Abstract:
Background: It is likely that lower limb ulceration, lower limb amputation, or their absence in diabetic subjects, indicate varying degrees of long-term diabetes and its complications, and that measures of atherosclerosis and neuropathy would reflect these differences. Objectives: To determine feasibility and, based on our results, make sample size estimates for future study: By comparing peripheral and central vasculature between diabetic subjects with lower extremity ulcers, diabetic subjects with lower extremity amputation and a group of diabetics without these complications — through evaluating toe blood pressure (TBP), toe-brachial index (TBI) and pulse wave velocity (PWV); also, by comparing peripheral and autonomic nervous system integrity between these groups — through sensory, nerve conduction, needle-examination and autonomic function assessment. Study design: A cross-sectional, descriptive and comparative pilot study. Setting: Pretoria Academic Hospital. Participants: Three groups of ten patients consecutively selected from diabetes and diabetic foot clinics — ten with chronic lower extremity ulcers, ten with healed lower extremity amputations and ten diabetic controls. Methods: Assessment of peripheral and autonomic neuropathy included evaluation of 5.07/10-g monofilament sensation, vibration perception (using a 128Hz tuning fork), nerve conduction and electromyography, cutaneous autonomic response and heart rate variability (expressed as an Expiration: Inspiration (E:I)-ratio). For evaluation of vascular status, we obtained the photo-plethysmographically-derived TBI and assessed carotid-femoral (CF) and carotid-radial (CR) PWV. Sample sizes for future studies were calculated through a nomogram for three-group comparisons, ANOVA, simulation and log-transformation of non-parametric data. Results: Absence of vibration perception in at least one leg, with significant p-values of 0.000 at toe-, and 0.027 at medial malleolus- level, occurred more frequently in the amputation, than in the control group. For the total bilateral monofilament count a statistically significant difference between groups was demonstrated (p-value 0.043). Peripheral neuropathy based on abnormality of at least one conduction attribute in at least two distinct nerves, the E:I-ratio, assessment of cutaneous autonomic responses and TBI, by worsening across groups, seemed to display a correlation with severity of lower limb complications, but without statistically significant results. For CF- and CR PWV, the lowest values were observed in the amputation group. Sample size calculations based on our TBP, TBI, vibration and monofilament results, lead to a proposed equal group size of between 34 and 103 for future three-group comparisons using these outcomes measures. Should PWV be included, the group size would have to be between 160 and 222. Conclusions: This study confirmed the usefulness of monofilament sensation and vibration perception assessment in identifying diabetic patients with differing degrees of lower extremity risk. Also, due to the large differences between groups, it demonstrated the effectiveness of these measures to display differences between groups, even in the event of very small sample sizes. The tendencies to worsen across the three groups, of the E:I -ratio, peripheral neuropathy based on nerve conduction, and the TBI, will have to be re-examined in a study with larger sample size. In order to demonstrate statistically significant CF- and CR PWV results, a larger sample size may also be required.
