Evaluation of indirect methods for estimating body fat in dogs using electrical bioimpedance and morphometry
Daniel Montoya-Barbosa, Clara Helena González-Correa, William Narváez-SolarteAbstract
Canine obesity causes metabolic issues and reduces quality of life. Assessing adiposity remains challenging due to limited indirect and reference methods. Electrical bioimpedance (BIA) is a proposed tool, but agreement between morphometric and BIA equations, and the body condition score (BCS) role, remains poorly explored in heterogeneous canine populations.
Objective
To evaluate agreement among indirect methods (BCS, morphometry, BIA) for estimating canine body fat and assess BIA's viability for nutritional evaluation.
Materials and methods
A cross-sectional study included 17 healthy dogs from Villamaría, Colombia. Body condition was assessed using 5- and 9-point scales. Morphometric variables and single-frequency 50 kHz BIA parameters (resistance, reactance, phase angle) were recorded. Body fat percentage (%BF) was estimated using four established equations. Method agreement was evaluated via Bland–Altman analysis and Lin's concordance correlation coefficient (CCC). Differences in bioelectrical parameters and lipid profiles between obese and lean dogs were analyzed using independent t-tests.
Results
Morphometric methods showed the highest agreement (CCC = 0.759). Conversely, comparisons involving BIA-derived equations exhibited low or negative agreement, with Bland–Altman analyses revealing wide limits of agreement. Obese dogs had higher resistance and lower phase angles than lean dogs, though differences were not statistically significant (p > 0.05). This lack of significance must be interpreted cautiously due to the small sample size, particularly the obese group (n = 4). HDL cholesterol was significantly lower in obese dogs (p = 0.030), with no differences in total cholesterol.
Conclusion
The evaluated indirect methods showed poor agreement and are not clinically interchangeable. Without a reference method like DEXA, accuracy cannot be determined. While raw BIA parameters aligned physiologically with body condition, their use remains preliminary and requires validation in larger, reference-controlled studies.