Background

Disparities in substance use and resulting harms between transgender and gender-diverse (TGD) people (i.e., individuals whose gender identity differs from societal expectations based on their sex assigned at birth) and cisgender people (i.e., individuals whose gender identity aligns with their sex assigned at birth) are well-established [1,2,3,4,5]. Such disparities are due, in part, to TGD individuals having an elevated risk of experiencing mental health problems [5,6,7], resulting from stigma, discrimination, and violence [8,9,10].

The COVID-19 pandemic has heightened mental health problems and reduced resilience to chronic stressors among TGD individuals due to delays in providing gender-affirming services, including gender-affirming interventions (e.g., transition-related surgeries, hormone therapy), and to reduced access to TGD support groups [11,12,13]. Substance use (i.e., alcohol drinking and cannabis use) may represent a coping strategy in response to elevated mental health problems [14,15,16,17]. This may result in acute and chronic harms, such as injury, substance use dependence, and death [18,19,20,21]. A better understanding of alcohol and cannabis use among TGD people during the pandemic is needed to inform public health interventions since TGD people may be more vulnerable to pandemic harms.

Recent research on the general population has shown an increase in alcohol [22,23,24,25] and cannabis [22, 26, 27] use during the pandemic, and that people experiencing high levels of mental health symptoms may be at risk of using substances [28, 29]. This pattern may have increased substance-related hospitalizations and deaths during the pandemic – in Canada, between March and September 2020, hospitalizations and deaths due to substance use increased by 5% and 13% compared to the same period in 2019 [30]. These increases are a pressing concern for the government and public health authorities working on a national recovery plan in response to the COVID-19 pandemic [31, 32]. Additionally, more national data on population health behaviours, such as substance use, are needed to address the needs of the total population [32], including more vulnerable subgroups [33]

Despite the disproportionate impact of the COVID-19 pandemic on the mental health of TGD people (compared to cisgender) [34,35,36], little research has examined the differential impacts of the pandemic on substance use among TGD versus cisgender individuals. There is a need for more research on this topic to build programs targeting the specific needs of TGD people in response to COVID-19 [37, 38]. Accordingly, this study aimed to assess differences between TGD and cisgender Canadian adults in their respective alcohol and cannabis use during the COVID-19 pandemic. To achieve our objective, we used repeat cross-sectional surveys conducted in Canada examining the impact of COVID-19 on mental health and substance use.

Methods

Participants

Data were obtained from seven successive Canada-wide cross-sectional web-based surveys of English-speaking Canadian adults aged ≥18 years. Survey data were collected by Delvinia using a proportional quota sampling methodology to approximate the English speaking population of Canada by age, sex, and region [39]. The surveys were conducted from May 2020 to March 2021: May 8-12, 2020 (Wave 1, n=1,005, response rate (RR)=15.9%), May 29-June 1, 2020 (Wave 2, n=1,002, RR=17.2%), June 19-23, 2020 (Wave 3, n=1,005, RR=16.4%), July 10-14, 2020 (Wave 4, n=1,003, RR=13.7%), September 18-22, 2020 (Wave 5, n=1,003, RR=17.6%), November 27-December 1, 2020 (Wave 6, n=1,003, RR=16.2%), and March 19-23, 2021 (Wave 7, n=1000, RR=15.8% (details of RR calculations are in Table S1 of the supplement). A pooled sample of 6,016 participants (Waves 2-7) was analyzed in this study. The questionnaires and data collected are provided in additional files. Wave 1 data were excluded as changes in alcohol and cannabis use were not measured for people who did not report drinking and cannabis use in the past week. The study received approval from the Research Ethics Board at the Centre for Addiction and Mental Health.

Measures

Gender identity was assessed using the question: “How do you describe your gender identity?” Response options included: “Man,” “Woman,” “Transgender man,” “Transgender woman,” “Two-Spirit,” “Non-binary (genderqueer, gender fluid),” “Questioning/Not sure of my gender identity,” “Identity not listed,” and “Prefer not to answer.” Individuals who self-identified as a man or woman were categorized as cisgender. Individuals who self-identified as transgender, two-spirit, or non-binary and those who selected “Questioning/Not sure of my gender identity” or did not find their gender listed were categorized as TGD. Due to sample size limitations, TGD sub-identities (e.g., transgender man, transgender woman, gender non-binary) were aggregated into one category [40,41,42]. Participants who responded “prefer not to answer” were excluded from the main analysis. This measure of gender identity may lead to misclassification as TGD participants who self-identified as a man or woman would have been classified as cisgender rather than TGD. We acknowledged that this limitation of the survey could potentially reduce the number of TGD in the sample. An alternative to accurately identify TGD persons in population-based surveys would be, for example, asking two separate questions (i.e., one for current gender identity and another for birth-assigned sex) [43,44,45].

Heavy episodic drinking (HED) was assessed using the question: “On how many of the past seven days did you drink five/four or more drinks on one occasion?” Cisgender men and TGD participants were asked about ≥5 standard drinks (≥68.0 grams of alcohol) and cisgender women were asked about ≥4 standard drinks (≥54.4 grams of alcohol). We used “5 or more” drinks on one occasion to screen TGD people for heavy drinking as recommended in a recent US study [46]. HED was defined as engaging in at least one HED occasion in the past 7 days. Subjective changes in alcohol use were assessed through the question: “In the past 7 days, did you drink more alcohol, about the same, or less alcohol overall than you did before the COVID-19 pandemic started?” Response options were: 1 (much less), 2 (slightly less), 3 (no change), 4 (slightly more), and 5 (much more).

Cannabis use was derived from the question: “During the past 7 days, on how many days did you use cannabis?” A binary variable was created to reflect any cannabis use (use on one or more days) in the past week versus no cannabis use. Subjective changes in cannabis use were assessed with the question: “In the past 7 days, did you use cannabis more often, about the same, or less often overall than you did before the COVID-19 pandemic started?” As was the case for changes in alcohol use, response options ranged from 1: much less to 5: much more.

The following individual and household covariates were included as confounders in all analyses: age groups (18-39, 40-59, and 60 years or more), ethnoracial background (White, Asian, Black/Indigenous/Arab/Latino, and other ethnicities), marital status (married/living with partner, separated/divorced/widowed, and single), education (high school or less, some post-secondary, college degree, and university degree), geographic location (urban, suburban, and rural), having children under 18 in the household, and whether or not a participant lived with others.

Statistical analyses

In order to ascertain whether an association existed between gender identity and alcohol/cannabis use during the COVID-19 pandemic, multivariate logistic regression models were used for binary dependent variables (i.e., HED and cannabis use), and ordinal logistic regression models were used for ordinal response variables (i.e., subjective changes in alcohol and cannabis use). Adjusted odds ratios (aORs) and adjusted proportional odds ratios (aPORs) were reported for logistic and ordinal regression models, respectively, adjusting for the above-mentioned covariates. To compare TGD and cisgender individuals in terms of their respective alcohol and cannabis use, alternating reference groups for gender were used in three sets of regression models; that is, TGD participants were compared to: 1) all cisgender participants (both men and women), 2) cisgender men, and 3) cisgender women. Stata (version 16.0) was used for all analyses.

Results

A total of 39 of 6,003 (0.7%) participants self-identified as TGD, 2,980 (49.6%) as cisgender men, and 2,980 (49.7%) as cisgender women. The TGD group was composed of eight transgender men, four transgender women, eleven two-spirits, nine non-binary (genderqueer and gender fluid), five participants selected “questioning/Not sure of my gender identity”, and two participants did not find their gender listed. Note that in Canada, the estimated percentage of the transgender population, including non-binary individuals, was 0.35% in 2019 [47]. We excluded thirteen participants for not answering the gender identity question. Table 1 presents the distribution of participants’ self-reported consumption of alcohol and cannabis, as well as their socio-demographic characteristics.

Table 1 Descriptive statistics: Substance use variables and covariates
Full size table

Table 2 presents the adjusted aORs and aPORs for the associations between gender and HED, cannabis use, and subjective changes in alcohol and cannabis use (Tables S2-S4 in the supplement present information regarding all parameters included in the regression models). No significant differences in HED were observed between the TGD group and cisgender participants (both women and men), cisgender men, and cisgender women; however, the odds ratios were greater than one as expected. TGD participants had higher odds of using cannabis at least once a week (aOR=3.78, 95%CI: 1.89, 7.53) relative to cisgender participants. Similar results were observed when TGD participants were compared to cisgender women (aOR=4.43, 95%CI: 2.21, 8.87) and men (aOR=3.20, 95%CI: 1.60, 6.41). Regarding subjective changes in alcohol and cannabis use compared to before the pandemic, the results showed that TGD participants were at greater odds of reporting subjective increases in their alcohol consumption (i.e., drinking much or slightly more versus no change, slightly less, or much less) relative to cisgender men and women combined (aPOR=2.00, 95%CI: 1.01, 3.95) and cisgender women (aPOR=2.05, 95%CI: 1.03, 4.05). Although the aPOR was greater than one when comparing TGD group to cisgender men regarding subjective changes in alcohol use, it was not statistically significant at 5% level. The TGD group was also at higher odds of reporting an increase in cannabis use than cisgender women and men combined (aPOR=4.56, 95%CI: 2.13, 9.78), cisgender women (aPOR=4.71, 95%CI: 2.18, 10.13), and cisgender men (aPOR=4.40, 95%CI: 2.04, 9.49).

Table 2 Associations between TGD status and HED, cannabis use, subjective changes in alcohol and cannabis use
Full size table

Participants who did not respond to the question on gender identity were excluded from the above results (13 participants were excluded). These participants were added into the TGD subsample to conduct a sensitivity analysis. Qualitatively similar results were observed when compared to the main findings (see Table S5 in the supplement).

Discussion

This study assessed disparities in HED, cannabis use, and subjective alcohol and cannabis use changes during the COVID-19 pandemic between TGD and cisgender participants. Results showed no significant differences in HED between TGD and any of the cisgender groups (i.e., both cisgender men and women, cisgender men or cisgender women), but interestingly the corresponding odds ratios were greater than one as expected. For all comparisons, TGD participants had higher odds of using cannabis. The results also revealed that TGD participants had a higher likelihood of reporting subjective increases in their use of cannabis since the start of the COVID-19 pandemic, compared with cisgender men and women combined, cisgender men, and cisgender women. Additionally, TGD participants were more likely to report subjective increases in their alcohol use during the pandemic than cisgender men and women combined and cisgender women. When compared to cisgender men, the proportional odds ratio of subjective change in alcohol was greater than one but not statistically significant.

These findings should be considered in the context of several limitations. First, the study used a non-probabilistic sample drawn from English-speaking Canadians, and therefore may not be generalizable to the Canadian population. Second, the cross-sectional nature of the study does not allow for causal inference. Third, the relatively small number of TGD respondents prevents the exploration of heterogeneity among TGD subgroups [48,49,50,51] and issues of intersectionality. It also resulted in wide confidence intervals for some estimates, thereby making the findings somewhat tentative. The measurement of gender identity in this survey may have impacted the number of TGD respondents. The gender identity question response options may not encompass all gender identities. To accurately classify transgender individuals, experts recommend including a second question measuring sex assigned at birth (i.e., "male" or "female") [44, 45], which was not done in this study. Therefore, some participants may have been grouped incorrectly. Finally, some individuals who do not identify as TGD may have been categorized as such, since those who selected: “Questioning/Not sure of my gender identity” and “Identity not listed” were included in the TGD group. Nevertheless, this preliminary and suggestive study offers useful insights into better understanding disparities in substance use between cisgender and TGD populations during the COVID-19 pandemic.

Our study’s findings are in line with recently published paper using data on Canadian adults [52] and previous research that reported increased risks for harmful drinking and HED [2, 4, 48, 49, 53, 54], and cannabis use [4, 49, 55] among TGD people, relative to their cisgender counterparts. These findings may be partially explained by social stigma and discrimination experienced by TGD individuals, resulting in heightened levels of stress, which may be heightened even further during the pandemic [56]. High levels of stress have been found to be associated with alcohol [57,58,59] and cannabis use [60,61,62]. Particularly, among TGD individuals, high levels of physical and psychological gender-related abuse have been associated with higher odds of alcohol and cannabis use [63, 64].

This study found that TGD participants increased their consumption of alcohol and cannabis during the COVID-19 pandemic compared to cisgender participants. These results are consistent with the fact that the implemented lockdown orders have exacerbated mental health symptoms among individuals [65,66,67,68,69], leading to more consumption of alcohol [22,23,24,25] and cannabis [22, 26]. The pandemic has exacerbated ongoing mental health disparities for TGD individuals [11, 12, 52, 70] due to reduced access to gender-affirming healthcare and treatment (as non-essential services) and the resulting increased psychological distress [11, 71]. This may have led to an increase in alcohol and cannabis use in TGD populations.

Conclusions

We identified disparities in HED, cannabis use, and subjective changes in alcohol and cannabis use during the COVID-19 pandemic between TGD and cisgender Canadian adults. Although our study highlights some differences in alcohol and cannabis use during the COVID-19 pandemic between TGD and cisgender people, more research is needed to fully understand disparities in substance use between TGD individuals and cisgender individuals during the COVID-19 pandemic. This will require collecting more data among TGD subgroups (i.e., transgender men, transgender women, two-spirit, and non-binary individuals). Identifying subgroups of TGD individuals at high-risk of engaging in substance use and misuse during the COVID-19 pandemic, and prioritizing gender-affirming medical and mental health care when reopening deferred services, are particularly important to increase general wellbeing among TGD individuals during and after the pandemic.