Introduction: There is a gap in the literature both in terms of the definition of the obesity component of sarcopenic obesity (SO) and the relationship between SO and functionality. We aimed to explore the ideal obesity definition in the context of SO for its relationship with functional measures.

Methods: We included community-dwelling adults aged > 60 between 2013–2021 in this retrospective cross-sectional study. We recorded age, sex, number of diseases & medications, and assessed nutritional status by mini-nutritional assessment-short form. We measured handgrip strength by Jamar dynamometer and defined the sarcopenia component as low muscle strength (probable sarcopenia). Obesity was determined with three different methods: body mass index (BMI), fat percentage estimated by bioimpedance analysis, and waist circumference. Functionality was assessed with Katz’s activities of daily living (ADL) and Lawton Brody’s instrumental ADL (IADL) scales. We performed multivariate analyses adjusted for age, sex, and nutritional status to determine the association between SO and the function of the measures.

Results: We included 1390 participants [mean age: 73.7 ± 6.8 (range: 60–98 years), 68.8% female]. The prevalences of SO (probable) were 12.4%, 11.7%, and 19%, by BMI, fat percentage, and waist circumference methods, respectively. For all probable SO definitions, impaired ADL was more prevalent in the “sarcopenic alone” and “probable SO” phenotypes compared to the reference standard “non-sarcopenic non-obese" phenotype (p < 0.01, for all). Impaired IADL was common in the “sarcopenic alone” and “probable SO” phenotypes compared to “non-sarcopenic non-obese" and “obese alone” phenotypes (p < 0.001, for all). “Probable SO” phenotype was independently associated with impaired ADL and IADL and had the highest odds ratio of 2.9, 2.2, and 2.8 for impaired ADL and 4.6, 3.3, and 3.8 for impaired IADL by BMI, fat percentage, and waist circumference methods, respectively when compared to “non-sarcopenic non-obese” phenotype in multivariate analyses.

Conclusions: Our study showed that sarcopenia demolished the beneficial functional associations of obesity in older adults. This study suggests that while considering obesity, sarcopenia should be a major component and determinant to be looked for. In addition, BMI and waist circumference methods which are much practical than the fat percentage method, came front while assessing the obesity component of SO in terms of its associations with functional measures.