Exploring the relationship between internalized stigma, insight and depression for inpatients with schizophrenia

International audienceBACKGROUND: Recent studies on insight in people with schizophrenia showed that insight level is linked with treatment compliance. Therefore, many therapies are aimed at increasing the insight level, such as psycho-education. However, insight level is also probably linked with depression level. So, improving insight is at risk of increasing the level of depression. Nevertheless, results on this topic are not consensual in the scientific literature. Presumably, this could be due to the concept of insight itself, although we could hypothesise that some confounding variables are implied in the interaction between insight and depression, such as internalized stigma. AIMS: to test the hypothesis that the relationship between insight and depression is mediated by internalized stigma in people with schizophrenia. METHOD: Sixty-two patients with schizophrenia (DSM-IV or ICD-10) recruited in mental health services in Île de France (75% male), aged 20 to 64 years (m=38.71, σ=0. 43), filled in questionnaires assessing internalized stigma (ISMI), depression (CDSS and BDI) and insight (SAIQ, Q8, IS), after giving their written informed consent. Correlations between insight, depression and different variables were made (Hypothesis 1). Then we ran multiple regressions and partial correlations to test the internalized stigma mediation (Hypothesis 2). RESULTS: Insight, internalized stigma and depression are statistically significantly correlated with each scales used (except Q8). Insight is correlated with depressed mood (correlations between IS and CDSS: r=0.27, P=0.04, and BDI, r=0.40, P=0.001). We also found negative correlations between SAIQ and CDSS (r=-0.35, P=0.005) and the BDI (r=-0.4265, P=0.000) which means that good insight is linked with depression. This result validates our hypothesis 1. The statistic tests reveal other complementary results: the association between insight and depression is mediated by the level of internalized stigma: when ISMI is controlled, the correlation between insight and depression decreased moderately with CDSS and with small intensity with SAIQ, but clearly. So, ISMI is probably a mediating variable between IS and BDI-CDSS. In conclusion, internalized stigma could be a mediating variable between insight and depression. This validates our second hypothesis. CONCLUSIONS: Our results suggest that the relationship between insight and depression is mediated by internalized stigma. Patients with good insight who internalize stigma seem to be more depressed than those who don't. This result could have important consequences in clinical practice: improving insight level should be completed by a specific attention to the level and evolution of internalized stigma to avoid increasing depression. Further studies need to be conducted to confirm these results