Long-Term Health Outcomes in Individuals With Confirmed Versus Unconfirmed Undiagnosed Diabetes Based on Oral Glucose Tolerance Test: Findings From the Tehran Lipid and Glucose Study.

Abstract

Background: To evaluate the prognostic performance of the single-sample confirmatory definition of undiagnosed diabetes on outcomes, including hypertension, chronic kidney disease (CKD), cardiovascular disease (CVD), and mortality, in the Tehran Lipid and Glucose Study cohort.

Methods: Among 6712 participants (3682 women) aged ≥ 30 years not on glucose-lowering medication, unconfirmed undiagnosed diabetes was defined as either fasting plasma glucose (FPG) ≥ 7 mmol/L or 2-h postload glucose (2-h PG) ≥ 11.1 mmol/L, whereas confirmed undiagnosed diabetes required both. Multivariable Cox models estimated hazard ratios (HRs) and 95% confidence intervals (CIs) for each outcome.

Results: At baseline, 3.6% and 3.3% of participants had confirmed or unconfirmed undiagnosed diabetes, respectively. During a median follow-up of over 10 years, both unconfirmed and confirmed undiagnosed diabetes were associated with a higher risk of developing treated diabetes, with HRs (95% CI) of 7.73 (6.35-9.44) and 17.30 (14.32-20.83), respectively. Confirmed undiagnosed diabetes was further associated with incident CKD (1.33 [1.02-1.73]), hypertension (1.26 [1.00-1.60]), first CVD (1.40 [1.01-1.93]), and hard CVD (1.65 [1.04-2.61]), whereas the association with all-cause mortality did not reach statistical significance (1.50 [0.96-2.33], p = 0.07). Unconfirmed undiagnosed diabetes was only associated with treated diabetes and first CVD (1.53 [1.11-2.12]). Isolated FPG and 2-h PG elevations showed similar risks for treated diabetes (9.00 [6.26-13.00] and 7.40 [5.92-9.24], respectively), but only isolated FPG was significantly associated with first CVD (2.55 [1.50-4.31]) and hard CVD (2.43 [1.06-6.00]), comparable with confirmed undiagnosed diabetes.

Conclusions: Confirmed undiagnosed diabetes was associated with a higher risk of cardiovascular and renal outcomes. The unconfirmed undiagnosed phenotype shared a similar CVD risk with confirmed diabetes but had a 50% lower risk of treated diabetes, with a stronger association observed for isolated FPG elevation than for isolated 2-h PG.