The WHO recently recommended earlier initiation of antiretroviral therapy (ART), in particular at CD4+ T-cell lymphocyte cell (CD4+) count above 500 cells/μl . These guidelines are based on evidence that earlier initiation reduces morbidity and mortality and lowers the risk of HIV transmission [2–4]. Earlier initiation may confer significant economic benefits as well. A growing body of work has shown that ART helps individuals resume employment after having been too sick to work [5–14]: earlier ART may prevent the pretreatment declines in socioeconomic status altogether . Moreover, earlier therapy may also augment subsequent economic recovery if those initiating ART at lower CD4+ cell counts have difficulty achieving their pretreatment economic status, perhaps due to reduced productivity from persistent morbidity or lower social mobility.
At present, the relationship between CD4+ cell count at initiation and the trajectory of economic status is not well understood. A recent study demonstrated that individuals with CD4+ cell count above 200 cells/μl in a rural Ugandan parish had similar labor force participation rates as HIV-negative individuals . However, this study did not explicitly focus on individuals on ART, nor did it examine differences in economic outcomes over time. To address this gap, we used data from an HIV cohort in rural Uganda to examine whether earlier initiation of ART was associated with a higher labor force participation rates and greater household asset ownership, both at ART initiation and through 6 years of follow-up.
Participants, setting, and data
We used data from the Uganda AIDS Rural Treatment Outcomes (UARTO) Study, an ongoing cohort study of HIV-infected individuals initiating ART in rural, south-western Uganda, started in 2005. Previously, ART-naive persons 18 years of age or older initiating ART at the Immune Suppression Syndrome Clinic of the Mbarara Regional Referral Hospital were eligible for enrollment. Survey instruments were translated into the local Bantu language Runyankole, with interviews conducted by a native Runyankole speaker. Ethical approval for the study was obtained from the Mbarara University of Science and Technology Institutional Review Committee, the Committee on Human Research at the University of California at San Francisco, and the Partners Healthcare Human Research Committee. Consistent with national guidelines, clearance for the study was granted by the Uganda National Council for Science and Technology and the Research Secretariat in the Office of the President.
Participants provided information on socioeconomic status and basic demographic characteristics at baseline and at yearly intervals thereafter. Our primary outcomes of interest were labor force participation and household asset ownership. Participants who reported engagement in any income-generating activity, whether in the informal (self-employment in trades, agriculture, etc.) or formal sectors at the time of survey were considered as participating in the labor force. For asset ownership, we created an index representing the number of reported assets owned by the household out of 16 different durable goods (see Table 1 notes). Our primary explanatory variable of interest was baseline CD4+ cell count, which was obtained via serum samples for all participants prior to initiating ART. We partitioned the sample into persons initiating ART at CD4+ cell count below 200 cells/μl vs. those initiating ART at CD4+ at least 200.
We first plotted unadjusted trends in labor force participation by CD4+ cell count at initiation. Second, we fitted a probit regression model specifying labor force participation as the outcome variable and the following explanatory variables: the baseline CD4+ cell count (<200 vs. ≥200 cells/μl); a set of binary indicators for each year since ART initiation; and interactions between the CD4+ cell count and year indicators, so as to test whether the labor supply response to ART differed between the CD4+ groups. We adjusted our models for age and age-squared interacted with sex, educational attainment, marital status, and season of interview (March–May and October–November rainy seasons). We presented labor force participation estimates as marginal effects (i.e. the percentage point increase in the probability of observing the dependent variable corresponding to a one-unit change in a continuous explanatory variable or a change from 0 to 1 for a dichotomous explanatory variable). For the household asset scores, we fitted regression models using ordinary least squares.
For both labor force participation and asset scores, we examined trends through 6 years after ART initiation. All analyses were conducted using Stata/SE 13.0 (Stata Corp, College Station, Texas, USA).
Our sample consisted of 505 participants: 325 initiated ART at CD4+ cell count below 200 cells/μl and 180 initiated ART at CD4+ at least 200. Within the latter group, the median CD4+ cell count at initiation was 284 cells/μl [interquartile range (IQR) 233–360], with 47 (26%) initiating ART at CD4+ at least 350 cells/μl. Women comprised 70% of the sample and participants were, on average, surveyed at five annual time points. At baseline, participants initiating ART at CD4+ cell count at least 200 cells/μl were more likely to be working compared to those initiating ART at CD4+ below 200 (70 vs. 56%; χ2 = 9.15, P < 0.01), but had similar asset index scores. Apart from marriage, there were no statistically significant differences in baseline characteristics (Table S1, http://links.lww.com/QAD/A465).
Figure 1 represents unadjusted trends in labor force participation and asset ownership. Although participants initiating ART at CD4+ cell count below 200 cells/μl were less likely to be working at baseline, within 1 year of treatment initiation, their average labor force participation rate converged to that of the participants initiating ART at CD4+ at least 200 cells/μl. There were moderate increases for both groups thereafter, with participation rates around 80% at 6-year follow-up (Fig. 1a). For asset scores, both groups started at similar levels and experienced gradual increases after initiating ART (Fig. 1b).
The multivariable regression results, shown in Table 1, were consistent with the patterns in Fig. 1. As indicated by the regression coefficients on the CD4+ cell count variable, participants initiating ART at CD4+ at least 200 cells/μl were 13 percentage points more likely to be working at baseline [column 1; b = 0.13, 95% confidence interval (CI) 0.06–0.21]. This group experienced little change in labor force participation rates over baseline in the ensuing years: the sum of the interacted and noninteracted yearly indicator variables (which recovers the total change in participation rates relative to baseline for participants initiating ART at CD4+ cell count at least 200 cells/μl) was effectively zero for most of the 6 follow-up years. Those initiating at CD4+ below 200 cells/μl, however, did experience increased labor force participation: the coefficients on the uninteracted yearly indicator variables indicate a 12–20 percentage point increase in the probability of working over baseline for each follow-up year (F = 61.24, P < 0.01).
For asset scores (column 2), participants in the high CD4+ cell count group owned a similar number of assets at baseline, and experienced gradual similar increases after starting ART, as participants in the low CD4+ cell count group: the time dummies were collectively statistically significant (F
= 6.25, P < 0.01), but the coefficients on the interactions with the CD4+ cell count dummy were not. For both models, the associations between the outcomes and schooling, sex, and age were in the expected directions (Table S2, http://links.lww.com/QAD/A465).
We additionally estimated models including individual fixed effects, to control for time-invariant individual level confounders; the results were unchanged. In addition, we also excluded from the analysis all women who reported being pregnant at baseline (since they may have been less likely to work and more likely to access ART at higher CD4+ cell counts). The substantive results again were unchanged. We also considered sample attrition and missing interview data, and ruled this out as a major source of bias in our comparisons of economic outcomes between the high and low CD4+ groups (see Table S3, http://links.lww.com/QAD/A465 and associated notes). Finally, we examined whether economic status at baseline and over time differed for those initiating at CD4+ cell count at least 350 cells/μl. As shown in Figure S1 and Table S4 (http://links.lww.com/QAD/A465), outcomes at baseline and at 1–2 years of follow-up were substantively similar to those for initiating at CD4+ cell count above 200 and below 350 cells/μl, although small sample sizes (only 47 participants initiated therapy at CD4+ at least 350 cells/μl) precluded any definitive interpretation.
Results from this cohort study of HIV-infected adults on ART in rural Uganda showed that participants initiating ART at CD4+ cell count at least 200 cells/μl started out with higher labor force participation rates relative to their counterparts, initiating ART at CD4+ below 200. Within 1 year after starting ART, however, those initiating ART at CD4+ below 200 cells/μl had caught up and thereafter maintained similar trajectories in labor force participation: the overall 6-year change in likelihood of working for adults initiating at CD4+ cell count below 200 cells/μl was 20 percentage points. Interestingly, despite being more likely to be working at baseline, participants with higher CD4+ cell counts at initiation reported similar household asset scores as the lower CD4+ cell count group, with both groups experiencing increases in asset ownership over the study period.
These results have several policy implications. Whereas those initiating ART at CD4+ cell count below 200 cells/μl caught up quickly and experienced similar trajectories thereafter, earlier ART initiation may have prevented households from experiencing job loss and economic hardship in the first place, consistent with recent findings by Thirumurthy et al. . Focusing on labor force participation alone, however, may mask more subtle impacts of HIV on individual and household socioeconomic outcomes. In particular, despite being more likely to work, participants with higher baseline CD4+ cell counts started with similar asset scores as those initiating ART at lower CD4+ cell counts. It may be that durable assets were being sold to make up for unmeasured declines in productivity or that participants, prior to treatment initiation, perceived shorter life expectancies and therefore did not undertake long-term investments in productive assets . These alternate mechanisms deserve further attention and motivate economic interventions at the time of diagnosis in order to prevent nonmorbidity-related economic decline. Finally, continued improvements in socioeconomic position for all study participants 6 years after ART initiation demonstrates that economic impacts of ART may persist well after immune reconstitution is achieved.
The study has several limitations. First, the nonexperimental study design limited our ability to make causal inferences. Second, we did not observe economic status prior to the baseline survey, limiting our ability to characterize the trajectory of pre-ART economic status. Third, we lacked data on alternate economic measures such as hours worked or wage earnings, which would have enabled us to better characterize subtle differences in productivity across the different CD4+ cell count groups over the course of ART. Finally, it is unclear whether our findings generalize to other settings, though similarities in ART-led economic recovery across different countries suggest that some degree of generalization is reasonable [6,10,11].
Understanding the relationship between timing of ART initiation and economic deterioration is important for the design of ART guidelines and the valuation of the economic benefits of early ART initiation. Our study demonstrates that initiating ART at CD4+ cell count above 200 cells/μl may have helped prevent job loss, though perhaps it did not help stave off losses in household assets. Future research should examine a broader set of socioeconomic outcomes across a wider range of baseline CD4+ cell counts (in particular the 350 and 500 cells/μl thresholds identified in the 2010 and 2013 WHO guidelines), most optimally in the setting of a randomized controlled trial, to better identify ART initiation thresholds in which economic status is not compromised.
We thank the UARTO participants and staff for making this study possible. We also thank three anonymous referees for their helpful comments.
Author contributions: A.S.V. took the lead on the conception and design of the analysis, performed data analysis, wrote the first draft of the article, and gave final approval of the version to be published.
H.T., J.E.H., Y.B.II, A.K., P.W.H., J.N.M. and D.R.B. assisted with study conception and design, interpretation of the results, made substantial edits and critical revisions to the article, and gave final approval of the version to be published.
M.J.S. assisted with the interpretation of the results, made substantial edits and critical revisions to the article, and gave final approval to the version to be published.
A.C.T. supervised study conception and design of the analysis, assisted with interpretation of the results, made substantial edits and critical revisions to the article, and gave final approval of the version to be published.
Funding sources: The Uganda AIDS Rural Treatment Outcomes (UARTO) study is funded by the US National Institutes of Health (NIH) R01MH054907 and P30A1027763. The authors also acknowledge salary support through K23MH087228, K24MH087227, K23 MH087228, K01HD061605, and K23MH096620.
Conflicts of interest
The authors have no conflicts of interest to report.
1. World Health OrganizationConsolidated guidelines on general HIV care and the use of antiretroviral drugs for treating and preventing HIV infection: recommendations for a public health approach
. Geneva:WHO; 2013.
2. Koenig S, Bang H, Severe P, Juste M, Ambroise A, Edwards A, et al. Cost-effectiveness of early versus standard antiretroviral therapy in HIV-infected adults in Haiti
. PLoS Med
3. Severe P, Juste M, Ambroise A, Eliacin L, Marchand C, Apollon S, et al. Early versus standard antiretroviral therapy for HIV-infected adults in Haiti
. N Engl J Med
4. Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour MC, Kumarasamy N, et al. Prevention of HIV-1 infection with early antiretroviral therapy
. N Engl J Med
5. Beard J, Feeley F, Rosen S. Economic and quality of life outcomes of antiretroviral therapy for HIV/AIDS in developing countries: a systematic literature review
. AIDS Care
6. Bor J, Tanser F, Newell M, Barninghausen T. In a study of a population cohort in South Africa, HIV patients on antiretrovirals had nearly full recovery of employment
. Health Affairs
7. Habyrimana J, Mbakile B, Pop-Eleches C. The impact of HIV/AIDS and ARV treatment on worker absenteeism: implications for African firms
. J Hum Resour
8. Nannungi A, Wagner G, Ghosh-Dastidar B. The impact of ART on the economic outcomes of people living with HIV/AIDS
. AIDS Res Treat
9. Rosen S, Larson B, Brennan A, Long L, Fox M, Mongwenyana C, et al. Economic outcomes of patients receiving antiretroviral therapy for HIV/AIDS in South Africa are sustained through three years on treatment
. PLoS One
10. Thirumurthy H, Graff-Zivin J, Goldstein M. The economic impact of AIDS treatment: labor supply in Western Kenya
. J Hum Resour
11. Thirumurthy H, Jafri A, Srinivas G, Arumugam V, Saravanan RM, Angappan SK, et al. Two-year impacts on employment and income among adults receiving antiretroviral therapy in Tamil Nadu, India: a cohort study
12. Lem M, Moore D, Marion S, Bonner S, Chan K, O’Connell J, et al. Back to work: correlates of employment among person receiving highly active antiretroviral therapy
. AIDS Care
13. Rosen S, Ketlhapile M, Sanne I, Desilva M. Differences in normal activities, job performance and symptom prevalence between patients not yet on antiretroviral therapy and patients initiating therapy in South Africa
14. Rosen S, Larson B, Rohr J, Sanne I, Mongwenyana C, Brennan A, et al. Effects of antiretroviral therapy on patients’ economic well being: five-year follow-up
2014; [Epub ahead of print].
15. Thirumurthy H, Chamie G, Jain V, Kabami J, Krawrisiima D, Clark T, et al. Improved employment and education outcomes in households of HIV-infected adults with high CD4 cell counts: evidence from a community health campaign in Uganda
16. Mimeo University of Pennsylvania, Baranov V, Kohler H-P. The impact of AIDS treatment on savings and human capital in Malawi