CROI 2025 Abstract eBook

Abstract eBook

Poster Abstracts

612

Epigenetic Age Advancement Predicts Cognitive Performance in Older People With HIV Carrie Johnston 1 , Raha M. Dastgheyb 2 , Leah H. Rubin 3 , Eran F. Shorer 4 , Eugenia Siegler 1 , Alina Pang 1 , Marshall Glesby 1 , Michael J. Corley 1 1 Weill Cornell Medicine, New York, NY, USA, 2 The Johns Hopkins University School of Medicine, Baltimore, MD, USA, 3 The Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA, 4 The Johns Hopkins University, Baltimore, MD, USA Background: Epigenetic studies have shown an accelerated epigenetic age phenotype in people with HIV (PWH). Accelerated epigenetic age has been linked to various disease outcomes including cognitive dysfunction. However, whether epigenetic age estimates predict cognitive measures in PWH remains understudied. Methods: PWH age 50 and older were recruited at an urban academic medical. Participants provided a blood sample at the baseline visit between 2016 2019, and epigenetic age was calculated from dried blood spots using first, second, and third generation epigenetic clocks. Participants then completed a follow-up visit including cognitive testing between 2022-2023. Timed outcomes were log transformed and reverse scored, so higher scores equated to better performance. We conducted dimension reduction on the cognitive outcome measures using Principal Components Analysis (with varimax rotation) which resulted in the following Rotated Components(RCs): learning and memory, verbal fluency, visual search, processing speed, executive function/working memory, and cognitive control/inhibition. Associations between epigenetic age and RCs were examined using partial correlation analyses accounting for sex, education, and CD4:CD8 ratio. Results: In total, 87 participants had measurement of baseline epigenetic age and completion of a follow up cognitive battery a median of 58 months [IQR: 53-63] later. The study population had a median age of 60 years (IQR:57.0,65.0), 39% female, and 42% identified as Black. First generation Horvath1 (r=0.30, p<0.01) and Horvath 2 (r= 0.31, p<0.01) residuals were associated with better processing speed. Second generation epigenetic clocks showed that greater PhenoAge (r= -0.23, p= 0.04) and GrimAge (r= 0.30, p=0.007) residuals were associated with poorer learning and memory. GrimAge residuals were also associated with better processing speed (r=0.24, p=0.026). The third generation epigenetic clock DunedinPACE was associated with poorer learning and memory (r= -0.26, p= 0.02) [Figure]. Conclusions: In this longitudinal study of a diverse cohort of older PWH, first generation clocks were associated with processing speed, second and third generation epigenetic aging clocks associated with poor learning and memory performance a median time of 5 years later. Epigenetic clocks may be a key predictive biomarker of cognitive performance spanning multiple domains in older PWH and warrant further study.

while associations with NBPs and subscale metrics were stronger. Current composite variables have value for clinical classification, but subscales and data driven NBPs may better reflect biological mechanisms and may better support new therapeutic strategies.

611

DNA Methylation Clocks Are Altered in Persons With Undetectable HIV and Correlate With Brain Aging Kalen J. Petersen 1 , Sarah Cooley 1 , Brittany Nelson 1 , Maurizio Caocci 2 , Tricia Burdo 3 , Beau Ances 1 1 Washington University in St Louis, St Louis, MO, USA, 2 Massachusetts General Hospital, Boston, MA, USA, 3 Rutgers Robert Wood Johnson Medical School, Piscataway, NJ, USA Background: DNA methylation (DNAm) clocks are quantitative markers of biological aging. The first-generation Horvath and Hannum clocks closely track actual age, second-generation clocks such as PhenoAge and GrimAge are sensitive to morbidity and mortality, and the third-generation DunedinPACE model estimates individual longitudinal “pace of aging.” Multiple studies have linked HIV infection with increased DNAm age; however, the clinical relevance of these clocks to brain health in virally suppressed persons with HIV (PWH) is unclear. We hypothesized that second-generation DNAm clocks would be significantly associated with HIV serostatus and would correlate with brain age acceleration (BAA), a machine learning model of overall brain health derived from structural MRI. Methods: PWH and people without HIV (PWoH) underwent phlebotomy and 3-Tesla brain imaging. DNA was extracted from peripheral blood mononuclear cells, and DNAm levels were quantified using MethylationEPIC 2.0 microarrays (Illumina). The R packages EWAStools , minfi , FlowSorted.Blood and methylclock were used for quality control, normalization, cell type inference, and DNAm age calculation. DeepBrainNet was used to calculate BAA from MRI. Hypotheses were tested using multiple linear regression, with age, sex at birth, and blood cell proportions as co-variates. Results: Participants included PWH ( n =61; sex=39/22 M/F; 34/27 Afr. Amer./ White; age=51.7±11.1 yr.) and PWoH ( n =50; 22/28 M/F; 28/20/2 A.A./W/ other; age=51.3±13.9 yr.). The Horvath and Hannum clocks were not different by serostatus ( p =0.51, p =0.053). PhenoAge showed a 4.8-year increase in PWH ( p =0.02; left panel), but GrimAge and DunedinPACE did not ( p =0.82, p =0.16). A consensus clock averaging all models showed a 1.9-year increase in PWH ( p =0.02). The Hannum, PhenoAge, and consensus clocks were positively associated with older brain-predicted age across serostatus ( p =0.002, p =0.008, p =0.005; right panel). Conclusions: These findings indicate that HIV infection is associated with increased DNAm age even in PWH with viral suppression. However, the Dunedin ‘pace of aging’ measure was not faster in these individuals, suggesting that biological aging is stabilized by antiviral therapy. Several DNAm clocks were correlated with older-appearing brain structure, suggesting that DNAm age reflects neurological health. More research is needed to understand the mechanisms linking altered DNAm in leukocytes to brain structural phenotypes, and potentially to neurocognitive impairment.

Poster Abstracts

The figure, table, or graphic for this abstract has been removed.

CROI 2025 166