Association Between the Nutritional Inflammatory Index and Obstructive Sleep Apnea Risk: Insights from the NHANES 2015–2020 and Mendelian Randomization Analyses

<b>Background/Objectives:</b> Current approaches to monitoring obstructive sleep apnea (OSA) risk primarily focus on structural or functional abnormalities, often neglecting systemic metabolic and physiological factors. Resource-intensive methods, such as polysomnography (PSG), limit their routine applicability. This study aimed to evaluate composite nutritional-inflammatory indices derived from routine blood markers to identify feasible indices for OSA management and explore their association with OSA risk. <b>Methods:</b> Data from 9622 adults in the NHANES (2015-2020) and GWAS datasets were analyzed using logistic regression, restricted cubic splines, machine learning, and Mendelian randomization (MR). These techniques were employed to identify nutritional-inflammatory indices associated with OSA risk. Random forest modeling identified body mass index (BMI) and albumin (ALB) as key components of the advanced lung cancer inflammation index (ALI). Causal relationships between ALI components and OSA were validated using MR. <b>Results:</b> ALI was significantly associated with OSA, with individuals in the highest ALI tertile exhibiting a 59% higher likelihood of OSA (OR = 1.59, 95% CI: 1.38-1.84; <i>p</i> < 0.001). BMI and ALB were identified as key contributors to ALI and confirmed as causal risk factors for OSA (BMI: OR = 1.91, 95% CI: 1.80-2.02; ALB: OR = 1.11, 95% CI: 1.04-1.19). Age, gender, and the neutrophil-to-lymphocyte ratio (NLR) were also significant predictors. <b>Conclusions:</b> This study identifies ALI as a potential composite index for assessing OSA risk. Integrating statistical modeling, machine learning, and causal inference techniques highlights the utility of nutritional-inflammatory indices in improving OSA monitoring and management in clinical practice.