The Effect of Diabetes and Limb Loss on Thermoregulation and Temperature Perception

Abstract

Elevated residual limb skin temperatures inside the prosthetic socket are a source of thermal discomfort and likely responsible for profuse sweating, both reported as significant problems for lower limb amputees (Hagberg & Branemark, 2001; Meulenbelt, Geertzen, Jonkman, & Dijkstra, 2009). Simply donning the prosthesis causes a slight increase in skin temperature, which rises markedly during activity (G. K. Klute, Berge, Orendurff, Williams, & Czerniecki, 2006). It takes doffing the prosthesis for the skin temperature to return to normal because poor thermal properties of the liner and socket severely impede heat transfer. Two local mechanisms that are important for maintaining thermal comfort are thermal perception (sensation of warm or cold) and control of peripheral blood flow (increasing or decreasing to dissipate or preserve heat). Diabetes is the reason for the majority of amputations each year (Dillingham, Pezzin, & Mackenzie, 2002) and has been strongly associated with altered thermoregulation (Wick et al., 2006). The goals of the proposed study are to design and build a research tool to measure thermal perception and blood flow and to use this device to determine if diabetic lower extremity amputees (LEAs) or traumatic LEAs: (1) are less sensitive to warm or cold and/or (2) have an altered microcirculatory response to a thermal stimulus. We have tested the thermal perception and microcirculatory reactivity of the lateral aspect of the distal residual limb of 2 traumatic LEAs and 3 diabetic LEAs, and the homologous location of 4 healthy, intact controls. We measured local blood flow using a laser Doppler flowmeter (LDF) and warm and cold detection thresholds (WDT, CDT) with quantitative sensory testing (QST). We started at four initial temperatures (30°C, 32°C, 34°C, and 35°C) and used the Method of Limits algorithm with a ramp rate of 0.2 C/s (Divert, 2001; Shy et al., 2003), but modified it by holding the initial temperature and perceived temperature for three minutes to collect blood flow data. A linear mixed effects regression was used to compare the threshold and blood perfusion data by subject group, and a likelihood ratio test was used to distinguish trends in the shape of the variables relationship to initial temperature. The average difference in CDT temperatures between the traumatic LEAs and controls was -1.6°C (SD = 2.6°C) and between diabetic LEAs and controls was -1.8°C (SD = 2.6°C). For the cold test, compared to controls, average change in blood perfusion was 0.5 (SD = 2.2) profusion unit (PU) more and 2.3 (SD = 2.2) PU less for diabetics and traumatic amputees, respectively. The two amputee groups were combined for the WDT and change in blood perfusion because of insufficient power in both groups. For the warm test, average difference in WDT temperatures and change in blood perfusion for the amputees was 2.1°C (SD = 1.2°C) greater and 13.7 (SE 18.0) PU less than the controls. Differences in warm detection thresholds were close to significance (p=0.069). There was insignificant difference in cold detection, vasoconstrictor response to cold stimuli, and vasodilator response to warm stimuli for traumatic LEAs and diabetic LEAs as compared to healthy non-amputee controls. We conclude that amputees may have decreased sensitivity to warm temperature and there was no noticeable trend towards a change in cold perception, vasodilation or vasoconstriction. The implications of these findings may mean that amputees have altered perception of warmth.