Quantitative measures of sympathetic skin response in diabetes: relation to sudomotor and neurological function.

The sympathetic skin response (SSR) at the foot to a deep inspiration was measured in 68 randomly selected diabetic patients and 46 age matched normal subjects and compared with other quantitative measures of neurological and sudomotor function. SSR was obtained in all but three diabetic patients. The upper limit of normal for the onset latency was 2202 ms and the lower limit for the amplitude of the first wave 92 microV. Ten diabetic patients had measurable but prolonged latencies, and 11 had measurable but low amplitudes. There were no significant associations between latency, height, and age, but in insulin dependent patients there was a significant diminution of response amplitude with increasing duration of diabetes. Latency was weakly associated with Marstock thermal thresholds, respiratory RR variation, and common peroneal nerve conduction velocity. SSR amplitude was associated with the density of pilocarpine activatable sweatspots in the same region of the foot. Patients with abnormal latencies were significantly older and had reduced thermal sensation than those with normal latencies. Median coefficients of variation for repeat testing in diabetic patients were 9% for latency and 13% for amplitude. The test is objective and reproducible, but latency measurements reflect conduction in a long multineuronal pathway and are not purely a measure of peripheral C fibre function; amplitude measurements reflect the density of spontaneously activable sweat glands and are therefore a valid measure of peripheral sympathetic activity, though they depend more on temperature than do latencies (mean change over the range 32-34 degrees C; 8.5% degrees C for amplitude, -2.5%/degrees C for latency).