Health Outcomes of Sexual-Minority Youth in Canada: An Overview

Abstract

Objectives

The objectives of this review are to (1) document health outcomes among sexual-minority youth (SMY) in Canada; and (2) identify sexual-minority-specific risk and protective factors

Data Sources

We conducted a review of Canadian data published after 2005 on the mental, physical and sexual health outcomes of SMY using relevant keywords. A total of 19 empirical studies and 2 research reports was included.

Data Synthesis

The study reviewed included 53 to 30 588 respondents (total = 81 567). SMY counted for 15.86% of the total sample. Overall, SMY in Canada show negative health outcomes in proportions varying from 7% to 69.4%, the most common issues being psychological distress and maladjustment. SMY are more likely than their heterosexual peers to report psychological distress/malfunctioning, suicidality, substance misuse, condomless intercourse, pregnancy involvement. Main SMY-specific risk factors were family rejection of one’s minority sexual orientation, homophobic bullying and victimization, and internalized homophobia. Among the few protective factors that were reported, school and family connectedness, school safety, parental support and sports involvement decreased the odds of negative health outcomes.

Conclusions

Canadian data show that SMY are more likely to experience negative health outcomes than their heterosexual peers. These results are consistent with data from around the world. Recommendations for research and intervention are discussed.

Introduction

As they grow up, sexual-minority youth (SMY) must face challenges typical of adolescence as well as challenges specific to their gender and sexual status. Sexual minorities, or gender and sexual minorities, refer to “people who identify as gay, lesbian, bisexual, transsexual, transgender, Two-spirited, queer, and questioning (LGBTQ people), as well as people who would choose other labels (or no labels) but whose sexual attractions or behaviors, gender identity or gender expression diverges from the heteronormative models that dominate most social contexts” ⁽¹⁾. While Canada is an international leader in promoting sexual minorities rights, homonegativity and homophobia still exist: gender and sexual variations still provoke negative reactions and SMY are overrepresented among bullying victims ^(2,3). Homonegativity and homophobia create a challenging environment for growing up as SMY, which contributes to higher proportions of SMY facing health issues compared to heterosexual adolescents ⁽⁴⁾. In this paper, the challenges and perspectives on adolescence in different socio-cultural contexts will be addressed by reviewing Canadian evidence on SMY health. The objectives of this paper are to (1) document health outcomes among sexual-minority youth in Canada; and (2) identify sexual-minority-specific risk and protective factors. In order to do so, we conducted a review of the relevant work that has been done in Canada.

Method

Data Sources and Studies Selection

Three steps were undertaken to select the papers included in this review. In a first step, we searched scholar databases (PsychInfo, Pubmed, Google Scholar) using terms related to physical, mental and sexual health (psychological distress, condomless sex, substance use, tobacco use, pregnancy involvement, etc.), adolescence (youth, emerging adult, adolescent), sexual minority youth (lesbian, gay, bisexual, trans, LGB, LGBT) and Canada. Only papers published after 2005 were included. A total of 2508 papers was identified in this first step. In a second step, titles and abstracts were reviewed and included for further analysis if they provide quantitative data on health among Canadian SMY aged 14 to 25 years old. When the titles and abstracts did not contain sufficient information, the full-text articles were reviewed. In a third step, we manually searched the reference lists of the papers for other relevant studies, including gray literature. For inclusion, studies had to report specific data (prevalence or odds ratios) on health outcomes among SMY recruited in Canada. Studies based on qualitative data were excluded given the limited space in this review. Following this procedure, in total, 19 papers and 2 research reports were retained for this review.

Quality assessment of research

The quality of research papers was assessed using four criteria proposed by and inspired from the Cochrane Handbook’s general guidance on non-experimental studies. Sampling was coded as non-random/convenient = 0 versus random = 1; representativeness was coded as response rates < 60% = 0 versus 60% or more = 1; population sample was defined as selected = 0 versus population = 1; and sample size was coded as <100 SMY = 0 and ≥ 100 SMY = 1. A score of 0 was given when information was not available. A summary percentage score was calculated for each study identified. The scores vary from 50% (n= 5) to 100% (n= 1), with an average score of 72.9%.

Studies and Participants’ Characteristics

Papers reviewed were based on data collected in British Colombia (n=9), Quebec (n=8), Ontario (n=3) and in the Atlantic provinces (n=1). All papers were based on cross-sectional studies. The vast majority of studies were multicentric (20) and only one study was conducted at a single center (n=1). Sample sizes vary from 53 to 30 588, for a total of 81 567 respondents. Papers based on the same sample (n =7) were counted only once in determining the total respondents.

Most studies (18/21) reported the gender of the participants. Overall, 42.43% of participants were described as male, men or boys and 42.84% were described as female, women or girls. Less than 1% (0.07) of the participants were described as trans, transgender, transidentified or transsexual. About 14% of the participants were described only by their sexual orientation with no gender breakdown. Sexual orientation was reported mostly according to a single dimension: sexual attraction (n= 9), self-definition (n= 9) or sexual behaviors (n= 1). Two studies included participants based on a combination of criteria.

Overall, SMY counted for 15.86% of the total sample (n = 11904; 7.83% non-exclusively heterosexual women; 5.9% non-exclusively heterosexual men). Among men, there were 87.2% of heterosexual men who have sex exclusively with women, or men who reported being sexually attracted only by women. SMY men were described as follows: questioning or unsure = 5.2%; bisexual being attracted by both men and women (including mostly, but not exclusively by women), or having sex with both men and women = 4.2%; and gay, having sex exclusively with other men, or being sexually attracted only by other men = 3.3%. The pattern looked quite similar for females, with 83.2% of women being described as exclusively heterosexual, having sex exclusively with men, or being sexually attracted only by men. SMY women were described as follows: questioning or unsure = 4.9%; bisexual, being attracted by both men and women (including mostly, but not exclusively by women), or having sex with both men and women = 11.7%; and lesbian, having sex exclusively with other women, or being sexually attracted only by other women = 0.2%.

Results

Mental Health Outcomes

The first objective was to document health outcomes among sexual-minority youth in Canada. Firstly, as shown in Table 1, mental health outcomes affected up to 70% of the SMY participants, compared to up to 54% of the heterosexual participants ^(3,5,6,7). Psychological malfunctioning or distress, including psychological maladjustment and feeling under pressure or sad, was the most prevalent problem, affecting 21% to 69.4% of SMY ^(3,5,6,7). Suicidal attempts and ideation were the less frequent outcomes (7% to 63%), although SMY were still 1.61 to 15.06 times more likely to report it compared to heterosexual youth ^(3,6,7,8,9).

Table 1.

Mental Health Outcomes among Canadian SMY

Sources	Sample type	Sample Characteristic	Health outcomes	Statistically significant risk factors (RF) and protective factors (PF)
Psychological functioning/distress
(30)	Cross-sectional; convenience sample	N = 3 876 aged 13–18; Distribution of SO groups (%):EHA = 92. 7; MHA = 3.2; BA = 3.2; SA = 0.9	Psychological functioning OR BA/SA vs. EHA = 0.33 (p < 0.001)
(5)	Cross-sectional; convenience sample	N = 194 aged 14–19; Distribution of SO groups (%): H = 50; LG = 4.12; B = 18.56; Q/U = 27.32	Psychological maladjustment Mean scores (SD) LG and B = 29.96 (12.11); Q/U = 27.71 (9.53); H = 24.83 (7.74)	RF: Bullying, sexual harassment; less companionship with their best friends PF: Closeness with their mother.
(2)	Cross-sectional; web-based survey; convenience sample	N = 300 aged 14–22 Distribution of SO groups (%):LG= 73.7; B=15.3; MH=7.7; Q/U=3.3 ; Trans=13	Self-esteem Std. Coeff. Trans (vs SMY men) = −0.78	RF: Internalized homophobia; homophobic victimization
(19)	Cross-sectional; cluster stratified convenience sample	N = 262 aged 14–22 Distribution of SO groups (%): H=14.5; LG = 37.8; B = 45; Q/U = 2.7 Trans = 9.2	Psychological distress (last week) Mean scores (SD) H = 9.70 (1.41); LG = 11.80 (0.85); B = 15.20 (0.94); Q/U = 23.60 (2.78); Trans= 19.50 (2.49); Cis= 12.70 (0.59)	RF: Homophobic victimization PF: Age; peer support
(20)	Cross-sectional; convenience sample	N = 164 aged 15–25 Distribution of SO groups (%): LG = 85; B = 11; Q/U = 4	Psychological distress (past month) Mean scores (SD) Undisclosed youth = 1.26 (.66) Disclosed youth = 1.29 (.73)	RF: Lower perceived mothers’ support; paternal rejection in childhood; current family members’ rejection.
(7)	Cross-sectional; stratified, probability sample	N 2003 = 30 588 (weighted = 289 412); Aged 12 to 19; Distribution of SO groups (%): MH = 12, B = 4; LG = 2; Q/U = 12	Felt under pressure (past month) Prevalence (%) HM = 30.00; MHM= 40.00; GM= 51.00; BM = 51.00 HW = 42.00; MHW = 58.00; BW = 65.00; LW= 61.00 Felt sad, discouraged, or hopeless (past month) Prevalence (%) HM= 13.00; MHM:= 21.00 ; BM= 29.00; GM= 35.00; HW= 22.00; MHW= 35.00; BW: 50.00; LW= 61.00
(3)	Cross-sectional; weighted cluster-stratified sample	N = 8 194 (weighted = 6540) Aged 14–22; Distribution of SO groups (%): H = 82.6; LG = 1.3; B = 10.6; Q/U = 5.5	Psychological distress (past week) Prevalence (%) HM= 31.20; GM= 43.60; BM= 41.20; Q/UM= 39.90 HW= 54.20; LW= 53.30; BW= 69.40; Q/UW= 49.50
(6)	Cross-sectional; convenience sample	N = 1856 aged 14 or older; Distribution of SO groups (%): H= 87.5; MH= 6.2; LGB= 3.1; Q/U= 3.2	Distressed mood (past year) Prevalence (%) LGB = 62.10; U/Q = 44.10; H = 32.50; MH = 44.30
(21)	Cross-sectional; convenience sample	N = 53 SMY, aged 15–25, and their parents (53) aged 40–67.	Psychological distress (past month) Means score (SD) Men = 1.07 (.64) Women = 1.34 (.73)	RF: Parental attempts to change their sexual orientation.
Suicide ideation/thoughts
(6)	Cross-sectional; convenience sample	N = 1856 aged 14 or older; Distribution of SO groups (%): H = 87.5; MH = 6.2; LGB = 3.1; Q/U = 3.2	Suicide ideation (past year) OR (95% CI) LGB= 2.31 (1.22–4.37); Q/U= 2.64 (1.38–5.08) Men= 0.60 (0.44–0.82)	RF: Depressed mood; physical abuse by family member; sexual abuse by anyone
(7)	Cross-sectional; stratified, probability sample	N 2003 = 30 588 (weighted = 289 412); Aged 12 to 19; Distribution of SO groups (%):MH = 12, B = 4, LG = 2, U/Q = 12	Seriously considered suicide (past year) Prevalence (%) HM= 10.00; MHM= 19.00; GM= 26.00; BM= 38.00; HW= 19.00; MHW= 34.00; BW= 52.00; LW= 63.00
(8)	Cross-sectional; stratified, probability sample	N = 29 315 aged 12 to 19; Distribution of SO groups (%): H=86; MH= 6.7; B= 2.1; LG= 0.7	Suicide ideation (past year) Prevalence (%) HM= 7.70; MHM= 22.90; BM= 37.90; GM= 33.90; HW= 11.50; MHW= 28.70; BW= 43.90; LW= 47.10
(3)	Cross-sectional; weighted cluster-stratified sample	N = 8 194 (weighted = 6540) Aged 14–22; Distribution of SO groups (%):H= 82.6; LG= 1.3; B= 10.6; Q/U= 5.5	Suicide ideation (lifetime) Prevalence (%; 95%IC) HM = 19.40 (17–21.7); GM = 34.40 (18.5–50.4); BM = 34.40 (29–39.9); Q/UM = 21.00 (12.2–29.7) HW = 28.20 (25.5–30.9); LW = 33.50 (17.3–49.8); BW = 50.40 (45.8–55); Q/UW = 26.70 (19.4–34)
(21)	Cross-sectional; convenience sample	N = 53 sexual minority youth aged 15–25 years old and their parents (53) aged 40–67 years old.	Suicide ideation (past year) Prevalence (%) Men = 25.00 Women = 24.10	PF: Parental support. RF: Father struggles with his child’s sexual orientation
(9)	Cross-sectional; cluster stratified random sample	N = 734 (weighted = 6905); Aged 12–19; Distribution of SO groups (%): BA women= 90.3; Men = 58.3; Others NHSO women = 9.7; men = 41.7	Suicide ideation (past year) Prevalence (%) SMY boys= 33.30	RF: Coming from a rural area for boys
Suicidal Attempts
(6)	Cross-sectional; convenience sample	N = 1856; aged 14 or older; Distribution of SO groups (%): H = 87.5; MH = 6.2; LGB = 3.1; Q/U = 3.2	Suicide attempts (past year) OR (95%CI) LGB= 2.23 (1.15–4.35) Men= 0.62 (0.42–0.90)	RF: Depressed mood; fighting; physical abuse.
(22)	Proxy-based interviews; cross-sectional;	N = 55; aged 11–18; 4 people were found to have NHSO or expressed concern regarding their sexual orientation.	SO among suicide completers Prevalence (%) SA = 7.27	RF: Anxiety disorders; having consulted a psychiatrist in the year preceding suicide
(9)	Cross-sectional; cluster stratified random sample	N = 734 (weighted = 6905); Aged 12–19; Distribution of SO groups (%): BA women = 90.3; BA men = 58.3; Others NHSO: Women = 9.7; Men = 41.7	Suicide attempts(past year) Prevalence (%) SMY boys = 1.20	RF: Coming from a rural area for boys
(7)	Cross-sectional; stratified, probability sample	N 2003 = 30 588 (weighted = 289 412); Aged 12 to 19; Distribution of SO groups (%): MH = 12, B = 4, LG= 2, U/Q = 12	Suicide attempts (past year) Prevalence (%) HM = 3.00; MHM=7.00; BM=13.00; GM=9.00 HW=8.00; MHW=17.00; BW=30.00; LW=38.00
(8)	Cross-sectional; stratified, probability sample	N = 29 315 aged 12 to 19; Distribution of SO groups (%): H=86; MH= 6.7; B= 2.1; LG=0.7	Suicide attempts (part year) Prevalence (%) HM= 2.40; MHM= 9.40; BM= 26.80; GM= 24.10 HW= 5.00; MHW= 12.50; BW= 29.00; LW=25.70 OR (95%CI) MHM= 4.23 (3.08–5.80); BM = 15.06 (10.34–21.95); GM = 13.07 (8.41–20.30) MHW = 2.89 (2.41–3.48); BW = 8.14 (CI = 6.54–10.12); LW = 7.15 (4.15–12.32) Serious suicide attempts (past year) Prevalence (%) HM = 3.70; MHM = 11.60; BM = 33.10 = GM: 33.80 HW = 6.50; MHW = 10.30; BW = 23.90; LW = 51.70	RF: Physical abuse; sexual abuse; physical victimization at school; Internet victimization; physical discrimination; sexual orientation discrimination; physical sexual harassment; verbal sexual harassment PF: Self-esteem; family connectedness; school connectedness; school safety

Notes: EHA = exclusive heterosexual attraction; BA = bisexual attraction; MHA = Mostly heterosexual attraction; SA = same-sex attraction; H = heterosexual; MH = Mostly heterosexual; B = bisexual; LG = Lesbian or Gay; Q/U = questioning/unsure; LGB = Lesbian, Gay and Bisexual. HM= heterosexual men; HW= heterosexual women; BM = bisexual men; BW= bisexual women; LW= lesbian women; GM = gay men; MHM = mostly heterosexual men; MHW = mostly heterosexual women; NHSO = non heterosexual sexual orientation; SMY = Sexual Minority Youth; OR = odds ratio.

^*p < 0.05;

^**p <0.01;

^***p < 0.001.

Regarding gender and sexual orientation variations among SMY, as is the case in the general non-SMY population, SMY women were more likely to report psychological distress compared to SMY men in the two studies testing for gender differences (35% to 69.4% vs. 21% to 51%) ^(3,7). Moreover, SMY women also reported higher prevalence of suicidal ideation (26.7% to 63% vs. 19% to 38%) and suicide attempts (10.3% to 51.7% vs. 7% to 33.8%) than SMY men in studies ^(3,7,8). Results on bisexual male youth suggest that they are more likely to report suicidality than other SMY males ^(3,7,8). Among female SMY, lesbian were more likely to report suicidality than other SMY subgroups ^(3,7,8). For psychological distress, lesbians and gays report the highest prevalence in all subgroups of sexual orientation^(3,6,7).

Physical Health Outcomes

Substance use, tobacco use, alcohol consumption, and being over/underweight were the four main physical health outcomes documented (see Table 2). Regarding drug use, SMY were 3.29 to 26.28 times more likely to have used “club” drugs such as ketamine, crystal methamphetamine and ecstasy ⁽¹⁰⁾. Across studies, marijuana use was reported among 21% to 62% of the SMY compared to 3.29% to 26.28% of the heterosexual respondents ^{(6,7,11–15)}. While two out of three studies suggest that SMY men have a higher prevalence of marijuana use ^(12,14), SMY women were more likely to have used drugs other than marijuana compared to SMY men in three out of three studies (W: 33.9% to 63%; M: 25.9% to 59.2%) ^(9,11,12). Among SMY, two studies (out of three) suggest that lesbian and gay show higher prevalence of hard drug use in the previous year ^(15,16), while others suggests that bisexual youth have a greater likelihood of marijuana use compared to other SMY ^(7,15).

Table 2.

Physical Health Outcomes among Canadian SMY

Sources	Study Design	Sample Characteristics	Health Outcomes	Risk Factors (RF)/Protective Factors (PF)
Drug use
(9)	Cross-sectional; cluster stratified random sample	N = 734 (weighted = 6905); Aged 12–19; Distribution of SO groups (%): BA women = 90.3; BA men = 58.3; Others NHSO: Women = 9.7; Men = 41.7	Other drug use (past month) Prevalence (%) LGB men= 35.10; LGB women = 49.50
(11)	Cross-sectional; cluster stratified random sample	N (weighted) = 5423; Distribution of SO groups (%):MHA men= 4.4; women: 7.7; LGB men= 1.6; women= 3.6.	Others drugs use (past year) Prevalence (%) LGB boys = 59.20; LGB girls = 63 Substance use problem (past year) Prevalence (%) LGB boys = 31.10; LGB girls = 20.90	RF: Enacted stigma; sexual abuse history PF: Sports involvement; school connectedness
(12)	Cross-sectional; Cluster-stratified random sampling; Included only East or Southeast Asian	N (weighted) = 51 349 aged 12–19; Distribution of SO groups (%):EHA= 79.2 MHA= 6.1; LGB= 2.6	Marijuana use (lifetime) Prevalence (%) GB men= 41.30; LB women= 21 EHA men= 18.87 and EHA women : 14.60% Other drugs use (lifetime) Prevalence (%) GB men= 25.90; LB women: 33.90 EHA men= 14.60 and EHA women= 15.
(7)	Cross-sectional; stratified, probability sample	N 2003 = 30 588 (weighted = 289 412); Aged 12 to 19; Distribution of SO groups (%): MH = 12, B = 4, LG = 2, U/Q = 12	Marijuana use (lifetime) Prevalence (%) BA men= 53; SA men= 40; BA women= 62; SA women= 51. EHA women= 39
(6)	Cross-sectional; Convenience sample	N = 1856 aged 14 or older; Distribution of SO groups (%): H = 87.5; MH = 6.2; LGB = 3.1; Q/U = 3.2	Hard drug use (past month) Prevalence (%) LGB= 17.20; MH= 13; U/Q= 23.70; H= 6.20 Marijuana use (past month) Prevalence (%) LGB= 50; MH= 35.70; U/Q= 37.30; H= 26.50
(10)	Cross-sectional ; Convenience sample	N= 509 aged 13–19; Distribution of SO groups (%):LGB= 2.14	Hard drug use (past year) OR (95%CI) for LGB Crystal Methamphetamine = 26.28 (6.13–112.57) Ecstasy = 3.29 (.98–11.05) Ketamine = 8.26 (1.98–34.34)
(15)	Cross-sectional; Convenience sample	N = 3876, aged 13–18 Distribution of SO groups (%):EHA = 92.7; MHA = 3.2; BA = 3.2; SA = 0.9.	Hard drugs use (past year) Prevalence (%) BA= 34; SA= 44; MHA= 16; EHA= 12 Marijuana use (past year) Prevalence (%) BA= 43; SA= 33; MHA= 25; EHA= 21
(20)	Cross-sectional; Convenience sample	N = 164 aged 15–25 Distribution of SO groups (%): LG = 85; B = 11; Q/U = 4	Drug use (past year) Prevalence (%) Undisclosed men = 63 ; women = 90 Disclosed men = 50; Women = 66	PF: Disclosure of sexual orientation to parents; RF: Current family rejection of sexual orientation.
(13)	Cross-sectional; Convenience sample	N = 1126, aged 18–25 Distribution of SO groups (%): GM = 84; Others NHSO = 16.	Drug use (lifetime) Prevalence (%) for GM Marijuana use= 46.80; ecstasy= 18; amphetamines/speed= 23.2; poppers = 17.70; Viagra= 3.40. Consumed at least one drug= 56.20
(14)	Cross-sectional; stratified, probability sample	N = 29 315 aged 12 to 19; Distribution of SO groups (%):H=86; MH= 6.70; B= 2.10; LG= 0.70	Heavy marijuana used (past 20 days or more) Prevalence (%) LGB men =29.8; H= 15.50; LGB women= 13.50; H=6.30
(16)	Cross-sectional; Convenience sample	N =680 aged 13–18; Distribution of SO groups (%): H= 25; LG= 25; B= 25; Q/U= 25	Substance use (past year) Prevalence (%) LG= 17; B= 3; Q/U= 2; H= 0.40
Alcohol Use
(7)	Cross-sectional; stratified, probability sample	N 2003 = 30 588 (weighted = 289 412); Aged 12 to 19; Distribution of SO groups (%): MH = 12, B = 4, LG = 2, U/Q = 12	Binge-drinking (lifetime) Prevalence (%) SA men = 28; BA men= 32 BA women= 43; SA women= 27
(14)	Cross-sectional; stratified, probability sample	N = 29 315 aged 12 to 19; Distribution of SO groups (%):H=86; MH= 6.70; B= 2.10; LG= 0.70	Frequent binge drinking (past month) Prevalence (%) LGB men= 20; H= 9.30 LGB women= 12.20; H= 7.40
(9)	Cross-sectional; cluster stratified random sample	N = 734 (weighted = 6905); Aged 12–19; Distribution of SO groups (%): BA women = 90.3; BA men = 58.3; Others NHSO: Women = 9.7; Men = 41.7	Binge drinking (past month) Prevalence (%) Boys= 30.70; Girls = 41.40	RF: Rural boys = 2.72 (1.95–3.79); rural girls = 1.39 (1.15–1.69)
(13)	Cross-sectional; Convenience sample	N = 1126, aged 18–25 Distribution of SO groups (%): GM = 84; Others NHSO = 16.	Consumption of 5+ glasses of alcohol/week Prevalence (%) GM= 9.30
(12)	Cross-sectional; Cluster-stratified random sampling; Included only East or Southeast Asian	N (weighted) = 51 349 aged 12–19; Distribution of SO groups (%):EHA= 79.20 MHA= 6.10; LGB= 2.60	Ever used alcohol Prevalence (%) GB men= 60.10; MHA men= 44.50; HM= 41.90 LB women: 54.70; MHA women= 51.70; H= 38 OR (95%CI) GB men= 1.75 (1.36–2.25); MHA men= 1.19 (1.02–1.38) LB women= 2.01 (1.68–2.40); MHA women= 1.90 (1.67–2.15)
(15)	Cross-sectional; Convenience sample	N = 3876, aged 13–18 Distribution of SO groups (%):EHA = 92.70; MHA = 3.20; BA = 3.20; SA = 0.90	Ever used alcohol Prevalence (%) BA= 50; H= 34; SA= 42; MHA= 42
(6)	Cross-sectional; Convenience sample	N = 1856 aged 14 or older; Distribution of SO groups (%): H = 87.5; MH = 6.2; LGB = 3.1; Q/U = 3.2	Ever used alcohol Prevalence (%) LGB= 77.60; MH= 62.60; U/Q= 50.80; H= 53.80
Tobacco use
(17)	Cross-sectional; two-stage stratified cluster sample;	N = 5994;Distribution of SO groups (%): H= 93.80; B= 4.40; LG= 1.80; LGB= 6.70	Lifetime cigarette use Prevalence (%) LGB= 16; H= 8 OR (95%CI) LB= 2.63 (1.83–3.78) Daily smoking OR (95%CI) LGB= 2 (1.35–2.58)
(7)	Cross-sectional; stratified, probability sample	N 2003 = 30 588 (weighted = 289 412); Aged 12 to 19; Distribution of SO groups (%): MH = 12, B = 4, LG = 2, U/Q = 12	Smoking (lifetime) Prevalence (%) BA men= 12; SA men= 20; BA women= 22; SA women= 7
(15)	Cross-sectional; Convenience sample	N = 3876, aged 13–18 Distribution of SO groups (%):EHA = 92.70; MHA = 3.20; BA = 3.20; SA = 0.90	Daily cigarette smoking Prevalence (%) SA= 25; MHA= 23; BA= 38; EHA= 12
Over/underweight
(7)	Cross-sectional; stratified, probability sample	N 2003 = 30 588 (weighted = 289 412); Aged 12 to 19; Distribution of SO groups (%): MH = 12, B = 4, LG = 2, U/Q = 12	Overweight Prevalence (%) GM= 28; LW= 23; HM= 23; HW= 11 Underweight Prevalence (%) GM= 9; LW= 11; HM= 4; HW= 4

Notes: EHA = exclusive heterosexual attraction; BA = bisexual attraction; MHA = Mostly heterosexual attraction; SA = same-sex attraction; H = heterosexual; MH = Mostly heterosexual; B = bisexual; LG = Lesbian or Gay; Q/U = questioning/unsure; LGB = Lesbian, Gay and Bisexual. HM= heterosexual men; HW= heterosexual women; BM = bisexual men; BW= bisexual women; LW= lesbian women; GM = gay men; MHM = mostly heterosexual men; MHW = mostly heterosexual women; NHSO = non heterosexual sexual orientation; SMY = Sexual Minority Youth; OR = odds ratio.

^*p < 0.05;

^**p <0.01;

^***p < 0.001.

Prevalence and odds ratios for alcohol use reveal a similar pattern. SMY were more likely to have ever used alcohol (44.5% to 60.1% vs. 34 to 41.9%) ^(6,12,15) and to have engaged in binge drinking (12.2% to 43% vs. 7.4% to 9.3%) ^(7,14) compared to heterosexual youth. However, mixed results emerged in regards to gender variations. Rates of binge drinking were higher among women than men in two studies ^(7,9), but the opposite was reported by Konishi et al, (2013). Moreover, lifetime prevalence of alcohol drinking was higher in men in general (Men = 41.9% to 60.1%; Women= 38% to 54.7%) ⁽¹²⁾. Two studies compared SMY subgroups and reported that bisexual youth were more likely to report alcohol consumption (32% to50% vs. 27% to 42%) ^(7,15).

Regarding tobacco use, SMY were 2 to 3.5 times more likely to be smokers and 7% to 38% of them were smoking cigarettes compared to 8% to 12% of heterosexual youth ^(7,15,17). None of the three studies addressing tobacco revealed gender differences in use (12% to 20% among men; 7% to 22% among women). Only one study examined problems relating to being under/overweight, revealing that SMY men were more likely to be overweight (23% to 28% vs. 11% to 23%) and SMY women, underweight (9% to 11% vs. 4%) compared to their heterosexual peers ⁽⁷⁾. No differences were reported between SMY subgroups regarding being under/overweight and tobacco use.

Sexual Health Outcomes

Condomless sex and pregnancy involvement were the two main sexual outcomes measured in the studies reviewed (see Table 3). SMY are more likely to have condomless sex than heterosexual youth (24.5% to 60.5% vs. 12% to 34%) ^(7,13,15,18). Moreover, SMY were 1.28 to 3.47 times more likely to report condomless sex at the last sexual event ⁽¹⁸⁾. In a sample of gay men aged 18 to 25 years old, a study reported a prevalence of condomless intercourse with a serodiscordant partner or an HIV-unknown partner varying from 17.9% to 24.9% ⁽¹³⁾. There was no clear gender difference between women and men regarding condomless sex, although it seems to be higher among women (34.2% to 60.5% vs. 25.3% to 53.5%) ^(7,15,18). One reason why women reported somewhat higher rates of condomless intercourse may be that lesbian and bisexual women’s last intercourse occurred with a female partner, which was not always specifically asked for in the studies who include SMY, rather than specifically targeting this population and adjusting the questions accordingly. Regarding sexual orientation variations, bisexual men seem to have higher prevalence rates of condomless sex than other subgroups ^(7,15,18).

Table 3.

Sexual Health Outcomes among Canadian SMY

Sources	Sample type	Sample Characteristics	Health outcomes	Statistically significant risk factors (RF) and protective factors (PF)
Condomless sex
(15)	Cross-sectional; Convenience sample	N = 3876, aged 13–18 Distribution of SO groups (%): EHA = 92.7; MHA = 3.2; BA = 3.2; SA = 0.9.	Condomless sex (last intercourse) Prevalence (%) BA= 38; SA= 31; EHA= 12
(13)	Cross-sectional; Convenience sample	N = 1126, aged 18–25 Distribution of SO groups (%): GM = 84; Others NHSO = 16.	Condomless intercourse (past 12 months) Prevalence (%) With a casual partner = 44.20; With a serodiscordant or HIV unknown partner = 24.90; In a relationship with a serodiscordant or HIV-unknown = 17.90	RF for condomless anal intercourse with a casual partner: Physical aggression; sensation seeking; over 10 sexual partners in the last year; marginal sexual activities (e.g. BDSM, fist fucking)
(7)	Cross-sectional; stratified, probability sample	N 2003 = 30 588 (weighted = 289 412), aged 12–19; Distribution of SO groups (%): MH = 12; BS = 4; LG = 2; Q/U = 12	Condom use (last intercourse) Prevalence (%) BM = 46; MHM = 62; G= 75; H= 74; BW= 62; MHW = 52; L= 39; HW= 66
(18)	Cross-sectional; cluster-stratified random sampling	N 2003 = 30 588 (weighted = 289 412), aged 12–19; Distribution of SO groups (%): MH = 12; BS = 4; LG = 2; Q/U = 12	Condomless intercourse (last intercourse) Prevalence (%) HM = 25.60; GM = 25.30; BM = 53.50; HW = 34.20; LW = 60.50; BW = 38.10 OR (95% CI) BW = 1.28 (1.17 1.40); LW = 2.94 (2.02 4.28); BM = 3.47 (2.92 4.12); GM = N/A
Pregnancy involvement
(18)	Cross-sectional; cluster-stratified random sampling	N 2003 = 30 588 (weighted = 289 412), aged 12–19; Distribution of SO groups (%): MH = 12; BS = 4; LG = 2; Q/U = 12	Pregnancy involvement (lifetime) Prevalence (%) HM= 5.30; BM= 16.80; GM= 16.80; HW= 5.30; LW= 12.60; BW= 8.80 OR (95% CI) BM= 3.61 (2.86–4.56); GM= 3.56 (2.58–4.92); LW= 2.63 (1.55–4.44); BW= 1.81 (1.55–2.10)	RF for LBW and BM: Experiencing discrimination in the past year based on sexual orientation; greater number of types of harassment and discrimination.
(9)	Cross-sectional; cluster stratified random sample	N = 734 (weighted = 6905); Aged 12–19; Distribution of SO groups (%): BA women= 90.3; Men = 58.3; Others NHSO women = 9.7; men = 41.7	Pregnancy involvement (lifetime among sexually experienced youth) Prevalence (% among sexually active youth) Boys: 16.80	RF: Coming from a rural area for boys
(7)	Cross-sectional; stratified, probability sample	N 2003 = 30 588 (weighted = 289 412), aged 12–19; Distribution of SO groups (%): MH = 12; BS = 4; LG = 2; Q/U = 12	Pregnancy involvement (life time among sexually experienced youth) Prevalence (%) HM = 5; MHM = 12; BM = 17; GM = 17 HW = 5; MHW = 7; BW = 9; LW = 13

Notes: EHA = exclusive heterosexual attraction; BA = bisexual attraction; MHA = Mostly heterosexual attraction; SA = same-sex attraction; H = heterosexual; MH = Mostly heterosexual; B = bisexual; LG = Lesbian or Gay; Q/U = questioning/unsure; LGB = Lesbian, Gay and Bisexual. HM= heterosexual men; HW= heterosexual women; BM = bisexual men; BW= bisexual women; LW= lesbian women; GM = gay men; MHM = mostly heterosexual men; MHW = mostly heterosexual women; NHSO = non heterosexual sexual orientation; SMY = Sexual Minority Youth; OR = odds ratio.

^*p < 0.05;

^**p <0.01;

^***p < 0.001.

Similarly, both female and male SMY were more likely to be involved in a pregnancy than heterosexual youth (7–17% vs. 5%), making them 1.81 to 3.61 times more likely than heterosexual youth of having participated in a pregnancy/been pregnant ^(7,18). Both SMY women and men reported a similar range of involvement in a pregnancy (5% to 13% of women vs. 5% to 17% of men) ^(7,18). Possible explanations are lower contraceptive use, unplanned sexual involvement with opposite-gender partners, engaging in heterosexual sexual behaviours or choosing pregnancy involvement to avoid being identified as SMY and targeted for homophobia, or a lack of sexual education that properly engages SMY by addressing their needs ⁽¹⁸⁾. Surprisingly, bisexual youth did not report a higher prevalence of pregnancy involvement compared to other sexual orientation subgroups ^(7,9,18). These can be due to measurement issues in the studies reviewed here, which have classified mostly homosexual youth in the same group as exclusively homosexual youth.

Risk and Protective Factors for Health Problems among Canadian SMY

The second study objective was to identify sexual-minority-specific risk and protective factors. Results show that lifetime risk factors for lower psychological functioning were cyber- and homophobic bullying ^(2,3), victimization ⁽⁵⁾, parental rejection, lower support from parents and/or peers, including parental attempts to change their child’s sexual orientation and lack of connectedness with significant others ^(5,19,20,21), internalized homophobia ⁽²⁾ and being younger ⁽¹⁹⁾. Suicidality was significantly associated with anxiety disorders ⁽²²⁾, depressed mood, fighting, physical and sexual abuse ⁽⁶⁾, living in rural areas (boys only) and fathers struggling with their child’s sexual orientation ⁽²¹⁾. Conversely, parental support ⁽²¹⁾, self-esteem, family and school connectedness as well as school safety were factors significantly protecting SMY from suicidal ideation and attempts ⁽⁸⁾.

Risk factors for substance use were high levels of stigmatization and sexual abuse ⁽¹¹⁾, disclosure of sexual orientation to the parents and family rejection of sexual orientation ⁽²⁰⁾, while sports involvement and school connectedness act as protective factors. Concerning binge drinking, living in a rural area was significantly associated with higher odds of binge drinking ⁽⁹⁾. No risk nor protective factors were identified for tobacco use and over/underweight problems.

Risk factors for condomless sex were reporting homophobic physical aggression, marginal sexual activities (e.g. BDSM, fist fucking), more than 10 casual sexual partners, and sexual sensation seeking ⁽¹³⁾. Pregnancy involvement was significantly associated with living in a rural area for boys ⁽⁹⁾, and, for lesbians and both bisexual women and men, reporting homophobic harassment and discrimination ⁽¹⁸⁾. No protective factor was identified for condomless sex and pregnancy involvement in the studies reviewed.

Discussion

This review aimedat documenting health outcomes among SMY in Canada and identifying sexual-minority-specific risk and protective factors. The available evidence regarding the first objective showed that SMY in Canada report negative health outcomes in proportions varying from 7% to 69.4%. SMY are more likely than their heterosexual peers to report psychological distress/malfunctioning, suicidality, substance misuse, condomless intercourse, pregnancy involvement. Gender variations suggest that SM men are particularly at-risk of being overweight, while women tend to be overrepresented in suicidality, psychological malfunctioning, use of hard drugs, binge drinking, and condomless sex. By their position in interacting systems of oppression (heterosexism, sexism and racism), some SMY subgroups, including women and ethnic minorities, face increasing odds of experiencing prejudice and negative health outcomes. Despite contradictory results, bisexual youth appear more likely to report higher prevalence of negative health outcomes. higher prevalence of negative health outcomes.

Canadian trends on SMY health are consistent with results from around the world. Marshal et al (2011), in their meta-analytic review, reported that SMY had higher rates of suicidality and depression disparities than heterosexual, and that bisexual youth reported higher scores of suicidality, which is in tune with the Canadian data. Moreover, those disparities may be affected by victimization and discrimination based on sexual orientation. Another meta-analysis of 25 international population based studies found that LGB respondents are at greater risk of mental disorder, suicidality, substance misuse/dependence than heterosexuals ⁽²⁴⁾. Similar results showing higher suicidality among SM have been reported in Denmark, New-Zealand and the Netherlands ⁽²⁵⁾.

The evidence regarding our second objective help to document and explain the challenges faced by SMY that contribute to these negative health outcomes. Among the main SMY-specific risk factors were family rejection of one’s minority sexual orientation, homophobic bullying and victimization, and internalized homophobia. Among the few protective factors explored, school/family connectedness, school safety, parental support and sports involvement seem to reduce the risk of negative health outcomes among the general youth population. Overall, this review highlights that few studies investigate risk and, more particularly, protective factors, let alone SMY-specific factors such as gender and sexual orientation integration.

The research included in this review suffer limitations worth mentioning. First, methodological challenges pertaining to the lack of uniformity in defining sexual minorities and sexual orientation and to the nonprobability samples affected most of the studies. Few studies share a common ground regarding sexual orientation measurement and only one used a multidimensional definition, which is more accurate in capturing respondents with minority sexual orientation, especially SMY. As such, some studies only assessed self-identification, which don’t take into account that sexual orientation can be fluid ⁽²⁶⁾, in the sense that it is not a fixed lifetime characteristic and that some people self-identifying as heterosexual can be attracted to same-sex people or have same-sex sexual behavior. Still, it is not clear yet which measure is more precise in capturing differences in health outcomes of sexual minority youth. For example, Zhao et al, (2010) did not find significant differences between heterosexual without same-sex attraction/fantasy or behavior and heterosexual with same-sex fantasy/attraction or behavior. Those limitations in the definition and measurement of sexual orientation should be taken into account when interpreting this review’s findings.

While non-probability sampling is more convenient for recruiting hidden population and costs less, this procedure may oversample people who are comfortable with their gender or sexual orientation variation, while underrepresenting those who prefer to stay hidden because of fear of stigma or a history of discrimination. New methodologies are developing and are promising in selection biases correction, such as respondent-driven sampling ^(27,28), even though they remain underused. Also, studies on SMY rarely address problems that are yet common in adolescence and emerging adulthood or that commonly arise at this time, such as body-image issues, eating disorders, behavioral problems or schizophrenia. Most of the studies reviewed assessed psychological distress or substance used and only one looked at issues of being under/overweight, but didn’t assess body image issues/satisfaction or eating disorders. Third, due to our inclusion and exclusion criteria (especially the fact that studies had to report prevalence or odds ratio of health outcomes), it is possible that many studies documenting the risk and protective factors in Canadian SMY were not included in this review. Nevertheless, most of the studies reviewed here have not assessed sexual-minority-specific risk and protective factors. As such, it is clear that documenting those factors needs to be one of the priorities of future research on SMY.

Recommendations for Intervention and Research

While social and political recognition of gender and sexual orientation variations has increased in Canada, SMY still face challenges such as school and peer connectedness, victimization, bullying and family rejection that negatively impact their health. The health outcomes faced by SMY and their persistence through adulthood have a high financial and social cost for Canada on the long term. To support the health of SMY, we need to target multiple environments (social policies, communities and schools, families, individuals) and facilitate linkages between them (see Table 4 for details).

Table 4.

Recommendations for intervention on SMY

Intervention	References
Policy level
Promote antibullying policies and implement SMY-sensitive, nonjudgmental programs that explicitly target homonegativity and homophobia in schools, communities and society. Implement policies and public campaigns that promote intolerance for discrimination, verbal or physical victimization, sexual harassment.	8; 2; 19; 30; 9; 11; 18; 7; 5.
Community and schools level
Adapt health intervention and promotion to location and developmental and cultural differences among SMY, adding additional support and services where needed.	17; 30; 9; 11; 7.
Provide training in sensitivity for gender and sexual variations for educators and clinicians in mental health services. Skills should cover questioning all youth about gender and sexual orientation, screening for environment adversity, distress and suicidality among SMY, and addressing disclosure - and report of - abuse (by peers, parents or romantic and sexual partners).	22; 18; 5; 6.
Implement early interventions aiming at protective factors for healthy development of SMY and homophobia management, such as self-esteem enhancement, coping skills acquisition (e.g. social support seeking), and social skills for peer and support group integration.	8; 19; 30; 20; 9; 11; 7; 5.
Family level
Promote family intervention aiming at family connectedness and support, acceptance of sexual diversity and parental implications with schools. Include both mother and father to reinforce their acceptance of their SMY.	8; 20; 18; 5.
Individual level
Promote inclusive and SMY-specific sexual education, including factual information about sexual orientation development and LGB-related sexual health issues such as safer sex, especially for male SMY, smoking and substance use, sexual abuse and healthy relationships.	17; 12; 10; 13; 9; 18.

Also, in order to improve the health on SMY, we need to better understand their lives, a goal that can be reached by improving research on SMY. First, there is a need to incorporate inclusive and comprehensive measures of gender and sexual orientation as well as sexuality-based prejudice in general health surveys. Second, the quality of research designs could be improved by multiplying sources of information (other than self-reports), using multivariate statistics and longitudinal design, developing innovative recruitment strategies to reach SMY to increase statistical power for underrepresented subgroups (e.g. transidentified youth, racialized SMY) and reinforcing the design of the qualitative studies to explore the meanings and the complex forms of experiences specific to SMY realities. Third, efforts should be made to fully understand the process of coming out in relation to social climate and culture of violence regarding gender and sexual variations in different environments, taking into account the social positions that intersect with gender and sexual variations, such as socioeconomic status and ethnic origin. Fourth, there is a need to deepen our understanding regarding protective and resilience factors specific to SMY, a research area that is still underdeveloped compared to the risk factors. Five, surveys should cover a wider range of issues among SMY, including non-intuitive ones (e.g. pregnancy involvement) and SMY-specific psychosexual development tasks (e.g. ages at awareness of same-sex sexual attractions, at self-identification as SMY and first sexual contact with same-sex partners, as well as the sexual scripts considered to be the markers of the transition to sexual activity for all subgroups). Finally, the characteristics and the impact of health services on SMY health should be documented in order to identify and develop promising and best practices. Table 5 provides details for these recommendations.

Table 5.

Recommendations for research on SMY

Recommendations	References
Incorporate measures of gender and sexual orientation as well as sexuality-based prejudice in general health surveys: sexual attraction, self-identified gender and sexual orientation, partners’ gender; prejudice based on gender nonconformity and sexual orientation, including among heterosexual youth	18.
Increase the quality of research among SMY: 1) multiply sources of information besides self-report (e.g. parents, friends, teachers, school records). 2) Use multivariate statistics to examine homophobic-based abuse among SMY, including gender and sexual orientation variations within SMY and controlling for potential confounders for homophobia in multivariate analyses, including social desirability. 3) Implement longitudinal design to examine trends in health outcomes and better understand the lives of youth reporting varying types of same-sex attraction unfold over time in relation to their developmental context, such as sexual orientation development among questioning youth, the dynamic of parental acceptance and rejection and its effects throughout adolescents’ development and the relationships between homophobic victimization and school achievement. 4) Develop recruitment strategies for hard to reach SMY. 5) Increase the statistical power for underrepresented subgroups: undisclosed SMY; questioning, mostly heterosexual, non-exclusively heterosexual youth, bisexual, and transidentified youth; racialized groups, including two-spirit, indigenous and first nations, SMY from urban centers, suburb and remote area, SMY in communities as well as clinical settings. 6) Reinforce the design of the qualitative studies to explore the meanings of sexual behavior, intentions, pregnancy experiences among SMY; complex forms of parental reactions.	8; 19; 15; 20; 18; 5; 12; 9; 11; 22; 5; 6.
Better understand the process of coming out in relation to social climate and culture of violence regarding gender and sexual variations in different environments (e.g. urban, suburban, rural and remote area, school), taking into account the social positions that intersect with gender and sexual variations (e.g. race/colonialism, socioeconomic status/classicism, sex/sexism) and factors that predict homophobic bullying (e.g. gender nonconformity).	9; 2; 12.
Document protective and resilience factors among SMY, such as coping strategies (e.g. social support, academic strength, belonging to LGBT associations), reasons why they choose to live and characteristics leading SMY to express their identity in an adverse environment.	8; 19; 6.
Better understand risk or protective factors of health outcomes, such as social support, gay-straight alliance effect; internalized homophobia, concealment of sexual orientation; physical and sexual abuse, discrimination and harassment; variations in sexual orientation and gender identity.	8; 22; 6; 20; 12; 10; 11.
Cover a wider range of issues in surveys among SMY, including pregnancy involvement, suicidality, sexual health behaviors, degree of psychosexual development (e.g. age at first awareness of sexual status, self-identification as SMY, first sexual contact with same-sex partners).	12; 18; 11.
Assess the characteristics and the impact of health services on SMY: barriers in accessing health services; influence of supportive sexual health education; impact of supportive school environment on health disparities for SMY.	8; 18.

Conclusion

Despite its limitations, this review highlights that Canadian SMY are more likely than their heterosexual peers to experience negative health issues. However, as Saewyc (2011) reminds us, most SMY transition successfully to adulthood, with similar levels of health and well-being than their heterosexual peers, despite facing homophobic prejudice during adolescence. The recommendations for research and for intervention emerging from the reviewed studies can contribute to establishing guidelines for better practices. As such, a better understanding of the lives of SMY is crucial for designing better intervention.