TABLE 1.

Summaries of Studies Included in This Review

Source	Intervention Description	Intervention Length	Evaluation Design	Sampling and Recruitment	Inclusion Criteria	Sample Characteristics	Outcome Measures	Results
Costa et al27	Aimed at relieving distress associated with puberty development in adolescents with gender dysphoria, the authors of this study examined the effects of psychological-only and psychological and puberty suppression interventions on the psychosocial functioning of youth. These interventions were delivered following guidelines set by the WPATH Standards of Care. Psychological support was provided to youth and their families (both together and separately) to support them through the early recognition and nonjudgmental acceptance of the gender identities of youth and ameliorate any behavior, emotion, or relationship problems. A variety of psychotherapeutic approaches were used and sometimes included social and educational interventions. Puberty suppression was provided by using GnRH analogs.	Immediately after baseline assessment, all youth received psychological support during the entire duration of the study at least once per mo. Nine mo. (on average) after baseline assessment, puberty suppression was initiated for the psychological and puberty suppression intervention group.	This was a nonrandomized comparison group pretest-posttest design using multiple posttests. Youth were assessed at baseline, 6, 12, and 18 mo (corresponding to time 0–3). At time 0, no intervention had taken place. At time 1, all participants had received psychological support. At times 2 and 3, some participants had received only the psychological intervention (ie, the psychological-only intervention group), and some participants had received psychological and puberty suppression interventions (ie, the psychological and puberty suppression intervention group). Participants were placed in the psychological and puberty suppression intervention group if they had a presence of gender dysphoria from early childhood on, an increase in gender dysphoria after their first puberty changes, an absence of psychiatric comorbidity that interferes with the diagnostic workup or treatment, adequate psychological and social support during treatment, and a demonstration of knowledge and understanding of the effects of puberty suppression, crossgender hormone treatment, surgery, and the social consequences of gender affirmation surgery. Otherwise, participants were placed in the psychological-only intervention group.	Youth were recruited from a population of youth with gender dysphoria who were referred to a gender identity clinic in London, England, from 2010 to 2014. All youth who completed the standard-of-care diagnostic assessments (∼6 mo after entry into the clinic) were invited to take part in the study.	At baseline, all participants were diagnosed with gender identity disorder (per the DSM-IV-TR criteria). Youth and their parents gave informed consent.	The total sample contained 201 youth; 101 youth were in the psychological-only intervention group, and 100 youth were in the psychological and puberty suppression intervention group. The 2 intervention groups did not differ with regard to natal sex, age, living arrangement, and education. Total sample: N = 201; mean age: 15.52 y at baseline (range: 12–17 y); assigned birth sex: 62.2% female and 37.8% male.	Psychosocial Functioning: Children’s Global Assessment Scale	In the total sample, compared to time 0 (57.73), psychosocial functioning significantly improved at time 1 (60.68; P < .001), time 2 (63.31; P < .001), and time 3 (64.93; P < .001).
								The psychological-only intervention group did not significantly differ from the psychological and puberty suppression intervention group in psychosocial functioning at any time point (P range: .14–.73).
								Among the psychological-only intervention group, compared to time 0 (56.63), psychosocial functioning significantly improved at time 1 (60.29; P = .05), time 2 (62.97; P = .005), and time 3 (62.53; P = .02). However, psychosocial functioning was not significantly different for time 1 vs 2 (P = .22), time 1 vs 3 (P = .37), or time 2 vs 3 (P = .88).
								Among participants of the psychological and puberty suppression intervention group, compared to time 0 (58.72), psychosocial functioning did not significantly differ at time 1 (60.89; P = .19) but was significantly higher at time 2 (64.70; P = .003) and time 3 (67.40; P < .001). Although psychosocial functioning significantly improved from time 1 vs 3 (P = .001), there were no significant differences for time 1 vs 2 (P = .07) and time 2 vs 3 (P = .35).
de Vries et al28	Aimed at enabling youth with gender dysphoria to explore their gender identity without the distress of physical puberty development, this intervention used puberty suppression via GnRH analogs.	Puberty suppression was conducted for 1.9 y (on average).	This was a 1-group pretest-posttest design. Youth were assessed at baseline and postintervention, which was 3.0 y on average after baseline (before the start of crossgender hormones). Puberty suppression was initiated 1.1 y, on average, after baseline assessment.	From 2000 to 2008, 140 of 196 referred youth were considered eligible for medical intervention at a gender identity clinic in Amsterdam, Netherlands. Of the 140 youth, 111 youth were given the intervention. Participants of this study were the first 70 children who had subsequently started crossgender hormone treatment.	Adolescents were eligible for puberty suppression when they were diagnosed with gender identity disorder, had shown persistent gender dysphoria since childhood, lived in a supportive environment, and had no serious comorbid psychiatric disorders that may have interfered with the diagnostic assessment. Youth and their parents gave informed consent.	The total sample contained 70 youth participants. Total sample: N = 70; mean age: 13.56 y at baseline (range: 11–17 y); assigned birth sex: 52.9% female and 47.1% male; sexual orientation: 88.6% had same-natal-sex attractions, 8.6% had both-natal-sex attractions, and 2.8% reported something else.	Depressive Symptoms: Beck Depression Inventory-II	Depressive symptoms decreased significantly from baseline to postintervention; 8.31 vs 4.95; F1,39 = 9.28; P = .004.
							Anxiety symptoms: State-Trait Anxiety Inventory	Anxiety symptoms did not significantly decrease from baseline to postintervention; 39.43 vs 37.95; F1,39 = 1.21; P = .276.
							Internalizing symptoms: Youth Self-Report	Internalizing symptoms decreased significantly from baseline to postintervention; 56.04 vs 49.78; F1,52 = 15.05; P < .001. The percentage of youth participants scoring in the clinical range for internalizing symptoms significantly decreased from baseline to postintervention; 29.6% vs 11.1%; χ21 = 5.71; P = .017.
							Internalizing symptoms: Child Behavior Checklist	Internalizing symptoms decreased significantly from baseline to postintervention; 61.00 vs 54.46; F1,52 = 22.93; P < .001.
							Externalizing symptoms: Youth Self-Report	Externalizing symptoms decreased significantly from baseline to postintervention; 53.30 vs 49.98; F1,52 = 7.26; P = .009.
							Externalizing symptoms: Child Behavior Checklist	Externalizing symptoms decreased significantly from baseline to postintervention; 58.04 vs 53.81; F1,52 = 12.04; P = .001.
de Vries et al29	Aimed at providing high-quality clinical care to youth with gender dysphoria, this intervention included puberty suppression, crossgender hormones, and gender affirmation surgery. Puberty suppression was provided by using GnRH analogs. Crossgender hormones were provided. Gender affirmation surgery included vaginoplasty for transwomen and mastectomy and hysterectomy with ovariectomy for transmen.	Participants started puberty suppression at a mean age of 14.8 y (range: 11.5–18.5 y), crossgender hormones at a mean age of 16.7 y (range: 13.9–19.0 y), and gender affirmation surgery at a mean age of 19.2 y (range: 18.0–21.3 y).	This was a 1-group pretest-midtest-posttest design. Youth were assessed at baseline (time 0), during intervention (time 1; ∼3.1 y after baseline; after initiation of puberty suppression and before initiation of crossgender hormones), and at postintervention (time 2; ∼7.1 y after baseline; and 1 y after gender affirmation surgery).	Participants were recruited from the first cohort of 70 children who had gender dysphoria, who were prescribed puberty suppression in Amsterdam, Netherlands, and who continued with gender affirmation surgery between 2004 and 2011.	Youth were eligible for puberty suppression, crossgender hormones, and gender affirmation surgery at the respective ages of 12, 16, and 18 y if they had a history of gender dysphoria, no psychosocial problems, adequate family or other support, and a good comprehension of the impact of medical interventions. At postintervention, from 2008 to 2012, young adults were eligible if they were ≥1 y past their gender affirmation surgery. Puberty suppression started after youth entered the first stages of puberty (Tanner stages 2–3). At baseline and time 1, youth and their parents provided consent. At time 3, only participants provided consent.	The total sample contained 55 participants. Total sample: N = 55; mean age: 13.6 y at baseline (range: 11.1–17.0 y); gender: 40.0% transwomen and 60.0% transmen.	Depressive symptoms: Beck Depression Inventory	Depressive symptoms had significant quadratic trends over time (P = .04), decreasing from baseline (7.89) to time 1 (4.10), and increasing at time 2 (5.44). Trends were similar by gender.
							Anxiety symptoms: State-Trait Anxiety Inventory	Anxiety symptoms did not have linear (P = .42) or quadratic (P = .47) trends over time. However, the linear trends were different by gender (P = .05): for transmen, symptoms decreased over time (44.41 at baseline; 41.59 at time 1; and 39.20 at time 2); for transwomen, average symptoms were lower at baseline and time 1 (31.87 and 31.71) than at time 2 (35.83).
							Psychosocial functioning: Children’s Global Assessment Scale	Psychosocial functioning increased linearly over time (P < .001). Psychosocial functioning was 71.13 at baseline, 74.81 at time 1, and 79.94 at time 2. Trends were similar by gender.
							Internalizing symptoms: Child and Adult Behavior Checklists	Internalizing symptoms linearly decreased over time (P < .001). Average internalizing symptoms were 60.83 at baseline, 54.42 at time 1, and 50.45 at time 2. Trends were similar by gender. Overall, prevalence of clinical levels of internalizing symptoms significantly decreased from baseline to time 1 (30.0% vs 12.5%), plateauing at time 3 (10.0%).
							Internalizing symptoms: Youth and Adult Self-Reports	Internalizing symptoms had quadratic trends over time (P = .008), decreasing from baseline to time 1 (55.47–48.65), and increasing at time 2 (50.07). Trends were similar by gender. Overall, prevalence of clinical levels of internalizing symptoms significantly decreased from baseline to time 1 (30.0% vs 9.3%), but time 2 prevalence (11.6%) was similar to both previous time points.
							Externalizing symptoms: Child and Adult Behavior Checklists	Externalizing symptoms decreased linearly over time (P < .001; 57.85 at baseline, 53.85 at time 1, and 47.85 at time 2). Trends were similar by gender. Overall, the prevalence of clinical levels of externalizing symptoms was not significantly different from baseline to time 1 (40.0% vs 25.0%) but was significantly lower at time 2 (2.5%).
							Externalizing symptoms: Youth and Adult Self-Reports	Externalizing symptoms did not have linear (P = .14) or quadratic (P = .09) trends. But linear trends differed by gender (P = .005): for transmen, there were linear decreases (57.16 at baseline; 52.64 at time 1; and 50.24 at time 2); for transwomen, symptoms were lower at baseline and time 1 (46.00 and 44.71) than at time 3 (50.24). Overall, prevalence of clinical levels of externalizing symptoms did not significantly change (21.0% at baseline; 11.6% at time 1; 7.0% at time 2).
Diamond et al30	Aimed at reducing suicidal ideation and depressive symptoms among SMY, this intervention tested a form of Attachment-Based Family Therapy specifically tailored to the needs of SMY and their families. Attachment-Based Family Therapy is an empirically informed, manualized family-based treatment, but this specific intervention was adapted by researchers and clinicians who had experience working with SMY. All therapy sessions were delivered in person by a PhD-level clinical psychologist. Sessions were provided to adolescents by themselves, parent(s) by themselves, and an adolescent and their parent(s) together. This intervention was guided by attachment theory, structural family therapy, multidimensional family therapy, and emotion-focused therapy.	Completing at least 8 sessions was considered a full intervention dosage. The No. sessions per participant ranged from 8 to 16, with an average of 12 sessions per family. Sessions were ∼60 min in length and were conducted on a weekly basis.	This was a 1-group pretest-midtest-posttest design. Research assistants naïve to the study purpose administered assessments at baseline, 6 wk later (halfway through intervention), and 12 wk later (postintervention).	Patients were recruited from 2 private psychiatric hospitals in Philadelphia, Pennsylvania, where participants had been admitted for suicidal ideation or attempts. Social work staff employed by the hospitals screened potential participants 1 wk before their discharge, and youth endorsing significant levels of suicidal ideation (per a score ≥31 on the Suicidal Ideation Questionnaire-Junior) were referred to the study.	Youth participants had to self-identify as gay, lesbian, or bisexual and had to report significant levels of suicidal ideation as evidenced by a score ≥31 on the Suicidal Ideation Questionnaire-Junior. Youth were excluded if they had current psychosis or mental retardation. Youth and their parents gave informed consent.	The total sample contained 10 youth participants. Regarding parental participation, 40% of youth completed the intervention with 2 parents, and 60% completed the intervention with their mother only. Total sample: N = 10; mean age: 15.10 y at baseline (range: 14–18 y); gender: 80% female and 20% male; sexual orientation: 30% identified as exclusively gay or lesbian, 10% identified as primarily gay and also attracted to girls, and 60% identified as primarily lesbian and also attracted to boys; race and/or ethnicity: 20% white, 50% African American, 20% multiracial, and 10% other.	Depressive Symptoms: Beck Depression Inventory-II	Average depressive symptoms decreased over the course of treatment; F2,18 = 4.59; P = .03; d = 0.90.
							Suicidal ideation symptoms: Suicidal Ideation Questionnaire-Junior	Average suicidal ideation decreased over the course of treatment; F2,18 = 18.78; P = .001; d = 2.10.
Lucassen et al31	Aimed at reducing depressive symptoms for SMY, this intervention used a 7-module computerized cognitive behavioral therapy intervention delivered via CD-ROM on personal computers and a paper-based user notebook. This intervention used the medium of a fantasy world where the user’s avatar is faced with a series of challenges to rid a virtual world of gloom and negativity. On the basis of cognitive behavioral therapy theories and adapted from an efficacious intervention (SPARX), this intervention was tailored to the needs SMY by having them contribute to the adaptation process. Participants could choose whether to complete the program at home, at a youth-led organization for SMY, at a selected high school, or on a dedicated computer where the study was based.	Each of the 7 modules took ∼30 min to complete. Participants were instructed to complete 1 or 2 modules per wk and to finish all modules within 2 mo.	This was a 1-group pretest-posttest-posttest design. Youth participants completed questionnaires at baseline, immediately postintervention, and 3 mo postintervention.	One youth-led organization for SMY and 4 high schools promoted the study in Auckland, New Zealand. The study was also advertised and endorsed by sexual-minority media.	Youth participants had to be attracted to the same sex, both sexes, or not sure of their sexual attractions; 13–19 y old; have depressive symptoms (i.e., Child Depression Rating Scale, Revised, raw score ≥30) at baseline; and living in Auckland, New Zealand. SMY with severe depressive symptoms or at risk for suicide or self-harm were eligible if they reported receiving support from a school guidance counselor, therapist, or general practitioner. Those receiving antidepressant medication or other relevant therapies were able to take part; these additional treatments were documented at the preintervention assessment. For youth <16 y, youth and their parents gave informed consent. For youth ≥16 y, only youth gave informed consent.	The total sample contained 21 youth. Total sample: N = 21; mean age: 16.5 y at baseline (range: 13–19 y); gender: 47.6% female and 52.3% male; sexual orientation: 47.6% had same-sex attractions, 47.6% had both-sex attractions, and 4.8% were not sure; race and/or ethnicity: 71.4% New Zealand European, 9.5% Māori, 4.8% Pacific ethnicity, and 14.3% Asian.	Depressive Symptoms: Children’s Depression Rating Scale, Revised	Depressive symptoms decreased significantly from baseline to immediate postintervention (mean change = −7.43; 95% CI: −10.79 to −4.07; P < .0001; d = 1.01). Depressive symptoms remained similar from immediate postintervention to 3 mo postintervention (mean change = −0.62; 95% CI: −5.82 to 4.58; P = .81).
							Depressive Symptoms: Reynolds Adolescent Depression Scale	Depressive symptoms decreased significantly from baseline to immediate postintervention (mean change = −7.90; 95% CI: −12.17 to −3.64; P = .001; d = 0.84). Depressive symptoms remained similar from immediate postintervention to 3 mo postintervention (mean change = −0.86; 95% CI: −5.41 to 3.70; P = .70).
							Depressive Symptoms: Mood and Feelings Questionnaire	Depressive symptoms decreased significantly from baseline to immediate postintervention (mean change = −6.19; 95% CI: −11.13 to −1.25; P = .02; d = 0.57). Depressive symptoms remained similar from immediate postintervention to 3 mo postintervention (mean change = 0.67; 95% CI: −5.58 to 6.92; P = .83).
							Anxiety symptoms: Spence Children’s Anxiety Scale	Anxiety symptoms decreased significantly from baseline to immediate postintervention (mean change = −7.86; 95% CI: −11.62 to −4.10; P < .0001; d = 0.95). Anxiety symptoms were not assessed 3 mo postintervention.
							Hopelessness: Kazdin Hopelessness Scale for Children	Hopelessness scores decreased significantly from baseline to immediate postintervention (mean change = −1.43; 95% CI: −2.43 to −0.43; P = .008; d = 0.65). Hopelessness was not assessed 3 mo postintervention survey.
Painter et al32	This program provided youth with services and supports through the Comprehensive Community Mental Health Services for Children with Serious Emotional Disturbances Program, more commonly known as the Children’s Mental Health Initiative. Guided by the “systems of care” framework, this program provided coordinated networks of community-based services tailored to youth. The interventions considered the unique strengths and needs of the youth target population and incorporated cultural, racial, ethnic, and linguistic diversity of the local environments, which included awareness of, sensitivity toward, and confidentiality for SGMY. The participants served by this program received a wide variety of specific interventions, including individual therapy, medication treatment, case management, group therapy, recreational activities, inpatient hospitalization, vocational training, family support, and residential treatment.	Interventions widely varied in length. For example, 6 mo after enrollment, those who received medication treatment had an average of 4.5 visits, those who received individual therapy had an average of 16.1 sessions, and those who were part of a therapeutic group home spent on average 124.7 d receiving the intervention. Treatment plans were created on an individual basis and constrained by the locally offered services and supports.	This was a 1-group pretest-midtest-posttest design. Youth and their caregivers completed questionnaires at baseline, 6 mo after baseline, and 12 mo after baseline.	Youth and caregivers were recruited within the 47 systems of care grantee communities from 2010 to 2014.	Youth had to be age 11–21 y, have a serious emotional disturbance, have entered Children’s Mental Health Initiative care services through 1 of 47 systems of care grantee communities from 2010 to 2014, have participated in the national evaluation, and identify as a sexual or gender minority.	The total sample contained 482 youth participants. Total sample: N = 482; age: 8.7% aged 11–12 y, 28.6% aged 13–15 y, 42.9% aged 15–17 y, and 19.7% aged 18–21 y; gender: 22.6% male, 68.3% female, 1.0% transgender female, 2.3% transgender male, 4.6% not sure, and 1.2% other; sexual orientation: 15.4% mostly heterosexual, 49.8% bisexual, 5.0% mostly homosexual, 12.0% homosexual, 5.4% other, and 12.4% questioning; race and/or ethnicity: 3.7% American Indian or Alaskan native, 25.5% African American, 0.4% native Hawaiian or other Pacific Islander, 42.9% white, 17.0% Hispanic or Latino, and 10.4% multiracial.	Anxiety symptoms: Revised Children’s Manifest Anxiety Scale, Second Edition	Anxiety symptoms significantly decreased across time: F2,167 = 5.59; P = .004.
							Depressive symptoms: Reynolds Adolescent Depression Scale, Second Edition	Depressive symptoms significantly decreased across time: F2,198 = 5.16; P = .006.
							Global functioning impairment: Columbia Impairment Scale	Global functioning impairment symptoms significantly decreased across time: F2,212 = 60.02; P < .001).
							Internalizing and externalizing symptoms: Child Behavior Checklist	Internalizing symptoms (F2,242 = 9.34; P < .001), externalizing symptoms (F2,214 = 15.73; P < .001), and total internalizing and externalizing symptoms (F2,251 = 15.78; P < .001) significantly decreased across time.
							Substance use and/or substance abuse symptoms: Substance Use and Abuse Scale-9 (GAIN Quick-R)	Substance use and/or substance abuse symptoms significantly decreased across time: F2,147 = 5.33; P = .006.
							Substance dependence symptoms: Substance Dependence Scale-7 from (GAIN Quick-R)	Substance dependence symptoms significantly decreased across time: F2,149 = 7.93; P = .001).
							Total substance use and/or substance abuse symptoms and substance dependence symptoms: GAIN Quick-R Total Substance Problems Scale	Total substance problems significantly decreased across time: F2,193 = 5.15; P < 0.01).
Raifman et al33	This intervention was the presence of a US state-level policy that granted same-sex couples equivalent marriage rights as opposite-sex couples.	This was a 1-time enactment of state-level policy.	This was an interrupted time series design using serial biennial cross-sectional data before and after policy intervention implementation.	Data were collected in the United States via the biennial state YRBSS from 1999 to 2015. YRBSS uses a 2-stage sampling of schools and classrooms to obtain a representative sample of students in grades 9–12 in public schools. A total of 47 states were included, and 25 states collected information about sexual identity in 2015.	Youth were included if they were in grades 9–12 in a sampled public school and classroom. YRBSS uses active or passive parental permission depending on the administering state. Student surveys are anonymous and voluntary.	The total sample contained 762 678 youth. The intervention and control groups differed by age and race and/or ethnicity but not gender. Information about group differences by sexual identity were not included. In 2015, 12.7% identified as sexual minorities: 2.4% as gay or lesbian, 6.4% as bisexual, and 4.0% as not sure. Intervention: n = 546 276; mean age: 15.9 y (SD: 1.2 y); gender: 49.7% male and 50.3% female; race and/or ethnicity: 58.4% white, 11.3% Hispanic, 14.3% African American, and 16.0% other. Control: n = 216 402; mean age: 16.0 (SD: 1.2 y); gender: 49.6% male and 50.4% female; race and/or ethnicity: 55.0% white, 16.1% Hispanic, 18.6% African American, and 10.3% other.	Suicide attempts: “During the past 12 months, how many times did you actually attempt suicide?” This was coded as any versus none.	Across all states before implementation of same-sex marriage policies, 28.5% of SMY and 8.6% of all youth reported having at least 1 past-year suicide attempt. After implementation among SMY, there was a significant decline in past-year suicide attempt prevalence (net change = −4.0; 95% CI: −6.9 to −1.2; P < .01), which is equivalent to a 14% relative decline in the proportion of SMY reporting at least 1 past-year suicide attempt. After implementation, there was also a significant decline in past-year suicide attempt prevalence among all youth (mean net change = −0.6; 95% CI: −1.2 to −0.1; P < .05).
Schwinn et al34	Aimed at reducing substance use among SMY via an online intervention, this intervention had an animated young adult narrator guide youth through interactive games, role-playing, and writing activities. Activities focused on skills for identifying and managing stress, making decisions, addressing drug use rates, and teaching drug refusal skills. This intervention was guided by a social competency skill-building strategy and minority stress theory.	Three sessions were completed throughout a 4-wk period. Youth completed each session in 14 min, on average.	This was a randomized controlled trial using a pretest-posttest design. Youth completed online questionnaires at baseline, immediately postintervention, and 3 mo postintervention. Youth completed follow-up questionnaires ∼1 mo and 4.5 mo after baseline. Authors only reported baseline and 3-mo postintervention results.	Youth were recruited from across the United States through Facebook advertisements posted to the pages of 15- and 16-y-old youth. Six advertisements ran for 9 d in the spring of 2014.	Youth were included if they were 15 or 16 y of age, a US resident, had access to a personal computer, and identified as gay, lesbian, bisexual, transgender, or questioning. Youth had to correctly answer a 5-question quiz on study procedures to participate. This study had a waiver of parental permission.	The total sample contained 236 youth. The intervention and control groups did not differ by demographics at baseline. Intervention: n = 119; mean age: 16.05 y at baseline (range: 15–16 y); gender: 32.1% male, 49.6% female, and 18.3% queer, fluid, or other; sexual orientation: 39.4% had same-sex attractions, 49.5% had both-sex attractions, 5.5% had opposite-sex attractions, and 5.6% were unsure; race and/or ethnicity: 66.1% white, 12.8% Hispanic, 7.3% African American, 6.4% Asian American, and 7.4% other. Control: n = 117; mean age: 16.10 y at baseline (range: 15–16 y); gender: 33.3% male, 52.2% female, and 4.5% queer, fluid, or other; sexual orientation: 37.9% had same-sex attractions, 49.1% had both-sex attractions, 6.9% had opposite-sex attractions, and 6.1% were unsure; race and/or ethnicity: 58.1% white, 13.7% Hispanic, 12.0% African American, 8.5% Asian American, and 7.7% other.	Alcohol use: No. times drank in past 30 d	At baseline, there was not a significant difference for intervention versus control groups (P = .09). At 3-mo follow-up, there was not a significant difference for intervention versus control groups in mean alcohol use frequency (1.29 vs 1.10; P ≥ .05; t = 0.66).
							Cigarette smoking: No. times smoked in past 30 d	At baseline, there was not a significant difference for intervention versus control groups (P = .82). At 3-mo follow-up, there was not a significant difference for intervention versus control groups in mean cigarette smoking frequency (0.72 vs 0.90; P ≥ .05; t = 0.59).
							Marijuana use: No. times used in past 30 d	At baseline, there was not a significant difference for intervention versus control groups (P = .51). At 3-mo follow-up, there was not a significant difference for intervention versus control groups in mean marijuana use frequency (1.63 vs 1.74; P ≥ .05; t = 0.41).
							Other drug use: No. times used in past 30 d	At baseline, there was not a significant difference for intervention versus control groups (P = .31). At 3-mo follow-up, intervention group participants had significantly lower mean other drug use frequency than control group participants (1.03 vs 1.09; P < .05; t = 2.16; d = 0.34).
							Perceived stress: scores ranged from 1 (low) to 5 (high)	At baseline, there was not a significant difference for intervention versus control groups (P = .72). At 3-mo follow-up, intervention group participants had significantly lower mean perceived stress than control group (3.05 vs 3.33; P < .05; t = 2.27; d = 0.34).
Seelman and Walker35	The 2 interventions were (1) the presence (versus absence) of a US state-level general antibullying law and (2) the presence (versus absence) of a US state-level antibullying law that enumerated sexual orientation as a protected class.	This was a 1-time enactments of the 2 different state-level laws.	This was an interrupted time series design using serial biennial cross-sectional data.	Data were collected in the United States via the biennial state YRBSS from 2005 to 2015. YRBSS uses a 2-stage sampling of schools and classrooms to obtain a representative sample of students in grades 9–12 in public schools. A total of 22 states were included. Only 3 states had data from both before and after enactment of the general antibullying law, and 4 states had data from both before and after enactment of the enumerated antibullying law.	Youth were included if they were in grades 9–12 in a sampled public school and classroom. YRBSS uses active or passive parental permission depending on the administering state. Student surveys are anonymous and voluntary.	The total sample contained 286 568 youth. Information about demographic differences by states with and without the intervention laws were not included. Total sample: N = 286 568; mean age: 16.0 y (SD: 0.02 y); gender: 50.6% male and 49.4% female; sexual orientation: 10.5% identified as lesbian or gay, bisexual, or not sure (henceforth referred to as questioning) and 89.5% identified as heterosexual.	Bullying victimization: “During the past 12 months, have you ever been bullied on school property?” This was coded as any versus none.	General antibullying laws were associated with reductions in bullying victimization among LGB youth (b = −0.055; SE = 0.023) and LGBQ youth (b = −0.072; SE = 0.024). In states with general antibullying laws, 6.4% fewer LGB youth and 7.5% fewer LGBQ youth were bullying victims. Enumerated antibullying laws were also associated with reductions in bullying victimization among LGB youth (b = −0.056; SE = 0.023) but not LGBQ youth by (b = −0.016; SE = 0.016). In states with enumerated antibullying laws, 5.1% fewer LGB youth were bullying victims. The protective associations of both general and enumerated antibullying laws were pronounced among LGB and LGBQ boys <16 y old.
							Threatened or injured with a weapon: “During the past 12 months, how many times has someone threatened or injured you with a weapon such as a gun, knife, or club on school property?” This was coded as any versus none.	Neither general nor enumerated antibullying laws were associated with being threatened or injured with a weapon among LGB or LGBQ youth (data not provided). However, there was a protective association for general antibullying laws among LGBQ boys <16 y old: in states with general antibullying laws, 13.8% fewer LGBQ boys <16 y reported being threatened or injured with a weapon. This protective association was not found when examining the same group of only LGB boys.
							Suicidal ideation: “During the past 12 months, did you ever seriously consider attempting suicide?” This was coded as any versus none.	Neither general nor enumerated antibullying laws were associated with suicidal ideation among LGB or LGBQ youth (data not provided). These associations were similar by sex and age.
							Suicide attempt: “During the past 12 months, how many times did you actually attempt suicide?” This was coded as any versus none.	General antibullying laws were not associated with suicide attempts among LGB youth (b = 0.009; SE = 0.023) or LGBQ youth (b = 0.005; SE = 0.019). Enumerated antibullying laws were associated with reductions in suicide attempts among LGBQ youth by 3.3% (b = −0.037; SE = 0.015; P < .05), but not among only LGB youth (b = 0.009; SE = 0.022).

CD-ROM, compact disc read-only memory; CI, confidence interval; DSM-IV-TR, Diagnostic and Statistical Manual of Mental Disorders-IV Text Revision; GAIN Quick-R, Global Appraisal of Individual Needs–Quick-Revised; LGBQ, lesbian, gay, bisexual, and questioning; SPARX, Smart, Positive, Active, Realistic, X-factor; YRBSS, Youth Risk Behavior Surveillance System.