a dose dependent manner (Tanda et al., 2000). In the US and Europe concentration of THC have more than doubled over the last 10 years in street cannabis (Chandra et al., 2019, Freeman et al., 2021) and high potency cannabis products have increasingly become dominant in new legal markets (Smart et al., 2017). Therefore, users of cannabis with higher concentrations of THC might be at increased risk of poor mental health and addiction.A systematic review found that use of high potency cannabis has been associated with an increased in the risk of psychosis and cannabis use disorder, while the evidence for depression and anxiety is currently mixed (Petrilli et al., 2022). In a case-control study of patients with first-episode psychosis in South London, users of high potency cannabis were three times as likely to receive a diagnosis of psychosis compared to non-users (OR = 2.91, 95% CI 1.52–3.60). This was not the case for users of low potency cannabis (OR = 0.83, 95% 0.52–1.77; Di Forti et al., 2015). These findings were replicated by a multi-country study across Europe and Brazil. High potency cannabis users show an increase in the risk of psychosis compared to non-users (OR = 1.6, 95% CI 1.2–2.2) but not low potency cannabis users (OR = 1.1, 95% CI 0.9–1.5; Di Forti et al., 2019). Cross-sectionally, high potency cannabis use has also been associated with an increased in the risk of psychosis diagnosis (OR = 1.28, 95% CI 1.07–1.53; Chan et al., 2017). However, two studies investigating the effects of high potency cannabis use on psychosis-like experiences have not found an increase in symptoms (Hines et al., 2020, Okey and Meier, 2020). Psychosis-like symptoms refer to symptoms that mimic those of psychosis but do not meet the clinical psychosis threshold.High potency cannabis use has also been associated with an increase in the risk of problematic cannabis use. In a UK study, frequent use of high potency cannabis was associated with greater severity of dependence (b = 0.254, 95% CI 0.161 – 0.357), while this association was not found in users of low potency cannabis (Freeman & Winstock, 2015). Another UK study found evidence for a four-fold increase in the risk of recent cannabis problems in users of high potency cannabis compared to low potency cannabis (OR 4.08, 95% CI 1.41 – 11.81; Hines et al., 2020). A seven times increased in the risk of dependence (OR 6.85, 95% 1.87 – 25.15) was also found in a sample of Japanese patients (Matsumoto et al., 2020).The association of high potency cannabis use with depression and anxiety has been investigated less often. High potency cannabis use has been associated with a two times increase in the risk of anxiety diagnosis in a study conducted across 20 countries (OR = 1.72, 95% CI 1.55 – 1.91; Chan et al., 2017) as well as in a study conducted in England (OR = 1.92, 95% CI 1.11–3.32; Hines et al., 2020). However, other studies have not found an association between high potency cannabis use and anxiety (Bidwell et al., 2018, Prince and Conner, 2019). Similarly, one study have reported a slight increase in the risk of depression in users of high potency cannabis compared to low potency cannabis (1.18, 95% CI 1.11 – 1.25; Chan et al., 2017), while other studies have not reported an association (Bidwell et al., 2018, Hines et al., 2020, Prince and Conner, 2019).While there is evidence to suggest high potency cannabis could be associated with poorer mental health outcomes and problematic cannabis use, these studies carry limitations in the measure of exposure (Petrilli et al., 2022). Cannabis potency is commonly measured using self-report data of different cannabis products as a proxy. However, self-report measures of cannabis potency only provide an approximation, instead of accurate quantification of THC, which could be a source of bias or measurement error. By measuring THC concentration in participants’ cannabis, it is possible to use an objective measure of cannabis potency in the sample. Nonetheless, use of a single cannabis sample for quantification of THC concentration might not provide a representative indication of the cannabis that the participant is most frequently exposed to over time. Therefore, we aim to use methodological triangulation (Salkind, 2010) to study the association between cannabis potency and mental health symptoms and addiction. We will test our hypothesis with both an objective measure of THC concentration obtained from a sample of participants’ own cannabis, and that individual’s preference for different cannabis products with varying potencies in separate models. By triangulating evidence from 1) a self- report measure of cannabis potency preference with 2) gas chromatography coupled to mass spectrometry for THC concentration in cannabis samples, we aim to mitigate the limitations and biases of each of these two measures.As registered on the Open Science Framework (Petrilli et al., 2022), our hypotheses were that users of high potency cannabis will have higher levels of (a) anxiety, (b) depression and (c) psychotic-like symptoms (d) cannabis dependence than users of low potency cannabis.

2. Methods2.1. Participants and procedure

The participants included in this study were from a repeated measures study of 410 users of cannabis conducted at the Clinical Psychopharmacology Unit, University College London (Curran et al., 2018), conducted between October 2008 and March 2012. Curran et al., 2018 used an exploratory approach to investigate cannabis use exposure variables. In this study we focused specifically on cannabis potency, using methodological triangulation across two measures. The inclusion criteria were for all participants to be aged 16–24 years, fluent in English, using cannabis at least once a month for at least a year, no learning impairments, no clinical diagnosis of a substance use disorder and no history of, or current psychotic illness. Recruitment was targeted to include 200 daily cannabis users and 200 less than daily users. A total of 410 participants were recruited, of whom 194 reported daily cannabis use and 216 reported using cannabis less frequently. Participants were tested on two separate days, 7 ± 2 days apart, in their own homes. They were tested once when acutely intoxicated with their own cannabis and once when non-intoxicated. On the intoxicated test day, participants provided a sample of 0.3 g of the cannabis they were about to smoke. Only outcome data of the non-intoxicated condition was used for this analysis.Ethical approval was obtained by the University College London (UCL) ethics committee, project ID 0052/003. The study was conducted in line with the Declaration of Helsinki, and all participants provided written, informed consent to participating.

2.2. Measures2.2.1. Predictor variables

THC in sample.Participants provided 0.3 g samples of their cannabis and this was analysed for concentrations (% of sample weight) of THC by gas chromatography coupled to mass spectrometry, resulting in the total concentration of THC (%).Cannabis potency preference.Using a modified Cannabis Experience Questionnaire (CEQ; Barkus et al., 2006) data on participant’s preference (yes/no) for “higher potency herbal cannabis” (skunk, sinsemilla), “standard herbal cannabis”, “Low quality hashish/resin”, “High quality hashish/resin” or “other” was collected (modified CEQ in supplementary materials). We categorised cannabis as either high potency (typically ≥ 10% THC; “higher potency herbal cannabis”) or low potency (typically < 10% THC; “standard herbal cannabis”, “Low quality hashish/resin”, “High quality hashish/resin”), consistent with previous research (Di Forti et al., 2019, Freeman et al., 2014, Hines et al., 2020). “Other” types of cannabis were categorised as high potency or low potency cannabis depending on the information provided by the participant such as name of the cannabis type.

2.2.2. Outcome variables

Spielberger Trait Anxiety Inventory (STAI) (Spielberger, 1983).Participants completed the 20 items in the trait anxiety inventory only. Items are rated in a 4- point scale from ‘Almost Never’ to ‘Almost Always’, with higher scores indicating greater anxiety.Beck Depression Inventory (BDI) (Beck, Steer, & Brown, 1996).The BDI is a 21-item self-report questionnaire which measures symptoms of depression. Each item is scored from 0 to 3 and higher scores indicate greater levels of depression.Brief Psychiatric Rating Scale (BPRS) (Overall & Gorham, 1962).The BPRS is a clinical semi-structured instrument used as a clinitian-rated measure of psychosis-like symptoms.DSM-IV-TR Cannabis dependence (American Psychiatry Association, 2013).The DSM-IV criteria were used as a clinical diagnosis of cannabis dependence. Cannabis dependence was defined as manifested by at least three of the following six criteria within the last 12 months: (1) Tolerance (2) Use of larger amount of cannabis or for longer periods of time than intended (3) Repeated attempts to quit or control use (4) Large amounts of time obtaining, using or recovering from the effects of cannabis (5) Giving up social, recreational or occupational activities (6) Continue to use cannabis despite negative effects.

2.2.3. Pre-defined covariates

Demographic variables.Age and sex data were recorded at screening.Alcohol use.Frequency of alcohol use was collected as days per month used.Tobacco use.Frequency of tobacco use was collected as days per month used.Other illicit drugs.Other illicit drugs data were collected as use in the past month (Yes/No).Cannabis use frequency.Frequency of cannabis use was collected as days per month used.

2.3. Statistical analysis

Statistical tests were conducted on RStudio 2021.09.0, using the ‘R Stats Package’ version 4.1.2. We ran linear regression models for the two exposures (a) THC concentration and (b) cannabis potency preference for each of the continuous outcomes (depression, anxiety, psychosis-like symptoms). For each model (a) and (b), we ran two separate models per outcome: (1) adjusted for sex, age, alcohol use, tobacco use, other illicit drugs use (2) additional adjustment for cannabis use frequency. For the dichotomous outcome of DSM-IV-TR cannabis dependence we ran two logistic regression models to calculate odd ratios, with (a) THC concentration and (b) cannabis potency preference as the exposure variables and cannabis dependence as the outcome variable. For each model (a) and (b), two separate models were performed (1) adjusted for sex, age, alcohol use, tobacco use, and other illicit drug use (2) additional adjustment for cannabis use frequency.Bootstrapping of 95% confidence intervals was conducted with 10,000 bootstrap re-samplings using the package ‘boot’ version 1.3–28. Bootstrapping is a non-parametric approach based on re-sampling with replacement from the data set to estimate the sampling distribution of an estimator, creating an empirically generated sampling distribution (Mooney & Duval, 1993).

3. Results

3.1. Participant characteristics

3.2. High potency cannabis preference, mental health symptoms and cannabis dependence

We found an unadjusted 1.44 increased in the BDI symptoms of depression in users of cannabis that preferred high potency cannabis compared to low potency cannabis (b = 1.44, 95% CI 0.12–2.81). After adjustements in model 2 there was weak evidence of an association, with no evidence that cannabis potency preference was associated with symptoms of depression after adjusting for frequency of cannabis use in model 3.High potency cannabis preference, compared to low potency cannabis preference was not associated with an increase in symptoms of anxiety or psychosis-like symptoms.

3.3. THC concentration, mental health symptoms and cannabis dependence

4. Discussion

This cross-sectional study of 410 people who use cannabis compared presence of cannabis dependence and mental health symptoms in users of high potency cannabis compared to low potency cannabis. Preference of high potency cannabis was associated with an increased risk of cannabis dependence, compared to low potency cannabis. However, this effect was not present when adjusting for frequency of cannabis use. In addition, THC concentration in cannabis was not associated with cannabis dependence symptoms. We found weak evidence of an association between both high potency cannabis preference and THC concentration with symptoms of depression. There was also weak evidence of a small association between THC concentration and anxiety, but this was not the case in the high potency cannabis preference model. High potency cannabis use, measured as THC concentration in cannabis and high potency cannabis preference were not associated with increased symptoms of psychosis-like symptoms.The results suggesting an increased risk in problematic cannabis use in users of high potency cannabis compared to low potency cannabis are in line with previous studies (Freeman and Winstock, 2015, Hines et al., 2020, Matsumoto et al., 2020). However, we found a lower increase in risk compared to previous studies, which have reported about 4 times increased in risk of problematic cannabis use in users of high potency cannabis compared to low potency cannabis (Hines et al., 2020, Matsumoto et al., 2020). While these results persisted after adjusting for demographics and use of other drugs, we did not find an increase in risk of cannabis dependence after including frequency of cannabis use in the model. The role of cannabis use frequency in the relationship between cannabis potency and mental health and addiction is not currently clear. One possibility is that cannabis use frequency might act as a confounder if it leads to increase risk of problematic cannabis use and increased use of cannabis with higher potency (Hines, 2020). If this is the case, model 3 (adjusted for cannabis frequency) should be considered as the final model. Alternatively, cannabis use frequency might be a mediator if use of high potency cannabis leads to an increased frequency in cannabis use due to tolerance (Hines, 2020), in which case model 2 (unadjusted for cannabis use frequency) should be the final model. As suggested by previous research, we have included both of these models for consideration in our analysis (Petrilli et al., 2022).We aimed to triangulate the evidence using two measures of exposure: THC concentration in participants’ cannabis and a self-report measure of high potency cannabis preference. Finding a consistent association with both of these measures would have increased our confidence in the robustness of the relationship between cannabis potency and cannabis dependence. Contrary to our hypothesis, we did not find an association between THC concentration in the samples of cannabis and cannabis dependence. There are different explanations for the discrepancy in the results. Due to the cross-sectional design of this study, having measures of THC from only one sample of cannabis might not capture the THC concentration that participants are regularly exposed to over time, if different to the cannabis sample provided during the study. While this study collected samples at a single time point it would not be feasible to collect multiple samples from the same individual over time. Therefore, it is possible that the measure of high potency preference is a better representation of the types of cannabis that the participants are more often exposed to. Standardised methods to assess cannabis use, such as the standard THC Unit (Freeman & Lorenzetti, 2020) and the iCannToolkit (Lorenzetti et al., 2022) are needed to improve methodological consistency and resolve discrepant findings in the field.We found weak evidence of an association between THC concentration in cannabis and high potency cannabis preference and symptoms of depression. Previous studies have found little evidence (Chan et al., 2017) or a lack of evidence (Bidwell et al., 2018, Hines et al., 2020, Prince and Conner, 2019) to suggest an association between high potency cannabis use and depression. Our results also show weak evidence of a small association between anxiety and THC concentration, but not high potency cannabis preference. Some studies have found users of high potency cannabis were twice as likely to experience anxiety compared to users of low potency cannabis (Chan et al., 2017, Hines et al., 2020). However, this has not been replicated across the literature and other studies have reported no difference in symptoms of anxiety between high potency and low potency cannabis users (Bidwell et al., 2018, Prince and Conner, 2019). We did not find evidence of an association between cannabis potency and psychosis-like symptoms. While previous studies have shown an association between high potency cannabis use and psychosis (Chan et al., 2017, Di Forti et al., 2015, 2019), there is a lack of evidence to suggest an association between high potency cannabis use and psychosis-like symptoms (Petrilli et al., 2022). Thus, our results are in line with previous studies investigating psychosis-like symptoms (Hines et al., 2020, Okey and Meier, 2020).Tobacco use may influence the association between cannabis potency and mental health and cannabis dependence. As such, we adjusted for days of tobacco use in our models. Cannabis and tobacco are commonly used by the same individuals (Hindocha et al., 2021) and studies show that people who co-use cannabis and tobacco are at greater risk of showing poorer mental health outcomes (Peters et al., 2014, Ramo et al., 2012) and severity of cannabis dependence (Dierker et al., 2018, Hindocha et al., 2015).An important limitation across the literature is the predominant use of only self-report measures of exposure (Petrilli et al., 2022). One major strength of this study is the use of laboratory analysed measures of THC concentration in cannabis. This provides an accurate measure of THC concentration in cannabis, and thus of cannabis potency. Another strength of the study is the use of methodological triangulation. By comparing the results on laboratory analysed THC concentration in cannabis and self-report measures of cannabis potency preference, we are able to reduce some of the bias and limitations that each of these methods have. In addition, we adjusted for a range of relevant confounders which is an important limitation lacking across the literature.This study also has some limitations. Since this data was collected in the context of an illegal cannabis market, participants’ cannabis potency preference might not always match their cannabis use due to issues accessing their preferred type of cannabis. In addition, overall THC concentration in cannabis was relatively low and varied, which could have impacted our results. In addition, the data were collected between 2008 and 2012. Since concentration of THC in cannabis has increased over the years (Chandra et al., 2019, Freeman et al., 2021), future studies are needed to investigate associations with recently emerging high potency products such as concentrated extracts. Our study was conducted over a similar time frame to previous studies investigating associations between cannabis potency and mental health, and with similar potencies of cannabis (e.g. median THC concentration in this study of 9.1%; cut-off for high versus low potency cannabis in Di Forti et al. (2019) of 10%). This suggests that associations can be detected according to variation in cannabis potency from previous years. Future studies investigating more modern high potency products should also seek to triangulate evidence from both self-report data and objectively verified analyses of THC concentrations, as this was a key strength of this study.Other limitations to consider include, first, we used DSM-IV criteria for cannabis dependence, instead of DSM-5 criteria for cannabis use disorder. Second, the sample of participants is not representative of the general population. However, due to low levels of drug use in the population this limitation is common in this field, and sampling people who use cannabis was necessary in order to obtain a large sample of individuals who were eligible to take and donate a sample of their own cannabis. Third, the exposure and outcomes were measured cross-sectionally. Longitudinal analyses with reliable measures of cannabis potency are needed to better understand the association between use of high potency cannabis, mental health and addiction.

5. Conclusion

Participants who preferred high potency cannabis might be at increased risk problematic cannabis use. However, these effects were not supported by triangulation with measures of THC concentration in cannabis. High potency cannabis preference and THC concentration might be associated with symptoms of depression but more evidence is needed to understand this relationship. High potency cannabis use, compared to low potency cannabis use, was not associated with psychosis-like symptoms and we found only week evidence for an small association between THC concentration and symptoms of anxiety. These findings highlight the importance of improving the methodology used to assess cannabis use in research and clinical settings.Roles of Funding Sources. KP was funded by the Economic and Social Sciences Research Council (ESRC) as part of the South West Doctoral Training Partnership (SWDTP). The study was funded by a UK Medical Research Council grant (G0800268) to HVC and CJAM. The funders had no role in the study design, collection, analysis or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication.Contributors.Authors CJM, HVC and TPM designed and conducted the study. Author KP wrote the analysis plan presented in this manuscript, conducted the analysis and wrote the first draft of the manuscript. All authors contributed to the design of the analysis plan. All authors contributed to and have approved the final manuscript.Ethical standards. Ethical approval was provided by the University College London (UCL) ethics committee when the study was conducted and reviewed by the Psychology Research and Ethics Committee at the University of Bath before data access and analysis. Samples of participants’ cannabis were collected under the UCL Clinical Psychopharmacology Unit’s Home Office Licence.

CRediT authorship contribution statement

Kat Petrilli: Conceptualization, Methodology, Formal analysis, Writing – original draft. Lindsey Hines: Conceptualization, Methodology, Writing – review & editing. Sally Adams: Conceptualization, Methodology, Writing – review & editing. Celia J. Morgan: Investigation, Project administration, Funding acquisition. H. Valerie Curran: Investigation, Project administration, Funding acquisition, Writing – review & editing. Tom P. Freeman: Conceptualization, Methodology, Writing – review & editing, Supervision.