E-Mail karger@karger.com Letter to the Editor of the Social Adjustment Scale-Self-Report (SAS-SR) [7] consist- ing of four subscales: (a) School (α = 0.77), (b) Friends (α = 0.81), (c) Family (α = 0.83), and (d) Overall Adjustment (α = 0.80). To longitudinally assess changes in study outcomes, mixed lin- ear regression models were fitted that account for multiple mea- surements per individual obtained at different time points. All analyses were conducted using a random coefficient model with the intercept being random and a covariance structure of variance components. Possible interactions of time with type of treatment were tested via regression models. Results regarding changes in CDI, CFSEI and SAS-SR scores as indicated from mixed-model regression analyses are shown in table 1. Similarly with previous studies [8–10], independently of the type of treatment, significant improvements in depression symptoms, self-esteem and social ad- justment were found from BL to EOT and from BL to FU. Chang- es from EOT to FU were significant for depression symptoms and self-esteem. No significant interaction of time with type of treat- ment was found for CDI and CFSEI, indicating that the two ap- proaches were equally effective in improving depression symp- toms and self-esteem. As far as SAS-SR is concerned, a significant interaction was found only for adjustment with friends, indicating that the effect of treatment increased further in patients that re- ceived PP. The overall change of CDI score in the PP group was signifi- cantly associated with changes in SAS-SR Friends (r = 0.49, p = 0.003), School (r = 0.44, p = 0.009), and Overall Adjustment (r = 0.52, p = 0.001) scales. On the other hand, change of CDI score in the FT group was significantly associated with change in the CFSEI scale (r = –0.53, p = 0.001). One possible explanation for this dis- crepancy may be that individual PP focuses on attachment-related issues, the relationship with the therapist and on the transference interpretation of these phenomena in the here and now. This fact could lead to modifications in attitudes and behaviours towards the therapist at first and subsequently to relationships with others (peers, parents, siblings, significant others). On the other hand, FT focuses on family issues (e.g. communication, relationships, roles, etc.) and not on the individual child per se. Thus, the depressed child is relieved from the burden of being the identified patient, the self-understanding for each member is enhanced, and the child starts to experience the family as more supportive and warm or al- lowing more autonomy and independence, as well as to perceive himself/ herself in a more realistic and positive way. Moreover, linear regression analyses showed that patients with co-morbidity had greater changes in the SAS-SR Friends subscale (β = –1.65, SE = 0.46, p = 0.001) and SAS-SR Overall Adjustment scale (β = –0.75, SE = 0.34, p = 0.034) compared to those without co-morbidity. Also, past episode of depression and/or dysthymia had a significant effect on changes in the Overall Adjustment scale (β = –0.81, SE = 0.29, p = 0.008). More difficult cases were substan- On the basis of available evidence, the effectiveness of psycho- logical and pharmacological interventions for youths’ depression cannot be established and further randomized controlled trials with larger sample sizes are required [1]. This is particularly true in the case of psychodynamic psychotherapy (PP) [2]. The aim of this study was to build on a previously reported multicentre ran- domized trial with depressed youths receiving PP or family thera- py (FT) [3], by exploring the role of self-esteem and social adjust- ment in the treatment outcome. Both these dimensions have been found to enable youths to cope adequately with stressful life situa- tions and act as protective factors against vulnerability to depres- sion [4]. The trial was conducted in London, Athens and Helsinki, with 72 patients aged 9–15 years randomly allocated to PP (n = 35) or FT (n = 37). Treatment was conducted over a 9-month period and consisted of: (i) 8–14 (mean = 11) 90-min sessions of systems in- tegrated FT with a focus on family dysfunction, but without spe- cific attention to unresolved intrapsychic conflicts and early child- hood, or (ii) 16–30 50-min sessions (mean = 24.7) of focused indi- vidual PP with a focus on interpersonal relationships, life stresses and dysfunctional attachments plus parent sessions (one per two child sessions), focused on the same areas, by a separate case work- er. Assessment took place at baseline (BL), at the end of treatment (EOT), and at 6 months’ follow-up (FU). All assessments were done by researchers who were not involved in treatment imple- mentation. Further details on the participants and procedures of the study are reported elsewhere [3]. Depressive symptoms were measured by the Children’s De- pression Inventory (CDI) [5]. The Cronbach α coefficient for the CDI in the present sample was 0.90. Self-esteem was measured by the Culture-Free Self-Esteem Inventory (CFSEI)-Form B [6], with an α of 0.76. Social adjustment was measured by a modified version Received: August 28, 2013 Accepted after revision: December 24, 2013 Published online: June 24, 2014 © 2014 S. Karger AG, Basel 0033–3190/14/0834–0249$39.50/0 www.karger.com/pps Psychother Psychosom 2014;83:249–251 DOI: 10.1159/000358289 Self-Esteem and Social Adjustment in Depressed Youths: A Randomized Trial Comparing Psychodynamic Psychotherapy and Family Therapy Gerasimos Kolaitis a , George Giannakopoulos a , Vlasis Tomaras a , Stelios Christogiorgos a , Valeria Pomini a , Effie Layiou-Lignos a , Chara Tzavara a , Maria Rhode b , Gillian Miles b , Ilan Joffe b , Judith Trowell b , John Tsiantis a a Department of Child Psychiatry, University of Athens Medical School, Aghia Sophia Children’s Hospital, Athens, Greece; b Tavistock Clinic, London, UK Dr. George Giannakopoulos Department of Child Psychiatry University of Athens Medical School 11527 Athens (Greece) E-Mail giannakopoulos.med  @  gmail.com