Social Media and Mental Health: Lessons Learned from the Psychology Research and Behavior Management Article Collection

Abstract

The article collection on social media and mental health attracted the interest of many researchers and resulted in 25 articles published in the collection. In this editorial, the guest advisors for the collection summarize the included studies and some of the most relevant findings from them. Five of the articles are given particular attention, representing both cross-sectional and longitudinal study designs. The article collection brings new and important insights into how mental health is shaped, and how mental health shapes behaviors, in the modern world of social media. It highlights mediational pathways from social media use to mental health problems through cyberbullying, social comparison, and cognitive overload, and from mental health problems to problematic social media use through self-referential processing. It is the guest advisors’ hope that researchers can use the collection, and indeed this editorial providing a synopsis and commentary to the collection, as a point of reference when choosing new research questions to explore and when deciding on certain aspects of design and methodology.

Introduction

The interest in studying social media in relation to mental health has grown markedly over the last years. This is not surprising, considering how social media has become a pervasive element of our daily lives and how social media is designed to occupy our time as much as possible. Adolescents and young adults spend considerable amounts of time using social media, and while several researchers have been concerned about the possibility that social media use may have a negative impact on their mental health, others have highlighted positive impacts, such as social support and community with others. However, answers to questions about impact are not straightforward and depend, to name some factors, on employed research designs and measurement choices, but also on the content of interactions on social media and the conceptualization of both concepts. For example, one longstanding debate in the field concerns the interpretation of statistical associations in the absence of longitudinal or experimental data. Thus, social media should not solely be considered a potential cause of mental health problems or improvements; researchers also argue that mental health factors indeed shape our behaviors on social media.

With an array of research approaches, the possible relationships between social media use and mental health are explored in 25 articles included in this article collection, indicative of the great interest in moving this research field forward. The current editorial provides an overview of the studies included and addresses some of the lessons learned from them. Table 1, also demonstrating the analytical lens applied to the articles, presents the studies’ characteristics, including their main results.

Table 1.

Overview of the Studies

Reference	Country	Time of Data Collection	Aim	Population	Sample Size	Study Design	Outcomes/Mediators	Main Results
Al-Abyadh et al (2024)1	Egypt and Saudi Arabia	2023	To determine how smartphone addiction and self-regulation failures affect students’ academic life satisfaction, taking into account the mediating role of mind wandering and cognitive failures.	Undergraduate students	950	Cross-sectional survey	Academic Life Satisfaction; Self-Regulation Failures; Mind Wandering; Cognitive Failures	Smartphone addiction and self-regulation failures negatively affected students’ academic life satisfaction, and positively affected students’ mind wandering and cognitive failures. Mind wandering mediated the relationship between smartphone addiction, self-regulation failure, and academic life satisfaction.
Alqarni et al (2024)2	Saudi Arabia	2021–2022	To estimate the prevalence of Problematic Internet Use (PIU) in medical students and explore their correlation with the medical students’ PTSD and Patient Health Questionnaire −9 (PHQ9) scores.	Medical students	505	Cross-sectional survey	Posttraumatic Stress Disorder (PTSD); Patient Health Questionnaire (PHQ-9)	The prevalence of problematic internet usage (PIU) was 30.6% in the sample. Moderate-to-moderately severe PTSD were 1.7 times more likely to have PIU. Students with moderate to severe depression were 2.2 times more likely to have PIU.
Babiker et al (2023)3	International	2022–2023	Examine the association between the need for affect and problematic social media use (PSMU), as well as the mediating role of fear of missing out in that relation.	Not specified, open recruitment through an online chat software	513; 262 from European countries and 251 from Arab countries	Cross-sectional survey	9-item Social Media Disorder Scale; a single-item FOMO scale	Affect approach and affect avoidance predicted PSMU in both samples. FOMO partially mediated the relationships between need for affect and PSMU.
Barragan et al (2024)4	USA	Not stated	The present research investigated the extent to which experiences of cyberbullying victimization mediate the link between greater social media use and poorer mental health in adults and whether such indirect effects are moderated by gender or age.	Adults, at least 18 years old.	502	Cross-sectional survey	The 7-item Depression Scale; 21-item Beck Anxiety Inventory (BAI) Mediator: Cyberbullying victimization	Robust indirect effect of cyberbullying victimization on the relation between degree of SM use and MH: greater SM use was associated with higher levels of cyberbullying victimization, and greater cyberbullying victimization was associated with increased depression, anxiety, and likelihood of substance use. Age moderated the mediation effects for anxiety and likelihood of substance use, with stronger mediation effects emerging for younger compared to older adults.
Du et al (2024)5	International	2018	Examine how internet use behaviors impact on adolescents’ mental health, and whether self-education expectations and parental support mediate this relationship	Students in grades 7–13	336,000	Cross-sectional survey (PISA)	Mental health scale composed of items assessing frequency of occurrence of positive and negative emotions	More internet use was related to better mental health, and self-education expectations and parental support mediated the relationship
Gong et al (2024)6	China	2022	Explore the relationship between adolescents’ problematic Internet use and loneliness and the mediating roles of social support and family communication.	12–17 years	2483	Cross-sectional survey	Loneliness. Social support Family communication	Positive effect of adolescents’ problematic Internet use on loneliness. Mediating effects of perceptual social support and family communication were found to play a chain effect between adolescents’ problematic Internet use and loneliness, respectively. Use and feelings of loneliness played a chain mediating role
Hu et al (2024)7	China	Not stated	To analyze the relationship between life satisfaction and learning burnout, and explore the mediating effect of mobile phone addiction and psychological capital	Chinese Medical Postgraduate Students:	628	Cross-sectional survey	Learning Burnout Questionnaire (LBQ); Mobile Phone Addiction; Psychological Capital	Life satisfaction and psychological capital had a negative predictive effect on learning burnout, and mobile phone addiction had a positive predictive effect on learning burnout
Jameel et al (2024)8	Saudi Arabia	2023	Examine how television advertisements and social media sites influence compulsive shopping behavior among university students, with a possible mediating impact of materialism.	University students	487	Cross-sectional survey	5-item measure to assess compulsive shopping behavior	Higher social media use was positively related to more compulsive shopping behavior. Materialism fully mediated the link between social media use and compulsive shopping behavior.
Li et al (2024)9	China	Not stated	Examine how mobile phone addiction is influenced by perceived discrimination, mediated by negative emotions and learning burnout	Vocational college students, age range 18–23, M=18.8 years.	1253	Cross-sectional survey	16-item Mobile Phone Addiction Tendency Scale for College Students	The relationship between perceived discrimination and mobile phone addiction was partially mediated by negative emotions and learning burnout.
Li et al (2024)10	China	2023	Examine the effects of new social media platforms, specifically short videos, on the prosocial behaviors of adolescents in relation to social value orientations	Mean age 16.29 years (SD=1.46)	Study 1 (N=148) Study 2 (N=152)	Two research laboratory experiments	The participant’s donation behavior.	Reward consequences bolster adolescents’ prosocial behavior, whereas punitive measures tend to adversely affect it.
Liu et al (2024)11	China	2023	Explore the impact of brief mindfulness training on college students’ mobile phone addiction and the mediating impact of sense of meaning in life.	University students, age range 17–22 (M=18.6 and 17.8 years in intervention and control groups, respectively)	44 (22 in each of the intervention and control groups)	Experiment with self-reports	16-item Mobile Phone Addiction Tendency Scale; 10-item Meaning in Life Questionnaire	Participants in the intervention group reported significantly increased mindfulness and sense of meaning in life, and decreased mobile phone addiction, whereas participants in the control group did not change significantly. Meaning in life fully mediated the intervention effect.
Liu et al (2024)12	China	Not stated	Examine the impact of social media interactions on MH. Uncover behavioral differences in information sharing between the general population and individuals with depression, while also elucidating the psychological mechanisms underlying these disparities	Mean age 30.92 years (SD= 6.62)	A pre-experiment (N=30), three experiments (Experiment 1a N=200, Experiment 1b N=180, Experiment 2 N=128)	A pre-experiment, three experiments	Intension to share (IS); The Beck Depression Inventory (BDI); Information quality (IQ)	Individuals with depression tend to favor sharing negative emotional content, a tendency that intensifies with the severity of their depression. Mediating role of self-referential processing in the relationship between content valence and information sharing
Luque-Reca et al (2024)13	Spain	Not stated	Develop and validate, through the completion of five different studies, a concise, comprehensive, and psychometrically sound Spanish self-assessed scale designed to evaluate patterns of smartphone and social media use that adversely affect the user’s personal and social functioning.	University students Young adults (18–35)	722	Mixed-methods approach across five stages: focus groups, cognitive debriefing, factor analysis	Problematic Smartphone and Social Network Use Scale (PSSNUS)	The PSSNUS manifested as a reliable instrument, comprising a structure with five key factors. The PSSNUS showed convergent and divergent validity through significant but modest correlations with daily smartphone usage hours, procrastination, emotional intelligence and mental health symptomatology
Mehmood et al (2024)14	India	Not stated	Explore the indirect (mediating) effect of internet addiction intervening between technostress and quality of life (QOL), and the moderating role of social support	University students aged mainly 18–30 years	Wave 1: 795 Wave 2: 703 Wave 3: 677	Time-lagged, longitudinal survey	4-item QOL scale	Internet addiction mediated the link between technostress and QOL and the impact of internet addition on QOL was moderated by social support
Pi et al (2024)15	China	2022	Examine latent profiles in problematic mobile social media use (PMSMU), and explore effects of gender, fear of missing out (FOMO), online positive feedback, and boredom proneness on PMSMU profiles	College students, age range 17–27 (M=20.2 years)	2591	Cross-sectional survey	20-item Problematic Mobile Social Media Use Scale	Three profiles were selected based on the PMSMU: no problems, mild problems and severe problems. Male students were less likely to have problems, while higher FOMO, more craving for online positive feedback, and more boredom proneness increased the risk.
Sang et al (2024)16	China	2022	Examine the mechanism underlying the association between intolerance of uncertainty and mobile phone addiction during the COVID-19 pandemic by examining the mediating roles of perceived stress and rumination.	18 years of age or older.	249	Cross-sectional survey	Intolerance of Uncertainty; Mobile Phone Addiction; Perceived Stress	Considerably high risk of mobile phone addiction, as three-fourths of the participants may have been susceptible to mobile phone addiction. Mediation analysis: intolerance of uncertainty affected mobile phone addiction
Shannon et al (2024)17	Canada	2020	Examine the relationship between changes over time in problematic SM use and MH outcomes in students. Whether resilience and loneliness moderated the relationship between SM use and MH.	Post-secondary students above the age of 18	104 (four weeks later: 78)	Longitudinal subgroup	Hamilton Depression Rating Scale (HDRS); General Anxiety Questionnaire 7; Loneliness Scale; Perceived Stress Scale; Brief Resilience Scale	Problematic SM use at baseline was negatively correlated with resilience and positively correlated with all other MH outcomes. Problematic SM use was associated with increased depressive symptoms and loneliness between visits. Frequency of SM use was not correlated with any MH measures at baseline
Sun et al (2024)18	China	2022	Examine the impact of recommendation algorithms on college students’ information sharing and internalizing and externalizing problem behaviors.	Chinese college students	34,752	Cross-sectional survey	The internalizing problem behaviors subscale, externalizing problems behaviors subscale	Four subgroups of college students identified in terms of social media information sharing: WeChat Moments low-frequency information sharing, middle-frequency comprehensive information sharing, TikTok high-frequency information sharing, and Sina Weibo high-frequency information sharing
Wang et al (2024)19	China	2021–2022	Examine the causal sequence of the relationship between self-esteem and problematic social media use (PSMU) over time, examine the trajectories of self-esteem and PSMU over time, and their mutual influence on each other	University students, M=18.4 years	Wave 1: 560 Wave 2: 401 Wave 3: 350 Paired across waves: 321	Longitudinal survey	Rosenberg 10-item self-esteem scale	Self-esteem negatively predicted PSMU at the following measurement, but PSMU did not predict subsequent self-esteem. Self-esteem decreased over time, while PSMU increased. The more rapid the decline in self-esteem over time, the more rapid the increase in PSMU.
Xiao et al (2024)20	China	2021	Examine the relationship between perceived social mobility and smartphone dependence, with a focus on the mediating role of hope and the moderating effect of family socioeconomic status (SES) underlying this relationship.	Mean age = 19.19, 70.2% female)	718	Cross-sectional survey	Smartphone dependence; Hope; Family Socioeconomic Status	Less perceived social mobility was linked with greater smartphone dependence. Hope mediated the relationship. Family SES moderated the first-stage path of the indirect effect through hope.
Xie et al (2024)21	China	2022–2023	Examine the longitudinal associations between cumulative ecological risk and smartphone addiction and sleep quality	University students (M=18.6 years)	Wave 1: 739 Wave 2: 653	Longitudinal survey	10-item Smartphone Addiction Scale Short Version; 18-item Pittsburgh Sleep Quality Index	Higher cumulative ecological risk was related to smartphone addiction and poorer sleep quality. Smartphone addiction fully mediated the relationship between cumulative ecological risk and sleep quality.
Xu et al (2024)22	China	Not stated	Examine electro-physiological correlates of problematic social media use and inhibitory control and explore“the role of ‘fear o” missing out’ (FOMO) in this relationship.	Not stated, open recruitment through an online chat software	66	Cross-sectional, survey and electro-physiological measurements	10-item Fear of Missing Out scale; 20-item Problematic Mobile Social Media Usage Questionnaire	PSMU appears to influence inhibitory control, and FOMO mediated their relationship.
Xu, Wang et al (2023)23	China	2022	Investigate the use of short video platforms and coping strategies during the COVID-19 pandemic lockdown.	Not specified	1569	Cross-sectional survey	7-item Generalized Anxiety Disorder Scale, change in viewing time and interaction on social media	People engaged in problem-focused and emotion-focus coping strategies in the use of short video social media for perceived stress during the lockdown.
Yan et al (2024)24	China	Not stated	Explore the relationship among the intensity of social media use, upward social comparison, cognitive overload and depressive mood.	College students	568 (125 males, 443 females)	Cross-sectional survey	The Brief Self-Rating Depression Scale PHQ-9 (Patient Health Questionnaire) Mediators: upward social comparison and cognitive load	The intensity of mobile social media use, social networking site upward social comparison, and social networking site cognitive overload were all positively correlated with depressive mood
Zhu et al (2023)25	China	Not stated	Explore the relationship between boredom proneness and bedtime procrastination in Chinese college students, and explore mediating effects of mobile phone addiction and negative emotions	College students, age range 18–32 (M=20.4 years)	668	Cross-sectional survey	9-item Bedtime Procrastination Scale; 16-item Mobile Phone Addiction Tendency Scale; 21-item Depression-Anxiety-Stress Scale	Mobile phone addiction positively predicted negative emotions and bedtime procrastination, and negative emotions predicted bedtime procrastination. The chain mobile phone addiction-negative emotions significantly influenced the relationship between boredom proneness and bedtime procrastination.

Notes: Authors written in bold denote the five studies identified by the guest advisors as most directly examining relationships between social media use and mental health, which are discussed in detail in the main text.

Study Characteristics and Suggestions for Future Research

The studies included in the collection were conducted in different countries representing both the Eastern and Western hemispheres, including China, India, Egypt, Saudi Arabia, Canada, USA, and Spain. The majority (n=16) were conducted in China. It is interesting to see so many research studies on social media use in relation to mental health in China, particularly considering the strict internet regulations in that country and that the social media frequently used (eg, WeChat, Douyin, Weibo) differ from those dominating in Western countries (eg, Facebook, Instagram, TikTok). We cannot help but wonder what this concentration of research might imply, although we know these are mere speculations. Could it be related to specific governmental or academic funding priorities in China? Or is there in China greater societal concern about the effects of technology on its youth? It appears that comparative cross-cultural research, not only on social media use but also on the political, cultural and economic contexts for conducting such research, is called for. In this article collection, only three studies used samples with participants from two or more countries, indicating that most research studies in this area are still firmly established in country-specific contexts. Conducting more cross-national research will allow for valuable between-country comparisons that may also consider the relevance of different types and brands of social media used in different contexts.

One of the included studies had data collected in 2018, while most of the others had data collected from 2020 onwards. However, although we may assume that the data in all studies were relatively recently collected, the time of data collection was not explicitly stated in nine of the studies. This is particularly vital in social media research, where platform features, user demographics, and societal norms can shift dramatically in a matter of months, potentially altering the very mechanisms being studied. Thus, we urge researchers to clarify the time of data collection in their studies, enabling readers to better understand the context of the study.

The target populations in the studies were most often young people, with three studies using samples of adolescents (up to about 18 years of age) and 16 studies using samples of young people between 18 and 30 years, commonly university students. The six remaining studies had samples comprised of adults 18 years and older. The age composition in the studies reflects that young people spend more time on social media than older people and may also be more prone to experiencing psychological impact from using them. However, in the years to come we will expect more people in the general adult and older adult age groups to be regular users of social media. Thus, patterns of social media use in the mature population, and how they relate to their mental health and wellbeing, are important lines of future research. As an example, how social media relates to loneliness and social support in older, homebound populations is an important question to clarify.

Sample sizes varied considerably, from 44 to 336,000. As expected, large surveys and registry-based studies were often able to obtain large sample sizes, while experiments had small sample sizes. With a view to study design, 18 of the studies were cross-sectional (including one mixed-methods study), indicating that the majority of studies employ relatively weak research designs, rendering associations between variables open to interpretation. Seven studies had longitudinal designs, of which three were experimental. Researchers using longitudinal study designs, and particularly experimental designs, is uplifting given that this is the way forward for establishing a clearer understanding of how social media use and mental health factors are related across time. Several of the studies used mediator or moderator analysis, which is also indicative of efforts to find out how and for whom certain factors are intrinsically connected. While experimental designs may be difficult to set in motion for a variety of reasons, we would like to point to the possibilities arising naturally, eg from the implementation of new policies. For example, in November 2024 Australia passed the Online Safety Amendment (Social Media Minimum Age) Act, which is believed to come into effect late 2025. How will this affect the mental health of children and adolescents? Researchers should see and exploit the opportunities that come with changes in policies. Given the mediational pathways through cyberbullying and social comparison, as described in two of the included articles,^4,24 policy interventions that aim to reduce these specific online harms – as opposed to just restricting access by age – could be particularly effective. Future research should be designed to evaluate the specific mechanisms through which such policies impact youth mental health.

The outcomes used in the included studies span widely. They include, amongst others, mental health and mental health problems (symptom scales), self-esteem, quality of life, mobile phone addiction, problematic social media use, sleep patterns and sleep quality, and compulsive shopping behavior. Thus, they can be summarized as including behaviors related to the use of internet, mobile phones, and social media, as well as behavioral and psychological aspects of mental health and well-being. Researchers in some studies treated mental health-related factors as an outcome predicted by social media use, whereas others treated social media use (or some other technology-related factor) as an outcome predicted by mental health. In our view, the magnitude of measures being used and the variation in modelling their associations very much reflect the state of the art, and we believe a diversity of research approaches is needed also in the future. When possible, however, researchers should try to align their research with existing research to enable comparisons and thereby increase the chances of knowledge progress. Related to the same point, future research should build on appropriate conceptual or theoretical frameworks, as such frameworks can help guide research and streamline hypothesis generation and testing. While all articles included in the article collection were well situated in their empirical context, the majority (ie, 18 articles) were also explicitly designed and executed based on relevant theory, such as Uses and Gratifications Theory,¹⁵ Interaction of Person-Affect-Cognition-Execution (I-PACE),⁹ the Stressor-Strain-Outcome framework,¹⁴ or Social Cognitive Theory.¹⁰ Thus, this article collection stands out as important in the way it clearly builds on the ideas examined in the past, enabling researchers to use it as a point of reference for their future work.

Study Results

The article collection had a broad approach to inclusion; thus, studies with very different ways of looking at social media, mental health, and related concepts were included in the collection. Our review (see Table 1) showed that only five of the 25 articles^{4,12,17,19,24} included measures specifically concerned with social media use and mental health, and we will look more closely at these.

Cross-Sectional Analyses

Results from the study by Shannon et al¹⁷ indicated that higher problematic social media use was related to higher depressive symptoms, anxiety, stress, and loneliness. Similarly, Yan et al²⁴ found that while greater intensity of social media use was related to higher levels of depressive symptoms, this relationship was fully mediated by social comparison and cognitive overload. Furthermore, Barragan et al⁴ explored the complexities surrounding this association and found that the links between greater social media use and higher levels of depression, anxiety, and substance use were mediated by cyberbullying victimization. Interestingly, this mediation effect was found to be stronger for younger individuals compared to older ones.

Longitudinal Analyses

Results from the included longitudinal studies showed that increases in problematic social media use were related to increases in depressive symptoms and loneliness.¹⁷ Conversely, self-esteem was found to negatively predict subsequent problematic social media use, implying that a more rapid decline in self-esteem can lead to a faster increase in problematic social media use.¹⁹

Experimental Analyses

Results from the experimental study conducted by Liu et al¹² suggested that, within contexts of low information quality, individuals with depression were more likely to share negative emotional content compared to the general population, a tendency that amplified with the severity of depression. Furthermore, self-referential processing was identified as a mediator between the emotional content and the intention to share. However, as the severity of depression escalated, the mediating effect diminished. A model based on the synthesized findings from the reviewed studies is shown in Figure 1.

Figure 1.

Overview of associations between social media use and mental health as reported in the five studies. The upper part of the figure displays social media use and social media intensity as predictor variables; cyberbullying, social comparison and cognitive overload as mediators, and depressive symptoms, anxiety symptoms, stress, loneliness and substance use as outcomes. The lower part of the figure displays depressive symptoms and self-esteem as predictors, self-referential processing as mediator, and problematic social media use and sharing of negative emotional content as outcomes. Arrows indicate the direction of the relationships. Solid arrows indicate direct effects while dashed arrows indicate indirect effects.

Conclusions and Implications for Research

The studies reviewed above showed that higher social media use was related to increased depressive symptoms, anxiety, stress, loneliness, and substance use, and that cyberbullying, social comparison, and cognitive overload mediated between social media use and the outcomes. These findings clearly suggest that relationships between social media use and mental health should be explored further via mediation analyses, to examine why such relationships exist. While time on social media alone is not irrelevant for understanding variations in mental health, more fruitful explanations may concern the events occurring on social media and individual vulnerabilities and resilience in the face of such events.

Moving beyond how social media impacts mental health, the collection also provides compelling evidence for the reverse pathway: how pre-existing mental health states influence online behavior. People with depression and self-esteem problems were found to be more inclined to share negative content on social media. Self-esteem inversely affected subsequent problematic social media use, and self-referential processing mediated between emotional content and sharing intentions among people with depression. These findings point in a new direction, suggesting that mental health and self-esteem problems can affect not only time use on social media, but also the contents of one’s sharing on social media. They also point to specific psychological processes that may contribute to explaining social media sharing intentions. Given the predominance of cross-sectional studies in the field, more longitudinal studies examining associations over time are warranted. Moreover, understanding sharing patterns and the contents of social media sharing in relation to mental health as well as contextual factors, appear to be interesting ways forward.

As guest advisors for the article collection, we would like to thank all authors for their contributions. In our opinion, the article collection brings new and important insights into how mental health is shaped, and how mental health shapes behaviors, in the modern world of social media. It is our hope that researchers can use the article collection, and indeed this synopsis and commentary to the collection, as a point of reference when choosing new research questions to explore and when deciding on certain aspects of design and methodology. Finally, it may be worth considering that the articles in this collection only represent a small fraction of the research on social media and mental health that evolves and is being published rapidly. Moreover, they generally focus on the pathological or negative aspects of social media use (eg, depression, anxiety, addiction). We may need to remind ourselves and our colleagues that people’s use of social media is many-faceted as are its impacts, and that this should be acknowledged in our collective research efforts.

Disclosure

The authors report no conflicts of interest in this work.