This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Mental Health, is properly cited. The complete bibliographic information, a link to the original publication on https://mental.jmir.org/, as well as this copyright and license information must be included.
Sexual and gender minority (SGM) individuals experience minority stress, especially when they lack social support. SGM young adults may turn to social media in search of a supportive community; however, social media use can become problematic when it interferes with functioning. Problematic social media use may be associated with experiences of minority stress among SGM young adults.
The objective of this study is to examine the associations among social media use, SGM-related internalized stigma, emotional social support, and depressive symptoms in SGM young adults.
Participants were SGM young adults who were regular (≥4 days per week) social media users (N=302) and had enrolled in Facebook smoking cessation interventions. As part of a baseline assessment, participants self-reported problematic social media use (characterized by salience, tolerance, and withdrawal-like experiences; adapted from the Facebook Addiction Scale), hours of social media use per week, internalized SGM stigma, perceived emotional social support, and depressive symptoms. Pearson correlations tested bivariate associations among problematic social media use, hours of social media use, internalized SGM stigma, perceived emotional social support, and depressive symptoms. Multiple linear regression examined the associations between the aforementioned variables and problematic social media use and was adjusted for gender identity.
A total of 302 SGM young adults were included in the analyses (assigned female at birth: 218/302, 72.2%; non-Hispanic White: 188/302, 62.3%; age: mean 21.9 years, SD 2.2 years). The sexual identity composition of the sample was 59.3% (179/302) bisexual and/or pansexual, 17.2% (52/302) gay, 16.9% (51/302) lesbian, and 6.6% (20/302) other. The gender identity composition of the sample was 61.3% (185/302) cisgender; 24.2% (73/302) genderqueer, fluid, nonbinary, or other; and 14.6% (44/302) transgender. Problematic social media use averaged 2.53 (SD 0.94) on a 5-point scale, with a median of 17 hours of social media use per week (approximately 2.5 h per day). Participants with greater problematic social media use had greater internalized SGM stigma (
Taken together, problematic social media use among SGM young adults was associated with negative psychological experiences, including internalized stigma, low social support, and depressive symptoms. SGM young adults experiencing minority stress may be at risk for problematic social media use.
Despite sweeping social change in many parts of the United States, many sexual and gender minority (SGM) youth and young adults face stigma, prejudice, and discrimination [
As a first step toward informing recommendations for health social media use for SGM young adults, it is important to identify those who may be at risk for problematic social media use (eg, preoccupation with social media use and difficulty with reducing or abstaining from use). In a sample of SGM young adults participating in clinical trials of 2 Facebook-delivered interventions for smoking cessation, this study examined the associations between problematic social media use and internalized SGM stigma, emotional social support, and depressive symptoms. We first review problematic social media use, including its theoretical underpinnings and correlates, among young adults. We then review the minority stress experienced by SGM young adults, social media use among SGM young adults, and how minority stress may influence the development of problematic social media use. Finally, we present potential correlates of problematic social media use in SGM young adults, informed by theoretical models of problematic social media use and minority stress.
Social media use is almost ubiquitous among young adults, with 91% of young adults in the United States reporting the use of at least one social media platform in a 2019 survey [
In the general young adult population, predisposing risk factors associated with vulnerability to problematic social media use include certain personality traits, mental health symptoms, and experiences. First, high extraversion [
SGM young adults face unique challenges and may have different predispositions than their non-SGM peers because of their experiences of minority stress. The minority stress model, as applied to sexual and gender minorities, posits that being a minority leads to chronic experiences of stigma, prejudice, and discrimination as well as more proximal stress processes, including vigilance toward future rejection, the concealment of one’s sexual or gender identity, and the internalization of negative societal attitudes toward SGM individuals (ie, internalized stigma) [
Social media may help undersupported individuals compensate for their lack of in-person emotional social support [
SGM young adults are vulnerable to internalized stigma, low emotional social support, and depressive symptoms as part of minority stress. The I-PACE model suggests that these predispositions may lead to problematic social media use in response to stress. Although research has examined predisposing factors for problematic social media use in the general young adult population, the effects and processes of minority stress as potential predispositions for problematic social media use in SGM young adults have not been explored.
This study is largely exploratory, testing the associations between problematic social media use and individual differences (informed by the I-PACE and minority stress models) in a convenience sample of SGM young adults enrolled in clinical trials of 2 smoking cessation interventions delivered on Facebook. Nonetheless, the extant literature and the minority stress model provided a framework for testing relevant factors that may predispose SGM young adults to problematic social media use. Specifically, we predicted that problematic social media use would be associated with greater internalized SGM stigma, lesser emotional social support, and greater depressive symptoms. Problematic social media use was expected to be positively associated, but not synonymous, with the time spent on social media [
Data were taken from the baseline assessments of 2 randomized controlled trials that tested the efficacy of SGM-tailored Facebook smoking cessation interventions compared with similar, nontailored interventions. Participants (N=302) were aged between 18 and 25 years, identifying as SGM (ie, nonheterosexual and/or noncisgender), having smoked at least 100 cigarettes in their lifetime, using Facebook for at least 4 days per week, living in the United States, and were well-versed in English. Participants in both trials were recruited between April and December 2018 with an SGM-tailored Facebook advertising campaign [
Problematic social media use was measured using the Bergen Social Media Addiction Scale [
Participants answered, “Approximately how many hours per week do you spend using social media?” The item was a free response; the natural log of responses was used in analyses.
A total of 9 items assessed the frequency of use of 8 social media platforms (Facebook, Instagram, Snapchat, Twitter, Pinterest, LinkedIn, Reddit, and Tumblr), plus a write-in “other” option, using 1-5 Likert-type scales (1=never; 2=monthly; 3=weekly; 4=once a day; 5=multiple times a day). The possible scores for each platform ranged from 1-5.
To measure internalized SGM stigma, participants completed the Revised Internalized Homophobia Scale, a 5-item measure adapted for all SGM identities [
The 8-item Emotional Support subscale of the National Institutes of Health (NIH) Toolbox for the assessment of neurological and behavioral functions measured the perceived emotional social support [
Depressive symptoms were measured using the 2-item Patient Health Questionnaire, a validated screener assessing the frequency of depressed mood and anhedonia in the past 2 weeks, with possible scores ranging from 0-6 [
The participants reported their age and sex assigned at birth (male or female). Participants selected all applicable terms to describe their gender identity (male/man, female/woman, trans male/trans man, trans female/trans woman, genderqueer/gender nonconforming, genderfluid, agender, nonbinary, or different identity), sexual identity (straight [heterosexual], lesbian/gay [homosexual], bisexual, pansexual, or not listed), and race and ethnicity (American Indian/Alaska Native, Asian, Black, Hispanic, Pacific Islander/Native Hawaiian, White, or other). Household income was presented in 7 ranges from “less than US $20,000” to “over US $200,000.” Subjective social status was measured on a 1-10 scale from “worst off” to “best off” compared with others in the United States and in their communities [
The analyses were conducted in 3 steps. First, we examined the distributions and reliability of the measures before creating composite scores. Second, we examined bivariate Pearson correlations between problematic social media use, social media use in hours per week, internalized SGM stigma, social support, and depressive symptoms. Third, significant correlates of problematic social media use identified in bivariate screening were entered into a multiple linear regression analysis as independent variables, with problematic social media use as the outcome dependent variable. As gender identity was significantly associated with problematic social media use, we adjusted for gender identity by entering it in step 1. The main correlates of interest were entered as independent variables in step 2. In the final model, gender identity (control variable) was entered in step 1; time spent on social media, internalized SGM stigma, emotional social support, and depressive symptoms (independent variables) were entered in step 2, and problematic social media use was entered as the dependent variable.
The sample characteristics are listed in
Participant characteristics (N=302).
Characteristics | Value | |
Age (years), mean (SD) | 21.9 (2.2) | |
Sex assigned at birth (female), n% | 218 (72.2) | |
|
||
|
Cisgender | 185 (61.3) |
|
Transgender | 44 (14.6) |
|
Genderqueer, fluid, and/or nonbinary | 73 (24.2) |
|
||
|
Gay | 52 (17.2) |
|
Lesbian | 51 (16.9) |
|
Bisexual/pansexual | 179 (59.3) |
|
Other | 20 (6.6) |
|
||
|
Non-Hispanic White | 188 (62.3) |
|
Native American/Alaskan Native | 3 (1) |
|
African American/Black | 3 (1) |
|
Asian | 10 (3.3) |
|
Hispanic/Latinx | 29 (9.6) |
|
Pacific Islander/Hawaiian Native | 2 (0.7) |
|
More than one race and ethnicity | 59 (19.5) |
|
Other | 8 (2.6) |
|
||
|
Less than US $20,000 | 75 (24.8) |
|
US $21,000-US $60,000 | 155 (51.3) |
|
US $61,000-US $100,000 | 52 (17.2) |
|
More than US $100,000 | 20 (6.6) |
|
||
|
Compared with others in the United States | 3.9 (1.6) |
|
Compared with others in your community | 4.3 (2.0) |
Years of education, mean (SD) | 13.5 (2.1) |
aParticipants selected all applicable terms from straight (heterosexual), lesbian/gay (homosexual), bisexual, pansexual, and not listed (please specify). Sexual identity was recoded into the categories presented here. When multiple identities were selected, “Gay/lesbian” took priority in coding, followed by “bisexual/pansexual” then “other” [
bParticipants selected all applicable terms from male/man, female/woman, trans male/trans man, trans female/trans woman, genderqueer/gender nonconforming, genderfluid, agender, nonbinary, or different identity (please specify). Gender identity was recoded into the categories presented here. When multiple identities were selected, “transgender” took priority, followed by “cisgender” then “other” [
The distributions of the key measures are presented in
Distributions of key measures.
Measure | Respondents, n (%) | Mean (SD) or median (IQR) | Observed range | Possible range | Scale | ||
Problematic social media use (score) | 302 (100) | 2.5 (0.9) | 1-5 | 1-5 | 1=very rarely; 2=rarely; 3=sometimes; 4=often; 5=very often | ||
Social media (hours per week) | 300 (99.3) | 17.0 (10-30) | 1-120 | ≥0 | Measured continuously | ||
Natural log (Ln) of hours per week | 300 (99.3) | 2.8 (0.7) | 0.7-4.8 | ≥0 | Measured continuously | ||
|
1=never; 2=monthly; 3=weekly; 4=once/day; 5=multiple times/day | ||||||
|
302 (100) | 4.7 (0.6) | 2-5 | 1-5 |
|
||
|
301 (99.7) | 4.0 (1.3) | 1-5 | 1-5 |
|
||
|
Snapchat | 300 (99.3) | 3.7 (1.4) | 1-5 | 1-5 |
|
|
|
297 (98.3) | 2.5 (1.6) | 1-5 | 1-5 |
|
||
|
292 (96.7) | 1.8 (1.1) | 1-5 | 1-5 |
|
||
|
293 (97) | 1.3 (0.7) | 1-4 | 1-5 |
|
||
|
292 (96.7) | 1.9 (1.3) | 1-5 | 1-5 |
|
||
|
Tumblr | 297 (98.3) | 2.6 (1.5) | 1-5 | 1-5 |
|
|
|
Other | 237 (78.5) | 1.2 (0.7) | 1-5 | 1-5 |
|
|
Internalized SGMa stigma (score) | 299 (99) | 1.7 (0.8) | 1-5 | 1-5 | 1=strongly disagree; 2=disagree; 3=neutral; 4=agree; 5=strongly agree | ||
Emotional social support (score) | 302 (100) | 30.7 (6.9) | 11-40 | 8-40 | 1=never; 2=rarely; 3=sometimes; 4=usually; 5=always (sum of 8 items) | ||
Depressive symptoms (score) | 302 (100) | 3.5 (1.8) | 0-6 | 0-6 | 0=not at all and 3=nearly every day (sum of 2 items) |
aSGM: sexual and gender minority.
On average, problematic social media use was around the midpoint of the 1-5 scale (mean 2.5, SD 0.9). Median use of social media was 17 hours per week (an average of approximately 2.5 h per day). The 3 most frequently used social media platforms were Facebook, Instagram, and Snapchat. Internalized SGM stigma scores, which can range from 1-5, averaged 1.7 (SD 0.8), which was comparable with previous research (mean 1.5, SD 0.7 [
Problematic social media use was significantly greater among participants who spent more time on social media (
Pearson correlations between key variables (N=302)a.
Measure | Problematic social media use | Social media hours per week (natural log [Ln]) | Internalized sexual and gender minority stigma | Emotional social support | Depressive symptoms | ||||||
|
|||||||||||
|
|
1 | 0.24b | 0.22b | −0.16c | 0.22b | |||||
|
—d | <.001 | <.001 | .007 | <.001 | ||||||
|
|||||||||||
|
|
0.24b | 1 | −0.05 | 0.02 | 0.15c | |||||
|
<.001 | — | .41 | .78 | .009 | ||||||
|
|||||||||||
|
|
0.22b | −0.05 | 1 | −0.23b | 0.13e | |||||
|
<.001 | .41 | — | <.001 | .02 | ||||||
|
|||||||||||
|
|
−0.16c | 0.02 | −0.23b | 1 | −0.28b | |||||
|
.007 | .78 | <.001 | — | <.001 | ||||||
|
|||||||||||
|
|
−0.22b | 0.15c | 0.13e | −0.28b | 1 | |||||
|
<.001 | .009 | .02 | <.001 | — |
an=300 for social media hours per week; n=299 for internalized sexual and gender minority stigma.
bSignificant at the
cSignificant at the
dNot applicable.
eSignificant at the
Time spent on social media (β=.24;
Adjusted multiple linear regression analysis with problematic social media use as the outcome variable.
Variable | β | |||||
|
||||||
|
|
|||||
|
|
Transgender (reference: cisgender) | .04 | .48 | ||
|
|
Nonbinary and/or other gender (reference: cisgender) | .18 | .002 | ||
|
||||||
|
|
|||||
|
|
Transgender (reference: cisgender) | .02 | .66 | ||
|
|
Nonbinary and/or other gender (reference: cisgender) | .15 | .009 | ||
|
Social media (hours per week; natural log [Ln]) | .24 | <.001 | |||
|
Internalized stigma | .20 | <.001 | |||
|
Emotional social support | −.06 | .28 | |||
|
Depressive symptoms | .11 | .05 |
In this study of SGM young adults enrolled in 2 Facebook smoking cessation intervention trials, those with greater problematic social media use had greater internalized SGM stigma and depressive symptoms and lower perceived emotional social support. Greater internalized SGM stigma remained significantly associated with greater problematic social media use after accounting for the time spent on social media and other correlates. In addition, participants with greater depressive symptoms had marginally greater problematic social media use. Overall, signs of problematic social media use were more likely to occur among SGM young adults who had internalized SGM stigma and depressive symptoms.
These results are consistent with our hypothesis that SGM young adults with greater internalized SGM stigma would be more likely to exhibit problematic social media use than those with lower internalized SGM stigma. According to the I-PACE model, personal characteristics can predispose individuals to problematic social media use. Internalized SGM stigma may be one such characteristic for SGM young adults. Individuals with poor self-concepts often seek social support on social media, and these attempts are often unsuccessful. Specifically, previous research found that adults with low self-esteem primarily posted negative thoughts [
SGM young adults with low emotional social support were more likely to have problematic social media use in bivariate analyses, thereby corroborating the associations between low social support and problematic social media use found in the general population [
Problematic social media use was associated with depressive symptoms in bivariate analyses, partially supporting our hypothesis. This is consistent with the I-PACE model and with extant research examining the link between depressive symptoms and problematic social media use in the general population [
Consistent with extant literature involving young adult social media users [
This pattern of results suggests multiple predisposing factors for problematic social media use among SGM young adults, especially internalized SGM stigma. Social media may appeal to young adults with poor self-concepts and insufficient emotional support, as social media presents opportunities to connect with others. However, social media may not be an adequate source of support and connection for SGM young adults with poor self-views. Our results suggest that SGM young adults who used social media intensively, to the point of problematic use, still had high internalized SGM stigma, high depressive symptoms, and low emotional support. Moreover, SGM young adults with problematic social media use may withdraw from offline social opportunities and thus receive less support. Importantly, the relationship between problematic social media use and social support was significant only in bivariate analyses, suggesting that internalized SGM stigma is more strongly associated with problematic use. Problematic social media use was also associated with depressive symptoms in bivariate analyses. Seeking social support and not receiving it may exacerbate depressive symptoms [
Conversely, spending more time on social media was not associated with internalized stigma or social support. SGM young adults who experience distress around their identities may be more vulnerable to problematic social media use, regardless of how frequently they use social media. Consistent with previous research, time spent on social media and problematic social media use were modestly correlated but distinct [
This study supports and extends prior research on SGM individuals’ social media use by demonstrating associations between problematic use and common challenges faced by SGM young adults. When used in moderation, social media offers benefits such as self-expression and connection with others [
First, this cross-sectional, correlational study design did not confirm causal inferences about the role of social media use in mental and physical health outcomes. Longitudinal data would enable the use of more sophisticated statistical modeling to understand the interplay among experiences of stigma, social support, and depressive symptoms. This study underscores the need for future research on the social media use of SGM young adults. Second, all participants were young adults who self-identified as SGM and smoked cigarettes. Other dimensions of sexual orientation (ie, sexual attraction and sexual behavior) were not assessed. SGM young adults who smoke may be a particularly vulnerable subset of SGM young adults, and it is unclear whether the results can be generalized to SGM individuals who do not smoke. Future research should aim to oversample non-White young adults and those assigned male at birth, as the sample in this study was mainly White (188/302, 62.3%) and assigned female at birth (218/302, 72.2%). Third, all participants were frequent Facebook users. Future research could aim to recruit SGM young adults from other social media platforms. Finally, this study was a secondary analysis that captured only a few potential correlates of problematic social media use. Measures of some constructs (eg, depressive symptoms and time spent on social media) were brief mental health symptoms aside from depression (eg, anxiety symptoms) were not measured, and “social media” was not defined. Future research could use more comprehensive measures to further our understanding of problematic social media use among SGM young adults.
Although social media use can be beneficial for SGM young adults [
fear of missing out
Interaction of Person-Affect-Cognition-Execution
lesbian, gay, bisexual, transgender, queer+
National Institutes of Health
sexual and gender minority
This work was supported by the National Institute on Minority Health and Health Disparities of the NIH (grant R21 MD011765), the National Institute on Drug Abuse of the NIH (grant K01 DA046697), and the Tobacco-Related Disease Research Program (grants 26IR-0004 and 28FT-0015). Sponsors had no role in the study design; collection, analysis, or interpretation of data; writing of the report; or decision to submit the paper for publication. The authors have no relevant financial interests to disclose.
EAV was involved in the conceptualization, formal analysis, investigation, and writing of the original draft. Investigation; writing, reviewing, and editing; supervision; and funding acquisition was conducted by DER. JJP and MCM were involved in conceptualization; investigation; and writing, reviewing, and editing. Conceptualization; investigation; data curation; and writing, reviewing, and editing was done by JFL. GLH was involved in conceptualization; investigation; writing, reviewing, and editing; and supervision.
None declared.