Examining the Relationship Between Proposed Anti-LGBTQ+ Legislation and LGBTQ+ College Student Mental Health: Findings from the Healthy Minds Study, 2021–2022

Abstract

Objective:

This study aimed to investigate the association between proposed anti-LGBTQ+ legislation and depressive symptoms among LGBTQ+ college students.

Participants:

Participants (N=72,135) included LGBTQ+ college students (N=21,466) from over 530 colleges and universities, including technical and community institutions, across the United States who took part in the Healthy Minds Study.

Methods:

Data on proposed anti-LGBTQ+ legislation introduced in 2021–2022 were collected from all 50 U.S. states, categorized, and analyzed. Individual-level data for 2021–2022 were obtained from the Healthy Minds Study, and a multilevel analysis was conducted to assess the association between proposed anti-LGBTQ+ legislation and depressive symptoms while considering individual factors.

Results:

A significant positive association between proposed anti-LGBTQ+ legislation and increased depressive symptoms among LGBTQ+ college students was found. This association remained significant after controlling for individual-level stressors and identities, including a sense of belonging, campus environment perceptions, first-generation student status, and transgender or gender non-conforming identity.

Conclusions:

The findings underscore the need for mental health support and inclusive policies for LGBTQ+ college students, particularly in regions where anti-LGBTQ+ legislation is proposed or enacted, to mitigate the potential negative impact of multiple factors on their mental well-being.

Introduction

Depression has increased among college students in the last decade, and depression combined with anxiety has doubled from 2013 to 2019.¹ The addition of anxiety to depression contributes to difficulties in treating depression.² With increasing rates of depression and other psychiatric disorders, the use of psychiatric medication among college students has also doubled in the last decade, with about one in four college students taking psychiatric medication.³ LGBTQ+ college students, especially, experience high rates of depression and other mental health problems and increased rates of suicide due to contributing factors such as family rejection, harassment, discrimination, prejudice, and stigma.⁴ Proposed anti-LGBTQ+ legislation is posing even more challenges to the health and well-being of this population. By March 2022, state lawmakers proposed 238 bills that limited the rights of LGBTQ+ Americans, with half of the bills focused on transgender youth.⁵ There is evidence that introducing these laws and the subsequent legislative debates about medical and supportive care for transgender people contributes to this population’s poorer mental health. Indeed, 66% of LGBTQ+ individuals and 85% of trans and/or non-binary individuals reported that recent debates focused on state laws that restrict the rights of transgender individuals had negatively impacted their mental health.⁶ With proposed anti-LGBTQ+ legislation increasing in 2023,⁷ more research examining the potential impact on the health of the LGBTQ+ population is needed.

College Student Mental Health

College student mental health has worsened over the past decade and appears to be especially poor among LGBTQ+ college students. The worsening trends of depression among college students, combined with increased psychiatric medication use, highlight the urgency to address mental health issues in this population.⁸ LGBTQ+ college students experience mental health disparities, with a disproportionate burden of depression and suicidality.^9,10 Gender minority students experience depression at twice the rate of cisgender students and rates of suicidal ideation, plans, and attempts are 3–4 times higher among gender minority students compared to their cisgender peers. ⁹ Further, 75% of gender minority students have one or more mental health problems—with more than half of gender minority students reporting they have experienced depression and more than one-third of gender minority students reporting suicidal ideation in the past year.⁹ Sexual minority young adults also experience significantly higher rates of depression compared to heterosexual young adults.^11,12 College students at the intersections of these identities, who identify as both a gender and sexual minority, reported significantly higher depressive symptoms compared with both cisgender heterosexuals and cisgender sexual minority individuals.¹⁰ The impact of proposed anti-LGBTQ+ legislation on LGBTQ+ individuals’ rights may further exacerbate mental health disparities. However, there is limited research examining the specific effects of proposed anti-LGBTQ+ legislation on the mental health of college students.

The minority stress model identifies distal and proximal stress processes that place LGBTQ+ individuals at greater risk for negative mental health outcomes.¹³ Distal stress processes include stressors established by institutions that impact LGBTQ+ individuals, such as proposed legislation that reduces access to gender-affirming care and LGBTQ+ content, as well as major life events (e.g. being the victim of violence) and everyday experiences of discrimination or microaggressions.¹⁴ These distal stressors may contribute to proximal processes such as LGBTQ+ individuals’ thwarted belongingness, which is a risk factor for suicide among LGBTQ+ college students.¹⁵ Further, perceptions of a negative campus climate for LGBTQ+ college students are linked to lower academic integration and social integration,¹⁶ which contributes to depressive symptoms among college students.¹⁷

Consistent with a protective factor in the minority stress model, research indicates that for LGBTQ+ individuals, feeling a sense of belonging to an LGBTQ+ group is associated with feeling a sense of belonging in the general community, which is associated with lower levels of depression.¹⁸ While the primary aim of this study is to evaluate associations between a distal stressor (i.e. proposed anti-LGBTQ+ legislation) and LGBTQ+ college student depressive symptoms, this study will also contribute to understanding how gender identity ^9,10 and proximal stressors (i.e., lacking a sense of belonging on campus,¹⁹ negative perceptions of the campus environment, first-generation student status²⁰) are associated with depressive symptoms among a representative sample of LGBTQ+ college students. The inclusion of individual-level variables such as lack of sense of belonging, perceptions of the campus environment, first-generation student status, and transgender or gender non-conforming identity will contribute to identifying how distal stressors, proximal stressors, and minoritized identities, represented in the minority stress model, contribute to mental health disparities within this population.^21–23

Anti-LGBTQ+ Legislation

Recent research supports linkages between proposed anti-LGBTQ+ legislation and poor mental health outcomes for LGBTQ+ populations. For instance, transgender individuals who knew the amount of anti-trans legislation in their state reported stronger external stressors, such as discrimination, and increased hopelessness, which was strongly associated with adverse mental health.²⁴ Proposing legislation aimed at restricting gender-affirming practices and banning LGBTQ+ content in schools, even if it is not passed into law, signals to LGBTQ+ populations that their rights and identities are not supported. Indeed, proposed anti-LGBTQ+ legislation in state legislatures was associated with an increase in texts to a Crisis Text Line from LGBTQ+ individuals.²⁵ Proposed anti-LGBTQ+ legislation threatens the healthcare and education system’s ability to provide social support and care.²⁶ Indeed, these legislative actions perpetuate hostile environments for LGBTQ+ physicians and patients and vilify those with beliefs contrary to those of the lawmakers.^27,28 The media exposure that accompanies the introduction of anti-LGBTQ+ legislation in state legislatures likely influences many of the distal stressors experienced by LGBTQ+ populations. For instance, transgender, genderqueer, and gender-nonconforming individuals, who were made political targets during the 2016 presidential election, reported elevated levels of hate speech and violence exposure.^9,29 Further, news consumption related to proposed anti-LGBTQ+ legislation was associated with increased persistent or unwanted thoughts about the legislation, while perceiving that people in one’s social network support such legislation was associated with greater depressive symptoms, physical health symptoms, and fear of disclosing one’s identity.³⁰ The stigmatizing rhetoric that accompanies proposed anti-trans legislation signals that trans people do not have rights that other people have and harms the entire LGBTQ+ community.^30,31

Political polarization itself has negative impacts on mental health. Attempts to pass anti-LGBTQ+ legislation could be perceived as hostile acts and contribute to a hostile political climate. The perception of this hostile climate may negatively affect the mental health and ultimate success of college students. By fall 2023, twice the number of anti-LGBTQ+ bills had been introduced compared to all of 2022, and 12 times the number of anti-LGBTQ+ bills introduced in 2018 in the U.S.³¹ This rapid increase in proposed anti-LGBTQ+ legislation indicates a pressing need to examine associations between attempts to pass this type of legislation and the health and well-being of populations affected by it. Therefore, this study explores how the magnitude of proposed anti-LGBTQ+ legislation is associated with depressive symptoms in a representative sample of college students while controlling for gender identity and proximal stressors within the minority stress model.

Methods

Anti-LGBTQ+ Legislation Coding

Data for this study were collected in two steps. Beginning in September 2022, researchers consulted the American Civil Liberties Union (ACLU) and Movement Advancement Project (MAP) websites to identify different types of anti-LGBTQ+ legislation in U.S. states introduced during the 2021–2022 legislative sessions. From a review of these websites, the researchers identified four categories of anti-LGBTQ+ legislation to search for and code for each U.S. state: single-sex facility restrictions, prohibiting health care for transgender individuals, excluding transgender individuals from athletics, and LGBTQ+ curricular laws. Using definitions from the MAP, inclusion criteria were developed for each category of legislation. Legislation categorized as “excluding transgender individuals from athletics” and “single-sex facility restrictions” included any legislative proposal that prohibited individuals from participating in school athletic teams or using single-sex public facilities (e.g., bathrooms, locker rooms) based on their gender identity. Legislation categorized as “prohibiting health care for transgender individuals” included any legislative proposal that limited access to gender-affirming health care or related procedures or treatment for minors under age 18. Bills categorized as “anti-LGBTQ+ curricular laws” included legislation like the “Don’t Say Gay” legislation passed in Florida²⁷ and legislation that allowed for parental opt-out of LGBTQ+-inclusive curricula.²⁸ This legislation included language that prohibited or restricted the teaching or discussion of LGBTQ+ topics and issues relating to gender identity or sexual orientation in public schools or required parents to be notified in advance of any gender identity or sexuality-related content and allowed parents to opt their children out of this instruction.

Next, the researchers consulted the ACLU and MAP websites and conducted web searches for news articles discussing each category of anti-LGBTQ+ legislation for each state to identify the legislative bill numbers corresponding to the four categories. Then, bill numbers were searched in each respective state’s legislative website to read the text of each bill to determine if it met the inclusion criteria for that category of proposed anti-LGBTQ+ legislation (i.e. single-sex facility restrictions, prohibiting health care for transgender individuals, excluding transgender individuals from athletics, and LGBTQ+ curricular laws). Within each category, researchers listed the year the anti-LGBTQ+ legislation was introduced, all bill numbers within that category (many states had bills in both the House and Senate), and the legislation’s status (i.e. in committee, passed, vetoed, failed). States were coded as 0 = no legislation that met the inclusion criteria for the anti-LGBTQ+ legislation category or 1 = legislation that was introduced AND had another action (e.g. referred to committee) that met the inclusion criteria for the category. Bills that were introduced and had additional legislative actions appeared to generate more attention from the media and advocacy groups, while bills that were solely introduced appeared to generate limited coverage in the media and with advocacy groups. Two researchers coded legislation by state, with a third coder coding 25% of the states to check for accuracy. Any discrepancies were resolved by consulting the language in the respective bills and discussions to reach a consensus. Once all legislation was coded into the spreadsheet, both coders independently checked each other’s coding and data collection as a final check of accuracy. Table 1 shows a sampling of the data collected for each state to illustrate the data and coding scheme.

Table 1.

Examples of anti-LGBTQ legislation coding for states within regions

State	Region	SSFR	# & Action	Legislative Text	THB	# & Action	Legislative Text	TSB	# & Action	Legislative Text	LGBTQ CL	# & Action	Legislative Text
AR	7 – South (West South Central)	0			1	HB1570 now ACT 626 (2021) Passed	“Prohibition of gender transition procedures for minors. (a) A physician or other healthcare professional shall not provide gender transition procedures to any individual under eighteen (18) years of age. (b) A physician, or other healthcare professional shall not refer any individual under eighteen (18) years of age to any healthcare professional for gender transition procedures. …”	1	SB354 now ACT 461 (2021) Passed	“Interscholastic, intercollegiate, intramural, or club athletic teams or sports that are sponsored by a school shall be expressly designated as one (1) of the following based on biological sex: (1) “Male”, “men’s”, or “boys; (2) (A) “Female”, “women’s”, or “girls”. (B) An interscholastic, intercollegiate, intramural, or club athletic team or sport that is expressly designated for females, women, or girls shall not be open to students of the male sex; or (3) “Coed” or “mixed”.”	1	SB389 now ACT 552 (2021) Passed	“A public school shall make the following, as they relate to sex education, sexual orientation, and gender identity, available for inspection by parents and legal guardians of participating public school students: (A) Curricula; (B) Materials; (C) Tests; (D) Surveys; (E) Questionnaires; (F) Activities; and (G) Instruction of any kind. (2) Information made available for inspection under subdivision (b)(1) of this section shall be made available regardless of whether any of the listed items are offered as part of a sex education class or program or as part of any other class, activity, or program.”
CT	1 – Northeast (New England)	0			0			0			0
IA	4 – Midwest (West North Central)	1	SF2025, SF224,HF184, HF405 (2021) - Introduced and referred to committee	“... It shall not be an unfair or discriminatory practice to prohibit a person from using a single or multiple occupancy toilet facility that does not correspond with the person’s biological sex as provided in section 280.32.11… A person shall not enter a school multiple occupancy toilet facility, or a single occupancy toilet facility designated only for persons of the same biological sex, that does not correspond with the person’s biological sex.”	1	HF193 (2021) - Introduced and referred to committee	“... it shall be unlawful for a medical professional to engage in any of the following practices upon a minor, or to cause such practice to be performed to facilitate a minor’s desire to present or appear in a manner that is inconsistent with the minor’s sex: a. Performing a surgery that sterilizes, including castration, vasectomy, hysterectomy…. b. Performing a mastectomy. c. Administering or supplying the following medications that induce transient or permanent infertility…”	1	HF2416 now H.J. 493 (2022) Passed	“1. a. An interscholastic athletic team, sport or athletic event that is sponsored or sanctioned by an educational institution or organization must be designated as one of the following, based on the sex at birth of the participating students: (1) Females, women, or girls. (2) Males, men, or boys. (3) Coeducational or mixed. b. Only female students, based on their sex, may participate in any team, sport, or athletic event designated as being for females, women, or girls.”	1	SF2024 (2022) - Introduced and referred to committee	“This bill establishes that the curriculum utilized by a school district or accredited nonpublic school in accordance with the state’s educational standards for elementary students shall not include instruction relating to gender identity… If a school does not obtain a parent’s or guardian’s written consent, the student may opt out of instruction relating to gender identity. Gender identity, for purposes of the bill, means a gender-related identity of a person, regardless of the person’s assigned sex at birth.”
IL	3 – Midwest (East North Central)	0			0			1	HB4082 (2022) - Introduced and referred to committee.	“Sec.1.25. Designation of team or sport based on biological sex. (a) Any athletic team or sport that is under the jurisdiction of an association or other entity that has as one of its purposes promoting, sponsoring, regulating, or in any manner providing for interscholastic athletics or any form of athletic competition among schools and students within this State must be expressly designated as being: (1) a male athletic team or sport; (2) a female athletic team or sport; or (3) a coeducational athletic team or sport. (b) An athletic team or sport designated as being female is available only to participants who are female, based on their biological sex.””	0
KY	6 – South (East South Central)	0			1	HB253 (2022) Introduced and referred to committee.	“A physician or other health care provider shall not provide gender transition procedures to any child under the age of eighteen (18) years. (2) A physician or other health care provider shall not refer any child under the age of eighteen (18) years to any health care provider for gender transition procedures. (3) Any referral for or provision of gender transition procedures to a child under the age of eighteen (18) years by a health care provider shall be considered unprofessional conduct and be subject to disciplinary action by the appropriate licensing or certifying entity.”	1	HB23 (2022) Introduced and referred to committee.	“The state board or any agency designated by the state board to manage interscholastic athletics shall promulgate administrative regulations or bylaws that provide that: 1. A member school shall designate all athletic teams, activities, and sports as one (1) of the following categories as: a. “Boys”; b. “Coed”; or b. “Girls”; 1 2. The sex of a student for the purpose of determining eligibility to participate in an athletic activity or sport or to use an athletic facility designated for the exclusive use of a single sex shall be determined by a. A studenťs biological sex as indicated on the studenťs original, unedited birth certificate issued at the time of birth; or b. An affidavit signed and sworn to by the physician, physician assistant, advanced practice registered nurse, or chiropractor that conducted the annual medical examination required by paragraph (e) of this subsection under penalty of perjury establishing the studenťs biological sex at the time of birth; 3. a. An athletic activity or sport designated as “girls” shall not be open to members of the male sex. b. Nothing in this section shall be construed to restrict the eligibility of any student to participate in an athletic activity or sport designated as “boys” or “coed”;“	0
VA	5 – South (South Atlantic)	1	HB1126 (2022) - Engrossed	“C. Each school board shall adopt policies to require each student and school board employee to have access to (i) restrooms, locker rooms, and other changing facilities in public school buildings that are shared only by members of the same biological sex;”	0			1	SB766 (2022) - Passed	“... B. Notwithstanding any other provision of law, all athletic teams or squads at a school, whether a school athletic team or an intramural team sponsored by such school, shall designate each such team based on biological sex as follows: 1. “Males,” “men,” or “boys”; 2. “Females,” “women,” or “girls”; or 3. “Coed” or “mixed.” Male students shall not be permitted to participate on any school athletic team or squad designated for “females,” “women,” or “girls.”...”	1	SB626 (2022)* - Passed	“SB 656: “§22.1–16.8… B. ...develop and make available to each school board model policies for ensuring parental notification of any instructional material that includes sexually explicit content and include information, guidance, procedures, and standards relating to: 3. Permitting the parent of any student to review instructional material that includes sexually explicit content and provide, as an alternative, nonexplicit instructional material and related academic activities to any student whose parent so requests.”*

Note: AR= Arkansas, CT=Connecticut, IA=Iowa, IL=Illinois, KY=Kentucky, VA=Virginia; SSFR=single-sex facility restrictions, THB=trans healthcare bans, TSB=trans sports bans, LGBTQ CL=LGBTQ Curricular laws

^*SB626 was signed into law in 2022 using a definition of sexually explicit content in Virginia code § 18.2–390 that included homosexuality in the definition of “sexual conduct”. Homosexuality was removed from the definition of “sexual conduct” in 2023.

Healthy Minds Data Collection

Individual-level data were collected in the 2021–2022 wave of the Healthy Minds Study (HMS),³² an annual web-based survey about mental health outcomes, resource availability/utilization, and related issues among undergraduate college students. The selection process involves inviting a random sample of students via email to participate in the study. The initial sample size is dependent on the size of the institution, with 4000 students recruited from campuses with an enrollment of > 4000 or more, and all enrolled students recruited from campuses with <4000 students enrolled. Participants must be at least 18 years old to participate. All research was approved by Institutional Review Boards at participating institutions. The 2021–2022 HMS survey was administered to 530 academic institutions.

Students who complete the survey are entered into cash sweepstakes to incentivize participation. Up to four follow-up reminders are sent during the month-long data collection period to encourage participation. The survey was administered using Qualtrics, a reliable survey software that ensures accurate and efficient data collection and asks questions about demographics, mental health status, and mental health service utilization/help-seeking. Additionally, the HMS network provides several elective surveys to institutions depending on their size, with topics concerning substance use, sleep, eating and body health, overall health, and peer support. The aggregated national data reports are made available to participating institutions to facilitate translating mental health research into practice.³²

Participants

The sample was established by a series of data cleaning and filtering steps. Participants over 45, international students, and those with missing data on gender identity or sexual orientation (less than 3% missingness for all) were excluded, recognizing documented disparities based on citizenship, like lower help-seeking rates among international students,³³ and the small sample size of participants over 45 years old as statistical outliers. Therefore, the sample is limited to U.S. citizens and permanent residents. For gender identity, participants were classified as either cisgender or transgender/gender non-conforming (TGNC) based on their responses to gender identity questions. Participants were excluded if they did not provide any information on gender identity. For sexual orientation, participants were classified as either heterosexual or LGBTQ+ based on their responses to sexual orientation questions. Participants were excluded if they did not provide any information on sexual orientation. This resulted in a total sample of 72,135 college students, including 21,466 LGBTQ+ college students.

Measures

Sexual Orientation.

In the HMS study, participants were asked to disclose their sexual orientation with the question: “How would you characterize your sexual orientation? (Please select all that apply).” Response options included heterosexual, lesbian, gay, bisexual, queer, questioning, asexual, pansexual, and an opportunity for participants to self-identify with a free-text prompt for further specification.³⁴ Responses of heterosexual were coded as a 0 and responses of lesbian, gay, bisexual, queer, questioning, asexual, and pansexual were coded as a 1.

Race/Ethnicity.

Participants were asked to report their race and ethnicity by selecting all that apply from a list of seven categories: American Indian/Alaskan Native (AI/AN), Arab/Arab American, Asian/Pacific Islander/Desi American (APIDA), Black/African American, Latino, White, and multiracial. Due to small cell sizes and for interpretability, racial/ethnic categories were combined into the following: APIDA, Black/African American, Hispanic/Latine, White, and other racial/ethnic groups.

Regional Level Variables

Region.

The HMS team supplements student data with institutional characteristics, including geography.³² In the HMS, the location of the institution (or geography) is coded as nine U.S. Census Regions: Region 1 (CT, ME, NH, RI, VT), Region 2 (NJ, NY, PA), Region 3 (IL, IN, MI, OH, WI), Region 4 (IA, KS, MN, MO, NE, ND, SD), Region 5 (DE, District of Columbia, FL, GA, MD, NC, SC, VA, WV), Region 6 (AL, KY, MS, TN), Region 7 (AR, LA, OK, TX), Region 8 (AZ, CO, ID, MT, NV, NM, UT, WY), and Region 9 (AK, CA, HI, OR, WA).

Anti-LGBTQ+ legislation.

State-level data were compiled into U.S. Census Regions aligned with regions used in the HMS. A cumulative score for each of the four legislative categories was calculated by adding the legislation codes (0,1) for each state (range = 0–4) and aggregating scores for the states in each region (range = 1–17). A proportion score was calculated by dividing the aggregated scores in each region by the number of states in that region.³⁵ This ensured that the regional anti-LGBTQ+ legislation score appropriately accounted for the varying number of states within each region (ranging from 3 to 8) and variability within each region related to proposed anti-LGBTQ+ legislation (e.g. in region 8 Colorado’s score was 0 and Arizona’s score was 3).

Individual Level Variables

Lack of sense of belonging and negative perceptions of campus environment.

To measure proximal stressors in the minority stress model, participants were asked about their agreement with statements about their belongingness to the campus community (e.g., I see myself as part of the campus community) and the impact of the campus environment on students’ mental and emotional well-being (e.g., At my school, I feel that the campus environment has a negative impact on student’s mental and emotional health). Responses were recorded on a six-point Likert scale (1 = strongly agree, 6 = strongly disagree). Higher scores on lack of sense of belonging indicate not feeling like a part of the campus community. Perceptions of campus environment were reverse coded so that higher scores indicated more agreement that the campus environment has a negative impact on student’s mental and emotional health.

First-generation student status.

To account for additional proximal stress factors in the minority stress model that may contribute to LGBTQ+ student depressive symptoms, first-generation student status was included in the analysis. The HMS questionnaire included two variables capturing the highest education level of participants’ parent(s), stepparent(s), or guardian(s). Participants reported the highest level of education of their parent(s), stepparent(s), or guardian(s) for each person separately from the following options: 8^th grade or lower, between 9^th and 12^th grade (but no HS degree), high school degree, some college (but no college degree, associate’s degree, and bachelor’s degree. Consistent with previous studies using HMS data, first-generation college student status was coded as 1 = no parent, stepparent, or guardian received a bachelor’s degree, and non-first-generation student status was coded as 0 = at least one parent, stepparent, or guardian received a bachelor’s degree.⁹

Transgender or gender non-conforming identity (TGNC).

The HMS questionnaire used binary variables to capture participants’ gender identity. Each variable represented a specific gender option, with 1 indicating selection and 0 representing non-selection. The options included male (1), female (2), trans male/Trans man (3), trans female/Trans woman (4), genderqueer/Gender non-conforming (5), self-identified (6), and gender non-binary (7). Responses of male or female (1–2) were coded as a 0 and responses indicating a TGNC identity (3–7) were coded as a 1.

Dependent Variable

Depression.

Participants’ levels of depression in the HMS were assessed using the Patient Health Questionnaire-9 (PHQ-9)³⁶ which consisted of nine items that measured the frequency of experiencing specific depressive symptoms over the past two weeks. These symptoms included a lack of interest or pleasure in activities, feelings of sadness or hopelessness, sleep disturbances, fatigue, changes in appetite, negative self-perception, difficulties with concentration, changes in motor function, and suicidal thoughts by rating the frequency of each symptom on a four-point scale (1 = Not at all, 2 = Several days, 3 = More than half the days, 4 = Nearly every day). Participant responses were summed with higher values indicating more depressive symptoms. The PHQ-9 had good reliability in the LGBTQ+ college student sample (α=0.89).

Data Analysis

The data analysis proceeded in several steps. First, descriptive statistics for the HMS sample and LGBTQ+ subsample were analyzed. Next, to account for data nested in regions, a series of multilevel fixed effects models were estimated with individual and regional-level variables. A baseline multilevel model with fixed effects and no covariates using the full HMS sample (N=71,235) was estimated. A baseline multilevel model with fixed effects and no predictors using the LGBTQ+ subsample (N=21,466) was also estimated. HMS includes the U.S. Census Regions as the geographic variable, which was used as the clustering level 2 variable for analysis purposes. Then, a series of hierarchical, fixed effects models were estimated by first adding the anti-LGBTQ+ legislation variable at level 2 and then adding each subsequent level 1 variable (lack of sense of belonging, negative perceptions of campus environment, first-generation student status, and TGNC) one at a time as predictors of depressive symptoms. All variables in the models were grand mean centered using the GRANDMEAN command in Mplus as our predictor of interest was at level 2.³⁷ Model 1 included total anti-LGBTQ+ legislation, the level 2 predictor of interest. Each model tested the inclusion of an additional level 1 predictor representing additional potential stressors in the minority stress model: lack of sense of belonging, negative perceptions of campus environment, first-generation student status, and TGNC. This hierarchical model building process was used to evaluate both model fit and the significance of each predictor represented in the minority stress model. Race/ethnicity were tested in models as control variables; however, were not included in the final models due to the number of parameters these dummy-coded variables added to the models and no improvement in model fit statistics (see supplementary tables 1 and 2). Model fit was evaluated with the addition of each predictor using the Akaike information criterion (AIC) and the Bayesian information criterion (BIC).³⁸ The BIC and AIC of the models were compared and the model with the lowest AIC and BIC was considered the best fitting model.³⁹ Each fixed effects model was clustered by region and adjusted by sampling weight. The sample probability weights constructed by HMS were used in the analyses. These weights adjust for potential differences between survey responders and non-responders by incorporating key demographic variables such as gender, race/ethnicity, academic level, and grade point average. Specifically, response weights were calculated as the inverse of the estimated probability of response, which was derived from logistic regression models that predicted response likelihood based on these demographic variables. (see Appendix A for further details). To ensure the robustness of findings, when applying these non-response weights in our analyses, the previously published standard HMS approach was followed.^9,40 Models were estimated using the MLR default in Mplus. All analyses were conducted in Mplus v8.⁴¹

Results

Descriptive Statistics

The total number of legislative bills coded for each U.S. Census Region is presented in Table 2. Of note, Region 5 had the highest number of attempts to pass anti-LGBTQ+ legislation, while Region 9 had the lowest number. Table 3 presents the demographic characteristics of the LGBTQ+ college student sample. Of note, mean depressive symptoms were higher in the LGBTQ+ subsample (M=12.15) compared to the larger HMS sample (M=9.53).

Table 2.

Total number of legislative bills coded for each U.S. Census Region

Region	# of States in Region	Single-Sex Facility Restrictions	Trans Healthcare Bans	Trans Sports Bans	LGBTQ Curricular Laws	Total anti-LGBTQ+ Legislation	% of Participants
1 – Northeast (New England)	6	0	1	3	2	6	29.1%
2 – Northeast (Mid-Atlantic)	3	0	1	2	2	5	6.4%
3 – Midwest (East North Central)	5	0	4	5	3	12	18.4%
4 – Midwest (West North Central)	7	2	3	6	3	14	2.4%
5 – South (South Atlantic)	8	1	4	7	5	17	14.0%
6 – South (East South Central)	4	3	4	4	3	14	5.5%
7 – South (West South Central)	4	1	4	4	3	12	2.1%
8 – West (Mountain)	8	0	4	6	3	13	8.3%
9 – West (Pacific)	5	0	0	1	0	1	13.6%

Table 3.

LGBTQ+ college student demographic characteristics (N=21,466)

Characteristic	Frequency	%
Age (Mean ± SD)	21.79 ± 4.25
Race/Ethnicity
APIDA	2067	9.6%
Black/African American	1671	7.8%
Hispanic/Latine	1954	9.1%
Multiracial	345	1.6%
White	15429	71.9%
First-Generation Student	6485	30.2%
Sexual minority	21367	99.5%
Transgender/gender non-conforming	3322	15.5%
Lack of sense of belonging (Mean ± SD)	3.29 ± 1.38
Negative perceptions of campus environment (Mean ± SD)	3.35 ± 1.38
Depressive symptoms (Mean ± SD)	12.15 ± 6.62

Notes: APIDA = Arab/Arab American, Asian/Pacific Islander/Desi American. Sexual minority includes college students who identified as lesbian, gay, bisexual, queer, questioning, asexual, or pansexual.

Multilevel Fixed Effects Models

A null multilevel model with fixed effects and no covariates using the HMS sample (N=71,235) was tested. There was no significant variance between regions (β=0.12, SE=0.07, p=0.084) indicating that there was no significant regional variation in depressive symptoms across the HMS sample. Then, a null multilevel model with fixed effects using the LGBTQ+ subsample from HMS (N = 21,466) was estimated. There was a small intraclass correlation (σ² = 0.005), indicating little variation at the regional level; however, there was significant variance between regions (β=0.23, SE=0.05, p<.001). Table 4 presents the model fit statistics. With the exception of the first predictor, each subsequent predictor resulted in reduced AIC and BIC indicating improved model fit.

Table 4.

Model fit statistics for multilevel fixed effects models

Model	AIC	BIC
Null	142133.64	142157.56
Model 1: + L2 Total anti-LGBTQ+ legislation	142134.36	142166.26
Model 2: + L1 Lack of sense of belonging	140612.45	140652.32
Model 3: + L1 Negative perceptions of campus environment	139966.12	140013.97
Model 4: + L1 First generation status	139772.03	139827.85
Model 5: + L1 Transgender/gender non-conforming identity	139452.42	139516.22

Note: L2 denotes the variable added at level 2 of the model and L1 denotes variables added at level 1 of the models.

Table 5 presents the standardized regression coefficients, standard errors, and p-values of the five models estimated. In Model 1, total anti-LGBTQ+ legislation was positively associated with increased depressive symptoms among LGBTQ+ college students (β=0.20, SE=0.09) indicating that as anti-LGBTQ+ legislation increased at the regional level so did depressive symptoms among LGBTQ+ college students. In Model 2, a lack of a sense of belonging on campus was significantly positively associated with increased depressive symptoms (β=1.25, SE=0.05) indicating that lacking a sense of belonging was associated with more depressive symptoms. In Model 3, more negative perceptions of the campus climate were significantly positively associated with increased depressive symptoms (β=0.80, SE=0.06), although this was a smaller effect than a lack of sense of belonging. In Model 4, being a first-generation student was significantly associated with an increase in depressive symptoms and had a larger effect than individual and regional variables estimated to this point (β=1.30, SE=0.15). Finally, in Model 5, TGNC identity was significantly associated with increased depressive symptoms (β=2.00, SE=0.08) and had the largest effect among all variables in the model. In Model 5, lack of sense of belonging (β=1.10, SE=0.06), negative perceptions of campus environment (β=0.80, SE=0.06), and first-generation status (β=1.35, SE=0.14) remained significant predictors of depressive symptoms. These findings indicate that LGBTQ+ students with a minoritized gender identity and first-generation student status and those lacking a sense of belonging and holding negative perceptions of the campus climate were at higher risk of experiencing increased depressive symptoms. Total anti-LGBTQ+ legislation, the variable of interest in this study, remained significant, albeit with small effects (β=0.23, SE=0.10), after the addition of individual-level stressors and a minoritized identity associated with increased depressive symptoms among LGBTQ+ college students.

Table 5.

Multilevel model fixed effects of regional and individual variables on depressive symptoms among LGBTQ+ college students (N = 21,466)

	Null Model β (SE)	Model 1 β (SE)	Model 2 β (SE)	Model 3 β (SE)	Model 4 β (SE)	Model 5 β (SE)
Intercept	12.58 (0.17)***	12.58 (0.16)***	12.42 (0.13)***	12.50 (0.16)***	12.49 (0.15)***	12.48 (0.16)***
Regional variables
Total anti-LGBTQ+ legislation		0.20 (0.09)*	0.21 (0.07)***	0.23 (0.11)*	0.21 (0.09)*	0.23 (0.10)*
Individual variables
Lack of sense of belonging			1.25 (0.05)***	1.16 (0.06)***	1.13 (0.06)***	1.10 (0.06)***
Negative perceptions of campus environment				0.80 (0.06)***	0.85 (0.06)***	0.80 (0.06)***
First generation status					1.30 (0.15)***	1.35 (0.14)***
TGNC						2.00 (0.08)***

Note: TGNC=Transgender or gender non-conforming

^*p<0.05

^**p<0.01

^***p<0.001

Discussion

This study found that proposed anti-LGBTQ+ legislation is associated with increased depressive symptoms among LGBTQ+ college students, who are already at increased risk of depression.¹⁰ Although the magnitude of this association was small, it held across models when individual-level predictors represented in the minority stress model and strongly associated with depressive symptoms were added. This finding supports recent research documenting how legislation restricting access to gender-affirming care also contributes to ambient and indirect harm among transgender individuals³⁰ and suggest that there is a harmful message inherent in anti-LGBTQ+ legislation that is contributing to depressive symptoms among LGBTQ+ college students. Although much of the legislation identified and coded in this study was aimed at public primary and secondary schools, the fact that legislators are debating the rights of LGBTQ+ youth and their parents to pursue gender affirming care and actively prohibiting the use of gender affirming practices and discussions of LGBTQ+ topics in public schools likely has implications for the broader LGBTQ+ community, including college students. Indeed, LGBTQ+ adults reported concerns for their safety and negative impacts on their mental health amid the expansion of proposed anti-LGBTQ+ legislation. In a recent poll, 79% of LGBTQ+ adults felt that their safety and that of their community was threatened and 80% of transgender and non-binary adults reported anti-LGBTQ legislation negatively impacted their physical and/or mental health.⁴² It is imperative that research continues to examine the effects of proposed anti-LGBTQ+ legislation on the mental and physical health of LGBTQ+ populations to inform policymakers and the public about potential harm.

TGNC individuals had a significantly increased risk of depressive symptoms in this study, with the highest magnitude of any other association. This finding adds additional support for targeting TGNC college students with mental health prevention and intervention due to their disproportionate risk for poor mental health outcomes.^43,44 Incoming TGNC freshmen experience depression, anxiety, and mental health disorders at significantly higher rates than cisgender freshmen.⁴⁵ With proposed legislation targeting gender affirming care for adults, these individuals may continue to experience disproportionate mental health burdens compared to their peers.⁴⁶ Additional research should focus on malleable risk and protective factors for TGNC college student depression as targets for prevention and intervention within colleges and universities.

The persistence of the association between anti-LGBTQ+ legislation and depressive symptoms among LGBTQ+ college students suggests that this type of legislation is detrimental to the entire LGBTQ+ community regardless of the age and subgroup targeted by the legislation. Proposed anti-LGBTQ+ legislation, that included legislation not specifically targeting youth such as religious exemptions from marriage ceremonies, religious exemptions in foster care and adoption, and religious exemptions in health care, was significantly associated with an increase in texts to the Crisis Text Line from LGBTQ+ youth in the days that the anti-LGBTQ+ legislation was proposed.²⁵ Therefore, while college students may not be explicitly impacted by proposed anti-LGBTQ+ legislation that targets their access to care or services, the rhetoric and media attention surrounding any attempts to pass legislation that infringes on access to gender affirming practices and the discussion of LGBTQ+ topics in schools appears to also be detrimental to their mental health. Indeed, evidence suggests that negative rhetoric from the 2016 presidential campaign contributed to fears of discrimination and violence, negatively impacting emotional well-being and physical health, in LGBTQ+ people aged 18–65.²⁹ Future research is needed to better understand how anti-LGBTQ+ legislation contributes to depressive symptoms among both the larger LGBTQ+ population and age groups, especially at risk within this population.

This study corroborates prior research⁴⁷ and supports the importance of belonging and community as protective factors within the minority stress model. The findings that lacking a sense of belonging on campus and negative perceptions of the campus environment were significantly associated with increased depressive symptoms among LGBTQ+ college students support the need for increasing adaptive coping skills and providing social support, both at the individual and community levels, to protect against distal stressors (i.e. proposed anti-LGBTQ+ legislation). A meta-analysis of student experiences in the U.S. found that LGBTQ+ college students were more likely to report feeling unsafe on campus and in the surrounding community; perceived campus as less supportive; were less likely to feel a sense of belonging on campus; and reported lower quality interactions with other students, advisors, faculty and staff.⁴⁵ Future research should examine interventions that improve sense of belonging and perceptions of the campus environment among college students as avenues for improving LGBTQ+ student mental health.

Data for this study was collected during the COVID-19 pandemic (2021–2022), and while this study did not directly address the influence of the COVID-19 pandemic on LGBTQ+ college students’ mental health, the mechanisms identified by Hager et al.⁴⁸ offer insights into how pandemic-induced changes in social interactions and daily routines may have influenced the association between proposed anti-LGBTQ+ legislation and depressive symptoms among LGBTQ+ college students. Specifically, loneliness among college students during the pandemic’s early stages was associated with increased depression, mediated by factors such as boredom and repetitive negative thinking. Repetitive negative thinking may also be influenced by proposed anti-LGBTQ+ legislation and, when coupled with social isolation and restricted activities during the pandemic, could have exacerbated feelings of loneliness potentially leading to heightened depressive symptoms. Future research should explore the complex interplay between COVID-19-related stressors, proposed anti-LGBTQ+ legislation, and mental health outcomes in LGBTQ+ college students.

Implications for Universities: Practitioners and Policies

People working in college mental health (e.g., counselors, psychiatric providers, primary care providers) should be aware of proposed legislation that may negatively impact the mental health of LGBTQ+ students and affect clinical care, such as gender-affirming care and group therapy, for these students. Practitioners should advocate for the retention of services that positively impact the well-being of LGBTQ+ college students, which could be lifesaving. Specifically, the Trevor Project has reported that LGBTQ+ college students with access to college mental health services had 84% lower odds of a suicide attempt compared to college students without access.⁴⁹ Maintaining and increasing access to mental health services is critical at a time when LGBTQ+ college students are under increased stress.

On an individual level, it is important for practitioners to compassionately hear what college students are experiencing and offer empathy. Practitioners report that their LGBTQ+ college students are experiencing increasing levels of stress related to proposed anti-LGBTQ+ legislation.^26,50,51 Encouraging students to take care of themselves, advocate where they feel they can safely do so on campus or through political groups, and stay hopeful about the future is essential to good clinical care. In addition, experts recommend that campus leaders support LGBTQ+ campus organizations and student groups, include LGBTQ+ topics in multicultural diversity and equity efforts on campus, offer stand-alone educational programs to provide opportunities for deeper exploration of issues affecting LGBTQ+ students, and offer faculty development focused on improving classroom climate for LGBTQ+ and other minoritized college students.⁵² Future research is needed to identify how these efforts may improve LGBTQ+ college students’ health outcomes in the face of increasing anti-LGBTQ+ legislation.

Strengths and Limitations

Strengths of this study include the comprehensive approach to coding several types of legislation that had been proposed, not just passed, to provide a more holistic picture of anti-LGBTQ+ legislation in particular regions. The large-scale, national HMS data allowed for a larger, more nationally representative sample of LGBTQ+ college students and enhanced the generalizability of these findings given the multisite nature of the HMS campuses, in addition to the randomized sampling design. However, the results should be interpreted considering the following limitations. Region represents the geography of the institution the college student attends, not the geography of their home state. The researchers assumed that college students are aware of the political environment around them and have some level of exposure, although the level of exposure was not explicitly measured. The coding scheme for proposed anti-LGBTQ+ legislation attempted to account for attention from media and advocacy groups by only considering bills that had a legislative action beyond introduction, as simply introducing the bill may not have garnered as much attention as a bill moving through the legislative process. The categorization of regions may overlook the variation in legislative climates among states within each region, potentially impacting the generalizability and accuracy of the results. For example, the South Atlantic Region (Region 5) includes eight states with different legislative contexts within each state. The use of a proportion score accounted for the variability in number of states within each region; however, future research should examine how proposed anti-LGBTQ+ legislation at the state level impacts LGBTQ+ individuals’ mental health within that state. Excluding participants who did not disclose their gender identity could affect the study’s representation of gender diversity and may bias associations rates by under-/over-estimation, which may limit the generalizability of this study. Future research should employ inclusive strategies to mitigate social desirability bias and ensure the inclusivity of all participants, regardless of their willingness to disclose sensitive personal information such as gender identity. College students may be exposed to different anti-LGBTQ+ legislation and rhetoric depending on where they go to college when compared to their permanent home residence, which may influence the impact of legislation on college students’ mental health. Future research should seek to measure exposure to media and rhetoric around proposed anti-LGBTQ+ legislation to determine if there is a dose-response relationship with mental health outcomes. Further, anti-LGBTQ+ legislation was coded as a yes/no binary which prevents interpretations of substantive differences among different pieces of legislation and how the ultimate legislative outcome (i.e. was the legislation vetoed?) was associated with depressive symptoms. Future research should examine these important considerations to better understand the mechanisms underlying associations between anti-LGBTQ+ legislation and mental health in LGBTQ+ populations. This study was also limited by the cross-sectional nature of the data, as both legislative data and mental health data came from the same year. Thus, this is a study of association only. Interpreting a causal relationship will require additional evidence including longitudinal tracking of proposed anti-LGBTQ+ legislation and student mental health.

Conclusion

This study enriches the expanding body of research on the impact of proposed anti-LGBTQ+ legislation on LGBTQ+ individuals’ mental health. Proposed anti-LGBTQ+ legislation that limited access to gender affirming care and policies and discussions of gender and sexual identity topics in schools was associated with increased depressive symptoms among LGBTQ+ college students. Furthermore, a lack of a sense of belonging on campus and a negative perception of the campus environment were significantly associated with increased depressive symptoms among LGBTQ+ college students. In addition, TGNC students were significantly more likely to report increased depressive symptoms. Already at increased risk of poor mental health outcomes, LGBTQ+ college students may need even more mental health support in regions with more proposed anti-LGBTQ+ legislation, and colleges and universities should be prepared to ensure access to inclusive college mental health services. Future research should focus on examining the influence of college mental health services in regions with increased anti-LGBTQ+ legislation on LGBTQ+ student mental health, which could pave the way for a more inclusive and supportive educational environment.

Appendix

Survey Weighting using Healthy Minds Study Data in Analysis

Survey studies are susceptible to the issue of non-representativeness among respondents compared to the overall population. In the case of the Healthy Minds Study (HMS), the invited participants are randomly selected from the complete list of currently enrolled students, ensuring a representative sample. However, it remains important to consider potential differences between survey completers and non-completers. Response rates across survey years varied, ranging from 13% to 27%. To address this concern, non-response weights were constructed using administrative data on the full student population, with the approval of Advarra and participating schools’ Institutional Review Boards. Sex, race/ethnicity, academic level, and grade point average were utilized, when available, to estimate the likelihood of response for different student types using multivariate logistic regressions. Propensity weights based on these response probabilities were then assigned to each survey completer. Higher weights were given to student types less likely to participate, ensuring that the weighted estimates represent the entire student population in relation to the available administrative variables at each institution. Importantly, these sample weights grant equal aggregate weight to each school in national estimates, as opposed to weighting proportional to school size. This decision was made to prevent the overrepresentation of schools with larger enrollments in the overall national estimates. The use of these weighted estimates in this study enhances the representativeness and generalizability of the findings from the Healthy Minds Study data analysis.

Data Sharing:

The data that supports the findings of this study are available from the Healthy Minds Study, but restrictions apply to the availability of these data and so are not publicly available. However, data are available from the authors upon reasonable request and with permission of the Healthy Minds Study.