Development and validation of two anthropometric models estimating abdominal fat percentage in Greek adult women and men

Background and aims: Abdominal obesity is associated with adverse health implications. However, the assessment of abdominal adiposity is either indirect, like in the case of waist circumference measurement, or difficult in daily clinical practice, like in the case of using reference methods such as computerized tomography, magnetic resonance imaging, etc. The aim of the current study was to develop and validate, using DXA as reference method, two easy and simple to use protocols, that estimate abdominal fat mass percentage (%AFM) in Greek females and males respectively, based on simple anthropometric measurements.

Methods: Circumferences and skinfolds were obtained from 709 Caucasian adults (437 women and 272 men) in order to develop and validate the two equations. The validation and the reliability of the equations were examined with Bland-Altman analysis and Intraclass Correlation Coefficient (ICC).

Results: The equations were: for females, %AFM = -121,368 - (17,322 × Height(m)) + (76,452 × log Abdominal Circumference(cm)) + (18,341 × log Suprailiac Skinfold(mm)) + (8802 × log Abdominal Skinfold(mm)), for males, %AFM = -174,117 - (15,247 × Height(m))+ (104,503 × log Abdominal Circumference(cm)) + (9907 × log Suprailiac Skinfold(mm)) + (7971 × log Abdominal Skinfold(mm)). The equations had no significant bias (0,203% and 0,406%, respectively), acceptable limits of agreement (±9.358% and ±8.780%, respectively) and high reliability (0,940 and 0,957, respectively).

Conclusion: Both equations were found to be valid and reliable for Greek population. Therefore, the utilization of these equations for abdominal fat assessment could be an alternative of labor-intensive, expensive and time-consuming reference methods.