J Diabetes. 2026 Aug;18(8):e70260. doi: 10.1111/1753-0407.70260.
ABSTRACT
BACKGROUND: The combined prognostic value of continuous glucose monitoring (CGM) metrics and circulating glucose biomarkers for predicting mortality in type 2 diabetes has not been fully established, particularly regarding residual risk in patients achieving glycemic targets.
METHODS: This cohort study included 3677 patients with type 2 diabetes for a median of 7.4 years follow-up. Cox proportional hazards models evaluated the associations of baseline CGM-derived time in range (TIR) (target 70%) and serum 1,5-anhydroglucitol (1,5-AG) (threshold 6.0 μg/mL) with all-cause and cardiovascular mortality in the overall population. We further examined the relationship between 1,5-AG and mortality within TIR subgroups and compared its performance with metrics of glycemic variability derived from CGM.
RESULTS: During follow-up, 522 all-cause deaths and 181 cardiovascular deaths occurred. TIR and 1,5-AG were moderately correlated and independently predicted mortality, with concurrent low TIR (≤ 70%) and low 1,5-AG (< 6.0 μg/mL) yielding the highest risk (hazard ratio [HR] 1.82, 95% CI 1.40-2.37). In stratified analyses, reduced 1,5-AG was significantly associated with increased mortality risk in patients with TIR > 70% (HR 1.69, 95% CI 1.25-2.28), whereas no significant association was observed in those with TIR ≤ 70%. Adding 1,5-AG improved traditional risk prediction in the former subgroup. Compared with 1,5-AG, CGM-derived glycemic variability indices such as mean amplitude of glycemic excursions, coefficient of variation, and standard deviation of glucose showed no significant association with mortality.
CONCLUSIONS: TIR and serum 1,5-AG offer independent and complementary value, with low 1,5-AG identifying residual mortality risk despite achieving TIR targets.
PMID:42552233 | DOI:10.1111/1753-0407.70260