PTSD is a known risk factor for increased substance use [7–9]. Cannabinoids in particular have relevancy to PTSD both as therapeutic products [10–13] as well as drugs of abuse [7, 9]. Since the beginning of the COVID- 19 pandemic, evidence has mounted suggesting that many factors such as stress about world events, long peri- ods of confinement, and changes in sleep patterns have contributed to increased stress-related mental health problems [14–22], a phenomenon that may influence patterns of substance use. While some work has exam- ined changes in substance use behaviours during the pan- demic [23], little is yet known about the mental health factors associated with changes in cannabis use patterns among individuals with PTSD during this period.

Mental health factors have long been known to affect substance use. For instance, past research has identified that exposure to a traumatic event is related to the initia- tion of cannabis consumption [24]. PTSD-related mental health issues have also been highlighted as possible risk factors for substance use disorders [25–27]. Recent work has shown that among Canadian and Australian samples, combined usage of substances including alcohol and can- nabis increased during the COVID-19 pandemic [28, 29] – and the perceived worsening of PTSD symptoms dur- ing the pandemic was associated with these increases [29]. Among Canadians, approximately one third of substance-using individuals also reported an increase in alcohol and cannabis intake during the pandemic [30]. Cannabis became widely available to Canadians in 2018 following the introduction of the Cannabis Act, which legalized the possession, purchase, sharing, and sale of cannabis for individuals over 18 years of age [31]. Thus, since that time point, cannabis has been sharing a similar availability and legal status to alcohol in Canada. There is a need to better understand the mental health correlates of changes in substance use in people with PTSD during this unique period where cannabis legislation was quickly followed by a major global stressor.

There is some evidence to suggest that psychiatric comorbidities can affect cannabis consumption in peo- ple with PTSD. Recent evidence has indicated that both depressive and anxious symptoms are associated with greater odds of cannabis use in people with PTSD [32]. Perceived stress is similarly associated with increased cannabis intake [33, 34], and the relationship between perceived stress and cannabis use is mediated by depres- sion and anxiety in people with PTSD [35]. Additionally, stress, anxiety, and depression are cited as some of the most common reasons for cannabis use among PTSD- prevalent groups such as veterans [36]. There is also a physiological basis to suspect that cannabinoids may directly modulate brain processes contributing to PTSD symptomology as well as comorbid stress, anxiety and

depression symptoms. Individuals with PTSD exhibit reduced availability of the endocannabinoid anandamide and a consequent upregulation of its receptor (CB1) [37], which plays a key role in the mediation of stress and fear responses at relevant brain sites [10, 38–43]. The endo- cannabinoid system is also believed to be implicated in stress [44, 45], anxiety [46, 47], and depressive symptoms [48, 49].

Due to its prevalent use either under medical authori- zation or as self-medication for PTSD, we speculated that cannabis use might change with increases in stress, anxiety, and depression symptoms co-occurring with PTSD during a global stressor like the COVID-19 pan- demic. The present study aimed to assess mental health factors in a sample of individuals with self-reported cur- rent diagnoses of PTSD in order to determine whether changes in the severity of stress, anxiety, and depression symptoms during the pandemic relative to pre-pandemic estimates were associated with changes in cannabis con- sumption. We addressed this objective first by compar- ing symptoms of stress, anxiety, and depression among groups of individuals with PTSD who reported different cannabis intake behaviours across the pandemic (those who increased their intake, those who did not change their intake, those who decreased their intake, and those who did not use cannabis in the month prior to the pan- demic or during the pandemic). Next, we compared the proportions of those whose stress, anxiety, and depres- sion symptoms underwent minimal clinically important differences (MCID, a framework for characterizing the minimum level of symptom change that would be con- sidered meaningful and that would mandate a change in illness management [50] between the cannabis groups).

Methods

Data collection

The data for this study were obtained as part of a larger web-based population survey on the psychological, social, and economic impacts of the COVID-19 pan- demic that was circulated via websites, social media, and multiple organizations and hospitals across Canada (please see ClinicalTrials.gov: NCT04369690 and [21]). The data included in this report contains survey entries from April 3rd until June 24th 2020. The survey was available in both official Canadian languages (English and French) and included custom-built questions regard- ing the COVID-19 pandemic (please see a copy of these items in previously published data supplement [21]), as well as validated questionnaires regarding mental health. It was developed in accordance with the Checklist for Reporting Results of Internet E-Surveys (CHERRIES; [51]). Data collected for the study was collected in an anonymous manner. Retrospective questions were used to estimate stress, anxiety and depression symptoms, as well as cannabis use frequency across two time-referents: in the last month before the beginning of the COVID- 19 pandemic (the pandemic declaration by the World Health Organization occurred on 11 March 2020), and in the 7 days prior to filling out the survey.

Sample

The main inclusion criterion for the current sample was a self-reported current diagnosis of PTSD (i.e. select- ing PTSD amongst a list of multiple mental disorders when answering the question: ‘Have you ever had a for- mal diagnosis of (Please select all that apply)’). 466 par- ticipants with self-reported current PTSD were identified from the overall sample of 6,981 participants (6.7%) who filled out the survey during that period. Four participants with self-reported current PTSD were excluded from the analysis as they met the following exclusion criteria: younger than 18 years of age or diagnosis of a psychotic disorder. The final sample consisted of 462 participants.

The sample of respondents was divided in four can- nabis groups based on changes in cannabis use patterns from pre-pandemic to during the pandemic. Change scores were calculated by subtracting estimated pre-pan- demic cannabis consumption frequencies from the fre- quencies reported during the pandemic. Individuals with a negative change score (i.e., a decrease in weekly canna- bis consumption frequency from pre-pandemic to during the pandemic) were included in the decreased use group. Those with a positive change score (i.e., an increase in weekly cannabis consumption frequency) were included in the increased cannabis use group. Those with a change score of 0 were included in the no-change group. A can- nabis non-user group was also formed to include the respondents who reported not having used cannabis at any of the sampled time referents.

(…)