Introduction

The COVID-19 pandemic prescribed strict mobility and social restrictions across the globe. Mexico was not the exception and implemented a social distancing policy on March, 2020, which included shutting down non-essential economic activities (schools, universities, restaurants, gyms, and other recreational spaces), as well strong recommendations to use facemasks, antibacterial gel and to socially isolate at home. As with other populations, such restrictions resulted in lesbian, gay, bisexual, trans, and queer’s (LGBTQ+) people spending more time in their homes, alone or with families and reduced access to their LGBTQ + spaces and communities.

Research on LGBTQ + mental health, shows that sexual and gender minorities are affected by unique stressors, such as discrimination, internalized homophobia, victimization, felt stigma, and expectation of rejection (Meyer et al., 2021), that have impact on depression, anxiety and drug use (Meyer, 2003). Studies on Mexican LGBTQ + populations suggest that mental health is strongly impacted by internalized homonegativity, which in turn is affected by homophobic discrimination (Lozano-Verduzco et al., 2017; Mendoza-Perez & Ortiz-Hernandez, 2020; Mendoza-Pérez & Ortiz-Hernández, 2019), which is frequently experienced in families, schools, workplaces, and in public spaces (Lozano-Verduzco & Salinas-Quiroz, 2016; Mendoza et al., 2015; Mendoza-Perez & Ortiz-Hernandez, 2019, 2020). The only available recent Mexican study exploring stigma found that 60% (Mendoza et al., 2015) of LGBTQ + people reported being discriminated against based on their sexual orientation and/or gender identity and expression.

LGBTQ + Mental Health, Social Support, and Personality Traits

There is very little research on Mexican LGBTQ + communities during the COVID-19 pandemic (Mendoza-Pérez, 2021; Ramírez-García et al., 2022). The little research regarding LGBTQ + mental health and social support in Mexico shows that community connectedness—as an indicator of social support– mediates the relationship between homonegative violence and mental health, showing that internalized homo and transnegativity correlate positively with adverse mental health outcomes (Lozano-Verduzco et al., 2017). According to Brown and Scheid (2010), social support refers to the role that family, friends and other important people play in helping a person cope with stress or in exposing them to more stressors. Social support can present itself in various forms: emotional support (love, care, empathy, etc.), structural support (money, housing, health services, etc.), or a combination of the two (Brown & Scheid, 2010). A meta-analysis showed that social support is a moderate mediator of stress and mental health indicators (Harandi et al., 2017).

Research outside of Mexico and Latin America suggests that depression and stress amongst LGBTQ + communities during the COVID-19 pandemic increased (Akré et al., 2021; Kneale & Bécares, 2021; Rodríguez-Seijas et al., 2020). Mental health problems were higher among LGBTQ + populations that reported harassment (prior to or during the confinement) because of their LGBTQ + identity, than their heterosexual and cisgender counterparts (Akré et al., 2021; Kneale & Bécares, 2021; Rodríguez-Seijas et al., 2020). Kamal and colleagues (2021), amongst a US sample, reported that sexual and gender minorities experienced higher levels of depression and post-traumatic stress disorder compared to participants who were not part of sexual and gender minority groups during the pandemic lockdown. In Portugal, research found that during the pandemic, a hostile family climate for LGBTQ + people was associated with higher levels of depression and anxiety (Gato, Leal & Scabra, 2020). In Chile, Barrientos, and colleagues (2021) reported that confinement negatively affected the emotional lives of sexual and gender minorities.

Previous research on Mexican LGBTQ + mental health shows that between 7 and 20% of participants report at least one symptom associated with depression; 17% reported suicidal ideation, and 4.6% reported a suicide attempt during the previous 12 months (Lozano & Salinas-Quiroz, 2016). Depressive symptoms are associated with discrimination, and alcohol use is associated with homonegative discrimination and violence, as well as moderate levels of community connectedness (Lozano-Verduzco et al., 2017)Mendoza-Perez (2021), reports that almost 25% of LGBTQ + respondents used drugs other than alcohol and tobacco during the pandemic, mostly marihuana, and to a much lesser degree, cocaine, and LSD. According to our findings, no differences or correlations have been reported between race/ethnicity and mental health in Mexican LGBTQ + groups. Researchers (Pucket et al., 2016), have turned to understand the role of personality traits to further inquire on the mental health of LGBTQ + people, particularly neuroticism. This trait refers to negative affect, high levels of self-consciousness, impulsiveness, and vulnerability (McCrea & Costa, 1991), and places individuals at risk of internalizing negative messages about their sexual orientation and/or gender identity, because neuroticism results in more frequent feelings of shame, guilt, and inferiority (Pucket et al., 2016).

In the lives of Mexican LGBTQ + populations, families (including extended family) and friends are the primary sources of social support (Medellín et al., 2012) and social support in turn decreases levels of depression (González et al., 2018), and increases levels of subjective wellbeing (Aranda et al., 2020). Díaz-Guerrero (2003), after fifty years of studying Mexican families, concluded that Mexican culture is a family centered one where family members depend on each other to carry out their daily lives (Torres et al., 2015): Mexican families not only provide a space for socialization, but provide the core value of individuals, and the very center of social life.

Data on the general Mexican population shows that during the pandemic confinement, anxiety and depression increased, quality of sleep decreased (Terán-Pérez et al., 2021), and stress was mainly derived from the family (Landa-Blanco et al., 2021). However, there is a void in knowledge on how the confinement affected LGTBQ + people, considering that this is a population that is already affected by discrimination that constantly affects their mental health. This study aims to describe the mediating role of perceived social support and its effects on depression, anxiety, and stress, as well as to understand the predictive power of social support on suicidal ideation, and suicide attempt.

Method

This study is part of a larger research project carried out in six other countries by researchers interested in understanding the contexts of the COVID-19 pandemic on LGBTQ + groups. The aim of this international project was to understand the dynamics of violence and other social determinants of health that affect LGBTQ + groups during the pandemic and compare these dynamics between countries. The project in Mexico aimed to understand the effects of confinement, discrimination, and personality traits on LGBTQ + sexual and mental health. In this text, we only present Mexican data. From the literature review, we identified a series of social and personal determinants that have proven to correlate with the mental health indicators that are explored here. Considering the gravity of the lockdown, we gave priority to the social elements that changed because and during the pandemic. Considering this, we hypothesized that (a) LGBTQ + participants with higher level of perceived social support from family would report better mental health; (b) reported isolation from family and community will negatively correlate with mental health indicators; (c) reported discomfort with family will correlate with mental health indicators; and (d) that neuroticism will correlate with mental health indicators.

Measures

Depression, anxiety, and stress scale 21 (DASS-21)

Daza and colleagues (2002) validated the scale for Spanish speakers. They found that it statistically and significantly correlated with other scales that measure the same construct and that it statistically and significantly discriminates, allowing for differentiation between groups (Gloster et al., 2008). For this study, the measure proved to have good reliability for each subscale: depression (α = 0.92; ω = 0.92), anxiety (α = 0.87; ω = 0.89) and stress (α = 0.89; ω = 0.89). It uses 21 items, seven for each subscale, with a 4-point Likert-type response format.

Neuroticism-Extraversion-Openness Five-Factor Inventory (NEO-FFI)

We used the NEO-FFI short subscale for neuroticism, validated in a Mexican population. It has six items with a 5-point Likert-style scale and, in this study, proved to have good reliability (α = 0.87; ω = 0.87). Construct validity is shown in the: (a) measure’s correlation with the NEO-FFI’s other four subscales (Meda Lara et al., 2015); (b) higher neuroticism scores correlating with lower levels of resilience (Solís-Cámara et al., 2017); (c) higher neuroticism scores predicting mental health issues (Vargas-Terrez et al., 2015).

Multidimensional Social Support Scale

Validated in the Mexican population (Arechabala & Miranda, 2002), this scale consists of 12 items on a 7-point Likert-type scale. For this study, both subscales proved to have good reliability: family support (α = 0.90; ω = 0.90) and friend support (α = 0.95; ω = 0.95). Literature reports that the measure positively correlates with similar measurements, such as self-efficacy (Mosqueda et al., 2015), and negatively with scales such as the DASS-21 (Barrera-Herrera et al., 2019).

Suicidal ideation and attempt

Two dichotomous questions from the Composite International Diagnostic Interview (Borges et al., 2010) were used to analyze LGBTQ + mental health (Mendoza-Perez & Ortiz-Hernandez, 2019, 2020). The first asked if the participant had “seriously considered committing suicide during the COVID-19 lockdown”, and the second asked if the participant had “tried to commit suicide during the COVID-19 lockdown”.

Internalized homo/transnegativity

A nine-item measure with a 7-point Likert-type answer scale that ranges from “completely agree” to “completely disagree” was built based on Mohr & Kendra’s (2011) work. For this study, the measure had decent reliability with α = 0.84; and ω = 0.85. Construct validity in Mexico shows that higher scores on the scale are positively associated with a higher risk of suicidal ideation and mental health issues (Ortiz-Hernández, 2005). Items were adapted to include transnegativity: “If I could choose, I would prefer to be heterosexual or cisgender.“

Isolation from Friends

we created two ad hoc questions to evaluate isolation from friends. Participants were asked how much they felt being isolated from their a) heterosexual and cisgender friends and b) from their LGBTQ + friends affected them. Responses varied from “did not affect me at all” (0) to “extremely affected me” (10). The reliability of these questions had a value of .79 according to McDonald’s omega and Cronbach’s alpha.

Isolation from Intimate Partner

we created a single ad hoc question to evaluate isolation from intimate partner. Participants responded how affected (0= “did not affect me at all” to 10= “extremely affected me”) they felt from being isolated from their intimate partner.

Discomfort with Family

participants responded to one question—created ad hoc for this study– that assessed how uncomfortable they felt with their families (0= “extremely comfortable” to 10= “extremely uncomfortable”) during quarantine, and one question that assessed how uncomfortable they felt expressing their LGBTQ + identity with their family. Each question was treated as a separate variable.

Fear of COVID-19

one ad hoc question explored how much fear participants felt when confronted with the possibility of getting infected with the virus (0= “no fear” to 10= “extreme fear”).

Drug use

questions from the National Survey on Drug, Alcohol, and Tobacco Use (ENCODAT, 2017) were presented to participants. The questions asked if they used illegal drugs during the previous month of answering the survey, a question that was answered with “yes” (1) or “no” (0).

Sexual orientation and gender identity

This was evaluated through one single question: “Which of the following best describes how you identify?“ which could be answered by choosing one of the following: gay/homosexual, lesbian, bisexual man, bisexual woman, trans man, trans woman, queer, non-binary or other, as proposed by Mendoza-Pérez (2021), and colleagues (2015).

Discrimination

This was evaluated through two questions to evaluate discrimination in lifetime and discrimination during the lockdown, following the recommendation from the National Council to Prevent Discrimination (CONAPRED, 2022). The first asked if, due to the participant’s sexual orientation or gender identity, they were: discriminated against, received verbal aggression, received physical aggression, or were discriminated against in another way. The second question asked the same thing, but during the lockdown, and had the same response options.

Sociodemographic Characteristics

These include age, sex assigned at birth (male, female or intersexual), race/ethnicity (Latinx/mestizx, white, native, European, Asian, black/afromexican), educational attainment (pre-school, elementary, middle, high, college, masters, Ph.D.), and monthly household income. These variables were used as predictors of mental health.

Participants and Procedure

This is a cross-sectional study with a non-probabilistic sample. The questionnaire was available online between August and October of 2020. Recruitment was carried out solely online, through the researchers’ personal and academic networks by email and on Facebook, Twitter, and Instagram, as well as paid advertisements on those outlets. Non-profit organizations throughout the country also shared the e-flyer and questionnaire link on their social media. Inclusion criteria included identifying as LGBTQ + and living in Mexico at the time of the study. Participants who did not meet these criteria were excluded from further questions. We received a total of 1,878 participations, of which 1,525 answered at least half of it, thus the size of the sample used was 1,525. The handling of missing data included the listwise exclusion technique, which consists of conducting the analyses only on cases that have complete data for the variables used in the analysis (Pepinsky, 2018). Therefore, in mediation models and other similar models, the sample size may have decreased, resulting in a final sample of 894 cases.

Analysis

Descriptive analysis was carried out to identify the general behavior of each variable in the current sample. Secondly, bivariate correlation analyses were carried out, and the results are presented in Table II. Correlations allowed for a mediation model using the method of Maximum verisimilitude as the estimation method. Perceived social support was introduced as a mediating variable, considering that previous literature reports that social support buffers the effect of stress on mental health (Harandi et al., 2017). These models controlled for LGBTQ + identity, age, race/ethnicity, and education attainment. Normal distribution of residuals through cuantil-cuantil graphics and the absence of multicollinearity between variables were analyzed for all models (Fox & Weisberg, 2018). Logistic regression models for suicidal ideation and attempt, controlling for the same variables as previously mentioned models, were also carried out. These models report estimations of the odds ratio for each variable (Hosmer & Lameshow, 2004). Unstandardized and standardized beta coefficients were also calculated for all models (only standardized betas are presented). Only when we found a unique correlation not reported in previous data, we carried out partial correlations to further inquire Analyses were carried out with Jamovi 1.6 software.

Ethical Considerations

The Ethics Committee (IRB) of the Faculty of Medicine of the National Autonomous University of Mexico approved this project’s protocol under project code FM/DI/044/2020. Informed consent was presented at the beginning of the questionnaire, which explained the study’s objectives and guaranteed that participation was confidential, anonymous, and voluntary. Parental consent was waived for participants under 18, as this type of consent is likely to put underage participants at risk (Craig et al., 2017; Mutanski, 2011).

Results

In this section, we present the main characteristics of the sample (Table I, N = 1,525), and five regression models, one per variable of interest: depression, anxiety, stress, suicidal attempt, and suicidal ideation (Tables II and III). Table IV shows the correlations between variables. All models were statistically significant and controlled for sex assigned at birth, age, race/ethnicity, LGBTQ + identity, and living context.

Table I Sociodemgrophic characteristics of participants
Full size table
Table II Predictive mediation models for depression, anxiety, and stress, mediated by social support
Full size table
Table III Predictive models of suicidal ideation and suicidal attempt, complete models
Full size table
Table IV Correlations between variables
Full size table

Descriptive analysis shows that on a scale from 1 to 10 (one being extremely low and ten extremely high), participants felt isolated from friends (M = 6.5, SD = 2.8), because of the pandemic. However, on the same scale, participants reported feeling more comfortable with their family (M = 4.6, SD = 3.3), moderate isolation from their intimate partner (M = 3.7, SD = 4.0), and moderate suffocation by not being able to freely express their LGBTQ + identities (M = 3.8, SD = 3.9). Participants scored 5.8 (SD = 2.9) on how scared they were to acquire COVID-19 and felt well informed on the topic (M = 7.8, SD = 1.8). Participants reported relatively high scores on neuroticism (M = 31.59 out of 50, SD = 10.04), perceived family support (M = 52.47 out of 70, SD = 12.92), and perceived social support from friends (M = 56.85 out of 70, SD = 13.83). Finally, participants reported relatively low scores on mental health indicators, such as depression (M = 19.40 out of 40, SD = 8.17), anxiety (M = 17.44 out of 40, SD = 7.14), stress (M = 21.43 out of 40; SD = 7.64), and internalized homonegativity (M = 26.18 out of 70, SD = 12.81).

Mediating models showed that the indirect effect of predictive variables on social support in all models explained 35% of the variance of social support. As shown in Table II and III, neuroticism was one of the most consistent predictors, and in general, had a direct effect on all models, except depression, where it had both a direct and indirect effect mediated by social support. In other words, predictive variables can have a direct effect on anxiety, stress, and depression, but in the case of depression, this effect can be mediated by the presence or absence of social support.

Depression

The model predicts 56% of the variance for depression. As shown in Table II, depression is mediated by social support, and it has both a direct and indirect effect on mental health. Results show that neuroticism, discomfort with family, and internalized homonegativity have positive correlations with depression, indicating that the higher the levels of these variables, the higher the probability of experiencing depressive symptoms. On the other hand, positive perception of family, and alcohol use lead to lower reported levels of depression.

Anxiety

This model also proved statistically significant and predicts 41% of anxiety’s variance. Table II shows that all effects on this variable were direct, they were not mediated by social support. Discomfort with family, isolation from family, friends and significant other, and fear of acquiring the virus show significant correlations with anxiety, indicating that higher levels of these variables are likely to lead to high levels of anxiety. However, isolation from LGBTQ + friends was inversely related to anxiety indicating that when isolation from LGBTQ + was higher, anxiety was lower. Because previous literature shows that connection to other LGBTQ + friends moderates and on occasions, alleviates anxiety (Lozano-Verduzco et al., 2017), we expected that higher levels of isolation from LGBTQ + friends would indicate higher levels of anxiety. To further inquire this novel association, partial correlations between anxiety and isolation from LGBTQ + friends, controlling for internalized homonegativity, neuroticism and isolation from straight and cis friends was carried out, showing that isolation from LGBTQ + friends is related to higher reported levels of anxiety.

Stress

The model predicts 46% of the variance in stress. As can be seen in Table II, all variables had direct effects on stress. Social support did not act as a mediating variable for stress. Discomfort with family, fear of acquiring the virus, neuroticism and having been discriminated against during the lockdown were directly correlated with stress.

Suicidal Ideation

This predictive model resulted to be statistically significant controlling for variables (χ2 = 103; gl = 18; p < .001), as well as without this control (χ2 = 328; gl = 37; p < .001), and predicts 45.8% of suicidal ideation variance, according to Nagelkerke’s pseudo- R2. The model presents an 85.1% precision of the area under the curve (AUC) ROC of 86.7%. No variables presented multicollinearity issues. Interestingly, age and fear of acquiring COVID-19 were inversely associated with suicidal ideation, indicating that young participants and participants with lower levels of fear, have a higher probability of reporting suicidal ideation. In contrast, discomfort with family and neuroticism were positively correlated to suicidal ideation, indicating that participants who felt more discomfort with their families and reported higher levels of neuroticism, were more likely to report suicidal ideation. Finally, a worrisome finding is that trans women report a 3.92 higher probability of presenting suicidal ideation in comparison to gay men.

Suicide Attempt

Lastly, the predictive model for suicide attempt was also statistically significant, but not when controlling for variables (χ2 = 22.2; gl = 18; p = .223), only without these controls (χ2 = 100.9; gl = 37; p < .001). Model change analyses indicated that the inclusion of controls was statistically significant (χ2 = 78.7; df = 19; p < .001), which indicates that control variables on their own, do not explain suicide attempt. Suicide attempt is explained by the interaction between control variables (age and educational attainment) and predictor variables. This last model explains 31.9% of the variance, according to Nagelkerke’s pseudo-R2, and reports a precision of 95.3% and an AUC of 87.8%. No variables presented issues regarding multicollinearity. As can be seen in Table III, neuroticism and discrimination during the lockdown were directly associated with a suicide attempt. Being discriminated based on sexual orientation or gender identity, increased the probability of suicide attempts by 2.49. However, the use of drugs was inversely correlated to a suicide attempt, indicating that the use of drugs other than tobacco and alcohol decreased the possibility of suicide by 72%.

Discussion

The objective of this study was to describe and predict five mental health indicators and understand the mediating role of social support on these indicators in a Mexican LGBTQ + sample during the COVID-19 pandemic. It is possible that the lack of association with some variables in the multivariate models is due to them being confounding variables of other factors. For instance, neuroticism, as a personality trait, may be influencing other variables involved in the expression of behavior in interpersonal relationships. Therefore, variables such as perception of victimization (like discrimination) or levels of internalized homonegativity may already be explained by personality traits such as neuroticism (Pucket et al., 2016). Findings suggest that neuroticism plays a central role in predicting each of the five mental health indicators, and that perceived social support plays a mediating role for depression, but not for any other indicator. LGBTQ + populations in this study reported feeling isolated from their LGBTQ + and straight/cisgender friends but did not feel isolated from their partner or family. This is probably because most participants reside with their families (57.4%) or returned to their family home during lockdown (7.3%). Almost 16% of participants live with their intimate partner, which may help explain why the level of felt isolation from these participants is low. Participants also reported high levels of social support from their families and friends, and positively perceived their family. This may be because, in a family-centered culture, Mexican families respond with solid support during crises and act as a strong network to fulfill each other’s needs, and if this is so, the lockdown could have been an opportunity for LGBTQ + people to connect with their families. At the same time, previous research shows that family is also a place for discrimination. Results suggest that social support is not all-or-nothing, but that it appears temporarily and circumstantially between certain family members. There is no previous research on Mexican LGBTQ + and their families, but data on community connectedness shows that feeling connected to others like them leads to low levels of depressive symptoms (Lozano-Verduzco et al., 2017).

The models presented here partially support the hypothesis that social support acts as a mediating variable on mental health. Social support only buffers depression, but not any other mental health indicator. This finding suggests that social support does not always influence mental health, but that it has multiple facets, and centers on two main axis: family and friends. In this sense, exploring discomfort with family, positive perception of family and isolation from friends, allowed a richer understanding of the relationships between social support, family, and mental health. Discomfort with the family is a variable that allowed to understand the negative aspect of family life, showing that it was directly related to anxiety, stress, depression, and suicidal ideation, proving to be a strong predictor of mental health, indicating that for LGBTQ individuals, feeling free to express their identity with their family is relevant to their mental wellbeing. Results suggest that discomfort had stronger effects on wellbeing than social support. This interpretation is further supported by the fact that positive perception of family (e.g., LGBTQ + participants seeing their family in a positive light), decreases levels of depression. In addition to family support, peer and partner relationships are also fundamental to social support and wellbeing. Data shows that these peer and partner relationships also impacted LGBTQ + participants’ mental health during lockdown. Specifically, isolation from friends and intimate partner have a direct effect on anxiety, increasing its levels. Contrary to what previous literature supports (Lozano-Verduzco et al., 2017; Meyer et al., 2021), internalized homonegativity only has an indirect (mediated by social support) effect on depression, suggesting that people who are uncomfortable with their sexual and gender minority status may feel more depressed, only if they are lacking social support.

This study found that experiencing discrimination during lockdown nearly triples the probability of suicide attempts. On the other hand, using drugs other than tobacco and alcohol decreases this probability. Drugs may be used during the lockdown to cope with isolation and discrimination, as drug use is part of gay culture, which allows for forms of socialization and community building, and normalizes drug use as a coping mechanism (Mendoza-Pérez, et al., 2022). Substances are usually available in LGBTQ + settings, and the negative correlation found in these results may even indicate possible abuse symptomatology in the studied sample (Felner et al., 2020), particularly considering that 85% of the sample reported using marihuana during the month prior to answering the survey. Furthermore, a significant predictor of suicidal ideation is discomfort with the family. The pandemic’s mandatory confinement thus poses severe issues for Mexican LGBTQ + people, as it requires them to be in constant interaction with their family, where discomfort appears, leading them to anxiety, stress, and suicidal ideation, at the same time, allows social support from family.

Conclusions

The data presented here, show some of the particularities, risks, and support that a Mexican LGBTQ + sample faced during the COVID-19 pandemic. An important finding is that participants who felt comfortable with their families reported better levels of mental wellbeing, which speaks to the importance that family has for this LGBTQ + sample in the Mexican context. In other words, those who felt comfortable with their families, who did not experience isolation from their important relationships (friends, peers, and intimate partners) reported little symptoms of depression, anxiety, and stress. Despite that homonegativity is most frequent between family members (Lozano-Verduzco et al., 2017; Mendoza-Pérez, 2015), this data points to understanding family as a space that supports the mental health of LGBTQ + folks. Data also shows that participants report higher levels of wellbeing when they can carry on with relationships outside of the family. Social support is a very relevant variable in understanding mental health of sexual and gender minorities, as it also ameliorates depression when participants report internalized homophobia. In other words, social support can help LGBTQ + people in feeling well, despite their discomfort with their sexual orientation and/or gender identity.

Furthermore, data supports that a personality trait—neuroticism—is the strongest predictor of poor mental health. Previous research shows that neuroticism is a mediating variable to understand poor mental health amongst Mexican Americans with childhood trauma (Knight et al., 2023), and that it poses emotion regulation difficulties for adolescents (Singh, 2022). Our results may indicate the development of neuroticism as a personality trait for LGBTQ + Mexicans. Neuroticism indicates emotional insecurity and instability, anxiety and constant tension which requires further research to understand if it’s related to violence exerted toward sexual and gender minority individuals. Data also supports previous findings, particularly that trans women are at the highest risk of suicidal ideation and attempt (Lozano-Verduzco & Melendez, 2019). A novel finding is that this LGBTQ + sample, relied on drug use to ameliorate mental health issues and suicidal attempts. Actions to decrease homonegative attitudes and discrimination against LGBTQ + communities are urgent in Mexico since this may lead to LGBTQ + individuals to create and support friendship networks. Clinical practice is fundamental for the well-being of LGBTQ + people in Mexico and must recognize the centrality of family and friendships for these minorities, as well as the connection between neuroticism and mental health. We suggest that clinical practice must focus on the effects of the homonegative stigma that arises in these two contexts but also recognize their importance in supporting LGBTQ + individuals. Clinical practice may also benefit from considering the LGBTQ + member’s family, friendships and their dynamic, as they are central to provide support. It is relevant that future research explores other variables and personality traits, particularly neuroticism, that can be associated with mental health. Future research may benefit from focus on the fluctuating character of social support and family dynamic perceived by LGBTQ + people.

Limitations

These results cannot be generalized to the Mexican LGBTQ + population, as the sample was not probabilistic. Notwithstanding, the results represent a significant sample. Online questionnaires can also imply a bias in participant selection because people without internet access, low educational attainment, and limited use of electronic devices can be underrepresented. In this regard, we must also recognize that almost half of participants identified as gay men, as is typical in Mexican LGBTQ + research. Gay men are usually of middle class, with high educational attainment, and access to information and communication technologies (Lozano & Salinas-Quiroz, 2016). Four of the measures were created ad hoc for this project and consist of only one question, which does not paint an in depth picture of this indicators. Future research must consider including lesbian, bisexual, trans and gender non-conforming participants, as well as explore the particularities of discomfort with families. It is also necessary to implement other methodological strategies to reach these groups. These limitations may pose threats to external validity, and results should not be generalized to the Mexican LGBTQ + population. However, the study can be considered adequate internal validity, as it allows us to understand the dynamics of unique minority stressors and their relationship to mental health for LGBTQ + people.

All procedures followed were under the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000 (5). Informed consent was obtained from all participants for being included in the study.