Background: Peripheral neuropathy (PN) is a diabetes complication but also occurs without diabetes. The effect of diabetes on PN is not well quantified.

Methods: Subjects ≥ 18 years old had a health exam, including an OGTT for diabetes. Distal sensory PN, assessed using 3 graded Semmes Weinstein monofilaments, was classified as none (felt all filaments), latent (insensate to 1 g filament), clinical [mild (insensate to 10 g) or severe (75 g)]. Associations with PN were assessed using ordinal logistic regression assuming equal coefficients across classes of PN.

Results: In 1564 subjects (median age 42 years, 50% women, 58% American Indians), PN was latent or worse in 68.9% (95% CI 66.7-71.2), clinical in 11.2% (CI 9.7-12.9), and associated with male sex, older age, and worse glucose regulation (Figure). Age-sex-race-adjusted prevalence was greater in people with than without diabetes (odds ratio; OR 2.2, CI 1.7-2.8), and in those with diabetes, with longer diabetes duration (OR 1.9 per decade, CI 1.5-2.4) or albuminuria, A/C ≥30 mg/g (OR 2.4, CI 1.6-3.6). In those without clinical PN, latent PN was associated with diabetes (OR 1.7, CI 1.2-2.2), longer diabetes duration (OR 2.2 per decade, CI 1.4-3.5) and albuminuria (OR 1.8, CI 1.0-3.2).

Conclusions: Although PN is common in people with no diabetes, diabetes greatly increases its risk. Use of 3 monofilaments rather than 1 adds precision and allows detection of latent PN.


M.T. Olaiya: None. R.L. Hanson: None. K.G. Kavena: None. M. Sinha: None. R. Nelson: None. W.C. Knowler: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at