Association of glycemic and metabolic biomarkers with cognitive impairment and dementia among elderly patients with diabetes: a systematic review and a dose-response meta-analysis

The association between glycemic and metabolic biomarkers and cognitive impairment and dementia among elderly diabetic patients remains a subject of debate. We systematically searched PubMed, Embase, Cochrane Library, and Web of Science from 2000 to March 13, 2026. We included cross‑sectional or cohort studies of older adults with diabetes aged ≥ 60 years without cognitive impairment or dementia at baseline that reported associations of glycated hemoglobin (HbA1c), fasting plasma glucose (FPG), the triglyceride‑glucose (TyG) index, or adiponectin with cognitive impairment and dementia. Random‑effects models were used to pool effect sizes for each biomarker in conventional meta‑analyses. For FPG, a non‑linear dose‑response meta‑analysis was conducted using restricted cubic splines to assess the dose‑dependent association with dementia risk. A total of 33 studies involving 16,926 participants were included. For HbA1c, 24 studies reported cognitive impairment ( N = 8473) and 6 reported dementia ( N = 4677); elevated HbA1c levels in elderly diabetic patients were significantly associated with dementia (OR = 1.60, 95% CI: 1.28–1.99) and cognitive impairment (OR = 1.22, 95% CI: 1.13–1.32). Significant publication bias was detected, as evidenced by asymmetry in the funnel plot, a significant Egger’s test (t = 5.38, P < 0.01), and 13 missing studies imputed by the trim‑and‑fill analysis. Therefore, the pooled estimate for HbA1c should be interpreted with caution. For FPG, 5 studies reported cognitive impairment ( N = 2291) and 6 reported dementia ( N = 3177); FPG levels were associated with dementia (OR = 1.36, 95% CI: 1.09–1.70), indicating a nonlinear dose-response association ( P non-linearity = 0.008, P overall < 0.001). At higher FPG levels (≥ 10.5 mmol/L), the RR increased more steeply, reaching 1.51 at 10.5 mmol/L and 1.77 at 12.5 mmol/L. For the TyG index, 1 study reported cognitive impairment ( N = 308) and 1 reported dementia ( N = 739); higher TyG index levels were associated with cognitive impairment (OR = 2.24, 95% CI: 1.47–3.42). For adiponectin, 2 studies reported cognitive impairment ( N = 251) and no study reported dementia; adiponectin was identified as a protective factor against cognitive impairment (OR = 0.67, 95% CI: 0.49–0.84). Substantial heterogeneity was observed across analyses, with I² values ranging from 58.7% to 69.5% ( P < 0.05), which limits confidence in these pooled estimates. Elevated levels of HbA1c, FPG, and the TyG index were associated with increased risk of cognitive impairment and dementia in elderly patients with diabetes, with modest effect sizes. The dose-response analysis for FPG and dementia in this study was based on only 3 studies, representing a very limited sample size. The TyG index and adiponectin analyses were each based on only 2 studies, and these findings should be viewed as exploratory and hypothesis‑generating, not confirmatory. Future studies incorporating more high-quality cohort investigations are still warranted to explore the long-term association between glycemic and metabolic biomarkers and cognitive impairment and dementia.