Neutrophil-to-Lymphocyte and Platelet-to-Lymphocyte Ratios and 10-Year Cardiovascular Risk in HIV.

Importance: People living with HIV (PLWH) are at increased risk for cardiovascular disease (CVD), which includes both cardiac and cerebrovascular outcomes. Current CVD prediction models underestimate risk in this population, highlighting the need for improved risk stratification tools. Objective: To assess whether adding neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR), emerging biomarkers of systemic inflammation derived from routine complete blood counts to the Atherosclerotic Cardiovascular Disease (ASCVD) risk score improves 10-year CVD prediction in PLWH. Design: Retrospective cohort study from 2009 to 2019 using electronic health records from the Bronx Regional Health Information Organization. Setting: Regional public health information exchange (Bronx, New York). Participants: Adults aged ≥18 years, including 11,334 PLWH and 31,276 demographically matched people without HIV (1:3 ratio). Individuals with pre-existing CVD were excluded. Mean age was 49 years, 45% female, 67% Black/African American, 39% Hispanic/Latino. Exposures: HIV status, ASCVD risk scores, and inflammatory markers (NLR, PLR quartiles) derived from routine complete blood count parameters. Main Outcomes and Measures: Incident CVD identified via ICD-9/10 diagnostic codes over 10-year follow-up. Multivariable logistic regression models assessed associations between HIV status, ASCVD risk, and inflammatory marker quartiles with outcomes. Model performance was compared using likelihood ratio tests. Results: PLWH were younger (47 vs. 49 years), more likely to be current smokers (53% vs. 33%), and had lower total cholesterol levels (173 vs. 187 mg/dL) (all p<0.001). PLWH had higher incident CVD rates (26% vs. 22%, p<0.001), including cerebrovascular disease (9.6% vs. 6.5%, p<0.001). In the fully adjusted model, HIV-positive status was associated with 32% higher odds of CVD (OR 1.316; 95% CI: 1.249-1.386). The highest NLR quartile was strongly associated with increased disease odds (OR 1.540; 95% CI: 1.430-1.658), while higher PLR quartiles showed protective effects. The full model achieved an AUC of 0.70, with likelihood ratio tests confirming significant improvements in predictive power (all p≤0.0004). Conclusions and Relevance: Adding NLR and PLR quartiles to ASCVD risk scores significantly improves 10-year CVD prediction in PLWH. These routine, low-cost, readily available biomarkers could enhance cardiovascular risk stratification for this high-risk population.