BackgroundRefugees and migrants are widely recognized for their considerable resilience in the face of traumatic events and adversities. However, epidemiological studies report an elevated risk of developing mental health problems among them. Despite this high prevalence, refugees and migrants underutilize mental health services. This underutilization is associated with both structural and cultural barriers. These include lack of language proficiency, lack of awareness of available services, negative attitudes toward professional help, and stigma. Culturally adapted internet-based psychological interventions are proposed as a possible way of overcoming some of the mentioned barriers.AimThe overarching aim of this thesis is to develop and evaluate a culturally adapted internet-based cognitive behavioral therapy (CaiCBT) targeting Arabic-speaking refugee and migrant youth and young adults with mild to moderate mental health problems. This was achieved by, firstly, exploring what Arabic-speaking refugee youth and young adults conceptualize as their mental health problems and in what way these problems impact functioning (Study I). Secondly, evaluating the feasibility of a culturally adapted internet-based cognitive behavioral therapy for mild to moderate mental health problems (Study II) and the efficacy of the intervention in a randomized clinical trial (Study IV). Thirdly, reporting the cultural adaptation process according to the RECAPT criteria (Study III). Finally, conducting an experimental study examining whether texts produced by generative AI can be perceived as culturally relevant and acceptable (Study V).MethodsThe thesis was guided by both qualitative and quantitative methods. Study I was a qualitative content analysis exploring the conceptualization of mental health problems among symptomatic refugee youth and young adults. The impact of the problems on functioning was categorized according to Alderfer's Existence- Relatedness-Growth Model. Study II was a feasibility study including qualitative interviews, which were analyzed using thematic analysis. Study III reported the cultural adaptation of Raha using the RECAPT criteria. Study IV was a randomized controlled trial involving 125 participants. Data were analyzed using a latent growth model. Study V utilized two-factor between-subjects analysis of variance to analyze the data.ResultsStudy I identified three main domains of mental health problems among young Arabic-speaking symptomatic refugees: challenges within the self, difficulties in exile, and challenges in the encounter with 'the other'. The identified problems impacted the youths' ability to fulfill existence and relatedness needs. No growth needs were identified, suggesting frustration-regression. Qualitative interviews in Study II revealed mixed experiences about the intervention, with themes including acceptability of the platform, beneficial therapist support components, and the need for format improvements such as video calls and increased awareness-raising about mental health. Study III documented the cultural adaptation process of Raha following RECAPT criteria, targeting Arabic-speaking refugee and migrant youth and young adults aged 15-29 with mild to moderate mental health problems. Adaptations included adding relationship and resilience modules, incorporating cultural concepts of distress, using WhatsApp for communication, and developing culturally relevant case examples. In Study IV, the latent growth curve model revealed a significant group-by-time interaction with a large between-group effect size (Cohen's d = 1.43, 95% CI [0.97, 1.89]). On average, the treatment group improved by 0.6 points per week on the primary outcome measurement, indicating an alleviation of symptoms of depression and anxiety. In the intervention condition, 57.7% achieved remission compared to 14.3% in the control condition, and 65.4% achieved clinical reduction compared to 16.1% in controls. Significant effects were observed for subjective problem areas, post-traumatic stress, insomnia, resilience, and well-being, sustained at 6- month follow-up. Study V examined AI-generated versus human-generated culturally adapted CBT texts among 128 raters. For cultural relevance, a significant main effect favoring AI was found in the first rating, but not the second rating. No significant differences were found for acceptability ratings between AI and human-generated texts.ConclusionThis thesis demonstrates that culturally adapted internet-based cognitive behavioral therapy (CaiCBT) is both effective and acceptable for Arabic- speaking refugee and migrant youth and young adults with mild to moderate mental health problems. The iterative cultural adaptation process, guided by RECAPT criteria, resulted in improvements in feasibility and outcomes. Furthermore, the finding that AI-generated therapeutic texts are perceived as equally culturally relevant and acceptable as human-generated texts opens promising opportunities for scaling culturally adapted psychological interventions. Taken together, these findings support culturally adapted internet- based interventions as a viable approach to reducing mental health disparities among refugee and migrant populations.List of scientific papersI. Demetry Y., Kanabi S., Meurling J., Amiri M.A., Andersson G., Shahnavaz S., 'A labyrinth with no way out' - The conceptualization of mental distress among Arabic-speaking refugee youth: A qualitative study, [Submitted]II. Demetry Y., Wasteson E., Lindegaard T. Abuleil A., Geranmayeh A., Andersson G., & Shahnavaz S., (2023). Individually tailored and culturally adapted internet-based cognitive behavioral therapy for Arabic-speaking youths with mental health problems in Sweden: Qualitative feasibility study, JMIR Formative Research, 7(1), e46253. https://doi.org/10.2196/46253III. Demetry Y., Geranmayeh A., Amiri M.A., Lindegaard T., Wasteson E., Abuleil A., Kutabi H., Shahnavaz S., & Andersson G., The design...