Discrimination reported by people with major depressive disorder--authors' reply.

Siddharth Sarkar and Rishab Gupta have been somewhat selective by mentioning the correlation between education and discrimination, since this association was found in univariate analyses only, but lost significance when variables were entered in a multivariate regression model.1Lasalvia A Zoppei S Bortel TV et al.Global pattern of experienced and anticipated discrimination reported by people with major depressive disorder: a cross sectional survey.Lancet. 2013; 381: 55-62Summary Full Text Full Text PDF PubMed Scopus (277) Google Scholar It should also be noted that the association found by Brohan and colleagues2Brohan E Gauci D Sartorius N Thornicroft G Self-stigma, empowerment and perceived discrimination among people with bipolar disorder or depression in 13 European countries: the GAMIAN-Europe study.J Affect Disord. 2011; 129: 56-63Summary Full Text Full Text PDF PubMed Scopus (182) Google Scholar was in the opposite direction to that suggested by Sarkar and Gupta, since higher education was significantly associated with lower self-stigma, consistent with other studies.3Yen CF Chen CC Lee Y et al.Self-stigma and its correlates among outpatients with depressive disorders.Psychiatr Serv. 2005; 56: 599-601Crossref PubMed Scopus (96) Google Scholar, 4Werner P Stein-Shvachman I Heinik J Perceptions of self-stigma and its correlates among older adults with depression: a preliminary study.Int Psychogeriatr. 2009; 21: 1180-1189Crossref PubMed Scopus (26) Google Scholar Our findings, therefore, still do not help to shed new light on this association, and so we are unwilling to accept that “ignorance is bliss”. Regarding the possible effect of comorbidity, we have discussed in our paper that this is indeed one of the study limitations that should be carefully addressed by future research. We could not provide a comprehensive account of cross-cultural validation of the Discrimination and Stigma Scale 12 (DISC-12), but full details are given elsewhere.5Brohan E, Clement S, Rose D, Sartorius N, Slade M, Thornicroft G. Development and psychometric validation of the discrimination and stigma scale (DISC-12). Psychiatry Res (in press).Google Scholar In brief, the scale was created by drawing on the experiences of discrimination as described directly by service users in the 27 national sites for our first study, and was then translated into the local languages using a forward and back translation approach. A focus group with local people was subsequently convened at each site, to make sure that the scale was comprehensible. A training manual was used by all sites to achieve consistent interpretation of all the scale items. In the main study, five of the interviews at each site were recorded verbatim, translated into English, and qualitatively analysed; the results provided strong support for the choice of the DISC-12 items. Adding a specific example for every type of discrimination provided data that validate for the occurrence, direction, and severity of the discrimination that was rated quantitatively. The concerns raised by Philipe de Souto Barreto can be summarised as “Did the conclusions remain the same when all variables were fitted in the model?” Our statistical analysis section seems to contain, in his view, two suspicious steps in the selection of the independent variables: (1) three potential explanatory variables were found to be highly correlated with three others and so were sacrificed in the regression modelling procedure; and (2) only potential explanatory variables significantly associated with the dependent variable in the univariable regression models were introduced in the multivariable one. We shall first discuss this second step. The final reduced multivariable model comes after the estimation of the full model, in which all 11 independent variables were entered into the model irrespective of significance. The findings remained consistent with the preceding ones. In turn, the full model derived from a complete model with all the 14 available predictors. Indeed, since years from first psychiatric treatment and lifetime number of episodes of depression, compulsory and voluntary admission, and living conditions and marital status were highly correlated, they conveyed essentially the same information and neither might contribute significantly to the model after the other one was included. The choice of which variables to drop could be based on the suggested variance inflation factor scores or on hypotheses to be tested: these variables were ignored because they did not seem conceptually essential to the hypotheses-driven modelling. We declare that we have no conflicts of interest. Discrimination reported by people with major depressive disorderAntonio Lasalvia and colleagues (Jan 5, p 55)1 did a multisite international study to assess the global pattern of experienced and anticipated discrimination. One of the most interesting findings of the paper is that lower education was associated with less stigma. The correlation stood out at a trend-level significance even in the multivariate analysis. Full-Text PDF Discrimination reported by people with major depressive disorderAntonio Lasalvia and colleagues1 show that almost four out of five people with major depressive disorder experience discrimination and that anticipated discrimination is not rare. However, the description of their statistical approach seems incomplete, which raises the question of its appropriateness. The most important concerns are listed below. Full-Text PDF Global pattern of experienced and anticipated discrimination reported by people with major depressive disorder: a cross-sectional surveyDiscrimination related to depression acts as a barrier to social participation and successful vocational integration. Non-disclosure of depression is itself a further barrier to seeking help and to receiving effective treatment. This finding suggests that new and sustained approaches are needed to prevent stigmatisation of people with depression and reduce the effects of stigma when it is already established. Full-Text PDF