Next Article in Journal
Temporal Associations Between Cognitive Impairment and Depression in Older Adults: A Longitudinal Analysis
Previous Article in Journal
Clarity and Emotional Regulation as Protective Factors for Adolescent Well-Being: A Moderated Mediation Model Involving Depression
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
EJIHPE
Volume 15
Issue 7
10.3390/ejihpe15070131
ejihpe-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Systematic Review

Loneliness as a Public Health Challenge: A Systematic Review and Meta-Analysis to Inform Policy and Practice

by
Ananda Zeas-Sigüenza
1,2,
Andreas Voldstad
3,
Pablo Ruisoto
1,4,5,6,*,
Ana Ganho-Ávila
7,
Raquel Guiomar
7,
Raúl Cacho
1,5,
Ferran Muntané
6,8 and
Joan Benach
6,8,*
1
Department of Health Sciences, Public University of Navarre, 31006 Pamplona, Spain
2
Department of Psychology, School of Psychology and Sport Science, Bangor University, Bangor LL57 2DG, Wales, UK
3
Department of Psychiatry, Warneford Hospital, University of Oxford, Oxford OX3 7JX, England, UK
4
I-COMMUNITAS, Institute for Advanced Social Research, Public University of Navarre, 31006 Pamplona, Spain
5
IdisNA, Navarra Institute for Health Research, Public University of Navarre, 31006 Pamplona, Spain
6
JHU-UPF Public Policy Center (JHU-UPF PPC), UPF Barcelona School of Management (UPF-BSM), Universitat Pompeu Fabra (UPF), 08005 Barcelona, Spain
7
Centre for Research in Neuropsychology and Cognitive Behavioural Intervention, University of Coimbra, 3030-788 Coimbra, Portugal
8
Research Group on Health Inequalities, Environment, and Employment Conditions (GREDS-EMCONET), Department of Political and Social Sciences, Universitat Pompeu Fabra, 08005 Barcelona, Spain
*
Authors to whom correspondence should be addressed.
Eur. J. Investig. Health Psychol. Educ. 2025, 15(7), 131; https://doi.org/10.3390/ejihpe15070131
Submission received: 17 April 2025 / Revised: 27 May 2025 / Accepted: 26 June 2025 / Published: 11 July 2025
Browse Figures
Versions Notes

Abstract

Loneliness is a recognized public health risk factor associated with increased morbidity and mortality. However, the effectiveness of interventions targeting loneliness remains unclear—particularly in relation to baseline severity. This systematic review and meta-analysis assessed intervention effectiveness and the influence of baseline severity and intervention characteristics. A total of 25 studies were included, of which 16 randomized controlled trials (RCTs; k = 21) were meta-analyzed. Interventions produced a moderate pooled effect at post-intervention (Hedge’s g = 0.65, 95% CI [0.05, 1.26], p = 0.037), though with high heterogeneity. Sensitivity analyses confirmed a moderate effect (g = 0.55, 95% CI [0.22, 0.88], p = 0.003). Higher baseline loneliness predicted greater intervention effects (b = 0.04, 95% CI [0.02, 0.07], Z = 3.36, p < 0.001), with cognitive-behavioral therapy (CBT) showing the largest effect size (g = 0.73). No significant effects were observed at follow-up. These findings underscore the need for dual strategies: targeted psychological interventions (e.g., CBT) for individuals with severe loneliness, and universal, context-based approaches for the broader population. This aligns with Geoffrey Rose’s distinction between individual-level treatment and population-level prevention and highlights the urgency of embedding loneliness interventions into public health frameworks and policy agendas focused on promoting social connectedness and equity.

1. Introduction

Loneliness is a well-established determinant of health, with risks comparable to those of smoking (Holt-Lunstad et al., 2010, 2015) and is increasingly recognized as a critical public health priority (Arsenijevic & Groot, 2018; Holt-Lunstad, 2017; Steptoe et al., 2013). Its prevalence surged following the COVID-19 pandemic, driven by lockdowns and distancing measures (WHO, 2020). As a potent psychosocial stressor, loneliness activates the hypothalamic–pituitary–adrenal axis (J. T. Cacioppo & Hawkley, 2009; S. Cacioppo et al., 2014) and is associated with a wide range of adverse health outcomes, including depression (Vanhalst et al., 2012), suicidal behavior (Mezuk et al., 2014), poor sleep quality (Hom et al., 2020), and cardiovascular disease (Valtorta et al., 2018). Conceptualized as a transdiagnostic factor, loneliness cuts across diagnostic categories and contributes to multiple mental and physical health conditions (Nowland et al., 2018; Ruisoto et al., 2022).
Unlike objective social isolation, loneliness is a subjective experience characterized by a discrepancy between desired and actual meaningful social relationships (J. T. Cacioppo & Patrick, 2008). It exists on a continuum of social connectedness—from low to high or severe loneliness (S. Cacioppo & Cacioppo, 2012)—and is thought to involve altered social information processing, where social interactions may become less rewarding or even aversive in individuals with severe loneliness (Tomova et al., 2020).
Despite its complexity, most intervention studies dichotomize individuals as either lonely or not, often neglecting the degrees in loneliness severity (Astell-Burt et al., 2022; Barreto et al., 2021; Cohen-Mansfield & Perach, 2015; Eccles & Qualter, 2021; Fu et al., 2022; Gardiner et al., 2018; Jarvis et al., 2020; Maes et al., 2019; Masi et al., 2011). Intervention effectiveness is likely influenced by a multitude of factors. These include the intervention focus—whether targeting individual capacities (e.g., enhancing coping or emotion regulation) or the broader context (e.g., addressing environmental or structural conditions) (J. T. Cacioppo & Patrick, 2008; J. T. Cacioppo et al., 2002)—as well as delivery mode (e.g., online vs. in-person), which can impact accessibility, engagement, and the quality of social interaction (Tan et al., 2024; Masi et al., 2011).
Additional moderating factors include intervention format (individual vs. group), which may influence the balance between personalization and social bonding (S. Cohen & Wills, 1985; Hoang et al., 2022), and specific active components such as cognitive restructuring or nature-based activities (Barton & Pretty, 2010; Leavell et al., 2019). Participant characteristics—such as age, gender, and cultural context—can also shape both the experience of loneliness and responsiveness to interventions. For instance, older adults often benefit from interventions that promote meaningful social engagement (de Oliveira et al., 2015), while cultural values (e.g., collectivism vs. individualism) and gender norms can influence how loneliness is conceptualized and mitigated (Hickin et al., 2021; Hoang et al., 2022; Hofstede et al., 2010).
Grounded in this complexity, this study conducts a systematic review and meta-analysis to assess the effectiveness of loneliness interventions and examine how outcomes are moderated by (1) baseline loneliness severity; (2) intervention characteristics (focus, type, delivery mode, format, and active components); and (3) demographic and cultural variables (age, gender, cultural context). This work is informed by Geoffrey Rose’s (1985) influential distinction between strategies that target “high-risk individuals” and those that aim to shift the risk distribution at the population level. By mapping loneliness interventions along this continuum, we aim to provide evidence not only for effective therapeutic approaches, but also for broader public health strategies. Our findings are intended to inform future research, guide mental health service design, and support the development of equity-oriented policies that recognize loneliness as a social determinant of health—shaped by both individual vulnerabilities and structural conditions.

2. Materials and Methods

This study was registered in June 2023 (PROSPERO CRD: 42023429666) and followed the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (Page et al., 2021) and the Cochrane Collaboration recommendations (J. P. Higgins et al., 2011; J. P. T. Higgins et al., 2023).

2.1. Eligibility Criteria

The primary outcome established for this review was post-intervention loneliness compared to a control group. Studies were eligible if they (1) were published between 2013 and 2023, due to the lack of extensive literature before 2013; (2) were in English or Spanish, the working languages of the review team; (3) employed a controlled quantitative design evaluating the effectiveness of loneliness interventions in adults; and (4) used validated original or short versions of either the gold standard University of Los Angeles Loneliness Scale (UCLA) or the De Jong–Gierveld Loneliness Scale (DJLS). See Supplement 1 for a detailed list of loneliness measures. Only studies involving healthy populations and non-pharmacological interventions were included to minimize confounding variables. Studies using non-standardized loneliness measures or transforming loneliness into binary outcomes were excluded due to limited conceptual comparability.

2.2. Search Strategy and Study Selection

Database searches were conducted on 23 June 2023, and repeated on 11 December 2023, in CINAHL, Embase, Medline, PsychINFO, Pubmed, and Web of Science Core Collection (see Supplement 1 for full syntax). Grey literature was identified via ProQuest and OpenAIRE (e.g., theses, dissertations). Backward citation tracking was performed by the first author. Study screening was conducted using Rayyan (Ouzzani et al., 2016). Titles and abstracts were independently screened by two reviewers (A.Z.S & A.V.). Full-text eligibility was independently assessed by two additional reviewers (A.Z.S & R.C.), with disagreements resolved by a third reviewer (P.R.).

2.3. Data Extraction and Quality Assessment

Data extraction was piloted and subsequently standardized across all included studies. Extracted information included authors, study design, sample size, baseline characteristics, country and cultural classification following the guidelines of Hofstede et al. (2010), whether loneliness was a primary or secondary outcome, and intervention characteristics (e.g., intervention focus, delivery mode) (see Supplement 1 for full extraction form). Two reviewers (A.Z.S. & R.C.) independently extracted data. Authors from eight studies were contacted for missing information; four provided the requested data and were retained, while the other four were excluded. Risk of bias was independently assessed by two reviewers (A.Z.S & R.C.) using the Cochrane Risk of Bias 2.0 tool (RoB) (Sterne et al., 2019), with disagreements resolved by a third reviewer (P.R.).

2.4. Statistical Analysis

All analyses were conducted using R (version v4.0.4) using the meta (v4.18.0) and metafor (v2.4.0) packages (Viechtbauer, 2010). Standardized mean differences (SMD) were calculated using Hedges’ g with 95% confidence intervals (CIs) for each comparison (k), subtracting the mean of the intervention group from the mean of the control group. Positive values indicated less loneliness in the intervention group compared to the control. Magnitude of the difference was classified as small (g = 0.20), moderate (g = 0.50), or large (g ≥ 0.80) (J. Cohen, 1992). A random-effects model was applied, as we assume the effect sizes come from different populations, applying the inverse variance method to calculate the pooled effect size, and standard errors were adjusted with the Hartung–Knapp Adjustment method (IntHout et al., 2016). To address the additional uncertainty of estimating the variance between studies (τ2), we used the DerSimonian–Laird method to provide the heterogeneity estimate τ2 (DerSimonian & Kacker, 2007; DerSimonian & Laird, 2015). Hedges’ g values were transformed into the number needed to treat (NNT) using the Kraemer and Kupfer (2006) method, where lower NNTs indicate greater effectiveness (Laupacis et al., 1988). Heterogeneity was assessed using Cochran’s Q statistic and the proportion of variability due to heterogeneity with the I2 index, interpreted as low (25%), moderate (50%), or high (75%) (J. P. Higgins et al., 2003). We conducted sensitivity analyses excluding (1) interventions without visual contact (e.g., phone-based or audio-only), due to the role of facial expressions in social information processing; (2) interventions with a combined format (group and individual), to control for the influence of mixed sessions; (3) studies identified with high risk of bias using the Cochrane Risk of Bias (RoB) tool; (4) a study with an effect size significantly larger than the rest in the data set (Cuijpers, 2016); and (5) both the extreme study and the high-RoB studies simultaneously. Publication bias was assessed via visual inspection of funnel plots on loneliness interventions, Begg’s test to assess potential differences between adjusted and observed effect sizes (Begg & Mazumdar, 1994), Egger’s regression test (Egger et al., 1997) of funnel plot asymmetry, and the trim-and-fill procedure (Duval & Tweedie, 2000) to address any potential publication bias. Begg’s and Egger’s tests were rerun post-adjustment to confirm results.
Post hoc subgroup and moderation analyses were conducted for cultural background (Hofstede categories: individualist vs. collectivist societies); intervention focus (person- vs. context-focused); intervention format (individual vs. group); delivery mode (in-person vs. virtual); and loneliness as the primary outcome (yes/no). Interventions were categorized as either single-component (e.g., CBT only) or multi-component interventions (e.g., CBT combined with other techniques).
Meta-regressions were conducted to explore the moderating effects of age, gender, and baseline loneliness severity on the efficacy of loneliness interventions. Moderator significance was assessed using the Q-test for moderators (QM). Significant Q-tests of categorical subgroups were interrogated with meta-regressions. For the baseline loneliness severity, all scores were standardized to align with the 20–80 range of the UCLA Loneliness Scale (Version 3). See Supplement 1 (Table S5).

3. Results

3.1. Characteristics of the Included Studies

A total of 25 randomized controlled trials (RCTs) were included in the systematic review. Of these, nine compared loneliness interventions to another intervention (intervention vs. intervention; I-I), and 16 compared loneliness interventions to passive control groups (intervention vs. control; I-C) (see PRISMA flowchart, Figure 1). Across all studies, the total sample at baseline comprised 4761 participants (n Intervention = 2443; n Control = 2318), of whom 71.28% (SD = 15.28%) were women. The mean age was 46.04 years (SD = 22.8) (see Table 1). All studies used the UCLA Loneliness Scale, with one study (Kahlon et al., 2021) additionally using the De Jong–Gierveld Loneliness Scale (DJLS). In 85% of the studies (n = 21), loneliness was reported as the primary outcome. Only three studies (12%) applied a pre-settled loneliness severity cut-off as part of their inclusion criteria. Additional information on sample characteristics, intervention description, and follow-up assessments is provided in Table 1 and Supplement 1 (Table S2). Regarding methodological quality, two studies (8%) were assessed as having low risk of bias, nineteen (76%) as moderate risk, and four (16%) as high risk. The primary source of bias (RoB) stemmed from the absence of pre-registered analysis plans. Further detail on risk of bias assessments is available in Supplements 1 and 2.
Narrative results are presented in the following paragraphs. We first present results from intervention–control (I-C) studies (n = 16), followed by intervention–intervention (I-I) studies (n = 9).
The 16 intervention–control (I-C) randomized controlled trials (RCTs) included a total of 2476 participants at baseline (n Intervention = 1271; n Control = 1205), of whom 68% (SD = 20.8) were women. The mean age across samples was 51.96 years (SD = 21.58). In terms of target populations, 44% (n = 7) of studies included older adults, 37.5% (n = 6) middle-aged adults, and 12.5% (n = 2) undergraduates. Regarding cultural setting, 67% (n = 11) of studies were conducted in individualistic cultural settings (Finland, Germany, Israel, Iran, Portugal, Sweden, and the United States) and 33% (n = 5) in collectivistic cultural settings (China, Iran, and Turkey). The most common active components were evidence-based psychological interventions, including Cognitive Behavioral Therapy (CBT [31%, n = 5]); Interpersonal Psychotherapy (IPT [6%, n = 1]); and Acceptance and Commitment Therapy (ACT [6%, n = 1]). These were followed by social support interventions (50%, n = 8) and reminiscence therapy (12.5%, n = 2). In terms of design complexity, 63% (n = 10) of studies employed a multiple-component approach: CBT (12.5%, n = 2); social support (SS [44%, n = 7]); and reminiscence therapy (6%, n = 1). Meanwhile, 37.5% (n = 6) used a single-component approach, including CBT (19%, n = 3), as well as ACT, art therapy, meditation, reminiscence therapy, physical activity, and mindful movement (each 6%, n = 1). Most interventions were person focused (81%, n = 13), with only 6% (n = 1) classified as context–person-focused. Regarding delivery mode, 44% (n = 7) were delivered in person, and 56% (n = 9) were delivered virtually. Format was evenly split: 50% (n = 8) used individual sessions, and 50% (n = 8) used group sessions. On average, interventions were delivered over 12.10 sessions (SD = 10.88) within an average period of eight weeks. One study (6%) consisted of a single session, meanwhile two studies (12.5%) spanned one month and six months, respectively.
The nine intervention–intervention (I-I) randomized controlled trials (RCTs) included a total of 2258 participants at baseline (n Intervention = 1158; n Control = 1100), of whom 73% (SD = 7.66) were women. The mean age across samples was 34.80 years (SD = 20.44). In terms of sample composition, 4 studies focused on undergraduate students, 3 on older adults, and 1 on middle-aged adults. Regarding cultural context, 6 studies were conducted in individualistic cultural settings (United Kingdom & United States), and 3 in collectivistic cultural settings (Taiwan). The most common active component was interventions that promoted social interactions (n = 2), followed by mindfulness-based interventions (n = 2). Other components—each represented in a single study—included art-therapy, volunteering, nature exposure, virtual interaction with strangers, and digital social support. Most interventions (n = 7) used a single-component approach, while 2 studies used a multiple-component approach. Only one study examined the effect of a context-based intervention. Delivery modes were predominantly virtual (n = 7); and two were delivered in-person. Most were individual-based (n = 8), with only one delivered in a group format. On average, I-I interventions were delivered in 12.10 sessions (SD = 10.88), over a span of approximately 13 weeks. Two studies represented temporal outliers, with interventions delivered within four days or extended across 12 months, respectively. Regarding outcomes, six of the I-I RCTs reported statistically significant pre-post reductions in loneliness, two reported no significant changes, and one did not report p-values. For detailed p-values, see Supplement 1.

3.2. Meta-Analysis

Due to the substantial heterogeneity in main active components across the intervention-intervention (I-I) RCTs, it was not feasible to group or compare them with consistent control types. As a result, only the 16 intervention-control (I-C) RCTs were included in the meta-analysis. The overall pooled effect size of loneliness interventions was moderate, with Hedges’ g = 0.65 (95% CI [0.05; 1.26]), p = 0.037), and a number needed to treat (NNT) of 2.82. However, heterogeneity was high, with Q(df=20) = 216.75, p < 0.001, and I2 = 91% (95% CI [87; 93]) (see Figure 2). One study (Li et al., 2022) exhibited an extremely large effect size (g = 8.81, 95% CI [7.14; 10.49]), prompting sensitivity analyses that excluded this study. Follow-up data were available in six studies (k = 8 comparisons). The pooled follow-up effect size was non-significant, g = 1.91 (95% CI [−1.87; 5.69], p = 0.251), with very high heterogeneity (Q(df=5) = 113.27, p < 0.001; I2 = 95.6%, 95% CI [93; 97]). For detailed results of the follow-up effects, see Supplement 1 (Table S3) and Supplement 3.
Three sensitivity analyses were conducted to assess the robustness of post-intervention effects. These included removing a study with a mixed intervention format, excluding two studies due to the nature of the online intervention, and excluding studies with high risk of bias (RoB) (Morgado et al., 2023; Myhre et al., 2017; Ristolainen et al., 2020; Thimmapuram et al., 2021). In all cases, the pooled effect sizes at post-intervention remained moderate. After excluding the study with an extremely large effect size (g = 8.81), the pooled effect was small to moderate, g = 0.46 (95% CI [0.19; 0.73], p = 0.002, NNT = 3.92), with high heterogeneity, Q(df=19) = 116.73, p < 0.001; I2 = 84% (95% CI [76; 89]). When both the extreme effect size study and the high RoB studies were excluded, the pooled effect size remained moderate, g = 0.55 (95% CI [0.22; 0.88], p = 0.003, NNT = 3.29), with similarly high heterogeneity, Q(df= 15) = 95.41, p < 0.001; I2 = 84% (95% CI [76; 89]). Follow-up effects remained non-significant across all sensitivity analyses. For full sensitivity analysis results, see Supplement 1 (Table S3) and Supplement 4.

3.3. Risk of Publication Bias

Publication bias was assessed using both the full sample of studies and a subset of 11 RCTs (after excluding the study with an extreme effect size and those rated as high risk of bias). In the full sample, both Begg’s rank correlation test and Egger’s regression test were statistically significant, indicating potential publication bias (see Supplement 1 for full details). The trim and fill method estimated that seven studies were potentially missing, resulting in an adjusted total of 23 studies. After adjustment, the corrected random-effects model showed a non-significant effect size, g = 0.14 (95% [−0.62, 0.91], p = 0.698), with high heterogeneity, Q(df=17) = 431.03, p < 0.001, I2 = 93.7%. After correction, Begg’s and Egger’s tests were no longer significant, further supporting the bias adjustment. In the subset analysis (excluding extreme- and high-RoB studies), Begg’s and Egger’s tests were again initially significant, suggesting small-study effects (see Supplements 1 and 5). The trim and fill method identified six potentially missing studies, increasing the total to 22 studies. The corrected effect size was again non-significant and small, g = 0.19 (95% CI [−0.20; 0.58], p = 0.328), with high heterogeneity, Q(df=21) = 194.1, p < 0.001; I2 = 89%. After this adjustment, both tests for publication bias became non-significant. In summary, these findings suggest that smaller studies with larger effect sizes may be overrepresented, and caution is warranted in interpreting pooled results. See Supplements 1 and 5 for additional details.

3.4. Moderation and Subgroup Analyses

Subgroup analyses and meta-regressions for age and gender were conducted using the subset of 11 studies (k = 16 comparisons) to strengthen the reliability of results. A significant difference was observed for cultural orientation, based on Hofstede’s classification (Q = 9.92, p < 0.001). In the subsequent subgroup analysis, interventions were found to be more effective in collectivistic cultures compared to individualistic ones (k = 5; b = 0.88, 95% CI [0.32, 1.45], p = 0.004). The pooled effect size was large and significant for collectivism-based studies (k = 5, g = 1.21, 95% CI [0.47; 1.95]), and small but significant for individualism-based studies (k = 11; g = 0.29, 95% CI [0.01; 0.56]). Significant differences were also found based on whether loneliness was assessed as a primary versus secondary outcome (Q = 16.80, p < 0.001). In subgroup analysis, interventions were more effective when loneliness was a secondary outcome rather than the primary focus (k = 14; b = −1.08, 95% CI [−1.96, −0.21, p = 0.018]). The pooled effect size was moderate and significant for studies measuring loneliness as a secondary outcome (k = 14; g = 0.41, 95% CI [0.11; 0.71]) but not significant for those with loneliness as the primary outcome (k = 2; g = 1.49, 95% CI [−1.35; 4.33]). Finally, a significant difference was found based on the main active component used in interventions (Q = 36.99, p < 0.001) (see Figure 3).
We interrogated the differences between the largest active component subgroups—CBT-based interventions (including ACT) and social support interventions. The meta-regression showed a trend favoring CBT, though it did not reach statistical significance (k = 7; b = 0.57, 95% [−0.04, 1.18], p = 0.062). A significant pooled effect was found for the CBT subgroup (k = 7; g = 0.73, 95% CI [0.07; 1.40]) but not for the social support subgroup (k = 6; g = 0.12, 95% CI [−0.21; 0.45]). Among smaller subgroups, significant effects were observed for art therapy (k = 1; g = 0.91, 95% CI [0.38; 1.45]), mindful movement (k = 1; g = 1.72, 95% CI [1.15; 2.30]), and physical activity (k = 1; g = 1.28, 95% CI [0.74; 1.81]). However, these results should be interpreted with caution, as they are derived from single-study estimates rather than pooled subgroup data. For the full set of subgroup analyses, see Supplement 1 (Table S4) and Supplement 6. No significant differences were found in subgroup analyses by intervention focus (person-focused vs. context-focused) (Q = 0.89, p = 0.346), delivery mode (in-person vs. virtual) (Q = 0.73, p = 0.391), or intervention format (group vs. individual) (Q = 1.39, p = 0.239) (see Supplements 1 and 6 for full results).
A significant negative association was found between the proportion of women participants and intervention effect size (b = −0.01, SE = 0.01, 95% CI [−0.03, 0.01]; t(df=14) = −1.11, p = 0.026). This indicates that a 1% increase in the proportion of women in a study sample was associated with a −0.01 reduction in the intervention effect size. However, the model explained a very small proportion of variance (R2 = 0.50%) and exhibited high residual heterogeneity (I2 = 84.37%, τ2 = 0.22, SE = 0.11) (see Supplement 7). Age was not a significant moderator (b = 0.00, SE = 0.01, 95% CI [−0.02, 0.01], t(df=14) = −0.34, p = 0.735). Two studies were excluded from the final meta-regression due to inconsistencies in their reported loneliness scores: Cohen-Mansfield et al. (2018) averaged across three different scales without providing UCLA loneliness raw scores, and Xiao et al. (2021) showed discrepancies between reported loneliness scores and the Likert scale used. The final model revealed a significant positive association between baseline loneliness scores and intervention efficacy (b = 0.04, SE = 0.01, 95% CI [0.02, 0.07]; Z = 3.36, p = < 0.001). Each one-unit increase in standardized baseline loneliness scores was associated with a 0.04-point increase in effect size. This model accounted for a substantial proportion of between-study variance (R2 = 59.04%), though residual heterogeneity remained high (QE(df=11) = 32.09, p = < 0.001; τ2 = 0.09 (SE = 0.06), (I2 = 70.19%) (see Figure 4 and Supplement 1 Table S5).

4. Discussion

This review yields three major findings with important implications for research, practice, and public health policy. First, it confirms the overall effectiveness of loneliness interventions in reducing subjective experiences of loneliness. Second, intervention outcomes vary substantially depending on the specific active components employed. Third—and most notably—interventions are significantly more effective for individuals experiencing higher baseline levels of loneliness. To our knowledge, this is the first meta-analysis to systematically examine this moderating effect across a diverse set of trials.
A key finding is the substantial heterogeneity in both intervention types and outcomes. Many interventions combined multiple active components, especially those focused on enhancing interpersonal relationships through social support. This complexity hinders the identification of potential causal mechanisms, as well as the isolation of intervention-specific effects. Moreover, considerable variation in intervention duration and intensity further challenges interpretation. While shorter interventions may benefit from novelty or greater accessibility, longer ones might gain effectiveness from providing sustained opportunities for meaningful connection—regardless of therapeutic content.
The absence of significant differences in outcomes based on delivery mode (in-person vs. virtual) and format (group vs. individual) warrants consideration. One possible explanation is the limited power of these analyses due to high heterogeneity and small sample sizes. Another potential explanation is that virtual and group-based interventions have improved in quality, accessibility, and engagement, particularly in the post-COVID-19 era. A further consideration is that some of the studies were conducted during the COVID-19 pandemic. In the context of limited social interactions, virtual interactions might have served as a temporary replacement for face-to-face interactions during the COVID-19 lockdowns. The scarcity of social interactions and the active component of the intervention may outweigh any effect of the delivery mode. The comparisons of virtual and in-person delivery modes, and group and individual intervention formats need more rigorous and controlled studies. Three-arm RCTs could be a viable design to compare the interventions with active and passive controls simultaneously, to address the potential influence of a placebo effect, and to determine which interventions are most effective for different individuals.
Among person-focused strategies, Cognitive Behavioral Therapy (CBT) showed the strongest and most consistent evidence of effectiveness, aligning with previous reviews (Masi et al., 2011; Hickin et al., 2021). Other individual-level approaches, such as art therapy and movement-based interventions, also showed promise—though evidence remains limited to single studies. Surprisingly, social support interventions—often regarded as central to loneliness mitigation—did not show a significant pooled effect. This may reflect their episodic nature, or the difficulty severely lonely individuals face in engaging in or benefiting from interpersonal interactions (Tomova et al., 2020; Gardiner et al., 2018).
In contrast, reminiscence therapies were constrained by moderate risk of bias and inflated effect sizes. One study (Li et al., 2022) reported an exceptionally large effect size and moderate risk of bias. This intervention was highly intensive and culturally tailored, combining reminiscence-based activities with traditional practices for older adults in rural China. Notably, participants were familiar with one another prior to the intervention, which may have enhanced its impact by fostering stronger group cohesion and emotional resonance. While we included this study in the main analysis for completeness, we conducted two of the five sensitivity analyses excluding it. The overall results remained robust, indicating that the main findings are not overly influenced by the study.
While contextual and nature-based interventions were underrepresented in the current evidence base, they offer considerable potential as scalable, population-level strategies. Rooted in the biophilia hypothesis (Wilson, 1986), these interventions leverage the innate human affinity for natural environments to promote psychological well-being and social connection. Emerging evidence suggests they may reduce loneliness by up to 28% (Hammoud et al., 2021). Their relevance is heightened in the context of the current ecosocial crisis, where climate change, biodiversity loss, and urban stressors intersect with growing psychosocial vulnerability. These strategies offer co-benefits for individual health, environmental restoration, and social cohesion (Whitmee et al., 2015; Lang & Rayner, 2012).
Such approaches are increasingly recognized in public health policy. The UK’s Campaign to End Loneliness and the NHS’s Social Prescribing initiative explicitly link community-based and green interventions to mental health and social integration (NHS England, 2019). Similarly, the WHO Urban Health Initiative promotes urban design that fosters social cohesion and well-being (WHO, 2016). These policy directions align with Geoffrey Rose’s (1985) classic distinction between high-risk and population-level strategies of prevention. From this perspective, CBT functions as a targeted secondary intervention for individuals experiencing severe loneliness, while universal and context-based interventions function as primary prevention, shifting the population distribution of loneliness. Importantly, integrating both levels is essential not only for clinical and economic impact, but also for ethical and public health reasons—addressing both individual suffering and structural determinants.
Our findings support this integrative model. Meta-regression results confirmed baseline loneliness severity as a strong moderator of intervention effectiveness. Although regression to the mean may partially account for this association, the analysis was restricted to a subset of studies with lower risk of bias and excluded studies with inconsistencies in their reported loneliness scores and extreme values. Tailoring intensity and delivery to individual needs is thus clinically, economically, and ethically justified. The presence of floor and ceiling effects also underscores the need to match interventions to baseline severity—both to prevent resource misallocation and to optimize measurable outcomes.
Measurement issues also warrant attention. The UCLA Loneliness Scale appeared more responsive to intervention effects than alternative tools (e.g., DJLS), consistent with previous reviews (Masi et al., 2011). A refined classification—for example, defining “severely lonely” as UCLA scores ≥65—could enhance targeting and outcome measurement by extending Perry’s (1990) classification. This recommendations are based on the loneliness baseline scores of the included studies, the scarcity of pre-settled loneliness inclusion criterion cut-offs, and the increasing rates of loneliness especially since 2020. Moreover, the frequent co-occurrence of depressive symptoms suggests a need for integrated assessment and intervention approaches.
Sociodemographic moderators were also relevant. Interventions were more effective in collectivistic cultural settings, suggesting that societal norms around interdependence may amplify intervention salience (Lykes & Kemmelmeier, 2014). One potential explanation is that collectivist cultures emphasize interdependence, group cohesion, and mutual support, which may enhance engagement with interventions that promote social connection by shaping how people respond emotionally and attitudinally to these experiences (Taras et al., 2010). Moreover, a recent meta-analysis (Wang et al., 2024) evidenced that these cultural dimensions can moderate the associations between loneliness and psychosocial outcomes, including social anxiety and stress. These findings suggest that cultural context not only shapes baseline experiences of loneliness but may also influence how individuals respond to interventions designed to alleviate it.
A modest gender effect was observed, with interventions yielding greater benefits in studies with a higher proportion of men, echoing earlier findings (Hickin et al., 2021; Masi et al., 2011). These findings highlight the importance of gender-sensitive and culturally informed intervention strategies, and future research should also consider socioeconomic position and material conditions, which are central to the social determinants of loneliness.
The higher average age in the I–C group may partly explain their greater responsiveness to certain interventions. As individuals age, their social goals shift toward emotionally meaningful interactions, a pattern described by Socioemotional Selectivity Theory (Carstensen et al., 2003). According to this framework, older adults tend to prioritize close, emotionally satisfying relationships over broader social networks. This theoretical framework may explain both the nature of loneliness in older populations and the effectiveness of interventions that emphasize emotional connection, suggesting that age-related differences in social priorities could enhance intervention outcomes.
Finally, several limitations must be acknowledged. First, the marked heterogeneity of interventions limits our ability to isolate specific mechanisms of change. Second, some studies presented dependent effect sizes (e.g., across time points), which may bias pooled estimates but could not be fully modelled due to dataset size. Third, the scarcity of follow-up data precludes firm conclusions about the long-term sustainability of intervention effects. Fourth, although the search strategy was broad, studies using newer loneliness measures or published after the review window may have been missed. Fifth, methodological quality varied across studies, with some exhibiting scoring discrepancies, extreme effect sizes, or inadequate reporting of randomization and analysis plans. For future reviews, the combined use of INSPECT-SR (Wilkinson et al., 2023) and the Cochrane Risk of Bias Tool is recommended to enhance trustworthiness and transparency.

5. Conclusions

This study offers four critical insights into the effectiveness and design of loneliness interventions. First, interventions are significantly more effective among individuals with higher levels of loneliness, underscoring the importance of stratified approaches that tailor intensity and content to severity levels. Second, CBT-based interventions emerge as robust secondary prevention strategies, particularly effective for individuals experiencing severe loneliness. Third, nature-based interventions show promise as universal primary prevention strategies, especially for mitigating mild loneliness and enhancing social connectedness; however, further evidence from well-controlled trials is needed. Fourth, outcomes appear to vary by gender, with preliminary evidence suggesting greater benefit among men, highlighting the need for culturally and demographically responsive intervention designs. Taken together, these findings support a dual-strategy public health model aligned with Rose’s prevention framework. Future research should integrate methodological rigor, equity considerations, and ecological awareness to enhance both the scientific validity and the societal relevance of loneliness interventions.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/ejihpe15070131/s1, Supplement 1: syntax, inclusion criteria, data extraction form, risk of publication bias details, tables, loneliness’ scales references; Supplement 2: Risk of Bias overall results, Figures S1 and S2; Supplement 3: Follow-up forest plot, Figure S3; Supplement 4: Forest plots of the five sensitivity analyses, Figures S4–S8; Supplement 5: Funnel plots, Figures S9 and S10; Supplement 6: Forest plots of the subgroup analyses, Figures S11–S15; Supplement 7: Association between gender and intervention efficacy, Figure S16.

Author Contributions

Conceptualization, A.Z.-S., A.G.-Á. and P.R.; data curation, A.Z.-S.; formal analysis, A.Z.-S., A.V., and R.G.; investigation, A.Z.-S., A.V., and R.C.; methodology, A.Z.-S., A.G.-Á., and P.R.; validation, F.M. and J.B.; writing—original draft preparation, A.Z.-S.; writing—review and editing, A.Z.-S., A.G.-Á., R.C., P.R., F.M., and J.B.; visualization, A.Z.-S. and A.V.; supervision, P.R. and J.B. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Fundación Caja Navarra (FBCAN) and the Public University of Navarre, Resolution number 589/2021, 2021 (A.Z.-S.); the Aker Scholarship & Saven European Scholarship (A.V.); the European Union-NextGeneration, Resolution 1692/2022 (P.R.); the Portuguese Foundation for Science and Technology Individual Call to Scientific Employment Stimulus—3rd Edition [2019-2020.02059-CEECIND], and by national funds of the Portuguese Foundation for Science and Technology and the Portuguese Ministry of Education and Science co-financed by FEDER through COMPETE2020 under the PT2020 Partnership Agreement [UID/PSI/01662/2013] (A.G.-Á.); and the Ph.D. Grant of the Portuguese Foundation for Science and Technology [SFRH/BD/5099/2020] (R.G.). The APC were waived by the journal following an invitation extended to the corresponding author.

Data Availability Statement

The raw data extracted from the included studies and R code are openly available in the Open Science Framework (OFS) and can be accessed at https://osf.io/d3rak/?view_only=3dd83d8426884b6885f547e78953e853 (accessed on 30 June 2025).

Conflicts of Interest

The authors declare no conflicts of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

Abbreviations

The following abbreviations are used in this manuscript:
ACMain Active Component used in the intervention
ACTAcceptance and Commitment Therapy
ATArt Therapy
CBTCognitive Behavioral Therapy
CIConfidence Interval
CGControl Group
CollCollectivistic
DADelayed Access
DJLSde Jong–Gierveld Loneliness scale
D-SSDigital Social Support
I-CIntervention vs. Control
I-IIntervention vs. Intervention
IndIndividualistic
INSPECT-SRINveStigating ProblEmatic Clinical Trials in Systematic Reviews
IPTInternet-Based Interpersonal Therapy
MMMindful Movement
MedtMeditation
NHSNational Health System
NINo Intervention
NNTNumber Needed to Treat
N-OnNon-online Interaction Group
N/rNumber of sessions not reported
n/rLength or age not reported
ODOnline Diary
OSFOpen Science Framework
PAPhysical Activity
PRISMAPreferred Reporting Items for Systematic Reviews and Meta-Analyses
RCTRandomized Controlled Trial
RoBRisk of Bias
SSSocial Support
SMDStandardized Mean Differences
UCLA (v3)University of Los Angeles Loneliness scale (version three)
UKUnited Kingdom
VIVirtual Interaction with Strangers
WLWaiting List
WHOWorld Health Organization
YT-iYouTube interaction

References

* articles included in the systematic review. ** articles also included in the meta-analysis.
  1. Arsenijevic, J., & Groot, W. (2018). Does household help prevent loneliness among the elderly? An evaluation of a policy reform in the Netherlands. BMC Public Health, 18(1), 1104. [Google Scholar] [CrossRef]
  2. Astell-Burt, T., Hartig, T., Putra, I. G. N. E., Walsan, R., Dendup, T., & Feng, X. (2022). Green space and loneliness: A systematic review with theoretical and methodological guidance for future research. Science of the Total Environment, 847, 157521. [Google Scholar] [CrossRef]
  3. **Aydın, M., & Kutlu, F. Y. (2021). The effect of group art therapy on loneliness and hopelessness levels of older adults living alone: A randomized controlled study. Florence Nightingale Journal of Nursing, 29(3), 271–284. [Google Scholar] [CrossRef]
  4. Barreto, M., Victor, C., Hammond, C., Eccles, A., Richins, M. T., & Qualter, P. (2021). Loneliness around the world: Age, gender, and cultural differences in loneliness. Personality and Individual Differences, 169, 110066. [Google Scholar] [CrossRef]
  5. Barton, J., & Pretty, J. (2010). What is the best dose of nature and green exercise for improving mental health? A multi-study analysis. Environmental Science & Technology, 44(10), 3947–3955. [Google Scholar] [CrossRef]
  6. Begg, C. B., & Mazumdar, M. (1994). Operating characteristics of a rank correlation test for publication bias. Biometrics, 50(4), 1088–1101. [Google Scholar] [CrossRef]
  7. **Bruehlman-Senecal, E., Hook, C. J., Pfeifer, J. H., FitzGerald, C., Davis, B., Delucchi, K. L., Haritatos, J., & Ramo, D. E. (2020). Smartphone app to address loneliness among college students: Pilot randomized controlled trial. JMIR Mental Health, 7(10), e21496. [Google Scholar] [CrossRef]
  8. Cacioppo, J. T., & Hawkley, L. C. (2009). Perceived social isolation and cognition. Trends in Cognitive Sciences, 13(10), 447–454. [Google Scholar] [CrossRef]
  9. Cacioppo, J. T., Hawkley, L. C., Crawford, L. E., Ernst, J. M., Burleson, M. H., Kowalewski, R. B., Malarkey, W. B., Van Cauter, E., & Berntson, G. G. (2002). Loneliness and health: Potential mechanisms. Psychosomatic Medicine, 64(3), 407–417. [Google Scholar] [CrossRef]
  10. Cacioppo, J. T., & Patrick, W. (2008). Loneliness: Human nature and the need for social connection. WW Norton & Company. [Google Scholar]
  11. Cacioppo, S., & Cacioppo, J. T. (2012). Decoding the invisible forces of social connections. Frontiers in Integrative Neuroscience, 6, 51. [Google Scholar] [CrossRef]
  12. Cacioppo, S., Capitanio, J. P., & Cacioppo, J. T. (2014). Toward a neurology of loneliness. Psychological Bulletin, 140(6), 1464–1504. [Google Scholar] [CrossRef]
  13. Carstensen, L. L., Fung, H. H., & Charles, S. T. (2003). Socioemotional selectivity theory and the regulation of emotion in the second half of life. Motivation and Emotion, 27, 103–123. [Google Scholar] [CrossRef]
  14. Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155–159. [Google Scholar] [CrossRef]
  15. Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98(2), 310–357. [Google Scholar] [CrossRef]
  16. **Cohen-Mansfield, J., Hazan, H., Lerman, Y., Shalom, V., Birkenfeld, S., & Cohen, R. (2018). Efficacy of the I-SOCIAL intervention for loneliness in old age: Lessons from a randomized controlled trial. Journal of Psychiatric Research, 99, 69–75. [Google Scholar] [CrossRef]
  17. Cohen-Mansfield, J., & Perach, R. (2015). Interventions for alleviating loneliness among older persons: A critical review. American Journal of Health Promotion, 29(3), e109–e125. [Google Scholar] [CrossRef]
  18. Cuijpers, P. (2016). Meta-analyses in mental health research: A practical guide (Vol. 15). Vrije Universiteit Amsterdam. [Google Scholar]
  19. de Oliveira, A. M., Radanovic, M., de Mello, P. C., Buchain, P. C., Vizzotto, A. D., Celestino, D. L., Stella, F., Piersol, C. V., & Forlenza, O. V. (2015). Nonpharmacological interventions to reduce behavioral and psychological symptoms of dementia: A systematic review. BioMed Research International, 2015, 218980. [Google Scholar] [CrossRef]
  20. DerSimonian, R., & Kacker, R. (2007). Random-effects model for meta-analysis of clinical trials: An update. Contemporary Clinical Trials, 28(2), 105–114. [Google Scholar] [CrossRef]
  21. DerSimonian, R., & Laird, N. (2015). Meta-analysis in clinical trials revisited. Contemporary Clinical Trials, 45, 139–145. [Google Scholar] [CrossRef]
  22. Duval, S., & Tweedie, R. (2000). Trim and fill: A simple funnel-plot–based method of testing and adjusting for publication bias in meta-analysis. Biometrics, 56(2), 455–463. [Google Scholar] [CrossRef]
  23. Eccles, A. M., & Qualter, P. (2021). Alleviating loneliness in young people—A meta-analysis of interventions. Child and Adolescent Mental Health, 26(1), 17–33. [Google Scholar] [CrossRef]
  24. Egger, M., Smith, G. D., Schneider, M., & Minder, C. (1997). Bias in meta-analysis detected by a simple, graphical test. British Medical Journal, 315(7109), 629–634. [Google Scholar] [CrossRef]
  25. Fu, Z., Yan, M., & Meng, C. (2022). The effectiveness of remote delivered intervention for loneliness reduction in older adults: A systematic review and meta-analysis. Frontiers in Psychology, 13, 935544. [Google Scholar] [CrossRef]
  26. Gardiner, C., Geldenhuys, G., & Gott, M. (2018). Interventions to reduce social isolation and loneliness among older people: An integrative review. Health & Social Care in the Community, 26(2), 147–157. [Google Scholar] [CrossRef]
  27. Hammoud, R., Tognin, S., Bakolis, I., Ivanova, D., Fitzpatrick, N., Burgess, L., Smythe, M., Gibbons, J., Davidson, N., & Mechelli, A. (2021). Lonely in a crowd: Investigating the association between overcrowding and loneliness using smartphone technologies. Scientific Reports, 11(1), 24134. [Google Scholar] [CrossRef]
  28. Hickin, N., Käll, A., Shafran, R., Sutcliffe, S., Manzotti, G., & Langan, D. (2021). The effectiveness of psychological interventions for loneliness: A systematic review and meta-analysis. Clinical Psychology Review, 88, 102066. [Google Scholar] [CrossRef]
  29. Higgins, J. P., Altman, D. G., & Sterne, J. A. (2011). Chapter 8: Assessing risk of bias in included studies. Cochrane Handbook for Systematic Reviews of Interventions Version, 5, 182–228. [Google Scholar]
  30. Higgins, J. P., Thompson, S. G., Deeks, J. J., & Altman, D. G. (2003). Measuring inconsistency in meta-analyses. British Medical Journal, 327(7414), 557–560. [Google Scholar] [CrossRef]
  31. Higgins, J. P. T., Thomas, J., Chandler, J., Cumpston, M., Li, T., Page, M. J., & Welch, V. A. (Eds.). (2023). Cochrane handbook for systematic reviews of interventions version 6.4. (Updated August 2023) (2nd ed.). Cochrane. Available online: www.training.cochrane.org/handbook (accessed on 20 March 2023).
  32. Hoang, P., King, J. A., Moore, S., Moore, K., Reich, K., Sidhu, H., Tan, C. V., Whaley, C., & McMillan, J. (2022). Interventions associated with reduced loneliness and social isolation in older adults: A systematic review and meta-analysis. JAMA Network Open, 5(10), e2236676. [Google Scholar] [CrossRef]
  33. Hofstede, G., Hofstede, G. J., & Minkov, M. (2010). Cultures and organizations: Software of the mind (3rd ed.). McGraw-Hill Professional. [Google Scholar]
  34. Holt-Lunstad, J. (2017). The potential public health relevance of social isolation and loneliness: Prevalence, epidemiology, and risk factors. Public Policy & Aging Report, 27(4), 127–130. [Google Scholar] [CrossRef]
  35. Holt-Lunstad, J., Smith, T. B., Baker, M., Harris, T., & Stephenson, D. (2015). Loneliness and social isolation as risk factors for mortality: A meta-analytic review. Perspectives on Psychological Science, 10(2), 227–237. [Google Scholar] [CrossRef]
  36. Holt-Lunstad, J., Smith, T. B., & Layton, J. B. (2010). Social relationships and mortality risk: A meta-analytic review. PLoS Med, 7(7), e1000316. [Google Scholar] [CrossRef]
  37. Hom, M. A., Chu, C., Rogers, M. L., & Joiner, T. E. (2020). A Meta-Analysis of the Relationship Between Sleep Problems and Loneliness. Clinical Psychological Science, 8(5), 799–824. [Google Scholar] [CrossRef]
  38. *Hussain, Z., Ferreira, R., & Kuss, D. J. (2023). The feasibility of smartphone interventions to decrease problematic use of social networking sites: A randomised controlled trial. Psychiatry Research Communications, 100132. [Google Scholar] [CrossRef]
  39. IntHout, J., Ioannidis, J. P., Rovers, M. M., & Goeman, J. J. (2016). Plea for routinely presenting prediction intervals in meta-analysis. BMJ Open, 6(7), e010247. [Google Scholar] [CrossRef]
  40. Jarvis, M. A., Padmanabhanunni, A., Balakrishna, Y., & Chipps, J. (2020). The effectiveness of interventions addressing loneliness in older persons: An umbrella review. International Journal of Africa Nursing Sciences, 12, 100177. [Google Scholar] [CrossRef]
  41. **Kahlon, M. K., Aksan, N., Aubrey, R., Clark, N., Cowley-Morillo, M., Jacobs, E. A., Mundhenk, R., Sebastian, K. R., & Tomlinson, S. (2021). Effect of layperson-delivered, empathy-focused program of telephone calls on loneliness, depression, and anxiety among adults during the COVID-19 pandemic: A randomized clinical trial. Journal of American Medical Association Psychiatry, 78(6), 616–622. [Google Scholar] [CrossRef]
  42. **Käll, A., & Andersson, G. (2023). Knowledge acquisition following internet-based cognitive behavioural therapy for loneliness—A secondary analysis of a randomised controlled trial. Journal of Behavior Therapy and Experimental Psychiatry, 81, 101872. [Google Scholar] [CrossRef]
  43. **Käll, A., Backlund, U., Shafran, R., & Andersson, G. (2020a). Lonesome no more? A two-year follow-up of internet-administered cognitive behavioral therapy for loneliness. Internet Interventions, 19, 100301. [Google Scholar] [CrossRef]
  44. **Käll, A., Bäck, M., Welin, C., Åman, H., Bjerkander, R., Wänman, M., Lindegaard, T., Berg, M., Moche, H., Shafran, R., & Andersson, G. (2021). Therapist-guided internet-based treatments for loneliness: A randomized controlled three-arm trial comparing cognitive behavioral therapy and interpersonal psychotherapy. Psychotherapy and Psychosomatics, 90(5), 351–358. [Google Scholar] [CrossRef]
  45. **Käll, A., Jägholm, S., Hesser, H., Andersson, F., Mathaldi, A., Norkvist, B. T., Shafran, R., & Andersson, G. (2020b). Internet-based cognitive behavior therapy for loneliness: A pilot randomized controlled trial. Behavior Therapy, 51(1), 54–68. [Google Scholar] [CrossRef]
  46. Kraemer, H. C., & Kupfer, D. J. (2006). Size of treatment effects and their importance to clinical research and practice. Biological Psychiatry, 59(11), 990–996. [Google Scholar] [CrossRef] [PubMed]
  47. Lang, T., & Rayner, G. (2012). Ecological public health: The 21st century’s big idea? BMJ, 345, e5466. [Google Scholar] [CrossRef] [PubMed]
  48. *Lanser, I., & Eisenberger, N. I. (2023). Prosocial behavior reliably reduces loneliness: An investigation across two studies. Emotion, 23(6), 1781. [Google Scholar] [CrossRef]
  49. Laupacis, A., Sackett, D. L., & Roberts, R. S. (1988). An assessment of clinically useful measures of the consequences of treatment. New England Journal of Medicine, 318(26), 1728–1733. [Google Scholar] [CrossRef] [PubMed]
  50. Leavell, M. A., Leiferman, J. A., Gascon, M., Braddick, F., Gonzalez, J. C., & Litt, J. S. (2019). Nature-Based Social Prescribing in Urban Settings to Improve Social Connectedness and Mental Well-being: A Review. Current Environment Health Report, 6, 297–308. [Google Scholar] [CrossRef]
  51. **Li, S., Dai, Y., Zhou, Y., Zhang, J., & Zhou, C. (2022). Efficacy of group reminiscence therapy based on Chinese traditional festival activities (CTFA-GRT) on loneliness and perceived stress of rural older adults living alone in China: A randomized controlled trial. Aging & Mental Health, 26(7), 1377–1384. [Google Scholar] [CrossRef]
  52. *Lindsay, E. K., Young, S., Brown, K. W., Smyth, J. M., & Creswell, J. D. (2019). Mindfulness training reduces loneliness and increases social contact in a randomized controlled trial. Proceedings of the National Academy of Sciences, 116(9), 3488–3493. [Google Scholar] [CrossRef]
  53. *Liu, C. W., Hsieh, P. L., Yang, S. Y., Lin, Y. L., & Wang, J. Y. (2023). The effectiveness of facilitator-led remote interactive intervention for loneliness, quality of life, and social support among seniors in communities during the COVID-19 pandemic: A randomized controlled experiment. Psychology & Health, 39, 1675–1688. [Google Scholar] [CrossRef]
  54. **Liu, S., Wegner, L., Haucke, M., Gates, J., Adam, M., & Bärnighausen, T. (2023). An entertainment-education video and written messages to alleviate loneliness in Germany: Pilot randomized controlled study. Journal of Medical Internet Research Formative Research, 7(1), e43036. [Google Scholar] [CrossRef]
  55. Lykes, V. A., & Kemmelmeier, M. (2014). What predicts loneliness? Cultural difference between individualistic and collectivistic societies in Europe. Journal of Cross-Cultural Psychology, 45(3), 468–490. [Google Scholar] [CrossRef]
  56. Maes, M., Qualter, P., Vanhalst, J., Van den Noortgate, W., & Goossens, L. (2019). Gender differences in loneliness across the lifespan: A meta–analysis. European Journal of Personality, 33(6), 642–654. [Google Scholar] [CrossRef]
  57. **Mahmoudpour, A., Rayesh, N., Ghanbarian, E., & Rezaee, M. (2021). Effectiveness of acceptance and commitment therapy (ACT) on emotional regulation and loneliness of divorced women in Iran. Journal of Marital and Family Therapy, 47(4), 831–842. [Google Scholar] [CrossRef]
  58. Masi, C. M., Chen, H. Y., Hawkley, L. C., & Cacioppo, J. T. (2011). A meta-analysis of interventions to reduce loneliness. Personality and Social Psychology Review, 15(3), 219–266. [Google Scholar] [CrossRef] [PubMed]
  59. Mezuk, B., Rock, A., Lohman, M. C., & Choi, M. (2014). Suicide risk in long-term care facilities: A systematic review. International Journal of Geriatric Psychiatry, 29(12), 1198–1211. [Google Scholar] [CrossRef]
  60. **Morgado, B., Fonseca, C., Afonso, A., Amaro, P., Lopes, M., & de Pinho, L. G. (2023). “Geriatric Proximity” Intervention in COVID-19 Context: Contribution to Reducing Loneliness and Improving Affectivity. Geriatrics, 8(2), 39. [Google Scholar] [CrossRef]
  61. **Myhre, J. W., Mehl, M. R., & Glisky, E. L. (2017). Cognitive benefits of online social networking for healthy older adults. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 72(5), 752–760. [Google Scholar] [CrossRef]
  62. NHS England. (2019). Social prescribing and community-based support: Summary guide. Available online: https://www.england.nhs.uk/personalisedcare/social-prescribing/ (accessed on 4 January 2025).
  63. Nowland, R., Robinson, S. J., Bradley, B. F., Summers, V., & Qualter, P. (2018). Loneliness, HPA stress reactivity and social threat sensitivity: Analyzing naturalistic social challenges. Scandinavian Journal of Psychology, 59(5), 540–546. [Google Scholar] [CrossRef]
  64. Ouzzani, M., Hammady, H., Fedorowicz, Z., & Elmagarmid, A. (2016). Rayyan—A web and mobile app for systematic reviews. Systematic Reviews, 5, 210. [Google Scholar] [CrossRef]
  65. Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R., Glanville, J., Grimshaw, J. M., Hróbjartsson, A., Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E., McDonald, S., … Moher, D. (2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. British Medical Journal, 372, n71. [Google Scholar] [CrossRef]
  66. Perry, G. R. (1990). Loneliness and coping among tertiary-level adult cancer patients in the home. Cancer Nursing, 13(5), 293–302. [Google Scholar] [CrossRef] [PubMed]
  67. *Razani, N., Morshed, S., Kohn, M. A., Wells, N. M., Thompson, D., Alqassari, M., Agodi, A., & Rutherford, G. W. (2018). Effect of park prescriptions with and without group visits to parks on stress reduction in low-income parents: SHINE randomized trial. PLoS ONE, 13(2), e0192921. [Google Scholar] [CrossRef] [PubMed]
  68. **Ristolainen, H., Kannasoja, S., Tiilikainen, E., Hakala, M., Närhi, K., & Rissanen, S. (2020). Effects of ‘participatory group-based care management’ on wellbeing of older people living alone: A randomized controlled trial. Archives of Gerontology and Geriatrics, 89, 104095. [Google Scholar] [CrossRef]
  69. Rose, G. (1985). Sick individuals and sick populations. International Journal of Epidemiology, 14(1), 32–38. [Google Scholar] [CrossRef]
  70. Ruisoto, P., López-Guerra, V. M., Lopez-Nunez, C., Sanchez-Puertas, R., Paladines-Costa, M. B., & Pineda-Cabrera, N. J. (2022). Transdiagnostic model of psychological factors and sex differences in depression in a large sample of Ecuador. International Journal of Clinical and Health Psychology, 22(3), 100322. [Google Scholar] [CrossRef]
  71. **Shapira, S., Yeshua-Katz, D., Cohn-Schwartz, E., Aharonson-Daniel, L., Sarid, O., & Clarfield, A. M. (2021). A pilot randomized controlled trial of a group intervention via Zoom to relieve loneliness and depressive symptoms among older persons during the COVID-19 outbreak. Internet Interventions, 24, 100368. [Google Scholar] [CrossRef] [PubMed]
  72. Steptoe, A., Shankar, A., Demakakos, P., & Wardle, J. (2013). Social isolation, loneliness, and all-cause mortality in older men and women. Proceedings of the National Academy of Sciences, 110(15), 5797–5801. [Google Scholar] [CrossRef]
  73. Sterne, J. A., Savović, J., Page, M. J., Elbers, R. G., Blencowe, N. S., Boutron, I., Cates, C. J., Cheng, H. Y., Corbett, M. S., Eldridge, S. M., Emberson, J. R., Hernán, M. A., Hopewell, S., Asbjørn, H., Junqueira, D. R., Jüni, P., Kirkham, J. J., Lasserson, T., Li, T., … Higgins, J. P. (2019). RoB 2: A revised tool for assessing risk of bias in randomised trials. British Medical Journal, 366. [Google Scholar] [CrossRef]
  74. **Tabrizi, F. M., Radfar, M., & Taei, Z. (2016). Effects of supportive-expressive discussion groups on loneliness, hope and quality of life in breast cancer survivors: A randomized control trial. Psycho-Oncology, 25(9), 1057–1063. [Google Scholar] [CrossRef]
  75. Tan, X. R., Wilson, I. M., Tay, P. K. C., Win, P. P. S., Song, C. F., & Wee, S. L. (2024). Mapping of technological strategies for reducing social isolation in homebound older adults: A scoping review. Archives of Gerontology and Geriatrics, 125, 105478. [Google Scholar] [CrossRef]
  76. Taras, V., Kirkman, B. L., & Steel, P. (2010). Examining the impact of Culture’s consequences: A three-decade, multilevel, meta-analytic review of Hofstede’s cultural value dimensions. Journal of Applied Psychology, 95(3), 405–439. [Google Scholar] [CrossRef]
  77. **Thimmapuram, J., Pargament, R., Bell, T., Schurk, H., & Madhusudhan, D. K. (2021). Heartfulness meditation improves loneliness and sleep in physicians and advance practice providers during COVID-19 pandemic. Hospital Practice, 49(3), 194–202. [Google Scholar] [CrossRef] [PubMed]
  78. Tomova, L., Wang, K. L., Thompson, T., Matthews, G. A., Takahashi, A., Tye, K. M., & Saxe, R. (2020). Acute social isolation evokes midbrain craving responses similar to hunger. Nature Neuroscience, 23(12), 1597–1605. [Google Scholar] [CrossRef] [PubMed]
  79. Valtorta, N. K., Kanaan, M., Gilbody, S., & Hanratty, B. (2018). Loneliness, social isolation and risk of cardiovascular disease in the English Longitudinal Study of Ageing. European Journal of Preventive Cardiology, 25(13), 1387–1396. [Google Scholar] [CrossRef]
  80. Vanhalst, J., Klimstra, T. A., Luyckx, K., Scholte, R. H., Engels, R. C., & Goossens, L. (2012). The interplay of loneliness and depressive symptoms across adolescence: Exploring the role of personality traits. Journal of Youth and Adolescence, 41, 776–787. [Google Scholar] [CrossRef]
  81. *Van Orden, K. A., Conwell, Y., Chapman, B. P., Buttaccio, A., VanBergen, A., Beckwith, E., Santee, A., Rowe, J., Palumbos, D., Williams, G., Messing, S., Sörensen, S., & Tu, X. (2022). The helping older people engage (HOPE) study: Protocol & COVID modifications for a randomized trial. Contemporary Clinical Trials Communications, 30, 101040. [Google Scholar] [CrossRef] [PubMed]
  82. Viechtbauer, W. (2010). Conducting Meta-Analyses in R with the metafor Package. Journal of Statistical Software, 36(3), 1–48. [Google Scholar] [CrossRef]
  83. Wang, J. A., Wang, H. F., Cao, B., Lei, X., & Long, C. (2024). Cultural dimensions moderate the association between loneliness and mental health during adolescence and younger adulthood: A systematic review and meta-analysis. Journal of Youth and Adolescence, 53(8), 1774–1819. [Google Scholar] [CrossRef]
  84. Whitmee, S., Haines, A., Beyrer, C., Boltz, F., Capon, A. G., de Souza Dias, B. F., Ezeh, A., Frumkin, H., Gong, P., Head, P., Horton, R., Mace, G. M., Marten, R., Myers, S. S., Nishtar, S., Osofsky, S. A., Pattanayak, S. K., Pongsiri, M. J., Romanelli, C., … Yach, D. (2015). Safeguarding human health in the Anthropocene epoch: The Rockefeller Foundation–Lancet Commission on planetary health. The Lancet, 386(10007), 1973–2028. [Google Scholar] [CrossRef]
  85. Wilkinson, J., Heal, C., Antoniou, G. A., & Kirkham, J. J. (2023, September 4–6). Introducing INSPECT-SR: A tool for detecting problematic randomised controlled trials in health systematic reviews. Workshop presented at the Cochrane Colloquium 2023, London, UK. Available online: https://abstracts.cochrane.org/2023-london/introducing-inspect-sr-tool-detecting-problematic-randomised-controlled-trials-health (accessed on 11 January 2025).
  86. Wilson, E. O. (1986). Biophilia. Harvard University Press. [Google Scholar]
  87. World Health Organization. (2016). Urban health initiative. Available online: https://www.who.int/initiatives/urban-health-initiative (accessed on 11 April 2025).
  88. World Health Organization. (2020). Mental health and psychological resilience during the COVID-19 pandemic. Available online: https://www.who.int/europe/news/item/27-03-2020-mental-health-and-psychological-resilience-during-the-covid-19-pandemic (accessed on 11 April 2023).
  89. **Xiao, T., Jiao, C., Yao, J., Yang, L., Zhang, Y., Liu, S., Grabovac, I., Yu, Q., Kong, Z., Yu, J., & Zhang, J. (2021). Effects of basketball and Baduanjin exercise interventions on problematic smartphone use and mental health among college students: A randomized controlled trial. Evidence-Based Complementary and Alternative Medicine, 2021, 8880716. [Google Scholar] [CrossRef]
  90. *Yang, S. Y., Yang, C. C., Lee, Y. C., Hsieh, P. L., & Lin, Y. L. (2023). Investigating the effectiveness of online interactive courses on loneliness and quality of life of older adults in the community during the COVID-19 pandemic: A pilot study and a randomized controlled trial. Geriatrics & Gerontology International, 23(2), 91–97. [Google Scholar] [CrossRef]
  91. **Zhang, W., Jhang, J., & Greenwell, M. R. (2023). Effects of replay and rehearsal expressive writing on mental health: A randomized controlled trial. Journal of Mental Health, 32(3), 582–591. [Google Scholar] [CrossRef] [PubMed]
Ejihpe 15 00131 g001
Figure 1. PRISMA flowchart of the study selection process. Of the 25 studies included in the narrative synthesis, 16 were retained for the meta-analysis due to the high heterogeneity of their control groups. * These 16 studies yielded a total of 21 effect sizes and are marked with double asterisks in the references list. “Citation searching” refers to backward reference checking of the included studies.
Figure 1. PRISMA flowchart of the study selection process. Of the 25 studies included in the narrative synthesis, 16 were retained for the meta-analysis due to the high heterogeneity of their control groups. * These 16 studies yielded a total of 21 effect sizes and are marked with double asterisks in the references list. “Citation searching” refers to backward reference checking of the included studies.
Ejihpe 15 00131 g001
Ejihpe 15 00131 g002
Figure 2. Forest plot of the effect sizes of interventions vs. control groups (16 Intervention–Control [I-C] RCTs; k = 21) at post intervention. A random-effects model with Hartung–Knapp adjustment was used to provide more accurate standard errors. Tabrizi et al. (2016) (k = 1); Myhre et al. (2017) a (k = 1); Cohen-Mansfield et al. (2018) (k = 1); Bruehlman-Senecal et al. (2020) (k = 1); Ristolainen et al. (2020) (k = 1); Aydın and Kutlu (2021) (k = 1); Kahlon et al. (2021) (k = 2; Kahlon et al., 2021 a,b); Li et al. (2022) (k = 1); Mahmoudpour et al. (2021) (k = 1); Shapira et al. (2021) (k = 1); Thimmapuram et al. (2021); Xiao et al. (2021) (k = 2; Xiao et al., 2021 a,b); Käll et al. (2021) (k = 2; Käll et al., 2021 a,b); Morgado et al. (2023) (k = 1); Käll and Andersson (2023) (k = 1); S. Liu et al. (2023) (k = 3; S. Liu et al., 2023 a–c). Abbreviations: AC = Main Active Component used in each intervention; AT = Art Therapy; CBT = Cognitive Behavioral Therapy; CI = Confidence Interval; D_SS = Digital Social Support; g Post = Hedges’ g at post-intervention; I2 = Heterogeneity; IPT = Internet-based Interpersonal Therapy; Medt = Meditation; MM = Mindful Movement; nExp = Experimental group sample size; nCon = Control group sample size; PA = Physical Activity; RT = Reminiscence Therapy; SS = Social Support; k = number of comparisons per intervention; a = first comparison, b = second comparison, c = third comparison.
Figure 2. Forest plot of the effect sizes of interventions vs. control groups (16 Intervention–Control [I-C] RCTs; k = 21) at post intervention. A random-effects model with Hartung–Knapp adjustment was used to provide more accurate standard errors. Tabrizi et al. (2016) (k = 1); Myhre et al. (2017) a (k = 1); Cohen-Mansfield et al. (2018) (k = 1); Bruehlman-Senecal et al. (2020) (k = 1); Ristolainen et al. (2020) (k = 1); Aydın and Kutlu (2021) (k = 1); Kahlon et al. (2021) (k = 2; Kahlon et al., 2021 a,b); Li et al. (2022) (k = 1); Mahmoudpour et al. (2021) (k = 1); Shapira et al. (2021) (k = 1); Thimmapuram et al. (2021); Xiao et al. (2021) (k = 2; Xiao et al., 2021 a,b); Käll et al. (2021) (k = 2; Käll et al., 2021 a,b); Morgado et al. (2023) (k = 1); Käll and Andersson (2023) (k = 1); S. Liu et al. (2023) (k = 3; S. Liu et al., 2023 a–c). Abbreviations: AC = Main Active Component used in each intervention; AT = Art Therapy; CBT = Cognitive Behavioral Therapy; CI = Confidence Interval; D_SS = Digital Social Support; g Post = Hedges’ g at post-intervention; I2 = Heterogeneity; IPT = Internet-based Interpersonal Therapy; Medt = Meditation; MM = Mindful Movement; nExp = Experimental group sample size; nCon = Control group sample size; PA = Physical Activity; RT = Reminiscence Therapy; SS = Social Support; k = number of comparisons per intervention; a = first comparison, b = second comparison, c = third comparison.
Ejihpe 15 00131 g002
Ejihpe 15 00131 g003
Figure 3. Forest plot of the effect sizes of the interventions vs. controls post-intervention: Subgroup analysis by active component (k = 16). A random-effects model with Hartung–Knapp adjustment was applied to provide more accurate standard error estimates. Tabrizi et al. (2016) (k = 1); Cohen-Mansfield et al. (2018) (k = 1); Bruehlman-Senecal et al. (2020) (k = 1); Aydın and Kutlu (2021) (k = 1); Kahlon et al. (2021) (k = 2; Kahlon et al., 2021 a,b); Li et al. (2022) (k = 1); Mahmoudpour et al. (2021) (k = 1); Shapira et al. (2021) (k = 1); Xiao et al. (2021) (k = 2; Xiao et al., 2021 a,b); Käll et al. (2021) (k = 2; Käll et al., 2021 a,b); Käll and Andersson (2023) (k = 1); S. Liu et al. (2023) (k = 3; S. Liu et al., 2023 a–c). Abbreviations: AC = Main Active Component used in each intervention; AT = Art Therapy; CBT = Cognitive Behavioral Therapy; CI = Confidence Interval; df = Degrees of Freedom; g Post = Hedges’ g at post-intervention; I2 = Heterogeneity; IPT = Internet-based Interpersonal Therapy; MM = Mindful Movement; nExp = Experimental group sample size; nCon = Control group sample size; PA = Physical Activity; SS = Social Support; k = number of comparisons per intervention; a = first comparison, b = second comparison, c = third comparison.
Figure 3. Forest plot of the effect sizes of the interventions vs. controls post-intervention: Subgroup analysis by active component (k = 16). A random-effects model with Hartung–Knapp adjustment was applied to provide more accurate standard error estimates. Tabrizi et al. (2016) (k = 1); Cohen-Mansfield et al. (2018) (k = 1); Bruehlman-Senecal et al. (2020) (k = 1); Aydın and Kutlu (2021) (k = 1); Kahlon et al. (2021) (k = 2; Kahlon et al., 2021 a,b); Li et al. (2022) (k = 1); Mahmoudpour et al. (2021) (k = 1); Shapira et al. (2021) (k = 1); Xiao et al. (2021) (k = 2; Xiao et al., 2021 a,b); Käll et al. (2021) (k = 2; Käll et al., 2021 a,b); Käll and Andersson (2023) (k = 1); S. Liu et al. (2023) (k = 3; S. Liu et al., 2023 a–c). Abbreviations: AC = Main Active Component used in each intervention; AT = Art Therapy; CBT = Cognitive Behavioral Therapy; CI = Confidence Interval; df = Degrees of Freedom; g Post = Hedges’ g at post-intervention; I2 = Heterogeneity; IPT = Internet-based Interpersonal Therapy; MM = Mindful Movement; nExp = Experimental group sample size; nCon = Control group sample size; PA = Physical Activity; SS = Social Support; k = number of comparisons per intervention; a = first comparison, b = second comparison, c = third comparison.
Ejihpe 15 00131 g003
Ejihpe 15 00131 g004
Figure 4. Association between baseline loneliness severity and intervention efficacy (k = 16). The bubble plot depicts the association between standardized baseline loneliness scores (x-axis) and intervention effect sizes (Hedges’ g) (y-axis). Each blue bubble corresponds to a study. Bubble size reflects the inverse of the study’s standard error, meaning larger bubbles indicate studies with greater precision. The gray trend depicts a positive association, revealing that higher levels of baseline loneliness are linked to larger intervention effects. The legend, “Study Weight (Inverse SE)”, categorizes study precision from “Low” to “High”. Tabrizi et al. (2016) (k = 1); Bruehlman-Senecal et al. (2020) (k = 1); Aydın and Kutlu (2021) (k = 1); Kahlon et al. (2021) (k = 2; Kahlon et al., 2021 a,b); Shapira et al. (2021) (k = 1); Mahmoudpour et al. (2021) (k = 1); Käll et al. (2021) (k = 2; Käll et al., 2021 a,b); Käll and Andersson (2023) (k = 1); S. Liu et al. (2023) (k = 3; S. Liu et al., 2023 a–c); k = number of comparisons per intervention; a = first comparison, b = second comparison, c = third comparison.
Figure 4. Association between baseline loneliness severity and intervention efficacy (k = 16). The bubble plot depicts the association between standardized baseline loneliness scores (x-axis) and intervention effect sizes (Hedges’ g) (y-axis). Each blue bubble corresponds to a study. Bubble size reflects the inverse of the study’s standard error, meaning larger bubbles indicate studies with greater precision. The gray trend depicts a positive association, revealing that higher levels of baseline loneliness are linked to larger intervention effects. The legend, “Study Weight (Inverse SE)”, categorizes study precision from “Low” to “High”. Tabrizi et al. (2016) (k = 1); Bruehlman-Senecal et al. (2020) (k = 1); Aydın and Kutlu (2021) (k = 1); Kahlon et al. (2021) (k = 2; Kahlon et al., 2021 a,b); Shapira et al. (2021) (k = 1); Mahmoudpour et al. (2021) (k = 1); Käll et al. (2021) (k = 2; Käll et al., 2021 a,b); Käll and Andersson (2023) (k = 1); S. Liu et al. (2023) (k = 3; S. Liu et al., 2023 a–c); k = number of comparisons per intervention; a = first comparison, b = second comparison, c = third comparison.
Ejihpe 15 00131 g004
Table 1. Characteristics of the 25 reviewed studies.
Table 1. Characteristics of the 25 reviewed studies.
Study
RoB: Overall Risk		Sample (Type of Sample)
% Women
Mean Age (SD)		Country
(Culture)		Contrast: Main Active Component vs. CG TypeDuration:
# Sessions (Length per
Session		Delivery-Mode; Format; Focus (Centered)Follow-Up (Months)Severity-Based Inclusion Criteria
Aydın and Kutlu (2021)
RoB: Some concerns		60 (Older Adults)
78.3% 72.56 (±1.01)		Turkey
(Coll)		Art therapy vs. NI×6 (60–90 min); length (n/r)In person; group; person-centeredNoYes
Bruehlman-Senecal et al. (2020)
RoB: Some concerns		221 (College Students)
59.3% 18.68 (0.35)		United States
(Ind)		CBT vs. DAn/r (n/r); 4 weeksVirtual; individual; person-centeredNoYes
Cohen-Mansfield et al. (2018)
RoB: Some concerns		74 (Older Adults)
CBT 79.49%
NI 82.86%
CBT 76.6 (6.8)
NI 79 (6.62)		Israel
(Coll)		CBT vs. NI×7 individual (n/r)
×10 group (n/r); length (n/r)		In person; group; person-centeredYes (3)No
Kahlon et al. (2021)
RoB: Low		240 (Older Adults)
79% SS 69.4 (11.5)
NI 68.7 (12.8)		United States
(Ind)		Social Support vs. NI×20 (~less than 10 min; 5 per week); 4 weeksVirtual; individual; person-centeredNoNo
Käll and Andersson (2023) c
RoB: Some concerns		73 (Adults)
71.2% 47.2 (17.63)		Sweden
(Ind)		CBT vs. DA×8 (n/r); 8 weeksVirtual; individual; person-centeredYes (24)Yes
Käll et al. (2021) b
RoB: Some concerns		170 (Adults)
75.9% 47.5 (16.4)		Sweden
(Ind)		CBT vs. WL
IPT vs. WL		×9 (n/r); 9 weeksVirtual; individual; person-centeredYes (4)No
Li et al. (2022)
RoB: Some concerns		64 (Older Adults)
63.3%
65.7 (3.69)		China
(Coll)		Reminiscence therapy vs. NI×8 (4 h); length (n/r)In person; group; person-centeredYes (3)No
S. Liu et al. (2023)
RoB: Some concerns		252 (Adults)
73%
33.93 (11.84)		Germany (Ind)Social support vs. NI×1 (~4 min); 1 dayVirtual; individual; person–context-centeredNoNo
Mahmoudpour et al. (2021)
RoB: Some concerns		83 (Divorced Women)
100%
32 (n/r)		Iran
(Coll)		ACT vs. NI×8 (90 min); 1 monthIn person; group; person-centeredNoNo
Morgado et al. (2023)
RoB: High		34 (Older Adults)
50%
81.97 (8.03)		Portugal (Ind)Reminiscence therapy vs. NI×10 (45 min); length (n/r)In person; group; person-centeredNoNo
Myhre et al. (2017) a
RoB: High		121 (Older Adults)
D-SS 36%
OD 31%
WL 21%
D-SS 80 (7.34)
OD 78.38 (7.32)
WL 79.29 (6.76)		United States
(Ind)		Digital-Social Support vs. WLnr (n/r); 8 weeksVirtual; individual; person-centeredNoNo
Ristolainen et al. (2020)
RoB: High		392 (Older Adults)
82.9% 76.8 (7.5)		Finland (Ind)Social support vs. NI×5 (2–3 h); 6 monthsIn person; group; person-centeredNoNo
Shapira et al. (2021)
RoB: Some concerns		82 (Older Adults)
CBT 81%
WL 78%
CBT 72.1 (5.3)
WL 71.7 (6.8)		Israel
(Ind)		CBT vs. WL×7 (60–90 min); ~4 weeksIn person; group; person-centeredNoNo
Tabrizi et al. (2016)
RoB: Some concerns		140 (Breast Cancer survivors)
100%
47.9 (11.4)		Iran
(Coll)		Social Support vs. NI×12 (90 min); 12 weeksIn person; group; person-centeredYes (2)No
Thimmapuram et al. (2021)
RoB: High		155 (Physicians)
66% 46 (11.03)		United States
(Ind)		Meditation vs. DA.×~28 Sessions (15 min); 4 weeksVirtual; individual; person-centeredNoNo
Xiao et al. (2021)
RoB: Some concerns		96 (University Students)
~26.04% PA 18.95 (0.89)
MM 19.21 (1.02)
NI 19.71 (1.77)		China
(Coll)		Physical Activity vs. NI
Mindful Movement vs. NI		×36 (90 min); 12 weeksVirtual; individual; person-centeredNoNo
Hussain et al. (2023)
RoB: Some concerns		110 (College students)
79%
20.97(4.83)		United Kingdom (Ind)Mindfulness vs. Happier Life Interventionnr (n/r); 4 weeksVirtual; individual; person-centeredYes (0.5)No
Lanser and Eisenberger (2023) **
RoB: Some concerns		300 (Undergraduate students)
79.7%
20.08 (n/r)		United States
(Ind)		Prosocial behavior-Giving vs. Logo viewing
Prosocial behaviour-Keeping vs. Logo viewing		×1 (task done only once); length (n/r)Virtual; individual; person-centeredNoNo
Lanser and Eisenberger (2023) **
RoB: Some concerns		300 (Undergraduate students)
77.1%
20.50 (n/r)		United States
(Ind)		Prosocial behavior-Gratitude vs. TV-show writing
Prosocial behaviour-Reflection vs. TV-show writing		×1 (task done only once); length (n/r)Virtual; individual; person-centeredNoNo
Lindsay et al. (2019)
RoB: Some concerns		153 (Young Adults)
67.32%
32.42 (13.68)		United States
(Ind)		Mindfulness vs. Coping control×14 sessions (20 min); 2 weeksVirtual; individual; person-centeredNoNo
C. W. Liu et al. (2023)
RoB: Some concerns		100 (Older Adults)
D-SS 74%
YT-i 70%
D-SS 72.90 (4.45)
YT-i 72.78 (±4.52)		Taiwan (Coll)Digital-Social Support vs. YouTube Interaction×60 (1 h); 12-weeksVirtual; group; person-centeredNoNo
Razani et al. (2018)
RoB: Some concerns		78 (Child-parent pairs)
87%
38 (n/r)		United States
(Ind)		Nature exposure vs. Independent Park prescription×3 (n/r); 3 weeksIn person; individual; context-centeredNoNo
Van Orden et al. (2022)
RoB: Low		291 (Older Adults)
75.3% 72 (9.07)		United States
(Ind)		Volunteering vs. Life Reviewn/r (n/r); 12 monthsIn person; individual; person-centeredNoNo
Yang et al. (2023)
RoB: Some concerns		90 (Older Adults)
64.04% VI 68.07 (6.68)
N-On 69.00 (6.04)		Taiwan (Coll)Virtual interaction with strangers vs. Non-online interaction group×40 (n/r); 8 weeksVirtual; individual; person-centeredNoNo
Zhang et al. (2023)
RoB: Some concerns		132 (University Students)
62.9% 24.4 (10.1)		United States
(Ind)		Art therapy vs. Hobbies-writing×4 (15 min); 4 daysVirtual; individual; person-centeredNoNo
Notes: Bolded study ID indicates Intervention–control (I-C) design-based studies selected in the meta-analysis (n = 16). a Only one comparison from this study was included in the meta-analysis. b Excluded from follow-up analyses as follow-up data was collected only for the experimental group. c Data from this intervention was also retrieved from (Käll et al., 2020a, 2020b). ** Indicates multiple studies reported in the same publication. Abbreviations: ACT = Acceptance and Commitment Therapy; CBT = Cognitive Behavioral Therapy; CG = Control Group; Coll = Collectivistic; DA = Delayed Access; D-SS = Digital Social Support; Ind = Individualistic; IPT = Internet-based Interpersonal Therapy; MM = Mindful Movement; NI = No Intervention; N-On = Non-online Interaction Group; n/r = number of sessions not reported, and/or length, and/or age not reported; OD = Online Diary; PA = Physical Activity; RoB = Risk of Bias assessment; SS = Social Support; VI = Virtual Interaction with Strangers; WL = Waiting List; YT-i = YouTube interaction; # = number of sessions in each intervention.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Zeas-Sigüenza, A.; Voldstad, A.; Ruisoto, P.; Ganho-Ávila, A.; Guiomar, R.; Cacho, R.; Muntané, F.; Benach, J. Loneliness as a Public Health Challenge: A Systematic Review and Meta-Analysis to Inform Policy and Practice. Eur. J. Investig. Health Psychol. Educ. 2025, 15, 131. https://doi.org/10.3390/ejihpe15070131

AMA Style

Zeas-Sigüenza A, Voldstad A, Ruisoto P, Ganho-Ávila A, Guiomar R, Cacho R, Muntané F, Benach J. Loneliness as a Public Health Challenge: A Systematic Review and Meta-Analysis to Inform Policy and Practice. European Journal of Investigation in Health, Psychology and Education. 2025; 15(7):131. https://doi.org/10.3390/ejihpe15070131

Chicago/Turabian Style

Zeas-Sigüenza, Ananda, Andreas Voldstad, Pablo Ruisoto, Ana Ganho-Ávila, Raquel Guiomar, Raúl Cacho, Ferran Muntané, and Joan Benach. 2025. "Loneliness as a Public Health Challenge: A Systematic Review and Meta-Analysis to Inform Policy and Practice" European Journal of Investigation in Health, Psychology and Education 15, no. 7: 131. https://doi.org/10.3390/ejihpe15070131

APA Style

Zeas-Sigüenza, A., Voldstad, A., Ruisoto, P., Ganho-Ávila, A., Guiomar, R., Cacho, R., Muntané, F., & Benach, J. (2025). Loneliness as a Public Health Challenge: A Systematic Review and Meta-Analysis to Inform Policy and Practice. European Journal of Investigation in Health, Psychology and Education, 15(7), 131. https://doi.org/10.3390/ejihpe15070131

Article Metrics

Back to TopTop