In this issue of International Psychogeriatrics, Lei et al. (Reference Lei, Matovic, Leung, Viju and Wuthrich2023) conducted a systematic review on the relationships between social media use and different psychosocial outcomes in older adults. The majority of papers were cross-sectional (59%), and many found positive associations between social media use and improved mental and social health, though study quality was questionable. Longitudinal and interventional studies found mixed results that differed by outcome. While cross-sectional papers generally found that less loneliness was associated with greater social media use, longitudinal studies were mixed (n = 1 of 2 were positive) as were interventional studies (n = 4 of 7 were positive). A similar trend was found for depression, where cross-sectional studies showed a strong association yet longitudinal (n = 2 of 4 were positive) and interventional studies (n = 4 of 6 were positive) revealed mixed findings. Fewer meaningful results were found for quality of life, anxiety, and social connectedness due to the low number of longitudinal studies and interventions on the topics. Importantly, the most successful social media interventions relied on video chatting/videoconferencing with family and friends to mimic face-to-face communication.

This review adds nuance to the literature that has long-focused on problematic social media use and its negative effects before and during the COVID-19 pandemic. Problematic social media use transiently increased during the COVID-19 pandemic due to restrictions in social activities (Casale et al., Reference Casale, Akbari, Seydavi, Benucci and Fioravanti2023) and has historically been a source of anxiety and depression in adolescents and young adults (Lee et al., Reference Lee, Jeon, Kang, Shin, Jung and Jung2022). Some studies report a dose-response relationship between problematic social media use and affective symptomatology, with a larger effect among girls and young women (Liu et al., Reference Liu, Kamper-DeMarco, Zhang, Xiao, Dong and Xue2022). However, a recent umbrella review on adolescent social media use and mental health found that some meta-analyses and systematic reviews reported weak and inconsistent findings while others reported substantial and deleterious effects, though most papers had concern with study design and heterogeneity (Valkenburg et al., Reference Valkenburg, Meier and Beyens2022). These themes mirror what Lei et al. (Reference Lei, Matovic, Leung, Viju and Wuthrich2023) found in the older adult population.

Lei et al. (Reference Lei, Matovic, Leung, Viju and Wuthrich2023) found that longitudinal and interventional studies focused on the review topic were scarce. Contrary to cross-sectional empirical findings, where social media use was often found to be a protective factor of adverse mental health outcomes, longitudinal studies suggest there could be null or even negative impacts. Although causal inference could not be achieved, longitudinal data compare changes within-person over time, while cross-sectional data contrast the associations of social media use and psychosocial outcomes at the between-person level (user vs. non-users or more frequent users vs. less frequent users). Older adults who choose to learn and use social media could have better mental health status originally. Future longitudinal and interventional studies could examine baseline differences in psychosocial well-being and potential bidirectional longitudinal relationships with social media use. Furthermore, Lei et al. (Reference Lei, Matovic, Leung, Viju and Wuthrich2023) found that social contact and competence mediated the associations between social media use and psychosocial outcomes in older adults, suggesting that social media use could enhance psychosocial well-being by promoting interpersonal relationships and improving self-efficacy. The underlying mechanisms are worth further exploration and can inform the development of social media-facilitated mental health interventions.

While heterogeneity is most often due to differences in study design, scales used, and populations sampled, it may also represent issues with combining several homogeneous effects from latent subgroups. In other words, the effect of social media use on mental health may not be consistent across groups (e.g., young adults vs older adults; men vs women; those environmentally or genetically predisposed to mental illness vs those without a predisposition) or across outcomes (e.g., depression vs anxiety vs loneliness), leading to inconsistency within the literature. This heterogeneity may be complicated by the fact that social media is used in different ways (e.g., self-comparison vs entertainment vs maintaining social connection), which may characteristically lead to different outcomes. For example, looking at pictures and watching videos on social media may lead to an adverse impact on one’s self-esteem (Saiphoo et al., Reference Saiphoo, Halevi and Vahedi2020) and an increase in risk-taking behaviors (Vannucci et al., Reference Vannucci, Simpson, Gagnon and Ohannessian2020), as users increasingly seek social acceptance and validation from their peers (Liu and Baumeister, Reference Liu and Baumeister2016). However, using social media to connect with existing friends/family by posting and videoconferencing may be beneficial for one’s social connectedness (Balki et al., Reference Balki, Hayes and Holland2022; Liu et al., Reference Liu, Ainsworth and Baumeister2016; Lei et al., Reference Lei, Matovic, Leung, Viju and Wuthrich2023).

Social media use is a unique social determinant of mental health (SDoMH) because of its ubiquitousness (Allen et al., Reference Allen, Balfour, Bell and Marmot2014). It is accessible in every country, it comes in every language, it is increasingly available across economic and social classes, and it allows for individuals to construct their identities and jumpstart their careers. However, it diverges from most SDoMHs (e.g., childhood adversities, domestic violence and intimate partner violence, environmental inequities and exposure to natural disasters, housing and food insecurity, incarceration, migration, poverty, etc.) due to its inconsistent and sometimes unexpected effects on mental health. Most SDoMHs affect mental health uniformly across domains and demographic groups (e.g., adverse impact of poverty on anxiety and depression is considerable, regardless of demographic characteristics) with little divergence. In addition, how one engages with social media is constantly changing – from posting text on public forums decades ago to uploading videos and live-streaming in present day.

As clinicians and researchers continue to grapple with an ever-changing digital world, work will be urgently needed to characterize the many nuanced effects of social media use on mental health, and importantly, how these effects may change by group, by utilization type, and over time.