Addi van Bergen

174 Table S1b: Description of observational studies on the association between SE/SI and mental health in high risk groups Reference Setting Population Sample size Indicators SE/ SI * Health measure ** Statistical analysis Case control study Flores et al. [42] Spain Adult men from various risk settings 105 S: family contact E: income level P: habitual domicile C: source of income (legal, illegal, work) SEPC: excluded on all 4 dimensions Outcome MH: personality features by DSM- III (9 scales) Mann-Whitney U Test comparing case groups with control (no AIDS, no drug addiction, no SE) Todd et al. [49] England Clients of Mental Health Services (MHS) and Drug & Alcohol Services (DAS) 590 Outcome: S: isolation E: employment P1: homelessness (2) P2: education C: contact with criminal justice system (4) MH: comorbidity of psychiatric and substance misuse disorders Conditional logit model Webber & Huxley [39] England Persons assessed for compulsory hospitalization 300 S: social support E1: income, E2: employment P1: insecure housing; P2: education; P3: neighbourhood deprivation a) SEP: 3 or more indicators above mean Outcome: MH: emergency compulsory hospitalization Uni and multivariate logistic regression MH: compulsory hospitalization Cross-sectional study Choi et al. [40] South Korea Torture survivors 206 S: exclusion by family and a cquaintances and not being able to reveal torture experiences (5) P: no support or help by government / institutions (1) SP: average rating on 6 items b) Post-traumatic stress disorder c) Hierarchical regression analysis Depression Anxiety Hostility Somatisation Interpersonal sensitivity