Available online at www.sciencedirect.

com

ScienceDirect

Review

Social media use and well-being: What we know and

what we need to know
Patti M. Valkenburg

Abstract on average five social media platforms in a comple-

Research into the impact of social media use (SMU) on well- mentary way [1], to interact privately with family
being (e.g., happiness) and ill-being (e.g., depression) has members and friends, and/or to interact publicly with
exploded over the past few years. From 2019 to August 2021, broader audiences of friends, acquaintances, and col-
27 reviews have been published: nine meta-analyses, nine leagues [2]. In parallel with this surging social media use
systematic reviews, and nine narrative reviews, which together (SMU), research into its potential impact on well-being
included hundreds of empirical studies. The aim of this um- (e.g., life satisfaction) and ill-being (e.g., depression)
brella review is to synthesize the results of these meta- has also accumulated dramatically [3]. As recent reviews
analyses and reviews. Even though the meta-analyses are demonstrate [4e6], the past years have witnessed at
supposed to rely on the same evidence base, they yielded least 300 studies on the impact of SMU on well- and
disagreeing associations with well- and ill-being, especially for ill-being.
time spent on SM, active SMU, and passive SMU. This um-
brella review explains why their results disagree, summarizes Together with the exponential increase in empirical
the gaps in the literature, and ends with recommendations for studies, reviews of the impact of SMU on well- and ill-
future research. being have also surged in the past few years. Because
this rapidly expanding research output makes it ever
Addresses
more difficult for researchers to keep track of it, an up-
Amsterdam School of Communication Research, University of
Amsterdam, the Netherlands to-date umbrella review of this literature is necessary
and important. An umbrella review, also called a meta-
Corresponding author: Valkenburg, Patti M (p.m.valkenburg@uva.nl) review, is a synthesis of existing reviews [7]. Three
earlier umbrella reviews have focused on the associa-
tions of SMU with well- and ill-being [3,8,9]. One of
Current Opinion in Psychology 2022, 45:101294
these focused on adolescents, thereby excluding reviews
This review comes from a themed issue on Social Media and Well- on adults [9], and neither of the two others included the
Being
22 reviews on the effects of SMU on well-/ill-being
Edited by Patti Valkenburg, Ine Beyens, Adrian Meier and Mariek published in 2020 and 2021.
Vanden Abeele
For a complete overview see the Issue and the Editorial
In this article, I first outline the search method of this
Available online 20 December 2021 umbrella review as well as the operational definitions of
https://doi.org/10.1016/j.copsyc.2021.12.006 SMU, well-being, and ill-being. To assess “what we
2352-250X/© 2021 The Author(s). Published by Elsevier Ltd. This is an know,” I use the meta-analyses to discuss the associa-
open access article under the CC BY license (http://creativecommons. tions of seven different types of SMU with well- and ill-
org/licenses/by/4.0/). being. The systematic and narrative reviews are used to
complement the meta-analytical results, as well as to
Keywords summarize the identified gaps in the literature and the
Review, Meta-analysis, Facebook, Instagram, Social media, Mental suggestions for future research. To assess “what we
health, Well-being, Depression, Idiographic approach, Social compari-
need to know,” the article ends with some general
son, Problematic social media use.
conclusions and three additional recommendations for
future research.
Introduction
It is almost a truism: In the past decade, social media Method and operational definitions
have become a massive and meaningful part of our daily Two coders used the same search strategy and terms as
existence. Individuals, adults and adolescents alike, use applied in the umbrella review of Valkenburg et al. [9],
except for the search terms related to adolescents (as
the current umbrella review excluded reviews focusing
This research was funded by an NWO Spinoza grant. Many thanks to
on adolescents). SMU was operationally defined as the
Wieneke Rollman for assisting with the literature search, and to Loes
Keijsers for her comments on the first draft. active (e.g., posting), passive (e.g., browsing), private,

www.sciencedirect.com Current Opinion in Psychology 2022, 45:101294

2 Social Media and Well-Being

and public uses of platforms such as Facebook, WeChat, use, or the frequency of SM checking). As for aggregated
and WhatsApp. As for the outcomes of SMU, the focus well-/ill-being, one meta-analysis yielded no association
lied on three well-being components (happiness; life [6], and another a small positive association with general
satisfaction; positive affect) and three ill-being compo- SMU [12]. As for aggregated well-being, one meta-
nents (depressive symptoms/depression; anxiety symp- analysis yielded a small negative [6], and another a
toms/anxiety; negative affect). Due to space small positive association with general SMU [15].
restrictions, other components of well-being (e.g., Remarkably though, his latter meta-analysis also re-
eudaimonic well-being) and ill-being (e.g., stress), as ported a small positive association with aggregated ill-
well as risk and resilience factors of well- and ill-being, being [15]. Finally, general SMU was consistently asso-
such as self-esteem, cyberbullying, and body image ciated with higher levels of depression/depressive
concerns, were not considered. symptoms [4,6,12,16] and anxiety [6,12], but, again
surprisingly, also with higher happiness levels [12].
Results: what we know …
The search yielded 27 reviews: nine meta-analyses Active versus passive social media use
[4,6,10e16], nine systematic reviews [5,17e24], and Three meta-analyses compared time spent on active and
nine narrative reviews [25e33] published from January passive SMU, again with highly inconsistent results.
2019 to August 2021. Except for five meta-analyses, Active SMU was not [12] or weakly positively associated
which included studies on both adolescents and [6] with aggregated well-/ill-being. Furthermore, it was
adults, none of the remaining 22 reviews were included not [15] or weakly positively associated [6] with aggre-
in the earlier umbrella review of Valkenburg et al. [9]. gated well-being, but not with aggregated ill-being [15].
Passive SMU, in contrast, was not [6] or negatively [12]
Seven social media activities associated with aggregated well-/ill-being, but not with
As Table 1 shows, some meta-analyses investigated (a) aggregated well-being [6,15] and not with aggregated
general time spent with SM, or (b) time spent with ill-being [15]. Yet, both active and passive SMU were
active and (c) passive SMU. Some (also) focused on associated with higher levels of depression/depressive
specific behaviors and mechanisms afforded by SM, symptoms [6] and anxiety [6].
including (d) the size of one’s SM network, (e) the in-
tensity of SMU, (f) problematic SMU (i.e., an enduring In all, the meta-analyses yielded scant support for both
preoccupation with SM, reflected in a persistent neglect the “active SMU hypothesis” and “passive SMU hy-
of one’s own health and important life areas), and (g) pothesis,” which respectively argue that active SMU
SM-induced social comparison (the tendency to observe elicits likes and support, which results in higher well-
others to assess how we are looking, thinking, or being/lower ill-being. And that passive SMU induces
behaving in comparison with these others) [34]. social comparisons and envy, which leads to lower well-
being/higher ill-being [37]. An elaborate explanation of
Conceptualizations of well-being and ill-being this lack of support can be found in a review by
Two out of the nine meta-analyses [6,12] reversed Valkenburg et al. [24].
different ill-being components (e.g., depression) and
combined them with well-being components (e.g., life Social comparison
satisfaction) to create an “aggregated well-/ill-being” Even though the direct meta-analytic associations of
outcome. Furthermore, five meta-analyses active and passive SMU with well- and ill-being were
[6,10,11,14,15] lumped together components like life inconsistent, two meta-analyses have addressed one part
satisfaction and self-esteem to create an “aggregated of the passive SMU hypothesis, which states that SM-
well-being” outcome. Likewise, they combined com- induced social comparison results in lower well-being/
ponents like depression and loneliness to create an higher ill-being [33]. Indeed, SM-induced social com-
“aggregated ill-being” outcome. However, because parison was associated with lower aggregated well-being
mental health theories agree that a low well-being (e.g., and life satisfaction [14] and higher depression [16]. It
low life satisfaction) does not necessary imply a high ill- must be noted, though, that 78% of SM users report
being (e.g., suffering from depression) and vice versa never feeling worse after comparing themselves to other
[8,35,36], this umbrella review investigated whether the users [38], that only a minority of SM users feel envious
three aggregated outcomes, that is, (a) aggregated well-/ while using SM [39], that they more often feel enjoy-
ill-being, (b) aggregated well-being and (c) aggregated ment [40], and that they sometimes also get inspired
ill-being, led to different associations with SMU. from SM-induced social comparisons [41].

General SMU Network size

The three types of aggregated well- and/or ill-being The results of the two meta-analyses focusing on network
outcomes yielded inconsistent associations with gen- size were rather consistent: The size of one’s SM
eral SMU (also called time spent using SM, general SNS network size was associated with higher aggregated well-

Current Opinion in Psychology 2022, 45:101294 www.sciencedirect.com

 Social Media Use and Well-Being: What We Know and What We Need to Know Valkenburg 3

being [11,15], happiness [11], and life satisfaction [11]. SMU [22,24] at the expense of more fine-grained
It was not [15] or weakly associated [11] with lower measures, such as the purpose of SMU or the type of
aggregated ill-being, and not with depression [11]. communication partners [12,29]. Finally, some reviews
Network size was not related to anxiety [11], but nega- criticized the over-reliance on self-reports [3,5,25,28e
tively to higher social anxiety [11]. However, this latter 30] and called for more objective measures of SMU,
association has mostly been investigated within the social such as log-based data obtained though screen-time
compensation framework [42], in which social anxiety is apps [28,29].
conceptualized as a predictor rather than an outcome of
SMU. Socially anxious people spend more time on SM Discussion: what we need to know …
[42], but particularly more time on passive SMU [22]. The nine meta-analyses in this umbrella review
Obviously, expanding one’s network does not occur via disagreed in their conclusions about the associations of
passive but via interactive SMU, which could explain why different types of SMU with well-being. This particu-
socially anxious users tend to have smaller SM networks larly applied to the time-based predictors and not or less
than their less socially anxious counterparts. to the other predictors. However, despite these in-
consistencies, all meta-analyses yielded pooled associa-
SM intensity and problematic SMU tions that were mostly small (for the time-based
Intensity of SMU refers to a mixture of users’ emotional predictors), occasionally moderate (for problematic
attachment to SM and the extent to which SMU is in- SMU), but never large. The conclusions of the meta-
tegrated into their lives [4]. It is mostly measured with analyses were largely supported by the narrative and
(adaptions of) the Facebook Intensity Scale (FIS) [43]. systematic reviews, which observed comparable gaps in
Even though the FIS was not designed as a measure of the literature and provided comparable suggestions for
problematic SMU, it is highly correlated with prob- future research. I end this article with three additional
lematic SMU (e.g., b = .57) [44], and in some studies, recommendations for future research.
intensity of SMU is even included as an indicator of
problematic SMU [19]. It is no surprise, therefore, that, Recommendation 1: don’t collapse across well- and
most meta-analytic effect sizes for intensity of SMU and ill-being outcomes
problematic SMU are not significantly different [4]. Meta-analyses of the effects of SMU can provide
Intensity of SMU was consistently associated with lower indispensable summaries of the evidence in this vastly
aggregated well-being [6], higher depression/depressive expanding literature [3]. But they can also suffer from
symptoms [4,6] and higher anxiety [6]. the same shortcomings as any other type of study. An
important shortcoming involves their arbitrary choices
Comparatively, problematic SMU was associated with to collapse across distinct well-being and ill-being
lower aggregated well-being [6,10], lower happiness components. In fact, the inconsistencies in effect sizes
[10], and life satisfaction [10]. And it was associated applied particularly to the six meta-analyses that created
with higher aggregated ill-being [10], depression/ well-being, ill-being, and well-/ill-being composites. As
depressive symptoms [4,6,10,13], and anxiety [6,10]. A Table 1 shows, these meta-analyses collapsed across a
likely explanation for the consistent associations of great variety of well- and ill-being components in addi-
problematic SMU with well- and ill-being outcomes may tion to a range of risk and resilience factors of well- and
lie in “construct overlap” [4]. After all, it should be no ill-being, such as envy, stress, self-esteem, self-harm,
surprise that well- and ill-being outcomes correlate with and suicidal ideation.
problematic SMU scales consisting of items like “How
often during the last year .” “did you use SM to escape This lumping together of different well- and ill-being
negative feelings” [45] and “have you become restless or components and their risk-resilience factors hampers
troubled if you were prohibited from using social the validity of the meta-analytic effect sizes for two
media?” [46]. reasons. First, ill-being is not simply the flip side of well-
being [36], as demonstrated, for example, by the posi-
Identified gaps and directions for future research tive meta-analytic associations of time spent on SM with
Seventeen out of the 27 reviews agreed that the evi- both the well-being component “happiness” and the ill-
dence on which their conclusions were based is primarily being component “depression” [12]. Second, compo-
cross-sectional and called for longitudinal and/or exper- nents within well- or ill-being composites also led to
imental studies to determine the causal direction of the different associations with SMU, as confirmed by the
effects of SMU [3,5,12,28,29,47], or for research sheer opposite meta-analytic associations of SM network
designed to investigate why and/or for whom SMU is size with anxiety versus social anxiety [11]. Therefore, a
associated with well- or ill-being [4,5,17,20,31,32]. first crucial step for future research is to avoid lumping
Other reviews observed an over-reliance on measures of together well- and ill-being components that deserve to
time spent on SM [4,6,22,28,29] and active and passive be investigated in their own right [30].

www.sciencedirect.com Current Opinion in Psychology 2022, 45:101294

4 Social Media and Well-Being

Table 1

Associations of different types of social media use (SMU) with indicators of well-being and ill-being.

Study # Articles Operational definitions of Main results

outcomes

Cunningham 62 Depressive symptoms r = .11* Time spent on SNS with depressive symptoms
et al. (2021) r = .09ns Intensity of SNS use with depressive symptoms
r = .29* Problematic SNS use with depressive symptoms
Hancock 256 Well-/ill-being = aggregate r = −.00ns General SMU with well-/Ill-being
et al. (2019) of anxiety, depression, r = .13* General SMU with depression
loneliness, & eudaimonic, r = .11* General SMU with anxiety
hedonic, and relational r = −.03* General SMU with (hedonic) well-being
well-being r = .11* Intensity of SMU with well-/ill-being
Hedonic well- r = .10* Intensity of SMU with depression
being = aggregate of r = .13* Intensity of SMU with anxiety
happiness, positive affect, r = −.16ns Intensity of SMU with (hedonic) well-being
subjective well-being, and r = .06* Active SMU with well-/ill-being
negative affect r = .08* Active SMU with depression
r = .06* Active SMU with anxiety
r = .06* Active SMU with (hedonic) well-being
r = −.03ns Passive SMU with well-/ill-being
r = .07* Passive SMU with depression
r = .21* Passive SMU with anxiety
r = .24ns Passive SMU with (hedonic) well-being
r = −.21* Problematic SMU with well-/ill-being
r = .34* Problematic SMU with depression
r = .32* Problematic SMU with anxiety
r = −.26* Problematic SMU with hedonic well-being
Huang (2020) 123 Well-being = aggregate of r = −.16* Problematic SMU with well-being
life satisfaction, self- r = −.30* Problematic SMU with happiness
esteem, happiness, and r = −.18* Problematic SMU with positive affect
positive affect r = −.11* Problematic SMU with life satisfaction
Ill-being (distress) r = .27* Problematic SMU with ill-being
= aggregate of r = .31* Problematic SMU with depression
depression, anxiety, r = .30* Problematic SMU with anxiety
loneliness, and negative
affect
Huang (2021) 90 Well-being = aggregate of r = .08* Network size with well-being
life satisfaction, self- r = .15* Network size with happiness
esteem, happiness r = .10* Network size with life satisfaction
Ill-being (distress) r = −.06* Network size with ill-being
= aggregate of r = .01ns Network size with depression
depression, loneliness, r = .08ns Network size with anxiety
social anxiety, and r = −.19* Network size with social anxiety
suicidal ideation
Liu et al. (2019) 93 Well-/ill-being = aggregate r = −.06* General SNS use with well-/ill-being
of life satisfaction, r = .09ns General SNS use with life satisfaction
happiness, self-esteem, r = .14* General SNS use with happiness
anxiety, depression, r = .13* General SNS use with depression
loneliness, and stress r = .10* General SNS use with anxiety
r = .02ns Active SNS use with well-/ill-being
r = −.14* Passive SNS use with well-/ill-being
Vahedi et al. (2021) 55 Depressive symptoms r = .11* General SNS use with depressive symptoms
r = .27* Problematic use with depressive symptoms
Yang et al. (2019) 13 Well-being = aggregate of r = −.20* Facebook social comparison with well-being
life satisfaction, self- r = −.21* Facebook social comparison with life satisfaction
esteem, and
psychological well-being
Yin et al. (2019) 63 Well-being = aggregate of r = .05* General SNS use with well-being
life satisfaction, well- r = .06* General SNS use with ill-being
being, self-esteem, and r = .13* Network size with well-being
positive affect r = −.03ns Network size with ill-being
Ill-being = aggregate of r = .04ns Active SNS use with well-being
depression, loneliness, r = .04ns Active SNS use with ill-being
anxiety, envy, and r = −.10ns Passive SNS use with well-being
negative affect r = .07ns Passive SNS use with ill-being

Current Opinion in Psychology 2022, 45:101294 www.sciencedirect.com

 Social Media Use and Well-Being: What We Know and What We Need to Know Valkenburg 5

Table 1 (continued )

Study # Articles Operational definitions of Main results

outcomes

Yoon et al. (2019) 45 Depression r = .11* Time spent on SNS with depression
r = .10* Frequency of checking SNS with depression
r = .23* Non-directional social comparison with depression
r = .33* Upward social comparison with depression

Notes. *Significant at least at p < .05. Table excludes effects for components of well-being (e.g., eudaimonic well-being) and ill-being (e.g., stress) that do not
fit within my operational definitions of well- and ill-being. SNS = Social network sites.

Recommendation 2: we need content-based SM research is to pay more attention to content-based SMU

predictors of well- and ill-being predictors [24].
The inconsistencies in the associations of the time-
based SM predictors may be caused by discrepancies Recommendation 3: we need a causal effect
in their operationalizations. For example, in some meta- heterogeneity paradigm
analyses “general SNS use” referred to time spent on Several reviews have pointed at the need for studies that
SNS [12], in others to a combination of time spent on allow for the investigation of within-person associations
SNS and the frequency of checking SNS [6,15], and in of SMU with well-being [24,27]. In recent years, a
yet others it was not defined [13]. growing number of such more rigorous studies have
appeared [51e56]. Butdagaindmost of these studies
Unfortunately though, the time-based predictors not found weak average associations with well- and ill-being
only led to heterogeneity across the meta-analyses but that were close to zero. What is still too often overlooked
also within the meta-analyses (e.g., I2s ranging from 57% in these studies is that such average associations are
for active SMU [12] to 97% for time spent on SM [4]), derived from heterogeneous populations of SM users
which could not or only partly be explained by moder- who differ in how they select and respond to SM [57], a
ators like age and gender. However, in case of consid- finding that has repeatedly been confirmed in qualita-
erable and (partly) unexplained heterogeneity, meta- tive studies [58]. To truly understand the effects of
analytic effect sizes may not be adequate and reliable SMU, researchers need to take the next step, that is,
[48]. A plausible explanation for the heterogeneity adopting a “causal effect heterogeneity” approach
within meta-analyses is that the time-based predictors [59,60], which enables them to better understand why
were operationalized differently in the included and how individuals differ in their responses to SMU.
empirical studies. This has been confirmed in a recent
scoping review, which revealed that of the 40 included To my knowledge, two communication research teams
survey-based studies, 90% used a unique, self-created have adopted a causal effect heterogeneity paradigm
operationalization of active and/or passive SMU, which [50,61], which led to the discovery of striking person-
led to a range of inconsistent associations with well- and specific effects of SMU on well-being. They found, for
ill-being components [24]. example, that about 20% of respondents experienced a
negative effect of passive SMU on happiness, 20% a
Yet even though the synchronization of time-based positive effect, and 60% no effect at all [51]. A causal
predictors in meta-analyses and empirical research may effect heterogeneity paradigm may not only help re-
be a first step, there are also conceptual concerns. Time- searchers resolve inconsistencies in findings (and repli-
based predictors may simply be too coarse to lead to cation failures) across studies [60], but it may also help
meaningful associations with well- and ill-being com- them to arrive at a better understanding of why in-
ponents [30]. Such predictors may be valuable for out- dividuals may or may not be affected by SMU.
comes like distraction or procrastination, which may be a
direct consequence of time spent using SM [49,50]. In A causal effect heterogeneity approach, sometimes
addition, they may be valuable when investigating time- called an idiographic or person-specific approach, can be
based hypotheses, such as the displacement hypothesis, applied in experimental designs [59,62], as well as in
which states that SMU takes away time that could non-experimental intensive longitudinal designs (e.g.,
otherwise be spent on activities that are more conducive experience sampling studies) [61,63]. The idiographic
to well-being than SMU. But since well- and ill-being approach has recently raised concerns among some
may be more amenable to the valence of SM in- communication scholars. One of these concerns is that
teractions (cf. humor vs hate, support vs neglect) than an idiographic approach in non-experimental settings
to their duration, a second important step for future would hinder inferences from an individual to a targeted

www.sciencedirect.com Current Opinion in Psychology 2022, 45:101294

6 Social Media and Well-Being

population. Another concern is that this approach hin- References

ders or even ignores the investigation of moderators to Papers of particular interest, published within the period of review,
have been highlighted as:
explain differences among subgroups in this targeted
population [64,65]. * of special interest
* * of outstanding interest

While these are valid concerns, they are well addressed

1. Waterloo SF, Baumgartner SE, Peter J, Valkenburg PM: Norms of
in recent idiographic modeling techniques, such as online expressions of emotion: comparing Facebook. Twitter
Dynamic Structural Equation Modeling (DSEM), which Instagram WhatsApp New Media Soc 2018, 20:1813–1831.
combine the strengths of “traditional” methods of 2. Bayer JB, Triệu P, Ellison NB: Social media elements, ecolo-
analysis (i.e., structural equation modelling and multi- * gies, and effects. Annu Rev Psychol 2020, 71:471–497.
This is an up-to-date conceptual review of computer-mediated
level analysis) with N = 1 time-series analysis [63]. communication and social media. It discusses how SM platforms
These modeling techniques require the same sizeable complicate the study of effects, as well as the challenges and oppor-
tunities to measure SMU across time.
samples to generalize to targeted populations as tradi-
tional (nomothetic) approaches do. Also, they can be 3. Appel M, Marker C, Gnambs T: Are social media ruining our
** lives? A review of meta-analytic evidence. Rev Gen Psychol
flexibly used to investigate the role of moderators in the 2020, 24:60–74.
person-specific effects of SMU on certain outcomes, see This is an earlier umbrella review of the effects of social media, based
on meta-analyses on well-being, academic performance, and narcis-
for example [40,66,67]. In fact, an important strength of ism. It includes a thorough discussion of possible pitfalls of meta-
these modelling techniques is that they allow for the analyses.
investigation of two types of moderators, (a) trait-like 4. Cunningham S, Hudson CC, Harkness K: Social media and
moderators, such as ethnicity and extraversion, and (b) depression symptoms: a meta-analysis. Res Child Adolesc
Psychopathol 2021, 49:241–253.
contextual moderators that are assumed to fluctuate
within participants, such as their motivations for using 5. Faelens L, Hoorelbeke K, Cambier R, van Put J, Van de Putte E,
De Raedt R, Koster EHW: The relationship between Instagram
SM, as well as their experience of envy or enjoyment use and indicators of mental health: a systematic review.
during SMU [40]. Comp Human Behav Rep 2021, 4:100121.
6. Hancock JT, Liu S, X, French M, Luo M, Mieczkowski H: Social
Idiographic approaches are thus complementing rather media use and psychological well-being: a meta-analysis. In
69th annual international communication association conference,
than replacing nomothetic approaches. They enable Washington, D.C.; 2019.
researchers to report aggregated between-person and 7. Aromataris E, Fernandez R, Godfrey CM, Holly C, Khalil H,
within-person associations of SMU with well- and ill- Tungpunkom P: Summarizing systematic reviews: methodo-
being. And in addition to these aggregated statistics, logical development, conduct and reporting of an umbrella
review approach. JBI Evi Implement 2015, 13:132–140.
they can demonstrate for how many participants an
8. Meier A, Reinecke L: Computer-mediated communication,
experimental treatment works or for how many partici- ** social media, and mental health: a conceptual and empirical
pants certain hypotheses hold [40]. Moreover, as argued meta-review. Commun Res 2020.
by Bryan et al. [60], a causal effect heterogeneity This is another recent umbrella review. It presents the extended two-
continua model of mental health, and gives an excellent review and
approach can improve interventions and “make them justification for why we cannot aggregate indicators of well-being and
effective for the diverse gamut of populations and con- ill-being.
texts policy must address” (p. 7). 9. Valkenburg PM, Meier A, Beyens I: The effects of social media
use on adolescents’ mental health: an umbrella review. Curr
Opin Psychol 2022, 44:58–68.
Conclusion 10. Huang C: A meta-analysis of the problematic social media
In sum, in addition to the wealth of valuable suggestions use and mental health. Int J Soc Psychiatr 2020.
for future research raised in the 27 reviews that have 0020764020978434.
appeared in the past 2.5 years, this umbrella review 11. Huang C: Correlations of online social network size with well-
being and distress: a meta-analysis. Cyberpsychology: J
showed why well-being and ill-being components Psychosoc Res Cyberspace 2021, 15. Article 3.
deserve to be investigated as two separate continuums
12. Liu D, Baumeister RF, Yang C-c, Hu B: Digital communication
(see also [8]). In addition, it made a case that we no media use and psychological well-being: a meta-analysis.
longer need additional meta-analyses reporting weak J Computer-Mediated Commun 2019, 24:259–274.
aggregate between-person effect sizes of time-based 13. Vahedi Z, Zannella L: The association between self-reported
SMU predictors, thereby reiterating the “one-size-fits- depressive symptoms and the use of social networking sites
(sns): a meta-analysis. Curr Psychol 2021, 40:2174–2189.
all approach” that has long characterized media effects
research. Indeed, “we have a bright future before us, and 14. Yang FR, Wei CF, Tang JH: Effect of Facebook social com-
parison on well-being: a meta-analysis. J Internet Technol
it begins where the average ends.” [ [68], p. 191]. 2019, 20:1829–1836.
15. Yin X-Q, De Vries DA, Gentile DA, Wang J-L: Cultural back-
Conflict of interest statement ground and measurement of usage moderate the association
between social networking sites (SNSs) usage and mental
Nothing declared. health: a meta-analysis. Soc Sci Comput Rev 2019, 37:631–648.

Current Opinion in Psychology 2022, 45:101294 www.sciencedirect.com

 Social Media Use and Well-Being: What We Know and What We Need to Know Valkenburg 7

16. Yoon S, Kleinman M, Mertz J, Brannick M: Is social network site 33. Verduyn P, Gugushvili N, Massar K, Täht K, Kross E: Social
usage related to depression? A meta-analysis of Face- comparison on social networking sites. Curr Opin Psychol
book–depression relations. J Affect Disord 2019, 248:65–72. 2020, 36:32–37.
17. Duradoni M, Innocenti F, Guazzini A: Well-being and social 34. Festinger L: A theory of social comparison processes. Hum
media: a systematic review of Bergen addiction scales. Relat 1954, 7:117–140.
Future Internet 2020, 12:24.
35. Keyes CLM: Mental illness and/or mental health? Investi-
18. Gilmour J, Machin T, Brownlow C, Jeffries C: Facebook-based gating axioms of the complete state model of health. J Consult
social support and health: a systematic review. Psychol Clin Psychol 2005, 73:539–548.
Popular Media 2020, 9:328–346.
36. Ryff CD, Dienberg Love G, Urry HL, Muller D, Rosenkranz MA,
19. Hussain Z, Wegmann E, Yang H, Montag C: Social networks Friedman EM, Davidson RJ, Singer B: Psychological well-being
use disorder and associations with depression and anxiety and ill-Being: do they have distinct or mirrored biological
symptoms: a systematic review of recent research in China. correlates? Psychother Psychosom 2006, 75:85–95.
Front Psychol 2020, 11:211.
37. Verduyn P, Gugushvili N, Kross E: The impact of social network
20. Karim F, Oyewande AA, Abdalla LF, Chaudhry Ehsanullah R, sites on mental health: distinguishing active from passive
Khan S: Social media use and its connection to mental health: use. World Psychiatr 2021, 20:133–134.
a systematic review. Cureus 2020, 12, e8627.
38. Burke M, Cheng J, de Gant B: Social comparison and Facebook:
21. Newman L, Stoner C, Spector A: Social networking sites and feedback positivity and opportunities for comparison, conference
the experience of older adult users: a systematic review. on human factos in computing systems. Honolulu, HI, USA: CHI;
Ageing Soc 2021, 41:377–402. 2020.
22. O’Day EB, Heimberg RG: Social media use, social anxiety, and 39. Krasnova H, Widjaja T, Buxmann P, Wenninger H, Benbasat I:
loneliness: a systematic review. Comp Human Behav Rep Why following friends can hurt you: an exploratory investi-
2021, 3:100070. gation of the effects of envy on social networking sites
among college-age users. Inf Syst Res 2015, 26:585–605.
23. Sharma MK, John N, Sahu M: Influence of social media on
mental health: a systematic review. Curr Opin Psychiatr 2020, 40. Valkenburg PM, Beyens I, Pouwels JL, van Driel II, Keijsers L:
33:467–475. Social media browsing and adolescent well-being: chal-
lenging the “passive social media use hypothesis,”.
24. Valkenburg PM, van Driel II, Beyens I: The associations of J Computer-Mediated Commun 2022, 27.
active and passive social media use with well-being: a critical
scoping review. New Media & Society 2021, https://doi.org/ 41. Meier A, Gilbert A, Börner S, Possler D: Instagram inspiration:
10.31234/osf.io/j6xqz. how upward comparison on social network sites can
contribute to well-being. J Commun 2020, 70:721–743.
25. AlBarashdi HS: Social networking (SNS) addiction among
university students: a literature review and research di- 42. Cheng C, Wang H-Y, Sigerson L, Chau C-L: Do the socially rich
rections. J Educ Soc Behav Sci 2020, 33:11–23. * get richer? A nuanced perspective on social network site use
and online social capital accrual. Psychol Bull 2019, 145:
26. Bettmann JE, Anstadt G, Casselman B, Ganesh K: Young adult 734–764.
depression and anxiety linked to social media use: assess- This meta-analysis is not included in Table 1, because it does not fit
ment and treatment. Clin Soc Work J 2021, 49:368–379. within a social media effects paradigm. It is an excellent theory-based
meta-analysis on the the poor-get-richer/rich-get richer hypotheses,
27. Carboni Jiménez A, Vaillancourt M, Zhu P, Seon Q: Social media
which conceptualize social media use as the outcome and extraver-
and mental health: what we know. McGill J Med 2021, 20.
sion, social anxiety, and loneliness as the predictors.
28. Hartanto A, Quek FYX, Tng GYQ, Yong JC: Does social media
* 43. Ellison NB, Steinfield C, Lampe C: The benefits of Facebook
use increase depressive symptoms? A reverse causation
“friends:” Social capital and college students’ use of online
perspective. Front Psychiatr 2021, 12.
social network sites. J Computer-Mediated Commun 2007, 12:
This is an insightful narrative review that includes an overview of “social
1143–1168.
media abstinence” studies, in which participants are asked to stay
away from social media for varying durations (e.g., days, weeks). It 44. Xie W, Karan K: Predicting Facebook addiction and state
discusses an important pitfall of most of these experiments: the anxiety without Facebook by gender, trait anxiety, Facebook
absence of a “placebo control” condition. Such a control condition intensity, and different Facebook activities. J Behav Addic
avoids that expectancy effects of participants in the abstinencer con- 2019, 8:79–87.
dition differ from those in the control condition. Everyone who intends to
conduct an social media abstinence experiment should read this 45. van den Eijnden RJJM, Lemmens JS, Valkenburg PM: The social
review, and its suggested references. media disorder scale. Comput Hum Behav 2016, 61:478–487.

29. Griffioen N, Rooij Mv, Lichtwarck-Aschoff A, Granic I: Toward 46. Andreassen CS, Pallesen S, Griffiths MD: The relationship be-
* * improved methods in social media research. Technol Mind tween addictive use of social media, narcissism, and self-
Behav 2020, 1. esteem: findings from a large national survey. Addict Behav
This review identifies itself as a narrative review, but is, in fact, one of 2017, 64:287–293.
the best systematic reviews of 94 empirical studies. It draws attention
47. Faelens L, Hoorelbeke K, Soenens B, Van Gaeveren K, De
to many important gaps in the literature and provides several important
Marez L, De Raedt R, Koster EHW: Social media use and well-
and well-argued suggestions for future research.
being: a prospective experience-sampling study. Comput
30. Kross E, Verduyn P, Sheppes G, Costello CK, Jonides J, Hum Behav 2021, 114:106510.
* Ybarra O: Social media and well-being: pitfalls, progress, and
48. Melsen WG, Bootsma MCJ, Rovers MM, Bonten MJM: The ef-
next steps. Trends Cognit Sci 2021, 25:55–66.
fects of clinical and statistical heterogeneity on the predictive
This is an accessible narrative review of the effects of social media use
values of results from meta-analyses. Clin Microbiol Infect
on well-being. It addresses some important pitfalls of the current liter-
2014, 20:123–129.
ature. It acknowledges that it is time to move beyond the active-passive
dichotomy, and that a refinement of this dichotomy is necessary. 49. Siebers T, Beyens I, Pouwels JL, Valkenburg PM: Social media
and distraction: an experience sampling study among ado-
31. Luo M, Hancock JT: Self-disclosure and social media: moti- lescents. Media Psychol 2021.
vations, mechanisms and psychological well-being. Curr Opin
Psychol 2020, 31:110–115. 50. Aalbers G, vanden Abeele MM, Hendrickson AT, de Marez L,
Keijsers L: Caught in the moment: are there person-specific
32. Mou J, Zhu W, Benyocef M, Kim J: Understanding the rela- associations between momentary procrastination and
tionship between social media use and depression: a sys- passively measured smartphone use? Mobile Media
tematic review. AMCIS 2020 Proceed 2020:15. Commun 2021, 2050157921993896.

www.sciencedirect.com Current Opinion in Psychology 2022, 45:101294

8 Social Media and Well-Being

51. Beyens I, Pouwels JL, van Driel II, Keijsers L, Valkenburg PM: effect heterogeneity can be modeled, and how it presents exciting
Social media use and adolescents’ well-being: developing a opportunities for theory and methods.
typology of person-specific effect patterns. Commun Res
2021. 60. Bryan CJ, Tipton E, Yeager DS: Behavioural science is unlikely
* to change the world without a heterogeneity revolution. Nat
52. Coyne SM, Rogers AA, Zurcher JD, Stockdale L, Booth M: Does Human Behav 2021, 5:980–989.
time spent using social media impact mental health? An Like Bolger et al. (see previous annotation), this study makes a strong
eight year longitudinal study. Comput Hum Behav 2020, 104: case for why we need a causal effects heterogeneity paradigm in the
106160. social and behavioral sciences.
53. Hall JA, Xing C, Ross EM, Johnson RM: Experimentally 61. Valkenburg PM, Beyens I, Pouwels JL, van Driel II, Keijsers L:
manipulating social media abstinence: results of a four-week Social media and adolescents’ self-esteem: heading for a
diary study. Media Psychol 2021:259–275. person-specific media effects paradigm. J Commun 2021, 71:
56–78.
54. Jensen M, George MJ, Russell MR, Odgers CL: Young adoles-
cents’ digital technology use and mental health symptoms: 62. Grice JW, Medellin E, Jones I, Horvath S, McDaniel H,
little evidence of longitudinal or daily linkages. Clin Psychol O’lansen C, Baker M: Persons as effect sizes. Adv Method Prac
Sci 2019, 7:1416–1433. Psychol Sci 2020, 3(4):443–455.
55. Orben A, Dienlin T, Przybylski AK: Social media’s enduring 63. McNeish D, Hamaker EL: A primer on two-level dynamic
effect on adolescent life satisfaction. Proc Natl Acad Sci Unit structural equation models for intensive longitudinal data in
States Am 2019, 116:10226–10228. Mplus. Psychol Methods 2020, 25:610–635.
56. Przybylski AK, Nguyen T-vT, Law W, Weinstein N: Does taking a 64. Parry D, Fisher J, Mieczkowski H, Sewall C, Davidson BI: Social
short break from social media have a positive effect on well- media and well-being: a methodological perspective. Curr Opin
being? Evidence from three preregistered field experiments. Psychol 2022, https://doi.org/10.1016/j.copsyc.2021.11.005.
J Technol Behav Sci 2021, 6:507–514.
65. Johannes N, Masur P, Vuorre M, Przybylski A: How should we
57. Valkenburg PM, Peter J, Walther JB: Media effects: theory and investigate variation in the relation between social media and well-
research. Annu Rev Psychol 2016, 67:315–338. being?. 2021.
58. Rideout V, Fox S: Digital health practices, social media use, and 66. Valkenburg PM, Pouwels JL, Beyens I, van Driel II, Keijsers L:
mental well-being among teens and young adults in the US. San Adolescents’ experiences on social media and their self-
Francisco, CA: HopeLab; 2018. esteem: a person-specific susceptibility perspective. Technol
Mind Behav 2021, 2(2).
59. Bolger N, Zee K, Rossignac-Milon M, Hassin R: Causal pro-
* cesses in psychology are heterogeneous. J Exp Psychol Gen 67. Siebers T, Beyens I, Pouwels JL, Valkenburg PM: Social media are
2019, 148:601–618. a powerful distractor for the vast majority of adolescents. 2021.
This paper argues that people differ in their responses to environmental
influences, but that most research has ignored this heterogeneity or 68. Rose T: The end of average: how to succeed in a world that
has treated it as uninteresting error. The paper shows how causal values sameness. New York: Harper Collins; 2016.

Current Opinion in Psychology 2022, 45:101294 www.sciencedirect.com