Table 2.
Detailed characteristics of included studies
| Author (year) | Country | N of participants (N = sex/gender) | Mean age in years (SD), range | Mean IQ (SD), range | ER measure (Informants) | Findings regarding emotion regulation (ER) |
|---|---|---|---|---|---|---|
| [49] Brewe et al. (2021)1 | USA | ASD group: N = 37 (29 M, 8 F) | 15.28 (2.21), 12–21 | 103.31 (17.13) | EDI (caregivers) | Therapeutic alliance was associated with improvements in the dysphoria subscale at post-treatment but not with the reactivity subscale. Internal consistencies for both subscales ranged from .83 to .95 at various time points. [Implementation] |
| [50] Charlton et al. (2020)2 | USA | ASD group: N = 27 (20 M, 7 F) | 18.54 (2.02), 16–24 | 103.88 (10.67) | RSQa (adolescents, caregivers) | Individuals with comorbid anxiety or mood disorders reported greater involuntary engagement (IE) and lower voluntary engagement (VE) than those without these disorders. However, caregiver reports showed no differences in IE or VE between individuals with and without anxiety or mood disorders. Moreover, IE was linked to internalizing and externalizing symptomatology only in caregiver ratings. Internal consistency was .86 for the VE subscale and .83 for the IE subscale. [Selection and Implementation] |
| [51] Chiu et al. (2024)2 | Hong Kong | ASD group: N = 23 (16 M, 7 F); non-ASD group: N = 32 (8 M, 24 F) |
ASD group: 15.78 (1.78), 13–18; non-ASD group: 16.38 (1.34), 13–18 |
ASD group: 90.30 (18.11); non-ASD group: 116.34 (12.79) | EDI (parents/ caregivers) HRV | The ASD group exhibited greater emotion dysregulation than the non-ASD group on both the reactivity and the dysphoria subscales; however, no significant differences in HRV were found between the groups. In the ASD group, higher levels of reactivity and dysphoria were both associated with lower HRV. Internal consistency was .98 for the reactivity subscale and .84 for the dysphoria subscale. [Implementation] |
| [52] Chou et al. (2023)1 | Taiwan | Experimental ASD group: N = 28 (23 M, 5 F); control ASD group: N = 38 (29 M, 9 F) |
All: 13.93 (0.93); experimental group: 13.89 (0.87); control group: 13.97 (0.99) |
n/a | ERS (adolescents, teachers) |
Adolescents in the experimental group showed improvement over time compared to the waitlist control group in teacher-reported overall ER. These findings were also evident at the subscale level of the ERS. Self-reports indicated similar trends, without reaching statistical significance. Internal consistencies of the ERS ranged from .92. to .95 [Identification and Selection] |
| [53] Coffman et al. (2024)1 | USA | ASD group: N = 15 (12 M, 3 F) | 15.62 (1.54), 13.14—18.32 | n/a | ABC-2-I, BRIEF-2, EDI (parents) | The ASD group showed improvement over time on the reactivity subscale of the EDI, the irritability subscale of the ABC-2 and in the emotion regulation index of the BRIEF-2. Additionally, post-hoc analyses revealed relevant improvements in these measures, as well as in the dysphoria subscale of the EDI, from the start of the intervention to multiple post-intervention assessments. No information on internal consistencies. [Implementation] |
| [54] Conner et al. (2023)2 | USA | ASD group: N = 57 (46 M, 11 F) | 18.56 (2.04), 16–24 | 105.82 (12.40) | DERS (adolescents) | Associations were found between ER, ASD traits, anxiety, and depression. In a regression model, ER accounted for a high percentage of the variance in depression and anxiety. Internal consistencies of DERS subscales ranged from .79 to .97. [Identification, Selection and Implementation] |
| [55] Conner et al. (2022)2 | USA | ASD group: N = 78 (61 M, 15 F, 2 O) | 14.87 (2.27), 12–20 | 102.44 (16.23) | EDI (caregivers) | Associations were found between the reactivity subscale and the dysphoria subscale with intolerance of uncertainty. Both EDI subscales mediated the association between intolerance of uncertainty, anxiety, and depression. Internal consistency was .93 for the reactivity subscale and .86 for the dysphoria subscale. [Implementation] |
| [56] Conner et al. (2019)1 | USA | ASD group: N = 17 (15 M, 2 F) | 14.94 (1.54), no age-range reported | 98.47 (11.71) |
EDI, ABC-I (parents), RSQb (adolescents, parents) |
Adolescents showed parent-reported improvements across various ER measures. Parent-reports on the RSQ indicated improved ER whereas self-reports did not. Internal consistency was .94 for the EDI and .85 for the ABC-I. RSQ internal consistencies ranged from .55 and .82 in the parent version and from .82 and .92 in the self-report version. [Selection and Implementation] |
| [57] Dağdelen et al. (2021)2 | Turkey | ASD group: N = 60 (30 M, 30 F); ADHD group: N = 60 (30 M, 30 F); control group: N = 60 (30 M, 30 F) | ASD group: 14.02 (1.44); ADHD group: 14.00 (1.43); control group: 13.55 (1.41), no age-range reported | ASD group: 95.47 (10.31); ADHD group: 95.47 (10.31); control group: 94.45 (10.82) | DERS (adolescents) | The ASD group had higher emotion dysregulation scores compared to both the ADHD group and the control group. Additionally, emotion dysregulation was associated with lower theory of mind scores. Internal consistency data were not reported. [Identification, Selection and Implementation] |
| [58] Fatta et al. (2024)1 | Italy | Experimental ASD group: N = 18 (12 M, 6 F); waitlist ASD group: N = 19 (14 M, 5 F) | Experimental ASD group: 15.14 (2.26), 12.3–18.2; waitlist ASD group: 15.5 (1.74), 12.2–18.2 | Experimental ASD group: 109.65 (10.51); waitlist ASD group: 106.43 (16.22) | BRIEF-2 (parents) | A group by time interaction effect was found for the BRIEF-2 emotion regulation index, indicating a relevant improvement in ER abilities in the experimental group. These improvements persisted at follow-up. Internal consistency was .90 for the emotion regulation index. [Implementation] |
| [59] Gerber, et al. (2024)1 | USA |
Experimental ASD group: N = 20 (16 M, 4 F); control ASD group: N = 20 (16 M, 4 F) |
Experimental ASD group: 14.27 (1.76); control ASD group: 14.17 (1.41), no age-range reported | Experimental ASD group: 103.47 (16.93); control ASD group: 107.47 (14.40) | BRIEF-2 (adolescents, parents) |
A group effect was found for the parent-reported BRIEF-2 emotion regulation index at the 3-month follow-up, whereas adolescent self-reports indicated marginal improvements Internal consistency ranged from .89 to .91 for the parent-report and from .85 to .88 for the adolescent-report. [Implementation] |
| [60] Goldfarb et al. (2021)3 | Israel |
ASD group: N = 10 (5 M, 5 F); N = 6 (3 M, 3 F) within the age-range 10–24 y |
Age range of included adolescents: 18–24 | n/a | In-depth interview, thematic analysis (adolescents) | Adolescents reported using self-harming behaviors to regulate their emotions, such as scratching with wire to shift away from emotional pain, self-hitting to prevent aggression towards others during anger and severe anxiety, and self-hitting to maintain self-control or regain focus in emotionally challenging situations. [Selection and Implementation] |
| [61] Gormley et al. (2022)2 | Ireland | ASD group: N = 43 (32 M, 11 F) | 13.1 (2.5), no age-range reported | IQ > 70 |
ERICA (adolescents,); ERC (parents) |
Adolescents exhibited more ER difficulties relative to normative data. A negative association was found between self-reported ER abilities and alexithymia. Regression analysis indicated that alexithymia accounted for a substantial proportion of variance in ER abilities. Parent reported emotional lability and negativity were strongly linked to ASD severity. Internal consistency was .84 for ERICA and ranged from .73 to .86 for ERC subscales. [Identification, Selection and Implementation] |
| [62] Khor et al. (2014)2 | Australia | ASD group N = 31 (26 M, 5 F) | 14.46 (1.83), 12–18 | 99.87 (14.33), 70–130 | RSQ (adolescents, parents) | The self-reports and parent-reports differed for most subscales, except for similar levels of reported disengagement coping and involuntary disengagement. The use of maladaptive ER strategies was associated with both self- and parent-reported behavioral and emotional problems. Internal consistencies for subscales and both respondents ranged from .64 to .89. [Selection and Implementation] |
| [63] Kim et al. (2021)1 | Republic of Korea | ASD group: N = 7 (7 M) | 20.29 (1.11), 19–22 | 92.57 (n/a), 76–120 | K-BDEFS (adolescents, caregivers) | Self-reports showed improvements on the regulation of emotions subscale from pre- to post-intervention. In contrast, caregiver-reports indicated a decline in scores on the same subscale during the same period. Internal consistency data were not reported |
| [64] Kose et al. (2023)2 | Turkey |
ASD group: N = 50 (38 M, 12 F); non-ASD group: N = 73 (35 M, 38 F) |
ASD group: 14.1 (2.2), 11–18; non-ASD group: 14.6 (2.6), 11–18 |
ASD group: 98.02 (15.24) |
ERQ (adolescents) | There were no differences in the ER subscales of cognitive reappraisal and expressive suppression between the two groups, nor was there a gender effect. High use of cognitive reappraisal was associated with better social skills and social functioning in the ASD group. Internal consistency was .78 for cognitive reappraisal and .73 for expressive suppression. [Selection] |
| [65] López-Pérez et al. (2018)2 | Spain |
ASD group: N = 30 (30 M); non-ASD group: N = 30 (30 M) |
ASD group: 10.97 (0.88), 10–12; non-ASD group: 11.07 (0.86), 10–12 |
ASD group: 101.50 (11.98); non-ASD group: 103.80 (10.71) | DERS (adolescents) | Group differences were observed, including an interaction between emotion dysregulation and group. Post-hoc analyses revealed greater difficulties in the ASD group regarding goal-directed behavior, emotional awareness, and emotional clarity, whereas no differences emerged in non-acceptance, impulse control, or regulation strategies. Internal consistency for subscales ranged from .71 to .78. [Identification, Selection and Implementation] |
| [66] Mazefsky et al. (2014)2 | USA |
ASD group: N = 25 (24 M, 1 F); non-ASD group: N = 23 (22 M, 1 F) |
ASD group: 15.22 (2.25), 12–19; non-ASD group: 15.56 (2.76), 12–19 |
ASD group: 110.48 (13.59); non-ASD group: 113.23 (12.87) | RSQ (adolescents) | No correlations were found between age, pubertal development, ASD traits and RSQ subscales. The ASD group reported greater use of both voluntary and involuntary disengagement strategies, as well as involuntary engagement strategies, compared to the non-ASD group. Distinct patterns of correlations among ER strategies were observed, with various strategies linked to both self- and parent-reported psychopathology. Internal consistency for the RSQ was .95 in the ASD group. [Selection and Implementation] |
| [67] Mazefsky et al. (2020)2 | USA |
ASD group: N = 25 (24 M, 1 F); non-ASD group: N = 23 (22 M, 1 F) |
ASD group: 14.95 (2.47), 12–19; non-ASD group: 15.5 (2.76), 12–19 |
ASD group: 115.00 (14.60); non-ASD group: 113.23 (12.87) | CBCL-EDI (parents) fMRI | Brain activity in specific regions (left and right insula, right pulvinar, left dorsolateral cortex) predicted emotion dysregulation scores in the ASD group. Prolonged activity in the prefrontal cortex and salience network in response to negative words was observed. Internal consistency for the CBCL-EDI was .87. [Implementation] |
| [68] Phillips et al. (2024)1 | UK |
ASD group: N = 27 (27 F); Non-ASD group: N = 215 (7 M, 208 F) Both groups with symptoms of BPD |
All: 16.48 (0.98), 13–17 | No intellectual disability in ASD group | DERS (adolescents) | The ASD group showed a reduction in emotion dysregulation scores from pre- to post intervention in an exploratory service evaluation. There were no differences in change scores between the ASD group and the non-ASD group. Internal consistency of the DERS ranged from .88 to .96. [Identification, Selection and Implementation] |
| [69] Pickard et al. (2020)2 | UK |
ASD group: N = 61 (42 M, 19 F); non-ASD group: N = 62 (26 M, 36 F) |
ASD group: 13.46 (1.77), 11–17; non-ASD group: 13.52 (1.57), 11–17 |
ASD group: 98.16 (13.99); non-ASD group: 100.76 (11.55) | CERQ (adolescents) | There were no differences between the ASD group and the non-ASD group regarding adaptive and maladaptive ER scores. An association between maladaptive ER scores and social anxiety was found in the ASD group. Internal consistency for the CERQ ranged from .84 to .89 in the ASD group. [Selection] |
| [70] Rosenthal et al. (2013)2 | USA |
ASD group 11–13 years: N = 50 (38 M, 12 F); ASD group 14–18 years: N = 36 (31 M, 5 F) |
ASD group 11-13y.: n/a (n/a) ASD group 14-18y.: n/a (n/a) |
ASD group 11-13y.: 109.38 (19.07); ASD group 14-18y.: 104.58 (17.16) |
BRIEF (caregivers) | Both groups exhibited clinically elevated scores (T > 65) on the shift subscale. Although T-scores on the emotional control subscale were also elevated, they did not reach clinical relevance. Post-hoc testing revealed no differences across the age groups. Internal consistency data were not reported. [Implementation] |
| [71] Russel et al. (2024)1 | USA | ASD group: N = 12 (7 M, 5 F); | 14.25 (1.22), 11–16; | 97.42 (19.0); | EDI (caregivers) | Post-hoc analysis revealed an increase of ER skills following the intervention. This improvement did not persist at the 10-week follow-up. Internal consistency data were not reported. [Implementation] |
| [72] Salem-Guirgis et al. (2019)1 | Canada | ASD group: N = 23 (19 M, 4 F) | 15.65 (2.57), 12–23 | 103.96 (12.98) | RRS, ERQ-CA (adolescents); ERC (parents) | Post-intervention, adolescents reported improvements in the RRS reflection subscale, which were maintained at the 10-week follow-up. Self-reported effects were observed on the ERQ-CA cognitive reappraisal subscale at the 1-week follow-up, and parents reported effects on the ERC emotion regulation subscale at 10-week follow-up. Internal consistency ranged from .81 to .90 for the RRS, .84 to .85 for the ERQ-CA and .75 to .85 for the ERC. [Selection and Implementation stages] |
| [73] Santomauro et al. (2016)1 | Australia |
Experimental ASD group: N = 10 (n/a); waitlist ASD group: N = 10 (n/a) |
Experimental ASD group: 16.00 (1.33); waitlist ASD group: 15.50 (1.43), no age-range reported |
Verbal IQ ≥ 85 | ERQ (adolescents) | No group differences were found on the cognitive reappraisal and suppression subscales. A time effect was observed for cognitive reappraisal with post-hoc tests indicating changes from baseline to post-intervention, and from post-intervention to 3-month follow up. Better theory of mind correlated with higher cognitive reappraisal scores. Internal consistency was .86 for cognitive reappraisal and .72 for suppression. [Selection] |
| [74] Santomauro et al. (2017)3 | Australia |
ASD group: N = 7 (6 M, 1 F); ASD group represented by parents: N = 9 (9 M); |
ASD group represented by parents: 15.43 (1.99), 13–19; ASD group: 20.71 (3.09), 14–23 | n/a | Focus groups and interviews (adolescents, parents, different professional groups) | The ASD group, caregivers and professional groups highlighted that ER difficulties significantly impact autistic adolescents’ lives. Key themes identified were: (1) triggers of distressing emotions (e.g. sensory sensitivities), (2) difficulties with emotional awareness (e.g. alexithymia and the need for external support), (3) ER strategies (both adaptive and dysfunctional strategies) and (4) consequences of distressing emotions (e.g. academic and social challenges). [Identification, Selection and Implementation] |
| [75] Shaffer et al. (2023)1 | USA | ASD group: N = 19 (n/a) | 14.80 (1.38), 13–18 | 96 (18.51), not specified by group (child and adolescents) | EDI, ABC-2-I, BRIEF-2 (caregivers) | The intervention showed an effect on the reactivity subscale, but not on the dysphoria subscale of the EDI. On the irritability subscale of the ABC-2, changes were observed across all time points. Additionally, the emotion regulation index of the BRIEF-2 indicated changes, both post-treatment and at the follow-up assessments. No information on internal consistencies. [Implementation] |
| [76] Skwerer et al. (2019)2 | USA | ASD group: N = 32 (22 M, 10 F) | 14.79 (1.9), 12–18 | NVIQ: 48.97(12.97) | EDI (caregivers) | Emotion dysregulation was not related to ASD traits, age group, NVIQ, or gender. Both reactivity and dysphoria scores were associated with the number of comorbid symptoms, while only dysphoria linked to generalized anxiety. Adaptive functioning scores did not correlate with EDI subscales; however, reactivity correlated with externalizing behaviors, while dysphoria showed no association with internalizing behaviors. In regression analysis, a composite of external and internal behaviors was predicted by the number of comorbid symptoms, but not by EDI scores. Internal consistency data were not reported. [Implementation] |
| [77] Timko et al. (2021)2 | USA |
ASD group: N = 29 (29 F) Anorexia nervosa (AN) group: N = 22 (22 F) |
ASD group: 12.30 (2.25), 10–17; AN group: 14.81 (2.33), 10–18 |
ASD group: 96.3 (24.69), 42–145 AN group: 105.84 (13.95), 83–142 |
BRIEF (caregivers) | The ASD group’s mean score on the shift subscale was in the clinical range (T > 65) compared to norms. A higher proportion of the ASD group scored in the clinical range on both the shift and emotional control subscales than the AN group. In the ASD group, the shift score predicted restricted repetitive behaviors in a linear regression analysis. Internal consistency data were not reported. [Implementation] |
| [78] Wieckowski et al. (2020)2 | USA | ASD group: N = 722 (576 M, 146 F) | 12.95 (3.41), 4–20 | NVIQ: 75.11 (28.23) | EDI (caregivers) | Autistic adolescents in specialized inpatient psychiatric units exhibited clinically significant emotion dysregulation relative to general population norms. Among those under age 13, no gender differences were observed in EDI scores; however, females aged 13 and older scored higher on the dysphoria subscale than males. Furthermore, being verbal correlated with higher dysphoria scores on the EDI. Internal consistency data were not reported. [Implementation] |
| [79] Woodcock et al. (2020)2 | UK | ASD group: N = 20 (16 M, 4 F); non-ASD group: N = 80 (64 M, 16 F) | ASD group: 13.3 (n/a), 11–17; non-ASD group: 13.3 (n/a), 10–17 | ASD group: 107.7 (18.1); non-ASD group: n/a |
CAMS, CSMS (adolescents); BRIEF (parents) |
In the ASD group, no correlations were found between self-reported anger and sadness regulation scores and decisions as proposers or responders in the ultimatum game. Average T-scores on the shift and emotional control subscales of the BRIEF were clinically elevated (> 65) relative to norms. Poorer emotional control was linked to lower acceptance rates of unfair offers in the ultimatum game. No information on internal consistencies. [Implementation] |
| [80] Yager et al. (2013)2 | Canada |
ASD group N = 22 (19 M, 3 F); non-ASD group; N = 22 (19 M, 3 F) |
ASD group: 14.17 (2.25), 11–18; non-ASD group: 14.12 (2.27), 11–18 | ASD group: 101.05 (10.53); non-ASD group: 100.18 (10.60) | MSCS (caregivers) | There was an association between the ER subscale of the MSCS and ASD traits. The ASD group exhibited lower scores in the ER subscale compared to the non-ASD group. Internal consistency was .89 for the ER subscale. [Implementation] |
ABC-I Aberrant Behavior Checklist—Irritability subscale, ABC-2-I Aberrant Behavior Checklist—Second Edition—Irritability subscale, ADHD Attention Deficit Hyperactivity Disorder, ASD Autism Spectrum Disorder, BRIEF Behavior Rating Inventory of Executive Function, specifically the subscales shift and emotional control, BRIEF-2 Behavior Rating Inventory of Executive Function—Second Edition, specifically the Emotional Regulation Index, BPD Borderline personality disorder, CAMS Children’s Anger Management Scale, CBCL-EDI The Child Behavior Checklist—Emotion Dysregulation Index, CERQ Cognitive Emotion Regulation Questionnaire, CSMS Children’s Sadness Management Scale, DERS Difficulties in Emotion Regulation Scale, EDI Emotion Dysregulation Inventory, ERC Emotion Regulation Checklist, ERICA Emotion Regulation Index for Children and Adolescents, ERQ Emotion Regulation Questionnaire, ERQ-CA Emotion Regulation Questionnaire for Children and Adolescents, ERS Emotion Regulation Scale, fMRI functional magnetic resonance imaging, F Female, HRV resting heart rate variability, ID Intellectual Disabilities, K-BDEFS Korean Version of Barkley Deficits in Executive Functioning Scale, specifically the factor regulation of emotion, M Male, MSCS Multidimensional Social Competence Scale, NVIQ Nonverbal IQ, Non-ASD Non-autistic adolescents, refers to adolescents with no history of neurological or psychological condition, if not otherwise stated, O other genders (including transgender, non-binary and agender), RRS Ruminative Response Scale, RSQ Response to Stress Questionnaire – Social stress version (RSQa only subscales involuntary engagement and voluntary engagement, RSQb only subscales involuntary engagement and involuntary disengagement). Information in square brackets in the 'Findings regarding ER' column indicates the stage(s) of the extended process model of emotion regulation to which the results predominantly pertain
1Interventional study; 2Cross-sectional study; 3Qualitative study