To the Editors:
The massive scale-up of antiretroviral therapy (ART) in sub-Saharan Africa (SSA) has resulted in substantial increases in life expectancy among people living with HIV (PLWH).1 This increase in life expectancy, coupled with a secondary HIV incidence peak observed in a multi-country study among older adults in the region,2,3 suggests that the number of PLWH ≥50 years will grow substantially in coming years. Indeed, the number of older PLWH in sub-Saharan Africa is projected to triple from 3.1 million in 2011 to 9.3 million by 2040.4
In a prior analysis using routinely collected data from a large multi-country HIV program in SSA, we found that adults aged ≥50 were increasingly represented among the population on ART.5 By 2010, older adults accounted for 12% of patients newly initiating ART and 16% of patients currently on ART. Additionally, the cumulative incidence of recorded death through one year after ART initiation was highest among patients aged ≥50 compared with younger individuals. Furthermore, the cumulative incidence of loss to follow-up (LTF) among patients on ART aged ≥50 was similar to that observed among those 40–49 years of age, but significantly lower than among those <40 years.
While those data highlighted the need to account for adults aged ≥50 years in HIV programs in SSA, age-specific data on determinants of mortality and LTF are also required so that programs can better support this growing population. Among the younger population of ART patients 15–49 years of age, patient sex has consistently emerged as an important determinant of poor outcomes, with men having a substantially higher risk of dying and being LTF than women across a range of programs and settings.6–8 To our knowledge, the relationship between sex and outcomes among older patients on ART has not been examined.
We used routinely collected de-identified patient-level data from 198 HIV clinics in Kenya (69 clinics), Mozambique (33 clinics), Rwanda (44 clinics), and Tanzania (52 clinics) to assess age- and sex-related differences in recorded all-cause mortality and LTF among adults ≥50 years of age who initiated ART between January 2005 and December 2010. All facilities provided HIV services per national guidelines, and received technical and financial support from ICAP-Columbia University through PEPFAR funding.
Patients were categorized into 5 strata reflecting age at the time of ART initiation: 50–54, 55–59, 60–64, 65–69, and ≥70 years. Given no prior research in this area, selection of age strata was not guided by pre-existing hypotheses but rather to facilitate exploratory analyses. Patients were considered LTF if they did not have a recorded visit for at least 6 months prior to the end of the observation period (December 2011) and were not known to have died or transferred to another facility. LTF patients were censored 15 days after their last recorded visit.
Cox Proportional Hazards models were used to describe the progression to death and LTF through 12 months after ART initiation. Sex and age differences in death and LTF were assessed in 3 ways. First, sex differences were examined for all patients. Second, sex differences were estimated in age-stratified models. Finally, differences by age were examined in separate models for males and females. All models controlled for the confounding effects of advanced HIV disease at ART initiation (CD4+ cell count <100 cells per cubic millimeter or WHO stage 4), entry point (inpatient vs. outpatient setting) and country; accounted for clustering of patients within clinics; and censored patients who were LTF (in mortality models) or died (in LTF models). Analyses were conducted using SAS 9.3. The Columbia University Medical Center, the US Centers for Disease Control and Prevention, and each of the participating countries provided ethical approval.
A total of 21,461 patients ≥50 years of age initiated ART between 2005 and 2010 at the 198 health facilities in this analysis: 10,800 were age 50–54, 5671 age 55–59, 2922 age 60–64, 1222 age 65–69 and 846 age ≥70. The proportion of female patients decreased with age from 53% of adults 50–54 years to 39% of those age ≥70. Across all age groups included, 6% of patients were diagnosed with HIV while receiving inpatient services, with a non-significant increasing trend by age (5% among those aged 50–54% to 8% among those aged ≥70 years) (P = 0.07). Median CD4+ cell count at ART initiation was 168 per cubic millimeter and similar across age groups. The incidence of mortality and LTF in the total population was 5.7% and 15.5%, respectively.
In unadjusted models among all patients on ART, men had significantly higher rates of recorded death through 12 months after ART initiation compared to women (HR = 1.49, 95% CI: 1.33 to 1.66) and LTF (HR = 1.25, 95% CI: 1.15 to 1.35). These sex differentials remained after adjusting for advanced HIV disease at initiation, point of entry, and country (Death: aHR = 1.38, 95% CI: 1.24 to 1.53; LTF: aHR = 1.23, 95% CI: 1.13 to 1.34) (Table 1). They also persisted when individual countries were removed from the analysis, and in country-specific models with the exception of Tanzania where no difference in the hazard of LTF was observed by sex (data not shown).
Models examining sex differences within age strata showed that men on ART also had significantly higher rates of death and LTF compared to women in the 50–54 age group (Death: aHR = 1.50, 95% CI: 1.29 to 1.76; LTF: aHR = 1.35, 95% CI: 1.18 to 1.53) and in the 55–59 age group (Death: aHR = 1.38, 95% CI: 1.07 to 1.78; LTF: aHR = 1.23, 95% CI: 1.08 to 1.39). Sex was not significantly associated with death or LTF within age strata among patients age ≥60.
Multivariable models stratified by sex also highlighted sex differences in the relationship between age and death and LTF. For women, there was an increasing association between age and recorded death among those 60–64 years of age (aHR = 1.40, 95% CI: 1.09 to 1.81), 65–69 years (aHR = 1.81, 95% CI: 1.26 to 2.60), and ≥70 years (aHR = 2.49, 95% CI: 1.70 to 3.65) compared to those aged 50–54 years. Similarly, the hazard of LTF was significantly higher for women aged 60–64 (aHR = 1.33, 95% CI: 1.14 to 1.54) and 65–69 (aHR = 1.31, 95% CI: 1.02 to 1.67) compared to those 50–54 years. No significant difference was observed in LTF between women 55–59 years (aHR = 1.08, 95% CI: 0.95 to 1.23) and those ≥70 years (aHR = 1.34, 95% CI: 0.97 to 1.85) compared to those 50–54 years. For men, while rates of recorded death were significantly higher among those 65–69 years of age (aHR = 1.47, 95% CI: 1.07 to 2.02), and ≥70 years (aHR = 1.52, 95% CI: 1.08 to 2.12) compared to those 50–54 years, differences in LTF were only observed among the oldest patients ≥70 years compared to those 50–54 years (aHR = 1.24, 95% CI: 1.03 to 1.51).
In this analysis of over 21,000 adults aged ≥50 years who initiated ART across diverse settings in 4 SSA countries, we found significant age- and sex-related differences in mortality and LTF. Men 50–59 years had higher rates of recorded death and LTF compared with women of the same age, echoing the pattern observed among PLWH aged 15–49 years.9–12 This may reflect similarities between younger men and those aged 50–59 years with regards to risk factors for poor outcomes such as frequent mobility,13,14 disease denial, and stigma.7,15,16 It may also result from overall higher non-HIV associated mortality among men relative to women due to higher prevalence of co-morbid conditions among men.17,18 However, the risk of death and LTF did not differ significantly between sexes aged ≥60 suggesting that factors contributing to sex differentials in death and LTF wane over the lifespan. Additionally, aging-related comorbidities may be more pronounced after age 60, thereby reducing sex differences as PLWH age.
As expected, mortality generally increased with age. While this can be partially explained by the natural aging process, we also observed sex differences in the relationship between age and death with elevated mortality rates beginning at age 60 for women versus age 65 for men. Further, the magnitude of the association between older age and reported mortality was substantially greater among women than men. This finding is surprising given that women have higher life expectancies than men in the 4 countries studied,19,20 and may reflect increased documentation of death among women than men.21
This study had several strengths including the use of routinely collected clinical data from a large multi-country program spanning more than 5 years. Also, sensitivity analyses examining sex differences in death and LTF by country and adjustment by country in multivariable analyses had little effect on the results, suggesting that our findings are robust across settings. Limitations include that 14% of participants were missing both CD4+ cell counts or WHO stage at ART initiation, and thus were included in the models as a “missing” category. However, there were no significant differences in missingness across older age groups suggesting non-differential misclassification bias. Additionally, as is common in studies of older patients, our study was subject to survivor bias which may have masked any sex differentials in outcomes after age 60.
In conclusion, our analysis suggests that sex differences in all-cause mortality and LTF observed among younger patients on ART are also observed among older patients aged 50–59, and that women on ART appear to have disproportionate risk of death as they age compared to men. As older adults will increasingly be represented in HIV care settings in SSA, further research is needed to understand determinants of mortality and retention in this important and growing population.
1. UNAIDS. Global Report: UNAIDS Report on the Global AIDS Epidemic. Geneva, Switzerland: UNAIDS; 2012.
2. Amin I. Social capital and sexual risk-taking behaviors among older adults in the United States. J App Gerontol. 2014.
3. Mutevedzi PC, Newell ML. A missing piece in the puzzle: HIV in mature adults in sub-Saharan Africa. Future Virol. 2011;6:755–767.
4. Hontelez JA, de Vlas SJ, Baltussen R, et al. The impact of antiretroviral treatment on the age composition of the HIV epidemic in sub-Saharan Africa. AIDS. 2012;26(suppl 1):S19–S30.
5. Eduardo E, Lamb MR, Kandula S, et al. Characteristics and outcomes among older HIV-positive adults enrolled in HIV programs in four sub-Saharan African countries. PLoS One. 2014;9:e103864.
6. Druyts E, Dybul M, Kanters S, et al. Male sex and the risk of mortality among individuals enrolled in antiretroviral therapy programs in Africa: a systematic review and meta-analysis. AIDS. 2013;27:417–425.
7. Govindasamy D, Ford N, Kranzer K. Risk factors, barriers and facilitators for linkage to antiretroviral therapy care: a systematic review. AIDS. 2012;26:2059–2067.
8. Nuwagaba-Biribonwoha H, Jakubowski A, Mugisha V, et al. Low risk of attrition among adults on antiretroviral therapy in the Rwandan national program: a retrospective cohort analysis of 6, 12, and 18 month outcomes. BMC Public Health. 2014;14:889.
9. Fox MP, Rosen S. Patient retention in antiretroviral therapy programs up to three years on treatment in sub-Saharan Africa, 2007-2009: systematic review. Trop Med Int Health. 2010;159(suppl 1):1–15.
10. Koole O, Kalenga L, Kiumbu M, et al. Retention in a NGO supported antiretroviral program in the Democratic Republic of Congo. PLoS One. 2012;7:e40971.
11. Odafe S, Torpey K, Khamofu H, et al. The pattern of attrition from an antiretroviral treatment program in Nigeria. PLoS One. 2012;7:e51254.
12. Weigel R, Estill J, Egger M, et al. Mortality and loss to followup in the first year of ART: Malawi national ART programme. AIDS. 2012;26:365–373.
13. Furin J, Miller AC, Lesia N, et al. Gender differences in enrolment in an HIV-treatment programme in rural Lesotho, 2006-2008: a brief report. Int J STD AIDS. 2012;23:689–691.
14. Siu GE, Wight D, Seeley J. How a masculine work ethic and economic circumstances affect uptake of HIV treatment: experiences of men from an artisanal gold mining community in rural eastern Uganda. J Int AIDS Soc. 2012;15(suppl 1):1–9.
15. Mburu G, Ram M, Siu G, et al. Intersectionality of HIV stigma and masculinity in eastern Uganda: implications for involving men in HIV programmes. BMC Public Health. 2014;14:1061.
16. Mugisha V, Teasdale CA, Wang C, et al. Determinants of mortality and loss to follow-up among adults enrolled in HIV care services in Rwanda. PLoS One. 2014;9:e85774.
17. Cornell M, Schomaker M, Garone DB, et al. Gender differences in survival among adult patients starting antiretroviral therapy in South Africa: a multicentre cohort study. PLoS Med. 2012;9:e1001304.
18. Streatfield PK, Khan WA, Bhuiya A, et al. Adult non-communicable disease mortality in Africa and Asia: evidence from INDEPTH Health and Demographic Surveillance System sites. Glob Health Action. 2014;7:25365.
19. Salomon JA, Wang H, Freeman MK, et al. Healthy life expectancy for 187 countries, 1990-2010: a systematic analysis for the Global Burden Disease Study 2010. Lancet. 2012;380:2144–2162.
20. Clark R, Peck BM. Examining the gender gap in life expectancy: a cross-national analysis, 1980–2005. Social Sci Q. 2012;93:820–837.
21. Groenewald P, Bradshaw D, Daniels J, et al. Local-level mortality surveillance in resource-limited settings: a case study of Cape Town highlights disparities in health. Bull World Health Organ. 2010;88:444–451.