Atients (5 in number) were not included in the final analysis. The research assistants also administered the sociodemographic questionnaire and the AIDSrelated stigma scale to all consented patients. The PHQ-9 was administered as part of a validation study, the findings of which will be reported elsewhere. Research assistants abstracted the most recent CD4 count level from patient’s charts. All the questionnaires were translated from English into the local TA 01 site language (Luganda), and administered in either language, depending on which of the two was understood by the patient.Data AnalysisData was analysed using STATA 11.2 [44]. We initially AN-3199 chemical information compared individuals with and without major depressive disorder using bivariate analysis; variables that were statistically significant at bivariate analysis were then entered into a multivariable model following a stepwise hierarchical format. Logistic regression was used to test for significant associations, controlling for age and gender. Major depressive disorder was the dependent variable, with AIDS related stigma, immune status and sociodemographic variables as independent variables. We also assessed whether the sociodemographic and clinical variables were confounding the association between AIDS related stigma and Major depression.ResultsThe mean age of the respondents was 38.8 years (SD 9.81; range 18?1). Of the 368 participants, 72 were female. About two thirds (65.5 ) had some form of employment, and onlyAids, Stigma, Depressive Disorder, UgandaTable 1. Sociodemographic and clinical characteristics of the respondents.Table 2. Bivariate analysis of the sociodemographic and clinical characteristics of the respondents.Variable Major depression Yes No Gender Male Female Age 18?9 30?9 40?9 50+ Age(Mean(sd)) AIDS stigma Yes No Marital status Not married Married Employment Employed Not employed Education status Primary education Secondary Education CD4 counts ,200 .201 CD4(Mean(sd)) Median Opportunistic Infection Yes No WHO stage I II III IV doi:10.1371/journal.pone.0048671.tFrequency( )VariableDepressed (frequency, )Not depressed (frequency, )P-value64(17.39 ) 304(82.61 )Gender Male Female 18 (16.82) 46(17.62) 89(83.18) 215(82.38) 0.107(29 ) 261(71 )Age 18?9 30?9 17(26.15) 27(18.75) 14(12.96) 6(11.76) 48 (73.85) 117(81.25) 94(87.04) 45(88.24) 0.068 0.014* 0.034*65(17.7 ) 144(39.1 ) 108(29.3 ) 51(13.8 ) 38.84(10.08)40?9 50+ AIDS stigma Yes No Marital status10 (34.48) 54(15.93)19(65.52) 285(84.07)0.002*29(7.88 ) 339(92.1 )Not married Married Employment37 (16.16) 27(19.42)192(83.84) 112(80.58) 0.229(62.2 ) 139(37.8 )Employed Not employed Education level34(14.11) 30(23.62)207(85.89) 97(76.38) 0.023*241(65.5 ) 127(34.5 )Primary Secondary CD4 counts39(22.94) 25(12.63)131(77.06) 173(87.37) 0.010*170(46.2 ) 198(53.8 )#200 201 Opportunistic16 (26.23) 48(15.64)45(73.77) 259(84.36) 0.049*61(16.6 ) 307(83.4 ) 328.22(182.36) 336 IQR 25?5 (250?56)Infection Yes No WHO stage I 13(14.13) 26(19.26) 15(15.79) 10(24.39) 79 (85.87) 109(80.74) 80(84.21) 31(75.61) 0.316 0.751 0.153 19(12.34) 45(21.03) 135(87.66) 169(78.97) 0.032*154(41.8 ) 214(58.2 )II III IV92(25 ) 135(36.7 ) 95(25.8 ) 41(11.1 )*denotes p value ,0.05 and statistical significance. doi:10.1371/journal.pone.0048671.t(53.8 ) had attained at least secondary education. Mean CD4+ was 329.70, (SD 183.95; range 1?99). Median CD4+ was 336, inter-quartile range 25?5 (250?56). The prevalence of major depressive disorder was 17.4 . Twenty nine out of the 368 part.Atients (5 in number) were not included in the final analysis. The research assistants also administered the sociodemographic questionnaire and the AIDSrelated stigma scale to all consented patients. The PHQ-9 was administered as part of a validation study, the findings of which will be reported elsewhere. Research assistants abstracted the most recent CD4 count level from patient’s charts. All the questionnaires were translated from English into the local language (Luganda), and administered in either language, depending on which of the two was understood by the patient.Data AnalysisData was analysed using STATA 11.2 [44]. We initially compared individuals with and without major depressive disorder using bivariate analysis; variables that were statistically significant at bivariate analysis were then entered into a multivariable model following a stepwise hierarchical format. Logistic regression was used to test for significant associations, controlling for age and gender. Major depressive disorder was the dependent variable, with AIDS related stigma, immune status and sociodemographic variables as independent variables. We also assessed whether the sociodemographic and clinical variables were confounding the association between AIDS related stigma and Major depression.ResultsThe mean age of the respondents was 38.8 years (SD 9.81; range 18?1). Of the 368 participants, 72 were female. About two thirds (65.5 ) had some form of employment, and onlyAids, Stigma, Depressive Disorder, UgandaTable 1. Sociodemographic and clinical characteristics of the respondents.Table 2. Bivariate analysis of the sociodemographic and clinical characteristics of the respondents.Variable Major depression Yes No Gender Male Female Age 18?9 30?9 40?9 50+ Age(Mean(sd)) AIDS stigma Yes No Marital status Not married Married Employment Employed Not employed Education status Primary education Secondary Education CD4 counts ,200 .201 CD4(Mean(sd)) Median Opportunistic Infection Yes No WHO stage I II III IV doi:10.1371/journal.pone.0048671.tFrequency( )VariableDepressed (frequency, )Not depressed (frequency, )P-value64(17.39 ) 304(82.61 )Gender Male Female 18 (16.82) 46(17.62) 89(83.18) 215(82.38) 0.107(29 ) 261(71 )Age 18?9 30?9 17(26.15) 27(18.75) 14(12.96) 6(11.76) 48 (73.85) 117(81.25) 94(87.04) 45(88.24) 0.068 0.014* 0.034*65(17.7 ) 144(39.1 ) 108(29.3 ) 51(13.8 ) 38.84(10.08)40?9 50+ AIDS stigma Yes No Marital status10 (34.48) 54(15.93)19(65.52) 285(84.07)0.002*29(7.88 ) 339(92.1 )Not married Married Employment37 (16.16) 27(19.42)192(83.84) 112(80.58) 0.229(62.2 ) 139(37.8 )Employed Not employed Education level34(14.11) 30(23.62)207(85.89) 97(76.38) 0.023*241(65.5 ) 127(34.5 )Primary Secondary CD4 counts39(22.94) 25(12.63)131(77.06) 173(87.37) 0.010*170(46.2 ) 198(53.8 )#200 201 Opportunistic16 (26.23) 48(15.64)45(73.77) 259(84.36) 0.049*61(16.6 ) 307(83.4 ) 328.22(182.36) 336 IQR 25?5 (250?56)Infection Yes No WHO stage I 13(14.13) 26(19.26) 15(15.79) 10(24.39) 79 (85.87) 109(80.74) 80(84.21) 31(75.61) 0.316 0.751 0.153 19(12.34) 45(21.03) 135(87.66) 169(78.97) 0.032*154(41.8 ) 214(58.2 )II III IV92(25 ) 135(36.7 ) 95(25.8 ) 41(11.1 )*denotes p value ,0.05 and statistical significance. doi:10.1371/journal.pone.0048671.t(53.8 ) had attained at least secondary education. Mean CD4+ was 329.70, (SD 183.95; range 1?99). Median CD4+ was 336, inter-quartile range 25?5 (250?56). The prevalence of major depressive disorder was 17.4 . Twenty nine out of the 368 part.