Table 1.
Study characteristics for longitudinal studies presented by exposure (school connectedness and school disconnectedness constructs)
| Author (Year) | N (% female), group characteristics, county | Mean age (years) ± SD (or range)* | School connectedness measure | Depression and anxiety measure | Relevant findings | ||||
|---|---|---|---|---|---|---|---|---|---|
| Exposure | Outcome | Direction of effect | Additional information | ||||||
| School Connectedness | |||||||||
| Arango et al. (2018) [43] | 142 (75%), USA |
T1: 13.41 (1.12) T2: 6 months later |
School Connectedness Scale | RADS-2:SF | School connectedness (T1) | Depression (T2) | Protective | ||
| Arora et al. (2017) [15] | 186 (49%), Asian American youth, USA |
T1: 12.50 (1.16) T2: 1 yr later |
School engagement (5 items), Teacher support (5 items) | CESD (adapted), State–Trait Anxiety Inventory for Children (adapted) | Teacher support (T1) | Depression (T2) | Protective | When teacher support was moderate-to-high at T1, high levels of anxiety at T1 were associated with increased levels of depressive symptoms at T2, an association that was not present under conditions of low teacher support | |
| School engagement (T1) | Depression (T2) | NS | |||||||
| Davis et al. (2019) [61] | 2,177 (48%), USA |
T1: 12.3 (0.7) T2: NR T3: NR T4: 13.8 (0.72) |
PSSM (4/20 items) | Orpinas Modified Depression Scale | School belonging (T1) | Depression (T2) | NS | Result across the whole sample | |
| School belonging (T2) | Depression (T3) | NS | Result across the whole sample | ||||||
| School belonging (T3) | Depression (T4) | NS | Result across the whole sample | ||||||
| School belonging (T1) | Depression (T2) | Protective | Result for females only | ||||||
| School belonging (T2) | Depression (T3) | Protective | Result for females only | ||||||
| School belonging (T3) | Depression (T4) | Protective | Result for females only | ||||||
| DeWit et al. (2011) [53] | 2,616 (54%), Canada |
T1: 13.77 (0.54) T2: ~ 6 months later T3: 1 yr later |
Social Support Appraisals Scale (SSAS) of the Survey of Children’s Social Support | CESD; Generalized Social Avoidance and Distress subscale of the Revised Social Anxiety Scale for Children | Classmate support (slope) | Depression (slope) | Protective | ||
| Teacher support (slope) | Depression (slope) | Protective | |||||||
| Classmate support (intercept) | Depression (slope) | NS | |||||||
| Teacher support (intercept) | Depression (slope) | NS | |||||||
| Classmate support (slope) | Social anxiety (slope) | Protective | |||||||
| Teacher support (slope) | Social anxiety (slope) | Protective | |||||||
| Classmate support (intercept) | Social anxiety (slope) | Risk | |||||||
| Teacher support (intercept) | Social anxiety (slope) | Risk | |||||||
| Fulco et al. (2019) [16] | 427 (50%), USA |
T1: 14 T2: 15 T3: 16 T4: 17 |
School engagement (9 items) | CESD (13 items) | Change in school engagement (T1 to T4, time-varying covariate) | Change in depressive symptoms (T1 to T4, non-significant) | Protective | Result for males only | |
| Change in school engagement (T1 to T4, time-varying covariate) | Change in depressive symptoms (T1 to T4, linear growth) | Protective | Result for females only | ||||||
| Gonzales et al. (2014)** [54] | 516 (51%), Mexican American adolescents, USA |
T1: 12.3 (0.54) T2: 2 yrs later T3: 5 yrs later |
School Engagement Scale—draws items from The School is Important Now Scale, the Academic Liking Scale, and the Importance of Education Scale | YSR at T1, T2, ASR at T3 | School engagement (T2) | Internalizing problems (T3) | Protective | T2 school engagement mediated the association between a family focused intervention and T3 internalising problems | |
| Hatchel et al. (2018) [55] | 404 (45.3% F; 51.8% M, 2.9% other), LGBTQ youth, USA |
T1: 15.27 (15–17) T2: 1 yr later T3: 2 yrs later |
PSSM (9 items) | Orpinas Modified Depression Scale (9 items) | School belonging (T1) | Depression (T2) | Protective | School belonging mediated the relationship between victimization and depression | |
| School belonging (T2) | Depression (T3) | Protective | |||||||
| Jiang et al. (2020) [45] | 2,041 (46%), Migrant adolescents, China |
T1 13.6 (0.71) T2: 1 yr later |
Emotional engagement (5 items) | CESD (5 items) | Emotional engagement (T1) | Depression (T2) | Protective | Emotional school engagement partially mediated the relationship between teacher discrimination and depression | |
| Joyce (2019) [35] | 13,120 (52%), USA |
T1: Grade 7–12 T2: 1 yr later |
Teacher support (2 items) | CESD (adapted) | Getting along with teachers (T1) | Depression (T2) | Protective | School connectedness at T2 partially mediated the effect between 1) getting along with teachers at T1 and depression at T2 and 2) feeling cared for by teachers at T1 and depression at T2 | |
| Feeling cared about by teachers (T1) | Depression (T2) | Protective | |||||||
| Klinck et al. (2020) [46] | 1,344 (51%), USA |
T1: 12.73 (11–14) T2: ~ 6 months later |
School Connectedness Scale | CESD; SCARED (total score and subscales) | School connectedness (T1) | Depression (T2) | Protective | ||
| School connectedness (T1) | Anxiety (T2) | NS | |||||||
| School connectedness (T1) | School avoidance (T2) | Protective | Association was not significant for other SCARED subscales (GAD, PD, SAD, SEP) | ||||||
|
Moderation analyses: Anxiety: Anxiety moderated the association between school connectedness and depression such that in adolescents at low risk of an anxiety disorder, higher school connectedness at T1 predicted lower levels of depressive symptoms at T2. Conversely, in adolescents at high risk of an anxiety disorder, there were no significant relationships between school connectedness and depressive symptoms. Gender: Time 1 associations between school connectedness and internalizing problems were stronger in magnitude for girls as compared with boys across all models. Race: In addition, race moderated the association, such that in adolescents identifying as non-Hispanic White, Hispanic, or Latinx, higher levels of school connectedness at T1 was associated with lower depression at T2, which was not the case for adolescents identifying as Black/African American | |||||||||
| Leonard et al. (2016) [30] | 769 (56%), Children in contact with CWS, USA |
T1: 12.69 (1.3) T4: 3 yrs later |
11 items from Drug-Free Schools and Communities Act Survey | YSR | School engagement (T1) | Internalizing problems (T4) | Protective | School engagement did not moderate the association between placement instability and internalizing problems | |
| Leonard & Gudiño (2016) [31] | 224 (58%), Children in out-of-home care during the study period, USA |
T1: 12.85 (1.25) T4: 3 yrs later |
11 items from Drug-Free Schools and Communities Act Survey | YSR | School engagement (T1) | Internalizing problems (T4) | NS | School instability prospectively predicted internalizing symptoms | |
| Lester et al. (2013) [63] |
T1: 1,054 (~ 54%) T2: 1,743 (additional students in secondary school [grade 8] that were not enrolled in the primary school at T1 [grade 7]), Australia |
T1: ~ 12 (end of grade 7) T2: ~ 12 (start of grade 8) T3: ~ 13 T4: ~ 14 |
School Connectedness Scale (4 items) | DASS-21 | School connectedness (T1) | Depression (T2) | NS | Result for males only | |
| School connectedness (T1) | Depression (T2) | Protective | Result for females only | ||||||
| School connectedness (T2) | Depression (T3) | Protective | Result for males only | ||||||
| School connectedness (T2) | Depression (T3) | Protective | Result for females only | ||||||
| School connectedness (T3) | Depression (T4) | Protective | Result for males only | ||||||
| School connectedness (T3) | Depression (T4) | NS | Result for females only | ||||||
| School connectedness (T1) | Anxiety (T2) | Protective | |||||||
| School connectedness (T2) | Anxiety (T3) | Protective | |||||||
| School connectedness (T3) | Anxiety (T4) | Protective | |||||||
| Lester & Cross (2015) [56] | 1616 (50% F) Australia |
T1: 12 T2: 13 T3: 14 |
Teacher connectedness (Teacher Connectedness Scale), School connectedness (School Connectedness Scale), The peer support at school scale (adapted from the 24-item Perceptions of Peer Social Support Scale) | Emotional symptoms (SDQ), Depression (DASS-21), Anxiety (DASS-21) | School connectedness (T1) | Depression (T2) | Protective | ||
| Anxiety (T2) | Protective | ||||||||
| Emotional problems (T2) | Protective | ||||||||
| Teacher connectedness (T1) | Depression (T2) | NS | |||||||
| Anxiety (T2) | NS | ||||||||
| Emotional problems (T2) | NS | ||||||||
| Peer support (T1) | Depression (T2) | Protective | |||||||
| Anxiety (T2) | Protective | ||||||||
| Emotional problems (T2) | Protective | ||||||||
| School connectedness (T2) | Depression (T3) | Protective | |||||||
| Anxiety (T3) | Protective | ||||||||
| Emotional problems (T3) | Protective | ||||||||
| Teacher connectedness (T2) | Depression (T3) | NS | |||||||
| Anxiety (T3) | NS | ||||||||
| Emotional problems (T3) | NS | ||||||||
| Peer support (T2) | Depression (T3) | Protective | |||||||
| Anxiety (T3) | Protective | ||||||||
| Emotional problems (T3) | Protective | ||||||||
| Li & Lerner (2011) [64] | 1,977 (43%), USA |
T1: 11 (0.52) T4: 3 yrs later |
Emotional school engagement (3 items) | CESD | Emotional school engagement | Depression | Protective | Four growth trajectories established for emotional school engagement (decreasing, moderate, high with decrease, and highest). Emotional engagement trajectory groups at T1 were associated with T4 depression. Members of the decreasing group of emotional engagement reported the highest levels of depression, whereas youth in the highest group were least depressed. Students who experienced high but decreasing emotional engagement were more depressed than youth in the highest group | |
| Loukas et al. (2016) [57] | 296 (50%), USA |
T1: 11.7 (0.76) T2: 12.3 (0.49) T3: 13.25 (0.44) |
5 items from the National Longitudinal Study of Adolescent Health | CDI | School connectedness (T1) | Depression (T2) | Protective | ||
| School connectedness (T1) | Depression (T3) | Protective | |||||||
| School connectedness (T2) | Depression (T3) | Protective | |||||||
| Markowitz (2016) [36] | 9,698 (53%), USA |
T1: 15.76 (1.57) T2: 1 yr later T3: 5 yrs later |
6 items | CESD (9 items) | School connection (T2) | Depression (T3) | Protective | There was an interaction between early adversity and school connection such that early adversity was associated with depressive symptoms only for boys with low levels of school connection | |
| McNeil et al. (2020) [32] | 627 (53%), Children in contact with CWS, USA |
T1: 12.5 (1.13) T2: 1.5 yrs later T3: 3 yrs later |
11 items from Drug-Free Schools and Communities Act Survey | CDI | School engagement (T1) | Depression (slope) | NS | ||
| School engagement (slope) | Depression (slope) | Protective | |||||||
|
Moderation analyses: Decreasing school engagement explained the association between parental non-involvement and increasing depression symptoms for Hispanic youth, but the indirect effect of parental non-involvement on depressive symptoms via school engagement was negative in White youth (increasing school engagement with low parental involvement led to decreasing depressive symptoms). The indirect effect was not significant for African American or Asian/other participants | |||||||||
| Moffa et al. (2016) [47] | 1,867 (51%, 1% other), USA |
T1: Grade 9–11 T2: 1 yr later |
5 items from the School Satisfaction subscale of the Multidimensional Students’ Life Satisfaction Scale | Internal distress (anxiety and depressive symptoms measured using 7 items) | School belonging (T1) | Internal distress (T2) | Risk | The authors noted that “the explained variance in internal distress was not substantial, Cohen’s f2 = .006. For this observed negligible effect size, the achieved power was not adequate (.75)”. A 1 standard deviation increase in school connectedness only predicted a 0.08 standard deviation increase in internal distress | |
| Okado et al. (2018) [58] | 209 (50%; survivors of pediatric cancer), USA |
T1: 12.48 (2.86) T2: 13.20 (2.93) T3: 15.64 (2.93) |
Hemingway Measure of Adolescent Connectedness | Behavior Assessment System for Children | School connectedness (T2) | Internalizing problems (T3) | Protective | ||
| Teacher connectedness (T2) | Internalizing problems (T3) | Protective | |||||||
| Peer connectedness (T2) | Internalizing problems (T3) | Protective | |||||||
| Pierre et al. (2020) [48] | 119 (0%) African-American males, USA |
T1: 15.33 (0.95) T2: 16.56 (0.97) |
PSSM | DASS | School belonging (T1) | Depression (T2) | NS | Sample included males only. T1 violence victimization and witnessing violence did not predict T2 depressive, anxiety, or stress symptoms at high levels of school belongingness | |
| School belonging (T1) | Anxiety (T2) | NS | |||||||
| School belonging (T1) | Stress (T2) | NS | |||||||
| Pössel et al. (2016) [66] | 2,545 (51%), Australia |
T1: Grade 8 (13.11 [0.56]) T2—T5: Grade 9–12 (1 yr intervals) |
Teacher-reported school climate (12 items)—obtained scores of two factors (teacher–student relationships and safe/orderly environment), which were averaged for analyses. The correlation between the two factors was 0.6, p < 0.001 | CESD | School climate (T1) | Depression (slope T1 to T5) | Risk | No difference between males and females | |
| School climate (T1 to T5 slope) | Depression (T1 to T5 slope) | NS | |||||||
| Sanders et al. (2020)**[49] | 294 (? F in 7th grade, but 54% at original recruitment in KG), USA |
T1: 7th Grade T2: 9th Grade |
School bonding and Affiliation with teacher subscales (People in My Life Questionnaire); General Adjustment subscale (SAQ) | SDQ | School bonding | Emotional symptoms | Protective | Estimated latent profiles of change in emotional symptoms and change in school bonding, resulting in three profiles (each) of both variables (high distress, medium distress and low distress for emotional symptoms, and strong school bond, average school bond, and weak school bond for school bonding). The profiles showed a moderate level of intercorrelation (r = 0.41), and 50% of the sample fell into a profile reflecting the same adjustment level (e.g., low, medium, or high) in both domains of emotional symptoms and school bonding | |
| Shochet & Smith (2014) [59] | 504 (45%), Australia |
T1: 13.3 (0.5) T2: 1 yr later T3: 1.5 yrs later |
PSSM | CDI | School connectedness (T1) | Depression (T2) | Protective | ||
| School connectedness (T1) | Depression (T3) | Protective | |||||||
| School connectedness (T2) | Depression (T3) | Protective | Mediated the association between classroom environment and depression | ||||||
| Shochet et al. (2011) [60] | 504 (45%), Australia |
T1: 13.3 (0.5) T2: 1 yr later T3: 1.5 yrs later |
School connectedness subscales: Caring Relations, Acceptance, and Rejection (PSSM) | CDI | Caring Relations (T2) | Depression (T3) | NS | Result for males only | |
| Acceptance (T2) | Depression (T3) | Protective | Result for males only | ||||||
| Rejection (T2) | Depression (T3) | NS | Result for males only | ||||||
| Caring Relations (T2) | Depression (T3) | NS | Result for females only | ||||||
| Acceptance (T2) | Depression (T3) | Protective | Result for females only | ||||||
| Rejection (T2) | Depression (T3) | NS | Result for females only | ||||||
| Stiles & Gudiño (2018) [33] | 2633 (52%), Youth in contact with CWS, USA |
T1: 10.04 (2.72) T2: 1.5 yrs later T3: 3 yrs later |
11 items adapted from the Drug-Free Schools and Community Act Survey | CBCL (subscale) | School engagement (T1) | Internalizing problems (T2) | NS | ||
| School engagement (T2) | Internalizing problems (T3) | NS | |||||||
| Wright & Wachs (2019) [51] | 416 (46%), USA |
T1: 13.89 (0.41) T2: 1 yr later |
School belongingness (18 items) | CESD; The Multidimensional Anxiety Scale for Children | School belongingness (T1) | Depression (T2) | Protective | For T2 depression and anxiety, there was a significant three-way interaction between cyber victimization, school-belongingness, and ethnicity. T2 depression/anxiety and cyber victimization were more strongly associated at lower levels of school-belongingness for Latinx adolescents | |
| School belongingness (T1) | Anxiety (T2) | Protective | |||||||
| Yu et al. (2016) [65] | 236 (58%), China |
T1: 7th Grade T2: 6 months later T3: 1 yr later T4: 14.34 (0.57); 1.5 yrs later |
School Engagement Scale at T2 and T3 | YSR at T3 and T4 (mean of 16 items) | School engagement (T2) | Anxiety & Depression (T3) | Protective | ||
| School engagement (T2) | Anxiety & Depression (T4) | Protective | |||||||
| School engagement (T3) | Anxiety & Depression (T4) | Protective | |||||||
| School Disconnectedness | |||||||||
| Benner et al. (2017) [44] |
252 (50%), Predominantly Latina/o and African American youth, USA |
T1: 14.38 (0.46) T2: 15.58 (0.51) |
Gottfredson’s measurement (5 items) | CDI | Decreasing school belonging (T1 to T2) (compared to stable and increasing school belonging) | Change in depressive symptoms (T1 to T2) | Risk | ||
| Boen et al. (2020) [34] | 20,475 (?), USA |
T1: Grade 7–12 T2: 1/2 yrs later T3: 5/6 yrs later T4: 12/13 yrs later |
Component obtained from Principal Component Analysis of interview and questionnaire items | CESD (9 items) | Low school connectedness (T1) | Depression (trajectory T1 to T4) | Risk | Low school connectedness was found to have a strong positive association with depressive risk, that diminished over time | |
| Cristini et al. (2012) [52] | 347 (53%), Italy | Data were collected at the end of each of the three middle school years (T1, T2, T3) | Teacher-student and student–student relationships using the School Situation Questionnaire | Depression and anxiety (5 items) | Socially isolated cluster (low on student–student relationships) at T1 | Depression/Anxiety (T2 & T3) | Risk (at T2 and T3) | Socially isolated group showed higher levels of emotional problems than the well-adjusted cluster at each wave | |
| Gunnarsódttir et al. (2021) [62] | 944 (48%), Sweden |
T1: 16 T2: 21 T3: 30 T4: 43 |
Principal Component Analysis on variables considered to capture interrelations occurring within the family and the school context | Depression (captured using six symptom measures) | Poor school connectedness (T1) | Depression (T2 to T4) | Risk | ||
| Tucker et al. (2011) [50] | 4,329 (52%), USA |
T1: 14.83 (95% CI 14.82 – 14.85) T2: ~ 21 |
School disengagement (5 items) | CESD (8 items) | School disengagement (T1) | Depression (T2) | Risk | ||
| Wickrama & Vazsonyi (2011) [37] | 20,745 (49%), USA |
T1: 13—19 yrs T2: NRT3: 6 yrs later |
School disengagement (4 items) | Depression (CESD; 8 items) | School disengagement (T1) | Depression (change in symptoms T1 to T3) | Risk | Interaction effects between race/ethnicity and school disengagement and between school minority concentration and school experiences were also statistically significant. For Hispanic American adolescents, school disengagement had a stronger influence on changes in depressive symptoms than for European American adolescents (reference group) | |
NS not significant, *School grade reported where age not provided, duration of follow-up timepoint compared to T1 (baseline); **Intervention studies classified as longitudinal for this review as the interventions were not designed to increasing school connectedness, PSSM Psychological sense of school membership scale, CESD Center for Epidemiological Studies – Depression scale, YSR Youth Self-Report, ASR Adult Self-Report, SCARED The Screen for Child Anxiety Related Disorders, DASS-21 Depression Anxiety and Stress Scale, SDQ Strengths and Difficulties Questionnaire, CDI Children’s Depression Inventory, CBCL Child Behavior Checklist, RADS-2:SF Reynolds Adolescent Depression Scale:Short-Form, GAD Generalised Anxiety Disorder, PD Panic Disorder, SAD Social Anxiety Disorder. SEP Separation Anxiety Disorder, T Time, CWS Child Welfare System