Cognitive impairment and frailty each independently doubled the death risk in a large Baltimore cohort of drug injectors with or without HIV . A separate study tied cognitive impairment to higher death risk in a Ugandan HIV group .
The Baltimore analysis involved 215 HIV-positive and 304 HIV-negative people with a drug-injection history . These members of the ALIVE cohort had standard neurocognitive testing and standard 5-factor frailty assessment in 2010-2012. Follow-up for all-cause mortality continued through 2016. Researchers assessed 5 cognitive domains and figured a global score by combining average z scores from each domain. The investigators used Cox proportional hazards models to estimate all-cause mortality risk.
The study group had a median age of 52 and about one third were women. An analysis adjusted for age, gender, race, premorbid IQ, number of comorbid conditions, and HIV status independently linked four cognitive variables to a higher all-cause death risk at the following adjusted hazard ratios (aHR) and 95% confidence intervals (95% CI):
-- Impaired information processing: aHR 1.41 per unit lower z score, 95% CI 1.06 to 1.88
-- Impaired motor processing: aHR 1.16 per unit lower z score, 95% CI 1.30 to 1.98
-- Global impairment: aHR 1.67, 95% CI 1.10 to 2.56
-- Severe global impairment: aHR 3.43, 95% CI 1.16 to 10.2
A model considering both cognitive impairment and frailty adjusted for the same variables as the preceding analysis. Both global cognitive impairment and frailty approximately doubled chances of death:
-- Global impairment: aHR 1.93, 95% CI 1.19 to 3.14
-- Frailty: aHR 2.32, 95% CI 1.03 to 5.20
Johns Hopkins University investigators concluded that frailty and cognitive impairment predict death independently of HIV and comorbid conditions in aging people who inject drugs.
The Ugandan study focused on 399 antiretroviral-naive adults in the rural Rakai cohort who had a detailed neurologic history and exam and underwent a comprehensive neurocognitive battery and depression screening. Researchers determined stage of HIV-associated neurocognitive disorder (HAND) with standard Frascati criteria and local normative data. All participants were immediately offered antiretroviral therapy. Health workers used phone calls and home visits to determine who died 2 and 5 years after the baseline visit.
Study group age averaged 35 years, 53% were men, and median initial CD4 count stood at 274. Cognitive testing determined that 41% of participants were normal, 6% had asymptomatic neurocognitive impairment, 38% had mild impairment, and 15% had dementia. After 2 years of follow-up, 4% had died and 11% were lost to follow-up. After 5 years those proportions were 5% and 56%.
Multivariate analysis adjusted for demographic and HIV factors linked every 1-stage worse HAND classification at baseline to about 60% higher odds of death after 2 years (adjusted odds ratio [aOR] 1.58), a trend that approached statistical significance (95% CI 0.97 to 2.57, P = 0.06). The link between every 1-stage worse HAND level and death did reach statistical significance after 5 years (aOR 1.83, 95% CI 1.13 to 2.96, P = 0.01). These associations were both significant in an analysis that classified people lost to follow-up as dead.
In the 5-year analysis, more education independently lowered chances of death and an initial CD4 count below 200 independently raised chances of death. But female versus male sex did not affect odds of death. Johns Hopkins investigators and African collaborators noted that this is the first study of HAND and mortality in a resource-limited locale in the antiretroviral therapy era.
1. Piggott DA, Selnes OA, Creighton J, et al. Neurocognition, frailty, and mortality among persons aging with HIV and substance abuse. Conference on Retroviruses and Opportunistic Infections (CROI). March 4-7, 2019. Seattle. Abstract 418.
2. Saylor D, Nakigozi G, Nakasujja N, et al. HIV-associated neurocognitive disorder leads to death. Conference on Retroviruses and Opportunistic Infections (CROI). March 4-7, 2019. Seattle. Abstract 425.