Skip to content
Medical InsightsKoreaMedical Insights Korea

Urinary Tract Infection

The diabetes drug that puts sugar in the urine did not raise urinary infection rates in 27,128 Koreans

A hazard ratio of 0.874, and a confidence interval running from 0.585 to 1.309. That interval is the reason this is reassurance rather than proof.

Reported by

Reporter

Sources
2 cited
Published
Reading time
6 minutes

In short

In a retrospective population-based cohort study using South Korean National Health Insurance Service data covering 27,128 patients with type 2 diabetes on metformin who added either a sodium-glucose cotransporter-2 inhibitor or another antidiabetic agent as dual therapy, of whom 1,274 received an SGLT2 inhibitor, the hazard ratio for urinary tract infection among SGLT2 inhibitor users compared with controls was 0.874 (95% CI 0.585 to 1.309, P = 0.510), with Kaplan-Meier analysis showing no significant difference in cumulative incidence between the groups.

Key takeaways

  • 27,128 Korean patients with type 2 diabetes already on metformin were studied using National Health Insurance Service data.
  • 1,274 of them added a sodium-glucose cotransporter-2 inhibitor; the rest added another antidiabetic agent.
  • The hazard ratio for urinary tract infection with SGLT2 inhibitor use was 0.874 (95% CI 0.585 to 1.309, P = 0.510).
  • Kaplan-Meier analysis found no significant difference in cumulative urinary tract infection incidence between the groups.
  • The confidence interval spans from a 41 percent reduction to a 31 percent increase, so a modest real effect in either direction is not excluded.
  • SGLT2 inhibitors work by increasing glucose excretion in urine, which is the mechanism behind the original concern.
  • The authors conclude the clinical impact on urinary tract infection risk is minimal while recommending continued monitoring and patient education.

The worry is mechanical and easy to state. SGLT2 inhibitors lower blood sugar by dumping glucose into the urine, and bacteria grow in sugar. Whether that translates into more urinary tract infections in practice is an empirical question, and Korean national data now answers it for a large routine-care population.

HR 0.874

hazard ratio for urinary tract infection among Korean SGLT2 inhibitor users against other second-line antidiabetic agents (95% CI 0.585 to 1.309)Jeon BJ et al., Journal of Korean Medical Science, 2025

A point estimate below 1, and why that is not the story

The hazard ratio came in at 0.874 — numerically fewer infections in the SGLT2 group — with a P value of 0.510. Nobody should read that as the drug protecting against urinary infection. The interval runs from 0.585 to 1.309, which is wide enough to contain a 41 percent reduction, a 31 percent increase and no effect at all.

The reason it is wide is visible in the design: 1,274 SGLT2 users against 25,854 comparators. An unbalanced cohort of that shape has limited power to detect a moderate effect, so the correct summary is that this study did not find an increase, not that it ruled one out.

What it does rule out

The comparison is also the right one. The control group was not untreated; it was patients adding a different second agent to metformin. That is the decision a clinician actually faces, and it removes the confounding that would follow from comparing treated diabetes against untreated diabetes.

What the claims data misses

Urinary tract infections that were self-managed, or treated without a coded diagnosis, do not appear. Nor does severity — a cystitis and a pyelonephritis enter the same way. And a channelling effect is plausible in both directions here: a clinician who already regarded a patient as infection-prone may have avoided prescribing an SGLT2 inhibitor to them, which would push the hazard ratio below 1 without the drug doing anything.

Read alongside the Korean finding that summer heat raises emergency admissions for urinary tract infection most sharply in people with diabetes — an odds ratio of 1.14 against 1.06 in the general population — the practical picture is that diabetes itself is the risk factor being managed, and the choice of second-line agent is a smaller lever than it has been made to sound.

Sources

  1. 1

    Jeon BJ, Ko YJ, Park JY, Bae JH, Tae BS

    Correlating Urinary Tract Infection Risks With Sodium-Glucose Cotransporter-2 Inhibitors: Insights From the National Health Insurance Data of Korea

    Journal of Korean Medical Science 2025;40(38):e244

    doi:10.3346/jkms.2025.40.e244 · PMID:41025341

    https://doi.org/10.3346/jkms.2025.40.e244
  2. 2

    Park J, Lee W, Kang D, Min J, Jang H, Kang C, Kwon D, Kwag Y, Ha E

    Summer temperature and emergency room visits due to urinary tract infection in South Korea: a national time-stratified case-crossover study

    BMC Public Health 2024;24(1):2274

    doi:10.1186/s12889-024-19454-1 · PMID:39169278

    https://doi.org/10.1186/s12889-024-19454-1