RESUMEN
Body image adjustments after a limb loss are essential to rehabilitation. This study evaluated the psychometric properties of the Brazilian Portuguese version of the Amputee Body Image Scale (ABIS). We analyzed data from a nonprobability sample of 100 participants (71% men, mean age 48.26 ± 18.35 years) by using a confirmatory factor analysis with partial least square path modeling. We evaluated the structural model by using Pearson's coefficient of determination, path coefficients, indicator of Stone-Geisser, and Cohen's indicator. We evaluated item factor loadings and average variance extracted for model measurement. We tested the internal consistency with Cronbach's α test and composite reliability and evaluated the discriminant validity by adopting Fornell and Larcker (1981) criteria. We tested the original and a new three-factor model and performed correlational and variance analyses as additional tests. The new three-factor theoretical model had stronger structural support and better discriminant and convergent validities than the original three-factor model. The ABIS total score was inversely correlated with body appreciation. There were significant body image score differences between the groups of sedentary and physically active amputees and among those with differences in years since amputation, but there were no differences between those with different causes of amputation nor for those with different amputated limbs.