Epidemiological Principles Relevant To The Study Of Diabetic Neuropathy

In order to understand published research on the epidemiology of diabetic neuropathy, certain principles of epidemiological study design must be taken into consideration. These principles guided these authors in the selection of relevant citations and data presentation. Cross-sectional or case-control studies conducted in a population-based sample (such as a defined community or health plan enrollment) were considered for this chapter based on review of Medline citations using the keywords "epidemiology," "diabetes," and "neuropathy" from 1966 to February 2005 review of bibliographies of the articles obtained from the Medline search for relevant citations, and review of the authors' files. Clinic-based cross-sectional or case-control studies have not been considered except in the case of rare conditions for which no other data exists, because of the potential problem of selection bias associated with these study designs (1). All prospective studies, and some randomized controlled trials, were considered. Prospective research is less likely to be biased because of differences in probability of subject selection based on disease (neuropathy) and risk factor presence. Prospective research is a stronger study design in inferring the possibility of causation, because the presence of risk factors may be determined before neuropathy onset.

The problem of measurement error in the assessment of the presence or absence of diabetic neuropathy is well recognized. Nerve conduction velocity, arguably the most objective and accurate test available for the diagnosis of this complication, is known to sometimes result in erroneous classification. For example, nerve conduction velocity may be normal in diabetic subjects with symptoms of distal symmetric polyneuropathy (2). This misclassification problem is even more problematic when a test result is used to formulate a clinical plan for an individual patient, in comparison with epidemiological analysis where population statistics are the result of interest. When misclassification of neuropathy or risk factor status occurs nondifferentially (randomly), the net result is bias of any observed difference toward the null value (1). Therefore observed differences found in an epidemiological analysis of risk factors for diabetic neuropathy validly reflect potential causative factors for this complication, but probably underestimate the magnitude of the risk increase. Epidemiological studies may draw valid conclusions regarding risk factors for diabetic neuropathy even if the techniques used to measure neuropathy and the potential risk factor are known to be inaccurate.


Dyck et al. (3) examined the prevalence of neuropathy among all clinically diagnosed diabetic subjects who resided in Rochester, Minnesota. Only 380 of 870 eligible subjects (44%) agreed to participate, possibly because of concern about the lengthy neurodiagnostic study protocol. Neuropathy was defined if two criteria were satisfied:

1. Abnormal nerve conduction in more than one nerve or abnormal test of autonomic function (low heart rate variation in response to breathing or the Valsalva maneuver).

2. Neuropathic symptom or sign or abnormal quantitative sensory testing.

Median duration of diabetes was 14.5 years for subjects with type 1 diabetes and 8.1 years for subjects with type 2 diabetes. Although the prevalence of neuropathy was high (Table 1), most subjects with neuropathy were asymptomatic (about 71%).

A community-based study in San Luis Valley, Colorado, measured prevalence of neuropathy in a bi-ethnic (Hispanic and Anglo) population (4,5). Neuropathy was defined if two of three criteria were satisfied:

1. Neuropathic discomfort in feet and legs.

2. Abnormal Achilles tendon reflexes.

3. Inability to feel an iced tuning fork on the dorsum of the foot (test of thermal sensation).

Subjects with type 2 diabetes had the highest prevalence of neuropathy, whereas subjects with IGT defined according to World Health Organization criteria had prevalence about midway between normal glucose tolerance (NGT) and type 2 diabetes (Table 1). No subjects with type 1 diabetes were included in this study. Significantly higher prevalence of neuropathy was found in relation to greater age, diabetes duration, glycosylated hemoglobin, male gender, and insulin use. Factors not associated with neuropathy prevalence included blood pressure, height, smoking, previous alcohol use, ankle-arm index, and serum cholesterol, lipid, and lipoprotein levels.

Table 1

Distal Symmetric Polyneuropathy: Prevalence, Incidence, and Risk Factors From Cross-Sectional Research Studies

Table 1

Distal Symmetric Polyneuropathy: Prevalence, Incidence, and Risk Factors From Cross-Sectional Research Studies

Delicious Diabetic Recipes

Delicious Diabetic Recipes

This brilliant guide will teach you how to cook all those delicious recipes for people who have diabetes.

Get My Free Ebook

Post a comment