Participants were 116 female adolescents (ages 13–20 years, M = 16.01; SD = 1.64). The sample included 45 adolescents with NSSI disorder, 27 adolescents with other mental disorders without NSSI (clinical controls, CCs), and 44 adolescents without current or past experience of mental disorders (nonclinical controls, NCs). Participants were similar with respect to age, F(2, 112) = 2.93, p > .05.
All adolescents were diagnosed using the Diagnostic Interview for Mental Disorders in Children and Adolescents (Kinder-DIPS) , a structured interview in German based on the DSM-IV-TR criteria .
The mean number of diagnoses was 3.36 (SD = 1.42) for adolescents with NSSI and 2.07 (SD = 0.92) for CC adolescents, which is a significant difference, t(70) = 7.27, p < .01. The most frequent diagnosis among adolescents with NSSI and CC adolescents was major depression, followed by social phobia. Posttraumatic stress disorder was diagnosed more often in the NSSI group (n = 10, 22.2 %) than in the CC group (n = 2, 7.4 %), and borderline personality disorder (n = 7, 15.6 %) and alcohol abuse (n = 2, 4.4 %) emerged only in the NSSI group.
A total of 116 parents including 92 mothers (ages 36–57 years, M = 45.67; SD = 4.91) and 24 fathers (ages 44–58 years, M = 48.74; SD = 3.13) participated. Participating fathers were significantly older than participating mothers, F(1, 103) = 7.79, p < .01. Parents’ education was assessed with the following scale: 0 (did not finish school), 1 (obligatory school), 2 (vocational training), 3 (Matur; slightly higher than a high school diploma), 4 (professional training), and 5 (university degree). Mothers’ mean education was 2.52 (SD = 1.23) in the NSSI group, 2.26 (SD = .87) in the CC group, and 3.12 (SD = 1.27) in the NC group, with a significant difference between the groups, F(2, 82) = 3.83, p < .05. Post hoc analyses indicated that this difference emerged between the CC and NC group. Fathers’ mean education was 4.00 (SD = .87) in the NSSI group, 4.75 (SD = .50) in the CC group, and 3.40 (SD = 1.51) in the NC group, with no significant difference between the groups, F(2,22) = 2.01, p > .05. The families’ average monthly income was assessed using a scale ranging from 1 (less than 2,000 Swiss francs per month) to 6 (more than 10,000 Swiss francs per month), with 2 = 2,000–4,000 and 3 = 4,001–6,000 Swiss francs per month. The mean income was 2.70 (SD = 1.45) in the NSSI group, 2.27 (SD = 1.03) in the CC group, and 2.23 (SD = 1.22) in the NC group, with no significant difference between the groups, F(2,82) = 1.26, p = .29.
The recruitment took place in Switzerland and Germany. The two clinical groups were recruited from different inpatient child and adolescent psychiatric units and the NC group from different schools. The inpatient clinics were responsible for the recruitment of the clinical groups. Therefore, we have no access to the demographic and clinical characteristics of patients excluded by the clinics. Our predefined exclusion criteria were current or past psychosis, schizophrenic symptoms, and acute substance abuse. The inpatient clinics were instructed to inform the participants at admission; in most cases it was not the therapist who did so. Adolescents and parents gave their written consent. The institutional review board (Ethikkommission beider Basel, EKBB) approved the study. Questionnaires were administered to the participating adolescents (Zurich Short Questionnaire on Parental Behavior, ZKE) and their parents (Depression Anxiety Stress Scale-21, DASS-21; Parental Stress Scale, PSS; Zurich Short Questionnaire on Parental Behavior, ZKE). The adolescents were paid 40 Swiss francs for participation.
Assessment of Axis I and Axis II diagnoses
To examine current and past DSM-IV-TR diagnoses a structured interview for mental disorders in children and adolescents  was conducted with each adolescent. The Kinder-DIPS assesses the most frequent mental disorders in childhood and adolescence, including anxiety disorders, depression, attention-deficit/hyperactivity disorder, conduct disorder, sleep disorders, and eating disorders. The interview has good validity and reliability [29, 30]. NSSI disorder was assessed with an interview using the DSM-5 criteria. The estimates of interrater reliability for the diagnosis of NSSI are very good (κ = 0.90) . Questions about triggers for NSSI were part of the sociodemographic questionnaire. Substance use disorder and borderline personality disorder were examined with the adult DIPS . Axis II personality disorders were obtained with the Structured Clinical Interview for DSM-IV Axis II Personality Disorders (SKID-II) .
Depression Anxiety Stress Scale-21 (DASS-21)
This 21-item questionnaire assesses depression, anxiety, and stress symptoms . Participants rate the frequency and severity of the symptoms over the last week on a 4-point Likert scale. The DASS-21 has a good internal consistency and convergent and discriminant validity . The internal consistency in the present sample was α = 0.92 for the depression scale, α = 0.86 for the anxiety scale, α = 0.86 for the stress scale, and α = 0.95 for the total scale.
Parental Stress Scale (PSS)
This instrument assesses parent satisfaction . It contains items representing positive themes of parenthood such as emotional benefits or self-enrichment and negative components such as demands on resources and restrictions. The questionnaire consists of the four subscales parental rewards, parental stressors, lack of control, and parental satisfaction. The PSS has satisfactory levels of internal consistency and convergent and discriminant validity . The internal consistency in the present sample was α = 0.76 for parental rewards, α = 0.51 for parental stressors, α = 0.68 for lack of control, and α = 0.59 for parental satisfaction.
The Zurich Short Questionnaire on Parental Behavior (ZKE)  assesses three aspects of parenting behavior from the parents’ and children’s perspective. Adolescents complete the questionnaire once for their mother and once for their father. The ZKE measures warmth and support, psychological pressure, and behavioral control (demands, rules, and discipline). The questionnaire demonstrated good psychometric properties. The internal consistency in the present sample was α = 0.93 for the subscale warmth and support, α = 0.88 for the subscale psychological pressure, and α = 0.72 for the subscale behavioral control.
Data were checked to insure that they met the assumptions for the analyses; no violations of assumptions were detected. We used multivariate analysis of variance (MANOVA) to investigate group differences in parenting behavior, parental psychopathology, and parental stress between the groups. Post hoc tests were conducted to analyze pairwise comparisons (NSSI vs. CC, NSSI vs. NC, and CC vs. NC). The Bonferroni–Holm correction was used to control for multiple comparisons. Effect sizes (Cohen’s d) are used to report differences between the groups. An effect size of 0.20 equates to a small effect, 0.50 to a medium effect, and 0.80 to a large effect. Parent–child agreement regarding parenting behavior was evaluated by calculating Pearson product–moment correlation coefficients. To compare correlations the coefficients were converted to z scores. Analyses were performed using SPSS version 21. Significance levels were set at α = 0.05.