<scp>BALLAST</scp> study: A multicentre, open‐label, randomized‐controlled, 52‐week clinical trial of the efficacy and safety of luseogliflozin in older Japanese adults with type 2 diabetes receiving leucine‐enriched amino acid supplementation and ph

Older adults are a rapidly growing segment of the diabetes population, with unique challenges including greater comorbidities and high risk of sarcopenia. Sodium–glucose co-transporter-2 inhibitors (SGLT2is) offer effective glycaemic control with weight loss and low hypoglycaemia risk, making them attractive for older adults with type 2 diabetes (T2D). However, concerns have been raised that SGLT2is may accelerate muscle loss and exacerbate sarcopenia.1, 2 In older T2D adults, sarcopenia is prevalent and associated with poor outcomes. Nutritional and exercise interventions — especially leucine-enriched amino acid supplementation (LEAAS), combined with resistance exercise — can help preserve muscle mass and function.3, 4 Current diabetes care guidelines in Japan emphasize strategies to prevent frailty and muscle loss in older adults with diabetes.5 Accordingly, we conducted the BALLAST study to evaluate the 52-week efficacy and safety of the SGLT2i luseogliflozin in older T2D adults, with all participants receiving LEAAS and an exercise program. BALLAST was a 52-week, multicentre, open-label randomized controlled trial conducted in Japan (jRCTs041210048). Eligible participants were Japanese adults ≥65 years old with T2D, body mass index (BMI) 18–30 kg/m2 and glycated haemoglobin (HbA1c) 7.0–9.0% if on diet/exercise alone or 6.0%–8.0% if on a single oral agent. Comprehensive eligibility criteria, study procedures and endpoints are described in the Supplemental methods. To avoid enrolling frail, sarcopenic individuals, we excluded those with low skeletal muscle mass (SMI <7.0 kg/m2 for men or <5.7 kg/m2 for women) and/or low muscle strength (grip strength <28 kg for men or <18 kg for women) at baseline, according to Asian Working Group for Sarcopenia 2019 consensus criteria.6 Participants were randomized 1:1 to the luseogliflozin group or the control group. Both groups received identical nutritional and exercise interventions: a daily LEAAS (AminoL®, Ajinomoto Co., Inc.) and a standardized home-based exercise program emphasizing resistance and balance training. Adherence to LEAAS and exercise was monitored via daily logbooks. The primary endpoint was the change in HbA1c from baseline to week 52. Key secondary endpoints included changes in body weight (BW), body composition and physical function, evaluated using handgrip strength and the time required to complete the 5-times sit-to-stand test. Dietary intake was assessed using a food frequency questionnaire and daily protein intake was calculated per ideal body weight (IBW), defined as height (m)2 x 22. Safety assessments, including the incidence of adverse events (AEs), serious AEs and predefined AEs of special interest (hypoglycaemia, diabetic ketoacidosis, lower-limb amputation and volume depletion), were performed in the safety set, which comprised all randomized participants who received LEAAS and exercise program at least once. Efficacy analyses were performed using the full analysis set (FAS), which included all randomized participants who received LEAAS and exercise program at least once, excluding those who were retrospectively deemed ineligible at the time of consent. Analyses were also conducted in the per-protocol set (PPS), a subset of FAS, which excluded participants who were randomized with an incorrect stratification factor (BMI or HbA1c), those who required rescue therapy and those with inappropriate modifications to their treatment regimen. Statistical analyses are described in the Supplemental methods. For the primary endpoint, a t test was used to compare the two groups. A two-way ANOVA was performed to test interactions, with treatment group, factor and their interaction included as covariates. A total of 91 individuals were screened and 85 were randomized (44 luseogliflozin, 41 control) (Figure S1). After excluding participants with withdrawal consent or with protocol violations, 77 participants (39 luseogliflozin, 38 control) were included in FAS and 65 participants (33 luseogliflozin, 32 control) in PPS. Baseline characteristics were comparable between groups (Tables S1 and S2). By design, all participants were relatively healthy older adults without baseline sarcopenia or frailty. Luseogliflozin improved HbA1c over 52 weeks in FAS and PPS. HbA1c declined in the luseogliflozin group while remaining stable in the control group (Figures 1 and S2–S4; and Tables 1 and S3). The mean HbA1c in FAS had decreased by 0.36% more in the luseogliflozin group than in the control group (between-group difference −0.36%, 95% confidence interval [CI] −0.60 to −0.12; p = 0.004). BW also decreased more with luseogliflozin in FAS and PPS (Figures 1 and S2–S4; and Tables 1 and S3): participants on luseogliflozin lost on average 1.15 kg more than those on control in FAS (between-group difference −1.15 kg, 95% CI −2.13 to −0.17; p = 0.025). The HbA1c reduction in FAS was consistent across baseline subgroups (age, T2D duration, eGFR, BMI), with a significant reduction among females (Figure S5). BW effects were also consistent across subgroups in FAS (Figure S5). Total appendicular muscle mass declined slightly in both groups over 52 weeks, with no significant difference between luseogliflozin and control in FAS and PPS (Figures 1 and S2–S4; and Tables 1 and S3). Regionally, luseogliflozin was associated with small but significant losses in upper-limb and trunk muscle mass versus control, while lower-limb muscle mass was preserved (Tables 1 and S3). Handgrip strength showed no significant change from baseline in either group and performance on the five-times sit-to-stand test was unchanged (Figures 1 and S2–S4; and Tables 1 and S3). Energy intake increased slightly in the luseogliflozin group and decreased slightly in the control group in FAS and PPS (Tables S4 and S5). Protein intake increased modestly in both groups during the study (Tables S4 and S5). At baseline, 37% of participants had protein intake ≥1.2 g/kg IBW/day in FAS; by Week 52 this proportion increased to about 50% in both groups, indicating improved nutritional intake over the study period. Median adherence to the LEAAS and exercise programme was high (Figure S3), supporting that the co-interventions were well implemented. Luseogliflozin was generally well tolerated. There were no occurrences of hypoglycaemia, ketoacidosis, lower-limb amputation or volume depletion events in either group (Table S10). Overall AEs were reported in 12.5% of participants in the luseogliflozin group versus 7.5% in the control group; most AEs were mild and unrelated to the study interventions. Over 52 weeks in older Japanese T2D adults, luseogliflozin combined with LEAAS and exercise produced significant improvements in HbA1c and BW without causing significant deterioration in muscle mass, strength, or physical function. These findings, in line with the recently published EMPA-ELDERLY trial,7 suggest that SGLT2is can be used safely in older T2D adults when accompanied by interventions to support muscle health. It should be emphasized that muscle-related outcomes in our study were secondary and exploratory; the trial was not powered to detect small changes in muscle metrics, and the absence of a significant between-group difference should not be interpreted as proof of no effect. Additionally, the BALLAST population consisted of relatively healthy, independent older adults and explicitly excluded those at high risk of sarcopenia; therefore, these results may not be generalizable to frailer or more sarcopenic populations.8 Changes in glucose-lowering therapy within 4 weeks prior to randomization, the open-label design, and the allowance of background medication adjustments represent additional limitations that may have introduced bias in glycaemic outcomes. In summary, our study provides evidence that SGLT2i can improve glycaemic control and induce weight loss in older T2D adults without precipitating muscle loss when implemented alongside nutritional and exercise interventions. Larger and longer term studies in more vulnerable populations are warranted to confirm these benefits and ensure safety with respect to sarcopenia and functional decline. Study Concept and Design: Y.T., K.M. and D.Y. Data Acquisition and Interpretation: Y.T., M.H., T.K., Y.H., S.T., H.K., Y.S., R.O., K.M., D.Y. and BALLAST Investigators. Manuscript Preparation: Y.T., K.M. and D.Y. Intellectual Contributions to Discussion: Y.T., M.H., T.K., Y.H., S.T., H.K., Y.S., R.O., K.M., D.Y. All authors reviewed and approved the final version of the manuscript. The authors thank BALLAST Investigators including Drs. M. Sakai, So. Kubota, Sa. Kubota, K. Takao, M. Mizuno and T. Hirota of Gifu University, as well as Drs. H. Niwa of Medical corporation Yuuaikai Iwasa Hospital Iwasa Maternity, S Kawachi of Japanese Red Cross Gifu Hospital, K. Takami of Central Japan International Medical Center, N. Fushimi of Ichinomiya Nishi Hospital, K. Nagata of Sawada Hospital, S. Ito of Keyaki Clinic, T. Shibata of Japanese Red Cross Takayama Hospital, T Shimada of Chuno Kosei Hospital, A. Fushimi of Ichinomiya Nishi Hospital, K. Sasaki of Asahi University Hospital, T. Horiya of Horiya Clinic, H. Daido of Gifu Prefectural General Medical Center, K. Yamada of Gifu Municipal Hospital. The authors also thank EviPro Co., Ltd. (Tokyo, Japan) for their professional support as the contract research organization in this study; Dr. R. Asada of Gifu University Hospital for his dedicated support in data management throughout this study; and Yuuka Fujiwara of Kansai Electric Power Medical Research Institute for her dedicated support in measuring IL-6 and TNF-α. The authors also thank M. Yato and Y. Ogiso for their valuable secretarial support. We extend our sincere gratitude to all healthcare professionals and study participants who contributed to this research. Their invaluable cooperation and commitment made this study possible. This study was funded by Taisho Pharmaceutical Co., Ltd. (Tokyo, Japan). D.Y. received clinically commissioned/joint research grants from Nippon Boehringer Ingelheim, Novo Nordisk, Ono Pharmaceutical, Taisho Pharmaceutical, Tanabe Pharma, Terumo and Arklay. D.Y. also received consulting or speaker fees from Sumitomo Pharma, Nippon Boehringer Ingelheim, Novo Nordisk, Ono Pharmaceutical, Eli Lilly, Sanofi and Tanabe Pharma. T.K. received consulting or speaker fees from Eli Lilly, Novo Nordisk and Tanabe Pharma. Y.H. received clinically commissioned/joint research grants from Nippon Boehringer Ingelheim. Y.H. also received consulting or speaker fees from Sumitomo Pharma. Y.S. received grants from Nippon Boehringer Ingelheim, ARKRAY, Taisho Pharmaceutical, Novo Nordisk, Terumo, and Sumitomo Pharma, and honoraria for lectures from Taisho Pharmaceutical, Nippon Becton Dickinson, Novo Nordisk, Eli Lilly, Sumitomo Pharma and Ono Pharmaceutical. Y.T., M.H., S.T., H.K., R.O. and K.M. declare no conflict of interest. Approval of the research protocol: The protocol for this research protocol was initially approved by the Nagoya University Clinical Research Review Board (Approval no. N0053) and subsequently by the independent ethics committees or institutional review boards at each participating site, and the research was conducted in accordance with the principles of the Declaration of Helsinki, with written informed consent. Informed Consent: Written informed consent was obtained. Approval date of Registry and the Registration No. of the study/trial: July 6, 2021 and Registry ID: jRCTs041210048. Animal Studies: N/A. The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/dom.70494. Data are available upon reasonable request. Table S1. Baseline demographic and clinical characteristics of participants in the full analysis set. Table S2. Baseline demographic and clinical characteristics of participants in the per protocol set. Table S3. Changes from baseline at week 52 in HbA1c, body weight, body composition and muscle strength in the per protocol set. Table S4. Changes from baseline in estimated energy intake in the full analysis set. Table S5. Changes from baseline in estimated energy intake in the per protocol set. Table S6. Stratified analysis by age on changes from baseline at week 52 in HbA1c, body weight, body composition and muscle strength in the full analysis set. Table S7. Stratified analysis by age on changes from baseline at week 52 in HbA1c, body weight, body composition and muscle strength in the per protocol set. Table S8. Changes from baseline at week 52 in laboratory parameters the full analysis set. Table S9. Changes from baseline at week 52 in laboratory parameters the per protocol set. Table S10. Summary of adverse events in the Safety Set. Supplemental methods. List of BALLAST Investigators. Figure S1. Flow of study participants. Figure S2. Time course of HbA1c, body weight, fat mass, muscle mass, total body water, grip strength and the five-times sit-to-stand test from baseline to Week 52 in the full analysis set. Figure S3. Changes from baseline at Week 52 in HbA1c, body weight, fat mass, muscle mass, total body water, grip strength and the five-times sit-to-stand test in the per protocol analysis set. Figure S4. Time course of HbA1c, body weight, fat mass, muscle mass, total body water, grip strength and the five-times sit-to-stand test from baseline to Week 52 in the per protocol set. Figure S5. Subgroup analyses of HbA1c and body weight change from baseline to week 52 in the full analysis set. Figure S6. Association of muscle mass and grip strength with AminoL® and exercise adherence in the full analysis set. Figure S7. Subgroup analyses of muscle mass and handgrip strength change from baseline to week 52 in the full analysis set. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.