The Alfred Hospital Depression in People Living with HIV/AIDS: Outcomes, Risks and Opportunities for Intervention Final Report August 2005 Chief Investigator Associate Professor Anne Mijch Infectious Diseases Department The Alfred Hospital This research project was funded by beyondblue
Final Report - A comprehensive report on whether the undertaking of the Project Services was in support of the Objectives and if not, why not - Appendix A; - A comparison between the achieved outcomes of the Project against the Objectives - Appendix A; - Where applicable, recommendations on how any Objectives that were not achieved could be achieved in the future: A study aim which was not met was to investigate neurocognitive performance in those who developed incident depression. This analysis was not performed as participant numbers were too low in the incident depression group (those who developed depression between baseline and follow-up, n = 10) for meaningful statistical analysis. For this Objective to be achieved in the future, a longer recruitment period would be required to allow for greater participant numbers. Also, an extended follow-up period would also benefit this Objective by allowing greater time periods for the development of incident depression. - Any statistics collected in the course of the Project - Appendix C. Conference Presentations and Publications Gibbie, T., Mijch, A., Ellen, S., Hoy, J., Hutchison, C., Wright, E., Chua, P. & Judd, F. (accepted). Depression and neurocognitive performance in individuals with HIV/AIDS: two year follow-up. HIV Medicine. Gibbie, T., Hay, M., Hutchison, C., & Mijch, A. (submitted). Depression, Social Isolation and Adherence to HAART We will present Depression, Social Isolation and Adherence to HAART at the 17 th Annual Conference of the Australasian Society for HIV Medicine (ASHM), Hobart, August 25 th 2005. with HIV/AIDS: Outcomes, Risks and Opportunities for Intervention Page 2 of 10
Appendix A Project Objectives 1. To examine the relationship between depression and quality of life, specifically whether depression is associated with increased at-risk behaviours, and whether depression is associated with poor adherence to treatment. 2. To examine the outcome of treatment (both HIV & depression) in individuals who are depressed by following them over time. 3. To determine the relative risk of developing neurocognitive defects and HIV dementia with and without incident depression or depressive symptoms. Achieved Outcomes In order to meet the Project Objectives, 80 HIV positive participants and 20 HIV negative participants were recruited for the follow-up phase of this study. Each participant completed a battery of questionnaires and neuropsychological tests. Each aim of the study and the Project Objectives has been successfully addressed and will be reported in turn: Depression in people living with HIV/AIDS and the associations of depression, HIV illness indicators, treatments and outcomes: At both baseline and follow-up there was a high rate of depressive symptoms with one third of the group scoring 14 or more on the Beck Depression Inventory (BDI). In the clinical interview (SCID), 27% of the group were found to have depression at baseline. On the whole, depression levels decreased between baseline and follow-up, with 14% of participants found to have depression diagnosed through clinical interview at follow-up (an average of two years later). There was no association between symptoms of depression and severity of HIV infection (such as CD4 cell counts, viral load and the duration of time since the diagnosis of HIV infection). In other words, those that had higher HIV viral loads or lower CD4 cell counts were not more likely to be depressed. Participants in close personal relationships had lower levels of depression than those who were not in a relationship. Similarly, people who were living alone reported significantly more symptoms of depression than those who lived with others. Higher levels of depression seemed to occur in those who were socially isolated and in poorer health. Depression in HIV positive participants and HIV negative controls and the relationship to increased at-risk behaviour. Research from the U.S.A, Europe and Australia has suggested men who have sex with men (msm) who use excessive drugs &/or alcohol are more likely to engage in at-risk sexual behaviour, increasing the likelihood of transmitting or contracting HIV and sexually transmitted infections (STI). The relationship between mental illness and at-risk behaviour is also under investigation, with early research suggesting increased high-risk sexual behaviour in people with mental illness. Many investigators have speculated on the links between depression and the spread of HIV infection. When depressed an individual may be less careful with regard to safe sex practice and risk behaviour. Depression and suicidality, often associated with indifference to health, may facilitate the extremely risky sexual behaviours and intravenous drug use practices reported by some of the most depressed subjects in surveys. with HIV/AIDS: Outcomes, Risks and Opportunities for Intervention Page 3 of 10
Objectives This analysis aimed to look for relationships between depression, drug and/or alcohol use and sexual risk behaviour in HIV positive msm. Method HIV positive participants (n = 80) completed questionnaires for depression (BDI), drug and alcohol use (frequency) and sexual risk behaviour (sex without a condom with a casual partner). Results Condom use with casual partners was not significantly correlated with depression scores or the amount of alcohol consumed on a weekly basis. However, illicit drug use was significantly correlated with unprotected anal sex with a casual partner (p <.05). Discussion These preliminary findings suggest HIV positive msm who use recreational illicit drugs are at-risk for HIV/STI transmission. Further investigation of these relationships is required to assess the nature of these relationships, which may then inform public health interventions. Depression in HIV positive participants and its relationship with poor adherence to Highly Active Antiretroviral Therapy (HAART). Estimates of the prevalence of depression in HIV-infected individuals range between 10% and 50%. Depression has been shown to have a negative impact on adherence to medication in chronic illness. Highly Active Antiretroviral Therapy (HAART) has resulted in a significant decline in HIV-related disease progression and mortality. To achieve the maximum benefit from HAART, patients must adhere to antiretroviral regimens for at least 95% of the time. These regimens are often complex and associated with significant sideeffects. Objectives The aim of this study was to investigate the prevalence of depression in HIV infected individuals compared to HIV negative controls and the association with depression and adherence to HAART. It was hypothesised that higher rates of depression would be found in people with HIV infection relative to controls, and that increased depression scores would be associated with reduced adherence to HAART. Method HIV positive (n = 80) and HIV negative (n = 20) participants were assessed for depression using the Beck Depression Inventory (BDI) and the Structured Clinical Interview for DSM-IV (SCID). Participants were identified as nonadherent if they reported less than 95% adherence to HAART (self-report). Results Overall, 14% of the HIV positive group met the criteria for SCID current mood disorder compared to 5% of controls. Similarly, 39% of HIV infected participants met the criteria for a past depressive episode compared to 15% of controls. HIV positive participants scored significantly higher on the BDI than controls (p <.05). Nonadherence to HAART was reported by 30.5% of those prescribed HAART. Multivariate analysis revealed that nonadherence to HAART was significantly associated with limited social support (living alone and not in a relationship) (p <.05) but not depression scores (p >.05). Discussion A higher prevalence of depression and past depressive episodes was identified in the group with HIV infection indicating the presence of compromised psychological health in people living with HIV infection. This study also found lower adherence to HAART in people living alone and not in a relationship. Health professionals should monitor and treat depression and should be particularly vigilant of adherence rates amongst those with limited social support. with HIV/AIDS: Outcomes, Risks and Opportunities for Intervention Page 4 of 10
To determine the relative risk of developing neurocognitive defects and HIV dementia with and without incident depression or depressive symptoms Objectives The aims of this study were to follow a cohort of HIV infected individuals for two years to: assess changes in depression and neuropsychological performance over time; explore the relationship between depression, HIV illness and neuropsychological performance; and to examine the natural history of Highly Active Antiretroviral Therapy (HAART) on depression and neurocognitive performance. Method HIV seropositive outpatients were assessed at baseline and at two-year follow-up. At each assessment patients were assessed for depression (using the Beck Depression Inventory (BDI) and Structured Clinical Interview (SCID-CV) and completed a battery of neuropsychological tests including the CANTAB and the Hopkins HIV Dementia Scale (HDS). Results Baseline results: 34.8% scored 14 on the BDI ( 14 suggests depressive symptoms (DS). The SCID-CV revealed 27% of participants met the criteria for current mood disorder. 7% of the participant s scores on the HDS indicated HIV associated cognitive changes. Follow-up results: 80 participants retested at two-year follow-up and were spilt into two groups based on BDI scores at baseline. CANTAB results revealed the cohort were significantly impaired on 9 of 10 measures compared with age-matched normative data. Neurocognitive performance significantly improved for participants with no DS at baseline, whereas participants with DS at baseline did not show as much improvement. Multivariate analysis revealed 40% of the change in cognitive performance was due to the variables age, AIDS and HAART regimens. Discussion These results suggest a significant decline in depression scores and an improvement in several neurocognitive domains overtime, with a relationship between HIV illness, HAART, symptoms of depression and neurocognitive performance. with HIV/AIDS: Outcomes, Risks and Opportunities for Intervention Page 5 of 10
Appendix C Statistics Table 1 HIV Illness Characteristics of HIV Seropositive Participants Variables HIV seropositive (n = 80) On HAART (%) 96% Mean time (years) HIV positive (SD) 13 (5) Previous AIDS diagnosis (%) 30% CD4 count cells/µl Median 522 Range 9-1325 25 th, 75 th percentile 311, 737 *HIV RNA viral load copies/ml Median 25 Range 25-100000 25 th, 75 th percentile 25, 2185 HIV RNA below detectability (%) 58% Note. * Undetectable viral load was calculated using the midpoint of the detectable limit of the assay (25 copies/ml). Table 2 Mean Depression Scores and Percent Diagnosed with Depression for HIV Seropositive (n = 80) and Seronegative Participants (n = 20) Variable HIV seropositive HIV seronegative BDI cognitive affective score Mean score (SD) 5.6 (5.3)* a 1.1 (2.5)* a SCID Depression Current (n) 14% (11) 5% (1) Lifetime (n) 39%* (31) 15%* (3) Note. *p < 0.05. a Levene s test of equality of variance not assumed p < 0.05. HIV seropositive participants had higher BDI scores and a greater prevalence of current and lifetime diagnoses of depression (Table 2). T-tests revealed that participants with HIV infection scored significantly higher (M = 5.6, SD = 5.3) than HIV negative controls (M = 1.1, SD = 2.5) on the BDI cognitive-affective items t(66) = 5.379, p =.000. There was no relationship between current depression diagnosis (SCID) and HIV status (χ 2 = 1.193, df = 1, p =.275). There was, however a significant difference between the groups on past depression (χ 2 = 4.159, df = 1, p =.041). with HIV/AIDS: Outcomes, Risks and Opportunities for Intervention Page 6 of 10
Intercorrelations between BDI cognitive affective scores and psychosocial variables were obtained for the HIV seropositive group. BDI scores correlated significantly with living situation (r =.31, p <.05) and relationship status (r =.27, p <.05). That is, increased BDI scores were significantly correlated with not being in a current relationship and living alone. Multiple regression analysis was undertaken to determine the independent influence of psychosocial factors and HIV illness on current diagnosis of depression for the HIV seropositive participants. Table 3 contains the results of this analysis. Using the enter method, the overall model was not significant for current depression (F 6,70 = 1.740, p > 0.05, adjusted R-squared =.059). Table 3 Multiple Linear Regression Analysis Predicting SCID Current Depression Diagnosis for HIV Seropositive Participants Current Depression Variable B SE t p HIV RNA -6.94.000-0.41.677 CD4 cell count 3.502.000 0.22.822 AIDS diagnosis 0.113.094 1.20.234 Living situation 0.210.106 1.98.051 Relationship status -2.24.100 0.22.823 Note. N = 80. It can be seen in Table 3 that no variables included in this model significantly predicted depression. Adherence to medication. Responses to the interviews regarding recent medication adherence patterns indicated that 30.5% (n = 22) of participants on HAART had completely missed at least one antiretroviral dose in the previous 7 days. Scores on depression measures, living situation and relationship status were used to model the likelihood of adherence. Table 4 Logistic Regression Analysis Predicting Adherence to HAART Regimens Variable B Wald p Exp(B) BDI -0.330 0.179.672 0.719 SCID current 1.048 0.843.359 2.851 SCID lifetime 0.084 0.018.894 1.088 Living situation -2.350 3.873.049 0.095 Relationship status -2.669 5.400.020 0.069 Note. N = 72. df = 1. with HIV/AIDS: Outcomes, Risks and Opportunities for Intervention Page 7 of 10
Depression and Neuropsychological Performance Table 5 Participant Neuropsychological Performance Compared to Age-matched Norms on the Grooved Pegboard and CANTAB at Baseline and Follow-up (z-scores). Depression Scores on the BDI and SCID at Baseline and Follow-up. Variable (domain) Baseline (n = 129) Follow-up (n = 78) Mean difference Std.Error 95% CI Mean difference Std.Error 95%CI from zero from zero Grooved Pegboard Psychomotor speed dominant hand -1.743* 0.09 (-1.50, -0.76) -0.47* 0.19 (-.85, 0.01) non-dominant hand -1.333* 0.18 (-1.48, -0.41) -0.35* 0.13 (-.62, 0.01) CANTAB -Mental flexibility -1.195* 0.16 (-1.50, -0.86) -0.54* 0.17 (-0.89, -.18) & attention (IED3) -Reaction Time (RTI1) -.709* 0.10 (-0.89, -0.49) 0.35* 0.00 (0.17, 0.52) (RTI2) -.654* 0.17 (-1.00, -0.33) -0.41* 0.14 (-0.68, -.13) -Spatial planning & problem (SOC1) 0.449* 0.08 (0.25, 0.59) 0.008 0.17 (-0.26, 0.46) solving (SOC2) -.233 0.13 (-0.54, -0.04) 0.007 0.11 (-0.13, 0.29) (SOC3) -.478* 0.09 (-0.71, -0.34) 0.006 0.10 (-0.13, 0.27) -Spatial working memory (SWM1) -.601* 0.09 (-0.80, -0.42) -0.31* 0.11 (-.530,.007) (SWM2) -.507* 0.08 (-0.70, -0.37) -0.40* 0.12 (-0.65, -.15) Beck Depression Inventory mean(sd) 12.3(8.2) 10.1(7.9) >14 34.8% 31% SCID - % current mood 27% (35/129) 14% (11/79) - % lifetime mood 26% (33/129) 39% (30/79) with HIV/AIDS: Outcomes, Risks and Opportunities for Intervention Page 8 of 10
Note. * Significantly (2-tailed) different from zero (p < 0.05) IED - measure of mental flexibility. RTI measure of psychomotor speed. SOC measure of problem solving and planning. SWM measure of spatial working memory SCID current mood diagnosis depressive symptoms within the past two weeks. SCID lifetime mood diagnosis past depressive episode, no symptoms within the past month. Table 4 Comparisons between Baseline Depressed and Not Depressed Groups on Neurocognitive Performance at Two Year Follow-up Depressed (n = 25) Not Depressed (n = 53) Mean change Std. Error 95% Mean change Std. Error 95% from BL of mean CI from BL of mean CI IED3a 0.84 0.43 (-1.7, 0.05) 0.63* 1.74 (-0.98, -0.28) RTI1a 0.67* 0.26 (-1.2, -0.12) 0.96* 0.09 (-1.13, -0.77) RTI2a 0.26 0.33 (-0.94, 0.43) 0.36 0.30 (-0.96, -0.24) SOC1a 0.11 0.13 (-0.16, 0.38) 0.37 0.23 (-0.09, 0.84) SOC2a 0.28* 0.11 (-0.52, -0.04) 0.75* 0.21 (-1.18, -0.30) SOC3a 0.42* 0.16 (-0.76, -0.07) 0.55* 0.13 (-0.82, -0.27) SWM1a 0.004 0.17 (-0.41, 0.32) 0.26* 0.11 (-0.49, -0.03) SWM2a 0.004 0.16 (-0.39, 0.31) -0.001 0.11 (-0.25, 0.22) Note. * Significant (2-tailed) change between baseline and follow-up (p <0.05). IED - measure of mental flexibility. RTI measure of psychomotor speed. SOC measure of problem solving and planning. SWM measure of spatial working memory with HIV/AIDS: Outcomes, Risks and Opportunities for Intervention Page 9 of 10
Table 5 Backward Elimination Regression Analysis examining Factors associated with changes in global neuropsychological performance scores over two years. Multivariate* FACTORS Coefficient Std. Error 95% Confidence Interval Lower Upper R-squared 0.405 p-value Adjusted R-squared 0.292 Age (years) -.014.005 -.024 -.003 0.013 Previous ADI.311.108.094.528 0.006 No Yes On Lamivudine at BL.154.101 -.049.357 0.135 CNS penetrating ART.306.148.0100.603 0.043 at follow-up No Yes On Stavudine at FU.136 -.010.536 0.059 On Didanosine at FU.133 -.023.507 0.077 On tenofovir at BL -.286.133 -.553 -.020 0.035 nevirapine at BL.263.110.042.484 0.020 nevirapine at FU -.232.123 -.480.015 0.065 indinavir at BL.206.136 -.066.479 0.135 abacavir at BL.157.106 -.056.370 0.147 lopinavir at FU -.149.116 -.382.082 0.202 Note. Included in the model were treatment, illness and lifestyle variables. with HIV/AIDS: Outcomes, Risks and Opportunities for Intervention Page 10 of 10