Abstract
AIMS: To determine whether recovery of fasting C-peptide (ΔFCP) reduces deterioration in sudomotor small-fibre function, whether a threshold exists, and whether associations are complication-specific. METHODS: Retrospective real-world cohort of 288 hospitalised adults with diabetes with repeated fasting C-peptide and Sudoscan electrochemical skin conductance (ESC). Deterioration in sudomotor small-fibre function was defined as worsening foot ESC. Large-fibre neuropathy (LFN) and diabetic kidney disease (DKD) progression were assessed by nerve conduction and kidney outcomes. Cox models, restricted cubic splines, maximally selected rank statistics, and subgroup/landmark/sensitivity analyses were performed. RESULTS: Over a median 2.2 years, higher ΔFCP was independently associated with lower risk of deterioration in sudomotor small-fibre function, but not with LFN or DKD progression. Highest versus lowest ΔFCP tertile: adjusted HR 0.31 (95% CI 0.14-0.69; p < 0.001). An optimal ΔFCP cutoff of 126.8 pmol/L was identified; ΔFCP ≥ 126.8 pmol/L was associated with a 66% lower risk of sudomotor deterioration (HR 0.34, 95% CI 0.17-0.69; p < 0.001), remaining robust in landmark and sensitivity analyses. Baseline C-peptide mediated little of the association. CONCLUSIONS: Endogenous fasting C-peptide recovery is linked to a clinically meaningful, threshold-dependent reduction in the progression of sudomotor small-fibre dysfunction, supporting β-cell preservation/recovery as targets in early diabetic neuropathy.