Illicit drug use and social victimization among Thai sexual and gender minority adolescents

Abstract

Objectives.

We examined the prevalence of lifetime illicit drug use and social victimization, and their association, among sexual and gender minority (SGM) and non-SGM Thai adolescents.

Methods.

In 2013, we conducted a school-based national survey among student’s grades 7–12 (aged 13–20 years) from 15 secondary schools (n=2,070) around Thailand. We classified adolescents with same-gender attraction, sexual or gender non-conforming identities as sexual and gender minority (SGM). Generalized estimating equations were used to estimate the odds of illicit drug use by SGM and non-SGM status.

Results:

Prevalence of lifetime illicit drug use was significantly higher among SGM adolescents than non-SGM adolescents (10.3% vs. 5.3%), but did not differ between those with same-gender attraction and SGM identity (10.3% vs. 10.8%). Among non-SGM adolescents, general social victimization, sexual experience and any school truancy were associated with lifetime illicit drug use (OR=2.59, 95% CI: 1.53, 4.38; OR=6.59, 95% CI: 4.90, 8.86; and OR=4.93, 95% CI: 3.13, 7.75, respectively). Among SGM adolescents, SGM based social victimization, depressive symptomology and suicidal ideation were associated with lifetime illicit drug use (OR=3.17, 95% CI: 2.03, 4.95; OR=5.03, 95% CI: 2.32, 10.90; and OR=5.03, 95% CI: 2.76, 9.16, respectively).

Conclusions:

SGM adolescents have higher burden of illicit drug use. Moreover, illicit drug use among SGM adolescents is indicative of depressive symptomology and suicidal ideation. Tailored and comprehensive programs are needed to reduce the gap in burden of illicit drug use between SGM and non-SGM adolescents.

Introduction

Prevalence of illicit drug use among Thai adolescents has nearly tripled over the past decade [1–5]. This is an urgent public health crisis because illicit drug use during adolescence predicts a range of concurrent and subsequent social vulnerabilities and health burden including depressive disorders, suicidal risk, sexual risk behaviors, and educational and socioeconomic disadvantages [2, 6–10]. Recent studies suggest that intervention to reduce illicit drug use targeting school-attending youths would be more effective, when coupled with reduction of peer victimization, particularly on reducing the disparity of drug use among heterosexual and non-heterosexual adolescents [4,11,12]. However, it is still unclear whether reduction of peer victimization in general, as opposed to reduction of sexual and gender minority (SGM)-based victimization in particular, can reduce the disparity of illicit drug use among SGM adolescents [13].

Studies have shown that illicit drug use in youths is substantially higher among SGM groups [9,13–15]. The excess prevalence of substance use among SGM youth is associated with elevated exposures to SGM-specific prejudice and victimizing experiences [13]. However, there is scarce evidence to support this early disparity on drug use by SGM populations and its linkage to SGM-specific victimization, particularly in lower- and middle-income countries [16,17]. To our knowledge, among the few studies focused on illicit drug use among younger age groups in Thailand [1,4–5,18], only one has examined illicit drug use by sexual minority status [16]. But none has explored the relationship between victimization experience and illicit drug use specifically among SGM youth.

In Thailand, there are multiple sexual and gender identities; some are intertwined and complex [19–25]. For example, the identity tom refers to a butch lesbian, which denotes both a masculine gender identity and same-gender sexual attraction. Similarly, the term dee refers to a feminine lesbian. Our team was among the first to investigate and classify these diverse Thai sexual and gender minority (SGM) identities in both adolescent and adult populations [21,25].

In this article, we aimed to characterize the patterns and correlates of illicit substance use by non-SGM and SGM Thai adolescents. We hypothesized that SGM-based peer victimization is significantly associated with illicit drug use. In addition, there is emerging evidence that gender or sexual identities alone do not capture the full spectrum of SGM individuals [26]. Therefore, utilizing a more nuanced measurement on sexual attraction, we hypothesized that Thai youth who are SGM-identified and those with same-gender attraction only would have similar patterns of illicit drug use; but when compared to non-SGM youth, both groups of SGM youth would have higher odds of illicit drug use than non-SGM youth.

Methods

Procedure and Participants

This article analyzed data from a cross-sectional, probability-based, multistage cluster sample of youth recruited from 15 secondary schools (aged 13–20 years) representing all regions in Thailand. The details of the sampling methodology can be found elsewhere [25]. Briefly, the study took into account all 4 regions of Thailand and Bangkok, public/private schools, and urban and rural settings. Schools in each region, classrooms within each school and then students within each classroom were randomly sampled. Students were invited to participate in a computer-assisted, self-administered survey in Thai, lasting 15–30 minutes each. Eligible criteria include being enrolled in a secondary school, at least 13 years of age, Thai speaking and able to provide informed consent or assent. The survey response rate was 90% and survey completion rate was 95% over the 1-month recruitment period between August and September 2013, leading to a final sample of 2,070 Thai youth. These proportions are typical of studies involving student participants in Thailand [3,4,27].

Informed assent and guardian permission.

Researchers obtained written informed consents from student participants 18 years and older and written informed assents from student participants 13 to 17 years. For participants who were younger than 18 years, additional written permission from guardians or proxies were obtained (homeroom teachers were proxies). All students were given participant information sheet, consent/assent and guardian permission forms that provided details of the study, at least one week before the research team arrived at the schools. Homeroom teachers explained the objective and details of the study to students and ask them to take these forms home to their guardians and to discuss with their guardians about study participation. If guardians have questions, they were encouraged to contact the homeroom teacher, the school administrator, the study PI, or the Chair of the University Institutional Review Board (IRB). Contact information was provided in the forms. For students under 18 years who were interested in participating in the study but did not return the forms to the homeroom teachers by the day the researchers arrived at the school to administer the surveys (usually because the students forgot to give the forms to their parents or forgot to bring the forms back to the school), homeroom teachers provided permission by proxy. Before all study activities, researchers went through the consent/assent process in detail to make sure that students were aware of the risks and benefits of study participation. Students were informed that study participation was voluntary and anonymous and each student received 50 baht (1.5 USD) for their participation. The IRB at Mahidol University reviewed and approved this study.

Measurements

We developed and piloted the study questionnaires by formative research that included systematic observations in schools, focus group discussions and in-depth interviews with students, teachers, and school administrators. The questionnaire began by asking student demographic characteristics such as sex, daily expenses, current part-time employment, and last term’s GPA.

Illicit drug and alcohol use.

Students were asked whether they have ever used the following substances: marijuana (ganja), methamphetamine-based drugs (pill form: ya ma or ya ba; crystal meth: ya ice), ecstasy (ya e), tranquilizers (sleeping pills or maew [cough syrups]), and injectable substances such as heroine and crystal meth.

Sexual and gender identities.

Based on previous studies conducted in Thailand, and this study’s previous qualitative findings [25], we asked participants to identify their gender that included locally-constructed masculine (e.g., chaay [man], gay, chai rak chai [man who loves men], tom [butch lesbian], phu chai kham phet [transgender man]) and feminine (e.g., ying [woman], dee [feminine lesbian], les [versatile lesbian], ying rak ying [woman who loves women], kathoey [transgender woman]) identities. We then categorized participants to either having a sexual and/or gender minority (SGM) identity using any identities above other than man and woman, for which we categorized as non-SGM identity.

Same-gender attraction.

Participants were asked about the preferred gender of their faen (steady) and kik (casual) partners. In this question, we aim to assess same-gender attraction among participants who identified as man or woman and may have not had any same-gender partnerships. For participants who identified as man/woman, those who were attracted to the gender identity of faen or kik partners other than woman/man were subsequently classified as same-gender attracted.

Social victimization.

Social victimization is a type of peer victimization. We categorized participants as having experienced social victimization if they answered yes to any of the 3 questions that assessed past month’s experience of whether other students have gossiped, spread rumors to make them look bad, excluded them from their social groups, or displayed degrading or nasty looks at them. These questions also assessed frequency of victimization in the past month as “not once,” “one to three times per month,” “once a week,” and “more than once a week.” Those who chose one of the above categories were then deemed as having experienced general social victimization. In addition, they were subsequently asked whether their victimization experience was related to their sexual and gender orientation following each item with a binary “yes/no” variable. Those who had a positive response in each of these items were categorized as having experienced SGM related social victimization.

Unauthorized school absences.

We categorized participants who reported having ever skipped school without any authorized reasons into a binary (yes, no) variable.

Depressive symptoms.

We used the Center for Epidemiologic Studies-Depression (CES-D) questionnaire Thai version which consisted of a 20-item psychometric scale to evaluate participants’ depressive symptoms in the past week with a sum score ranging from 0 to 60 points. A higher score equates to a higher level of depressive symptoms. We used sum score of 22 or higher to categorize those with a high level of depressive symptomology, this version and cut-off have previously validated using adolescent samples from Thailand [28,29].

Suicidal ideation.

We asked participants whether they have had any thoughts to commit suicide in the past month and recorded their binary (yes/no) response.

Sexual intercourse.

We asked participants whether they have had any sexual intercourse (yes/no).

Statistical analyses

We conducted all analyses using SAS 9.3, North Carolina, USA with a two-sided, 10% alpha level for hypothesis testing. We estimated proportions using multi-stage sampling weights. The associations between the prevalence of illicit drug use and same-sex attraction and LGBT identity were evaluated using Pearson’s chi square test for categorical variables or Fisher’s exact test when expected values for these variables were less than 5. For both bivariate and multivariable analyses, we corrected the clustering effect by schools using generalized estimating equations (GEE) with robust variance [30]. In the first part of the analyses, we described the patterns of illicit drug use and SGM-based social victimization by categories of non-SGM, same-gender attracted, and SGM-identified adolescents using design-weighted proportions. We estimated the adjusted odds ratios of illicit drug use and SGM-based social victimization comparing across these three categories. Then, separately for those with non-SGM and SGM adolescents, we estimated crude odds radios of any illicit drug use by socio-demographic variables, social victimization, alcohol use, any sexual intercourse, unauthorized absences from school, depressive symptoms and suicidal ideation using logistic regression procedures. Finally, we constructed multivariable logistic models to identify independent correlates of illicit drug use among non-SGM and SGM adolescents. Complete case analyses were used in statistical procedures since none of the responses had aggregated missing data exceeding 10% of the analytic samples.

Results

The study sample of 2070 students was drawn from a national sample of 15 public and private secondary schools (grades 7 to 12) from urban and rural areas in all five regions of Thailand (Bangkok, Central, Northern, Northeastern and Southern regions). Student’s age ranged from 13 to 18 years. Table 1 shows lifetime prevalence of marijuana, methamphetamine, ecstasy, tranquilizers, and injection drug use by three categories: 1) non-SGM youths, 2) SGM youths with same-gender attraction only and 3) SGM-identified adolescents. Specific patterns in lifetime use of each category were not significantly different (p>0.05) between those with SGM identities and same-gender attraction. After collapsing these two groups and compared to non-SGM group, most categories of drug use were significantly higher (p<0.05) among SGM youths, compared to non-SGM youths, with the exception of marijuana use (p=0.27). After adjusting for sex assigned at birth, those with same-gender attraction only (10.4%, AOR= 1.94, 95% CI: 1.15, 3.26) and those with SGM identities (10.8%, AOR= 2.27, 95% CI: 1.38, 3.73) had approximately twice the odds of reporting any illicit drug use, compared to non-SGM participants (6.6%). SGM-identified youths had more than twice the odds of having SGM-based social victimization, compared to SGM youths with same-gender attraction (37.0% vs. 20.7%, AOR= 2.19, 95% CI: 1.40, 3.42).

Table 1.

Patterns of lifetime illicit drug use by same-sex attraction and LGBT identity among Thai youths (N=2070)

			Non-SGM		Same-gender attraction		LGBT identified
	N	(%)	n	(%)	n	(%)	n	(%)
Overall	2070	--	1581	76.4	243	11.7	246	11.9
Illicit drug use
Marijuana	51	2.5	37	2.3	9	3.7	5	2.0
Methamphetamine	32	1.5	18	1.1*	7	2.9	7	2.8
Ecstasy	7	0.3	2	0.1*	2	0.8	3	1.2
Tranquilizers	73	3.5	36	2.3**	16	6.6	21	8.5
Injection drug use	8	0.4	2	0.1**	2	0.8	4	1.2
Any illicit substance use	126	6.1	77	4.9**	24	9.9	25	10.2
AOR (95% CI)			Ref		1.94 (1.15, 3.26)		1.69 (1.36, 2.11)
Any SGM based social victimization	327	15.8	187	11.8**	51	21.0	89	36.2**
AOR (95% CI)			0.51 (0.38, 0.69)		1		2.19 (1.40, 3.42)

Note.

^*= p ≤ 0.05

^**=p ≤ 0.01

AOR=adjusted odds ratio, models adjusted for sex at birth and clustering by school

Table 2 summarizes the bivariate associations of socio-demographic and psychosocial variables with any illicit drug use by sexual and gender minority (SGM) status. Among non-SGM youths, those who were males (OR= 2.02, 95% CI: 1.46, 2.79), who had GPA below 3.0 (OR= 2.82, 95% CI: 1.83, 4.33), had any casual or steady partners (OR= 3.34, 95% CI: 2.22, 5.13), had sexual intercourse (OR= 6.59, 95% CI: 4.90, 8.86), had unauthorized absences from school in the past month (OR= 4.93, 95% CI: 3.13, 7.75), had any suicidal ideation in the past month (OR= 2.13, 95% CI: 1.39, 3.26), had clinically significant depressive symptoms (OR= 2.85, 95% CI: 1.51, 5.36), and had been socially victimized in the past month (OR= 2.59, 95% CI: 1.53, 4.38) were significantly more likely to have higher odds of life-time illicit drug use. Among sexual minorities, only those who had unauthorized absence from school in the past month (OR= 2.69, 95% CI: 1.40, 5.15), suicidal ideation in the past month (OR= 5.03, 95% CI: 2.76, 9.16), clinically significant depressive symptoms (OR= 5.03, 95% CI: 2.32, 10.90), and experienced non SGM-based social victimization and SGM-based social victimization in the past month had higher odds of lifetime illicit drug use (OR= 4.06, 95% CI: 1.77, 9.31; OR= 3.17, 95% CI: 2.03, 4.95, respectively).

Table 2.

Correlates of illicit drug use among SGM and non-SGM Thai youths (N=2070)

	Illicit Drug Use by Sexual and Gender Minority (SGM) Status
	Non-SGM (N=1581)				SGM (N=489)
	n*	%	OR	(95% CI)	n*	%	OR	(95% CI)
Sex
Male	47/704	6.7	2.02	(1.46, 2.79)	18/153	11.8	1.31	(0.61, 2.93)
Female	30/887	3.4	1		31/336	9.2	1
Grade level
7th- 9th	32/799	4.0	1		22/229	9.6	1
10th- 12th	45/781	5.8	1.47	(0.92, 2.35)	27/260	10.4	1.09	(0.58, 2.04)
Missing	0/1	0.0			0
Daily expenses
60 baht or less	25/755	3.3	1		19/207	9.2	1
61 baht or more	52/826	6.3	1.85	(0.86, 4.04)	30/282	10.6	0.97	(0.42, 2.25)
Part-time employment, current
Yes	13/237	5.5	1.19	(0.66, 2.15)	15/101	14.9	1.66	(0.70, 3.95)
No	64/1344	4.8	1		34/388	8.8	1
Last term’s GPA
3.0 or above	15/631	2.4	1		16/187	8.6	1
Below 3.0	62/950	6.5	2.82	(1.83, 4.33)	33/302	10.9	1.58	(0.79, 3.16)
Any casual or steady partners, ever
None	16/747	2.1	1		12/173	6.9	1
Opposite gender only	61/817	7.5	3.34	(2.22, 5.13)	13/105	12.4	1.34	(0.39, 4.57)
Same and opposite gender	--				24/211	11.4	1.77	(0.75, 4.18)
Any sexual experience, ever
Yes	31/164	18.9	6.59	(4.90, 8.86)	13/77	16.9	1.84	(0.87, 3.90)
No	46/1417	3.2	1		36/412	8.7	1
Any unauthorized absence at school, past month
Yes	48/425	11.3	4.93	(3.13, 7.75)	24/156	15.4	2.69	(1.40, 5.15)
No	29/1156	2.5	1		25/333	7.5	1
Any suicidal ideation, past week
Yes	11/117	9.4	2.13	(1.39, 3.26)	20/76	26.3	5.03	(2.76, 9.16)
No	66/1464	4.5	1		29/413	7.0	1
Depressive symptoms (CESD ≥22), past week
Yes	22/192	11.5	2.85	(1.51, 5.36)	24/99	24.2	5.03	(2.32, 10.90)
No	55/1389	4.0	1		25/390	6.4	1
Non-sexual minority based social victimization, past month
Yes	20/723	2.8	2.59	(1.53, 4.38)	42/323	13.0	4.06	(1.77, 9.31)
No	57/858	6.6	1		7/166	4.2	1
Sexual minority based social victimization, past month
Yes	11/187	5.9	1.15	(0.58, 2.27)	24/140	17.1	3.17	(2.03, 4.95)
No	66/1394	4.7	1		25/349	7.2	1

^*The numerator is the number of students who have ever used any of the illicit drugs listed in Table 1. The denominator is the total number of students who have ever used illicit drugs and those who have not used illicit drugs for that condition (e.g., 47/704 represents 47 non-SGM male students who have used illicit drugs and 704 is the total number of non-SGM male students in the sample). Due to missing values, the denominator for each variable is displayed.

Table 3 presents the multivariable logistic regression models by SGM status. Among non-SGM students, ever had sexual experience was independently associated with odds of lifetime illicit drug use (OR= 4.03, 95% CI: 2.83, 5.75). In addition, unauthorized school absences and social victimization (general) in the past month were independently associated with odds of lifetime illicit drug use (OR= 2.76, 95% CI: 1.69, 4.51 and OR= 2.05, 95% CI: 1.28, 3.28, respectively). Among SGM students, those who reported suicidal ideation (OR= 2.65, 95% CI: 1.25, 5.64), clinically significant depressive symptoms (OR= 3.06, 95% CI: 1.10, 8.53) and SGM-based social victimization (OR= 2.06, 95% CI: 1.10, 3.86) were more likely to have higher odds of lifetime illicit drug use.

Table 3.

Adjusted odds of illicit drug use by sexual and gender minority (SGM) status among Thai youths

	SGM status
	Non-SGM		SGM
	Model A		Model B
	AOR	(95% CI)	AOR	(95% CI)
Any sexual experience, ever
Yes	4.03	(2.83, 5.75)	1.22	(0.48, 3.07)
No	1		1
Any unauthorized absence at school, past month
Yes	2.76	(1.69, 4.51)	1.57	(0.85, 2.92)
No	1		1
Any suicidal ideation, past week
Yes	1.33	(0.67, 2.65)	2.65	(1.25, 5.64)
No	1		1
Depressive symptoms (CESD ≥22), past week
Yes	1.89	(0.92, 3.89)	3.06	(1.10, 8.53)
No	1		1
Non-SGM based social victimization, past month
Yes	2.05	(1.28, 3.28)	1.81	(0.84, 3.89)
No	1		1
SGM based social victimization, past month
Yes	1.13	(0.57, 2.23)	2.06	(1.10, 3.86)
No	1		1

Discussion

The present study, for the first time, utilized a diverse classification of Thai sexual and gender identities, coupled with same-gender attraction, to examine lifetime prevalence of illicit drug use among secondary school-attending Thai adolescents. We found that SGM youths had more than two-fold higher odds of illicit drug use, compared to non-SGM youths. And although SGM-identified youths had twice the odds of SGM-based social victimization, compared with adolescents with same-gender attraction, the two groups did not differ on burden of illicit drug use. Similarly, Brewster et al [26] found that among adolescent sexual minorities, those who identified as LGBT had similar prevalence of substance use than those with same-gender attraction and/or sexual contact only. Collectively, this evidence supports the psychosexual development perspective, which proposed that SGM identity formation follows the development of sexual attraction [31]. Future studies should investigate the trajectories of sexual and gender attraction and identity development during adolescence, and explore the contextual and resilient factors that may mitigate the negative social consequences associated with SGM identity formation [32].

In addition, we illustrated that risk factors for drug use differed significantly across non-SGM and SGM adolescents. Pengid and Peltzer [4] found that school truancy and having had sexual experiences were associated with higher prevalence of illicit drug use. We found similar association between truancy, sexual intercourse experience and illicit drug use among non-SGM adolescents. However, among SGM adolescents, illicit drug use was also associated with adverse mental health outcomes, including depressive symptoms and suicidal ideation. First, this evidence suggests that existing programs focused on addressing risk factors of illicit drug use on heterosexual and gender-conforming adolescents may not benefit non-heterosexual and non-gender-conforming adolescents. Secondly, it demonstrated the universality of multiple co-occurring psychosocial adversities faced by SGM adolescents referred to as syndemics [21,33,34], and highlighted the urgent need for comprehensive and tailored health services for SGM youths in Thailand [35]. Third, we found that SGM related victimization, but not general victimization, was associated with illicit drug use among SGM adolescents. A recent meta-analysis found a stronger association between general victimization on substance use than LGBT-based victimization among sexual minority adolescents [13]. Our results therefore provide further support [11] that these reviewed studies using general measures of victimization may not have adequately captured LGBT-based victimization.

Results of our study should be interpreted in light of the following limitations. Due to the lack of temporal sequence of our cross-sectional design, illicit drug use could have preceded victimization and the development of identity and sexual attraction and other covariates included, however, the majority of cohort studies do not support these reverse pathways [13,36,37]. We acknowledge potential information bias on reporting sensitive information such as SGM attraction and identities, illicit drug use and victimization, however, the use of computer-assisted self-administered survey and multiple reminders of anonymity may have minimized this bias [27,38]. The low reporting of illicit drug use among adolescents in this study may limit potential generalizability. Additionally, we used three single-item questions to assess social victimization that have not been validated. We acknowledge that thoroughly validated scales may have higher reliability and validity in the statistical sense than concise scales (like ours). That said, many large-scale studies (e.g., HBSC, EU Kids Online, TIMMS) routinely inquire about bullying, cyberbullying, or aggression experiences using single-item measurements, and data gained from these measurements are reported in the research literature despite the limitations of the measurement strategy. In our case, we conducted a qualitative pilot test of our measurement to ensure that the three single-item measurements were culturally relevant types of social victimization in the Thai context, and easily understood by our pilot testers, who were equivalent in their characteristics to those who completed the actual survey. Finally, our sample was limited to only secondary school-attending Thai adolescents, and therefore, results may not be generalizable to all Thai adolescents, particularly those who were institutionalized, attending vocational schools, dropped out of school or were absent at the day the survey was implemented. For example, students who had high truancy levels, and so may not be included in the study, may potentially contribute to selection bias.

Despite these limitations, our study has important implications. Firstly, we found that illicit drug use is indicative of higher burdens of depressive symptomology and suicidal thoughts among SGM adolescents, but not among non-SGM adolescents. Educators and mental health professionals in Thailand should provide timely screening and referral to appropriate services for SGM adolescents. Secondly, we demonstrated, for the first time, the application of a nuanced and culturally sensitive approach in categorizing sexual and gender minority adolescents in Asia [11,26,39]. This implies that existing research and services focused on adolescent’s self-identities and particularly those based on western contexts (gay, lesbian, bisexual and transgender identities) may not sufficiently identify the full spectrum of sexual and gender minority adolescents. Further, our study demonstrated that including SGM students under 18 years in surveys is possible when studies are conducted in schools and allow proxy permission by homeroom teachers. Sometimes SGM students under 18 years are reluctant to participate in research studies because guardian permission are often required. By having homeroom teachers provide proxy permission, SGM students are able to participate in sexual and gender minority-related research studies. We have previously showed through a qualitative study with Thai parents of students under 18 years that guardians are willing to waive permission for their children to participate in research studies if the said studies are conducted on school grounds and with permission by school teachers [40]. Lastly, we confirmed previous findings from other countries that SGM-based social victimization is associated with illicit drug use, further strengthening this evidence, particularly from a middle-income country in Asia. The United Nations Educational, Scientific and Cultural Organization (UNESCO) has been bringing worldwide attention to this issue and has been working to decrease LGBT-based bullying victimization in schools. However, these efforts require responsive and dedicated efforts from local governments such as the ministries of education and ministries of health so that preventive intervention programs are implemented, targeted, culturally sensitive, sustainable and galvanize the dignity of young individuals [11,13].