Serum Lipidomics Profiling Identify Novel Biomarkers of Distal Symmetrical Polyneuropathy in Type 1 Diabetes

Distal symmetrical polyneuropathy (DSPN) is a common complication of type 1 diabetes (T1D), yet validated biomarkers for early detection or prognosis are lacking. Metabolic disturbances driven by chronic hyperglycemia and dyslipidemia contribute to DSPN development and progression.

We used untargeted serum liquid chromatography-mass spectrometry lipidomics to identify DSPN biomarkers. The discovery cohort included 153 individuals with T1D (109 with and 44 without DSPN) and 50 non-diabetic controls. The independent validation cohort included 99 individuals with T1D with established DSPN status. Key lipids were identified using multivariate modelling, followed by ANCOVA and adjusted post hoc tests.

A total of 543 lipid species were identified in the discovery cohort. Among these, 14 lipids were associated with DSPN. Of which, 6 lipids, namely Cer(d42:1), PC(36:4), LPC(16:0), LPE(18:1), PE(36:2), and PE(O-40:5) or PE(P-40:4), showed a clear pattern among non-diabetic individuals, people affected by T1D with and without DSPN. In the validation cohort, 3 of the 6 initially identified lipids showed the same directional changes as in the discovery cohort. A logistic regression model combining the six lipid biomarkers with HbA1c, diastolic blood pressure, and age achieved AUCs of 0.83 (95% CI, 0.76-0.91) in the discovery cohort and 0.81 (95% CI, 0.73-0.90) in the validation cohort for detecting DSPN in T1D. AUC improvement was not significant by DeLong's test in either cohort (p = 0.086 and 0.089) but was significant when combined using Fisher's method (χ² = 9.7, p = 0.045). Additionally, 22 lipid species differed significantly between painful and painless DSPN.

We identified a reproducible lipidomic signature discriminating DSPN in type 1 diabetes. Painful DSPN was characterized by reduced unsaturated PC/PE and increased sphingolipids, offering new insight into disease pathophysiology.