Background
While persons with HIV (PWH) can live into old age on ART, tailoring interventions to optimize health and lifespan requires ongoing study of determinants of survival.
Methods
We studied HOPS participants with >2 visits between 1/1/2007 and 12/31/2022 who received ART continuously for >6 months and for at least 75% of time under observation. PWH were followed from baseline (1/1/2007, or date of first HOPS visit thereafter) until death, last contact date plus 6 months, or 12/31/2022. Deaths included occurred within 6 months of last contact. Univariate Poisson regression generated overall and non-AIDS-related mortality rates by demographic and select risk factors. For longitudinal trends, we used univariate and multivariable Poisson regression with generalized estimating equations (GEE) for crude and adjusted mortality rates and rate ratios (RR). GEE models updated age in years and other death risks annually.
Results
Of 4,859 participants, median age (Interquartile range (IQR) at baseline was 43 years (yrs) (33, 49), 20% were women; 396 died, most (n=307) from non-AIDS causes during a median of 7.0 yrs of observation (interquartile range: 3.4–12.0). Median age (-IQR) at death was 56 yrs (49,63). Mortality rates (95% CI) were 1.3 (0.9,1.9) and 1.1 (0.7,1.8) per 100 person-years in 2007 and 2022, respectively. In multivariable models, having public insurance versus all other types of insurance (excluding private), was associated with greater all-cause mortality RR = 1.55 (95% CI: 1.13, 2.13) (P<0.01). Non-Hispanic Black people, compared to White people, had higher non-AIDS-related mortality rates (P<0.05), but not all-cause mortality. Other factors positively associated with overall mortality (P<0.05) were a history of injection drug use, and any of cardiovascular disease, renal disease, hypertension, smoking or cancer. However, the magnitude of association with mortality was greatest for HIV-related clinical factors (Figure) i.e., low CD4 cell count, and included HIV non-suppression. Sex and obesity were not associated with death rates in adjusted analyses.
Conclusions
HOPS mortality rates remained low among ART-treated PWH. Non-AIDS causes of death predominated but the greatest risks for death remained HIV-related factors i.e. lower CD4; even low-level viremia remained a significant factor. Interventions to further optimize lifespan among aging PWH will require improvement in prevention and management of comorbidities and tight maintenance of viral control and immune health.