Social Media Abstinence Did Not Improve Mood or Life Satisfaction in 10-Study Meta-Analysis of 4,674 Adults

TL;DR: A 2025 systematic review and meta-analysis in Scientific Reports pooled 10 social media abstinence experiments (N = 4,674 adults; abstinence 1–28 days) and found no significant effect on positive affect, negative affect, or life satisfaction.

Source: Scientific Reports (2025) | Lemahieu et al.

10 Abstinence Experiments and 4,674 Adults Were Pooled

Researchers at the University of Antwerp and Ghent University ran a preregistered systematic review of social media detox experiments.

They searched PubMed, Scopus, Web of Science, Communication Source, the Cochrane Library, and Google Scholar for studies that asked adults to completely stop using one or more social media platforms.

The final pool included 10 studies and 4,674 participants, yielding 38 separate effect sizes across the three outcomes. About 65% of participants were women, and most studies recruited college students or community adults from Western countries.

The team measured three well-being outcomes:

• Positive affect: the frequency and intensity of pleasant emotions such as enthusiasm and energy, usually measured with the Positive and Negative Affect Schedule (PANAS).
• Negative affect: unpleasant emotions such as anger, fear, and guilt, also measured with the PANAS in most studies.
• Life satisfaction: a person’s overall evaluation of their life, usually measured with the Satisfaction With Life Scale (SWLS).

Effect sizes were pooled with a random-effects model and reported as Hedges’ g, a standardized mean difference where positive numbers favor the abstinence group.

Pooled Effect Sizes Were Near Zero for All Three Outcomes

The pooled effects were near zero across all three outcomes:

• Positive affect: Across 9 studies, the pooled effect was g = 0.03 with a 95% confidence interval of [−0.11, 0.16].
• Negative affect: Across the same 9 studies, the pooled effect was g = −0.01 with a 95% confidence interval of [−0.13, 0.10].
• Life satisfaction: Across 6 studies, the pooled effect was g = 0.03 with a 95% confidence interval of [−0.17, 0.22].

None of the three pooled effects reached statistical significance, and every confidence interval crossed zero.

Heterogeneity between studies was substantial: I² was 60.7% for positive affect, 63.7% for negative affect, and 58.8% for life satisfaction. That means the studies disagreed with each other more than chance would predict, even though their pooled average was close to zero.

Removing Within-Subjects Studies Nudged Positive Affect to a Borderline Effect

Researchers ran a sensitivity analysis that excluded studies with within-subjects designs and kept only the 10 between-subjects effect sizes. For positive affect, the pooled estimate moved from g = 0.03 (p = 0.69) to g = 0.13 (95% CI [0.00, 0.27], p = 0.047), a borderline effect.

The same exclusion did not change the negative-affect result. The pooled estimate moved from g = −0.01 to g = −0.04 and remained non-significant.

The borderline positive-affect signal in between-subjects studies is small in absolute terms. A Hedges’ g of 0.13 sits well below conventional thresholds for a small effect (g = 0.20).

Researchers also noted that the included studies were generally underpowered to detect small effects, which makes any single contrast hard to interpret.

Abstinence Length Did Not Predict Better Affect or Life Satisfaction

Most studies asked participants to give up social media for about a week. The most common assigned duration was 7 days (k = 6), but the included experiments ranged from 1 day to 28 days.

Researchers ran a meta-regression for each outcome with abstinence days as the moderator. None of the three slopes were significant:

• Positive affect: γ = 0.0032, SE = 0.0103, p = 0.76, R² = 0%.
• Negative affect: γ = −0.0056, SE = 0.0088, p = 0.54, R² = 0%.
• Life satisfaction: γ = 0.004, SE = 0.01, p = 0.70, R² = 0%.

Longer breaks did not produce reliably better mood or higher life satisfaction. The team noted that very few of the pooled studies tested abstinence durations beyond 4 weeks, so the comparison mostly contrasts 1-day, 1-week, 2-week, 3-week, and 4-week breaks rather than truly long-term disconnection.

Compliance Was Often Poor and Most Studies Did Not Blind Participants

Nearly all of the included trials checked whether participants actually stayed off social media, but compliance varied widely:

• Stieger and Lewetz (2018): 59% of participants visited social media at least once during the assigned 7-day break.
• Przybylski et al. (2021): overall compliance across three experiments was 49.5%.
• Wadsley and Ihssen (2023): only 13.7% of participants completed a full week of abstinence, although the group still cut its overall use by 83.4%.

None of the trials blinded participants to the intervention. That is hard to do in a behavioral abstinence study, but it leaves room for demand effects, where participants change their self-reported mood because they know what the researchers expect to see.

Most studies also did not run a power analysis, and the firepower analysis in this meta-analysis indicated low statistical power across all three outcomes.

Sample Size, Western Student Skew, and Self-Reported Outcomes Are Real Limits

The review named several constraints on how far the null result should travel:

• Few included studies: 10 trials is a small base for moderator analyses, so the team could not run subgroup tests by platform, age, or country.
• WEIRD sampling: participants were mostly Western, Educated, Industrialized, Rich, and Democratic adults, often students. Adolescents and older adults are barely represented.
• Inconsistent platform definitions: some trials did not specify whether instant messaging apps counted as social media, which makes the abstinence dose hard to compare across studies.
• Self-reported outcomes: affect and life satisfaction came from self-report scales, which can be influenced by participant expectations during an unblinded intervention.
• Short follow-up windows: most experiments ran for a week or less, so longer-term effects of stepping away from social media were not really tested.

Reducing use may be a better test than total abstinence: The non-significant pooled effects do not prove that social media has no impact on well-being.

The pooled effects suggest that short, complete breaks from social media are not, on average, the lever that moves mood or life satisfaction in the adults studied so far.

Researchers argued for a shift in the question. They recommended longer studies, intensive longitudinal designs that capture day-level satisfaction, and tests of partial-reduction strategies such as built-in screen-time limits or app-blockers.

They also noted that voluntary, self-imposed breaks may behave differently than researcher-imposed ones, because motivation and mindset toward social media likely shape the outcome.

For people considering a personal detox, the meta-analysis offers a calmer takeaway than either “quit social media to feel better” or “quitting will make you miserable.” In the pooled data, neither was true on average, and the duration of the break did not matter.