Table 3.
Longitudinal Studies Examining Social Isolation/Loneliness and Subsequent Mental Health Outcomes
| Author (year), country | Sample (selection criteria) | Total N (% male participants) | Child (≤11 y)/adolescent (12−18 y)/young adult (≥19 y) | Age range, y | Mean age (SD) at T1 | Social isolation/ loneliness measure | Mental health measures | Cross-Sectional associations r (p) | Length of follow-up, y | Is social isolation/loneliness associated with later mental health? |
|
|---|---|---|---|---|---|---|---|---|---|---|---|
| Depression | Anxiety | ||||||||||
| Boivin et al. (1995),61 Canada |
School students | 774 (51.8) | Child | 9–12 | 10.8 (NS) | LSDQ | CDI | Lon-Dep 0.53 (p < .001) | 1 | T1 Lon – T2 Dep: r = 0.36 (p < .01) T1 Lon accounted for 8.3% of variance in T2 Dep |
|
| Christ et al. (2017),62 USA |
National Survey of Child and Adolescent Well-being (child welfare cohort) | 2776 (47) | Adolescent | 11–17 | 13.5 (NS) | LDSQ 7 peer isolation items | 4 items from YSR | NS | 7 | Controlling for caregiver neglect and covariates, a 1-SD increase in peer isolation was associated with a 0.49-SD increase in depression | |
| Danneel et al. (2019),63 Belgium |
Longitudinal cohorts | Sample 1: 1116 (51.1), Sample 2: 1423 (47.6), Sample 3: 549 (37.33) |
Adolescent | Sample 1: 11–17 Sample 2: 11–18 Sample 3: 12–17 |
Sample 1: 13.79 (0.94) Sample 2: 13.59 (0.98) Sample 3: 14.82 (0.79) |
LACA peer-related loneliness subscale | Samples 1 and 3 – SAS-A; CES-D. Sample 2 - CDI |
Lon-Social anxiety 0.58 ≤ r ≤ 0.67 Lon-Dep 0.48 ≤ r ≤ 0.56 (all <.01) |
1 | Not significant | Lon → Social anxiety β = 0.10 (p < .001) |
| Fontaine et al. (2009),64 USA |
School students (longitudinal cohort) | NS (52) | Child | NS Estimated 5–9 | NS | LSDQ (T2) | Internalizing items from: CBCL (mother T1 and T3); TRF (teacher T1 and T2); YSR (self T2 and T3) |
NS | 2-3 | T2 Lon → Anx/Dep symptoms at T3 γ2 = 0.18, z = 2.60 (p < .01) |
|
| Jones et al. (2011),65 USA |
Longitudinal cohort | 889 (50) | Child | 6 | NS | LSDQ | CDI short form | NS | 9 | Indirect effects T1 Lon → T2 Suicidal thoughts through Dep (β = 0.06, p < .001) | |
| Ladd and Ettekal (2013),66 USA |
School students (longitudinal cohort) | 478 (50) | Adolescent | 12–18 | 12.0 (n.s) | LSDQ – revised - 3 items | Depression items CBCL (parent); TRF (teacher); YSR (self) |
Lon-Dep 0.19 (p < .01) (parent), 0.38 (p < .001) (teacher) 0.62 (p < .001) (self) |
7 | Changes in Lon associated with changes in Dep reported by teachers (r = 0.63, p < .001) and adolescents (r = 0.65, p < .001), but not parents (r = 0.18,p = .13) |
|
| Lalayants and Prince (2015),67 multiple countries | National Survey of Child and Adolescent Wellbeing (child welfare cohort) | 356 (0) | Adolescent | 11–12 | NS | LSDQ | CDI | NS | 1.5 | T1 Lon → T2 Dep AOR = 2.93, CI = 1.74−4.91 (p < .001) T1 lonely female participants were 5.09 times more likely (CI 2.24−11.56 (p < .001) to be depressed at T2 |
|
| Lapierre et al. (2019),68 USA |
College Students | 346 (33.6) | Young adult | 17–20 | 19.11 (0.75) | UCLA Loneliness Scale | 10-Item CES-D | Lon-Dep 0.628 (T1), 0.666 (T2) (p < .001) | 0.25 | T1 Lon – T2 Dep (r = 0.524, p < .001) T1 Lon → T2 Dep b = 0.21, SE = .05 (p < .001) |
|
| Lasgaard et al. (2011b),69 Denmark | School students | T1: 1009 (43) T2: 541 (40) |
Adolescent/Young adult | 15–26 | 17.11 (1.11) | SELSA-short form; MSPSS | BAI-Y, BDI-Y |
Lon-Dep 0.61 (p < .0005) Lon-Anx 0.51 (p < .0005). Soc support– Dep r = −0.12, −0.18, −0.28 (all p < .0005) |
1 | T1 Lon→ T2 Dep (r = 0.37, p < .0005) Cross-lagged structural equation modeling found T1 Lon did not predict Dep at T2 |
|
| Liu et al. (2020),70 China |
College students | 741 (28.3) | Young adult | NS (estimated 18−20) | 18.47 (0.87) | 6 item index of social isolation based on only child status, number of friends, frequency of contact with friends and family; UCLA Loneliness Scale | SDS | NS | 3 | Female participants: T1 isolation associated with increased Dep (β = 0.22, p < .001) Lon associated with increased Dep (β = 0.23, p < .001) Male participants: T1 isolation associated with increased Dep (β = 0.25, p < .01) Lon did not predict Dep (β = 0.14, p > .05) |
|
| Mak et al. (2018),71 USA |
School students (randomized trial) | 687 (47.7) | Adolescent | NS (estimated 11–14) | 11.27 (0.49) | LSDQ | SAS-A | Lon-social anxiety 0.41−0.45 (p < .01) | 1.5 (T2), 3 (T3) | T1 Lon → T2 Social Anxiety (β = 0.09, p < .05). T2 Lon → T3 Social Anxiety (β = 0.12, p < .01) By gender: T2 Lon → T3 Social Anx: Boys (β = 0.22, p < .001) Girls (β = 0.01 p = .79) |
|
| Matthews et al. (2015),72 UK |
Twin birth cohort | 2232 (NS) | Child | 5 | NS | 6 items from CBCL (parent) and TRF (teacher) | MASC | NS | 7 | T1 social isolation failed to predict T2 Anx, controlling for T1 Anx | |
| Qualter et al. (2010),73 UK |
School students | 296 (49.3) | Child | 5 | NS | T1 and T2: Peer and Parent subscales LACA | T1: T-CARS T2 and T3: DDPCA |
T1 Peer Lon-internalizing symptoms 0.32 (p < .01) Parent Lon-Internalizing Symptoms 0.09. T2 Peer Lon- Dep 0.13 (p < .05) Parent Lon-Dep 0.12 (p < .05) |
8 | T1 Peer Lon-T2 Dep r = 0.07 T1 Peer Lon-T3 Dep r = 0.06 T2 Peer Lon – T3 Dep r = 0.12 (p < .05) T1 Parent Lon – T2 Dep r = 0.19, p < .01 T1 Parent Lon-T3 Dep r = 0.13 (p < .05) T2 Parent Lon-T3 Dep r = 0.08 Structural model: Duration of Peer Lon → T3 Dep T1 and T2 Peer Lon, Parent Lon (T1, T2, and duration) did not independently predict T3 Dep |
|
| Schinka et al. (2013),74 USA |
Longitudinal cohort study | 832 (53) | Child | 9 | NS | LDSQ | T1: CBCL (mother) T3: CDI−Short form; Suicide items from CBCL and YSR |
T3 Lon-Dep -0.10 (p < .01) Lon− suicidal ideation r = 0.02 Lon− suicide attempt r = 0.4 |
2 (T2), 6 (T3) |
T1 Lon-T3 Dep r = 0.01 T2 Lon-T3 Dep r = −0.01 T1 Lon-T3 Suicidal ideation r = 0.00 T2 Lon-T3 suicidal ideation r = 0.03 T1 Lon-T3 suicide attempt r = 0.02 T2 Lon-T3 suicide attempt r = −0.01 |
|
| Vanhalst, Goosens et al. (2013)75 and Vanhalst, Klimstra et al. (2012),76 Netherlands | Community sample via municipality registers | 389 (53) | Adolescents | 15 | 15.22 (0.60) | LACA Peer-related loneliness subscale | 6 item depression questionnaire; SCARED generalized anxiety, panic and social anxiety subscales. | Lon-Dep 0.34−0.50 (p < .001) Lon- Perceived stress 0.23 (p < .001). Lon- Generalized Anx 0.40 (p < .001), Lon-Panic 0.13 (p < .05), Lon− Social phobia 0.47 (p < .001) |
5 | T1 Lon → T2 Dep symptoms (B = 0.13, p < .001) | |
| Vanhalst, Luyckx et al. (2012)77 Belgium | University students | Sample 1: 514 (10.9) Sample 2: 437 (17) |
Young adults | Sample: 19.62 (0.62) Sample 2: 18.22 (1.21) |
NS | Sample 1: 8-item revised UCLA Loneliness Scale. Sample 2: LACA Peer-related loneliness subscale |
Sample 1: 12-item CES-D Sample 2: 20-item CES-D |
Sample 1: Lon-Dep 0.49−0.52 (p < .001) Sample 2: Lon-Dep r = 0.40−0.60 (p < .001) |
2 | Sample 1: T1 lon – T2 Dep r = 0.35 (p < .001) T1 lon – T3 Dep r = 0.36 (p < .001) Lon → associated with Dep across both time intervals. Sample 2: cross-lagged path from Lon associated with Dep (b = 0.12, p < .05) |
|
| Wang et al. (2020),78 China |
School students | 921 (48.3) | Adolescents | 12–15 | 12.98 (0.66) | Revised UCLA Loneliness Scale (T1 and T2) | SCARED; DSRSC (T1 and T3) |
T1 Lon- Anx 0.40 (p < .001) Lon-Dep 0.57 (p < .001) |
1 | T1 Lon-T3 Dep 0.36 (p < .001) T2 Lon-T3 Dep 0.46 (p < .001) |
T1 Lon-T3 Anx 0.29, p<.001. T2 Lon-T3 Anx 0.36 (p < .001) |
| Zhou et al. (2020),79 China | School students | 866 (49) | Adolescents | 11–15 | 12.98 (0.67) | UCLA Loneliness Scale (T1 and T2) | DSRSC (T3) | T1 Lon-Dep r = 0.56 (p < .001) |
2 | T1 Lon-T3 Dep r = 0.38 (p < .001) Controlling for age, sex, and SES, T2 Lon-T3 Dep adjusted b = 0.34 (p < .001) |
|
Note: Anx = Anxiety; BAI-Y = Beck Anxiety Inventory for Youth; BDI-Y = Beck Depression Inventory for Youth; CBCL = Child Behaviour Checklist; CDI = Children’s Depression Inventory; CES-D = Center for Epidemiologic Studies Depression Scale; DDPCA = Depression Profile for Children and Adolescents; Dep = depression; DSRSC = Birleson Depression Self-Rating Scale for Children; Lon = Loneliness; LSDQ = Loneliness and Social Dissatisfaction Questionnaire; LACA = Loneliness and Aloneness Scale for Children and Adolescents; MASC = Multidimensional Anxiety Scale for Children; MSPSS = Multidimensional Scale of Perceived Social Support; NS = not specified; SAS-A = Social Anxiety Scale for Adolescents; SCARED = Scale for Child Anxiety Related Emotional Disorders; SES = socioeconomic status; SDS = Zung Self-rating Depression Scale; SELSA = Social and Emotional Loneliness Scale for Adults; T-CARS = Teacher-Classroom Adjustment Rating Scale; T1 = Time 1; T2 = Time 2; T3 = Time 3; TRF = Teacher Rating Form; YSR = Youth Self-Report Form.