Associations of Body Composition with OCT-Derived Optic Nerve Head and RNFL Parameters in Healthy Adults

To determine whether anthropometric and bioelectrical impedance-derived body composition indices are independently associated with optic nerve head (ONH) structural parameters and peripapillary retinal nerve fiber layer (RNFL) thickness in healthy adults.

This cross-sectional study included 480 healthy adults from the PERSIAN Cohort Study of Mashhad. Body composition was assessed by bioelectrical impedance analysis (InBody 770), and ONH/RNFL parameters by spectral-domain OCT (RTVue XR Avanti). Each body composition index was evaluated as the exposure in separate multivariable linear regression models, with each OCT parameter as the dependent variable, adjusting for age and sex. False discovery rate (FDR) correction was applied within four anatomically defined outcome families; unadjusted Spearman correlations served as a sensitivity analysis.

The study included 271 men and 209 women. Anthropometric and body composition measures differed significantly by sex, while global ONH parameters did not. Women showed slightly greater RNFL thickness in two temporal-superior sub-sectors (FDR-adjusted P = 0.034). After adjusting for age and sex, none of the 29 body composition indices showed an independent association with any ONH or RNFL parameter in any outcome family; a few adiposity-related indices showed nominal, temporal-superior associations (|β| ≤ 0.12) that did not survive FDR correction.

In healthy adults, OCT-derived ONH morphology and RNFL thickness appear largely independent of body composition after accounting for age and sex, providing normative evidence against body composition as a clinically meaningful confounder of OCT interpretation in the absence of disease.