Previous reviews of mental health interventions have focused on adolescents (10–19 years), with a paucity of comprehensive evidence syntheses on preventive interventions for school-aged children (5–10 years).
To summarize and synthesize the available evidence from systematic reviews of mental health and positive development interventions for children aged 5–14.9 years in both high-income (HIC) and low- and middle-income countries (LMIC), with a focus on preventive and promotive strategies.
This overview includes all relevant reviews from OVID Medline, The Cochrane Library, and Campbell Systematic Reviews through December 2020.
We included systematic reviews that synthesized empirical studies using experimental or quasi-experimental designs to evaluate the effectiveness of interventions in children aged 5–14.9 years.
Data extraction and quality assessment were completed independently and in duplicate by two review authors. The AMSTAR2 tool was used to assess methodological quality.
We included 162 reviews. The greatest evidence was found in support of school-based universal and anti-bullying interventions in predominantly HIC. Moderate evidence was found for the use of substance abuse prevention, and early learning and positive development interventions in mixed settings. In LMIC-only contexts, the most promising evidence was found for positive youth development programs.
The review was primarily limited by paucity of high-quality research due to methodological issues and high heterogeneity.
This overview of reviews highlights the need for further research to consolidate findings and understand the specific criteria involved in creating positive mental health and development outcomes from the various interventions considered.
The Sustainable Development Goals have brought attention to early childhood development and the intergenerational benefits of investing in early life.1 Since early childhood development underpins several of the Sustainable Development Goals, increased focus on this area is imperative to reach these global goals by 2030.1 It is estimated that 10% to 20% of children and adolescents worldwide experience mental health disorders, of which ∼50% manifest before age 14.1 In Europe and the Americas, mental health disorders are among the leading causes of disability-adjusted life-years among children aged 5 to 14 years.2 The burden of poor mental health and development prevents children from achieving their full potential, and has downstream impacts on the human capital of entire countries.
Most mental health conditions are underdiagnosed and undertreated, and children and adolescents face barriers to recovery, including stigma, shame, and inaccessibility of interventions.3 Intervening at an early age is critical because the consequences of mental illness extend into adulthood, leading to mental and physical harm across the life course.3 Additionally, multidirectional linkages exist between mental health conditions and other developmental concerns, such as risk-taking behaviors, poor education, and additional health problems.4 For example, depression and anxiety are leading causes of illness and disability among adolescents worldwide, with the potential to lead to self-harm, substance abuse, and risk-taking behaviors.3,5 Among adolescents aged 15 to 19, about 6% of all deaths are attributed to suicide and self-harm.5 Inadequate mental health support within schools and families highlights the need for more preventive and promotive efforts that are delivered earlier in life, and which improve children’s resilience, emotional regulation, and ability to avoid risky behaviors. More than ever before, the dramatic rise in mental health issues among school-aged children and adolescents (SACA) in the wake of the global coronavirus disease 2019 pandemic has increased awareness of this issue as it manifolds, with interest in mitigation strategies.6
Why It Is Important to Do This Review
Previous reviews of mental health interventions in childhood have been focused on adolescents (10–19 years),7,8 and there is a dearth of comprehensive evidence syntheses specifically focused on school-aged children aged 5 to 10 years, and on preventive and promotive interventions that start in childhood and continue into adolescence. It is important to consider the needs of the school-age children age group in intervention efforts for several reasons. It is recognized that preventive interventions are more effective when implemented before the age of disease onset.9 Because of the early age of onset for many mental health conditions, these preventive interventions could be more effective at preventing mental health conditions in adolescents and young adults if they were implemented during the school-aged years.3 Despite this, programs designed, for example, to prevent substance use initiation are often delivered to children who are aged >12 years.10 Younger children, particularly those aged <8 years, may be better suited to receive nonsmoking messages because their perception of smoking is still largely negative.10 Likewise, primary school children (grades 1–5) may be better suited to receive antibullying interventions because the incidence of perpetration and victimization is highest in these children.11 Additionally, it is important to consider a life course perspective when designing mental health and positive development interventions because those implemented before adolescence can impact both adolescence and adulthood.7 Furthermore, in the current landscape of school closures and lockdowns, evidence continues to build on the impact of coronavirus disease 2019 on mental health in young children.12–14 For example, early evidence from China illustrates a prevalence of anxiety and depressive symptoms of 19% and 23%, respectively, in primary school children after home confinement.15 For these reasons, this review will focus on interventions delivered to children aged <10 years, with the recognition that their delivery may often extend into early adolescence (10–14.9 years).
Methods
Objectives
This review aims to summarize and synthesize the available evidence from systematic reviews of mental health and positive development interventions for children aged 5 to 14.9 in both high-income countries (HIC) and low- and middle-income countries (LMIC), with a focus on preventive and promotive strategies.
Eligibility Criteria
We considered systematic reviews published until December 2020 that synthesized empirical studies using either experimental or quasi-experimental designs to evaluate the effectiveness of interventions.
For this overview, we primarily focused on those reviews that covered ages 5 to 14.9 years, which encompasses the period of “school-age” and includes late childhood and early adolescence. Although the primary focus is on ages 5 to 14.9 years, we still included reviews in which most of the sample’s target age range falls within our selected age limit. For example, we included reviews where the target age range of the intervention was 5 to 19 years or where the stated average age of participants fell between 5 and 14.9 years. In reviews that reported large age ranges (eg, 5–19 years), we were unable to parse out effects of younger age groups unless the data were disaggregated by age bands. In reviews that reported disaggregated data by age, we have prioritized children aged 5 to 9 years. We chose to include both HIC and LMIC because we did not expect to find evidence for all interventions in the LMIC context alone. However, when possible, we highlight LMIC evidence in this article where it existed. The reviews that focused on interventions delivered to special populations of children that were amenable to prevention (eg, those with chronic illness such as epilepsy) or nongeneralizable contexts (eg, conflict and humanitarian settings) were considered separately and are listed in Supplemental Information. This narrowing of scope was performed to allow for generalizability of effective interventions to the greatest number of children possible.
We focused on universal, generalizable, preventive interventions that aim to promote positive development, prevent mental disorders, improve mental health and well-being, or modify its determinants. We focused exclusively on preventive interventions because intervening before the onset of mental health problems is particularly beneficial in the context of young children who could potentially be spared of psychological distress later in adolescence or adulthood.
Several delivery strategies were considered, including school-based, community-based, digital, and mixed settings. An expanded listing of intervention types and associated delivery strategies can be found in Supplemental Information. Interventions that can be categorized into 1 or more of the following domains described below are the focus of this article. The domains were selected through expert discussion, the scope of previous overviews, and with focus on the greatest disease burden. Although some reviews could be categorized into multiple domains, we took a pragmatic approach to organizing them in the different sections of the narrative synthesis to ensure reviews of similar intervention types were considered together. Universal prevention interventions were placed in this category if the authors described the intervention as universal or covered a broad range of outcomes. The remaining interventions were categorized into each domain if either the intervention or outcome was specific to that domain. See Fig 1 for a general overview of the interventions, delivery strategies, and outcomes assessed.
Conceptual framework of mental health and positive development interventions, delivery strategies, and outcomes assessed in this overview.
Conceptual framework of mental health and positive development interventions, delivery strategies, and outcomes assessed in this overview.
Which universal prevention interventions are effective in improving multiple domains of SACA mental health status, development, and well-being?
Which interventions are effective in targeting the following key issues in SACA:
o bullying prevention interventions;
o substance use prevention interventions; and
o self-harm prevention interventions?
Which interventions are effective in supporting the following aspects of nurturing care in SACA:
o supporting safety and security, and preventing abuse and maltreatment;
o promoting responsive caregiving and parenting practices; and
o providing learning opportunities and promoting positive child development?
Which targeted interventions are effective for special populations of SACA that are amenable to prevention or nongeneralizable contexts?
Information Sources and Search Strategy
Keyword searches were conducted in the following electronic databases: OVID Medline, The Cochrane Library, and Campbell Systematic Reviews. Each domain of interest described above used its own search strategy to identify relevant systematic reviews of interventions (see Supplemental Information). We also hand-searched the reference lists of key overviews to identify additional relevant systematic reviews. Although initial searches were conducted in 2019, an updated search was conducted until December 2020 to identify newer reviews.
Screening and Selection
At both title/abstract and full-text screening stages, a team of reviewers independently screened all records retrieved by the searches for relevance based on predefined eligibility criteria (see Supplemental Information). Any disagreements were resolved through discussion or by a third reviewer when necessary.
Data Items, Collection, and Quality Assessment
A standardized data extraction form was used to extract data from included reviews. The extraction form (see Supplemental Information) included general review characteristics, description of the target population, intervention types, delivery strategies, and data on outcomes and intervention effectiveness. The AMSTAR2 tool was used to assess the methodological quality of the included systematic reviews.16 Full-text data extraction and quality assessment for all included reviews was conducted independently by 2 reviewers. Pooled estimates (eg, risk ratios, odds ratios, and mean differences) from meta-analyses were extracted for a variety of mental health and positive development outcomes.
Results
Study Selection
After removal of duplicates, 1926 reviews were title/abstract-screened for inclusion on the basis of the eligibility criteria previously described. Two-hundred twenty-nine citations were screened at full text and 162 studies were ultimately included (Fig 2). Sixty-seven studies did not meet our inclusion criteria and are presented in the Table of Excluded Studies (see Supplemental Information). The 162 reviews selected for final analysis were composed of 5 domains: nurturing care (n = 45),17–62 self-harm (n = 3),63–65 bullying (n = 18),11,66–82 substance use (n = 22),83–104 and universal prevention interventions (n = 74).105–176
Study Characteristics of Included Reviews
Twenty-nine reviews covered targeted interventions and were synthesized separately in Supplemental Information H. Four reviews included both universal and targeted populations.63,93,107,108 Study characteristics for the 137 included reviews are summarized in Table 1 and additional detail is provided in Supplemental Information.
Study Characteristics Summary Table
| Domain | Number of Reviews, Excluding Special Populations | Number of Special Population Reviews | Number of LMIC-Only Reviews | Number of LMIC + HIC Reviews | Number of HIC-Only Reviews | Age Range | Publication Date Range | Number of RCT-Only Reviews |
|---|---|---|---|---|---|---|---|---|
| Universal prevention | 61 | 15 | 5 | 18 | 19 | <1 to 25 y | 2002 to 2020 | 20 |
| Bullying | 14 | 4 | 1 | 5 | 6 | 0 to 19 y | 2007 to 2021 | 3 |
| Substance abuse | 22 | 1 | 1 | 11 | 9 | 5 to 26 y | 2008 to 2020 | 10 |
| Self-harm | 2 | 2 | 0 | 0 | 1 | 0 to 18 y | 2019 | 0 |
| Nurturing care | 38 | 7 | 1 | 15 | 18 | 6 mo to 21 y | 2006 to 2021 | 10 |
| Totals | 137 | 29 | 8 | 49 | 53 | 0 to 26 y | 2002 to 2021 | 43 |
| Domain | Number of Reviews, Excluding Special Populations | Number of Special Population Reviews | Number of LMIC-Only Reviews | Number of LMIC + HIC Reviews | Number of HIC-Only Reviews | Age Range | Publication Date Range | Number of RCT-Only Reviews |
|---|---|---|---|---|---|---|---|---|
| Universal prevention | 61 | 15 | 5 | 18 | 19 | <1 to 25 y | 2002 to 2020 | 20 |
| Bullying | 14 | 4 | 1 | 5 | 6 | 0 to 19 y | 2007 to 2021 | 3 |
| Substance abuse | 22 | 1 | 1 | 11 | 9 | 5 to 26 y | 2008 to 2020 | 10 |
| Self-harm | 2 | 2 | 0 | 0 | 1 | 0 to 18 y | 2019 | 0 |
| Nurturing care | 38 | 7 | 1 | 15 | 18 | 6 mo to 21 y | 2006 to 2021 | 10 |
| Totals | 137 | 29 | 8 | 49 | 53 | 0 to 26 y | 2002 to 2021 | 43 |
Methodological Quality (AMSTAR2) of Included Reviews
AMSTAR2 assessments were performed for 78 reviews that reported meta-analyses.17,21–31, 33–35,37,39,40,45,46,48,50,51,56,58,59,61,63,64, 67–69,76,81,84,85,88–90,94,95,97,98, 100–102,110,112,114,118–120,123,125,129,138, 141,143,145,148–150,154,155,157–160,162, 165–167,169,172–175,177 Thirty-one studies (39.7%) were graded as “critically low,” 21 studies (26.9%) were graded as “low,” 6 studies (7.8%) were graded as “moderate,” and 20 studies (25.6%) were graded as “high” quality. Additional detail of assessments can be found in Supplemental Information.
Summary of the Effects of Interventions
Effects of interventions with quality ratings from all included reviews that were able to meta-analyze data on any mental health and/or positive development outcome have been summarized in Tables 2–6, by domain and delivery strategy.
Universal Prevention Meta-Analysis Results
| Source | Comparison | Outcomes | Pooled Effect Estimates (95% CI)) | AMSTAR Rating | Total # of People | Total # of Studies |
|---|---|---|---|---|---|---|
| School-based interventions | ||||||
| Langford et al138 (2015) | Antibullying interventions (Friendly Schools and KiVa programs) versus no intervention | Being bullied | OR = 0.83 (0.72 to 0.96) | Critically low | 26 256 | 6 |
| Antibullying interventions versus no intervention | Bullying others | OR = 0.9 (0.78 to 1.04) | 26 176 | 6 | ||
| Tobacco interventions versus no intervention | Tobacco use; tobacco only | OR = 0.77 (0.64 to 0.93) | 4747 | 3 | ||
| Multiple risk behaviors interventions versus no intervention | Tobacco use; multiple risk behavior | OR = 0.84 (0.76 to 0.93) | 9992 | 5 | ||
| Alcohol interventions versus no intervention | Alcohol use; alcohol only | OR = 0.72 (0.34 to 1.52) | 7481 | 2 | ||
| Multiple risk behaviors interventions vs no intervention | Alcohol use; multiple risk behavior | OR = 0.75 (0.55 to 1.02) | 8140 | 4 | ||
| Multiple risk behaviors interventions vs no intervention | Substance use; multiple risk behavior | OR = 0.57 (0.29 to 1.14) | 6820 | 3 | ||
| Alcohol interventions vs no intervention | Substance use; alcohol only | OR = 0.94 (0.78 to 1.12) | 7481 | 2 | ||
| Dray et al118 (2017) | Usual care | Depressive symptoms | SMD = −0.08 (−0.14 to −0.01) | Low | 30 | — |
| Anxiety symptoms | SMD = −0.14 (−0.28 to 0.00) | 22 | — | |||
| Hyperactivity | SMD = −0.07 (−0.18 to 0.05) | 5 | — | |||
| Conduct problems | SMD = 0.01 (−0.11 to 0.12) | 4 | — | |||
| Internalizing problems | SMD = −0.21 (−0.36 to −0.06) | 4 | — | |||
| Externalizing problems | SMD = −0.18 (−0.34 to −0.01) | 4 | — | |||
| General psychological distress | SMD = −0.11 (−0.21 to −0.01) | 6 | — | |||
| Franklin et al125 (2017) | Usual care | Internalizing | G = 0.133 (0.002 to 0.263) | Critically low | — | 27 |
| Externalizing | G = 0.015 (−0.037 to 0.066) | — | 96 | |||
| Internalizing versus externalizing | G = 0.118 (0.034 to 0.202) | — | 123 | |||
| Taylor et al157 (2017) | Usual care | Social and emotional assets: SEL skills | ES = 0.23 (0.15 to 0.31) | Critically low | — | 29 |
| Social and emotional assets: attitudes | ES = 0.13 (0.05 to 0.21) | — | 26 | |||
| Positive and negative indicators of well-being: positive social behavior | ES = 0.13 (0.05 to 0.21) | — | 28 | |||
| Positive and negative indicators of well-being: academic performance | ES = 0.33 (0.17 to 0.49) | — | 8 | |||
| Positive and negative indicators of well-being: conduct problems | ES = 0.14 (0.07 to 0.21) | — | 34 | |||
| Positive and negative indicators of well-being: emotional distress | ES = 0.16 (0.08 to 0.23) | — | 35 | |||
| Positive and negative indicators of well-being: drug use | ES = 0.16 (0.09 to 0.24) | — | 28 | |||
| van Genugten et al160 (2017) | Usual care | Internalizing | ES = 0.20 (0.03 to 0.38) | Low | — | 13 |
| Short-term internalizing | ES = 0.09 (−0.01 to 0.20) | — | 6 | |||
| Long-term self-esteem | ES = 0.29 (0.18 to 0.41) | — | 14 | |||
| Short-term self-esteem | ES = 0.16 (0.05 to 0.26) | — | 5 | |||
| Watson et al162 (2017) | Usual care | Physical activity on classroom behavior | SMD = 0.60 (0.20 to 1.00) | Critically low | 804 | 4 |
| Physical activity on cognitive function | SMD = 0.33 (-0.11 to 0.77) | 1081 | 5 | |||
| Sanchez et al149 (2018) | School-based mental health programs versus no services, peer tutoring or psychologist program | All outcomes | ES = 0.39 (0.29 to 0.49) | Low | — | 43 |
| Externalizing problems | ES = 0.50 (0.35 to 0.63) | — | 27 | |||
| Internalizing problems | ES = 0.30 (0.16 to 0.43) | — | 18 | |||
| Attention problems | ES = 0.10 (0.03 to 0.17) | — | 7 | |||
| Substance use | ES = 0.18 (−0.15 to 0.50) | — | 7 | |||
| Tejada-Gallardo et al158 (2020) | Active control groups (n = 5) and nonactive/waitlist (n = 4) | Subjective well-being (studies postintervention) | ES = 0.24 (0.11 to 0.37) | High | — | 6 |
| Psychological well-being (studies postintervention) | ES = 0.25 (−0.01 to 0.51) | — | 5 | |||
| Depression (studies postintervention) | ES = 0.28 (0.13 to 0.43) | — | 4 | |||
| Anxiety (studies postintervention) | ES = 0.14 (0.04 to 0.24) | — | 4 | |||
| Subjective well-being (studies postintervention excluding low quality studies) | ES = 0.21 (0.05 to 0.37) | — | 4 | |||
| Psychological well-being (studies postintervention excluding low quality studies) | ES = 0.31 (−0.03 to 0.67) | — | 4 | |||
| Depression (studies postintervention excluding low quality studies) | ES = 0.34 (0.24 to 0.44) | — | 3 | |||
| Anxiety (studies postintervention excluding low quality studies) | ES = 0.15 (0.05 to 0.25) | — | 3 | |||
| Subjective well-being (follow-up studies) | ES = 0.13 (0.03 to 0.23) | — | 4 | |||
| Psychological well-being (follow-up studies) | ES = 0.44 (-0.45 to 1.31) | — | 3 | |||
| Depression (follow-up studies) | ES = 0.31 (0.20 to 0.41) | — | 3 | |||
| Anxiety (follow-up studies) | ES = 0.15 (0.05 to 0.26) | — | 3 | |||
| Psychological well-being (follow-up studies excluding low quality studies) | ES = 0.66 (-0.39 to 1.72) | — | 2 | |||
| Depression (follow-up studies excluding low quality studies) | ES = 0.29 (0.11 to 0.47) | — | 2 | |||
| Anxiety (follow-up studies excluding low quality studies) | ES = 0.21 (0.10 to 0.33) | — | 2 | |||
| Caldwell et al112 (2019) | Active control, waitlist, curriculum control, no intervention, or attention control. | Universal primary: depression (waitlist) | SMD = −0.09 (−0.77 to 0.54) | High | 56 620 | 12 (Direct trials: 0) |
| Universal primary: anxiety (waitlist) | SMD = 0.02 (−0.20 to 0.22) | 15 (Direct trials: 0) | ||||
| Universal primary: depression (no intervention) | SMD = 0.13 (−0.40 to 0.65) | 12 (Direct trials: 0) | ||||
| Universal primary: anxiety (no intervention) | SMD = 0.23 (−0.15 to 0.60) | 15 (Direct trials: 0) | ||||
| Universal primary: depression (attention control) | SMD = −0.07 (−0.79 to 0.62) | 12 (Direct trials: 0) | ||||
| Universal primary: anxiety (attention control) | SMD = −0.17 (−0.51 to 0.17) | 15 (Direct trials: 0) | ||||
| Universal primary: depression (CBT) | SMD = −0.13 (−0.44 to 0.17) | 12 (Direct trials: 6) | ||||
| Universal primary: anxiety (CBT) | SMD = −0.07 (−0.23 to 0.05) | 15 (Direct trials: 6) | ||||
| Universal primary: depression (behavioral therapy) | SMD = −0.10 (−1.04 to 0.80) | 12 (Direct trials: 0) | ||||
| Universal secondary: depression (waitlist) | SMD = 0.00 (−0.19 to 0.19) | 34 (Direct trials: 0) | ||||
| Universal secondary: anxiety (waitlist) | SMD = −0.05 (−0.28 to 0.18) | 21 (Direct trials: 0) | ||||
| Universal secondary: depression (no intervention) | SMD = 0.03 (−0.15 to 0.21) | 34 (Direct trials: 0) | ||||
| Universal secondary: anxiety (no intervention) | SMD = −0.07 (−0.34 to 0.20) | 21 (Direct trials: 0) | ||||
| Universal secondary: depression (attention control) | SMD = 0.07 (−0.12 to 0.25) | 34 (Direct trials: 1) | ||||
| Universal secondary: anxiety (attention control) | SMD = −0.15 (−0.51 to 0.16) | 21 (Direct trials: 0) | ||||
| Universal secondary: depression (CBT) | SMD = −0.04 (−0.16 to 0.07) | 34 (Direct trials: 11) | ||||
| Universal secondary: anxiety (CBT) | SMD = −0·15 (−0·34 to 0·04) | 21 (Direct trials: 3) | ||||
| Universal secondary: depression (third wave) | SMD = −0.03 (−0.21 to 0.14) | 34 (Direct trials: 1) | ||||
| Universal secondary: anxiety (third wave) | SMD = 0.03 (−0.14 to 0.20) | 21 (Direct trials: 3) | ||||
| Universal secondary: depression (IPT+CBT) | SMD = −0.19 (−0.46 to 0.08) | 34 (Direct trials: 0) | ||||
| Universal secondary: depression (IPT) | SMD = −0.03 (−0.36 to 0.29) | 34 (Direct trials: 1) | ||||
| Universal secondary: depression (psychoeducation) | SMD = −0.13 (−0.49 to 0.22) | 34 (Direct trials: 1) | ||||
| Universal secondary: depression (behavioral therapy) | SMD = −0.02 (−0.40 to 0.37) | 34 (Direct trials: 1) | ||||
| Universal secondary: anxiety (mindfulness and relaxation) | SMD = −0.65 (−1.14 to −0.19) | 21 (Direct trials: 0) | ||||
| Feiss et al123 (2019) | Active control, usual care, or no control | Stress | t (10) = −0.36; dexp = −0.05 (−0.58 to 0.48); dctrl = 0.05 (−0.18 to 0.28) | Critically low | 420 | 4 |
| Anxiety | t (54) = −3.72; dexp = −0.70 (−0.94 to −0.46); dctrl = −0.14 (−0.26 to −0.01) | 2166 | 20 | |||
| Depression | t (116) = −3.120; dexp = −0.62 (−0.81 to −0.43); dctrl = −0.22 (−0.34 to −0.10) | 6741 | 38 | |||
| Mixed setting interventions | ||||||
| Ciocanel et al114 (2017) | — | Behavioral adjustment, positive social behaviors | ES = 0.04 (-0.11 to 0.21) | Critically low | — | 7 |
| Behavioral adjustment, problem behaviors | ES = 0.05 (−0.00 to 0.11) | — | 16 | |||
| Psychological adjustment, all combined | ES = 0.17 (0.04 to 0.31) | — | 8 | |||
| Psychological adjustment, emotional distress | ES = 0.14 (-0.002 to 0.29) | — | 3 | |||
| Psychological adjustment, self-perceptions | ES = 0.19 (0.02 to 0.37) | — | 6 | |||
| Academic/school outcomes, academic achievement | ES = 0.22 (0.07 to 0.38) | — | 10 | |||
| Academic/school outcomes, academic adjustment | ES = 0.09 (−0.02 to 0.20) | — | 5 | |||
| Sexual health outcomes, risky sexual behavior | ES = 0.05 (−0.00 to 0.12) | — | 11 | |||
| Melendez-Torres et al141 (2016) | — | All substance use outcomes: all time points | d = 0.079 (−0.025 to 0.183) | Moderate | 54 | 9 |
| All substance use outcomes: short-term time points | d = 0.086 (−0.025 to 0.197) | 36 | 9 | |||
| Omnibus substance use outcomes: short-term time points | d = 0.169 (0.012 to 0.326) | 10 | 7 | |||
| Bennett et al110 (2015) | — | Anxiety symptoms | SMD = 0.22 (0.14 to 0.29) | Low | — | 65 |
| Currier et al177 (2007) | — | General helpfulness | ES = 0.14 (0.00 to 0.28) | Critically low | — | 13 |
| Ekeland et al120 (2004) | — | Self-esteem | SMD = 0.51 (0.10 to 0.9) | Low | 161 | 4 |
| Pratt et al143 (2010) | — | Eating attitudes/behaviors and adolescent issues. BMI at 12 − 14-mo follow-up | MD = −0.10 (−0.45 to 0.25) | Low | 1235 | 4 |
| Eating attitudes/behaviors and adolescent issues. Eating attitudes test (EAT) total at 6 − 12 mo follow-up | SMD = 0.01 (−0.13 to 0.15) | 792 | 4 | |||
| Eating attitudes/behaviors and adolescent issues. Eating disorder inventory (EDI), bulimia at 12 − 14-mo follow-up | SMD = −0.03 (−0.16 to 0.10) | 955 | 2 | |||
| Media literacy and advocacy. SATQ awareness at 3 mo | SMD = 0.18 (−0.05 to 0.41) | 297 | 2 | |||
| Media literacy and advocacy. SPA and SPPC global self-worth at 3 mo | SMD = −0.05 (−0.44 to 0.34) | 104 | 2 | |||
| Media literacy and advocacy. Body image assessment at 3 mo | SMD = 0.02 (−0.37 to 0.42) | 127 | 2 | |||
| Self-esteem approach; SPA close friendships at 3 mo | MD = −0.01 (−0.09 to 0.06) | 524 | 2 | |||
| Self-esteem approach; SPA social acceptance at 3 mo | MD = −0.03 (−0.10 to 0.04) | 531 | 2 | |||
| Raposa et al145 (2019) | — | Average effect of youth mentoring on all outcomes | g = 0.21 (0.14 to 0.28) | Critically low | 25 286 | 70 |
| Schleider et al150 (2017) | — | Overall SSI effect | ES = 0.32 (0.17 to 0.46) | Critically low | — | 50 (299 ESs) |
| Spruit et al155 (2016) | — | Internalizing problems | MD = 0.316 (0.073 to 0.558) | Critically low | — | 14 |
| Self-concept | MD = 0.297 (0.127 to 0.468) | — | 24 | |||
| Academic achievement | MD = 0.367 (0.380 to 0.697) | — | 33 | |||
| Overall effect size | ES = 0.33 (0.14 to 0.53) | — | 10 | |||
| Externalizing problems | MD = 0.320 (0.023 to 0.616) | — | 7 | |||
| Salazar de Pablo et al148 (2020) | Universal/selective interventions for good mental health versus a control group | Mental health literacy | ES = 0.685 (P < .001) | Low | 37 533 | 45 |
| Emotions | ES = 0.541 (P < .001) | 7593 | 37 | |||
| Self-perceptions and values | ES = 0.49 (P < .001) | 37 183 | 92 | |||
| Quality of life | ES = 0.457 (P = .001) | 31 276 | 48 | |||
| Cognitive skills | ES = 0.428 (P < .001) | 12 368 | 20 | |||
| Social skills | ES = 0.371 (P < .001) | 62 274 | 94 | |||
| Physical health | ES = 0.285 (P < .001) | 7642 | 25 | |||
| Sexual health | ES = 0.257 (P = .017) | 14 202 | 13 | |||
| Academic/occupational performance | ES = 0.211 (P < .001) | 19 324 | 30 | |||
| Attitude toward mental disorders | ES = 0.177 (P = .006) | 8741 | 16 | |||
| Improve behaviors | ES = 0.065 (P = .23) | 10 474 | 11 | |||
| Improve family/significant relationships | ES =-0.046 (P = .674) | 32 639 | 45 | |||
| Self-management strategies | ES = 0.107 (P = .09) | 41 437 | 86 | |||
| Vallis et al159 (2020) | Cognitive behavior interventions compared with passive control groups | Change in anxiety from pretreatment to postintervention | SMD = −1.34 (−1.59 to −1.09) | Moderate | 1965 | 41 |
| Efficacy of interventions compared with control conditions | SMD = −0.81 (−1.00 to −0.63) | 1525 | 19 | |||
| Anxiety decreased from preintervention to follow-up | SMD = −1.40 (−1.75 to −1.06) | 1525 | 19 | |||
| Efficacy of intervention, reported by assessor | SMD = −0.86 (SE = 0.12) | — | — | |||
| Efficacy of intervention, reported by parent | SMD = −0.83 (SE = 0.15) | — | — | |||
| Efficacy of intervention, participants with anxiety disorders | SMD = −0.91 (SE = 0.13) | — | — | |||
| Efficacy of intervention, participants with anxiety symptoms | SMD = −0.84 (SE = 0.14) | — | — | |||
| Efficacy of intervention, participants with behavioral inhibition | SMD = −0.47 (SE = 0.16) | — | — | |||
| Efficacy of intervention, parent focused | SMD = −0.72 (SE = 0.18) | — | — | |||
| Efficacy of intervention, both parent and child focused | SMD = −0.91 (SE = 0.12) | — | — | |||
| Efficacy of intervention, child focused | SMD = −1.02 (SE = 0.31) | — | — | |||
| Efficacy of intervention, delivered individual in person | SMD = −0.85 (SE = 0.10) | — | — | |||
| Efficacy of intervention, delivered in group in person | SMD = −0.89 (SE = 0.18) | — | — | |||
| Efficacy of intervention, delivered over the internet | SMD = −0.45 (SE = 0.24) | — | — | |||
| Hetrick et al129 (2016) | Overall evidence‐based psychological therapy versus control | Depression diagnoses | RD = −0.05 (−0.08 to −0.02) | High | 3232 | 36 |
| Overall evidence‐based depression prevention program versus control | Depression symptoms self-reported | SMD = −0.21 (−0.27 to −0.15) | 13 829 | 73 | ||
| Universal depression prevention programs compared with any comparator | Depression symptoms self-reported | SMD = −0.11 (−0.17 to −0.05) | 9013 | 31 | ||
| Overall evidence‐based depression prevention program versus control | Depression symptoms clinic related | SMD = −0.23 (−0.41 to −0.05) | 2175 | 11 | ||
| Universal depression prevention programs compared with any comparator | Anxiety symptoms | SMD = −0.09 (−0.17 to −0.01) | 3130 | 8 | ||
| Overall evidence‐based depression prevention program versus control | General and social functioning | SMD = 0.24 (0.06 to 0.41) | 2067 | 10 | ||
| Universal depression prevention programs compared with any comparator | General and social functioning | SMD = 0.16 (0.04 to 0.28) | 1046 | 1 | ||
| Dunning et al119 (2019) | No contact, waitlist, active or attention placebo control | All measures | d = 0.19 (0.14 to 0.23) | High | 3666 | 33 |
| Mindfulness | d = 0.24 (0.01 to 0.46) | 1475 | 11 | |||
| Social behavior | d = 0.16 (−0.05 to 0.37) | 1247 | 10 | |||
| Negative behavior | d = 0.27 (0.07 to 0.47) | 970 | 11 | |||
| Depression | d = 0.27 (0.06 to 0.49) | 1529 | 13 | |||
| Anxiety/stress | d = 0.16 (0.04 to 0.27) | 2319 | 20 | |||
| Executive functions | d = 0.30 (0.12 to 0.49) | 1691 | 15 | |||
| Attention | d = 0.19 (0.04 to 0.34) | 1158 | 8 | |||
| Digital-based interventions | ||||||
| Skeen et al154 (2019) | Universally delivered psychosocial interventions versus control | Positive mental health (face-to-face) | ES = 0.257 (0.097 to 0.416) | Moderate | — | 129 |
| Positive mental health (digital and combined) | ES = 0.197 (0.016 to 0.379) | — | 29 | |||
| Depression and anxiety symptoms (face-to-face) | ES = −0.088 (−0.151 to −0.025) | — | 129 | |||
| Depression and anxiety symptoms (digital and combined) | ES = −0.054 (−0.81 to 0.074) | — | 29 | |||
| Violence, aggression, and bullying (face-to-face) | ES = −0.294 (−0.564 to −0.024) | — | 129 | |||
| Violence, aggression, and bullying (digital and combined) | ES = −0.075 (−0.249 to 0.099) | — | 29 | |||
| Substance use (face-to-face) | ES = −0.04 (−0.117 to 0.037) | — | 129 | |||
| Substance use (digital and combined) | ES = −0.114 (−0.199 to −0.029) | — | 29 |
| Source | Comparison | Outcomes | Pooled Effect Estimates (95% CI)) | AMSTAR Rating | Total # of People | Total # of Studies |
|---|---|---|---|---|---|---|
| School-based interventions | ||||||
| Langford et al138 (2015) | Antibullying interventions (Friendly Schools and KiVa programs) versus no intervention | Being bullied | OR = 0.83 (0.72 to 0.96) | Critically low | 26 256 | 6 |
| Antibullying interventions versus no intervention | Bullying others | OR = 0.9 (0.78 to 1.04) | 26 176 | 6 | ||
| Tobacco interventions versus no intervention | Tobacco use; tobacco only | OR = 0.77 (0.64 to 0.93) | 4747 | 3 | ||
| Multiple risk behaviors interventions versus no intervention | Tobacco use; multiple risk behavior | OR = 0.84 (0.76 to 0.93) | 9992 | 5 | ||
| Alcohol interventions versus no intervention | Alcohol use; alcohol only | OR = 0.72 (0.34 to 1.52) | 7481 | 2 | ||
| Multiple risk behaviors interventions vs no intervention | Alcohol use; multiple risk behavior | OR = 0.75 (0.55 to 1.02) | 8140 | 4 | ||
| Multiple risk behaviors interventions vs no intervention | Substance use; multiple risk behavior | OR = 0.57 (0.29 to 1.14) | 6820 | 3 | ||
| Alcohol interventions vs no intervention | Substance use; alcohol only | OR = 0.94 (0.78 to 1.12) | 7481 | 2 | ||
| Dray et al118 (2017) | Usual care | Depressive symptoms | SMD = −0.08 (−0.14 to −0.01) | Low | 30 | — |
| Anxiety symptoms | SMD = −0.14 (−0.28 to 0.00) | 22 | — | |||
| Hyperactivity | SMD = −0.07 (−0.18 to 0.05) | 5 | — | |||
| Conduct problems | SMD = 0.01 (−0.11 to 0.12) | 4 | — | |||
| Internalizing problems | SMD = −0.21 (−0.36 to −0.06) | 4 | — | |||
| Externalizing problems | SMD = −0.18 (−0.34 to −0.01) | 4 | — | |||
| General psychological distress | SMD = −0.11 (−0.21 to −0.01) | 6 | — | |||
| Franklin et al125 (2017) | Usual care | Internalizing | G = 0.133 (0.002 to 0.263) | Critically low | — | 27 |
| Externalizing | G = 0.015 (−0.037 to 0.066) | — | 96 | |||
| Internalizing versus externalizing | G = 0.118 (0.034 to 0.202) | — | 123 | |||
| Taylor et al157 (2017) | Usual care | Social and emotional assets: SEL skills | ES = 0.23 (0.15 to 0.31) | Critically low | — | 29 |
| Social and emotional assets: attitudes | ES = 0.13 (0.05 to 0.21) | — | 26 | |||
| Positive and negative indicators of well-being: positive social behavior | ES = 0.13 (0.05 to 0.21) | — | 28 | |||
| Positive and negative indicators of well-being: academic performance | ES = 0.33 (0.17 to 0.49) | — | 8 | |||
| Positive and negative indicators of well-being: conduct problems | ES = 0.14 (0.07 to 0.21) | — | 34 | |||
| Positive and negative indicators of well-being: emotional distress | ES = 0.16 (0.08 to 0.23) | — | 35 | |||
| Positive and negative indicators of well-being: drug use | ES = 0.16 (0.09 to 0.24) | — | 28 | |||
| van Genugten et al160 (2017) | Usual care | Internalizing | ES = 0.20 (0.03 to 0.38) | Low | — | 13 |
| Short-term internalizing | ES = 0.09 (−0.01 to 0.20) | — | 6 | |||
| Long-term self-esteem | ES = 0.29 (0.18 to 0.41) | — | 14 | |||
| Short-term self-esteem | ES = 0.16 (0.05 to 0.26) | — | 5 | |||
| Watson et al162 (2017) | Usual care | Physical activity on classroom behavior | SMD = 0.60 (0.20 to 1.00) | Critically low | 804 | 4 |
| Physical activity on cognitive function | SMD = 0.33 (-0.11 to 0.77) | 1081 | 5 | |||
| Sanchez et al149 (2018) | School-based mental health programs versus no services, peer tutoring or psychologist program | All outcomes | ES = 0.39 (0.29 to 0.49) | Low | — | 43 |
| Externalizing problems | ES = 0.50 (0.35 to 0.63) | — | 27 | |||
| Internalizing problems | ES = 0.30 (0.16 to 0.43) | — | 18 | |||
| Attention problems | ES = 0.10 (0.03 to 0.17) | — | 7 | |||
| Substance use | ES = 0.18 (−0.15 to 0.50) | — | 7 | |||
| Tejada-Gallardo et al158 (2020) | Active control groups (n = 5) and nonactive/waitlist (n = 4) | Subjective well-being (studies postintervention) | ES = 0.24 (0.11 to 0.37) | High | — | 6 |
| Psychological well-being (studies postintervention) | ES = 0.25 (−0.01 to 0.51) | — | 5 | |||
| Depression (studies postintervention) | ES = 0.28 (0.13 to 0.43) | — | 4 | |||
| Anxiety (studies postintervention) | ES = 0.14 (0.04 to 0.24) | — | 4 | |||
| Subjective well-being (studies postintervention excluding low quality studies) | ES = 0.21 (0.05 to 0.37) | — | 4 | |||
| Psychological well-being (studies postintervention excluding low quality studies) | ES = 0.31 (−0.03 to 0.67) | — | 4 | |||
| Depression (studies postintervention excluding low quality studies) | ES = 0.34 (0.24 to 0.44) | — | 3 | |||
| Anxiety (studies postintervention excluding low quality studies) | ES = 0.15 (0.05 to 0.25) | — | 3 | |||
| Subjective well-being (follow-up studies) | ES = 0.13 (0.03 to 0.23) | — | 4 | |||
| Psychological well-being (follow-up studies) | ES = 0.44 (-0.45 to 1.31) | — | 3 | |||
| Depression (follow-up studies) | ES = 0.31 (0.20 to 0.41) | — | 3 | |||
| Anxiety (follow-up studies) | ES = 0.15 (0.05 to 0.26) | — | 3 | |||
| Psychological well-being (follow-up studies excluding low quality studies) | ES = 0.66 (-0.39 to 1.72) | — | 2 | |||
| Depression (follow-up studies excluding low quality studies) | ES = 0.29 (0.11 to 0.47) | — | 2 | |||
| Anxiety (follow-up studies excluding low quality studies) | ES = 0.21 (0.10 to 0.33) | — | 2 | |||
| Caldwell et al112 (2019) | Active control, waitlist, curriculum control, no intervention, or attention control. | Universal primary: depression (waitlist) | SMD = −0.09 (−0.77 to 0.54) | High | 56 620 | 12 (Direct trials: 0) |
| Universal primary: anxiety (waitlist) | SMD = 0.02 (−0.20 to 0.22) | 15 (Direct trials: 0) | ||||
| Universal primary: depression (no intervention) | SMD = 0.13 (−0.40 to 0.65) | 12 (Direct trials: 0) | ||||
| Universal primary: anxiety (no intervention) | SMD = 0.23 (−0.15 to 0.60) | 15 (Direct trials: 0) | ||||
| Universal primary: depression (attention control) | SMD = −0.07 (−0.79 to 0.62) | 12 (Direct trials: 0) | ||||
| Universal primary: anxiety (attention control) | SMD = −0.17 (−0.51 to 0.17) | 15 (Direct trials: 0) | ||||
| Universal primary: depression (CBT) | SMD = −0.13 (−0.44 to 0.17) | 12 (Direct trials: 6) | ||||
| Universal primary: anxiety (CBT) | SMD = −0.07 (−0.23 to 0.05) | 15 (Direct trials: 6) | ||||
| Universal primary: depression (behavioral therapy) | SMD = −0.10 (−1.04 to 0.80) | 12 (Direct trials: 0) | ||||
| Universal secondary: depression (waitlist) | SMD = 0.00 (−0.19 to 0.19) | 34 (Direct trials: 0) | ||||
| Universal secondary: anxiety (waitlist) | SMD = −0.05 (−0.28 to 0.18) | 21 (Direct trials: 0) | ||||
| Universal secondary: depression (no intervention) | SMD = 0.03 (−0.15 to 0.21) | 34 (Direct trials: 0) | ||||
| Universal secondary: anxiety (no intervention) | SMD = −0.07 (−0.34 to 0.20) | 21 (Direct trials: 0) | ||||
| Universal secondary: depression (attention control) | SMD = 0.07 (−0.12 to 0.25) | 34 (Direct trials: 1) | ||||
| Universal secondary: anxiety (attention control) | SMD = −0.15 (−0.51 to 0.16) | 21 (Direct trials: 0) | ||||
| Universal secondary: depression (CBT) | SMD = −0.04 (−0.16 to 0.07) | 34 (Direct trials: 11) | ||||
| Universal secondary: anxiety (CBT) | SMD = −0·15 (−0·34 to 0·04) | 21 (Direct trials: 3) | ||||
| Universal secondary: depression (third wave) | SMD = −0.03 (−0.21 to 0.14) | 34 (Direct trials: 1) | ||||
| Universal secondary: anxiety (third wave) | SMD = 0.03 (−0.14 to 0.20) | 21 (Direct trials: 3) | ||||
| Universal secondary: depression (IPT+CBT) | SMD = −0.19 (−0.46 to 0.08) | 34 (Direct trials: 0) | ||||
| Universal secondary: depression (IPT) | SMD = −0.03 (−0.36 to 0.29) | 34 (Direct trials: 1) | ||||
| Universal secondary: depression (psychoeducation) | SMD = −0.13 (−0.49 to 0.22) | 34 (Direct trials: 1) | ||||
| Universal secondary: depression (behavioral therapy) | SMD = −0.02 (−0.40 to 0.37) | 34 (Direct trials: 1) | ||||
| Universal secondary: anxiety (mindfulness and relaxation) | SMD = −0.65 (−1.14 to −0.19) | 21 (Direct trials: 0) | ||||
| Feiss et al123 (2019) | Active control, usual care, or no control | Stress | t (10) = −0.36; dexp = −0.05 (−0.58 to 0.48); dctrl = 0.05 (−0.18 to 0.28) | Critically low | 420 | 4 |
| Anxiety | t (54) = −3.72; dexp = −0.70 (−0.94 to −0.46); dctrl = −0.14 (−0.26 to −0.01) | 2166 | 20 | |||
| Depression | t (116) = −3.120; dexp = −0.62 (−0.81 to −0.43); dctrl = −0.22 (−0.34 to −0.10) | 6741 | 38 | |||
| Mixed setting interventions | ||||||
| Ciocanel et al114 (2017) | — | Behavioral adjustment, positive social behaviors | ES = 0.04 (-0.11 to 0.21) | Critically low | — | 7 |
| Behavioral adjustment, problem behaviors | ES = 0.05 (−0.00 to 0.11) | — | 16 | |||
| Psychological adjustment, all combined | ES = 0.17 (0.04 to 0.31) | — | 8 | |||
| Psychological adjustment, emotional distress | ES = 0.14 (-0.002 to 0.29) | — | 3 | |||
| Psychological adjustment, self-perceptions | ES = 0.19 (0.02 to 0.37) | — | 6 | |||
| Academic/school outcomes, academic achievement | ES = 0.22 (0.07 to 0.38) | — | 10 | |||
| Academic/school outcomes, academic adjustment | ES = 0.09 (−0.02 to 0.20) | — | 5 | |||
| Sexual health outcomes, risky sexual behavior | ES = 0.05 (−0.00 to 0.12) | — | 11 | |||
| Melendez-Torres et al141 (2016) | — | All substance use outcomes: all time points | d = 0.079 (−0.025 to 0.183) | Moderate | 54 | 9 |
| All substance use outcomes: short-term time points | d = 0.086 (−0.025 to 0.197) | 36 | 9 | |||
| Omnibus substance use outcomes: short-term time points | d = 0.169 (0.012 to 0.326) | 10 | 7 | |||
| Bennett et al110 (2015) | — | Anxiety symptoms | SMD = 0.22 (0.14 to 0.29) | Low | — | 65 |
| Currier et al177 (2007) | — | General helpfulness | ES = 0.14 (0.00 to 0.28) | Critically low | — | 13 |
| Ekeland et al120 (2004) | — | Self-esteem | SMD = 0.51 (0.10 to 0.9) | Low | 161 | 4 |
| Pratt et al143 (2010) | — | Eating attitudes/behaviors and adolescent issues. BMI at 12 − 14-mo follow-up | MD = −0.10 (−0.45 to 0.25) | Low | 1235 | 4 |
| Eating attitudes/behaviors and adolescent issues. Eating attitudes test (EAT) total at 6 − 12 mo follow-up | SMD = 0.01 (−0.13 to 0.15) | 792 | 4 | |||
| Eating attitudes/behaviors and adolescent issues. Eating disorder inventory (EDI), bulimia at 12 − 14-mo follow-up | SMD = −0.03 (−0.16 to 0.10) | 955 | 2 | |||
| Media literacy and advocacy. SATQ awareness at 3 mo | SMD = 0.18 (−0.05 to 0.41) | 297 | 2 | |||
| Media literacy and advocacy. SPA and SPPC global self-worth at 3 mo | SMD = −0.05 (−0.44 to 0.34) | 104 | 2 | |||
| Media literacy and advocacy. Body image assessment at 3 mo | SMD = 0.02 (−0.37 to 0.42) | 127 | 2 | |||
| Self-esteem approach; SPA close friendships at 3 mo | MD = −0.01 (−0.09 to 0.06) | 524 | 2 | |||
| Self-esteem approach; SPA social acceptance at 3 mo | MD = −0.03 (−0.10 to 0.04) | 531 | 2 | |||
| Raposa et al145 (2019) | — | Average effect of youth mentoring on all outcomes | g = 0.21 (0.14 to 0.28) | Critically low | 25 286 | 70 |
| Schleider et al150 (2017) | — | Overall SSI effect | ES = 0.32 (0.17 to 0.46) | Critically low | — | 50 (299 ESs) |
| Spruit et al155 (2016) | — | Internalizing problems | MD = 0.316 (0.073 to 0.558) | Critically low | — | 14 |
| Self-concept | MD = 0.297 (0.127 to 0.468) | — | 24 | |||
| Academic achievement | MD = 0.367 (0.380 to 0.697) | — | 33 | |||
| Overall effect size | ES = 0.33 (0.14 to 0.53) | — | 10 | |||
| Externalizing problems | MD = 0.320 (0.023 to 0.616) | — | 7 | |||
| Salazar de Pablo et al148 (2020) | Universal/selective interventions for good mental health versus a control group | Mental health literacy | ES = 0.685 (P < .001) | Low | 37 533 | 45 |
| Emotions | ES = 0.541 (P < .001) | 7593 | 37 | |||
| Self-perceptions and values | ES = 0.49 (P < .001) | 37 183 | 92 | |||
| Quality of life | ES = 0.457 (P = .001) | 31 276 | 48 | |||
| Cognitive skills | ES = 0.428 (P < .001) | 12 368 | 20 | |||
| Social skills | ES = 0.371 (P < .001) | 62 274 | 94 | |||
| Physical health | ES = 0.285 (P < .001) | 7642 | 25 | |||
| Sexual health | ES = 0.257 (P = .017) | 14 202 | 13 | |||
| Academic/occupational performance | ES = 0.211 (P < .001) | 19 324 | 30 | |||
| Attitude toward mental disorders | ES = 0.177 (P = .006) | 8741 | 16 | |||
| Improve behaviors | ES = 0.065 (P = .23) | 10 474 | 11 | |||
| Improve family/significant relationships | ES =-0.046 (P = .674) | 32 639 | 45 | |||
| Self-management strategies | ES = 0.107 (P = .09) | 41 437 | 86 | |||
| Vallis et al159 (2020) | Cognitive behavior interventions compared with passive control groups | Change in anxiety from pretreatment to postintervention | SMD = −1.34 (−1.59 to −1.09) | Moderate | 1965 | 41 |
| Efficacy of interventions compared with control conditions | SMD = −0.81 (−1.00 to −0.63) | 1525 | 19 | |||
| Anxiety decreased from preintervention to follow-up | SMD = −1.40 (−1.75 to −1.06) | 1525 | 19 | |||
| Efficacy of intervention, reported by assessor | SMD = −0.86 (SE = 0.12) | — | — | |||
| Efficacy of intervention, reported by parent | SMD = −0.83 (SE = 0.15) | — | — | |||
| Efficacy of intervention, participants with anxiety disorders | SMD = −0.91 (SE = 0.13) | — | — | |||
| Efficacy of intervention, participants with anxiety symptoms | SMD = −0.84 (SE = 0.14) | — | — | |||
| Efficacy of intervention, participants with behavioral inhibition | SMD = −0.47 (SE = 0.16) | — | — | |||
| Efficacy of intervention, parent focused | SMD = −0.72 (SE = 0.18) | — | — | |||
| Efficacy of intervention, both parent and child focused | SMD = −0.91 (SE = 0.12) | — | — | |||
| Efficacy of intervention, child focused | SMD = −1.02 (SE = 0.31) | — | — | |||
| Efficacy of intervention, delivered individual in person | SMD = −0.85 (SE = 0.10) | — | — | |||
| Efficacy of intervention, delivered in group in person | SMD = −0.89 (SE = 0.18) | — | — | |||
| Efficacy of intervention, delivered over the internet | SMD = −0.45 (SE = 0.24) | — | — | |||
| Hetrick et al129 (2016) | Overall evidence‐based psychological therapy versus control | Depression diagnoses | RD = −0.05 (−0.08 to −0.02) | High | 3232 | 36 |
| Overall evidence‐based depression prevention program versus control | Depression symptoms self-reported | SMD = −0.21 (−0.27 to −0.15) | 13 829 | 73 | ||
| Universal depression prevention programs compared with any comparator | Depression symptoms self-reported | SMD = −0.11 (−0.17 to −0.05) | 9013 | 31 | ||
| Overall evidence‐based depression prevention program versus control | Depression symptoms clinic related | SMD = −0.23 (−0.41 to −0.05) | 2175 | 11 | ||
| Universal depression prevention programs compared with any comparator | Anxiety symptoms | SMD = −0.09 (−0.17 to −0.01) | 3130 | 8 | ||
| Overall evidence‐based depression prevention program versus control | General and social functioning | SMD = 0.24 (0.06 to 0.41) | 2067 | 10 | ||
| Universal depression prevention programs compared with any comparator | General and social functioning | SMD = 0.16 (0.04 to 0.28) | 1046 | 1 | ||
| Dunning et al119 (2019) | No contact, waitlist, active or attention placebo control | All measures | d = 0.19 (0.14 to 0.23) | High | 3666 | 33 |
| Mindfulness | d = 0.24 (0.01 to 0.46) | 1475 | 11 | |||
| Social behavior | d = 0.16 (−0.05 to 0.37) | 1247 | 10 | |||
| Negative behavior | d = 0.27 (0.07 to 0.47) | 970 | 11 | |||
| Depression | d = 0.27 (0.06 to 0.49) | 1529 | 13 | |||
| Anxiety/stress | d = 0.16 (0.04 to 0.27) | 2319 | 20 | |||
| Executive functions | d = 0.30 (0.12 to 0.49) | 1691 | 15 | |||
| Attention | d = 0.19 (0.04 to 0.34) | 1158 | 8 | |||
| Digital-based interventions | ||||||
| Skeen et al154 (2019) | Universally delivered psychosocial interventions versus control | Positive mental health (face-to-face) | ES = 0.257 (0.097 to 0.416) | Moderate | — | 129 |
| Positive mental health (digital and combined) | ES = 0.197 (0.016 to 0.379) | — | 29 | |||
| Depression and anxiety symptoms (face-to-face) | ES = −0.088 (−0.151 to −0.025) | — | 129 | |||
| Depression and anxiety symptoms (digital and combined) | ES = −0.054 (−0.81 to 0.074) | — | 29 | |||
| Violence, aggression, and bullying (face-to-face) | ES = −0.294 (−0.564 to −0.024) | — | 129 | |||
| Violence, aggression, and bullying (digital and combined) | ES = −0.075 (−0.249 to 0.099) | — | 29 | |||
| Substance use (face-to-face) | ES = −0.04 (−0.117 to 0.037) | — | 129 | |||
| Substance use (digital and combined) | ES = −0.114 (−0.199 to −0.029) | — | 29 |
—, not reported.
Bullying Prevention Meta-Analysis Results
| Source | Comparison | Outcomes | Pooled Effect Estimates (95% CI) | AMSTAR Rating | Total # of People | Total # of Studies |
|---|---|---|---|---|---|---|
| School-based interventions | ||||||
| Gaffney et al68 (2019) | School-based sessions versus no intervention | Cyberbullying perpetration (random effects model) | OR = 1.233 (1.04 to 1.46) | Critically low | — | 18 |
| Cyberbullying perpetration (fixed effects model) | OR = 1.144 (1.06 to 1.23) | — | 18 | |||
| Cyberbullying perpetration (multiplicative variance adjustment) | OR = 1.144 (0.99 to 1.33) | — | 18 | |||
| Victimization (random effects model) | OR = 1.227 (1.05 to 1.44) | — | 19 | |||
| Victimization (fixed effects model) | OR = 1.231 (1.15 to 1.32) | — | 19 | |||
| Victimization (multiplicative variance adjustment) | OR = 1.231 (1.08 to 1.40) | — | 19 | |||
| Jiménez-Barbero et al70 (2016) | Curriculum-based intervention versus no intervention | Bullying or school violence frequency | d = −0.12 (−0.17 to −0.06) | Critically low | — | 14 |
| Victimization frequency | d = −0.09 (−0.18 to 0.01) | — | 8 | |||
| Attitudes favoring bullying or school violence | d = −0.18 (−0.30 to −0.06) | — | 3 | |||
| Attitudes against bullying or school violence | d = 0.06 (0.03 to 0.10) | — | 4 | |||
| School climate | d = −0.03 (−0.08 to 0.02) | — | 3 | |||
| Ttofi and Farrington76 (2009) | — | Bullying (randomized experiments) | OR = 1.10 (0.97 to 1.26) | Critically low | — | 14 |
| Bullying (before-after, experimental-control) | OR = 1.60 (1.45 to 1.77) | — | 14 | |||
| Bullying (other experimental-control) | OR = 1.20 (1.04 to 1.38) | — | 4 | |||
| Bullying (age-cohort designs) | OR = 1.51 (1.35 to 1.70) | — | 9 | |||
| Bullying (total weighted, ean) | OR = 1.36 (1.26 to 1.47) | — | 41 | |||
| Victimization (randomized experiments) | OR = 1.17 (1.00 to 1.37) | — | 11 | |||
| Victimization (before-after, experimental-control) | OR = 1.22 (1.06 to 1.37) | — | 17 | |||
| Victimization (other experimental-control) | OR = 1.43 (1.11 to 1.85) | — | 4 | |||
| Victimization (age-cohort designs) | OR = 1.44 (1.21 to 1.72) | — | 9 | |||
| Victimization (total weighted mean) | OR = 1.29 (1.18 to 1.42) | — | 41 | |||
| Gaffney et al69 (2021) | — | School-bullying perpetration | Multivariance adjustment model (MVA): OR = 1.324 (1.27 to 1.38). Random effects model (RE): OR = 1.308 (1.24 to 1.38) | Low | — | 81 evaluations |
| School-bullying victimization | MVA: OR = 1.248 (1.20 to 1.29). RE: OR = 1.242 (1.18 to 1.30) | — | 84 evaluations | |||
| Fraguas et al67 (2021) | — | Overall bullying (end of intervention) | Cohen d = -0.150 (-0.191 to -0.109) | Critically low | Intervention: 46 847; Control: 45 744 | 45 |
| Overall bullying (follow-up) | d = −0.171 (−0.243 to −0.099) | 11 020; 11 977 | 21 | |||
| Bullying perpetration (end of intervention) | d = −0.111 (−0.146 to −0.077) | 43 199; 42 991 | 35 | |||
| Bullying perpetration (follow-up) | d = −0.175 (−0.276 to −0.073) | 7889; 7993 | 17 | |||
| Bullying exposure (end of intervention) | d = −0.158 (−0.225 to −0.092) | 37 190; 37 001 | 32 | |||
| Bullying exposure (follow-up) | d = −0.122 (−0.173 to −0.071) | 6971; 7629 | 13 | |||
| Cyberbullying (end of intervention) | d = −0.135 (−0.201 to −0.069) | 3271; 2472 | 5 | |||
| Attitudes that discourage bullying (end of intervention) | d = 0.195 (0.145 to 0.245) | 20 537; 17 778 | 25 | |||
| Attitudes that discourage bullying (follow-up) | d = 0.143 (0.083 to 0.202) | 5517; 4596 | 14 | |||
| Attitudes that encourage bullying (end of intervention) | d = −0.115 (−0.184 to −0.046) | 15 884; 14 037 | 15 | |||
| Attitudes that encourage bullying (follow-up) | d = −0.123 (−0.197 to −0.048) | 3329; 3299 | 7 | |||
| Mental health problems (end of intervention) | d = −0.205 (−0.277 to −0.133) | 14 543; 14 649 | 20 | |||
| Mental health problems (follow-up) | d = −0.202 (−0.347 to −0.056) | 1605; 1621 | 6 | |||
| School climate (end of intervention) | d = 0.07 (0.044 to 0.096) | 11 417; 11 995 | 12 | |||
| School climate (follow-up) | d = 0.135 (0.037 to 0.233) | 2647; 2978 | 5 | |||
| Lee et al72 (2015) | No intervention | Bullying victimization | d = −0.151 (−0.201 to −0.101) | Critically low | — | 13 |
| Subgroup analysis: school grade level | Secondary school: d = −0.315; Primary school: d = −0.135; P < .05 | — | — |
| Source | Comparison | Outcomes | Pooled Effect Estimates (95% CI) | AMSTAR Rating | Total # of People | Total # of Studies |
|---|---|---|---|---|---|---|
| School-based interventions | ||||||
| Gaffney et al68 (2019) | School-based sessions versus no intervention | Cyberbullying perpetration (random effects model) | OR = 1.233 (1.04 to 1.46) | Critically low | — | 18 |
| Cyberbullying perpetration (fixed effects model) | OR = 1.144 (1.06 to 1.23) | — | 18 | |||
| Cyberbullying perpetration (multiplicative variance adjustment) | OR = 1.144 (0.99 to 1.33) | — | 18 | |||
| Victimization (random effects model) | OR = 1.227 (1.05 to 1.44) | — | 19 | |||
| Victimization (fixed effects model) | OR = 1.231 (1.15 to 1.32) | — | 19 | |||
| Victimization (multiplicative variance adjustment) | OR = 1.231 (1.08 to 1.40) | — | 19 | |||
| Jiménez-Barbero et al70 (2016) | Curriculum-based intervention versus no intervention | Bullying or school violence frequency | d = −0.12 (−0.17 to −0.06) | Critically low | — | 14 |
| Victimization frequency | d = −0.09 (−0.18 to 0.01) | — | 8 | |||
| Attitudes favoring bullying or school violence | d = −0.18 (−0.30 to −0.06) | — | 3 | |||
| Attitudes against bullying or school violence | d = 0.06 (0.03 to 0.10) | — | 4 | |||
| School climate | d = −0.03 (−0.08 to 0.02) | — | 3 | |||
| Ttofi and Farrington76 (2009) | — | Bullying (randomized experiments) | OR = 1.10 (0.97 to 1.26) | Critically low | — | 14 |
| Bullying (before-after, experimental-control) | OR = 1.60 (1.45 to 1.77) | — | 14 | |||
| Bullying (other experimental-control) | OR = 1.20 (1.04 to 1.38) | — | 4 | |||
| Bullying (age-cohort designs) | OR = 1.51 (1.35 to 1.70) | — | 9 | |||
| Bullying (total weighted, ean) | OR = 1.36 (1.26 to 1.47) | — | 41 | |||
| Victimization (randomized experiments) | OR = 1.17 (1.00 to 1.37) | — | 11 | |||
| Victimization (before-after, experimental-control) | OR = 1.22 (1.06 to 1.37) | — | 17 | |||
| Victimization (other experimental-control) | OR = 1.43 (1.11 to 1.85) | — | 4 | |||
| Victimization (age-cohort designs) | OR = 1.44 (1.21 to 1.72) | — | 9 | |||
| Victimization (total weighted mean) | OR = 1.29 (1.18 to 1.42) | — | 41 | |||
| Gaffney et al69 (2021) | — | School-bullying perpetration | Multivariance adjustment model (MVA): OR = 1.324 (1.27 to 1.38). Random effects model (RE): OR = 1.308 (1.24 to 1.38) | Low | — | 81 evaluations |
| School-bullying victimization | MVA: OR = 1.248 (1.20 to 1.29). RE: OR = 1.242 (1.18 to 1.30) | — | 84 evaluations | |||
| Fraguas et al67 (2021) | — | Overall bullying (end of intervention) | Cohen d = -0.150 (-0.191 to -0.109) | Critically low | Intervention: 46 847; Control: 45 744 | 45 |
| Overall bullying (follow-up) | d = −0.171 (−0.243 to −0.099) | 11 020; 11 977 | 21 | |||
| Bullying perpetration (end of intervention) | d = −0.111 (−0.146 to −0.077) | 43 199; 42 991 | 35 | |||
| Bullying perpetration (follow-up) | d = −0.175 (−0.276 to −0.073) | 7889; 7993 | 17 | |||
| Bullying exposure (end of intervention) | d = −0.158 (−0.225 to −0.092) | 37 190; 37 001 | 32 | |||
| Bullying exposure (follow-up) | d = −0.122 (−0.173 to −0.071) | 6971; 7629 | 13 | |||
| Cyberbullying (end of intervention) | d = −0.135 (−0.201 to −0.069) | 3271; 2472 | 5 | |||
| Attitudes that discourage bullying (end of intervention) | d = 0.195 (0.145 to 0.245) | 20 537; 17 778 | 25 | |||
| Attitudes that discourage bullying (follow-up) | d = 0.143 (0.083 to 0.202) | 5517; 4596 | 14 | |||
| Attitudes that encourage bullying (end of intervention) | d = −0.115 (−0.184 to −0.046) | 15 884; 14 037 | 15 | |||
| Attitudes that encourage bullying (follow-up) | d = −0.123 (−0.197 to −0.048) | 3329; 3299 | 7 | |||
| Mental health problems (end of intervention) | d = −0.205 (−0.277 to −0.133) | 14 543; 14 649 | 20 | |||
| Mental health problems (follow-up) | d = −0.202 (−0.347 to −0.056) | 1605; 1621 | 6 | |||
| School climate (end of intervention) | d = 0.07 (0.044 to 0.096) | 11 417; 11 995 | 12 | |||
| School climate (follow-up) | d = 0.135 (0.037 to 0.233) | 2647; 2978 | 5 | |||
| Lee et al72 (2015) | No intervention | Bullying victimization | d = −0.151 (−0.201 to −0.101) | Critically low | — | 13 |
| Subgroup analysis: school grade level | Secondary school: d = −0.315; Primary school: d = −0.135; P < .05 | — | — |
—, not reported.
Substance Abuse Prevention Meta-Analysis Results
| Source | Comparison | Outcomes | Pooled Effect Estimates (95% CI) | AMSTAR Rating | Total # of People | Total # of Studies |
|---|---|---|---|---|---|---|
| School-based interventions | ||||||
| Faggiano et al89 (2008) | Skills-based interventions compared with usual curricula | Marijuana use | RR = 0.82 (0.73 to 0.92) | Critically low | 7287 | 2 |
| Skills-based interventions compared with usual curricula | Drug use | RR = 0.81 (0.64 to 1.02) | 2371 | 2 | ||
| Skills-based interventions compared with usual curricula | Hard drug use | RR = 0.45 (0.24 to 0.85) | 746 | 2 | ||
| Skills-based interventions compared with usual curricula | Decision-making | SMD = 0.78 (0.46 to 1.09) | 1229 | 2 | ||
| Affective versus usual curricula | Drug knowledge | SMD = 1.88 (1.27 to 2.50) | 63 | 2 | ||
| Affective versus usual curricula | Decision making skills | SMD = 1.35 (0.79 to 1.91) | 63 | 2 | ||
| Knowledge versus usual curricula | Drug knowledge | SMD = 0.91 (0.42 to 1.39) | 220 | 3 | ||
| Knowledge versus usual curricula | Decision making skills | SMD = −0.06 (−0.60 to 0.47) | 55 | 2 | ||
| Skills versus knowledge | Drug knowledge | SMD = 0.02 (−0.18 to 0.22) | 522 | 2 | ||
| Skills versus knowledge | Self-efficacy | SMD = 0.13 (−0.37 to 0.63) | 522 | 2 | ||
| Affective versus knowledge | Drug knowledge | SMD = 0.60 (0.18 to 1.03) | 91 | 2 | ||
| Affective versus knowledge | Decision making skills | SMD = 1.22 (0.33 to 2.12) | 64 | 2 | ||
| Interactive versus passive techniques | Drug knowledge | SMD = 0.02 (−0.18 to 0.22) | 522 | 2 | ||
| Interactive versus passive techniques | Self-efficacy | SMD = 0.13 (−0.37 to 0.63) | 522 | 2 | ||
| Faggiano et al90 (2014) | Social competence versus usual curricula | Marijuana use <12 mo | RR = 0.9 (0.81 to 1.01) | Low | 9456 | 4 |
| Social competence versus usual curricula | Any drug use <12 mo | RR = 0.27 (0.14 to 0.51) | 2512 | 2 | ||
| Social influence versus usual curricula | Marijuana use <12 mo | RR = 0.88 (0.72 to 1.07) | 10c716 | 3 | ||
| Combined versus usual curricula | Marijuana use <12 mo | RR = 0.79 (0.59 to 1.05) | 8701 | 3 | ||
| Combined curricula versus usual curricula | Marijuana use ≥12 mo | RR = 0.83 (0.69 to 0.99) | 26c910 | 6 | ||
| Combined curricula versus usual curricula | Hard drug use ≥12 mo | RR = 0.86 (0.39 to 1.90) | 1066 | 2 | ||
| Espada et al88 (2015) | — | Program effectiveness: global | d = 0.16 (0.10 to 0.22) | Critically low | Total in post-test = 10c956; total in follow-up = 9149 | 36 |
| Knowledge of drugs | d = 0.34 (0.23 to 0.45) | 18 | ||||
| Attitudes toward drugs | d = 0.44 (0.33 to 0.54) | 15 | ||||
| Intention | d = 0.23 (0.14 to 0.32) | 14 | ||||
| Drug consumption | d = 0.18 (0.11 to 0.26) | 20 | ||||
| Alcohol | d = 0.38 (0.27 to 0.49) | 20 | ||||
| Tobacco | d = 0.20 (0.10 to 0.30) | 12 | ||||
| Cannabis | d = 0.19 (0.05 to 0.32) | 11 | ||||
| Other drugs | d = 0.19 (0.10 to 0.28) | 23 | ||||
| Thomas et al101 (2013) | Intervention curricula versus control | Smoking prevention at 1 y follow up | OR = 0.94 (0.85 to 1.05) | High | 142 447 | 49 |
| Social competence and social influences curricula versus control | Smoking prevention at 1 y follow up | OR = 0.49 (0.28 to 0.87) | — | 6 | ||
| Social influences only versus control | Smoking prevention at 1 y follow up | OR = 1.00 (0.88 to 1.13) | — | 16 | ||
| Multimodal interventions versus control | Smoking prevention at 1 y follow up | OR = 0.89 (0.73 to 1.08) | — | 3 | ||
| Intervention curricula versus control | Smoking prevention at longest follow up | OR = 0.88 (0.82 to 0.96) | — | — | ||
| Social competence and social influences curricula versus control | Smoking prevention at longest follow up | OR = 0.50 (0.28 to 0.87) | — | 8 | ||
| Social influences only versus control | Smoking prevention at longest follow up | OR = 0.52 (0.30 to 0.88) | 5 | |||
| Intervention curricula versus control | Change in smoking behavior over time, at 1 y follow up | SMD = 0.04 (0.02 to 0.06) | — | 8 | ||
| Social influences only versus control | Change in smoking behavior over time, at 1 y follow up | SMD = 0.04 (0.03 to 0.06) | — | 6 | ||
| Intervention curricula versus control | Change in smoking behavior over time, at longest follow up | SMD = 0.01 (0.00 to 0.02) | — | 15 | ||
| Social influences only versus control | Change in smoking behavior over time, at longest follow up | SMD = 0.05 (0.03 to 0.06) | — | 10 | ||
| Peer- versus adult-led interventions | Point prevalence of smoking at 1 y follow up | OR = 0.46 (0.26 to 0.84) | — | — | ||
| Point prevalence of smoking at longest follow up | OR = 0.88 (0.81 to 0.96) | — | — | |||
| Social competence versus control | Point prevalence of smoking at longest follow up | OR = 0.52 (0.30 to 0.88) | — | — | ||
| Effect of adding booster sessions | Point prevalence of smoking at 1 y follow up | OR = 0.94 (0.85 to 1.05) | — | 36 | ||
| Point prevalence of smoking at longest follow up | OR = 0.90 (0.83 to 0.97) | — | 66 | |||
| Community-based interventions | ||||||
| Gilligan et al94 (2019) | Any family-based intervention versus no intervention/standard care | Prevalence of alcohol use | SMD = 0.00 (−0.08 to 0.08) | Low | 7490 | 12 |
| Universal family-based intervention versus no intervention/standard care | Prevalence of alcohol use | SMD = 0.02 (−0.06 to 0.11) | 189 | 10 | ||
| Selective or indicated family-based intervention versus no intervention/standard care | Prevalence of alcohol use | SMD = −0.16 (−0.36 to 0.05) | 357 | 2 | ||
| Prevalence of alcohol use, minority ethnic group | SMD = −0.20 (0.42 to 0.02) | 325 | 3 | |||
| Any family-based intervention versus no intervention/standard care | Frequency of alcohol use | SMD = −0.31 (−0.8 to 0.21) | 1835 | 8 | ||
| Universal family-based intervention versus no intervention/standard care | Frequency of alcohol use | SMD = 0.18 (−0.40 to 0.75) | 1090 | 3 | ||
| Selective or indicated family-based intervention versus no intervention/standard care | Frequency of alcohol use | SMD = −0.65 (1.64 to 0.33) | 745 | 5 | ||
| Frequency of alcohol use, ethnic minority groups | SMD = −1.19 (2.83 to 0.46) | 1037 | 3 | |||
| Any family-based intervention versus no intervention/standard care | Vol of alcohol use | SMD = −0.14 (−0.27 to 0.00) | 1825 | 5 | ||
| Universal family-based intervention versus no intervention/standard care | Vol of alcohol use | SMD = −0.21 (−0.32 to −0.10) | 1481 | 3 | ||
| Selective or indicated family-based intervention versus no intervention/standard care | Vol of alcohol use | SMD = 0.06 (−0.15 to 0.27) | 344 | 2 | ||
| Vol of alcohol use, ethnic minority groups | SMD = −0.24 (−0.36 to −0.12) | 1081 | 3 | |||
| Any family-based and adolescent interventions versus interventions with young people alone | Prevalence of alcohol use | SMD = −0.39 (−0.91 to 0.14) | 5640 | 4 | ||
| Universal family-based and adolescent interventions versus interventions with young people alone | Prevalence of alcohol use | SMD = −0.44 (−1.08 to 0.20) | 5351 | 3 | ||
| Frequency of alcohol use | SMD = −0.16 (−0.42 to 0.09) | 915 | 4 | |||
| Carson et al85 (2011) | Multicomponent community-based interventions in influencing smoking behavior versus no intervention | Smoking, daily intervention duration 13 mo or more | OR = 0.89 (0.69 to 1.15) | High | 1304 | 2 |
| Smoking, weekly | OR = 1.00 (0.90 to 1.11) | 11c363 | 6 | |||
| Smoking, monthly | OR = 0.98 (0.84 to 1.14) | 18c677 | 8 | |||
| Smoking, ever smoked. Intervention duration 12 mo or less | OR = 0.82 (0.39 to 1.74) | — | 3 | |||
| Smokeless tobacco use | OR = 0.78 (0.50 to 1.22) | 7667 | 3 | |||
| Intervention duration 13 mo or more. Behaviors rules on smoking | OR = 1.10 (1.02 to 1.18) | — | 2 | |||
| Perceptions, peer smoking | OR = 0.98 (0.78 to 1.24) | — | 2 | |||
| Thomas et al100 (2011) | Mentoring versus no intervention | Alcohol use. Nondrinkers at baseline (12-mo or 18-mo follow-up period) | RR = 0.71 (0.57 to 0.90) | Low | 1116 | 2 |
| Thomas et al102 (2015) | Family-based interventions versus no intervention control. Family plus school intervention versus a school intervention only | New smoking at follow-up. Baseline never smokers only | RR = 0.76 (0.68 to 0.84) | Low | 4810 | 9 |
| New smoking at follow-up. Baseline never smokers only | RR = 0.85 (0.75 to 0.96)) | 2301 | 2 | |||
| O’Connor et al98 (2020) | General prevention trials | Primary drug use outcome (KQ2) for general prevention | SMD = −0.08 (−0.16 to 0.00) | Critically low | 12 801 | 24 |
| Any illicit drug use | OR = 0.82 (0.67 to 1.04) | 9031 | 11 | |||
| Any cannabis use | OR = 0.78 (0.64 to 0.95) | 6520 | 6 | |||
| Times used in previous 3 mo | MD = −0.21 (−0.44 to 0.02) | 3651 | 12 | |||
| Times used cannabis in previous 3 mo | MD = −0.23 (−0.48 to 0.01) | 3616 | 10 | |||
| Primary alcohol outcome | SMD = −0.11 (−0.16 to −0.07) | 12 307 | 24 | |||
| Any alcohol use | OR = 0.79 (0.64 to 0.96) | 5854 | 6 | |||
| Risky alcohol use | OR = 0.92 (0.72 to 1.17) | 5078 | 5 | |||
| Times used alcohol in previous 3 mo | MD = −0.29 (−0.53 to −0.05) | 3192 | 8 | |||
| Primary tobacco outcome | SMD = −0.09 (−0.15 to −0.03) | 8366 | 16 | |||
| Any tobacco use | OR = 0.91 (0.73 to 1.14) | 5373 | 7 | |||
| Times used tobacco in previous 3 mo | MD = −0.30 (−0.58 to 0.02) | 2893 | 8 | |||
| Digital-based interventions | ||||||
| Boumparis et al84 (2019) | Digital intervention to reduce cannabis use versus nonactive control | Cannabis use | OR 0.33 (0.13 to 0.54) | Critically low | 2564 | 6 |
| Mixed setting interventions | ||||||
| Hefler et al95 (2017) | Incentives for preventing smoking | Smoking initiation among children and adolescents- RCTs | RR = 1.00 (0.84 to 1.19) | High | 1108 | 3 |
| Smoking initiation among children and adolescents. Controlled trials | RR = 0.82 (0.63 to 1.08) | 1377 | 3 | |||
| MacArthur et al97 (2018) | Universal school sessions versus no intervention | Tobacco use, short-term usage | ES = 0.77 (0.60 to 0.97) | High | 15 354 | 9 |
| Targeted school sessions versus no intervention | Tobacco use, long-term usage | ES = 0.6 (0.33 to 1.09) | 879 | 2 | ||
| Universal school sessions versus no intervention | Alcohol use, short-term usage | ES = 0.72 (0.56 to 0.92) | 8751 | 8 | ||
| Targeted school sessions versus no intervention | Alcohol use, long-term usage | ES = 1.34 (0.55 to 3.27) | 566 | 1 | ||
| Universal school sessions versus no intervention | Illicit drug use, short-term usage | ES = 0.96 (0.79 to 1.18) | 1299 | 2 | ||
| Targeted school sessions versus no intervention | Illicit drug use, long-term usage | ES = 0.74(0.55 to 1.00) | 11c058 | 5 | ||
| Universal school sessions versus no intervention | Cannabis use, short-term usage | ES = 1.1 (0.69 to 1.76) | 126 | 2 | ||
| Universal school sessions versus no intervention | Cannabis use, short-term usage | ES = 1.02 (0.52 to 2.02) | 380 | 3 | ||
| Targeted school sessions versus no intervention | Cannabis use, long-term usage | ES = 0.79 (0.62 to 1.01) | 4140 | 5 | ||
| Universal school sessions versus no intervention | Cannabis use, long-term usage | ES = 0.82 (0.51 to 1.32) | 806 | 2 | ||
| No intervention/ usual practice | Cannabis use, long-term usage | ES = 1.13 (0.40 to 3.21) | 566 | 1 | ||
| Targeted family sessions versus no intervention | Tobacco use, short-term usage | OR = 0.78 (0.40 to 1.53) | 313 | 2 | ||
| Targeted family sessions versus no intervention | Tobacco use, long-term usage | OR = 0.82 (0.32 to 2.14) | 1177 | 2 | ||
| Universal family sessions versus no intervention | Tobacco use, long-term usage | OR = 0.82 (0.38 to 1.78) | 237 | 1 | ||
| Targeted family sessions versus no intervention | Alcohol use, short-term usage | OR = 0.83 (0.47 to 1.46) | 417 | 3 | ||
| Targeted family sessions versus no intervention | Alcohol use, long-term usage | OR = 0.73 (0.52 to 1.03) | 762 | 2 | ||
| Targeted family sessions versus no intervention | Illicit drug use, short-term usage | OR = 0.94 (0.71 to 1.25) | 638 | 3 | ||
| Universal family sessions versus no intervention | Illicit drug use, short-term usage | OR = 0.74 (0.42 to 1.31) | 69 | 1 | ||
| Targeted family sessions versus no intervention | Illicit drug use, long-term usage | OR = 1.07 (0.19 to 6.21) | 819 | 2 | ||
| Targeted family sessions versus no intervention | Cannabis use, short term usage | OR = 0.69 (0.46 to 1.04) | 362 | 2 | ||
| Targeted family sessions versus no intervention | Cannabis use, long-term usage | OR = 0.53 (0.28 to 1.02) | 340 | 2 | ||
| Universal family sessions versus no intervention | Cannabis use, long-term usage | OR = 0.8 (0.44 to 1.45) | 237 | 1 | ||
| Targeted sessions versus no intervention | Tobacco use, short-term usage | OR = 0.98 (0.35 to 2.73) | 521 | 2 | ||
| Universal sessions versus no intervention | Tobacco use, short-term usage | OR = 1.03 (0.32 to 3.27) | 1549 | 2 | ||
| Targeted sessions versus no intervention | Tobacco use, long-term usage | OR = 1.08 (0.56 to 2.11) | 397 | 1 | ||
| Targeted sessions versus no intervention | Alcohol use, short-term usage | OR = 1.02 (0.80 to 1.31) | 2044 | 4 | ||
| Targeted sessions versus no intervention | Alcohol use, short-term usage | OR = 0.8 (0.58 to 1.11) | 1911 | 4 | ||
| Targeted sessions versus no intervention | Alcohol use, long-term usage | OR = 1.24 (0.69 to 2.24) | 1417 | 3 | ||
| Targeted sessions versus no intervention | Alcohol use, long-term usage | OR = 0.86 (0.47 to 1.55) | 237 | 1 | ||
| Targeted sessions versus no intervention | Illicit drug use, long- term usage | OR = 0.8 (0.52 to 1.24) | 2032 | 4 |
| Source | Comparison | Outcomes | Pooled Effect Estimates (95% CI) | AMSTAR Rating | Total # of People | Total # of Studies |
|---|---|---|---|---|---|---|
| School-based interventions | ||||||
| Faggiano et al89 (2008) | Skills-based interventions compared with usual curricula | Marijuana use | RR = 0.82 (0.73 to 0.92) | Critically low | 7287 | 2 |
| Skills-based interventions compared with usual curricula | Drug use | RR = 0.81 (0.64 to 1.02) | 2371 | 2 | ||
| Skills-based interventions compared with usual curricula | Hard drug use | RR = 0.45 (0.24 to 0.85) | 746 | 2 | ||
| Skills-based interventions compared with usual curricula | Decision-making | SMD = 0.78 (0.46 to 1.09) | 1229 | 2 | ||
| Affective versus usual curricula | Drug knowledge | SMD = 1.88 (1.27 to 2.50) | 63 | 2 | ||
| Affective versus usual curricula | Decision making skills | SMD = 1.35 (0.79 to 1.91) | 63 | 2 | ||
| Knowledge versus usual curricula | Drug knowledge | SMD = 0.91 (0.42 to 1.39) | 220 | 3 | ||
| Knowledge versus usual curricula | Decision making skills | SMD = −0.06 (−0.60 to 0.47) | 55 | 2 | ||
| Skills versus knowledge | Drug knowledge | SMD = 0.02 (−0.18 to 0.22) | 522 | 2 | ||
| Skills versus knowledge | Self-efficacy | SMD = 0.13 (−0.37 to 0.63) | 522 | 2 | ||
| Affective versus knowledge | Drug knowledge | SMD = 0.60 (0.18 to 1.03) | 91 | 2 | ||
| Affective versus knowledge | Decision making skills | SMD = 1.22 (0.33 to 2.12) | 64 | 2 | ||
| Interactive versus passive techniques | Drug knowledge | SMD = 0.02 (−0.18 to 0.22) | 522 | 2 | ||
| Interactive versus passive techniques | Self-efficacy | SMD = 0.13 (−0.37 to 0.63) | 522 | 2 | ||
| Faggiano et al90 (2014) | Social competence versus usual curricula | Marijuana use <12 mo | RR = 0.9 (0.81 to 1.01) | Low | 9456 | 4 |
| Social competence versus usual curricula | Any drug use <12 mo | RR = 0.27 (0.14 to 0.51) | 2512 | 2 | ||
| Social influence versus usual curricula | Marijuana use <12 mo | RR = 0.88 (0.72 to 1.07) | 10c716 | 3 | ||
| Combined versus usual curricula | Marijuana use <12 mo | RR = 0.79 (0.59 to 1.05) | 8701 | 3 | ||
| Combined curricula versus usual curricula | Marijuana use ≥12 mo | RR = 0.83 (0.69 to 0.99) | 26c910 | 6 | ||
| Combined curricula versus usual curricula | Hard drug use ≥12 mo | RR = 0.86 (0.39 to 1.90) | 1066 | 2 | ||
| Espada et al88 (2015) | — | Program effectiveness: global | d = 0.16 (0.10 to 0.22) | Critically low | Total in post-test = 10c956; total in follow-up = 9149 | 36 |
| Knowledge of drugs | d = 0.34 (0.23 to 0.45) | 18 | ||||
| Attitudes toward drugs | d = 0.44 (0.33 to 0.54) | 15 | ||||
| Intention | d = 0.23 (0.14 to 0.32) | 14 | ||||
| Drug consumption | d = 0.18 (0.11 to 0.26) | 20 | ||||
| Alcohol | d = 0.38 (0.27 to 0.49) | 20 | ||||
| Tobacco | d = 0.20 (0.10 to 0.30) | 12 | ||||
| Cannabis | d = 0.19 (0.05 to 0.32) | 11 | ||||
| Other drugs | d = 0.19 (0.10 to 0.28) | 23 | ||||
| Thomas et al101 (2013) | Intervention curricula versus control | Smoking prevention at 1 y follow up | OR = 0.94 (0.85 to 1.05) | High | 142 447 | 49 |
| Social competence and social influences curricula versus control | Smoking prevention at 1 y follow up | OR = 0.49 (0.28 to 0.87) | — | 6 | ||
| Social influences only versus control | Smoking prevention at 1 y follow up | OR = 1.00 (0.88 to 1.13) | — | 16 | ||
| Multimodal interventions versus control | Smoking prevention at 1 y follow up | OR = 0.89 (0.73 to 1.08) | — | 3 | ||
| Intervention curricula versus control | Smoking prevention at longest follow up | OR = 0.88 (0.82 to 0.96) | — | — | ||
| Social competence and social influences curricula versus control | Smoking prevention at longest follow up | OR = 0.50 (0.28 to 0.87) | — | 8 | ||
| Social influences only versus control | Smoking prevention at longest follow up | OR = 0.52 (0.30 to 0.88) | 5 | |||
| Intervention curricula versus control | Change in smoking behavior over time, at 1 y follow up | SMD = 0.04 (0.02 to 0.06) | — | 8 | ||
| Social influences only versus control | Change in smoking behavior over time, at 1 y follow up | SMD = 0.04 (0.03 to 0.06) | — | 6 | ||
| Intervention curricula versus control | Change in smoking behavior over time, at longest follow up | SMD = 0.01 (0.00 to 0.02) | — | 15 | ||
| Social influences only versus control | Change in smoking behavior over time, at longest follow up | SMD = 0.05 (0.03 to 0.06) | — | 10 | ||
| Peer- versus adult-led interventions | Point prevalence of smoking at 1 y follow up | OR = 0.46 (0.26 to 0.84) | — | — | ||
| Point prevalence of smoking at longest follow up | OR = 0.88 (0.81 to 0.96) | — | — | |||
| Social competence versus control | Point prevalence of smoking at longest follow up | OR = 0.52 (0.30 to 0.88) | — | — | ||
| Effect of adding booster sessions | Point prevalence of smoking at 1 y follow up | OR = 0.94 (0.85 to 1.05) | — | 36 | ||
| Point prevalence of smoking at longest follow up | OR = 0.90 (0.83 to 0.97) | — | 66 | |||
| Community-based interventions | ||||||
| Gilligan et al94 (2019) | Any family-based intervention versus no intervention/standard care | Prevalence of alcohol use | SMD = 0.00 (−0.08 to 0.08) | Low | 7490 | 12 |
| Universal family-based intervention versus no intervention/standard care | Prevalence of alcohol use | SMD = 0.02 (−0.06 to 0.11) | 189 | 10 | ||
| Selective or indicated family-based intervention versus no intervention/standard care | Prevalence of alcohol use | SMD = −0.16 (−0.36 to 0.05) | 357 | 2 | ||
| Prevalence of alcohol use, minority ethnic group | SMD = −0.20 (0.42 to 0.02) | 325 | 3 | |||
| Any family-based intervention versus no intervention/standard care | Frequency of alcohol use | SMD = −0.31 (−0.8 to 0.21) | 1835 | 8 | ||
| Universal family-based intervention versus no intervention/standard care | Frequency of alcohol use | SMD = 0.18 (−0.40 to 0.75) | 1090 | 3 | ||
| Selective or indicated family-based intervention versus no intervention/standard care | Frequency of alcohol use | SMD = −0.65 (1.64 to 0.33) | 745 | 5 | ||
| Frequency of alcohol use, ethnic minority groups | SMD = −1.19 (2.83 to 0.46) | 1037 | 3 | |||
| Any family-based intervention versus no intervention/standard care | Vol of alcohol use | SMD = −0.14 (−0.27 to 0.00) | 1825 | 5 | ||
| Universal family-based intervention versus no intervention/standard care | Vol of alcohol use | SMD = −0.21 (−0.32 to −0.10) | 1481 | 3 | ||
| Selective or indicated family-based intervention versus no intervention/standard care | Vol of alcohol use | SMD = 0.06 (−0.15 to 0.27) | 344 | 2 | ||
| Vol of alcohol use, ethnic minority groups | SMD = −0.24 (−0.36 to −0.12) | 1081 | 3 | |||
| Any family-based and adolescent interventions versus interventions with young people alone | Prevalence of alcohol use | SMD = −0.39 (−0.91 to 0.14) | 5640 | 4 | ||
| Universal family-based and adolescent interventions versus interventions with young people alone | Prevalence of alcohol use | SMD = −0.44 (−1.08 to 0.20) | 5351 | 3 | ||
| Frequency of alcohol use | SMD = −0.16 (−0.42 to 0.09) | 915 | 4 | |||
| Carson et al85 (2011) | Multicomponent community-based interventions in influencing smoking behavior versus no intervention | Smoking, daily intervention duration 13 mo or more | OR = 0.89 (0.69 to 1.15) | High | 1304 | 2 |
| Smoking, weekly | OR = 1.00 (0.90 to 1.11) | 11c363 | 6 | |||
| Smoking, monthly | OR = 0.98 (0.84 to 1.14) | 18c677 | 8 | |||
| Smoking, ever smoked. Intervention duration 12 mo or less | OR = 0.82 (0.39 to 1.74) | — | 3 | |||
| Smokeless tobacco use | OR = 0.78 (0.50 to 1.22) | 7667 | 3 | |||
| Intervention duration 13 mo or more. Behaviors rules on smoking | OR = 1.10 (1.02 to 1.18) | — | 2 | |||
| Perceptions, peer smoking | OR = 0.98 (0.78 to 1.24) | — | 2 | |||
| Thomas et al100 (2011) | Mentoring versus no intervention | Alcohol use. Nondrinkers at baseline (12-mo or 18-mo follow-up period) | RR = 0.71 (0.57 to 0.90) | Low | 1116 | 2 |
| Thomas et al102 (2015) | Family-based interventions versus no intervention control. Family plus school intervention versus a school intervention only | New smoking at follow-up. Baseline never smokers only | RR = 0.76 (0.68 to 0.84) | Low | 4810 | 9 |
| New smoking at follow-up. Baseline never smokers only | RR = 0.85 (0.75 to 0.96)) | 2301 | 2 | |||
| O’Connor et al98 (2020) | General prevention trials | Primary drug use outcome (KQ2) for general prevention | SMD = −0.08 (−0.16 to 0.00) | Critically low | 12 801 | 24 |
| Any illicit drug use | OR = 0.82 (0.67 to 1.04) | 9031 | 11 | |||
| Any cannabis use | OR = 0.78 (0.64 to 0.95) | 6520 | 6 | |||
| Times used in previous 3 mo | MD = −0.21 (−0.44 to 0.02) | 3651 | 12 | |||
| Times used cannabis in previous 3 mo | MD = −0.23 (−0.48 to 0.01) | 3616 | 10 | |||
| Primary alcohol outcome | SMD = −0.11 (−0.16 to −0.07) | 12 307 | 24 | |||
| Any alcohol use | OR = 0.79 (0.64 to 0.96) | 5854 | 6 | |||
| Risky alcohol use | OR = 0.92 (0.72 to 1.17) | 5078 | 5 | |||
| Times used alcohol in previous 3 mo | MD = −0.29 (−0.53 to −0.05) | 3192 | 8 | |||
| Primary tobacco outcome | SMD = −0.09 (−0.15 to −0.03) | 8366 | 16 | |||
| Any tobacco use | OR = 0.91 (0.73 to 1.14) | 5373 | 7 | |||
| Times used tobacco in previous 3 mo | MD = −0.30 (−0.58 to 0.02) | 2893 | 8 | |||
| Digital-based interventions | ||||||
| Boumparis et al84 (2019) | Digital intervention to reduce cannabis use versus nonactive control | Cannabis use | OR 0.33 (0.13 to 0.54) | Critically low | 2564 | 6 |
| Mixed setting interventions | ||||||
| Hefler et al95 (2017) | Incentives for preventing smoking | Smoking initiation among children and adolescents- RCTs | RR = 1.00 (0.84 to 1.19) | High | 1108 | 3 |
| Smoking initiation among children and adolescents. Controlled trials | RR = 0.82 (0.63 to 1.08) | 1377 | 3 | |||
| MacArthur et al97 (2018) | Universal school sessions versus no intervention | Tobacco use, short-term usage | ES = 0.77 (0.60 to 0.97) | High | 15 354 | 9 |
| Targeted school sessions versus no intervention | Tobacco use, long-term usage | ES = 0.6 (0.33 to 1.09) | 879 | 2 | ||
| Universal school sessions versus no intervention | Alcohol use, short-term usage | ES = 0.72 (0.56 to 0.92) | 8751 | 8 | ||
| Targeted school sessions versus no intervention | Alcohol use, long-term usage | ES = 1.34 (0.55 to 3.27) | 566 | 1 | ||
| Universal school sessions versus no intervention | Illicit drug use, short-term usage | ES = 0.96 (0.79 to 1.18) | 1299 | 2 | ||
| Targeted school sessions versus no intervention | Illicit drug use, long-term usage | ES = 0.74(0.55 to 1.00) | 11c058 | 5 | ||
| Universal school sessions versus no intervention | Cannabis use, short-term usage | ES = 1.1 (0.69 to 1.76) | 126 | 2 | ||
| Universal school sessions versus no intervention | Cannabis use, short-term usage | ES = 1.02 (0.52 to 2.02) | 380 | 3 | ||
| Targeted school sessions versus no intervention | Cannabis use, long-term usage | ES = 0.79 (0.62 to 1.01) | 4140 | 5 | ||
| Universal school sessions versus no intervention | Cannabis use, long-term usage | ES = 0.82 (0.51 to 1.32) | 806 | 2 | ||
| No intervention/ usual practice | Cannabis use, long-term usage | ES = 1.13 (0.40 to 3.21) | 566 | 1 | ||
| Targeted family sessions versus no intervention | Tobacco use, short-term usage | OR = 0.78 (0.40 to 1.53) | 313 | 2 | ||
| Targeted family sessions versus no intervention | Tobacco use, long-term usage | OR = 0.82 (0.32 to 2.14) | 1177 | 2 | ||
| Universal family sessions versus no intervention | Tobacco use, long-term usage | OR = 0.82 (0.38 to 1.78) | 237 | 1 | ||
| Targeted family sessions versus no intervention | Alcohol use, short-term usage | OR = 0.83 (0.47 to 1.46) | 417 | 3 | ||
| Targeted family sessions versus no intervention | Alcohol use, long-term usage | OR = 0.73 (0.52 to 1.03) | 762 | 2 | ||
| Targeted family sessions versus no intervention | Illicit drug use, short-term usage | OR = 0.94 (0.71 to 1.25) | 638 | 3 | ||
| Universal family sessions versus no intervention | Illicit drug use, short-term usage | OR = 0.74 (0.42 to 1.31) | 69 | 1 | ||
| Targeted family sessions versus no intervention | Illicit drug use, long-term usage | OR = 1.07 (0.19 to 6.21) | 819 | 2 | ||
| Targeted family sessions versus no intervention | Cannabis use, short term usage | OR = 0.69 (0.46 to 1.04) | 362 | 2 | ||
| Targeted family sessions versus no intervention | Cannabis use, long-term usage | OR = 0.53 (0.28 to 1.02) | 340 | 2 | ||
| Universal family sessions versus no intervention | Cannabis use, long-term usage | OR = 0.8 (0.44 to 1.45) | 237 | 1 | ||
| Targeted sessions versus no intervention | Tobacco use, short-term usage | OR = 0.98 (0.35 to 2.73) | 521 | 2 | ||
| Universal sessions versus no intervention | Tobacco use, short-term usage | OR = 1.03 (0.32 to 3.27) | 1549 | 2 | ||
| Targeted sessions versus no intervention | Tobacco use, long-term usage | OR = 1.08 (0.56 to 2.11) | 397 | 1 | ||
| Targeted sessions versus no intervention | Alcohol use, short-term usage | OR = 1.02 (0.80 to 1.31) | 2044 | 4 | ||
| Targeted sessions versus no intervention | Alcohol use, short-term usage | OR = 0.8 (0.58 to 1.11) | 1911 | 4 | ||
| Targeted sessions versus no intervention | Alcohol use, long-term usage | OR = 1.24 (0.69 to 2.24) | 1417 | 3 | ||
| Targeted sessions versus no intervention | Alcohol use, long-term usage | OR = 0.86 (0.47 to 1.55) | 237 | 1 | ||
| Targeted sessions versus no intervention | Illicit drug use, long- term usage | OR = 0.8 (0.52 to 1.24) | 2032 | 4 |
—, not reported.
Self-Harm Meta-Analysis Results
| Source | Comparison | Outcomes | Pooled Effect Estimates (95% CI) | AMSTAR Rating | Total # of People | Total # of Studies |
|---|---|---|---|---|---|---|
| Mixed setting interventions | ||||||
| Morken et al63 (2019) | Children and adolescents between the ages of 10 and 23. School-based suicide prevention programs versus TAU, alternative interventions, wait list or no intervention | Suicidal ideation | RR = 0.67 (0.48 to 0.93) | Critically low | 13936 | 5 |
| Suicide attempts (3- to 12-mo follow-up) | RR = 0.53 (0.36 to 0.80) | 14 042 | 5 | |||
| Adolescents, 12- to 19-y-olds, with a history of multiple episodes of self-harm. Dialectical behavior therapy for adolescents (DBT-A). Control: TAU or enhanced TAU | Repetition of self-harm; between 16 wk and 6 mo follow-up-period | OR = 0.72 (0.12 to 4.40) | 105 | 2 | ||
| Frequency of self-harm; between 16 wk and 6 mo follow-up-period | MD = −0.79 (−2.78 to 1.20) | 104 | 2 | |||
| Suicidal ideation; between 16 wk and 12 mo follow-up-period | SMD = −0.62 (−1.07 to −0.16) | 100 | 2 | |||
| Adolescents, 12- to 17-y-olds, referred to child and adolescent services after an episode of intentional self-injury or self-poisoning, irrespective of intent. Developmental group therapy versus TAU | Repetition of self-harm; 6-mo follow-up period | OR = 1.72 (0.56 to 5.24) | 430 | 2 | ||
| Repetition of self-harm; 12-mo follow-up period | OR = 0.80 (0.22 to 2.97) | 490 | 3 | |||
| Depression (scale not reported); 12-mo follow-up period | MD = −0.93 (−4.03 to 2.17) | 473 | 3 | |||
| Suicidal ideation (scale not reported); 12- mo follow-up period | MD = −1.51 (−9.62 to 6.59) | 471 | 3 |
| Source | Comparison | Outcomes | Pooled Effect Estimates (95% CI) | AMSTAR Rating | Total # of People | Total # of Studies |
|---|---|---|---|---|---|---|
| Mixed setting interventions | ||||||
| Morken et al63 (2019) | Children and adolescents between the ages of 10 and 23. School-based suicide prevention programs versus TAU, alternative interventions, wait list or no intervention | Suicidal ideation | RR = 0.67 (0.48 to 0.93) | Critically low | 13936 | 5 |
| Suicide attempts (3- to 12-mo follow-up) | RR = 0.53 (0.36 to 0.80) | 14 042 | 5 | |||
| Adolescents, 12- to 19-y-olds, with a history of multiple episodes of self-harm. Dialectical behavior therapy for adolescents (DBT-A). Control: TAU or enhanced TAU | Repetition of self-harm; between 16 wk and 6 mo follow-up-period | OR = 0.72 (0.12 to 4.40) | 105 | 2 | ||
| Frequency of self-harm; between 16 wk and 6 mo follow-up-period | MD = −0.79 (−2.78 to 1.20) | 104 | 2 | |||
| Suicidal ideation; between 16 wk and 12 mo follow-up-period | SMD = −0.62 (−1.07 to −0.16) | 100 | 2 | |||
| Adolescents, 12- to 17-y-olds, referred to child and adolescent services after an episode of intentional self-injury or self-poisoning, irrespective of intent. Developmental group therapy versus TAU | Repetition of self-harm; 6-mo follow-up period | OR = 1.72 (0.56 to 5.24) | 430 | 2 | ||
| Repetition of self-harm; 12-mo follow-up period | OR = 0.80 (0.22 to 2.97) | 490 | 3 | |||
| Depression (scale not reported); 12-mo follow-up period | MD = −0.93 (−4.03 to 2.17) | 473 | 3 | |||
| Suicidal ideation (scale not reported); 12- mo follow-up period | MD = −1.51 (−9.62 to 6.59) | 471 | 3 |
Nurturing Care Meta-Analysis Results
| Source | Comparison | Outcomes | Pooled Effect Estimates (95% CI) | AMSTAR Rating | Total # of People | Total # of Studies |
|---|---|---|---|---|---|---|
| School-based interventions | ||||||
| Bastounis et al17 (2016) | School-based Penn Resiliency Program (PRP) compared with active control, nonintervention, and waiting list | Depression | MD =−0.23 (−1.09 to 0.62) | Critically low | 4744 | 9 |
| Anxiety | SMD = 0.13 (0.00 to 0.26) | 1518 | 4 | |||
| Explanatory style | MD = 0.80 (−1.04 to 2.63) | 1063 | 3 | |||
| Durlak et al23 (2010) | After-school programs (ASP) compared with controls | Overall effect of ASP | SMD = 0.22 (0.16 to 0.29) | Critically low | — | 68 |
| Child self-perception | SMD = 0.34 (0.23 to 0.46) | — | 23 | |||
| School bonding | SMD = 0.14 (0.03 to 0.25) | — | 28 | |||
| Positive social behavior | SMD = 0.19 (0.10 to 0.29) | — | 36 | |||
| Problem behaviors | SMD = 0.19 (0.10 to 0.27) | — | 43 | |||
| Drug use | SMD = 0.10 (0.00 to 0.20) | — | 28 | |||
| Achievement test scores | SMD = 0.17 (0.06 to 0.29) | — | 20 | |||
| School grades | SMD = 0.12 (0.01 to 0.23) | — | 25 | |||
| School attendance | SMD = 0.10 (−0.01 to 0.20) | — | 21 | |||
| Durlak et al24 (2011) | School-based universal social and emotional learning (SEL) programs compared with controls | Grand study-level mean for improvement in students’ skills, attitudes, and behaviors | ES = 0.30 (0.26 to 0.33) | Critically low | 270 034 | 213 interventions |
| SEL skills | ES = 0.57 (0.48 to 0.67) | — | 68 interventions | |||
| Attitudes | ES = 0.23 (0.16 to 0.30) | — | 106 interventions | |||
| Positive social behavior | ES = 0.24 (0.16 to 0.32) | — | 86 interventions | |||
| Conduct problems | ES = 0.22 (0.16 to 0.29) | — | 112 interventions | |||
| Emotional distress | ES = 0.24 (0.14 to 0.35) | — | 49 interventions | |||
| Academic performance | ES = 0.27 (0.15 to 0.39) | — | 35 interventions | |||
| Hodder et al25 (2017) | Universal school-based resilience intervention compared with no intervention, usual practice, attention only or an alternate intervention | Tobacco: overall analysis | OR = 0.96 (0.85 to 1.08) | Moderate | — | 12 |
| Alcohol: overall analysis | OR = 0.86 (0.73 to 1.02) | — | 13 | |||
| Illicit substances: overall analysis | OR = 0.78 (0.66 to 0.93) | — | 10 | |||
| Klingbeil et al27 (2017) | Mindfulness-based interventions compared with control | Average treatment effect in pretest-posttest | g = 0.305 (0.223 to 0.387) | Low | 1075 | 30 |
| Average treatment effect in controlled studies | g = 0.322 (0.242 to 0.402) | 4811 | 48 | |||
| Average treatment effect in pretest-posttest: follow-up | g = 0.462 (0.202 to 0.723) | 532 | 12 | |||
| Average treatment effect in controlled studies: follow-up | g = 0.402 (0.220 to 0.584) | 1501 | 12 | |||
| Mindfulness | pretest-posttest: g = 0.174 (0.032 to 0.316). Controlled: g = 0.510 (0.082 to 0.939) | pretest-posttest: 213; Controlled: 1108 | pretest-posttest: 12; Controlled: 15 | |||
| Attention | pretest-posttest: g = 0.324 (0.232 to 0.415). Controlled: g = 0.291 (0.124 to 0.456) | 557; 1243 | 8; 10 | |||
| Meta-cognition and cognitive flexibility | n/a; Controlled: g = 0.404 (0.011 to 0.798) | 33; 806 | 3; 9 | |||
| Emotional or behavioral regulation | pretest-posttest: g = 0.297 (0.213 to 0.381); Controlled: g = 0.322 (0.180 to 0.464) | 976; 1404 | 17; 16 | |||
| Academic achievement and school functioning | pretest-posttest: g = 0.321 (0.132 to 0.511); Controlled: g = 0.393 (−0.153 to 0.939) | 516; 434 | 7; 5 | |||
| Externalizing problems | pretest-posttest: g = 0.145 (0.032 to 0.258); Controlled: g = 0.296 (0.087 to 0.505) | 388; 871 | 10; 9 | |||
| Internalizing problems | Pretest-posttest: g = 0.262 (0.161 to 0.362); Controlled: g = 0.392 (0.267 to 0.391) | 527; 2941 | 17; 29 | |||
| Negative emotions and subjective distress | Pretest-posttest: g = 0.323 (0.125 to 0.521); Controlled: g = 0.254 (0.123 to 0.385) | 408; 2342 | 12; 21 | |||
| Positive emotions and self-appraisal | Pretest-posttest: g = 0.342 (0.165 to 0.519); Controlled: g = 0.280 (0.092 to 0.468) | 487; 2321 | 15; 22 | |||
| Physical health | Pretest-posttest: g = 0.492 (−0.130 to 1.113); Controlled: g = 0.282 (0.173 to 0.391) | 308; 994 | 7; 14 | |||
| Social competence and prosocial behaviors | Pretest-posttest: g = 0.214 (0.092 to 0.335); Controlled: g = 0.368 (0.165 to 0.570) | 905; 1105 | 12; 12 | |||
| Melendez-Torres et al30 (2018) | Integrated academic and health education compared with treatment as usual | Alcohol (school grade7–9 y; aged 11–14 y) | SMD = −0.11 (−0.23 to 0.004) | High | 23 | 5 |
| Smoking (school grade 7–9; aged 11–14 y) | SMD = −0.05 (−0.12 to 0.02) | 13 | 5 | |||
| Illicit drug use: marijuana (school grade 7–9; aged 11–14 y) | SMD = −0.10 (−0.16 to −0.04) | 10 | 5 | |||
| Illicit drug use (school grade 7–9; aged 11–14 y) | SMD = −0.07 (−0.14 to −0.01) | 14 | 5 | |||
| All drug use outcomes (school grade 7–9; aged 11-14 y) | SMD = −0.09 (−0.17 to −0.01) | 55 | 5 | |||
| Alcohol (school grade 10–11; aged 14–16 y) | SMD = −0.01 (−0.09 to 0.06) | 15 | 3 | |||
| Smoking (school grade 10–11; aged 14–16 y) | SMD = −0.08 (−0.15 to −0.01) | 9 | 3 | |||
| Illicit drug use: marijuana (school grade 10–11; aged 14–16 y) | SMD = −0.10 (−0.17 to −0.03) | 10 | 3 | |||
| All drug use outcomes (school grade 10–11; aged 14–16 y) | SMD = −0.06 (−0.09 to −0.02) | 34 | 3 | |||
| Maynard et al29 (2017) | Mindfulness‐based interventions compared with wait list control, no treatment, treatment-as-usual or alternative treatment groups | Cognitive outcomes | ES = 0.25 (0.06 to 0.43) | High | — | 10 |
| Academic outcomes | ES = 0.27 (-0.04 to 0.58) | — | 5 | |||
| Behavioral outcomes | ES = 0.14 (-0.02 to 0.30) | — | 13 | |||
| Socioemotional outcomes | ES = 0.22 (0.14 to 0.30) | — | 28 | |||
| Walsh et al56 (2018) | School-based education programs for the prevention of child sexual abuse compared with wait-listed control | Protective behaviors | OR = 5.71 (1.98 to 16.51) | High | 102 | 2 |
| Questionnaire-based knowledge | SMD = 0.61 (0.45 to 0.78) | 4657 | 18 | |||
| Vignette-based knowledge | SMD = 0.45 (0.24 to 0.65) | 1688 | 11 | |||
| Retention of knowledge over time | SMD = 0.78 (0.38 to 1.17) | 956 | 4 | |||
| Disclosure of previous or current sexual abuse | OR = 3.56 (1.13 to 11.24) | 1788 | 3 | |||
| Wilson et al34 (2006) | No-treatment or wait-list control. | Aggressive and disruptive behavior | Random effects mean 0.21 (P < .01) | Critically low | — | 73 |
| Wilson et al35 (2006) | No-treatment or wait-list Control. | Aggressive and disruptive behavior | Random effects mean 0.26 (P < .01) | Critically low | — | 47 |
| Community-based interventions | ||||||
| Dowdall et al21 (2020) | Shared picture book readings interventions compared with Passive and active controls | Expressive language outcomes | d = 0.41 (0.20 to 0.61) | Low | 1664 | 16 |
| Receptive language outcomes | d = 0.26 (0.12 to 0.40) | 1871 | 16 | |||
| Caregiving competence outcomes | d = 1.01 (0.40 to 1.63) | 481 | 6 | |||
| Jackson et al26 (2016) | Exercise programs compared with normal activity or waitlist control | Inhibitory control (RCTs) | d = 0.2 (0.03 to 0.37) | Critically low | 770 | 8 |
| Liu et al28 (2015) | Physical activity (PA) interventions compared with nonphysical activity control or comparison group | General self-outcomes (RCTs) | g = 0.29 (0.14 to 0.45) | Critically low | — | 18 |
| Self-concept (RCTs) | g = 0.49 (0.10 to 0.88) | — | 18 | |||
| Self-worth (RCTs) | g = 0.31 (0.13 to 0.49) | — | 18 | |||
| General self-outcomes (non-RCTs) | g = 0.33 (−0.35 to 1.01) | — | 6 | |||
| Macbeth et al46 (2015) | Mellow parenting interventions versus wait-list control or treatment as usual | Effect of mellow parenting on child outcomes: Dersimonian-Laird random effects model | SMD = −0.40 (−0.77 to −0.02) | Critically low | — | 3 |
| Effect of mellow parenting on child outcomes: fixed effects model | SMD = −0.40 (−0.77 to −0.02) | — | 3 | |||
| McGinn et al60 (2020) | Traditional care case processing compared with no intervention or alternative treatment | Reunification of children with families or maintenance of in-home care | OR = 1.69 (1.03 to 2.78) | High | 86 305 | 10 |
| Continued maltreatment, effects from nonrandomized studies | OR = 0.73 (0.48 to 1.11) | 1779 | 5 | |||
| Continued maltreatment, effects from RCTs | OR = 1.29 (0.85 to 1.98) | 1158 | 4 | |||
| Kinship placements | OR = 1.29 (0.94 to 1.76) | 85 537 | 5 | |||
| Leijten et al45 (2016) | Transported or homegrown interventions compared with no-treatment, waitlist, minimal intervention, or care as usual. | Effectiveness between transported or homegrown interventions to reduce disruptive child behavior | MD = 0.10 (−0.08 to 0.29) | Moderate | 374 | 129 |
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: Incredible years program | MD = 0.12 (−0.15 to 0.40) | 82 | 31 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: triple P program | MD = 0.27 (−0.02 to 0.56) | 113 | 39 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: PCIT program | MD = 0.40 (−0.26 to 1.05) | 51 | 11 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: Parent Management Training (Oregon) program | MD = 0.39 (−0.93 to 1.70) | 24 | 9 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: US and Canada | MD = 0.19 (−0.60 to 0.98) | 155 | 51 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: Australia | MD = −0.19 (−0.63 to 0.26) | 101 | 30 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: non-English speaking European countries (Continental Europe and Iceland) | MD = 0.38 (−0.11 to 0.86) | 60 | 22 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: English-speaking European countries (UK/Ireland) | MD = −0.07 (−0.42 to 0.28) | 38 | 16 | |||
| Sanders et al48 (2014) | Multilevel triple P-positive pParenting program system compared with nonactive control group (ie, a waitlist control group or usual care) | Child SEB outcomes: all levels combined | SMD = 0.525 (0.358 to 0.692) | Low | — | 56 |
| Parenting practices: all levels combined | SMD = 0.498 (0.362 to 0.634) | — | 48 | |||
| Parenting satisfaction and efficacy: all levels combined | SMD = 0.551 (0.372 to 0.730) | — | 41 | |||
| Parental adjustment: all levels combined | SMD = 0.481 (0.321 to 0.641) | — | 45 | |||
| Parental relationship: all levels combined | SMD = 0.230 (0.136 to 0.325) | — | 37 | |||
| Child observation: all levels combined | SMD = 0.400 (0.070 to 0.730) | — | 13 | |||
| Parent observation: all levels combined | SMD = 0.249 (0.031 to 0.467) | — | 11 | |||
| Valentine et al50 (2019) | Effects of families and schools together (FAST), relative to waiting list, usual or alternative services, or no intervention. | Child school performance | SMD = −0.02 (−0.11 to 0.08) | High | 6276 | 4 |
| Child internalizing behavior: long-term follow-up, parent report | SMD = 0.03 (−0.11 to 0.17) | 908 | 4 | |||
| Child internalizing behavior: long-term follow-up, teacher report | SMD = −0.06 (−0.19 to 0.07) | 912 | 4 | |||
| Child externalizing behavior: long-term follow-up, parent report | SMD = −0.19 (−0.32 to −0.05) | 754 | 4 | |||
| Child externalizing behavior: long-term follow-up, teacher report | SMD = −0.10 (−0.24 to 0.04) | 912 | 4 | |||
| Child school attendance | SMD = 0.03 (−0.10 to 0.16) | 1430 | 2 | |||
| Parental engagement with education | SMD = 0.03 (−0.07 to 0.12) | 1232 | 2 | |||
| Family relationships | SMD = 0.08 (−0.03 to 0.19) | 2569 | 4 | |||
| Vlahovicova et al51 (2017) | Manualized interventions against treatment as usual | Risk of re-abuse | RD = −0.11 (-0.22 to -0.004) | Low | — | 4 |
| Risk of re-abuse: sensitivity analysis | RR = 0.76 (0.54 to 1.07) | — | 4 | |||
| Zief et al37 (2006) | After-school programs compared with no intervention | Self-care | SMD = 0.503 (0.097 to 0.910) | Critically low | 101 | 3 |
| Athletic activities | RD = 0.073 (0.02 to 0.144) | 661 | 2 | |||
| Art/music/drama dance activities | RD = 0.083 (0.01 to -0.153) | 661 | 2 | |||
| School attendance | SMD = −0.001 (−0.125 to 0.122) | 1072 | 2 | |||
| Television viewing | SMD = −0.066 (−0.1224 to 0.091) | 661 | 2 | |||
| Reading scores | SMD = 0.028 (−0.101 to 0.157) | 983 | 2 | |||
| Grade point average (GPA) | SMD = 0.083 (−0.032 to 0.199) | 1004 | 5 | |||
| Mixed setting interventions | ||||||
| Burkey et al40 (2018) | No treatment, wait list controls, treatment as usual, groups or inactive controls. | Child behavior problems (all interventions) | SMD = −0.38 (−0.51 to −0.24) | High | 4441 | 24 |
| Child behavior problems, child-focused interventions | SMD = −0.39 (−0.62 to −0.16) | — | 13 | |||
| Child behavior problems, parent-focused interventions | SMD = −0.43 (−0.66 to −0.20) | — | 8 | |||
| Child behavior problems, classroom based | SMD = −0.49 (−0.71 to −0.28) | — | 2 | |||
| Child behavior problems, multicomponent interventions | SMD = −0.19 (−0.50 to 0.12) | — | 3 | |||
| Child behavior problems, children aged 0–5 | SMD = −0.36 (−0.66 to −0.06) | — | 4 | |||
| Child behavior problems, children aged 6–12 | SMD = −0.42 (−0.65 to −0.19) | — | 11 | |||
| Child behavior problems, children aged 13–18 | SMD = −0.43 (−0.76 to −0.10) | — | 6 | |||
| Child behavior problems, prevention focused | SMD = −0.25 (−0.41 to −0.09) | — | 11 | |||
| Child behavior problems, treatment focused | SMD = −0.56 (−0.51 to −0.24) | — | 11 | |||
| Child behavior problems, specialist delivered | SMD = −0.58 (−0.75 to −0.43) | — | 12 | |||
| Child behavior problems, nonspecialist delivered | SMD = −0.15 (−0.27 to −0.03) | — | 11 | |||
| Durlak and Taylor22 (2007) | Positive youth development (PYD) programs compared with control groups | Systemic change efforts at psychosocial environment: classroom | Post: ES = 0.47; Follow-up: ES = 0.42 | Critically low | — | Post: 10; Follow-up: 1 |
| Systemic change efforts at psychosocial environment: school-wide | Post: ES = 0.74; Follow-up: — | — | Post: 8; Follow-up: 0 | |||
| Systemic change efforts at classroom level | Post: ES = 0.78; Follow-up::— | — | Post: 4; Follow-up: 0 | |||
| Parenting practices | Post: ES = 0.41; Follow-up: ES = 0.49 | — | Post: 33; Follow-up: 17 | |||
| Family environment | Post: ES = 0.34; Follow-up: ES = 0.94 | — | Post: 25; Follow-up: 3 | |||
| Bonding to community adults | Post: ES = −0.26; Follow-up: — | — | Post: 2: Follow-up: 0 | |||
| Family-school relationships | Post: ES = 0.49; Follow-up: ES = −0.09 | — | Post: 6: Follow-up: 1 | |||
| Morton and Montgomery31 (2011) | Youth empowerment programs (YEPs) compared with no intervention, waitlist, and alternative intervention | General self-efficacy | SMD = 0.19 (-0.12 to 0.49) | High | 167 | 2 |
| Ulferts et al33 (2019) | Initiatives aimed at enhancing the quality of early childhood education | Global process quality on children's outcomes (language/ literacy and mathematics) | ES = 0.11 (0.04 to 0.18) | Low | — | 13 |
| Domain-specific process quality on children's outcomes (language/literacy and mathematics) | ES = 0.10 (0.05 to 0.15) | — | 11 |
| Source | Comparison | Outcomes | Pooled Effect Estimates (95% CI) | AMSTAR Rating | Total # of People | Total # of Studies |
|---|---|---|---|---|---|---|
| School-based interventions | ||||||
| Bastounis et al17 (2016) | School-based Penn Resiliency Program (PRP) compared with active control, nonintervention, and waiting list | Depression | MD =−0.23 (−1.09 to 0.62) | Critically low | 4744 | 9 |
| Anxiety | SMD = 0.13 (0.00 to 0.26) | 1518 | 4 | |||
| Explanatory style | MD = 0.80 (−1.04 to 2.63) | 1063 | 3 | |||
| Durlak et al23 (2010) | After-school programs (ASP) compared with controls | Overall effect of ASP | SMD = 0.22 (0.16 to 0.29) | Critically low | — | 68 |
| Child self-perception | SMD = 0.34 (0.23 to 0.46) | — | 23 | |||
| School bonding | SMD = 0.14 (0.03 to 0.25) | — | 28 | |||
| Positive social behavior | SMD = 0.19 (0.10 to 0.29) | — | 36 | |||
| Problem behaviors | SMD = 0.19 (0.10 to 0.27) | — | 43 | |||
| Drug use | SMD = 0.10 (0.00 to 0.20) | — | 28 | |||
| Achievement test scores | SMD = 0.17 (0.06 to 0.29) | — | 20 | |||
| School grades | SMD = 0.12 (0.01 to 0.23) | — | 25 | |||
| School attendance | SMD = 0.10 (−0.01 to 0.20) | — | 21 | |||
| Durlak et al24 (2011) | School-based universal social and emotional learning (SEL) programs compared with controls | Grand study-level mean for improvement in students’ skills, attitudes, and behaviors | ES = 0.30 (0.26 to 0.33) | Critically low | 270 034 | 213 interventions |
| SEL skills | ES = 0.57 (0.48 to 0.67) | — | 68 interventions | |||
| Attitudes | ES = 0.23 (0.16 to 0.30) | — | 106 interventions | |||
| Positive social behavior | ES = 0.24 (0.16 to 0.32) | — | 86 interventions | |||
| Conduct problems | ES = 0.22 (0.16 to 0.29) | — | 112 interventions | |||
| Emotional distress | ES = 0.24 (0.14 to 0.35) | — | 49 interventions | |||
| Academic performance | ES = 0.27 (0.15 to 0.39) | — | 35 interventions | |||
| Hodder et al25 (2017) | Universal school-based resilience intervention compared with no intervention, usual practice, attention only or an alternate intervention | Tobacco: overall analysis | OR = 0.96 (0.85 to 1.08) | Moderate | — | 12 |
| Alcohol: overall analysis | OR = 0.86 (0.73 to 1.02) | — | 13 | |||
| Illicit substances: overall analysis | OR = 0.78 (0.66 to 0.93) | — | 10 | |||
| Klingbeil et al27 (2017) | Mindfulness-based interventions compared with control | Average treatment effect in pretest-posttest | g = 0.305 (0.223 to 0.387) | Low | 1075 | 30 |
| Average treatment effect in controlled studies | g = 0.322 (0.242 to 0.402) | 4811 | 48 | |||
| Average treatment effect in pretest-posttest: follow-up | g = 0.462 (0.202 to 0.723) | 532 | 12 | |||
| Average treatment effect in controlled studies: follow-up | g = 0.402 (0.220 to 0.584) | 1501 | 12 | |||
| Mindfulness | pretest-posttest: g = 0.174 (0.032 to 0.316). Controlled: g = 0.510 (0.082 to 0.939) | pretest-posttest: 213; Controlled: 1108 | pretest-posttest: 12; Controlled: 15 | |||
| Attention | pretest-posttest: g = 0.324 (0.232 to 0.415). Controlled: g = 0.291 (0.124 to 0.456) | 557; 1243 | 8; 10 | |||
| Meta-cognition and cognitive flexibility | n/a; Controlled: g = 0.404 (0.011 to 0.798) | 33; 806 | 3; 9 | |||
| Emotional or behavioral regulation | pretest-posttest: g = 0.297 (0.213 to 0.381); Controlled: g = 0.322 (0.180 to 0.464) | 976; 1404 | 17; 16 | |||
| Academic achievement and school functioning | pretest-posttest: g = 0.321 (0.132 to 0.511); Controlled: g = 0.393 (−0.153 to 0.939) | 516; 434 | 7; 5 | |||
| Externalizing problems | pretest-posttest: g = 0.145 (0.032 to 0.258); Controlled: g = 0.296 (0.087 to 0.505) | 388; 871 | 10; 9 | |||
| Internalizing problems | Pretest-posttest: g = 0.262 (0.161 to 0.362); Controlled: g = 0.392 (0.267 to 0.391) | 527; 2941 | 17; 29 | |||
| Negative emotions and subjective distress | Pretest-posttest: g = 0.323 (0.125 to 0.521); Controlled: g = 0.254 (0.123 to 0.385) | 408; 2342 | 12; 21 | |||
| Positive emotions and self-appraisal | Pretest-posttest: g = 0.342 (0.165 to 0.519); Controlled: g = 0.280 (0.092 to 0.468) | 487; 2321 | 15; 22 | |||
| Physical health | Pretest-posttest: g = 0.492 (−0.130 to 1.113); Controlled: g = 0.282 (0.173 to 0.391) | 308; 994 | 7; 14 | |||
| Social competence and prosocial behaviors | Pretest-posttest: g = 0.214 (0.092 to 0.335); Controlled: g = 0.368 (0.165 to 0.570) | 905; 1105 | 12; 12 | |||
| Melendez-Torres et al30 (2018) | Integrated academic and health education compared with treatment as usual | Alcohol (school grade7–9 y; aged 11–14 y) | SMD = −0.11 (−0.23 to 0.004) | High | 23 | 5 |
| Smoking (school grade 7–9; aged 11–14 y) | SMD = −0.05 (−0.12 to 0.02) | 13 | 5 | |||
| Illicit drug use: marijuana (school grade 7–9; aged 11–14 y) | SMD = −0.10 (−0.16 to −0.04) | 10 | 5 | |||
| Illicit drug use (school grade 7–9; aged 11–14 y) | SMD = −0.07 (−0.14 to −0.01) | 14 | 5 | |||
| All drug use outcomes (school grade 7–9; aged 11-14 y) | SMD = −0.09 (−0.17 to −0.01) | 55 | 5 | |||
| Alcohol (school grade 10–11; aged 14–16 y) | SMD = −0.01 (−0.09 to 0.06) | 15 | 3 | |||
| Smoking (school grade 10–11; aged 14–16 y) | SMD = −0.08 (−0.15 to −0.01) | 9 | 3 | |||
| Illicit drug use: marijuana (school grade 10–11; aged 14–16 y) | SMD = −0.10 (−0.17 to −0.03) | 10 | 3 | |||
| All drug use outcomes (school grade 10–11; aged 14–16 y) | SMD = −0.06 (−0.09 to −0.02) | 34 | 3 | |||
| Maynard et al29 (2017) | Mindfulness‐based interventions compared with wait list control, no treatment, treatment-as-usual or alternative treatment groups | Cognitive outcomes | ES = 0.25 (0.06 to 0.43) | High | — | 10 |
| Academic outcomes | ES = 0.27 (-0.04 to 0.58) | — | 5 | |||
| Behavioral outcomes | ES = 0.14 (-0.02 to 0.30) | — | 13 | |||
| Socioemotional outcomes | ES = 0.22 (0.14 to 0.30) | — | 28 | |||
| Walsh et al56 (2018) | School-based education programs for the prevention of child sexual abuse compared with wait-listed control | Protective behaviors | OR = 5.71 (1.98 to 16.51) | High | 102 | 2 |
| Questionnaire-based knowledge | SMD = 0.61 (0.45 to 0.78) | 4657 | 18 | |||
| Vignette-based knowledge | SMD = 0.45 (0.24 to 0.65) | 1688 | 11 | |||
| Retention of knowledge over time | SMD = 0.78 (0.38 to 1.17) | 956 | 4 | |||
| Disclosure of previous or current sexual abuse | OR = 3.56 (1.13 to 11.24) | 1788 | 3 | |||
| Wilson et al34 (2006) | No-treatment or wait-list control. | Aggressive and disruptive behavior | Random effects mean 0.21 (P < .01) | Critically low | — | 73 |
| Wilson et al35 (2006) | No-treatment or wait-list Control. | Aggressive and disruptive behavior | Random effects mean 0.26 (P < .01) | Critically low | — | 47 |
| Community-based interventions | ||||||
| Dowdall et al21 (2020) | Shared picture book readings interventions compared with Passive and active controls | Expressive language outcomes | d = 0.41 (0.20 to 0.61) | Low | 1664 | 16 |
| Receptive language outcomes | d = 0.26 (0.12 to 0.40) | 1871 | 16 | |||
| Caregiving competence outcomes | d = 1.01 (0.40 to 1.63) | 481 | 6 | |||
| Jackson et al26 (2016) | Exercise programs compared with normal activity or waitlist control | Inhibitory control (RCTs) | d = 0.2 (0.03 to 0.37) | Critically low | 770 | 8 |
| Liu et al28 (2015) | Physical activity (PA) interventions compared with nonphysical activity control or comparison group | General self-outcomes (RCTs) | g = 0.29 (0.14 to 0.45) | Critically low | — | 18 |
| Self-concept (RCTs) | g = 0.49 (0.10 to 0.88) | — | 18 | |||
| Self-worth (RCTs) | g = 0.31 (0.13 to 0.49) | — | 18 | |||
| General self-outcomes (non-RCTs) | g = 0.33 (−0.35 to 1.01) | — | 6 | |||
| Macbeth et al46 (2015) | Mellow parenting interventions versus wait-list control or treatment as usual | Effect of mellow parenting on child outcomes: Dersimonian-Laird random effects model | SMD = −0.40 (−0.77 to −0.02) | Critically low | — | 3 |
| Effect of mellow parenting on child outcomes: fixed effects model | SMD = −0.40 (−0.77 to −0.02) | — | 3 | |||
| McGinn et al60 (2020) | Traditional care case processing compared with no intervention or alternative treatment | Reunification of children with families or maintenance of in-home care | OR = 1.69 (1.03 to 2.78) | High | 86 305 | 10 |
| Continued maltreatment, effects from nonrandomized studies | OR = 0.73 (0.48 to 1.11) | 1779 | 5 | |||
| Continued maltreatment, effects from RCTs | OR = 1.29 (0.85 to 1.98) | 1158 | 4 | |||
| Kinship placements | OR = 1.29 (0.94 to 1.76) | 85 537 | 5 | |||
| Leijten et al45 (2016) | Transported or homegrown interventions compared with no-treatment, waitlist, minimal intervention, or care as usual. | Effectiveness between transported or homegrown interventions to reduce disruptive child behavior | MD = 0.10 (−0.08 to 0.29) | Moderate | 374 | 129 |
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: Incredible years program | MD = 0.12 (−0.15 to 0.40) | 82 | 31 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: triple P program | MD = 0.27 (−0.02 to 0.56) | 113 | 39 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: PCIT program | MD = 0.40 (−0.26 to 1.05) | 51 | 11 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: Parent Management Training (Oregon) program | MD = 0.39 (−0.93 to 1.70) | 24 | 9 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: US and Canada | MD = 0.19 (−0.60 to 0.98) | 155 | 51 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: Australia | MD = −0.19 (−0.63 to 0.26) | 101 | 30 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: non-English speaking European countries (Continental Europe and Iceland) | MD = 0.38 (−0.11 to 0.86) | 60 | 22 | |||
| Effectiveness between transported or homegrown interventions to reduce disruptive child behavior: English-speaking European countries (UK/Ireland) | MD = −0.07 (−0.42 to 0.28) | 38 | 16 | |||
| Sanders et al48 (2014) | Multilevel triple P-positive pParenting program system compared with nonactive control group (ie, a waitlist control group or usual care) | Child SEB outcomes: all levels combined | SMD = 0.525 (0.358 to 0.692) | Low | — | 56 |
| Parenting practices: all levels combined | SMD = 0.498 (0.362 to 0.634) | — | 48 | |||
| Parenting satisfaction and efficacy: all levels combined | SMD = 0.551 (0.372 to 0.730) | — | 41 | |||
| Parental adjustment: all levels combined | SMD = 0.481 (0.321 to 0.641) | — | 45 | |||
| Parental relationship: all levels combined | SMD = 0.230 (0.136 to 0.325) | — | 37 | |||
| Child observation: all levels combined | SMD = 0.400 (0.070 to 0.730) | — | 13 | |||
| Parent observation: all levels combined | SMD = 0.249 (0.031 to 0.467) | — | 11 | |||
| Valentine et al50 (2019) | Effects of families and schools together (FAST), relative to waiting list, usual or alternative services, or no intervention. | Child school performance | SMD = −0.02 (−0.11 to 0.08) | High | 6276 | 4 |
| Child internalizing behavior: long-term follow-up, parent report | SMD = 0.03 (−0.11 to 0.17) | 908 | 4 | |||
| Child internalizing behavior: long-term follow-up, teacher report | SMD = −0.06 (−0.19 to 0.07) | 912 | 4 | |||
| Child externalizing behavior: long-term follow-up, parent report | SMD = −0.19 (−0.32 to −0.05) | 754 | 4 | |||
| Child externalizing behavior: long-term follow-up, teacher report | SMD = −0.10 (−0.24 to 0.04) | 912 | 4 | |||
| Child school attendance | SMD = 0.03 (−0.10 to 0.16) | 1430 | 2 | |||
| Parental engagement with education | SMD = 0.03 (−0.07 to 0.12) | 1232 | 2 | |||
| Family relationships | SMD = 0.08 (−0.03 to 0.19) | 2569 | 4 | |||
| Vlahovicova et al51 (2017) | Manualized interventions against treatment as usual | Risk of re-abuse | RD = −0.11 (-0.22 to -0.004) | Low | — | 4 |
| Risk of re-abuse: sensitivity analysis | RR = 0.76 (0.54 to 1.07) | — | 4 | |||
| Zief et al37 (2006) | After-school programs compared with no intervention | Self-care | SMD = 0.503 (0.097 to 0.910) | Critically low | 101 | 3 |
| Athletic activities | RD = 0.073 (0.02 to 0.144) | 661 | 2 | |||
| Art/music/drama dance activities | RD = 0.083 (0.01 to -0.153) | 661 | 2 | |||
| School attendance | SMD = −0.001 (−0.125 to 0.122) | 1072 | 2 | |||
| Television viewing | SMD = −0.066 (−0.1224 to 0.091) | 661 | 2 | |||
| Reading scores | SMD = 0.028 (−0.101 to 0.157) | 983 | 2 | |||
| Grade point average (GPA) | SMD = 0.083 (−0.032 to 0.199) | 1004 | 5 | |||
| Mixed setting interventions | ||||||
| Burkey et al40 (2018) | No treatment, wait list controls, treatment as usual, groups or inactive controls. | Child behavior problems (all interventions) | SMD = −0.38 (−0.51 to −0.24) | High | 4441 | 24 |
| Child behavior problems, child-focused interventions | SMD = −0.39 (−0.62 to −0.16) | — | 13 | |||
| Child behavior problems, parent-focused interventions | SMD = −0.43 (−0.66 to −0.20) | — | 8 | |||
| Child behavior problems, classroom based | SMD = −0.49 (−0.71 to −0.28) | — | 2 | |||
| Child behavior problems, multicomponent interventions | SMD = −0.19 (−0.50 to 0.12) | — | 3 | |||
| Child behavior problems, children aged 0–5 | SMD = −0.36 (−0.66 to −0.06) | — | 4 | |||
| Child behavior problems, children aged 6–12 | SMD = −0.42 (−0.65 to −0.19) | — | 11 | |||
| Child behavior problems, children aged 13–18 | SMD = −0.43 (−0.76 to −0.10) | — | 6 | |||
| Child behavior problems, prevention focused | SMD = −0.25 (−0.41 to −0.09) | — | 11 | |||
| Child behavior problems, treatment focused | SMD = −0.56 (−0.51 to −0.24) | — | 11 | |||
| Child behavior problems, specialist delivered | SMD = −0.58 (−0.75 to −0.43) | — | 12 | |||
| Child behavior problems, nonspecialist delivered | SMD = −0.15 (−0.27 to −0.03) | — | 11 | |||
| Durlak and Taylor22 (2007) | Positive youth development (PYD) programs compared with control groups | Systemic change efforts at psychosocial environment: classroom | Post: ES = 0.47; Follow-up: ES = 0.42 | Critically low | — | Post: 10; Follow-up: 1 |
| Systemic change efforts at psychosocial environment: school-wide | Post: ES = 0.74; Follow-up: — | — | Post: 8; Follow-up: 0 | |||
| Systemic change efforts at classroom level | Post: ES = 0.78; Follow-up::— | — | Post: 4; Follow-up: 0 | |||
| Parenting practices | Post: ES = 0.41; Follow-up: ES = 0.49 | — | Post: 33; Follow-up: 17 | |||
| Family environment | Post: ES = 0.34; Follow-up: ES = 0.94 | — | Post: 25; Follow-up: 3 | |||
| Bonding to community adults | Post: ES = −0.26; Follow-up: — | — | Post: 2: Follow-up: 0 | |||
| Family-school relationships | Post: ES = 0.49; Follow-up: ES = −0.09 | — | Post: 6: Follow-up: 1 | |||
| Morton and Montgomery31 (2011) | Youth empowerment programs (YEPs) compared with no intervention, waitlist, and alternative intervention | General self-efficacy | SMD = 0.19 (-0.12 to 0.49) | High | 167 | 2 |
| Ulferts et al33 (2019) | Initiatives aimed at enhancing the quality of early childhood education | Global process quality on children's outcomes (language/ literacy and mathematics) | ES = 0.11 (0.04 to 0.18) | Low | — | 13 |
| Domain-specific process quality on children's outcomes (language/literacy and mathematics) | ES = 0.10 (0.05 to 0.15) | — | 11 |
—, not reported.
The key effects of interventions on depression, anxiety, and externalizing symptoms/behavior have been summarized in Table 7, by delivery strategy.
Key Effects of Interventions on Depression, Anxiety, and Externalizing Symptoms/Behaviors
| Delivery platform | Depression | Anxiety | Externalizing symptoms/behaviors |
|---|---|---|---|
| School-Based | • Universal CBT interventionsa112 | • Universal CBT interventionsa112 | • Universal resilience-focused interventionsa118 |
| o SMD = −0.13 (−0.44 to 0.17) | o SMD= −0.07 (−0.23 to 0.05)b | o SMD = −0.18 (−0.34 to −0.01) | |
| • Universal resilience-focused interventionsa118 | • Universal mindfulness and relaxation-based interventionsa112 | • Teacher delivered psychosocial interventionsc125 | |
| o SMD = −0.08 (−0.14 to −0.01) | o SMD = −0.65 (−1.14 to −0.19) | o G = 0.015 (−0.037 to 0.066) | |
| • Multicomponent positive psychology interventionsdd158 | • Universal resilience-focused interventionsa118 | • Universal school-based interventionsc149 | |
| o ES = 0.28 (0.13 to 0.43) | o SMD = −0.14 (−0.28 to 0.00) | o ES = 0.50 (0.35 to 0.63) | |
| • School based Penn Resiliency Programd17 | • Multicomponent positive psychology interventionsd158 | • Self-regulation interventionsd160 | |
| o MD = −0.23 (−1.09 to 0.62) | o ES = 0.14 (0.04 to 0.24) | o ES = 0.01 (−0.38 to0.39) | |
| • Universal screening programsd106 | • School based Penn Resiliency Programd17 | • Group-based mindfulness interventionsa27 | |
| o detection of depression (n = 3) | o SMD = 0.13 (0.00 to 0.26) | o G = 0.145 (0.032 to 0.258) | |
| • Mental health promotion interventionse108 | • School based pull-out programs on aggressive behavior | ||
| o depression (n = 14) | o Random effects mean (REM) = 0.21, P < .01d34 | ||
| • Early intervention programs for depressionc113 | o REM = 0.26, P < .01a35 | ||
| o depressive symptoms (n = 42) | • Universal screening programsd | ||
| • School based, universal prevention interventionsa | o ↑ detection of anxiety (n = 3)106 | ||
| o depression scores in the intervention group compared with control (n = 24) | • Mental health promotion interventionse108 | ||
| o ↓ depression at postprogram and long-term follow-up periods, but not at short-term follow-up (n = 14)132 | o reduced anxiety (n = 14) | ||
| • Interventions describing modules of mental health programse112 | • School based, universal prevention interventionsa | ||
| o depression (n = 11) | o ↓ anxiety scores in the intervention group compared with control (n = 24)116 | ||
| • Peer outreach, counseling, and education interventionse147 | • Interventions describing modules of mental health programse126 | ||
| o depression, improvements were found for depressive symptoms (n = 7) | o anxiety (n = 11) | ||
| • Physical activity interventionsc156 | |||
| o depression, improvements were found for measures of depression symptoms (n = 5) | |||
| Community-Based | • Evidence based youth mental and behavioral health disorder preventive interventionsd146 | • Parenting interventions based on behavioral/social learning theorya | |
| o depression, significant improvements were found for depression compared with comparison conditions (n = 5) | o MD = 0.10 (−0.08 to 0.29)45 | ||
| • Developmental group therapyc63 | • Families and schools together (FAST) programd | ||
| o MD = −0.93 (−4.03 to 2.17) ¶ | o SMD = −0.19 (−0.32to −0.05)50 | ||
| Digital-Based | • Universal psychosocial interventionsa154 | • Universal psychosocial interventionsa154 | |
| o ES = −0.054 (−0.181 to 0.074) | o ES = −0.054 (−0.181 to 0.074) | ||
| • Computerized CBTd115 | • Computerized CBTd115 | ||
| o depressive symptoms (n = 7) | o anxiety symptoms (n = 7) | ||
| • Digital health interventionsc131 | • Digital health interventionsc131 | ||
| o depression outcomes (n = 12) | o anxiety (n = 12) | ||
| Mixed Settings | • Mindfulness based interventionsc119 | • Prevention interventions for any DSM-5 anxiety disorderd110 | • Physical activity interventionsc |
| o d = 0.27 (0.06 to 0.49) | o SMD = 0.22 (0.14 to 0.29) | o MD = 0.32 (0.023 to 0.616)156 | |
| • CBT & interpersonal therapy interventionsa129 | • Mindfulness based interventionsc119 | • Psychosocial interventionsa | |
| o SMD = −0.21 (−0.27 to -0.15) | o d = 0.16 (0.04 to 0.27) | o SMD = −0.38 (−0.51 to −0.24)40 | |
| • Cognitive-behavioral or behavioral interventionsa159 | |||
| o SMD = −0.81 (−1.00 to −0.63) |
| Delivery platform | Depression | Anxiety | Externalizing symptoms/behaviors |
|---|---|---|---|
| School-Based | • Universal CBT interventionsa112 | • Universal CBT interventionsa112 | • Universal resilience-focused interventionsa118 |
| o SMD = −0.13 (−0.44 to 0.17) | o SMD= −0.07 (−0.23 to 0.05)b | o SMD = −0.18 (−0.34 to −0.01) | |
| • Universal resilience-focused interventionsa118 | • Universal mindfulness and relaxation-based interventionsa112 | • Teacher delivered psychosocial interventionsc125 | |
| o SMD = −0.08 (−0.14 to −0.01) | o SMD = −0.65 (−1.14 to −0.19) | o G = 0.015 (−0.037 to 0.066) | |
| • Multicomponent positive psychology interventionsdd158 | • Universal resilience-focused interventionsa118 | • Universal school-based interventionsc149 | |
| o ES = 0.28 (0.13 to 0.43) | o SMD = −0.14 (−0.28 to 0.00) | o ES = 0.50 (0.35 to 0.63) | |
| • School based Penn Resiliency Programd17 | • Multicomponent positive psychology interventionsd158 | • Self-regulation interventionsd160 | |
| o MD = −0.23 (−1.09 to 0.62) | o ES = 0.14 (0.04 to 0.24) | o ES = 0.01 (−0.38 to0.39) | |
| • Universal screening programsd106 | • School based Penn Resiliency Programd17 | • Group-based mindfulness interventionsa27 | |
| o detection of depression (n = 3) | o SMD = 0.13 (0.00 to 0.26) | o G = 0.145 (0.032 to 0.258) | |
| • Mental health promotion interventionse108 | • School based pull-out programs on aggressive behavior | ||
| o depression (n = 14) | o Random effects mean (REM) = 0.21, P < .01d34 | ||
| • Early intervention programs for depressionc113 | o REM = 0.26, P < .01a35 | ||
| o depressive symptoms (n = 42) | • Universal screening programsd | ||
| • School based, universal prevention interventionsa | o ↑ detection of anxiety (n = 3)106 | ||
| o depression scores in the intervention group compared with control (n = 24) | • Mental health promotion interventionse108 | ||
| o ↓ depression at postprogram and long-term follow-up periods, but not at short-term follow-up (n = 14)132 | o reduced anxiety (n = 14) | ||
| • Interventions describing modules of mental health programse112 | • School based, universal prevention interventionsa | ||
| o depression (n = 11) | o ↓ anxiety scores in the intervention group compared with control (n = 24)116 | ||
| • Peer outreach, counseling, and education interventionse147 | • Interventions describing modules of mental health programse126 | ||
| o depression, improvements were found for depressive symptoms (n = 7) | o anxiety (n = 11) | ||
| • Physical activity interventionsc156 | |||
| o depression, improvements were found for measures of depression symptoms (n = 5) | |||
| Community-Based | • Evidence based youth mental and behavioral health disorder preventive interventionsd146 | • Parenting interventions based on behavioral/social learning theorya | |
| o depression, significant improvements were found for depression compared with comparison conditions (n = 5) | o MD = 0.10 (−0.08 to 0.29)45 | ||
| • Developmental group therapyc63 | • Families and schools together (FAST) programd | ||
| o MD = −0.93 (−4.03 to 2.17) ¶ | o SMD = −0.19 (−0.32to −0.05)50 | ||
| Digital-Based | • Universal psychosocial interventionsa154 | • Universal psychosocial interventionsa154 | |
| o ES = −0.054 (−0.181 to 0.074) | o ES = −0.054 (−0.181 to 0.074) | ||
| • Computerized CBTd115 | • Computerized CBTd115 | ||
| o depressive symptoms (n = 7) | o anxiety symptoms (n = 7) | ||
| • Digital health interventionsc131 | • Digital health interventionsc131 | ||
| o depression outcomes (n = 12) | o anxiety (n = 12) | ||
| Mixed Settings | • Mindfulness based interventionsc119 | • Prevention interventions for any DSM-5 anxiety disorderd110 | • Physical activity interventionsc |
| o d = 0.27 (0.06 to 0.49) | o SMD = 0.22 (0.14 to 0.29) | o MD = 0.32 (0.023 to 0.616)156 | |
| • CBT & interpersonal therapy interventionsa129 | • Mindfulness based interventionsc119 | • Psychosocial interventionsa | |
| o SMD = −0.21 (−0.27 to -0.15) | o d = 0.16 (0.04 to 0.27) | o SMD = −0.38 (−0.51 to −0.24)40 | |
| • Cognitive-behavioral or behavioral interventionsa159 | |||
| o SMD = −0.81 (−1.00 to −0.63) |
Bolded indicates significant; italicized indicates t= nonsignificant.
HIC and LMIC.
Significance not reported.
Country context not reported.
HIC context.
LMIC.
Universal Prevention Interventions
Sixty-one reviews synthesized data or narratively reported on universal prevention interventions. Overall, we found evidence to support the effectiveness of psychosocial and exercise interventions to prevent negative mental health outcomes118,120,125,129,150,154, 155,159 and multicomponent positive psychology (PP), social–emotional learning (SEL), mindfulness, and mentoring-based interventions to promote well-being119,145,157,158 in school-aged children.
School-Based Interventions
Thirty reviews synthesized data on a range of school-based interventions. The greatest effects on mental health were found for psychosocial, internal resilience, SEL, and multicomponent, PP-based interventions. For example, resilience-focused interventions for children aged 5 to 18 years were effective relative to control for reducing depressive symptoms (standardized mean difference [SMD] = −0.08, 95% confidence interval [CI]: −0.14 to −0.01), internalizing (SMD = −0.21, 95% CI: −0.36 to −0.06) and externalizing problems (SMD = −0.18, 95% CI: −0.34 to −0.01), and general psychological distress (SMD = −0.11, 95% CI: −0.21 to −0.01).118 Similarly, psychosocial interventions that were delivered by teachers to school-aged children were effective in reducing internalizing outcomes (d = 0.133, 95% CI: 0.002 to 0.263).125
Community-Based Interventions
Six reviews synthesized data on community-based interventions; however, the diversity of intervention components and outcomes assessed prevented the ability to synthesize results and make any definitive statements about effectiveness. In all cases, the findings were encouraging but required more research to make firm conclusions.118,131,137,140,145,147 Of note, cognitive behavioral therapy (CBT) was found to be an encouraging intervention at the parental/family context to prevent adolescent depression.118 Childhood pet ownership had a wide range of emotional health benefits, in particular for self-esteem and loneliness.144
Mixed-Setting Interventions
Twenty reviews synthesized data on a broad range of interventions delivered in mixed settings. The greatest effects on mental health were found for exercise, CBT, mentoring, single-session therapy, and mindfulness-based interventions. For example, exercise-based interventions for school-aged children were effective in improving self-esteem (SMD = 0.51, 95% CI: 0.15 to 0.88),121 reducing externalizing problems (d = 0.320, 95% CI: 0.023 to 0.616), internalizing problems (d = 0.316, 95% CI: 0.073 to 0.558), improving self-concept (d = 0.297, 95% CI: 0.127 to 0.468), and academic achievement (d = 0.367, 95% CI: 0.038 to 0.697).156 CBT interventions reduced anxiety for children aged 3 to 8 years (SMD = −1.34, 95% CI: −1.59 to −1.09)160 and self-rated depression in children aged 8 to 24 years (SMD −0.11, 95% CI: −0.17 to −0.05).130
Digital-Based Interventions
Five studies synthesized data on digital-based interventions, and all found encouraging results for the effectiveness of these interventions to achieve positive mental health outcomes. For example, results from online prevention interventions indicated a significant positive effect of computerized CBT on adolescents’ anxiety and depressive symptoms.115 Furthermore, universal psychosocial interventions delivered digitally were effective in improving mental health, in particular, interpersonal skills, emotional regulation, and alcohol and drug education, for children aged 10 to 19 years (effect size [ES] = 0.197, 95% CI: 0.016 to 0.379).154
Bullying Prevention
School-Based Interventions
Fourteen reviews examined school-based interventions. Evidence suggests that antibullying interventions are effective for a range of bullying-related outcomes. For example, interventions significantly reduced bullying (ES = −0.150; 95% CI: −0.191 to −0.109) and improved mental health (ES = −0.205; 95% CI: −0.277 to −0.133) at study endpoint.67 However, more research is needed to determine whether anti-cyberbullying interventions delivered in schools reduce cyberbullying behaviors.68,71,73 Reviews disagreed on which age group benefited most. Several reviews reported that bullying and victimization steadily increase with age, suggesting that intervening in elementary schools through preventive interventions may significantly reduce bullying-related outcomes.11,78
Digital Interventions
One review also examined information and communication technology-mediated interventions delivered in schools, finding that more research is required.74
Substance Use Prevention
Twenty-two reviews synthesized data or narratively reported on substance use prevention interventions. Most evidence supported the use of family- and school-based interventions to prevent substance use initiation in school-aged children.
Community-Based Interventions
Eight reviews examined community-based interventions and concluded that more research is needed. However, promising evidence was found for Youth Participatory Action Research to increase community awareness and multicomponent interventions to prevent smoking.85,104 Four reviews found that family-based interventions were effective for preventing substance abuse initiation.83,91,93,102 For example, interventions prevented smoking initiation (relative risk [RR] = 0.76; 95% CI: 0.68 to 0.84).102 One review concluded that family-based interventions were more effective when delivered in early adolescence.93
School-Based Interventions
Five reviews examined school-based interventions. Combining social competence and influence components had small but consistent positive effects on drug use prevention (odds ratio [OR] = 0.50; 95% CI: 0.28 to 0.87).90,101 Health education and social learning components in combination with oral, written, and audiovisual support material improved attitudes toward drugs (d = 0.44; 95% CI: 0.33 to 0.54).88 Skills-based components improved self-efficacy compared with affective components, but had the same effect as knowledge-based components on drug knowledge. Affective components improved decision-making and drug knowledge compared with knowledge-based components.89
Digital-Based Interventions
Three reviews found little, though promising, evidence to support digital-based interventions. One review reported reduced cannabis use (OR = 0.33; 95% CI: 0.13 to 0.54) persisting to 12 months.84
Mixed-Setting Interventions
Other Interventions
Self-Harm Prevention
Two reviews synthesized data or narratively reported on self-harm prevention interventions. Self-harm is more common in adolescents (12–18 years) than children.63 However, interventions targeted at school-aged children may help to prevent adolescent self-harm.
Mixed-Setting Interventions
One review examined interventions that were designed to improve the ability of gatekeepers to prevent suicide-related crises in young people by recognizing risks and responding appropriately.65 Gatekeeper training is associated with significant moderate-to-large improvements in suicide literacy outcomes (eg, knowledge of suicide, confidence to intervene, and attitudes toward suicide).65 The second review examined primary prevention interventions, such as knowledge promotion, reducing access to means, local media reporting, local suicide plans, and screening for suicide risk.63 The review assessed multiple settings and found that school-based interventions prevent suicidal ideation and attempts short term, and possibly suicide attempts long term.63
Nurturing Care (Responsive Caregiving, Safety and Security, and Early Learning to Support Positive Development)
Thirty-eight reviews synthesized data or narratively reported on nurturing care interventions. Twenty-nine studies were identified which reported mixed effects of interventions on children's social, emotional, and behavioral outcomes in a wide range of delivery strategies. Nine out of 29 studies reported safety and security interventions, which focused on prevention of child and adolescent maltreatment, and promotion of mental and physical well-being in abused or neglected children. Early learning and positive development interventions were reported in 16 reviews, which focused on academics, positive development, and positive connections to improve developmental outcomes and a positive transition into adulthood. School- and community-based interventions had positive effects on school-aged children’s mental health and developmental outcomes.
Community-Based Interventions
Eighteen reviews reported family-based interventions. Improved parenting responsiveness resulted in better health and development in younger children, especially in poor settings.41 The Multilevel Triple P-Positive Parenting Program system showed improvements in children's social, emotional, and behavioral outcomes (SMD = 0.525; 95% CI: 0.358 to 0.692); parenting practices (SMD = 0.498; 95% CI: 0.362 to 0.634); parenting satisfaction and efficacy (SMD = 0.551; 95% CI: 0.372 to 0.730); parental adjustment (SMD = 0.481; 95% CI: 0.321 to 0.641); and parental relationship (SMD = 0.230; 95% CI: 0.136 to 0.325).48 Physical activity interventions showed improvements in motor skills and cognitive development in children aged <6 years.36 After-school programs promoted positive behaviors, including self-care (SMD = 0.503; 95% CI: 0.097 to 0.910) and reduced negative behaviors.37 Physical activity interventions improved self-concept (hedges g = 0.49; 95% CI: 0.10 to 0.88), self-worth (g = 0.31; 95% CI: 0.13 to 0.49), and inhibitory control (d = 0.2; 95% CI: 0.03 to 0.37) in children aged 0 to 19 years.26,28
School-Based Interventions
School-based programs for sexual abuse prevention improved protective behaviors (OR = 5.71; 95% CI: 1.98 to 16.51) and knowledge (SMD = 0.61; 95% CI: 0.45 to 0.78).56 School-based universal SEL interventions significantly improved social and emotional distress (mean effect [ME] = 0.24; 95% CI: 0.14 to 0.35); attitudes (ME = 0.23; 95% CI: 0.16 to 0.30); positive behavior (ME = 0.24; 95% CI: 0.16 to 0.32); conduct problems (ME = 0.22; 95% CI: 0.16 to 0.29); and academic performance (ME = 0.27; 95% CI: 0.15 to 0.39) in children aged 5 to 18 years.24 After-school programs demonstrated a significant increase in overall effect (SMD = 0.22; 95% CI: 0.16 to 0.29); self-perceptions (SMD = 0.34; 95% CI: 0.23 to 0.46); positive behaviors (SMD = 0.19; 95% CI: 0.10 to 0.29); levels of academic achievement (SMD = 0.17; 95% CI:0.06 to 0.29); and reductions in problem behaviors (SMD = 0.19 95% CI: 0.10 to 0.27).23 School‐based universal programs34 and pull-out programs35 targeting aggressive behavior in children and adolescents reported positive effects with short, intensive interventions compared with extended yearlong programs. Universal school-based resilience interventions25 and health education curriculums30 showed reduction in illicit substance use.
Mixed-Setting Interventions
Fifteen reviews reported on mixed settings including schools, day care, communities, and households. Community-based parenting interventions showed positive effects on measures of good parenting, with positive flow-on effects to some aspects of challenging adolescent behaviors including parent–child communication, smoking reduction, and parental reactions.43,47,49 Individual and group-based, child-focused, and behavioral parenting interventions showed reduction in disruptive behavior problems (SMD = −0.38; 95% CI: −0.51 to −0.24) among school-aged children in LMIC.40 Use of activity- and occupation-based interventions, including video game interventions, showed positive effects on mental, behavioral, and physical health in children and youth.19,32 Positive youth development (PYD) programs showed positive effects on behaviors, including substance use and risky sexual activity, and/or more distal developmental outcomes.20,22
Evidence from LMIC
Forty-nine reviews covered populations in both HIC and LMIC; however, most studies captured by these reviews were in HIC and, thus, makes generalizability of findings to LMIC difficult. We found 8 reviews which encompassed LMIC studies only. Interventions considered were PYD programs, school tobacco policies, school-based antibullying or mental health promotion interventions, and peer-facilitated, community-based interventions for mental health disorders and substance use. There was promising evidence for PYD programs, and findings from multiple reviews suggest that interventions that promote mental health in young people can be implemented effectively in LMIC school and community settings. However, overwhelmingly, all reviews point to the dearth of evidence that exists for interventions in the LMIC context, where some reviews only captured 1 study for inclusion.
Discussion
Summary of the Evidence
We summarize findings from 162 reviews, where 78 studies meta-analyzed data. Findings for targeted interventions (n = 29) have been summarized in Supplemental Information. Of the included studies, 39.7% had critically low-quality ratings, 26.9% were low, 25.6% were high, and 7.8% were moderate. Evidence from universal prevention interventions showed overall positive effects on youth development and prevention of poor mental health outcomes, where most interventions were school-based. Antibullying interventions reduced bullying and victimization where the majority were school-based, with limited evidence for digital interventions. Community- and school-based interventions were found to prevent substance abuse; however, the evidence for digital interventions was limited. Only 2 reviews examined interventions for the prevention of self-harm and found evidence for gatekeeper training and school-based interventions. School- and community-based interventions had positive effects on children and adolescents’ early learning and positive development. Lastly, there was mixed evidence overall for the effectiveness of home-based interventions to enhance responsive caregiving and child safety.
Evidence for key impacts on the greatest causes of mental illness burden in the school-age group suggest that school-based multicomponent PP158 and universal prevention interventions116,132 were effective in reducing depression and anxiety in predominantly HIC. Furthermore, universal, resilience-focused interventions118 were found to be effective in reducing depression and externalizing symptoms in HIC. School-based mental health promotion interventions108 and interventions describing modules of mental health126 were effective in reducing depression and anxiety in LMIC. Furthermore, peer outreach, counseling, and education interventions were effective in reducing depression in LMIC.147
In community-based settings, evidence-based youth mental and behavioral health disorder prevention interventions146 were effective in reducing depression, and the FAST program50 was effective in reducing externalizing symptoms in HIC. Computerized CBT and digital health interventions were effective in reducing depression and anxiety115,131 in HIC. Mindfulness-based interventions119 delivered through mixed settings were effective in reducing depression and anxiety in LMIC and HIC.
When discussing evidence specifically from LMIC, school-based PYD programs provided the strongest evidence for positive development (eg, reduced risky sexual behavior, substance use, and interpersonal violence) and mental health outcomes (eg, increased emotional stability and self-esteem) in children; however, more research is urgently required to identify effective preventive interventions for children exposed to a range of adversities and types of mental disorders in the LMIC context.
Limitations
Our overview of reviews was primarily limited by our inability to synthesize and interpret research findings from the included systematic reviews because of inadequate or varied reporting of pooled data from the primary studies.
Implications for Research
Despite the breadth of reviews found, there remain several significant gaps in the literature. Firstly, there were a limited number of studies focused on low education, low-income, and conflict-struck settings. Some outcome measures (eg, child abuse and neglect) may be more prevalent and underreported in these contexts.178 Furthermore, there were a paucity of studies conducted across all delivery strategies. In particular, digital interventions were understudied compared with school-based interventions which were most often discussed across all domains. Given there are many school-aged children who never attend school, there is a need for interventions conducted outside of the school setting to improve coverage to all school-aged children. Only 2 studies were captured in the self-harm domain, highlighting the need for more research focused on self-harm prevention. Lastly, many of the included reviews captured children into adolescence (eg, from 0 to 18 years old). There is need for more research targeting younger children (<10 years) given that intervening at an early age is critical to prevent mental illness in adolescence and adulthood.
Implications for Policy
Given the identified gaps in the literature and methodological and clinical heterogeneity of included reviews, we are limited in our ability to make definitive statements about effectiveness of interventions, rendering policy interpretation and implementation challenging. However, there were some common recommendations that emerged from the literature. First, a common recommendation across several domains is to implement multicomponent and multilevel interventions.41,48,78,90,112,158 Several reviews suggest adding school-based components to comprehensive antibullying interventions78,90 and integrating home-level responsive parenting interventions into child survival programs.41,48 Furthermore, groups involved in intervention implementation, such as school health personnel, should be involved in intervention development and supported during implementation.78 Several reviews also recommend training these groups for effective intervention programming.41,65,78 Finally, given most universal prevention interventions were delivered in schools, there is evidence for the use of school-based universal prevention programs.
Key Messages
Positive youth development and after-school programs had encouraging effects on academic achievements and problematic behaviors, including substance use and risky sexual activity, and/or more distal developmental outcomes, such as employment and health indicators. Interventions targeting positive development showed improvements at different contextual levels including school, homes, and community settings.
School-based antibullying interventions reduced bullying perpetration and victimization outcomes.
Community- and school-based interventions prevented substance use initiation.
A limited number of reviews addressed self-harm prevention. Evidence suggests that gatekeeper training improves suicide literacy outcomes in gatekeepers, and school-based interventions prevent suicidal ideation and attempts in youth.
Responsive caregiving and parenting programs supported nurturing care through improvements in children's social, emotional, and behavioral outcomes.
The largest number of included studies had consistent findings that showed positive effects from universal prevention interventions on a large range of mental health outcomes.
Overall, more research is required in LMIC contexts that focuses on digital interventions, interventions provided outside of the school context, prevention of self-harm, and especially interventions targeting young children (aged <10 years).
Conclusions
High levels of statistical heterogeneity and diversity of study design, instruments, populations, interventions, and settings were common. This limited our ability to make comparisons and definitive statements about effectiveness of interventions, rendering policy interpretation and implementation challenging. Overall, this overview of reviews highlights the need for further research to consolidate findings and understand the specific criteria involved in creating positive mental health and development outcomes from the various interventions considered. Furthermore, there is urgent need for more research specifically in the LMIC context to improve coverage and extend findings to all SACA.
Acknowledgments
We thank Nabeela Ilyas and Christine Leung for their contributions to this article.
Dr Bhutta conceptualized and designed the study; Mr Vaivada conceptualized and designed the study and drafted the initial manuscript; Ms Harrison, Dr Irfan, Ms Sharma, and Ms Zaman screened the search results, screened the retrieved papers against the inclusion criteria, appraised the quality of papers, extracted the data, completed data tabulation and synthesis, and drafted the initial manuscript; and all authors reviewed, revised, and approved the final manuscript as submitted and agreed to be accountable for all aspects of the work.
FUNDING: This work was supported by a grant from the International Development Research Centre (#109010-001). The funder did not participate in the work. Core funding support was also provided by the SickKids Centre for Global Child Health in Toronto.
CONFLICT OF INTEREST DISCLAIMER: The authors have no conflicts of interest relevant to this article to disclose.
Comments