Table 3 Characteristics of included studies
Cross-sectional studies | |||||||||
Study | N (depressed) | % Female | Age in years | Recruitment | Self Terminology | Self Measure | Depression Status | Effects | |
 Auerbach et al. (2015) [13] | 52 (22) | 100 | 13–18 | Community | Self-referential biases and self-criticism | CTIC-S & endorsement of positive and negative words | Diag. (KSADS; DSM-IV) | Depression associated with more positive and fewer negative words. | |
 Becker-Weidman et al. (2009) [32] | 439 | 54 | 12–17 | TADS trial participants | View of self | CTIC-S | Diag. & Elev. (DSM-IV and CDRS-R ≥ 45) | Depression and hopelessness associated with view of self. | |
 Bennett et al. (1997) [33] | 328 | 58 | 11–19 | Outpatient and inpatient | Self-attitude | BDI (negative self-attitude factor) | Diag. (KSADS; DSM-III-R) | Depression associated with worse negative self-attitude than anxiety, disruptive behaviour and other psychiatric controls. | |
 Bradley et al. (2016) [34] | 41 (23) | 56 | 12–20 | Outpatient and community | Self-perception | Self-referent judgement | Diag. (KSADS; DSM-IV) | Depression associated with more positive and fewer negative descriptions. | |
 Cooper et al. (2005) [35] | 272 (28) | 100 | 17–18 | Community | Core beliefs | YSQ and negative self-beliefssubscale of the EDBQ | Elev. (Median split BDI) | Depression associated with more negative self-beliefs than eating disorder and healthy control group. Endorsement of negative word types varied between groups. | |
 Dozois et al. (2012) [36] | 47 (22) | 60 | 13–17 | CAMH program and community | Core beliefs and self-concept | YSQ–Short Form & HSPCA | Diag. (DICA-IV; DSM-IV) | Mixed effects. Depression associated with some self-concepts compared to non-psychiatric controls. | |
 Grilo et al. (1999) [37] | 127 (53) | 70 | 12–18 | Inpatient | Self-criticism | DEQ-A | Diag. & Elev. (DSM-III-R and BDI ≥18) | Mixed effects. Depression associated with higher self-criticism than abused group, but no difference on dependency subscale. | |
 Heath et al. (1999) [38] | 104 (29) | 47 | M = 170.9 Months (SD = 8.74) | School students | Self-concept | SPPC | Elev. (CDI ≥12) | Mixed effects. Depression associated with worse academic and non-academic self-concept. Age differences identified. | |
 Kendall et al. (1990) [39] | 34 | 55 | 11–13 | School students | Self-evaluation | My Standards Questionnaire | Diag. (KSADS; DSM-III) | Depression associated with lower evaluation of performance on personal domains. | |
 Koenig (1988) [40] | 721 (213) | 50 | 12–19 | School students and inpatients | Self-image | OSIQ | Diag. (Patient medical record; DSM-III) | Recurrent depression associated with poorer self-image than dysthymic disorder or atypical depression, but better self-image than single episode of depression for younger participants aged 12–15 | |
 Korhonen et al. (2001) [41] | 107 (68) | 73 | M = 17.9 (SD = 2.3) | Outpatient facility | Self-image | OSIQ | Diag. (SCID; DSM-III-R) | Mixed effects. Depression associated with worse self-image according total scores. Majority of subscales worse in depressed group. | |
 Lopez Molina et al. (2014) [42] | 137 | 74 | 18–24 | Community | Self-criticism | BDI items | Diag. (MINI; DSM-IV) | Mixed effects. Depression in females associated with some higher scores of self-criticism than depression in males. | |
 Marton et al. (1993) | 103 (38) | 52 | 15–19 | Outpatient facility and schools | Self-Perception | HSPPA | Diag. (KSADS; DSM-III-R) | Mixed effects. Depression associated with some lower scores of self-perception compared to control clinical and healthy groups. | |
 McClure et al. (1997) [43] | 31 (14) | 100 | 12–17 | School students | Self-Perception | HSPPA | Diag. (DICA-R-A) | Mixed effects. Depressed group rated themselves as less competent on some subscales of self-perception. | |
 Morey-Nase et al. (2019) [22] | 11 | 64 | 17–24 | Outpatient facility | Self-esteem | Qualitative interview | Diag. (Clinician diagnosis) | Young people described feeling like they are letting people down, not meeting own expectations, disappointing others, and self-loathing. | |
 Ofonedu et al. (2013) [44] | 10 | 60 | 13–17 | African American School students | Self | Qualitative interview | Diag. (K-SADS; DSM-IV) | Themes emotional sense of self, survival self and healing self. This included experiences of feeling worthless, inadequate, stupid, ugly. | |
 Orchard et al. (2017) [11] | 100 (43) | 85 | 12–17 | NHS CAMHS | Self-perception | K-SADS | Diag. (KSADS) | Depression associated with higher negative self-perception than other clinical group and no diagnosis group. | |
 Orchard et al. (2019) [45] | 291 (33) |  | 12–18 | Outpatient facility and schools | Self-evaluation | Self-description questionnaire | Diag. (KSADS; DSM-IV) | Depression associated with more negative and fewer positive words. Factor analysis revealed pro-social words which were equally endorsed by depressed and community adolescents. | |
 Pilecki et al. (2008) [46] | 90 (36) | 100 | Adolescent girls | Outpatient unit and school students | Self-image | OSIQ | Diag. (DSM-IV) | Mixed effects. Depression associated with worse self-images on majority of subscales, and worse self-image than anorexia on some subscales. | |
 Pinto et al. (1996) [47] | 40 (21) | 100 | 13–17 | Adolescent inpatient unit | Self-concept | PHCSCS | Diag. (DICA-R-A; DSM-III-R) | Mixed effects. Depression associated some negative self-concepts. BPD with depression group reported lower self-concept on some scales compared to depressed non-BPD group. | |
 Qian et al. (2002) [48] | 79 (40) | 56 | M = 20 | Undergraduate | Self-evaluation | SAI | Elev. (BDI ≥ 13) | Depression associated with lower self evaluation and lower perceived efficacy. | |
 Quevedo et al. (2017) [49] | 121 (86) | 50 | M = 14.75 (SD =1.64) | Inpatient unit and community | Self-attributions | Self-appraisal task | Diag. (KSADS; DSM-IV) | Depression associated with more negative and fewer positive self-evaluations. No difference between depressed groups. | |
 Robinson et al. (1992) [50] | 50 | 56 | 11–17 | Inpatient | Self-concept | PHCSCS | Diag. (Hospital records) | Mixed effects. Depression associated with some subscales of self-concept. | |
 Ross (1989) [51] | 33 (18) | 51 | 18–22 | University students | Self-traits | Self-referent judgement | Elev. (BDI ≥10) | Depression associated with more unstable positive and negative endorsements, i.e. endorsement of traits was more likely to change. | |
 Ross et al. (1986) [52] | 72 | – | 18–22 | University students | Self-traits | Self-referent judgement | Elev. (BDI ≥14) | Mixed effects. Depression associated with more negative traits, but not less positive traits. | |
 Rotundo et al. (1985) [53] | 84 (22) |  | 12–16 | Inpatient, outpatient and schools | Self-esteem and self-perception | PHCSCS | Diag. (DSM-III) | Depression associated with worse self-concept compared to clinical controls. | |
 Savilahti et al. (2018) [54] | 409 (206) | 71 | 13–17 | Inpatient and community | Self-image | OSIQ | Diag. (DSM-IV) | Depression associated with worse self-image. | |
 Wixom et al. (1993) [55] | 52 (17) | 100 | 14–18 | Inpatient | Self-criticism | DEQ | Diag. (medical chart; DSM-III) | BPD associated with more self-criticism than depression. | |
 Woo et al. (2004) [56] | 480 (238) | 40 | 13–19 | Outpatient facility and schools | Self-evaluation | Asian Adolescent Depression Scale | Diag. (KSADS; DSM-IV) | Depression associated with greater negative self-evaluation compared to community and clinical controls | |
Longitudinal studies | |||||||||
Lead Author | N (depressed) | % Female | Age in years | Sample | Self Terminology | Self Measure | Depression Status | Findings | |
 Ames et al. (2018) [57] | 662 | 52 | 12–18, followed up at 22–29 years | Community based prospective | Physical self-concept | Items from ‘HealthBehaviour in School-Aged Children scale’ | Elev. (B-CFPI; ‘Persistent high’ class based on latent class growth analysis’) | Mixed effects. Self-concept differed between different classes at time 1 and over time. | |
 Carbonell et al. (1998) [58] | 108 | – | Data at 5,9,15 and 18 years | Community | Self-concept | PHCSCS | Diag. (DSM-III-R) | Mixed effects. Some self-perceptions at age 9 associated with some impaired behavioural academic and psychosocial functioning at age 15. | |
 Ferro et al. (2015) [59] | 2825 | 49 | Data at multiple time points from 10 to 25 years | National Longitudinal Survey of Children and Youth | Self-concept | General Self-Image subscale of the Self-Description Questionnaire (Marsh 1992) | Elev. (CES-D) | Young people on a trajectory of sub-clinical and clinical symptoms showed poorer self-concept over time compared to those with minimal symptoms. | |
 Fine et al. (1993) [60] | 47 | 85 | M = 15.2 (SD = 1.1) | Outpatient facility | Self-concept | OSIQ | Diag. (K-SADS; DSM-III-R) | Self image predicted depressive symptoms/recovery from depression at 3 months and a year. Self-image was better predictor of depression than depression was of self image. | |
 Franko et al. (2005) [61] | 1727 (246 moderate, 209 mild) | 100 | 16–18 | Longitudinal cohort study | Self-worth | HSPPA | Elev. (CES-D - 16-23 mild depression, ≥24 moderate depression) | Mild and moderate depressed groups had lower self-worth than the non-depressed group 3 years later. | |
Cross-sectional and intervention | |||||||||
Study | N (depressed) | % Female | Age | Sample | Self Terminology | Self Measure | Depression Status | Intervention | Effects |
 King et al. (1993) [62] | 60 (30) | 76 | Inpatient M = 15.8 (SD = 1.1) Outpatient M = 15.7 (SD = 1.2) | Inpatient and community | Self-worth | HSPPA | Diag. (KSADS; DSM-III-R) | multidisciplinary programme | Mixed effects. Depression associated with global self-worth and some subscales. Improvement in depression was marked by increases in global self-worth and some subscales. |
Intervention studies | |||||||||
Study | N (depressed) | % Female | Age | Sample | Self Terminology | Self Measure | Depression Status | Intervention | Effects |
 Alavi et al. (2018) [63] | 15 | 86 | 14–17 | Outpatient facility | Self-concept | BSCI-Y | Diag. (DSM-IV) | Face to face vs e-CBT | Mixed effects. e-CBT and F2F did not differ on post-treatment self-concept. Pre- to post-treatment scores did not change in F2F group but did improve in e-CBT. |
 Fine et al. (1991) [64] | 66 | 83 | 13–17 | Inpatient | Self-concept | OSIQ | Diag. (KSADS; DSM-III-R) | Social skills training vs therapeutic group support | Mixed effects. Improvements in self-concept at post-treatment for therapeutic support group, no change for social skills group. At 9-month follow up social skills group improved. |
 Gottlieb et al. (2016) [65] | 439 | 54 | 12–17 | TADS treatment trial | Self-concept | CTIC-S | Diag. & Elev. (MDD & CDRS-R ≥ 45 | Fluoxetine vs CBT vs combination vs placebo | Mixed effects. Over 12 weeks, combined treatment group outperformed other groups on self-concept. Over 36 weeks combined group only outperformed the fluoxetine group. |
 Hintikka et al. (2003) [66] | 39 | 61 | 13–17 | Inpatient | Self-image | OSIQ | Diag. (SCID; DSM-III) | Individualised inpatient treatment programmes | Mixed effects. Improvements after treatment on some aspects of self-image. |
 Kurdziel et al. (2018) [67] | 1 | 100 | 14 | University clinic | Self-esteem and self-criticism | N/A | Diag. (DSM-IV) | Psychodynamic psychotherapy | Long-term psychodynamic therapy discussed as a method for targeting self-criticism amongst problems. |
 Le Noury et al. (2015) [68] | 275 | 37 | 12–18 |  | Self-Perception | HSPPA | Diag. (KSADS; DSM-III-R) | Paroxetine (20–40 mg), imipramine (200–300 mg), or placebo. | No effect. Paroxetine and imipramine did not improve self-perception compared to placebo. |
 Lusk et al. (2011) [69] | 15 | 60 | 12–17 | Outpatient facility | Self-concept | BSCI-Y | Diag. (DSM-IV-TR) | Cognitive–behavioural skills building intervention | Self-concept improved |
 Nasstasia et al. (2019) [70] | 68 | 53 | M = 20.75 (SD = 2.59) | Community and university populations | – | BDI (sub scales and items) | Diag. (SCID-1; DSM-IV) | Initial session of motivational interviewing followed by 12-week, multi-modal exercise program | Mixed effects. Improvement after intervention on some items from the cognitive subscale. |
 Rickhi et al. (2015) [71] | 62 | 70 | 12–24 (split into young 12–18 and older 19–24) | Community | Self-concept | Piers Harris II (younger) and (older) SFSCS | Diag. & Elev. (DSM-IV-TR & CDRS-R 40–70 or HAMD 12–24) | Spirituality informed e-mental health tool | Mixed effects, age differences. Self-concept improved for younger participants immediately after the intervention compared to waitlist, and over time. In older participants, change only in one of six factors. |
 Riley et al. (2011) [72] | 7 | 28 | 12–16 | Inpatient and Outpatient | Self-concept | TSCS-II short form | Diag. (Clinician assigned) | Group therapy. Based on adventure and problem-solving. | Mixed effects. Four out of six participants that completed treatment showed improvement in self-concept. More change in self-concept was seen towards the end of treatment. |
 Rossello et al. (2008) [73] | 112 | 55 | 12–18 | School students | Self-concept | PHCSCS | Diag. & elev. (DISC-2.1; DSM-III-R or CDI > 13) | Individual/group CBT or IPT | Mixed effects. Self-concept improved in both group and individual CBT but not in the IPT conditions. |