Development of a sex-specific visceral fat area estimation using discrete multi-wavelength near-infrared spectroscopy measurements in Korean individuals.

Abstract

Background/objectives: A high level of visceral fat area (VFA) is associated with obesity and cardiometabolic risk factors. VFA measured by computer tomography (CT) scan is accurate but has limitations for everyday use. Meanwhile, near-infrared (NIR) light penetrates the superficial layers of the human body so that fat content can be measured just as CT imaging measures fat accumulation.

Subjects/methods: This study evaluated whether discrete multi-wavelength NIR spectroscopy (DMW-NIRS) can be used to measure abdominal fat as a satisfactory alternative to a CT scan. 290 subjects were enrolled in this study, and each subject underwent DMW-NIRS NIR measurement and CT scan. A sex-specific DMW-NIRS-based VFA estimation formula was developed by multiple linear regression, including lipid density, age, and body mass index (BMI).

Results: The model of DMW-NIRS estimated VFA gave the least Akaike Information Criterions (AIC), Root Mean Squared Errors (RMSE), and the greatest Coefficient of determination (R²) to predict VFA (1199, 29.5, 0.544 in female, and 1714, 41.3, 0.504 in male, respectively). Also, the DMW-NIRS estimated that VFA was highly performed to determine visceral obesity, which is comparable with other obesity surrogates.

Conclusions: This study suggested that lipid density can be used as a valid, noninvasive method to determine visceral obesity.