BMJ Open · Oct 27, 2026

Association of urine pH with the progression in the individuals with prediabetes: a population-based cohort study in China

By Luo S, Wen Z, Huang W et al.

OBJECTIVES: To explore the association between urine pH (uPH) and the progression of pre-diabetes. DESIGN: Retrospective observational cohort study. SETTING: A health examination centre at the Third Affiliated Hospital of Sun Yat-sen University in Guangzhou, China, 2019–2023. PARTICIPANTS: A total of 6165 individuals with pre-diabetes were enrolled in the study. The diagnosis of pre-diabetes was based on the 2023 criteria of the American Diabetes Association. Individuals with normal blood glucose, diabetes mellitus (DM), abnormal renal function or conditions affecting urine test results were excluded. PRIMARY OUTCOME MEASURES: The primary outcome was progression to diabetes. uPH was measured using semiquantitative dipstick testing and categorised as <6.0, ≥6.0 to <6.5 or ≥6.5 at baseline. Repeated measurements were used to classify five longitudinal patterns: sustained ≥6.0, sustained <6.0, rise, drop and fluctuation. RESULTS: Over a median follow-up of 1.95 years (IQR 1.04–3.00), 616 participants (9.9%) progressed to DM. Baseline low uPH (<6.0) was independently associated with a significantly increased risk of progression (HR 3.15, 95% CI 2.33 to 4.25, p<0.001) compared with high uPH (≥6.5). Sustained low uPH over time conferred the highest risk (HR 1.97, 95% CI 1.31 to 2.96, p=0.001). The corresponding subdistribution HRs in the competing risk analysis were 2.25 (95% CI 1.63 to 3.09) and 1.81 (95% CI 1.24 to 2.65), respectively. Subgroup analyses revealed stronger associations in males, individuals with non-alcoholic fatty liver disease and those without hypertension. CONCLUSIONS: Low and sustained acidic uPH are independently associated with an increased risk of progression from pre-diabetes to diabetes in Chinese adults. As a routinely available, non-invasive marker, uPH may offer additional information for risk assessment in individuals with pre-diabetes.

doi.org/10.1136/bmjopen-2026-123363Open accessPubMed 42946369Populations