Table 2.
Data extraction and quality assessment of included studies (N=42).
| Author, year, country, quality | Study design | Setting and participants | Digital intervention | Outcomes and measures |
| Alvarez-Jimenez et al [58], 2013, Australia, high quality | Quantitative and qualitative, uncontrolled single-group, observation, questionnaire, and semistructured interview | Setting: early psychosis prevention and intervention center; sample: 20 patients (50% female; aged 15-25 years; 45% Anglo-Australian, 25% Asian, 10% biracial, and 5% African); presenting condition: first episode psychosis | Peer-to-peer web-based social networking, individually tailored web-based psychosocial interventions, and expert moderation: HORYZONS | Outcomes: depression and anxiety reduced; measures: BPRSa, CDRSb, and BAIc |
| Alvarez-Jimenez et al [30], 2018, Australia, medium quality | Quantitative, uncontrolled single-group, observation, and semistructured interview | Setting: PACEd clinic for ultrahigh-risk psychosis; sample: 14 patients (79% female; aged 15-25 years; ethnicity unknown; all Australia-born); presenting condition: ultrahigh risk for psychosis | Web-based social networking, peer-to-peer and professional moderation: MOMENTUM | Outcomes: depression reduced and psychological well-being improved; measures: SWLSe, MADRSf, and PSSSg |
| Bailey et al [22], 2020, Australia, high quality | Quantitative, uncontrolled single-group pre- and posttest, observation, and semistructured interview | Setting: tertiary-level mental health service; sample: 20 patients (55% female; aged 16-25 years; ethnicity unknown; country of birth: 75% Australia, 20% Asia, and 5% United Kingdom); presenting condition: suicidal ideation | Enhanced web-based social networking intervention: Affinity | Outcome: depression reduced; measure: PHQ-9h |
| Bhuvaneswar and Gutheil [49], 2008, United States, high quality | Qualitative, retrospective case study, and observation | Setting: psychodynamic psychotherapy clinic; sample: 1 patient (female; aged 17 years; ethnicity unknown); presenting condition: depression | Instant messenger | Outcome: psychological well-being worsened; measure: self-report |
| Blackwell et al [24], 2012, United States, high quality | Quantitative, randomized controlled trial, and questionnaire | Setting: general; sample: 100 adolescents (62% female; mean age 15.69 years, SD 2.91 years; 57% White, 16% Hispanic, 9% African American, and 18% ethnicity unknown); presenting condition: cystic fibrosis | Web-based social networking peer support program: CFfone.com | Outcomes: Depression and anxiety reduced; measure: HADSi |
| Campbell et al [55], 2019, Australia, medium quality | Qualitative and quantitative, participatory action research design, observation, and questionnaire | Setting: Kids Helpline family discord service; sample: 105 callers to helpline (82% female; aged 13-25 years; ethnicity unknown); presenting condition: mild to moderate depression or anxiety (not high risk) | Social networking site for peer-to-peer and counsellor-to-peer group support | Outcomes: depression and anxiety—data quality too low to assess; measures: CES-Dj and RCMASk |
| Canady [25], 2018, United States, high quality | Quantitative, cross-sectional study, questionnaire, and interview | Setting: general; sample: 1300 adolescents (gender unknown; aged 14-22 years; ethnicity unknown); presenting condition: none in particular | Web-based health information and digital health tools in general, including peer-to-peer health exchange networks | Outcomes: Depression and anxiety reduced; measures: PHQ-9 and self-report |
| Chyzzy et al [59], 2020, Canada, high quality | Qualitative and quantitative, uncontrolled single-group design, questionnaire, and semistructured interview | Setting: MPPSl intervention group; sample: 21 mothers (100% female; aged 17-24 years, mean age 21.3, SD 1.8, years; ethnicity unknown; country of birth: 66.7% Canada), presenting condition: generally healthy, 14.3% with prior history of depression | Individualized peer mentor support through telephone call and SMS text messaging: MPPS intervention | Outcomes: depression and anxiety reduced; measure: self-report |
| Clarke [45], 2018, Ireland, high quality | Qualitative, retrospective case study, and observation | Setting: clinical; sample: 1 patient (male; aged 16 years; ethnicity unknown); presenting condition: Asperger syndrome with comorbid depression | Telepsychiatry | Outcome: depression treatment engagement improved; measure: observation |
| Colder Carras et al [28], 2017, United States, medium quality | Quantitative, cross-sectional study, and questionnaire | Setting: 30 US schools; sample: 9733 students (51% female; aged 13-16 years, average age 14.1 years; 82.1% Dutch); presenting condition: none in particular | Web-based video gaming | Outcomes: depression and social anxiety reduced for social engaged gamers compared with problematic, at-risk, or extensive gamers; measures: depressive mood list and SASC-Rm |
| Cole et al [36], 2017, United States, —n | Quantitative, uncontrolled single-group design, and questionnaire | Setting: private university; sample: 231 undergraduate students (72% female; average age 19.28, SD 1.15, years; 67.1% White, 23.4% Asian American, 10.4% African American, 5.2% Hispanic or Latino, and 0.4% Other); presenting condition: none in particular | Web-based social networks in general | Outcomes: depression worsened; measures: DASSo, CTIp, and BDI-IIq |
| Dhesi [51], 2019, United Kingdom, high quality | Qualitative, cross-sectional, and web-based semistructured interviews | Setting: Kooth digital mental health care service; sample: 13 Kooth users (69% female; aged 14-18 years; 69.2% White British, 15.4% White and Asian, and 15.4% Other); presenting condition: none in particular | Web-based counseling (text) | Outcomes: anxiety reduced; measure: thematic analysis of interviews |
| Dolev-Cohen and Barak [48], 2013, Israel, high quality | Qualitative, case-control design, questionnaire, textual analysis, and observation | Setting: general; sample: 150 instant messaging users (63% female; aged 14-18 years; ethnicity unknown); presenting condition: distressed vs nondistressed groups of participants | Regular use of instant messaging | Outcome: psychological well-being improved; measure: PANASr |
| Ellis et al [56], 2011, Australia, — | Qualitative and quantitative, comparative randomized controlled trial, and questionnaire | Setting: university students not receiving mental health treatment; sample: 39 students (77% female; aged 18-25 years, mean age 19.67, SD 1.66, years; ethnicity unknown); presenting condition: anxiety or depression but none severe | Web-based cognitive behavior therapy self-help program (MoodGYM) compared with web-based support group (MoodGarden) | Outcomes: depression and anxiety reduced; measures: DASS and ATQs |
| Feinstein et al [26], 2012, United States, high quality | Quantitative, short-term prospective cohort study, and questionnaire | Setting: undergraduate university students; sample: 301 students (62% female; mean age 19.44, SD 2.05, years; 41% Asian or Pacific Islander, 41% White, 6% Latino, 6% African American, and 6% Other); presenting condition: some participants had raised depression, anxiety, or social anxiety at baseline | Social networking in general | Outcome: depression resulted in poor-quality social connections, which in turn worsened depression and anxiety; measures: DASS and BFNEt |
| Felnhofer et al [34], 2018, Austria, — | Quantitative, randomized controlled trial, and questionnaire | Setting: public university; sample: 95 students (87% female; mean age 23.34, SD 2.727, years; ethnicity unknown); presenting condition: none in particular | Avatars (virtual entities controlled by another human being) and agents (virtual entities controlled by a computer) | Outcome: social interaction anxiety unchanged; measure: SIASu |
| Frison and Eggermont [37], 2016, Belgium, medium quality | Quantitative, uncontrolled cross-sectional, and questionnaire | Setting: 18 randomly selected high schools in Flanders, Belgium; sample: 910 students with Facebook account (52% female; average age 15.44, SD 1.71, years; ethnicity unknown; country of birth: 96.1% Belgium, 1.8% Europe, and 2.1% non-European country); presenting condition: none in particular | Outcome: depression reduced; measure: CES-DCv | |
| Garrido et al [43], 2019, Australia, medium quality | Qualitative and focus groups | Setting: high schools and universities in Western Australia; sample: 23 students (65% female; aged 13-25 years; ethnicity unknown); presenting condition: DASS score <15 (severely depressed excluded) | A total of 6 currently available smartphone apps for mental health (Mood Mission, Music eScape, Pacifica, Mindshift, Headspace, and What’s Up) | Outcome: helpful and unhelpful aspects of smartphone apps for mental health; measure: thematic analysis of focus group content |
| Horgan et al [54], 2013, Ireland, medium quality | Qualitative and quantitative, pre- and posttest and qualitative descriptive designs, extraction of posts from website, and questionnaire for CES-D scores | Setting: University of Cork; sample: 118 students (36% female; aged 18-24 years; 98.3% White and 1.7% Asian or Asian Irish); presenting condition: depression | Depression support website with peer support forum | Outcome: depression reduced; measure: CES-D |
| Horgan and Sweeney [18], 2010, Ireland, medium quality | Quantitative, descriptive study, and questionnaire | Setting: university; sample: 922 students (62% female; aged 18-24 years; ethnicity unknown); presenting condition: none in particular | Internet use for mental health support | Outcome: reasons for use of internet-based mental health support; measure: self-developed questionnaire |
| Lim et al [57], 2019, Australia, high quality | Qualitative and quantitative, descriptive design, pre- and posttest questionnaires, mood tracker, and semistructured interview | Setting: local youth health service (participants with social anxiety disorder) and Australian university (participants without social anxiety disorder); sample: 20 participants (45% female; aged 18-23 years; 91% White and 9% multiracial or other); presenting disorder: with or without social anxiety disorder | +Connect, a digital smartphone app with video material | Outcomes: depression and anxiety reduced; measures: CES-D and SIAS |
| Liu and Yu [35], 2013, Taiwan, medium quality | Quantitative, cross-sectional study, and questionnaire | Setting: college; sample: 330 Facebook-using students (63% female; aged 18-23 years; ethnicity unknown); presenting condition: none in particular | Outcome: psychological well-being improved; measure: Ryff scales of psychological well-being | |
| McCloskey et al [23], 2015, United States, medium quality | Quantitative, uncontrolled single-group design, and questionnaire | Setting: university; sample: 633 undergraduate students with Facebook page (70% female; aged ≥18 years, median age 21 years; 64.8% White); presenting condition: none in particular; participants on average had mild levels of depression at baseline | Outcome: depression reduced; measure: PHQ-9 | |
| Mikami [38], 2010, United States, high quality | Quantitative, longitudinal, observation, and questionnaire | Setting: public middle school; sample: 92 social networking site users (58% female; mean age 20.92, SD 1.11, years; 58% White, 29% African American, and 13% Other or Mixed); presenting condition: none in particular | Web-based social networking | Outcome: depression—no outcomes reported; measure: CDIw |
| Ozcan and Buzlu [39], 2007, Turkey, high quality | Quantitative, uncontrolled single-group design, and questionnaire | Setting: university; sample: 730 undergraduate students who use the internet (53% female; mean age 20.84, SD 1.95, years; ethnicity unknown); presenting condition: none in particular | Internet use in general | Outcome: depression reduced; measure: BDI |
| Poppelaars [33], 2018, The Netherlands, medium quality | Quantitative, randomized controlled trial, and questionnaire | Setting: university; sample: 146 undergraduate students who play video games (71% female; mean age 20.2, SD 1.74, years; ethnicity unknown; nationality: 76% Dutch, 23% German, and 1% Other); presenting condition: none in particular; some with higher depressive symptoms at outset | Video game that included cooperation with other players and with mental health messaging vs video game without mental health messaging | Outcome: psychological well-being improved, with larger improvement for those higher in depressive symptoms; measures: BDI-II, SAMx, and International PANAS short form |
| Radovic [44], 2017, United States, United States, high quality | Qualitative, randomized controlled trial, semistructured interviews, think aloud, advisory boards, and focus groups | Setting: academic adolescent medicine clinic and specialty psychiatry clinic; sample: 23 patients (78% female; aged 13-20 years, mean age 16, SD 2.3, years); presenting condition: depression | Social media website for depressed adolescents | Outcome: adolescent-informed design of social media website for depression; measure: thematic analysis from semistructured interviews |
| Radovic [53], 2017, United States, medium quality | Qualitative, uncontrolled cross-sectional study, and semistructured interview | Setting: academic adolescent medicine clinic and specialty psychiatry clinic; sample: 23 patients (78% female; aged 13-20 years, mean age 16, SD 2.3, years; 87% White); presenting condition: depression | Social media | Outcomes: depressive symptoms either made participants reach for social media as a distraction or avoid it to avoid bringing down others. Psychological well-being improved; measure: thematic analysis from semistructured interviews |
| Rice et al [29], 2018, Australia, medium quality | Quantitative, uncontrolled single-group pilot, structured clinical interview, and questionnaire | Setting: mental health clinic; sample: 42 patients (50% female; aged 15-25 years, mean age 18.5, SD 2.1, years; ethnicity unknown; country of birth: 95.2% Australia); presenting condition: previous depression sufferers | Novel, moderated web-based social therapy intervention: Rebound | Outcomes: depression reduced and anxiety unchanged; measures: MADRS and DASS |
| Rice et al [31], 2020, Australia, high quality | Quantitative, single-group uncontrolled pre-post design, and questionnaire | Setting: 4 Headspace early intervention centers in northwestern Melbourne; sample: 89 patients (47% female; aged 14-25 years; ethnicity unknown); presenting condition: social anxiety | Social networking platform for socially anxious young people (Entourage): a wall function allows posting and commenting publicly | Outcomes: depression and social anxiety reduced and psychological well-being improved; measures: PHQ-9, MDRS-22y, LSASz, BFNE, SIAS, and SWEMWBSaa |
| Santesteban-Echarri et al [46], 2017, Australia, medium quality | Qualitative, uncontrolled single-group pilot, semistructured interview, and focus group data | Setting: mental health clinic; sample: 42 patients (50% female; aged 15-25 years, mean age 18.5, SD 2.1, years; ethnicity unknown; country of birth: 95.2% Australia); presenting condition: previous depression sufferers | Novel, moderated web-based social therapy intervention: Rebound | Outcome: efficacy and usability evaluation of web-based social therapy intervention; measure: thematic analysis from semistructured interviews |
| Saulsberry et al [40], 2013, United States, medium quality | Quantitative, randomized controlled trial, and telephone interview | Setting: 12 primary care sites across southern and midwestern United States; sample: 58 patients (57% female; mean age 17.26, SD 1.85, years; 61% White, 24% Black, 6% Asian, 5% Hispanic, and 4% Other); presenting condition: depression | Primary care provider motivational interview+CATCH-IT internet program vs primary care provider brief advice+CATCH-IT internet program | Outcome: depression reduced; measures: CES-D-10, DSM-IV-TRab, and PHQ-Aac |
| Selkie et al [47], 2020, United States, high quality | Qualitative, uncontrolled single-group design, and semistructured interviews | Setting: pediatric gender clinic; sample: 25 transgender adolescents with social media profile (44% trans-feminine; aged 15-18 years, mean age 16 years; 80% White non-Hispanic, 4% African American, 8% American Indian, and 8% Asian); presenting condition: none in particular | Social media platforms, including YouTube, Instagram, Facebook, Twitter, and Tumblr | Outcomes: positive and negative outcomes of using social media for mental health support; measure: — |
| Sharabi and Margalit [32], 2011, Israel, medium quality | Quantitative, cross-sectional crossover, and questionnaire | Setting: middle to high socioeconomic families vs those who failed in school (mostly from low socioeconomic families); sample: 716 students (48% female; aged 16-18 years; ethnicity unknown); presenting condition: with or without learning disabilities | Internet communication | Outcomes: psychological well-being negatively correlated with loneliness. Loneliness reduced by internet communication with people known offline; measure: Hebrew adaptation of Mood Scale |
| Sharabi and Margalit [19], 2011, Israel, medium quality | Quantitative and cross-sectional case-control | Setting: 3 high schools in urban Israel; sample: 887 students grades 10-12 (50% female; aged 16-18 years; ethnicity unknown); presenting condition: with (n=213) or without (n=674) learning disabilities | Internet communication | Outcome: psychological well-being reduced; measure: Hebrew adaptation of Affect Scale |
| Siriaraya et al [52], 2011, Cyprus, medium quality | Qualitative, cross-sectional study, and content analysis | Setting: general; sample: 400 messages from teenagers using web-based discussion forum (gender unknown; age range unknown; ethnicity unknown); presenting condition: none in particular | Web-based anonymous discussion forum | Outcome: level of support provided among adolescents; measure: Content analysis of web-based forum messages |
| Stockdale and Coyne [27], 2020, United States, high quality | Quantitative, longitudinal, and questionnaire | Setting: longitudinal study of intrafamily life participants; sample: 385 participants who use smartphones (53% female; aged 17-19 years; 70% European-American, 10% African American, 12% Multiethnic, 5% Asian American, and 2% Other); presenting condition: none in particular | Social media use | Outcomes: depression unchanged and anxiety worsened; measures: CES-DC and SCASad |
| van Rensburg et al [50], 2015, United States, high quality | Qualitative, uncontrolled single-group design, and semistructured interviews | Setting: Yale Psychiatric Hospital Intensive Outpatient Program; sample: 20 patients (75% female; aged 14-19 years; 80% White, 15% Hispanic, and 5% Mixed); presenting condition: combination of ADHDae, mood disorder NOSaf, MDDag, anxiety, PTSDah, psychosis, and ODDai | Social media for patient-provider interactions | Outcomes: positive (including safety) and negative (including anxiety) outcomes of patient-provider interactions through social media; measure: thematic analysis of semistructured interviews |
| van Zalk et al [41], 2011, Sweden, high quality | Quantitative, uncontrolled single-arm longitudinal study, and questionnaire | Setting: university in Utrecht; sample: 197 psychology freshmen (78% female; mean age 18.9, SD 1.6, years; ethnicity unknown; 92% Dutch origin); presenting condition: none in particular | Web-based chatting with friends through web-based social networking site | Outcome: depression unchanged; measure: BDI Dutch short version |
| Van Zalk and Tillfors [21], 2017, Sweden, high quality | Quantitative, longitudinal study, and questionnaire | Setting: Swedish school; sample: 526 students from grades 7-9 (68% female; aged 13-15 years; ethnicity unknown; 12.1% first-generation immigrants); presenting condition: none in particular | Web-based chatting with friends through web-based social networking site | Outcome: Reduced depression among those with higher, but not lower, social anxiety; measures: CES-D and SPSQ-Caj |
| Wright et al [42], 2013, United States, medium quality | Quantitative, cross-sectional observational study, and questionnaire | Setting: undergraduate university; sample: 361 students who use Facebook (53% female; mean age 20.26, SD 2.72, years; 77% White, 8.6% Native American, 4.4% Latino, 3.6% Asian American, 3.3% African American, and 3.3% Other); presenting condition: none in particular | Facebook use | Outcome: depression reduced; measure: CES-D |
| Yeh et al [20], 2008, Taiwan, medium quality | Quantitative, cross-sectional, and questionnaire | Setting: project of mental health survey; sample: 3477 college students (55% female; mean age 22.45, SD 1.56, years; ethnicity unknown); presenting condition: none in particular | Social support on the web | Outcome: depression worsened by higher web-based and lower actual social support; measure: Ko Depression Inventory |
aBPRS: Brief Psychiatric Rating Scale.
bCDRS: Children’s Depression Rating Scale.
cBAI: Beck Anxiety Inventory.
dPACE: Personal Assessment and Crisis Evaluation.
eSWLS: Satisfaction With Life Scale.
fMADRS: Montgomery–Åsberg Depression Rating Scale.
gPSSS: Perceived Social Support Scale.
hPHQ-9: Patient Health Questionnaire Depression Scale.
iHADS: Hospital Anxiety and Depression Scale.
jCES-D: Center for Epidemiological Studies Depression Scale.
kRCMAS: Revised Children’s Manifest Anxiety Scale.
lMPPS: Mothers’ Perceptions of Mobile Phone–Based Peer Support.
mSASC-R: Social Anxiety Scale for Children-Revised.
nNot available.
oDASS: Depression Anxiety Stress Scales.
pCTI: Cognitive Triad Inventory.
qBDI-II: Beck Depression Inventory II.
rPANAS: Positive and Negative Affect Scale.
sATQ: Automatic Thoughts Questionnaire.
tBFNE: Brief Fear of Negative Evaluation.
uSIAS: Social Interaction Anxiety Scale.
vCES-DC: Center for Epidemiological Studies Depression Scale for Children.
wCDI: Children’s Depression Inventory.
xSAM: Self-Assessment Manikin.
yMDRS-22: Male Depression Risk Scale.
zLSAS: Liebowitz Social Anxiety Scale.
aaSWEMWBS: Short Warwick–Edinburgh Mental Well-being Scale.
abDSM-IV-TR: Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Text Revision.
acPHQ-A: Patient Health Questionnaire-9 modified for Adolescents.
adSCAS: Spence Children’s Anxiety Scale.
aeADHD: attention-deficit/hyperactivity disorder.
afNOS: not otherwise specified.
agMDD: major depressive disorder.
ahPTSD: posttraumatic stress disorder.
aiODD: oppositional defiant disorder.
ajSPSQ-C: Social Phobia Screening Questionnaire for Children and Adolescents.