Endothelin-1 predicts incident diabetic peripheral neuropathy in Type 2 Diabetes: a cohort study

in European Journal of Endocrinology
View More View Less
  • 1 Clinical Research Unit, Khoo Teck Puat Hospital, Singapore, Singapore
  • 2 Diabetes Centre, Admiralty Medical Centre, Singapore, Singapore
  • 3 Division of Endocrinology. Khoo Teck Puat Hospital, Singapore, Singapore
  • 4 Yishun Polyclinic, National Healthcare Group, Singapore, Singapore
  • 5 Saw Swee Hock School of Public Health, Singapore, Singapore

Correspondence should be addressed to S Tavintharan; Email: subramaniam.tavintharan@ktph.com.sg
Restricted access


Diabetic peripheral neuropathy (DPN) is a common microvascular complication in patients with type 2 diabetes (T2D). Apart from hyperglycemia, few modifiable risk factors have been identified. Endothelin-1 is a potent vasoconstrictor peptide, implicated in the causal pathway of microangiopathy. We investigated whether baseline plasma endothelin-1 and other metabolic and vascular risk factors predicted the incidence of DPN.


This is a 3-year observational, cohort study.


In patients with T2D (n = 2057), anthropometric data, fasting blood, and urine were collected for biochemistry and urine albumin/creatinine measurements. Forearm cutaneous endothelial reactivity was assessed by iontophoresis and laser Doppler flowmetry/imaging. Measurements were repeated on follow-up. Incident DPN was considered present if an abnormal finding in monofilament (<8 of 10 points) or neurothesiometer testing was ≥25 volts on either foot at 3-year follow-up, but normal at baseline. Plasma endothelin-1 was assessed by ELISA.


At baseline, mean age of patients was 57.4 ± 10.8 years old and prevalence of DPN was 10.8%. Of the 1767 patients without DPN, 1250 patients returned for follow-up assessment ((2.9 ± 0.7) years), with a 10.7% incidence of DPN. Patients with incident DPN had significantly higher baseline endothelin-1 (1.43 (1.19–1.73) vs 1.30 (1.06–1.63)) pg/mL, P < 0.0001. Multivariable Cox proportional hazards ratio showed a 1-s.d. increase in log endothelin-1 (adjusted HR: 4.345 (1.451–13.009), P = 0.009), systolic blood pressure (per 10-unit) (adjusted HR: 1.107 (1.001–1.223), P = 0.047) and diabetes duration (adjusted HR: 1.025 (1.004–1.047), P = 0.017) predicted incident DPN, after adjustment for glycemic control, eGFR, albuminuria, peripheral arterial disease and retinopathy status.


Higher baseline endothelin-1, blood pressure and diabetes duration were significant and independent predictors for incident DPN. Validation of our findings in independent cohorts and molecular mechanistic studies will help better our understanding on the role of endothelin-1 in DPN.

Supplementary Materials

    • Supplementary Figure 1 Figure showing patients with and without DPN at baseline and those who returned on follow-up
    • Supplementary Table 1 Characteristics of patients at 3-year follow-up, with no DPN at baseline, comparing incident DPN and those who remain free-of neuropathy, and DPN at baseline
    • Supplementary Table 2 Multivariable generalized linear regression of plasma endothelin-1 (log transformed) with anthropometric and biochemical parameters, where endothelial-dependent (top) and independent (bottom) association were analysed separately
    • Supplementary Table 3 Patients on medications and corresponding plasma Endothelin-1 (pg/ml)
    • Supplementary Table 4 Multivariable Cox regression analysis predicting incident Diabetic Peripheral Neuropathy


     European Society of Endocrinology

Sept 2018 onwards Past Year Past 30 Days
Abstract Views 1404 1404 70
Full Text Views 83 83 7
PDF Downloads 65 65 4