Skip to main content
NCBI home page
Springer logoLink to Springer
. 2024 Oct 12;54(4):849–862. doi: 10.1007/s10964-024-02093-5

An Investigation of the Longitudinal Bidirectional Associations Between Interactive Versus Passive Social Media Behaviors and Youth Internalizing Difficulties. A Within-Person Approach

Maria Tibbs 1,2,✉, Sonya Deschênes 1, Peter van der Velden 3,4, Amanda Fitzgerald 1
PMCID: PMC11933239  PMID: 39395918

Abstract

Ongoing concerns about the mental health of young people have intensified interest in the role of social media, with research suggesting that the nature of social media behaviors—whether interactive or passive—may differentially impact mental health. However, the bidirectional relationships between specific types of social media use and internalizing difficulties (anxiety and depression) remain underexplored, particularly at the within-person level over time. Data were extracted from the Dutch population-based Longitudinal Internet Studies for the Social Sciences (LISS) panel. Four yearly studies assessing time spent on interactive (communication) and passive use of social media in October (2019–2022) and four annual studies evaluating internalizing difficulties (anxiety and depression) in November (2019–2022) were used. Respondents who were 16- 25 years old in October 2019 (N = 321; M = 20.49; SD = 2.94; 61.7% female). Random Intercept Cross Lagged Path Models were used to analyze the data. There were no significant associations between passive social media use and internalizing difficulties at the within-person level over time. Within-person interactive use was associated with decreases in internalizing difficulties at one time point (2021 to 2022). The results provide marginal evidence that distinct social media behaviors are prospective factors associated with internalizing difficulties in young people.

Introduction

The rapid evolution of social media has profoundly impacted adolescent psychosocial development. With a substantial increase in social media use among young people, evidenced by a two-fold rise in the number of adolescents online ‘almost constantly’ (Pew Research Centre, 2022) since 2014, there has been an increasing interest in how these platforms influence youth mental health. Despite growing research, inconsistencies remain in linking social media behaviors with mental health outcomes, often due to the oversimplification of effects by examining social media use and neglecting the bidirectional relationship between mental health and social media behaviors. This study aims to address these gaps by employing random intercept cross-lagged panel models to explore the bidirectional relationships between different social media behaviors—interactive (two-way communication such as calling, messaging, or chatting) and passive use (one-way engagement by reading and viewing content without interaction Kaye, 2021)—and internalizing difficulties among young people.

Youth Psychosocial Development in the Context of Social Media

Young people are disproportionately affected by mental health concerns, due to a combination of biological and social developmental changes occurring during adolescence and early adulthood (Ogden & Hagen, 2018). Recent evidence suggests that approximately 62.5% of all serious mental health difficulties emerge before the age of 25 (Solmi et al., 2022), with reports that the prevalence of youth mental health difficulties has grown steadily over the past two decades (Collishaw, 2015; Fusar-Poli, 2019; Wiens et al., 2020). Given the enduring functional impact of mental health difficulties on the lives of young people (Scott et al., 2014), it is crucial to enhance our understanding of the factors influencing youth mental health. One such factor that has gained particular attention in recent decades is social media use (Keles et al., 2019; Valkenburg et al., 2022).

Social media use, encompassing the use of platforms for online content creation, sharing, and interaction (Kietzmann et al., 2011) is continually evolving. In recent years, the exponential rise of social media has fundamentally transformed how young people engage with the world around them. Young people are among the highest users of social media platforms, with recent data showing a two-fold increase in the number of 13–17-year-olds who report being online ‘almost constantly’ from 2014–2015 to 2022 (24 to 46%; Pew Research Centre, 2022). Social media now plays a pivotal role in the social development of adolescents and emerging adulthood by facilitating communication and interaction crucial for friendship formation. These interactions are essential for young people’s social development, providing opportunities to build, maintain, and strengthen peer relationships (Nesi et al., 2018b, 2018a). Moreover, emerging research identifies social media as a critical facilitator of identity exploration and self-concept development (Andalibi et al., 2017; Lee & Borah, 2020; Pérez-Torres, 2024) offering a platform for experimenting with roles, receiving feedback on self-presentation, and navigating the core developmental tasks of adolescence and emerging adulthood. However, this increase in social media use has also raised concerns and is suggested to be a contributing factor to the recent reported increases in youth mental health difficulties (Twenge et al. 2018).

The Mental Health Impact of Social Media and The Need for Specificity in Terminology

A growing body of research has explored the relationship between social media use and mental health. Cross-sectional studies have provided insight into associations between overall social media use and internalizing difficulties, often highlighting negative effects, particularly for young girls (Svensson et al., 2022; Twenge & Martin, 2020). These studies, however, fail to address causal relationships and cannot account for within-person dynamic associations over time.

Still, longitudinal studies in this area have produced mixed findings. Some have demonstrated small to moderate positive associations between total time spent on social media and internalizing difficulties, such as anxiety and depression, during adolescence and emerging adulthood, especially for young girls (Leo et al., 2021; Riehm et al., 2019; Thorisdottir et al. 2019). Conversely, other longitudinal studies have found no significant associations between overall social media use and internalizing difficulties over time (Jensen et al., 2019; Nesi et al., 2017). These inconsistencies likely arise from oversimplified definitions of total social media use.

Although understanding that social media behaviors often co-occur, recent literature distinguishes between several types of social media behaviors, including interactive, broadcasting (posting), reactive (liking, commenting), and passive use (Kaye, 2021). Moving from the more simplistic dual distinguishing of active and passive social media use (Verduyn et al., 2017), this approach acknowledges the user as either a sender or recipient of information online. While users can never be truly “passive” (Livingstone, 2014), newer classifications describe passive social media use as a one-way interaction, where users primarily consume information by reading and viewing pages, feeds, sites, and timelines. In contrast, truly interactive behaviors include two-way engagements such as calling, messaging, or chatting. Despite these recent distinctions in social media behavior, the body of longitudinal multi-wave studies examining these behaviors and mental health remains scarce and where there is an investigation, findings have again been mixed to some extent. For example, passive social media use has been associated with increased levels of depression over time (Wang et al., 2019). On the other hand, a prior investigation using 2016 and 2017 data from the LISS longitudinal panel under investigation in the current study in years previous, showed that passively viewing, and interactively communicating, were not associated with depression and anxiety one month or one year later in a subgroup of 16–24-year-olds (van der Velden et al., 2019).

A number of theories dominate the literature regarding social media use and mental health difficulties. The social displacement hypothesis (Kraut et al., 1998) proposes that social media use poses a risk to youth mental health, because time spent on internet technologies may displace protective and health-promoting activities such as in-person time spent socially engaging with friends and family (Dworkin et al., 2018; Twenge et al., 2019). In particular, passive viewing of feeds and timelines on social media use has been shown to contribute to feelings of missing out (FOMO), increased upward social comparison, exposure to harmful material online, and in turn internalizing difficulties (Hall et al., 2019; Oberst et al., 2017). In contrast, the enhanced self-disclosure hypothesis (Valkenburg & Peter, 2009), argues that online communication relative to face-to-face interactions allows for increased uninhibited self-disclosure. This increased self-disclosure is thought to enhance a young person’s social network by increasing perceived social support and connectedness (Huang, 2016), and in turn, decreasing internalizing difficulties (Niederhoffer & Pennebaker, 2009; Pennebaker, 1997). Social media’s impact on mental health can be complex and multifaceted, with both displacement and enhanced self-disclosure contributing to different aspects of adolescent adjustment. While passive use may pose risks, interactive use can offer benefits, and both may interact in ways that affect mental health outcomes differently.

Bidirectional Associations Between Social Media Behaviors and Youth Mental Health

Evidence suggests a bidirectional relationship between social media use and internalizing difficulties, with young people experiencing higher levels of depression and anxiety often engaging more with social media (Orsolini et al., 2022; Primack et al., 2011; Toma, 2022; Weidman et al., 2012). However, many longitudinal that have investigated this relationship have focused predominantly on overall social media use (Coyne et al., 2020; Kelly et al., 2018; Li et al., 2018; Tang et al., 2021; Tian et al., 2017). Among the studies that have disaggregated social media behaviors, findings have been inconsistent. For instance, Cheng et al., 2023 found that passively viewing friend’s social media sites predicted higher levels of depression but this did not apply in reverse. Similarly, Wang et al. (2018) found bidirectional associations between passive social media use and subjective wellbeing, such that increases in passive use predicted decreases in subjective wellbeing, and decreases in subjective wellbeing predicted increased passive use.

Furthermore, the importance of interaction for youth social development cannot be overstated, as evidence supports positive peer, friend, romantic partner, and family interaction as associated with reductions in internalizing difficulties such as anxiety and depression (Long et al., 2020; Zapf et al., 2024). Interactive communication on social media, such as messaging and chatting can play a crucial role in fostering these social connections, providing emotional support, and enhancing self-esteem (Bonetti et al., 2010; Davis, 2012; Dolev-Cohen & Barak, 2013; Prescott et al., 2017). Similarly, the social compensation hypothesis (Kraut et al., 1998) suggests that those young people experiencing depression or anxiety might prefer to communicate online in an attempt to compensate for the potential lack of offline social interactions (Desjarlais & Joseph, 2017). Within the longitudinal literature, very few studies have examined the specific impacts of social media behaviors. However, some research has indicated that online chatting can predict higher self-esteem and reduced depression, particularly among more introverted young people (Van Zalk et al., 2011). Similarly, private messaging on platforms like Facebook has been identified as a protective factor against internalizing difficulties in adolescents, largely due to increased perceived social support (Frison et al., 2019). These findings suggest that interactive social media behaviors, such as chatting and private messaging, may play a beneficial role in mitigating internalizing difficulties.

The Predominant Focus on Between-Person Associations

The longitudinal literature to date on social media use has predominantly focused on between-person approaches, that is, examining how average trends evolve across the total study sample over time. This approach, often referred to as group-level or between-person analysis, neglects the potential for individual-level variations over time, obscuring the potential impact of social media use on internalizing difficulties at the individual level. Within-person approaches, on the other hand, statistically account for additional variance at the individual level, that is, how an individual’s scores on a particular construct vary when compared to their scores at previous time points. Within the context of social media research, examining these within-person associations is critical due to the potential significant individual variation in both social media usage patterns and mental health profiles. Each young person’s unique engagement with social media and their distinct mental health experiences may lead to different outcomes, thus capturing these individual variations may provide better insights into the relationship between social media behaviors and internalizing difficulties. In the limited studies disaggregating between-person and within-person associations, again the focus has been on overall social media or internet use rather than distinct social media activities (Coyne et al., 2020; Fitzpatrick et al., 2023; Marciano et al., 2022).

Current Study

Research on the bidirectional associations between specific social media behaviors and internalizing difficulties in youth remains limited, particularly regarding within-person associations over time. No study has yet investigated the directionality of within-person associations between distinct social media behaviors—interactive use (e.g., messaging, chatting) and passive use (e.g., reading, viewing without interaction)—and internalizing difficulties such as depression and anxiety in youth. Furthermore, many previous studies have focused on timeframes from the early to mid-2010s, potentially overlooking the impact of newer and evolving social media platforms. To address these gaps, this study aims to examine the bidirectional associations between specific social media behaviors and internalizing difficulties. While prior research has often been cross-sectional or focused on overall social media use, this study explores how interactive communication and passive content consumption differentially impact mental health over time. We hypothesize that interactive communication will reduce internalizing difficulties (Hypothesis 1), whereas passive use will increase them (Hypothesis 2). Employing random intercept cross-lagged panel models, this study provides a nuanced understanding of these associations at the within-person level.

Method

Sampling Procedure

Data were extracted from the Longitudinal Internet Studies for the Social Sciences (LISS) panel by Centerdata, Tilburg, Netherlands. The LISS panel is based on a probability sample of households drawn from the Dutch register by Statistics Netherlands, comprising of approximately N = 5000 households (Scherpenzeel, 2011). The panel consists of all households where individuals are currently living at addresses randomly drawn from a sample provided by Statistics Netherlands (CBS). Participants aged 16+ are invited to participate in the panel. The panel has been collecting data since 2007. Respondents are given a small incentive of €15 per hour of participation and provided explicit consent for the use of their data for research purposes.

For the present study, data was extracted from 8 studies conducted with the LISS panel between 2019 and 2022). Social media use data (interactive communication and passive reading and viewing content) was extracted from four annual Social Integration and Leisure surveys, administered in October each year (T1a, T2a, T3a, T4a). Mental health data was extracted from four annual Health surveys, administered in November or each year (T1b, T2b, T3b, T4b). To be included in the current sample, respondents must have been between 16 and 25 years of age at the first wave (T1a; Oct-Nov 2019; N = 684). Further, participants must have completed all baseline measures of interest (2019) and all measures of interest in at least one follow-up wave (Oct–Nov 2020 to Oct- Nov 2022). This process resulted in a final sample of N = 321, or 46.9% of all those 16–25-year-olds in the first wave (see Supplementary Materials for demographic comparisons of those young people aged 16–25 who did not complete measures at baseline and one follow-up time point). Of the total analytic sample, missingness on measures of interest at T2-T4 ranged between 7.5% to 29.6%. At T1a, the analytical sample included 61.7% (n = 198) females, and the mean age was 20.49 (SD = 2.94).

Measures

Internalizing behaviors

Internalizing behaviors were assessed using the Mental Health Inventory-5 (MHI-5). The MHI-5 is a short five-item subscale of the Medical Outcomes Study (MOS) Short Form Health Survey (Ware & Sherbourne, 1992) which assesses depressive and anxiety symptoms in the previous month. Items are scored on a six-point Likert scale (0 = never, 5 = continuously) and include items such as “Have you been a very nervous person?” and “Have you felt so down in the dumps that nothing could cheer you up?”. As outlined in the scoring manual, following reverse scoring of negatively worded items, total scores were linearly transformed (i.e., multiplied by four), representing a range of 0–100 with lower scores indicative of higher levels of anxiety and depression. The omega coefficient has been shown to outperform Cronbach’s alpha in cases where assumptions of tau equivalence are violated (Dunn et al., 2014), and in cases where this assumption is met, omega has been reported to perform as well as Cronbach’s alpha (Zinbarg et al., 2005). As such, the Omega coefficient was used to assess the internal consistency of the MHI-5 across all waves (2019–2022; ω = 0.86–0.87).

Interactive Communication and Passive Reading and Viewing Social Media

Respondents’ active communication and passive reading and viewing on social media were assessed by asking the “average number of hours per week spent on: reading and viewing social media (e.g., Facebook, Instagram, Twitter, YouTube, LinkedIn, Google + , Pinterest, Flickr, or similar services” (passive use; reading and viewing content) and the “average number of hours per week spent on chatting, video calling or sending messages via WhatsApp, Telegram, Snapchat, Skype or similar services” (interactive use; online communication).

Covariates

Given the age range of respondents is inclusive of two distinct developmental phases, age was included as a covariate to control for the potential differences in communication and mental health across time. As outlined, gender differences are thought to exist between social media use and youth. Consequently, gender was included as a covariate across both models. Finally, to account for the potential confounding impact of perceptions of social media on communication patterns, respondents were asked to rate the extent to which they agreed with the following statement on a 5-point Likert scale (1 = completely disagree, 5 = completely disagree): “Because of social media sites it is easier for people to maintain friendships”. All covariates were measured at T1 (October 2019). Family level covariates, such as family income as a proxy for socioeconomic status were not included as the proportion of missing data was deemed too considerable.

Data Analytic Plan

Results indicated that data were missing completely at random (MCAR; χ2 = 576.14, p = 0.90). As such, full Information Maximum Likelihood (FIML) with robust standard errors was used to impute missing values. FIML is known to provide less biased estimates than pairwise or listwise deletion in structural equation models (Enders & Bandalos, 2001; Lee et al., 2019).

Preliminary analyses examined the descriptive statistics, distributions, and bivariate correlations of all key variables. Intra-class correlation coefficients (ICCs) were calculated for internalizing difficulties and interactive and passive social media behaviors across the four included waves. ICCs estimate the proportion of between-person explained variance relative to the total variance explained.

In line with the procedure outlined by Mulder and Hamaker (2021), two random intercept cross-lagged panel models (RI-CLPM) were estimated to model bidirectional associations between interactive communication and internalizing difficulties, as well as bidirectional associations between passive social media use and internalizing difficulties over time. Compared to the conventional Cross Lagged Path Model (CLPM), the RI-CLPM disentangles an individual’s observed variance into three components. Specifically, variance is separated into grand means, stable between-person effects as represented by random intercepts, and dynamic within-person variance, determined by an individual’s deviation from their expected score (i.e., the difference between individual scores and the sum of the grand mean at each wave and the random intercept) (Hamaker et al., 2015). To account for the time lag of approximately one month between administered surveys of social media behaviors and mental health, unidirectional regression paths were added to the model from social media behaviors to mental health. All models were estimated using the “Lavaan” package (Rosseel, 2012) in R statistical software (R Core Team, 2021).

Measurement error variances of the observed scores were constrained to zero to improve parsimony and simplicity of interpretation. In this way, variation in the observed variables was entirely accounted for by the within-person and between-person latent factor structure. The models were built iteratively across several steps. First, a fully unconstrained model was tested. This was followed by the addition of equality constraints to autoregressive and cross-lagged associations as well as directional regression paths from social media behaviors to internalizing difficulties across each wave. Finally, variances, and covariances of latent variables across T2a (Oct-Nov 2020), T3a (Oct–Nov 2021), and T4a (Oct-Nov 2022) were constrained to be equal.

Accounting for deviations in normality, robust maximum likelihood estimation was employed to estimate the models and the Satorra-Bentler scaled chi-square test was used to assess fit. Global model fit was assessed using the chi-square statistic, comparative fit index (CFI), Root Mean Square Error of Approximation (RMSEA), and Standardised Root Mean Square Residual (SRMR; Kline, 2016). For Goodness of Fit indices based on maximum likelihood estimation, models were considered a good fit if CFI > 0.90 for acceptable and >0.95 for excellent fit; RMSEA < 0.05 for good and < 0.08 for acceptable fit; and SRMR < 0.08 (Hu & Bentler, 1999). The Satorra–Bentler scaled chi-square difference test (Satorra & Bentler, 2001) ∆CFI, ∆RMSEA, and reductions in AIC and BIC were used to compare the nested model.

Results

Preliminary Analyses

The means, standard deviations, and correlations for all study variables across each wave are presented in Table 1. ICC estimations revealed that 24% (ICC = 0.24) of the variance in passive social media use, 27% (ICC = 0.27) of the variance in interactive communication on social media, and 61% (ICC = 0.61) of the variance in internalizing difficulties could be attributed to between-person effects. In other words, a substantial proportion of the observed variations across all key variables (39 to 76%) were due to individual differences in young people’s reports of their social media behaviors and mental health over time. Considering the ratio of within-person dynamic variance, it was deemed appropriate to run the RI-CLPM (Hamaker et al., 2015).

Table 1.

Means, standard deviations, and bivariate correlation coefficients among all study variables

M (SD) 1 2 3 4 5 6 7 8 9 10 11 12
1. Internalizing difficulties T1b 69.40 (16.26) -
2. Internalizing difficulties T2b 69.23 (16.06) 0.65*** -
3. Internalizing difficulties T3b 68.51 (16.16) 0.55*** 0.65*** -
4. Internalizing difficulties T4b 68.32 (15.62) 0.63*** 0.66*** 0.67*** -
5. Interactive Use T1a 6.20 (8.70) −0.12* −0.09 −0.14* −0.12 -
6. Interactive Use T2a 6.48 (8.51) −0.09 0.03 −0.09 0.01 0.25*** -
7. Interactive Use T3a 5.87 (7.26) −0.12 0.04 −0.10 −0.03 0.27*** 0.45*** -
8. Interactive Use T4a 5.75 (7.50) −0.18*** −0.14* −0.13 −0.11 0.27*** 0.53*** 0.35*** -
9. Passive Use T1a 7.50 (10.61) −0.17*** −0.14* −0.14* −0.17*** 0.61*** 0.23*** 0.23*** 0.17* -
10. Passive Use T2a 8.14 (7.52) −0.21*** −0.11 −0.16* −0.18*** 0.23*** 0.43*** 0.27*** 0.31*** 0.41*** -
11. Passive Use T3a 7.95 (8.27) −0.09 −0.06 −0.11 −0.14* 0.13* 0.27*** 0.42*** 0.15* 0.43*** 0.51*** -
12. Passive Use T4a 7.35 (7.88) −0.13* −0.13 −0.19** −0.17* 0.16* 0.30** 0.21*** 0.45*** 0.33*** 0.50*** 0.42*** -
Open in a new tab

T = Time, 1 = 2019, 2 = 2020, 3 = 2021, 4 = 2022. a October-November, b November-December. Spearman Rank Order Correlations are presented due to non-normality

*p < 0.05, ***p < 0.001. High scores on the MHI-5 are indicative of lower levels of internalizing difficulties

Random Intercept Cross-Lagged Panel Model (RI-CLPM) Results

Interactive communication on social media

The model-building procedure indicated that model fit did not improve significantly upon adding constraints over time (i.e., autoregressive, and cross-lagged associations constrained to be equal over time, and variances and covariances of observed variables constrained to be equal over time). Consequently, the unconstrained model was fit. See Supplementary Material for further details. Model 1 examining associations between interactive use and internalizing difficulties over time demonstrated excellent model fit (χ2 = 25.00, df = 27, p < 0.001, CFI = 1.00, RMSEA = 0.00 [90% CI = 0.00–0.05], SRMR = 0.03).

At the within-person level, no significant autoregressive effects were observed for communication or mental health indicating that previous levels of these variables were not predictive of their values in the subsequent year. There was only one significant cross-lagged path, where communication on social media at T3a predicted reductions in internalizing difficulties at T4b (β = 0.32, p = 0.00). At the between-person level, a significant correlation between the random intercepts of communication and internalizing difficulties was observed, (r = −0.50, p <0.01), indicating that on average, young people who spent more time communicating experienced higher levels of internalizing difficulties (lower MHI-5 scores indicate more internalizing difficulties). Although the primary hypotheses investigated the bidirectional associations over a longer period, unidirectional within-person associations between interactive use in October 2020 was predictive of lower levels of internalizing difficulties in November 2020 (β = 0.26, p < 0.001). No other unidirectional regression paths were significant. All model results, including unstandardized estimates and standard errors are presented in Table 2. A visual representation of the final statistical model is presented in Fig. 1.

Table 2.

Parameter estimates for final random intercept cross-lagged path model associating interactive SM use and internalizing difficulties

B SE β
Interactive social media use
Associations
Between-person association −17.40 6.17 −0.50
Within-person cross-lagged associations
Inter T1a → ID T2a −0.02 0.10 −0.05
Inter T2a → ID T3a 0.03 0.03 0.07
Inter T3a → ID T4a 0.50 0.04 0.32
ID T1b → Inter T2a 0.37 0.22 0.13
ID T2b → Inter T3a 0.49 0.27 0.21
ID T3b → Inter T4a 0.33 0.23 0.14
Within-person autoregressive associations
Inter T1a → Inter T2a 0.16 0.17 0.17
Inter T2a → Inter T3a 0.20 0.10 0.24
Inter T3a → Inter T4a 0.07 0.15 0.07
ID T1b → ID T2b 0.16 0.11 0.17
ID T2b → ID T3b 0.21 0.11 0.19
ID T3b → ID T4b 0.18 0.11 0.21
Inter T1a → ID T1b (1-month lag) 0.00 0.03 0.01
Inter T2a → ID T2b (1-month lag) 0.09 0.02 0.26
Inter T3a → ID T3b (1-month lag) 0.02 0.03 0.05
Inter T4a → ID T4b (1-month lag) 0.03 0.03 0.09
Between-Person (Random Intercept) Covariates
Gender T1→ InterT1 −0.06 0.61 −0.02
Age T1 → InterT1a −0.29 0.11 −0.09
SM Maintain Friendships T1 → InterT1 0.24 0.52 0.08
Gender T1 → ID T1 −1.16 0.34 −0.40
Age T1 → ID T1 0.10 0.06 0.04
SM Maintain Friendships T1 → ID T1 0.14 0.22 0.05
Open in a new tab

Bold inidcates a significant path. Inter = Interactive communication on social media (h/week); ID Internalizing difficulties, N = 312, N = 4 measurements, Standardized effects (β) are indicators of effect size, B unstandardized regression coefficient, SE standard error. Higher scores on the IDI-5 are indicative of lower internalizing difficulties. a October-November, b November-December Bolded figures indicate statistical significance

Fig. 1.

Fig. 1

Open in a new tab

Statistical RI-CLPM model of interactive communication on social media and internalizing difficulties. Inter = Interactive communication on social media; ID = internalizing difficulties. Standardized effects (β) are presented. Dashed lines represent non-significant paths. Bolded lines represent significant paths

Passive reading and viewing on social media

In line with the preceding model, the process of model building for passive reading and viewing suggested that model fit did not significantly improve upon adding constraints over time. As a result, the unconstrained model was fit (details presented in Supplementary Material). Model fit was excellent (χ2 = 26.34, df = 27, p = 0.50, CFI = 1.00, RMSEA = 0.00 [90% CI = 0.00–0.05], SRMR = 0.04). See Fig. 2 for conceptual model.

Fig. 2.

Fig. 2

Open in a new tab

Statistical RI-CLPM model of passive reading and viewing social media and internalizing difficulties. Pass = Passive reading and viewing on social media; standardized effects (β) are presented. ID = internalizing difficulties. Standardized effects (β) are presented. Dashed lines represent non-significant paths. Bolded lines represent significant paths

At the within-person level, no significant autoregressive or cross-lagged effects were observed for passive reading and viewing social media and internalizing difficulties. At the between-person level, a significant correlation was observed between the random intercepts of passive use and mental health, (r = −0.21, p = 0.03), indicating that on average those young people who spent more time passively engaging on social media experienced higher levels of internalizing difficulties. Although the primary hypotheses investigated the bidirectional associations over a longer period, unidirectional within-person associations between passive reading and viewing and mental health one month later were not significant across all waves. All model results, including unstandardized estimates and standard errors, are presented in Table 3. A visual representation of the final statistical model is presented in Fig. 2.

Table 3.

Parameter estimates for final random intercept cross-lagged path model associating passive reading and viewing social media and internalizing difficulties

B SE β
Passive reading and viewing on social media
Associations
Between-person association −2.62 1.23 −0.21
Within-person cross-lagged associations
Passive T1a → ID T2b −0.05 0.01 −0.06
Passive T2a → ID T3b −0.27 0.06 −0.13
Passive T3a → ID T4b 0.05 0.04 0.04
ID T1b → Passive T2a −0.04 0.17 −0.02
ID T2b → Passive T3a 0.40 0.43 0.14
ID T3b → Passive T4a −0.06 0.22 −0.03
Within-person autoregressive associations
Passive T1a → Passive T2a 0.06 0.07 0.10
Passive T2a → Passive T3a −0.04 0.20 −0.03
Passive T3a → Passive T4a −0.02 0.05 −0.02
ID T1b → ID T2b 0.13 0.11 0.14
ID T2b → ID T3b 0.20 0.11 0.17
ID T3b → ID T4b 0.15 0.10 0.19
Passive T1a → ID T1b (1-month lag) −0.00 0.02 −0.01
Passive T2a → ID T2b (1-month lag) 0.03 0.04 0.07
Passive T3a → ID T3b (1-month lag) 0.01 0.05 0.03
Passive T4a → ID T4b (1-month lag) −0.01 0.04 −0.02
Between-Person (Random Intercept) Covariates
Gender T1 → Passive T1 0.28 0.62 0.06
Age T1 → Passive T1 −0.54 0.12 −0.12
SM Maintain Friendships T1 → Passive T1 0.41 0.45 0.10
Gender T1 → ID T1 −1.19 0.39 −0.39
Age T1 → ID T1 0.10 0.06 0.03
SM Maintain Friendships T1 → ID T1 0.01 0.22 0.03
Open in a new tab

Bold inidcates a significant path. Passive = Passive reading and viewing content on social media (h/week); ID Internalizing difficulties, N = 312, N = 4 measurements, Standardized effects (β) are indicators of effect size, B unstandardized regression coefficient., SE standard error. a October-November, b November-December. Higher scores on the IDI-5 are indicative of lower internalizing difficulties

Associations Between Random Intercepts and Covariates

Active Communication and Internalizing Difficulties

Age was a significant predictor of the random intercept of communication, such that those who were younger spent more time communicating online (β = −0.09, p = 0.01). Gender significantly predicted the random intercept of internalizing difficulties, revealing that, on average, females experienced higher levels of internalizing difficulties (β = −0.40 p = 0.00). No remaining covariates were significant predictors of the random intercepts of active communication on social media and internalizing difficulties.

Passive Reading and Viewing and Internalizing Difficulties

Age was a significant predictor of the random intercept of passive use, such that those who were younger spent more time communicating online (β = −0.12, p < 0.001). Gender significantly predicted the random intercept of internalizing difficulties, revealing that on average, females experienced higher levels of internalizing difficulties (β = −0.40 p = 0.00). No remaining covariates were significant predictors of the random intercepts of passive engagement on social media and internalizing difficulties.

Sensitivity Analysis

To ensure robustness, both models, that is those which examined interactive online communication and passive reading and viewing were re-run with complete cases (N = 174) rather than Full Information Maximum Likelihood Imputation (FIML) and results remained consistent across both models.

Discussion

To date, there is no consensus on the relationship between distinct social media behaviors and youth mental health. Much of the existing research has been cross-sectional, focused broadly on overall social media use, and has often overlooked potential bidirectional effects and within-person variations. This study aimed to address these gaps by longitudinally examining the bidirectional associations between two distinct social media behaviors—interactive communication and passive use—and internalizing difficulties during the transition to emerging adulthood. This association was examined from a within-person perspective, examining whether fluctuations in social media behaviors and internalizing difficulties were reciprocally associated. A random intercept cross-lagged panel model approach was used to assess the potential directionality of these effects. Guided by both the empirical and theoretical literature on social media use and mental health, the current study hypothesized that interactive use would be associated with reductions in internalizing difficulties over time, whereas passive reading and viewing on social media would be associated with increases in internalizing difficulties over time. Furthermore, the present research hypothesized that the associations between passive use and interactive communication would exhibit a bidirectional association with internalizing difficulties.

Overall, findings suggested that both models fit the data sufficiently in line with guidelines for Goodness of Fit. Findings from both random-intercept cross-lagged path models highlighted some evidence for the differential associations between interactive and passive reading and viewing social media and internalizing difficulties. Specifically, results revealed that active online communication and internalizing difficulties at the within-person level were only significant at only one-time point (from 2021 to 2022). This association was unidirectional, such that individuals’ interactive social media use was predictive of lower levels of internalizing difficulties in the following wave. In contrast, there were no associations between passive reading and viewing content on social media and internalizing difficulties. Findings at the between-person (random intercept) level revealed that on average, increases in both interactive and passive use were associated with increases in internalizing difficulties.

Our findings diverge from most previous research using random intercept cross-lagged panel models (Boer et al., 2021; Fitzpatrick et al., 2023; Houghton et al., 2018; Marciano et al., 2022), which have established bidirectional associations between total time spent on social media and depression in young people, such that increases in overall time spent were associated with increases in depression. While prior evidence suggests gender differences in the relationship between social media use and mental health, our results did not reveal such effects. Consistent with Marciano et al. (2022), the current study revealed no associations between the random intercepts of passive and interactive communication and gender. Although some studies have reported mixed gender effects in cross-lagged models (Fitzpatrick et al., 2023; Houghton et al., 2018), the current disaggregation of social media behaviors into interactive communication and passive scrolling may explain these differing findings, as it provides a more focused investigation of specific behaviors. However, it remains important to acknowledge that this specificity still does not capture the breadth and variety of online activities relevant to youth mental health.

Associations Between Interactive Social Media Use and Youth Mental Health

Inconsistent findings were revealed regarding interactive social media and youth mental health difficulties, such that there was evidence for one unidirectional association between interactive communication and mental health between 2021 and 2022. Further, and although not the primary aim of the current research, a unidirectional shorter-term (one-month lag) association between interactive use and mental health in 2020 was present. The inconsistent associations in cross-lagged associations were not expected. Previous literature has highlighted the positive mental health effects of chatting online in adolescence and emerging adulthood (Karsay et al., 2023; Wenninger et al., 2019). Moreover, this positive effect has been observed to hold in the inverse direction, indicating that those with pre-existing internalizing difficulties may derive even more benefit from online communication (Dolev-Cohen & Barak, 2013).

There are several potential explanations for these inconsistent findings. For example, and not captured in the current research, the positive effects of online communication may be contingent upon the modality chosen. When compared to asynchronous modes of communication, synchronous (real-time) communication has been shown to have a positive impact on youth self-esteem (Gonzales, 2014) and friendship quality (Desjarlais & Joseph, 2017). The beneficial impact of reciprocity in communication on social media has also been demonstrated by Wenninger et al., (2019), where activities that promoted reciprocity, such as chatting, giving, and receiving feedback on social media sites were associated with increased well-being in adolescents. Further research should endeavor to disentangle the potential effects of synchronous versus asynchronous communication and youth mental health over time.

Similarly, conversation quality may account for the inconsistent within-person associations between interactive social media use and internalizing difficulties over time in this study. The quality of peer relationships is a critical developmental factor known to influence overall youth well-being (McMahon et al., 2020; Victor et al., 2019). Negative social interactions, peer victimization, or perceived peer rejection offline have long been linked with decreased youth well-being and increased depression and anxiety (Lopez & DuBois, 2005; Nepon et al., 2021; Platt et al., 2013). This pattern extends to online interactions. Specifically, a study examining the effects of both the frequency and quality, i.e., self-reports of positive or negative social interactions on social media, found that self-reported interaction quality, rather than quantity, was associated with depression in emerging adulthood (Davila et al., 2012). Thus, the potential adverse impact of poor interaction quality online may, in certain cases offset the positive aspects of communication on social media, potentially explaining the inconsistent findings in within-person associations over time in the present study.

Associations Between Passive Reading and Viewing Content on Social Media and Youth Mental Health

Contrary to the study hypotheses, passive reading and viewing of content on social media were not related to within-person changes in internalizing difficulties over time. When compared to more interactive social media use, passive behavior has been shown to contribute to depression and anxiety in young people (Aalbers et al., 2019; Frison & Eggermont, 2020; Thorisdottir et al. 2019). Furthermore, there have been numerous indications across the literature that these associations may be bidirectional (Cheng et al., 2023; Wang et al., 2018), with some studies indicating that young people are more likely to passively engage with social media when experiencing reduced wellbeing (Kross et al., 2013). However, current study findings indicate that there was little evidence for these associations. While our study could not capture the specific passive content that young people engaged with, it is important to note that exposure to passive social media content may not be wholly negative. Indeed, many young people passively use social media as a source of entertainment and a place to consume enjoyable content (Vaingankar et al., 2022).

Discrepancies in Within-and Between-Person Effects

Our results show that young people who spend more time engaging in interactive two-way communication and passively engaging with social media have more internalizing difficulties on average, and that individual changes in young people’s mental health and online behaviors were not associated. The emphasis on distinguishing within-person and between-person effects aligns with previous reports by Hamaker et al. (2015) and Zhang & Wang (2014). It may be suggested that the small but significant findings from previous cross-sectional and longitudinal literature implicating social media in the increases in internalizing difficulties may be accounted for in part by this focus on between-person effects. Coyne et al’s (2020) similar findings, that is, no observed associations between overall social media use and mental health underscore the considerable within-person variability in these associations, hinting at potential content-related or contextual confounding factors contributing to the between-person effects.

Strengths, Limitations, and Future Directions

The current study findings should be interpreted considering some limitations. Firstly, all endogenous variables captured in the current study were self-reported. Considering previous research findings that people tend to underreport their social media usage (Coyne et al., 2020), it cannot be assured that the young people included in the current sample reported their average usage without some error.

Considering the limited and non-significant findings, it is important to state that the time lags within the current study (11 months, 12 months, and unidirectional associations of 1 month) may produce differing results to studies using shorter and momentary measurement occasions in intensive longitudinal research designs. To fully capture intra-individual differences, research should also consider the potential for daily or moment-to-moment state-like fluctuations in psychological processes and interactive and passive social media use studies through novel approaches such as digital phenotyping and ecological momentary assessments. Potentially intensive longitudinal methods such as these have been shown to better capture subtle changes in within-person associations (Bolger & Laurenceau, 2013). Consideration of the specific modality used for online communication, and the associated quality of interactions would progress understanding in this area. Similarly, future investigations might also consider enhancing our understanding of ‘passive’ use by assessing the longitudinal impact of specific passive content on social media such as viewing others’ photos, scrolling through newsfeeds, watching videos, or observing online discussions without direct involvement and whether different components of passive use are implicated in longitudinal associations with mental health outcomes.

Finally, the temporal variations in associations between interactive and passive social media use and mental health observed in the current findings may be attributed to, at least in part, the dynamic context shaped by the timing of data collection in relation to Covid-19 lockdowns. It cannot be denied that the evolving circumstances during this period may have influenced individual patterns of social media use, as well as individual fluctuations in internalizing difficulties and may have contributed to these differences in within and between person associations. For example, research has reported that while digital communication was associated with lower levels of internalizing difficulties during Covid-19, the association between face-to-face communication and lower levels of internalizing difficulties was stronger during this time (Stieger et al., 2023).

Notwithstanding the limitations, the present study contributes to research on social media and youth mental health by examining two distinct behaviors, interactive use and passive reading and viewing content rather than the predominant focus on overall time spent using social media. In addition, another strength of this research is the disaggregation of findings into between within- and between-person effects. Consistent with Coyne et al., (2020), this study reveals that there are indeed potential differences in these effects. Regardless, the current results indicate that the associations between distinct social media behaviors and internalizing difficulties of young people are limited.

Conclusion

The existing literature on social media use and youth mental health has largely relied on cross-sectional data, with limited longitudinal evidence examining the reciprocal associations between specific social media behaviors and internalizing outcomes. This study addressed these gaps by longitudinally investigating the bidirectional associations between interactive communication, passive content consumption on social media, and internalizing difficulties during the transition to emerging adulthood. The findings revealed contrary to previous research, that neither interactive communication nor passive content consumption demonstrated consistent within-person associations with internalizing difficulties over time. Notably, while increases in interactive communication were associated with to reductions in internalizing difficulties at one time point, passive social media use showed no significance across any time points. These results underscore the need to move beyond generalized social media use to better understand the potential nuanced effects of different social media behaviors on youth mental health. This study contributes to a more refined understanding of how specific social media activities are associated with youth mental health, highlighting the importance of distinguishing between various online behaviors when assessing their impacts.

Supplementary Information

Supplementary Materials (24.1KB, docx)

Acknowledgements

The authors thank the individuals who provided their data as part of the LISS panel.

Biographies

Maria Tibbs

is a PhD candidate at the School of Psychology, University College Dublin, and a researcher at Jigsaw—the National Centre for Youth Mental Health, with research interests in digital interventions and digital risk and resilience for adolescents.

Sonya Deschénes

is an Assistant Professor at the School of Psychology, University College Dublin. Her major research interests focus on mental health epidemiology, and the intersection of health psychology and epidemiology, including the biopsychosocial mechanisms of the associations between physical and mental health conditions.

Peter van der Velden

is a Professor at Tranzo, the Scientific Center for Care and Wellbeing, Tilburg University, and a senior researcher at Centerdata. He specializes in research related to stress, experiences of collective and individual trauma, and mental health.

Amanda Fitzgerald

is an Associate Professor at the School of Psychology, University College Dublin and co-founder of the UCD Youth Mental Health Lab. Her research focuses on risk and protective factors that shape young people’s mental health and service innovations in mental health delivery.

Author’s Contributions

M.T. conceived the study, performed and interpreted statistical analyses, and drafted the manuscript; S.D. advised on the design, statistical analyses, and interpretation of findings and contributed to drafting the manuscript; P.V.D.V. advised on the design, statistical analyses, and interpretation of findings and contributed to drafting the manuscript; A.F. advised on the design and contributed to drafting the manuscript. All authors read and approved the final version of this manuscript.

Funding

The preparation of this manuscript was supported by the COORDINATE Transnational Access Visits (TAV) and the Irish Research Council Employment Based Postgraduate Program (Project No: EBPPG/2020/91), both awarded to M.T.

Data Sharing Declaration

The datasets used in the current study are open-access and available upon request on the Longitudinal Internet Studies for the Social Sciences website: https://www.lissdata.nl/. Lavaan codes used in the current study are available from the corresponding author upon reasonable request.

Compliance with Ethical Standards

Conflict of Interest

The authors declare no competing interests.

Compliance with Ethical Standards

This study was performed in line with the principles of the Declaration of Helsinki. Informed consent was obtained from all respondents included in the LISS panel. As the study uses secondary analyses of panel data, additional ethical approval was not sought.

Informed Consent

All respondents provided informed consent to take part in the Longitudinal Internet Studies for the Social Sciences.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

The online version contains supplementary material available at 10.1007/s10964-024-02093-5.

References

  1. Aalbers, G., McNally, R. J., Heeren, A., de Wit, S., & Fried, E. I. (2019). Social media and depression symptoms: A network perspective. Journal of Experimental Psychology: General, 148(8), 1454–1462. 10.1037/xge0000528. [DOI] [PubMed] [Google Scholar]
  2. Andalibi, N., Ozturk, P., & Forte, A. (2017). Sensitive self-disclosures, responses, and social support on Instagram: The case of# depression. In Proceedings of the 2017 ACM conference on computer supported cooperative work and social computing (pp. 1485-1500).
  3. Boer, M., Stevens, G. W. J. M., Finkenauer, C., de Looze, M. E., & van den Eijnden, R. J. J. M. (2021). Social media use intensity, social media use problems, and mental health among adolescents: Investigating directionality and mediating processes. Computers in Human Behavior, 116, 106645 10.1016/j.chb.2020.106645. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bolger, N. (2013). Intensive longitudinal methods: An introduction to diary and experience sampling research. The Guilford Express.
  5. Bonetti, L., Campbell, M. A., & Gilmore, L. (2010). The relationship of loneliness and social anxiety with children's and adolescents' online communication. Cyberpsychology, behavior, and social networking, 13(3), 279–285. [DOI] [PubMed] [Google Scholar]
  6. Cheng, W., Nguyen, D. N., & Nguyen, P. N. T. (2023). The association between passive social network usage and depression/negative emotions with envy as a mediator. Scientific Reports, 13(1), 1 10.1038/s41598-023-37185-y. Article. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Collishaw, S. (2015). Annual Research Review: Secular trends in child and adolescent mental health. Journal of Child Psychology and Psychiatry, 56(3), 370–393. 10.1111/jcpp.12372. [DOI] [PubMed] [Google Scholar]
  8. Coyne, S. M., Rogers, A. A., Zurcher, J. D., Stockdale, L., & Booth, M. (2020). Does time spent using social media impact mental health?: An eight year longitudinal study. Computers in human behavior, 104, 106160. [Google Scholar]
  9. Davis, K. (2012). Friendship 2.0: Adolescents' experiences of belonging and self-disclosure online. Journal of adolescence, 35(6), 1527–1536. [DOI] [PubMed] [Google Scholar]
  10. Desjarlais, M., & Joseph, J. J. (2017). Socially interactive and passive technologies enhance friendship quality: An investigation of the mediating roles of online and offline self-disclosure. Cyberpsychology, Behavior, and Social Networking, 20(5), 286–291. 10.1089/cyber.2016.0363. [DOI] [PubMed] [Google Scholar]
  11. Dolev-Cohen, M., & Barak, A. (2013). Adolescents’ use of Instant Messaging as a means of emotional relief. Computers in Human Behavior, 29(1), 58–63. 10.1016/j.chb.2012.07.016. [Google Scholar]
  12. Dunn, T. J., Baguley, T., & Brunsden, V. (2014). From alpha to omega: A practical solution to the pervasive problem of internal consistency estimation. British Journal of Psychology, 105(3), 399–412. 10.1111/bjop.12046. [DOI] [PubMed] [Google Scholar]
  13. Dworkin, J., Rudi, J. H., & Hessel, H. (2018). The state of family research and social media. Journal of Family Theory & Review, 10(4), 796–813. [Google Scholar]
  14. Enders, C., & Bandalos, D. (2001). The relative performance of full information maximum likelihood estimation for missing data in structural equation models. Structural Equation Modeling: A Multidisciplinary Journal, 8(3), 430–457. 10.1207/S15328007SEM0803_5. [Google Scholar]
  15. Fitzpatrick, C., Lemieux, A., Smith, J., West, G. L., Bohbot, V., & Asbridge, M. (2023). Is adolescent internet use a risk factor for the development of depression symptoms or vice-versa? Psychological Medicine, 53(14), 6773–6779. 10.1017/S0033291723000284. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Frison, E., & Eggermont, S. (2020). Toward an integrated and differential approach to the relationships between loneliness, different types of Facebook use, and adolescents’ depressed mood. Communication Research, 47(5), 701–728. 10.1177/0093650215617506. [Google Scholar]
  17. Frison, E., Bastin, M., Bijttebier, P., & Eggermont, S. (2019). Helpful or harmful? The different relationships between private Facebook interactions and adolescents’ depressive symptoms. Media Psychology, 22(2), 244–272. [Google Scholar]
  18. Fusar-Poli, P. (2019). Integrated mental health services for the developmental period (0 to 25 years): A critical review of the evidence. Frontiers in Psychiatry, 10. 10.3389/fpsyt.2019.00355. [DOI] [PMC free article] [PubMed]
  19. Gonzales, A. L. (2014). Text-based communication influences self-esteem more than face-to-face or cellphone communication. Computers in Human Behavior, 39, 197–203. 10.1016/j.chb.2014.07.026. [Google Scholar]
  20. Hall, J. A., Kearney, M. W., & Xing, C. (2019). Two tests of social displacement through social media use. Information. Communication & Society, 22(10), 1396–1413. 10.1080/1369118X.2018.1430162. [Google Scholar]
  21. Hamaker, E. L., Kuiper, R. M., & Grasman, R. P. P. P. (2015). A critique of the cross-lagged panel model. Psychological Methods, 20(1), 102–116. 10.1037/a0038889. [DOI] [PubMed] [Google Scholar]
  22. Houghton, S., Lawrence, D., Hunter, S. C., Rosenberg, M., Zadow, C., Wood, L., & Shilton, T. (2018). Reciprocal relationships between trajectories of depressive symptoms and screen media use during adolescence. Journal of Youth and Adolescence, 47(11), 2453–2467. 10.1007/s10964-018-0901-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. 10.1080/10705519909540118. [Google Scholar]
  24. Huang, H.-Y. (2016). Examining the beneficial effects of individual’s self-disclosure on the social network site. Computers in Human Behavior, 57, 122–132. 10.1016/j.chb.2015.12.030. [Google Scholar]
  25. Jensen, M., George, M. J., Russell, M. R., & Odgers, C. L. (2019). Young adolescents’ digital technology use and mental health symptoms: Little evidence of longitudinal or daily linkages. Clinical Psychological Science, 7(6), 1416–1433. 10.1177/2167702619859336. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Karsay, K., Matthes, J., Schmuck, D., & Ecklebe, S. (2023). Messaging, posting, and browsing: A mobile experience sampling study investigating youth’s social media use, affective well-being, and loneliness. Social Science Computer Review, 41(4), 1493–1513. 10.1177/08944393211058308. [Google Scholar]
  27. Kaye, L. K. (2021). Exploring the “socialness” of social media. Computers in Human Behavior Reports, 3, 100083. [Google Scholar]
  28. Keles, B., McCrae, N., & Grealish, A. (2019). A systematic review: the influence of social media on depression, anxiety and psychological distress in adolescents. International Journal of Adolescence and Youth, 25(1), 79–93. 10.1080/02673843.2019.1590851. [Google Scholar]
  29. Kelly, Y., Zilanawala, A., Booker, C., & Sacker, A. (2018). Social media use and adolescent mental health: Findings from the UK Millennium Cohort Study. EClinicalMedicine, 6, 59–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kietzmann, J. H., Hermkens, K., McCarthy, I. P., & Silvestre, B. S. (2011). Social media? Get serious! Understanding the functional building blocks of social media. Business Horizons, 54(3), 241–251. 10.1016/j.bushor.2011.01.005. [Google Scholar]
  31. Kline, R. B. (2016). Principles and practice of structural equation modeling. Guilford Publications.
  32. Kraut, R., Patterson, M., Lundmark, V., Kiesler, S., Mukophadhyay, T., & Scherlis, W. (1998). Internet paradox: A social technology that reduces social involvement and psychological well-being? American Psychologist, 53(9), 1017–1031. 10.1037/0003-066X.53.9.1017. [DOI] [PubMed] [Google Scholar]
  33. Kross, E., Verduyn, P., Demiralp, E., Park, J., Lee, D. S., Lin, N., Shablack, H., Jonides, J., & Ybarra, O. (2013). Facebook Use Predicts Declines in Subjective Well-Being in Young Adults. PLOS ONE, 8(8), e69841 10.1371/journal.pone.0069841. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Lee, D. K. L., & Borah, P. (2020). Self-presentation on Instagram and friendship development among young adults: A moderated mediation model of media richness, perceived functionality, and openness. Computers in Human Behavior, 103, 57–66. 10.1016/j.chb.2019.09.017. [Google Scholar]
  35. Lee, D. Y., Harring, J. R., & Stapleton, L. M. (2019). Comparing Methods for Addressing Missingness in Longitudinal Modeling of Panel Data. The Journal of Experimental Education, 87(4), 596–615. 10.1080/00220973.2018.1520683. [Google Scholar]
  36. Leo, K., Kewitz, S., Wartberg, L., & Lindenberg, K. (2021). Depression and social anxiety predict internet use disorder symptoms in children and adolescents at 12-month follow-up: Results from a longitudinal study. Frontiers in psychology, 12, 787162. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Li, J. B., Mo, P. K., Lau, J. T., Su, X. F., Zhang, X., Wu, A. M.,... & Chen, Y. X. (2018). Online social networking addiction and depression: The results from a large-scale prospective cohort study in Chinese adolescents. Journal of behavioral addictions, 7(3), 686–696. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Livingstone, S. (2014). Developing social media literacy: How children learn to interpret risky opportunities on social network sites. Communications, 39(3), 283–303. [Google Scholar]
  39. Long, E., Gardani, M., McCann, M., Sweeting, H., Tranmer, M., & Moore, L. (2020). Mental health disorders and adolescent peer relationships. Social Science & Medicine, 253, 112973. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Lopez, C., & DuBois, D. L. (2005). Peer Victimization and Rejection: Investigation of an Integrative Model of Effects on Emotional, Behavioral, and Academic Adjustment in Early Adolescence. Journal of Clinical Child & Adolescent Psychology, 34(1), 25–36. 10.1207/s15374424jccp3401_3. [DOI] [PubMed] [Google Scholar]
  41. Marciano, L., Schulz, P. J., & Camerini, A.-L. (2022). How do depression, duration of internet use and social connection in adolescence influence each other over time? An extension of the RI-CLPM including contextual factors. Computers in Human Behavior, 136, 107390 10.1016/j.chb.2022.107390. [Google Scholar]
  42. McMahon, G., Creaven, A. M., & Gallagher, S. (2020). Stressful life events and adolescent well‐being: The role of parent and peer relationships. Stress and Health, 36(3), 299–310. [DOI] [PubMed] [Google Scholar]
  43. Mulder, J. D., & Hamaker, E. L. (2021). Three Extensions of the Random Intercept Cross-Lagged Panel Model. Structural Equation Modeling: A Multidisciplinary Journal, 28(4), 638–648. 10.1080/10705511.2020.1784738. [Google Scholar]
  44. Nepon, T., Pepler, D. J., Craig, W. M., Connolly, J., & Flett, G. L. (2021). A Longitudinal Analysis of Peer Victimization, Self-Esteem, and Rejection Sensitivity in Mental Health and Substance Use Among Adolescents. International Journal of Mental Health and Addiction, 19(4), 1135–1148. 10.1007/s11469-019-00215-w. [Google Scholar]
  45. Nesi, J., Miller, A. B., & Prinstein, M. J. (2017). Adolescents’ depressive symptoms and subsequent technology-based interpersonal behaviors: A multi-wave study. Journal of Applied Developmental Psychology, 51, 12–19. 10.1016/j.appdev.2017.02.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Nesi, J., Choukas-Bradley, S., & Prinstein, M. J. (2018a). Transformation of adolescent peer relations in the social media context: Part 1—A theoretical framework and application to dyadic peer relationships. Clinical child and family psychology review, 21, 267–294. 10.1007/s10567-018-0261-x. [DOI] [PMC free article] [PubMed]
  47. : Nesi, J., Choukas-Bradley, S., & Prinstein, M. J. (2018b). Transformation of adolescent peer relations in the social media context: Part 2—application to peer group processes and future directions for research. Clinical child and family psychology review, 21, 295–319. 10.1007/s10567-018-0262-9. [DOI] [PMC free article] [PubMed]
  48. Niederhoffer, K. G., & Pennebaker, J. W. (2009). Sharing One’s Story: On the Benefits of Writing or Talking About Emotional Experience. In S. J. Lopez & C. R. Snyder (Eds.), The Oxford Handbook of Positive Psychology (p. 0). Oxford University Press. 10.1093/oxfordhb/9780195187243.013.0059.
  49. Oberst, U., Wegmann, E., Stodt, B., Brand, M., & Chamarro, A. (2017). Negative consequences from heavy social networking in adolescents: The mediating role of fear of missing out. Journal of adolescence, 55, 51–60. [DOI] [PubMed] [Google Scholar]
  50. Ogden, T., & Hagen, K. A. (2018). Adolescent mental health: Prevention and intervention. Routledge.
  51. Orsolini, L., Volpe, U., & Albert, U., et al. (2022). Use of social network as a coping strategy for depression among young people during the COVID-19 lockdown: findings from the COMET collaborative study. Ann Gen Psychiatry, 21, 44 10.1186/s12991-022-00419-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Pennebaker, J. W. (1997). Writing About Emotional Experiences as a Therapeutic Process. Psychological Science, 8(3), 162–166. 10.1111/j.1467-9280.1997.tb00403.x. [Google Scholar]
  53. Pérez-Torres, V. (2024). Social media: a digital social mirror for identity development during adolescence. Curr Psychol, 43, 22170–22180. 10.1007/s12144-024-05980-z. [Google Scholar]
  54. Pew Research Centre. (2022). Teens, Social Media, and Technology. https://www.pewresearch.org/internet/2022/08/10/teens-social-media-and-technology-2022/.
  55. Platt, B., Kadosh, K. C., & Lau, J. Y. F. (2013). The Role of Peer Rejection in Adolescent Depression. Depression and Anxiety, 30(9), 809–821. 10.1002/da.22120. [DOI] [PubMed] [Google Scholar]
  56. Prescott, J., Hanley, T. & Ujhelyi, K. (2017). Peer communication in online mental health forums for young people: directional and nondirectional support. JMIR Ment Health, 4(3), e29. 10.2196/mental.6921. [DOI] [PMC free article] [PubMed]
  57. Primack, B. A., Silk, J. S., & DeLozier, C. R., et al. (2011). Using Ecological Momentary Assessment to Determine Media Use by Individuals With and Without Major Depressive Disorder. Arch Pediatr Adolesc Med, 165(4), 360–365. 10.1001/archpediatrics.2011.27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. R Core Team. (2021). R: A language and environment for statistical computing. [Computer software]. R Foundation for Statistical Computing. https://www.R-project.org/.
  59. Riehm, K. E., Feder, K. A., Tormohlen, K. N., Crum, R. M., Young, A. S., Green, K. M., Pacek, L. R., La Flair, L. N., & Mojtabai, R. (2019). Associations Between Time Spent Using Social Media and Internalizing and Externalizing Problems Among US Youth. JAMA Psychiatry, 76(12), 1266–1273. 10.1001/jamapsychiatry.2019.2325. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Rosseel, Y. (2012). lavaan: An R Package for Structural Equation Modeling. Journal of Statistical Software, 48, 1–36. 10.18637/jss.v048.i02. [Google Scholar]
  61. Satorra, A., & Bentler, P. M. (2001). A scaled difference chi-square test statistic for moment structure analysis. Psychometrika, 66(4), 507–514. 10.1007/BF02296192. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Scherpenzeel, A. (2011). Data collection in a probability-based internet panel: how the LISS panel was built and how it can be used. Bulletin of Sociological Methodology/Bulletin de Méthodologie Sociologique, 109(1), 56–61. [Google Scholar]
  63. Scott, J., Scott, E. M., Hermens, D. F., Naismith, S. L., Guastella, A. J., White, D., Whitwell, B., Lagopoulos, J., & Hickie, I. B. (2014). Functional impairment in adolescents and young adults with emerging mood disorders. The British Journal of Psychiatry, 205(5), 362–368. 10.1192/bjp.bp.113.134262. [DOI] [PubMed] [Google Scholar]
  64. Solmi, M., Radua, J., Olivola, M., Croce, E., Soardo, L., Salazar de Pablo, G., Il Shin, J., Kirkbride, J. B., Jones, P., Kim, J. H., Kim, J. Y., Carvalho, A. F., Seeman, M. V., Correll, C. U., & Fusar-Poli, P. (2022). Age at onset of mental disorders worldwide: Large-scale meta-analysis of 192 epidemiological studies. Molecular Psychiatry, 27(1), 1 10.1038/s41380-021-01161-7. Article. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Stieger, S., Lewetz, D., & Willinger, D. (2023). Face-to-face more important than digital communication for mental health during the pandemic. Scientific Reports, 13(1), 8022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Svensson, R., Johnson, B., & Olsson, A. (2022). Does gender matter? The association between different digital media activities and adolescent well-being. BMC Public Health, 22, 273 10.1186/s12889-022-12670-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Tang, S., Werner-Seidler, A., Torok, M., Mackinnon, A. J., & Christensen, H. (2021). The relationship between screen time and mental health in young people: A systematic review of longitudinal studies. Clinical psychology review, 86, 102021. [DOI] [PubMed] [Google Scholar]
  68. Thorisdottir, I. E., Sigurvinsdottir, R., Asgeirsdottir, B. B., Allegrante, J. P., & Sigfusdottir, I. D. (2019). Active and Passive Social Media Use and Symptoms of Anxiety and Depressed Mood Among Icelandic Adolescents. Cyberpsychology, Behavior, and Social Networking, 22(8), 535–542. 10.1089/cyber.2019.0079. [DOI] [PubMed] [Google Scholar]
  69. Tian, Y., Bian, Y., Han, P., Gao, F., & Wang, P. (2017). Associations between psychosocial factors and generalized pathological internet use in Chinese university students: A longitudinal cross-lagged analysis. Computers in Human Behavior, 72, 178–188. 10.1016/j.chb.2017.02.048. [DOI] [PubMed] [Google Scholar]
  70. Toma, C. L. (2022). Online dating and psychological wellbeing: A social compensation perspective. Current Opinion in Psychology, 46, 101331. [DOI] [PubMed] [Google Scholar]
  71. Twenge, J. M., & Martin, G. N. (2020). Gender differences in associations between digital media use and psychological well-being: Evidence from three large datasets. Journal of adolescence, 79, 91–102. [DOI] [PubMed] [Google Scholar]
  72. Twenge, J. M., Spitzberg, B. H., & Campbell, W. K. (2019). Less in-person social interaction with peers among U.S. adolescents in the 21st century and links to loneliness. Journal of Social and Personal Relationships, 36(6), 1892–1913. 10.1177/0265407519836170. [Google Scholar]
  73. Twenge, J. M., Joiner, T. E., Rogers, M. L., & Martin, G. N. (2018). Increases in Depressive Symptoms, Suicide-Related Outcomes, and Suicide Rates Among U.S. Adolescents After 2010 and Links to Increased New Media Screen Time. Clinical. Psychological Science, 6(1), 3–17. 10.1177/2167702617723376. [Google Scholar]
  74. Vaingankar, J. A., Van Dam, R. M., Samari, E., Chang, S., Seow, E., Chua, Y. C.,... & Subramaniam, M. (2022). Social media–driven routes to positive mental health among youth: Qualitative enquiry and concept mapping study. JMIR pediatrics and parenting, 5(1), e32758. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Valkenburg, P. M., & Peter, J. (2009). Social Consequences of the Internet for Adolescents: A Decade of Research. Current Directions in Psychological Science, 18(1), 1–5. 10.1111/j.1467-8721.2009.01595.x. [Google Scholar]
  76. Valkenburg, P. M., Meier, A., & Beyens, I. (2022). Social media use and its impact on adolescent mental health: An umbrella review of the evidence. Current opinion in psychology, 44, 58–68. 10.1016/j.copsyc.2021.08.017. [DOI] [PubMed] [Google Scholar]
  77. Velden, P. G., van der, Setti, I., Das, M., & van der Meulen, E. (2019). Does social networking sites use predict mental health and sleep problems when prior problems and loneliness are taken into account? A population-based prospective study. Computers in human behavior, 93, 200–209. 10.1016/j.chb.2018.11.04. [Google Scholar]
  78. Verduyn, P., Ybarra, O., Résibois, M., Jonides, J., & Kross, E. (2017). Do Social Network Sites Enhance or Undermine Subjective Well-Being? A Critical Review. Social Issues and Policy Review, 11(1), 274–302. 10.1111/sipr.12033. [Google Scholar]
  79. Victor, S. E., Hipwell, A. E., Stepp, S. D., & Scott, L. N. (2019). Parent and peer relationships as longitudinal predictors of adolescent non-suicidal self-injury onset. Child and adolescent psychiatry and mental health, 13, 1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Wang, H. Z., Yang, T. T., Gaskin, J., & Wang, J. L. (2019). The longitudinal association between passive social networking site usage and depressive symptoms: The mediating role of envy and moderating role of life satisfaction. Journal of Social and Clinical Psychology, 38(3), 181–199. [Google Scholar]
  81. Wang, J.-L., Gaskin, J., Rost, D. H., & Gentile, D. A. (2018). The Reciprocal Relationship Between Passive Social Networking Site (SNS) Usage and Users’ Subjective Well-Being. Social Science Computer Review, 36(5), 511–522. 10.1177/0894439317721981. [Google Scholar]
  82. Ware, J. E., & Sherbourne, C. D. (1992). The MOS 36-Item Short-Form Health Survey (SF-36): I. Conceptual Framework and Item Selection. Medical Care, 30(6), 473–483. [PubMed] [Google Scholar]
  83. Weidman, A. C., Fernandez, K. C., Levinson, C. A., Augustine, A. A., Larsen, R. J., & Rodebaugh, T. L. (2012). Compensatory internet use among individuals higher in social anxiety and its implications for well-being. Personality and individual differences, 53(3), 191–195. [DOI] [PMC free article] [PubMed] [Google Scholar]
  84. Wenninger, H., Krasnova, H., & Buxmann, P. (2019). Understanding the role of social networking sites in the subjective well-being of users: A diary study. European Journal of Information Systems, 28(2), 126–148. 10.1080/0960085X.2018.1496883. [Google Scholar]
  85. Wiens, K., Bhattarai, A., Pedram, P., Dores, A., Williams, J., Bulloch, A., & Patten, S. (2020). A growing need for youth mental health services in Canada: Examining trends in youth mental health from 2011 to 2018. Epidemiology and Psychiatric Sciences, 29, e115 10.1017/S2045796020000281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  86. Van Zalk, M. H., Branje, S. J., Denissen, J., Van Aken, M. A., & Meeus, W. H. (2011). Who benefits from chatting, and why? The roles of extraversion and supportiveness in online chatting and emotional adjustment. Personality and Social Psychology Bulletin, 37(9), 1202–1215. [DOI] [PubMed] [Google Scholar]
  87. Zapf, H., Boettcher, J., Haukeland, Y., Orm, S., Coslar, S., & Fjermestad, K. (2024). A systematic review of the association between parent‐child communication and adolescent mental health. JCPP advances, 4(2), e12205. [DOI] [PMC free article] [PubMed] [Google Scholar]
  88. Zhang, Q., & Wang, L. (2014). Aggregating and testing intra-individual correlations: Methods and comparisons. Multivariate Behavioral Research, 49(2), 130–148. 10.1080/00273171.2013.870877. [DOI] [PubMed] [Google Scholar]
  89. Zinbarg, R. E., Revelle, W., Yovel, I., & Li, W. (2005). Cronbach’s α, Revelle’s β, and Mcdonald’s ωH: Their relations with each other and two alternative conceptualizations of reliability. Psychometrika, 70(1), 123–133. 10.1007/s11336-003-0974-7. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Materials (24.1KB, docx)

Articles from Journal of Youth and Adolescence are provided here courtesy of Springer

RESOURCES