HIV disease progression to Acquired Immune Deficiency Syndrome (AIDS) is one of the greatest contributors of mortality in Africa [1]. A large number of infections occur among serodiscordant couples [2]. The progressive loss of CD4+ T lymphocytes during HIV infection eventually results in an inability to mount an adequate immune response to opportunistic pathogens resulting in death [3].

Socioeconomic determinants such as age, sex, and income are reported as contributors to disease progression [4–8] in HIV infected individuals who are not on highly active antiretroviral therapy (HAART). These factors may still continue to influence the uptake of care and treatment even, where these services are provided at no cost to the individual. There is a paucity of data on the effects of socioeconomic determinants of disease progression among HIV infected individuals in sub-Saharan Africa.

The World Health Organization (WHO) recommends the treatment of all HIV positive individuals in a discordant relationship regardless of their CD4 count and all adults and adolescents with CD4 cell counts below 500 cells/mm³ [9] should receive treatment. The implementation of these guidelines [9] in low-income countries like Kenya [10] may be hindered due to the financial considerations on already strained health systems. To alleviate this, low and middle-income countries commonly use a phasic approach in the implementation of HAART scaling up [11]. Integration of clinical and sociodemographic criteria would be useful in the in prioritization of vulnerable populations. The objective of this study was to identify socioeconomic correlates of disease progression in HIV infected adults in serodiscordant relationships.

This study was conducted in a sample of HIV positive individuals drawn from a randomized clinical trial (RCT) carried out in a serodiscordant couple’s clinic in Nairobi. The RCT was investigating Acyclovir in the prevention of HIV transmission (HSV2/HIV1 study) - clinical trials no: NCT00194519 [12]. We analysed data from 312 HIV positive individuals drawn from a cohort of 416 study participants for a period of two years beginning in November 2011 and ending in June 2013. Participants were selected from the main cohort based on CD4 count of above 350 cells/mm³ and WHO clinical stage 1 or 2 at enrolment into the clinical trial. We excluded HIV positive individuals who seroconverted during the clinical trial [10]. Out of the total cohort initially enrolled in the clinical trial, 86 were excluded due to CD4 counts of ≤350 cells/mm³ at enrollment, because they were eligible for HAART according to the Kenya national guidelines 2011 [13]. End point outcomes for 18 participants (5 %) were unavailable because they were reported as lost to follow up (n = 11) or dead (n = 7) during the clinical trial. A summary of the cohort is illustrated in Fig. 1.

Disease progression was defined as a composite outcome of WHO clinical staging criteria stages 3 and 4 or CD4 count of < 350 cells/mm^3. Outcomes were evaluated at six month intervals over the two years. Socio-demographic and economic predictors of disease investigated included daily income available for expenditure, level of education, housing settlement, age, and sex.

In this study there were more HIV- infected individuals, who showed progression based on WHO clinical staging criteria compared to those showing progression based on CD4 count criteria. This is similar to studies in the PRE-HAART era [15] where disease progression was reported earlier when defined by laboratory criteria rather than by the development of an opportunistic infection.

This study found a strong association between daily incomes and disease progression portrayed by a decrease in CD4 counts to below 350 cells/mm³ (p = 0.026 CI 95 %). This progression occurred in spite of the optimal provision of free HIV care and treatment for opportunistic infections and prophylaxis using cotrimoxazole or dapsone. Income as a determinant of disease progression was strengthened by finding of an association between higher daily income available for expenditure and delayed disease progression. The acyclovir arm had been reported as having displayed delayed disease progression in the larger study [12]. Studies in America and Canada [5, 6] have reported a similar association between low incomes and more rapid HIV disease progression. Additional evidence shows, higher rates of disease progression after seroconversion among individuals with low incomes levels prior to infection [9]. The association between income and disease progression can therefore not be explained by a reduction of income due to HIV infection and morbidity. We postulate that micronutrient deficiencies which are common in low-income countries and have been proven to compromise the immune systems of HIV infected individuals [16] may play a role in the rate of disease progression observed [16]. Even with most treatment programmes providing multivitamin supplementation to individuals in care and treatment, People Living with HIV (PLWHIV) still bear a heavy burden of dietary micronutrient supplementation which is affected by the daily income available for expenditure. This contributes to the [16] weakening of the immune system and the depletion of CD4 cells resulting in faster disease progression. Micronutrient deficiency in the diet may explain the strong association found in this study between income available for daily expenditure and disease progression. Even with HAART provision, low income levels continue to be associated with poor health outcomes. In Kenya the inability to pay for transport to a HAART provision centre results in poor uptake of HAART [17]. Similarly, a study in British Columbia also reported the inability to pay for transport to a treatment centre as a cause of poor treatment outcomes among individuals with low-income [18]. The cost of transport to a treatment centre may therefore be another reason for the association between HIV disease progression and daily income available for expenditure.

Studies in Tanzania, Uganda and France [5, 19, 20] have found higher rates of disease progression in individuals above 40 years. This study did not find an association (p = 0.68; p < 0.05) between age, and disease progression. The finding of age as a non-determinant of disease progression in this study was most likely due to only 9 % of the participants’ being above 40 years. This study also found no association (p = 0.06; p < 0.05) between sex and disease progression measured using WHO criteria or a reduction in CD4 counts. This finding was corroborated by findings of various studies [6, 7] including a meta-analysis of 23 cohorts from Europe, Australia, and Canada which reported no association between sex and HIV disease progression [7]. The lack of association between level of education and disease progression in this study corresponds to findings in other similar studies of HAART naive individuals [5, 8]. Associations have been found in individuals already on HAART [21] with increased level of education. This delayed progression is attributed to the empowered attitude towards treatment and care resulting from higher levels of education. This did not apply to this study which focused on the period before initiation of HAART.

This study is unusual as it integrated widely accepted definitions of poverty [14] during analysis allowing comparison of results among other low and middle income countries within a global context. The study was conducted in a cohort of HIV positive individuals in serodiscordant unions who were all participating in a controlled trial. This may have introduced some level of selection bias though it is important to note that progression of disease is not expected to be different from other HIV positive individuals. The retrospective study design limited the socioeconomic indicators available to those captured during data collection for HSV2/HIV1 study clinical trial as this study was retrospective [12].

The data presented here however indicates that poverty, defined as the level of daily income available for expenditure, influences pre-HAART HIV disease progression and deserves consideration as a contextual factor in HIV disease progression. Additionally, impoverished populations may benefit from prioritization during implementation of WHO guidelines.

The association reported between daily income available for daily expenditure and HIV disease progression indicates that economic empowerment should be considered as a possible contributor to better health outcomes in HIV infected individuals. We recommend prioritization of populations with the lowest daily available incomes for expenditure in low and middle income countries during the implementation of HAART scale up.