The SARC-F questionnaire has been suggested by the European Working Group on Sarcopenia in Older People (EWGSOP) as a first-step screening tool for sarcopenia. However, the sensitivity to SARC-F is low among community-dwelling older adults. Therefore, this study aimed to develop a new prediction model for sarcopenia screening in the community setting. We conducted a cross-sectional analysis of data from the Taiwan Integration of Health and Welfare (TIHW) study. Covariates including comorbidities, socioeconomic status, social support, health behaviors, body composition, and serum biomarkers were collected for analysis. Sarcopenia was defined using handgrip strength and gait speed cut-off values suggested by the Asian Working Group for Sarcopenia. Risk scores for sarcopenia were estimated by stepwise logistic regression. Among 1025 participants (mean age, 71.95 ± 6.89 years), 179 (17.5%) had sarcopenia. Seven items, including age, female sex, receiving social assistance pension, absence of exercise, being underweight, abnormal fasting glucose levels, and abnormal creatinine levels were selected for the Taiwan Risk Scores for Sarcopenia (TRSS) with a cutoff value of 76 (sensitivity, 71.8%; specificity, 71.1%) and area under the curve (AUC) of 0.757. Our results suggested that the TRSS model could be applied cost-effectively in the community for early detection of sarcopenia.
Keywords: SARC-F; body composition; handgrip; sarcopenia.