Post-secondary student adoption of mental health labels through their use of social media: a scoping review

Abstract

Background

Social media is a significant source of information for post-secondary students, who are usually at the age at which many common mental disorders first express themselves. Social media can have a role in the way post-secondary students identify and act on mental health issues.

Objectives

Explore how the use of social media influences post-secondary students’ adoption of mental health labels.

Eligibility criteria

We included empirical studies on mental health labelling in the context of social media use among post-secondary students published in English between January 1995 and April 2025.

Sources of evidence

The review includes references from five databases: Scopus, PubMed, Ovid MEDLINE (to access APA PsycINFO), Web of Science and ProQuest Global Dissertations and Theses. Based on the included studies from the initial search, we built a complementary search strategy using Research Rabbit artificial intelligence.

Charting methods

We present a table listing characteristics of the studies and brief summaries of their findings. A narrative synthesis compiled the information from each study to answer the research questions.

Results

The search identified 7551 references and 1099 additional records from Research Rabbit. 11 studies published since 2011 met the inclusion criteria with qualitative, mixed methods and quantitative designs, without major quality concerns. Approaches to measuring social media exposure varied, including platform reports of user activity and self-reported indicators. Individuals adopted labels themselves or received labels from peers or researchers. Most research focused on self-presentation and symptom disclosure rather than labelling itself. The accuracy of self-diagnosis was higher for common disorders and lower for complex conditions such as mania or panic disorders. Labelling varied across social media platforms. Online interaction revealed issues that students were reluctant to share face-to-face. Label use appeared to influence help-seeking and peer support, with effects shaped by social stigma.

Conclusions

The adoption of mental health labels via social media among post-secondary students remains largely unexplored. The concept of labelling and its operationalisation vary across research. Future studies should provide more formal definitions, investigate mechanisms driving labelling and assess its potential effects on human health.

STRENGTHS AND LIMITATIONS OF THIS STUDY

• Artificial intelligence (AI) tools and a range of databases, including various types of publications, increased the number of considered papers.

• Our search strategy included topics related to mental health labelling to compensate for its lack of precise definition.

• The heterogeneity of the literature limits quantification of the effect of labelling on health outcomes, although it does allow general theories on mechanisms of action.

Introduction

Mental health labels result from assigning patients to categories that discriminate and identify mental health disorders.¹ Labelling can be a formal diagnostic process, an informality by peers or a self-driven process. Labelling by others can follow from self-presentation or disclosure, and it is often linked to self-image and stigma. Labelling patients can impact one’s sense of self, the resources we access, and our ability to build relationships and social support networks.^{2 3}

Sociologist Thomas Scheff argued that when society labels someone as ‘deviant’ (eg, mentally ill), it alters the way they are treated and pressures the individual to behave in accordance with the label, further perpetuating their condition in a loop.⁴ While there can be consequences for self-esteem and social networks, labelled individuals can mitigate these effects through solidarity with other labelled individuals or by withdrawing from society.⁵ Labels can enhance access and care from health services, and they also provide social currency⁶ or leverage among peers. Labelling might also generate stigma, an intelligible attribute that can be deeply discrediting when revealed to the general public and yet might provide solidarity and sense of community among other stigmatised persons.^{7 8}

The meaning and consequences of mental health labels depend on the context in which they are displayed.⁹ According to Ian Hacking, our behaviour and cognition change in response to psychological classification through a series of loops.¹⁰ Serial socialisation can lead to an evolution of labels. If one’s lived experience feels congruent with the label, in contrast, its effect tends to stabilise.⁷ While stabilising loops allow one to maintain and even reinforce a diagnosis, destabilising loops cause individuals to seek out new labels as the old ones no longer apply.¹¹ Looping effects are a possible consequence of adopting labels from social media.

Post-secondary students are of special interest due to the rising global prevalence of mental health problems in this population.¹² Many experience mental health disorders for the first time as post-secondary students, and they are also heavy users of social media.¹³ Post-secondary education overlaps with emerging adulthood, a developmental stage between 18 and 29 years that is marked by social and financial instability.¹⁴ Emerging adults have a greater capacity for independent living than adolescents do, but they experience more uncertainty than young adults.¹⁵

In European universities, 88% of students reported using the internet several times per day and no less than 50% used the internet to search for mental health information.¹⁶ A literature review of student use of digital mental health resources in the USA found one-third reported searching for mental health information online at least once.¹⁷ The internet is attractive for this purpose as it provides anonymity, asynchrony, new ways of communicating (status updates, microblogging) and the formation of online communities around shared interests, including mental health concerns.18,20 Algorithms play a strong role in creating and reinforcing online mental health silos, as they determine what users see on social media by combining their interests and engagement patterns.

Some researchers suggest the spread of medical knowledge to the broader population leads to decreased well-being and even iatrogenesis (unintended harm from medication or diagnostic errors).^{21 22} We know very little about how social media labelling involves Hacking’s looping dynamics. While some users benefit from online health information, others might develop cyberchondria (escalating health anxiety from digital health information) or engage with communities that normalise psychiatric conditions like anorexia.²³ Concerns also arise over the cultural overinterpretation of mental illness, potentially encouraging its adoption as identity.^{24 25} There is also compelling evidence, however, that social media users can use mental health information appropriately.^{9 26}

Rationale

We found no published systematic review discussing looping effects and labelling related to social media use. Research to date has focused on those with formal diagnoses and how to encourage treatment seeking through social media,²⁷ and ways that health professionals can deliver conversational interventions through TikTok.²⁸ Literature reviews have covered the ability of artificial intelligence (AI) to detect depression and label users,^{29 30} and avoiding ‘judgement or labelling’ when seeking mental health help online.³¹

This literature review contributes to understanding informal labelling processes among post-secondary students. Its objective is to examine how social media use influences students’ use of psychiatric language to categorise themselves and others using mental health labels and the consequences of this type of labelling. We also highlight how researchers conceptualise and operationalise mental health labels in their work on social media usage. A greater understanding of how these delicate medical, digital and social ecosystems interact with each other can inform new research strategies and interventions.

Methods

This review follows a published protocol,³² and we report the study according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Scoping Reviews (PRISMA-ScR).³³ The overall research question leading this literature review was: How do social media platforms influence post-secondary students’ adoption of mental health labels?

Specific questions included:

• How do researchers operationalise mental health labelling?

• How do researchers measure social media exposure on social media platforms?

• How do university students, researchers and mental health professionals use mental health labels?

• What are the reported effects of using mental health labels?

• What mechanisms have been reported to lead to the acquisition of mental health labels from social media?

Information sources

We searched five databases: Scopus, PubMed, Ovid MEDLINE (APA PsycINFO 1967–2023), Web of Science and ProQuest Global Dissertations and Theses. The initial search for each of these databases was conducted in June 2023 and was updated in April 2025. An example search strategy is attached as online supplemental appendix 1. We also used Research Rabbit, an AI tool, to identify references that the initial database searches might have missed. Research Rabbit generates a list of papers similar to a manually entered base list by analysing related references and topics.³⁴ For this search, we created a base list from papers selected after full-text review of references in the five databases (online supplemental appendix 2). We conducted the Research Rabbit search in September 2023. Each identified reference received a tag and went to Covidence to allow comparison with references identified by other methods. The process for selecting references from AI was identical to that used for those retrieved from the databases.

Eligibility criteria

Eligibility criteria and the search strategy were developed with a specialised librarian at McGill University. The search was limited to English-language articles published between 1995 and 2025. The wide time frame covers the entire period following the emergence of social media in the late 1990s and its rapid growth during the early 2000s.³⁵ Our aim was to capture possible changes in labelling patterns during the variations of social media landscape. We included references reporting empirical studies on mental health labelling in the context of social media. Our population of interest was post-secondary students, including those attending universities, colleges or other institutes with the goal of attaining an academic degree. Studies that did not present disaggregated data on students enrolled in post-secondary institutions, mental health labelling and social media exposure were excluded.

Selection

Covidence literature review management software facilitated the initial screening based on title and abstract. The package identified an initial list of duplicated references. The lead author (EDA) and a second reviewer (AY or VC) screened all references; an independent reviewer (AY, VC or IS) resolved any conflicts.

Data extraction

Two reviewers extracted data about the included studies and relevant content to answer the research questions. Online supplemental appendix 3 shows the items used during the data extraction. They solved disagreements by consensus. A third reviewer was brought in to resolve conflicts.

Critical appraisal

We used the Mixed Methods Appraisal Tool (MMAT).³⁶ It allowed assessing the quality of included studies based on their type (qualitative/quantitative-randomised, quantitative-non-randomised, quantitative descriptive and mixed methods). We did not exclude studies based on quality.

Synthesis of results

Following PRISMA-ScR guidelines, the synthesis includes both a summary table and a narrative integration of findings. The table presents key information from each study, including citation details, study design, participant characteristics, social media platform, mental health measures and main findings. The narrative synthesis draws on the findings of included studies to answer the research questions by identifying recurring themes, differences between populations and platforms, and proposed labelling mechanisms.

Results

Characteristics of sources of evidence

Searching five academic databases returned 7551 papers after removing 5688 duplicates. Research Rabbit generated a list of 1165 papers, of which 66 were duplicates. A total of 128 references were assessed for eligibility based on a full-text review. A PRISMA flowchart maps the selection process (figure 1). Eleven studies met the inclusion criteria, one of which came from Research Rabbit. We describe their general characteristics in table 1 (see online supplemental appendix 4 for more details).

Figure 1. Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Scoping Reviews (PRISMA-ScR) flow diagram.

Table 1. General characteristics of included studies (see online supplemental appendix 4 for more details).

Authors (publication date)	Title	Setting	Study design	Sample size	Type of labelling
Moreno et al45 (2011)	Feeling bad on Facebook: depression disclosures by college students on a social networking site	USA College	Observational, cross-sectional content analysis study	200	Formal
Moreno et al39 (2012)	A Pilot Evaluation of Associations Between Displayed Depression References on Facebook and Self-reported Depression Using a Clinical Scale	USA College	Quantitative, cross-sectional, self-report survey study	224	Formal
Michikyan41 (2020)	Depression symptoms and negative online disclosure among young adults in college: a mixed-methods approach	USA College	Mixed methods, cross-sectional, self-report study	261	Formal
Kim and Hong40 (2021)	Understanding University Students’ Experiences, Perceptions, and Attitudes Toward Peers Displaying Mental Health–Related Problems on Social Networking Sites: Online Survey and Interview Study	South Korea University	Mixed methods, two-phase study, survey followed by semistructured interviews	226	Peer
Zengaro37 (2021)	“The World Ain’t All Sunshine and Rainbows”: Enacting the Athlete Identity Narrative in Stigmatising Mental Health Communication	USA College	Qualitative, cross-sectional, interviews	30	Informal
Rutter et al34 (2023)	“I Haven’t Been Diagnosed, but I Should Be”—Insight Into Self-diagnoses of Common Mental Health Disorders: Cross-sectional Study	USA University	Quantitative, cross-sectional, self-report study	2337	Self-diagnosis
Ye et al46 (2023)	Relationship between university students’ emotional expression on tweets and subjective well-being: Considering the effects of their self-presentation and online communication skills	Japan University	Quantitative, cross-sectional with retrospective content analysis	577	Formal
Mohamad et al44 (2024)	The Role of Instagram in Promoting Mental Health Awareness and Help-Seeking Attitudes Among Malaysian University Students	Malaysia University	Quantitative, cross-sectional, correlation study	364	Formal
Zhang38 (2024)	Digital technology and mental health: Chinese university students’ perspectives on the impact of social media	China University	Qualitative, cross-sectional, interviews	16	Informal
Sump et al42 (2025)	Self-diagnosis and help-seeking behaviours: Impact of self-diagnosis in seeking counselling	USA Community college	Quantitative, cross-sectional, self-report study	363	Self-diagnosis
Ye and Ho43 (2025)	University Students’ Subjective Well-Being in Japan Between 2021 and 2023: Its Relationship with Social Media Use	Japan University	Quantitative, longitudinal	1681 (year 1) 1292 (year 2) 851 (year 3)	Formal

Most studies (6/11) were in the USA, two papers in Japan and one each in China, Malaysia and South Korea. The studies were published between 2011 and 2025. Two used an exclusively qualitative approach,^{37 38} two used mixed methods^{31 32} and the rest used quantitative methods. The most common quantitative design in the studies was cross-sectional.3439,44 Mixed methods designs used surveys and semistructured interviews.^{40 41} Three studies accessed social media accounts directly.^{39 45 46} Most participants across all studies were undergraduate students. For more information on the studies included, see table 1.

Critical appraisal within sources of evidence

While quality was not considered during the selection process, all included studies met the requirements of the Mixed Methods Appraisal Tool 2018³⁶ for the respective research designs used. For all studies, the research questions were clear, and the study designs were justified as appropriate to address the research questions (online supplemental appendix 5)

How do social media platforms influence post-secondary students’ adoption of mental health labels?

Included studies found social media was an important outlet for students’ disclosure of mental health issues. Most of the research emphasises the act of disclosing symptoms rather than the labelling process itself.^{39 41 43 45 46} Social media disclosures have been correlated with formal diagnostic outcomes. Students who display symptoms of depression on platforms like Facebook might meet formal diagnostic criteria reflected in higher Patient Health Questionnaire 9 (PHQ-9) scores.³⁰ Approximately 25% of college students disclose comments that meet criteria for depression symptoms.⁴⁵ Students with higher depression symptoms are more likely to disclose various life hassles online.⁴¹

Student-athletes choose social media to disclose mental health issues because they lack confidence in face-to-face communication. This disclosure can be direct or through more subliminal messages and attempts to avoid stigma from coaches/teammates.³⁷ A qualitative study from China found that students navigate mental health language on social media based on their cultural identity while balancing mental health awareness and stigma.³⁸

In the table (table 1), we separated studies by the types of labelling they focused on. We defined formal labelling as labelling based on diagnostic criteria or assessment, while informal labelling was done without any reported diagnostic comparison. Peer labelling is based on students’ observations of their colleagues’ social media mood disclosures and inconsistent offline and online behaviours, which can be subtle and recognised only by those who know the poster and the context. These inconsistencies might lead peers to believe their friends have mental health problems.⁴⁰ In many cases, the labelling process was influenced by participants’ existing experiences with their peers’ offline identity.⁴⁰

The choice of social media platform influences the likelihood of disclosing mental health symptoms and being labelled with a mental health condition. For example, Twitter/X users might be more readily labelled as depressed compared with Instagram users.⁴⁶ This could be due to different microcultures on each platform or to different affordances, such as only sharing text or images.¹⁹

A follow-up study on social media usage patterns in Japan showed a link between disclosing various aspects of one’s identity and positive subjective well-being.⁴³ Instagram use for information seeking and emotional support was found to positively influence mental health awareness and increase help-seeking attitudes in Malaysia.⁴⁴ In the USA, however, a study of self-diagnosing on social media was found to have no impact on willingness to seek professional help.⁴²

How do researchers operationalise mental health labelling?

Two approaches involved participants self-reporting whether they had been diagnosed with certain disorders or whether they believed they should be.^{34 42} Other approaches relied more on the interpretation of researchers to categorise social media users based on social media posts. Researchers used the DSM-IV criteria for major depressive episodes, which includes symptoms like depressed mood, loss of interest, appetite changes, sleep problems and more.³⁹ Posts were considered references to depression if they matched these criteria through keywords or synonyms.⁴⁵ Researchers analysed social media posts reflecting negative emotional states using thematic analysis to highlight different kinds of negative life experiences disclosed by participants.⁴¹ Researchers also measured self-reported tendencies to depression using a validated Japanese scale.^{43 47} In one study, participants selected images and texts from their own post history that best described themselves and explained why these posts were important to them.⁴¹ In studies involving athletes, researchers used the communication theory of identity, which suggests that identities are expressed through core symbols, meanings and labels. None of the studies explicitly distinguished between “mental health label” and formal diagnosis of a mental illness.

How do researchers assess exposure on social media platforms?

In the study of peer labelling, participants reported whether they had witnessed someone signalling mental health problems on social networking sites.⁴⁰ A more formal approach involved examining the ‘information section’ of participants’ profiles to gather demographic information and descriptive variables about Facebook use, such as time since the last Facebook activity.³⁹ This included the number of Facebook friends and the number of days since the last Facebook activity, with fewer days indicating more frequent use.⁴⁵ In the Facebook Use Questionnaire, participants reported their daily Facebook use, including average time spent, number of logins, status updates, wall posts and perceived activeness. The researchers converted these activities to scores which they summed to create a Facebook Activities variable.⁴¹ Two studies measured mental health information seeking behaviour on social media in relation to help-seeking.^{42 44}

Two studies identified patterns of platform-specific use.^{43 46} Participants were considered users of a particular platform if they spent at least 20% of their social media time on it. The study observed nine patterns of use among 15 possible combinations, such as LINE (a social networking site popular in Japan) only, Twitter/X only, Instagram only and various combinations of these platforms.⁴⁶ A follow-up study added TikTok and Discord.⁴³

A study measured sports media use with eight items adapted from the Motivation Scale for Sport Online Consumption. This scale was developed to study how sports organisations engage with fans through websites.⁴⁸ Participants respond to questions about their use of team social media sites (eg, Facebook, Instagram, Twitter/X, Snapchat) on a 7-point Likert scale (1=strongly disagree; 7=strongly agree).³⁷ Sample questions include reasons for using the team’s social media site, such as escaping from reality or learning how to get along with others. Only usage of team social media sites was recorded, not personal accounts.

How do university students, researchers and mental health professionals use mental health labels?

University students who recognised signals of vulnerability from peers deliberately intervened in the problematic situations.⁴⁰ It was not always easy for them to take action in response, and sometimes they opted to remain silent. Peer interaction on social media might also encourage disclosure. On Facebook, for example, students who received responses to their depression-related posts were more likely to continue discussing their depressive symptoms publicly.⁴⁵ Students who disclosed positive and negative aspects of their identity had increased subjective well-being across usage patterns and gender.⁴³ Students experiencing depressive symptoms might use social networking sites as a safe and indirect outlet for their emotions. This online disclosure can be a call for aid, with those reporting higher depression symptoms more likely to share their life hassles and negative emotions online.⁴¹

Cultural identity also plays a strong role in whether mental health disclosure leads to increased awareness or increased stigma.³⁸ Student-athletes used social media as a ‘step towards communicating mental illness’ when they did not feel confident in their resources or feared disclosing issues to coaches and teammates.³⁷ Professional athletes, who used their platforms to raise awareness about mental health, helped student-athletes overcome barriers to communicating their mental health problems.³⁷ Professionals could reduce the stigma associated with mental illness and show that it is possible to maintain an athletic identity without being labelled as weak by sharing their own experiences.

Self-diagnosis can accurately identify disorders like major depression³⁹ and generalised anxiety disorder.³⁴ These disorders are common, and individuals can often detect significant symptoms without formal training. However, self-diagnosis accuracy was lower for less common disorders like mania and panic disorders³⁴ and self-diagnosis was not associated with increased help-seeking behaviour.⁴²

What are the reported effects of using mental health labels?

Peer labelling led to internal discussions on how individuals respond to peers in distress and whether to intervene or to remain silent and avoid involvement.⁴⁰ In some cultures, stigma played a strong role in students’ decisions to reach out for help.^{38 44} Athletes tended to avoid being labelled as ‘sick’ due to the perceived expectations from coaches and teammates regarding their toughness and strength. Fears of labelling were associated with fears of losing their positions in the team (being benched).³⁷

What mechanisms have been reported to lead to the acquisition of mental health labels from social media?

The studies identified focused primarily on the reasons behind disclosure, which differ from the mechanisms used to adopt a label. Studies also focused on the accuracy³⁴ and consequences⁴² of self-diagnosis without investigating the mechanism of self-diagnosis itself. Disclosing symptoms on social media was the primary mechanism for label assignment by peers or researchers.^{34 39 41 43 45} Young adults might use terms like depression more commonly now, even without embodying the true diagnostic meaning, due to increased recognition of these conditions.^{20 39} There was an association between patterns of social media use and emotional expressions that affect subjective well-being, although it was unclear if these patterns influence the perception of well-being or vice versa.^{43 46} Among athletes, the adoption and disclosure of labels were influenced by fears of stigma and perceptions of others’ expectations regarding their image.³⁷ Overall, research has yet to explore the mechanisms for label adoption.

Discussion

This review explored how use of social media influences the adoption of mental health labels among post-secondary students. According to the included literature, the field remains underdeveloped, with an emphasis on symptom disclosure and inconsistent operational definitions of labelling. Available studies have not examined the mechanisms underlying label adoption or quantified its impact on health outcomes. Studies showed that post-secondary students share personal information on social media that feeds into the development of mental health labels as viewed by themselves, other students or researchers. The tendency to share this information and the content has been found to vary across platforms.⁴⁶ This suggests that different expectations across social media platforms inform how users share and respond to mental health information. However, the available information does not allow a formal definition of the effects that different aspects of social media have on label adoption. The confusion of mental health labels with general diagnostics or self-disclosure reduced interpretability, making it difficult to attribute the role of social media in the adoption of labels.

The labels and the associated lexicon users learn online can correspond to their formal diagnosis, for example, through assessments like the PHQ-9,³⁴ but not invariably so (eg, in cases of self-diagnosing bipolar disorder³⁴). Labels can also be self-identified or come from peers or researchers. Future research could look at labelling typologies such as formal, applied as part of sanctioned clinical or institutional processes, or informal, which emerge through day-to-day social interactions.⁴⁹ Other typologies can distinguish between labels with positive societal or medical views and those negatively sanctioned (pejorative).⁵⁰

The full consequences of self-diagnosis and mental health self-labelling are poorly understood regarding mental health progression or the potential outcomes they can generate. Labels might or might not encourage help-seeking.^{42 44} While social media offers the promising benefits of tailored and accessible mental healthcare for students, inconsistent use of mental health language can affect their formal mental health assessment in unpredictable ways.

A promising sign for this research is a growing interest in the topic, demonstrated by early career researchers in the form of theses and dissertations.2537 51,53 Future work might apply findings from offline studies of mental health labelling in post-secondary students to empirical studies of social media. For example, one ecological study concluded students might use mental health labels to communicate distress and access support from professors and administrators, in addition to their use for gaining social capital.⁶ Other studies show students attribute nuanced and shifting meaning to labelling mental health experiences.^{54 55} An empirical study of help-seeking behaviour found that using more accurate mental health labels predicted an increase in health-seeking behaviours in post-secondary students compared with lay terms for mental distress.⁵⁶

We could find no published review exploring this topic and the evidence we can report on remains general. Yet the field is of considerable importance to mental health, especially for people immersed in the world of social media. Our review contributes elements of a psychology proto-theory⁵⁷ that could be relevant for family medicine, psychological counselling and therapy, and for social media users themselves. Additional studies could establish the foundation for a labelling theory of social media use by focusing on correlations between specific features of a given social media platform and mental health labelling processes to support evolution of evidence-based online mental health services and accurate sharing of mental health information.

Strengths and limitations

There is no accepted definition of mental health labels in the literature, so our search strategy had to rely on related topics. The breadth of the search increased the volume of references for screening; it spanned a wide time interval to reflect changes in the social media landscape. Inclusion of theses and dissertations provided useful preliminary insights into this developing field. AI tools contributed only one reference to the final list of included studies and complicated the selection instead of streamlining it. Our results help to formalise an early understanding of mental health labels and the internet; they also identify gaps for future research.

Conclusions

The adoption of mental health labels via social media among post-secondary students remains largely unexplored. Our findings identified only a small number of studies that address this topic. Because the concept of labelling and its operationalisation lack clarity and vary across research, future studies should offer more formal definitions and investigate the mechanisms driving labelling, as well as quantify its frequency and potential effects on human health.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.