Cognitive impairment is an increasingly recognized complication of type 2 diabetes mellitus (T2DM), but reliable circulating indicators of early neurocognitive vulnerability remain limited. A case-control study published in Diabetes, Obesity and Metabolism evaluated whether circulating short-chain fatty acids (SCFAs), metabolites derived from gut microbiota, differ across healthy individuals, people with T2DM, and those with diabetic cognitive impairment (DCI).
The analysis included 180 older adults, with 60 participants in each group: healthy controls, T2DM, and DCI. Serum concentrations of acetate, propionate, and butyrate were measured. Parallel multinomial regression models were constructed using T2DM as the reference group. Two models were evaluated: Model 0, including SCFAs only, and Model 1, including SCFAs with covariates. Diagnostic discrimination across groups was assessed using multiclass and pairwise receiver operating characteristic (ROC) analyses, with bootstrap internal validation using 2000 resamples to generate optimism-corrected area under the curve (AUCc) estimates with 95% confidence intervals.
In adjusted analyses, higher acetate levels corresponded to higher odds of healthy control status compared with T2DM (OR 1.153; 95% CI 1.004-1.324; p = 0.044). Conversely, lower acetate levels corresponded to higher odds of DCI compared with T2DM (OR 0.797; 95% CI 0.650-0.977; p = 0.029). Propionate and butyrate showed positive associations with healthy control status versus T2DM, but were not independently related to DCI versus T2DM.
Model discrimination improved when covariates were incorporated. SCFA-only models yielded modest discrimination with Hand-Till AUCc values of 0.585-0.629, whereas covariate-adjusted models achieved AUCc values of 0.733-0.740. Among pairwise comparisons, distinguishing DCI from T2DM remained the most difficult contrast, although acetate demonstrated the highest discriminative performance across models. These findings highlight differences in circulating SCFA profiles across metabolic and cognitive states in older adults with diabetes.
