Hypertension is a marker of the micro-epidemiologic transition in ageing HIV populations in Kenya, Uganda and Tanzania (AFRICOS, 2013–2023)

Abstract

Background: Although mortality among people living with HIV (PLWH) in sub-Saharan Africa has decreased with antiretroviral therapy (ART) scale-up, the demographic consequences remain underexamined. Methods: Using ten years of longitudinal data from the African Cohort Study (AFRICOS; 2013–2023) in Kenya, Tanzania, and Uganda, we estimated the association of hypertension with all-cause mortality among adults aged ≥ 40 years. We combined decremental life-table analysis with discrete-time logistic regression, applying lagged and exponentially weighted moving-average (EWMA) exposure models to capture cumulative risk. Results: At enrollment, 18.6% were hypertensive; 60.1% experienced hypertension during follow-up, and all-cause mortality was 6.4%. Life-table estimates showed cumulative excess mortality became statistically significant from year 7 onward, reaching 3.4% (95% CI: 1.3–5.6) by year 9 before plateauing. Excess mortality was largest among adults aged 50–59 years (4.9% [1.8-8.0]), men (6.0% [1.4–10.5] by year 8), and underweight participants (27.1% [5.4–48.8] by year 9). Participants with suppressed viral load exhibited significant excess throughout follow-up (5.5% [3.3–7.8] by year 9). Tanzania showed the earliest and most persistent excess (≈ 5% from year 2 onward). Among underweight participants, hypertensives had a 3.6-fold higher mortality rate than non-hypertensives (58.3 vs. 16.2 per 1000 PY; p = 0.009). Similarly, among those with suppressed HIV viral load, hypertension was associated with a 2.3-fold higher mortality rate (11.4 vs. 4.9 per 1000 PY; p = 0.025). In adjusted models, hypertension was associated with higher mortality under standard lag (aOR = 2.04; 1.10–3.80) and EWMA specifications (aOR = 3.25; 1.26–8.40 at α = 0.3; aOR = 2.51; 1.16–5.44 at α = 0.7). Mortality odds were higher among participants aged ≥ 60 years (aOR = 2.23–2.40) and those with high viral load (aOR = 2.36–2.44), while overweight and obese participants had substantially lower odds of death (aOR = 0.29; 0.10–0.85 and aOR = 0.14; 0.04–0.56). Conclusions: These findings highlight a demographic transformation of the HIV epidemic in East Africa, where mortality among PLWH increasingly reflects chronic disease influence. Hypertension has become a key driver of excess mortality and a demographic indicator of the region’s compressed health transition.