The ASTRA study (Antiretrovirals, Sexual Transmission Risk and Attitudes) is a UK multicentre prospective study using self-completed questionnaires relating to HIV and depression in over 3000 HIV positive people. Importantly, the study also correlates responses to adherence and clinical outcomes.
The PHQ-9 questionnaire used in the study classifies participants according to presence/absence of depressive disorder (DD) or major depressive disorder (MDD) and a depression severity score (DSS) that divided diagnoses (from a scale of 0-27) into none, minimal, mild, moderate or severe.
Results from the first 2175 responses (from February to November 2010) were presented by Fiona Lampe from University College London in an oral abstract session. Baseline demographics for this analysis included: 16% women; 73% MSM; 10% heterosexual men. Mean age was 44 (range 18-80). Ethnic classification was: 70% white; 15% Black African; 14% other ethnicity. Approximately 86% were on ART, 9% of whom were not suppressed (VL >50 copies/mL).
For the primary endpoints, the prevalence was 26.6% (95% CI 24.8-28.5) for DD and 19.1% (95% CI 17.4-20.7%) for MDD, with 27% participants having moderate or severe depression by DSS (>/=10/27).
Perhaps surprisingly, in multivariate analysis, there were no significant differences in DD by gender, ethnicity, country of birth or ART. Younger age was associated with higher depression rates (32% age <30 to 17% at age >60, p=0.028).
Significant associations (p<0.001 for each trend) were seen with socio-economic factors including employment status (15% employed, 43% unemployed, 52% sick/disabled), education level (19% university vs. 32% other), income (defined as "money to cover basis needs": 13% "yes mostly" through to 53% "no") and social support -- a measure of supporting relationships (9% "high" through to 66% "low").
Depression also correlated positively with duration of infection (20% <2 years, 24% 2-10 years, 30% 10-20 years and 35% >30 years), which is interesting given that depression was higher with younger age. Current CD4 count or ART status had no relationship to depression scores.
Higher rates of depression were closely related to lower adherence and lower rates of viral suppression. This ranged from 24% in people who had not missed ART in the past two weeks to 29% with one missed dose, 34% with two missed doses and 42% with 3 or more missed doses.
The percentage of patients with detectable viral load by depression index 14% for those with DD vs. 7% for those with no DD, 13% vs. 8% for MDD versus no MDD, and ranged from 7% none to 18% severe (DSS) in the non-depressed vs. severely depressed groups respectively. The association between depression and viraemia remained significant after adjusting for clinic and self-reported adherence.
The study also indicated that these symptoms may be largely undiagnosed and untreated for many patients. Of the 579 people with depressive disorder (DD) by questionnaire 241 were receiving medication or other treatment and 338 were not. Conversely, of the 1596 people without depression, 200 were on (presumably effective) treatment and 1396 were not.
The total prevalence of depression (symptoms or treatment) was 35.8% (779/2175) in this study, of whom 43% (338/779) were untreated.
This is a large study that included six different UK centres (the Royal Free, Mortimer Market, Homerton, North Manchester, Brighton and Eastbourne) with a broad patient demographic.
It is probably the largest UK study to date to comprehensively look at HIV and depressive symptoms and highlights very high rates of depression with a strong indication that this is likely to be under diagnosed and under treated.
These results, especially if confirmed in the full analysis (enrollment is now completed) support the important of identification and management of depression as an important part of HIV care.
Lampe F et al. Depression and virological status among UK HIV outpatients: results from a multicentre study. 18th BHIVA Conference, 18-20 April 2012, Birmingham. Oral abstract O10.
Further details, including the slides from this presentation are on the ASTRA study website.
Links to external websites are current at time of posting but not maintained.