Background & Aims: Body weight is one of the essential indicators of nutritional status, and body weight management is vital in nutritional care. In addition, low body mass index (BMI) was included as a phenotypic criterion in the Global Leadership Initiative on Malnutrition (GLIM) criteria. Furthermore, low BMI has been used in grading the severity of malnutrition (moderate or severe malnutrition) in the GLIM criteria. A recent cross-sectional study reported that muscle mass of the quadriceps in older inpatients with severely low BMI is less than those of older inpatients with moderately low BMI and non-low BMI. However, the longitudinal relationship between body weight and muscle mass of the quadriceps in older inpatients in each BMI category according to the GLIM criteria remains unclear. This study aimed to examine the longitudinal relationship between body weight and muscle mass of the quadriceps in older inpatients in each BMI category according to the GLIM criteria., Methods: This retrospective cohort study included 179 older inpatients (aged ≥ 70 years) (median [IQR] age: 84.0 [79.0-89.0]). The period of this study was between January 2017 and March 2020. In accordance with the cut-off value of a low BMI for patients aged ≥ 70 years in the Asian population according to the GLIM criteria, the participants were divided into the following three groups: the severely low BMI group (< 17.8 kg/m 2 ) (n = 47), moderately low BMI group (≥ 17.8 to < 20.0 kg/m 2 ) (n = 38), and non-low BMI group (≥ 20.0 kg/m 2 ) (n = 94). The medians (IQR) of the length of hospital stay of the severely low BMI, moderately low BMI, and non-low BMI groups were 71.0 (49.0-98.0) days, 71.0 (50.0-98.0) days, and 50.5 (36.5-103.0) days, respectively. The primary outcome was a change in muscle mass of the quadriceps. The muscle mass of the quadriceps was examined using ultrasound images (i.e., quadriceps thickness). The changes in quadriceps thickness and body weight were calculated by subtracting the quadriceps thickness and body weight at admission from those values at discharge. Multiple linear regression analysis adjusting for confounding factors was used to determine whether the change in body weight was independently and significantly related to the change in quadriceps thickness in the severely low BMI, moderately low BMI, and non-low BMI groups., Results: The means (SD) of the change in quadriceps thickness of the severely low BMI group, moderately low BMI group, and non-low BMI group were 0.0 ± 0.3 cm, 0.1 ± 0.3 cm, and 0.1 ± 0.5 cm, respectively. The means of the change in body weight in those groups were 0.4 ± 2.8 kg, - 1.1 ± 2.7 kg, and - 1.3 ± 4.3 kg, respectively. In the severely low BMI group, the change in body weight (β = 0.34, p = 0.006) and quadriceps thickness at admission (β = -0.62, p < 0.001) were significantly and independently related to the change in quadriceps thickness (R 2 = 0.645, f 2 = 1.817, statistical power = 1.000). In the moderately low BMI and non-low BMI groups, there were no factors that were significantly and independently related to the change in quadriceps thickness., Conclusions: The results of this study suggest that change in body weight is positively related to the change in muscle mass of the quadriceps in older inpatients with severely low BMI according to the GLIM criteria. These results imply the importance of body weight management for older inpatients with severely low BMI perspective from the muscle mass of the quadriceps., (© 2024. The Author(s).)