Anonymized data are publicly available for purchase from JMDC, Inc.

This retrospective cohort study used the JMDC Claims Database, a large-scale administrative claims database [16,17,18] that uses data from annual employee health checkups which are a legal requirement in Japan. The JMDC includes annual health checkup data (e.g., blood tests and anthropometric measurements) and health insurance records between 2005 and 2022. Medical diagnoses were coded according to the International Classification of Diseases, 10th revision (ICD-10). This study adopted an active comparator, a new user design to account for confounding by indication and unmeasured confounders (Supplementary Fig. 1) [19]. The active comparator group was composed of individuals who initiated treatment with dipeptidyl peptidase-4 inhibitors (DPP4i). Considering the relatively high prescription rate of DPP4i for individuals with diabetes in Japan [20], individuals who were newly prescribed DPP4i were set as the control group in this study. We extracted the data of 21,492 individuals with diabetes (ICD-10 codes E10–E14), without a prior diagnosis of hypertension (ICD-10 code: I10-I15) and without antihypertensive medications, who had newly initiated SGLT2i or DPP4i treatment. To exclude previous users, new usage was defined as starting either drug class in those who had never used either drug class in the past year. Among the 21,492 individuals, we excluded participants for the following reasons: age <20 years (n = 1); prior diagnosis of cardiovascular diseases such as myocardial infarction, angina pectoris, stroke, heart failure, and atrial fibrillation (n = 1163); missing cigarette smoking data (n = 226); missing alcohol consumption data (n = 1058); and missing physical activity data (n = 444). Ultimately, 18,600 individuals were included in this study (Fig. 1).

Flowchart. We extracted the data of 21,492 individuals with diabetes, without a prior diagnosis of hypertension and without antihypertensive medications, who newly initiated sodium-glucose cotransporter 2 inhibitors (SGLT2i) or dipeptidyl peptidase-4 inhibitors (DPP4i). Among 21,492 individuals, we excluded participants for the following reasons: aged <20 years (n = 1); a prior diagnosis of cardiovascular diseases such as myocardial infarction, angina pectoris, stroke, heart failure, and atrial fibrillation (n = 1163); missing cigarette smoking data (n = 226), missing alcohol consumption data (n = 1058), and missing physical activity data (n = 444) Finally, 18,600 individuals were included in this study

This study was approved by the Ethics Committee of the University of Tokyo (approval number:2018-10862), and informed consent was waived because all data included in the JMDC Claims Database were anonymized and de-identified.

We obtained the following data from the health checkups: body mass index (BMI), BP, laboratory data (fasting plasma glucose, hemoglobin A1c [HbA1c], low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides), cigarette smoking (current or noncurrent/never), alcohol consumption (daily or not every day), and physical activity (active or inactive). Cigarette smoking and alcohol consumption were assessed using a self-report questionnaire during health checkups. Physical inactivity was defined as not exercising for 30 min ≥twice a week or not walking for more than an hour per day. Based on the ICD-10 codes, we obtained data on the presence of diabetic nephropathy (ICD-10 codes E102, E112, E122, E132, and E142), diabetic retinopathy (ICD-10 codes E103, E113, E123, E133, and E143), and diabetic neuropathy (ICD-10 codes E104, E114, E124, E134, and E144) at the date of prescription of SGLT2i or DPP4i. Data on concomitant medications at the prescription date of SGLT2i or DPP4i were extracted from the administrative claims records.

A propensity score-matching algorithm was used to generate a matched cohort to compare the benefits of SGLT2i and DPP4i use. We estimated the propensity scores of the SGLT2i users using a logistic regression model. To estimate the propensity score, we included the following variables: age, sex, BMI, SBP, diastolic blood pressure (DBP), fasting plasma glucose, HbA1c, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, cigarette smoking, alcohol consumption, physical inactivity, diabetic nephropathy, diabetic retinopathy, diabetic neuropathy, use of medications (insulin, glucagon-like peptide-1 receptor agonist, biguanide, sulfonylurea, α-glucosidase inhibitor, thiazolidine, glinide, and statins), and year of SGLT2i or DPP4i prescription. SGLT2i and DPP4i users were matched using a 1:1 matching protocol (caliper width equal to 0.2 standard deviations of the logit score).

The primary outcome was incident hypertension (ICD-10 codes I10-I15). We followed the study participants from the index date (i.e., initiation of SGLT2i or DPP4i) to the incidence of hypertension, discontinuation of insurance, death, or study end date (May 2022). We did not complete the follow-up even if an SGLT-2i or DPP4i is initiated and the other is prescribed additionally.

Descriptive statistics were reported as median (interquartile range) and number (percentage). Standardized mean differences were used to compare the clinical characteristics of SGLT2i and DPP4i users. The hazard ratios (HRs) and 95% confidence intervals (95% CI) of the incidence of hypertension in SGLT2i users versus DPP4i users were estimated using a Cox proportional hazards regression model. We also conducted subgroup analysis by age (≥50 and <50 years), sex, median BMI value (≥26.4 and <26.4 kg/m2), median HbA1c level (≥7.5 and <7.5%), and median SBP value (≥127 and <127 mmHg).

Seven sensitivity analyses were conducted. First, we examined the incidence of hypertension in individuals who continued to use SGLT2i or DPP4i for >3 months. Second, we analyzed individuals with a diagnosis of type 2 diabetes (ICD-10 code E11). Third, we excluded individuals with SBP ≥ 140 mmHg or DBP ≥ 90 mmHg at the initial health checkup. Fourth, the outcome was redefined as a diagnosis of hypertension with a prescription for antihypertensive medications (World Health Organization Anatomical Therapeutic Chemical [WHO-ATC] codes C02, C03, C04, C07, C08, and C09) in the months before and after the diagnosis of hypertension. Fifth, we excluded individuals with glucagon-like peptide-1 receptor agonist at the index date. Sixth, we repeated the primary analyses after excluding individuals who had any antidiabetic medications at the index date. Finally, we performed an analysis using overlap weighting to balance the exposure groups (SGLT2i and DPP4i). Overlap weights were defined as 1−propensity score among SGLT2i users and as the propensity score among DPP4i users. All statistical analyses were performed using STATA version 17 (StataCorp LLC, College Station, TX, USA).