The clinical characteristics of the FAS at baseline are shown in Table 1. After matching, there were no significant differences between the two groups in age, sex, smoking history, current alcohol consumption, obesity, duration of diabetes mellitus, or percentage of cases with hypertension or hyper-LDL-C. There were also no significant differences in the HbA1c, weight, BMI, blood pressure, eGFR, uACR, AST, ALT, high-density lipoprotein cholesterol (HDL-C), and LDL-C values. Baseline uric acid levels averaged 4.9 ± 0.9 mg/dL in the low and 6.7 ± 0.5 mg/dL in the high group and were significantly higher in the high group. There were no differences between the two groups in the number of diabetes medications used, the percentage of each diabetes medication used, or the percentage of RAS inhibitors used. In the FAS, no patients in either group were using diuretics. Only allopurinol and febuxostat were used as uric acid-lowering medications, and there was no difference in their use between the two groups of 7 patients. Regarding urinary albumin, the proportion of cases with microalbuminuria was 39.8% in the low group and 39.0% in the high group, while the proportion of cases with macroalbuminuria was 5.2% in the low group and 4.5% in the high group, with no difference between the two groups. The low group had a significantly higher number of patients with coronary artery disease (p = 0.041), and although this difference was not significant, the low group also tended to have a higher number of patients with cerebrovascular disease (p = 0.097).

Table 2 shows the changes in clinical parameters of the FAS from baseline to 12 months after the initiation of luseogliflozin. Both groups showed significant reductions in uric acid levels at 12 months (low group: p < 0.001, high group: p < 0.001) compared with pre-treatment levels. Furthermore, both groups showed significant improvements in AST, ALT, γ-GTP, uACR, and HDL-C, in addition to decreases in HbA1c, weight, and BMI. However, with regard to blood pressure, only the high group showed a significant decrease in both systolic and diastolic blood pressure, while the low group showed no significant change. In terms of eGFR, propensity score matching showed that baseline eGFR before luseogliflozin was 77.3 mL/min/1.73m2 in the low group and 75.6 mL/min/1.73m2 in the high group, with no difference between the two groups (p = 0.462). Despite this, the 12-month change in eGFR (Δ eGFR) in the high group was − 0.8 ± 1.3 mL/min/1.73m2 (p = 0.667 vs. baseline eGFR), showing no significant change over 12 months, while the Δ eGFR from baseline in the low group was − 3.1 ± 0.9 mL/min/1 0.73m2 (p < 0.001 vs. baseline eGFR), a significantly greater decrease. However, as shown in Table 3, no significant differences were observed between the two groups in the degree of change from baseline for eGFR (p = 0.667). Similarly, changes in HbA1c, body weight/BMI, blood pressure, and uACR did not differ significantly between the groups. The change in uric acid levels (Δ uric acid) was the only parameter that showed a significant difference between the groups, with a greater reduction observed in the high group compared to the low group (low group: − 0.3 ± 0.1 mg/dL; high group: − 1.0 ± 0.2 mg/dL; p < 0.001).

Figure 2 shows the trend of uric acid levels and eGFR during 12 months in both groups. In both groups, uric acid levels decreased significantly after 1 month of luseogliflozin administration and remained so for 12 months.

In the two subgroups (A: low group, B: high group), serum uric acid concentrations were measured at baseline and at multiple time points during the 12-month period. Both the low and high group showed significant reductions in uric acid levels starting at 1 month and persisting at 12 months. Data are presented as mean ± SD; error bars indicate SD. *P < 0.05, **P < 0.01, ***P < 0.001 vs. baseline

Figure 3 shows the trend of eGFR. The high group showed no significant change in eGFR during the 12-month period, while the low group showed a significant decrease in eGFR from the first month of treatment and maintained the decreased level throughout the 12-month period.

In the two subgroups (A: low group, B: high group), eGFR were measured at baseline and at multiple time points during the 12-month period. The high group showed no significant change in eGFR throughout the 12 months, while the low group showed a significant decrease in eGFR beginning at 1 month. Data are presented as mean ± SD; error bars indicate SD. *P < 0.05, **P < 0.01, ***P < 0.001 vs. baseline

Table 4 shows the results of a multivariate analysis of the FAS regarding the relationship with the baseline clinical parameters, using Δ eGFR as the objective variable. Multiple regression analysis was performed on the following explanatory variables: sex, age, smoking history, current alcohol consumption, BMI, blood pressure, baseline HbA1c, Δ HbA1c, baseline eGFR, Δ eGFR, baseline uric acid level, uACR, Δ uACR, and the use of antihyperuricemic agents. Δ eGFR was significantly affected by baseline eGFR, uACR, Δuric acid, current drinker as independent variables.

Table 5 is used as in Table 4, the results of a multivariate analysis of the FAS regarding the relationship with the baseline clinical parameters, using Δ uric acid as the objective variable. Δ Uric acid was significantly affected by pre-administration uric acid levels, Δ eGFR, and sex as independent variables. Δ eGFR and Δuric acid showed no correlation with the degree of change in HbA1c (Δ HbA1c). Similarly, it did not correlate with the degree of change in body weight.

In Fig. 4, panel A shows the relationship between Δ eGFR and Δ uric acid. The regression coefficient was − 1.08 (95% CI − 1.32 to − 0.84; p = 0.008; R2 = 0.362), indicating that greater increases in Δ uric acid were associated with smaller Δ eGFR. Panel B shows the correlation between Δ uric acid and baseline uric acid levels before luseogliflozin administration. The regression coefficient was − 0.56 (95% CI − 0.77 to − 0.37; p < 0.001; R2 = 0.479), indicating that higher baseline uric acid levels were associated with greater Δ uric acid.

A Relationship between the degree of change in uric acid level (Δ uric acid) and the degree of change in eGFR (Δ eGFR). B Relationship between the degree of change in uric acid level and baseline uric acid level before luseogliflozin administration. (The Δ uric acid and Δ eGFR presented in A showed a significant correlation. Similarly, B Δ uric acid and baseline uric acid levels showed a significant correlation. R2 presented shows the correlation coefficient

We also conducted analyses excluding cases who were using antihyperuricemic agents from both the low and high uric acid groups.

In these cases, the baseline serum uric acid levels were 4.9 ± 1.2 mg/dL in the low group and 6.9 ± 0.8 mg/dL in the high group, with significantly higher levels observed in the high group (p < 0.001). Both groups showed significant reductions in uric acid levels at 12 months compared with their respective baseline values (low group: p < 0.001, high group: p < 0.001).

Regarding eGFR, baseline values were 76.5 ± 17.4 mL/min/1.73 m2 in the low group and 75.1 ± 19.8 mL/min/1.73 m2 in the high group, with no significant difference between the two. At 12 months, however, only the low group showed a significant change in eGFR (low group: p < 0.001, high group: p = 0.730).

In the comparison of changes in clinical parameters between the two groups shown in Table 3, the Δ uric acid was − 0.4 ± 0.2 mg/dL in the low group and − 1.0 ± 0.4 mg/dL in the high group, with a significantly greater reduction observed in the high group (p < 0.001). As for Δ eGFR, it was − 2.9 ± 0.9 mL/min/1.73 m2 in the low group and − 0.9 ± 1.4 mL/min/1.73 m2 in the high group, with no statistically significant difference between the groups (p = 0.101). Multivariate analysis, as shown in Table 4, revealed that Δ eGFR was significantly associated with baseline eGFR, baseline uACR, Δ uric acid, and alcohol consumption status as explanatory variables.

Furthermore, when Δ uric acid was used as the dependent variable in the multivariate analysis, as in Table 5, it was significantly affected by baseline uric acid levels, Δ eGFR, and sex as independent variables.

In the SAS, luseogliflozin was discontinued in 24 patients (14%) because of the development of AEs. Within 100 days of starting luseogliflozin, AEs were observed that could be related to luseogliflozin administration, including urogenital infection, increased urine output, decreased fluid volume, skin rash, and gastrointestinal symptoms. In addition, luseogliflozin was stopped in 10 patients (4%) because it was judged to be ineffective at the discretion of the attending physician, but the frequency of discontinuation did not differ between the high and low groups.