Abstract
BACKGROUND AND PURPOSE: Diabetes mellitus (DM) has been proposed as a risk factor for carpal tunnel syndrome (CTS), but this remains controversial. We investigated the association between DM and CTS using both ultrasonography (US) and nerve conduction study (NCS) data. METHODS: We analyzed a prospectively recruited database of neuromuscular US and medical records of subjects who had undergone NCSs and electromyography for symptoms suggestive of CTS. Subjects were assigned to the follow groups: Group I, CTS with DM; Group II, CTS without DM; Group III, no CTS with DM; and Group IV, no CTS without DM. US cross-sectional area (CSA) and NCS measurements at the median nerve (MN) were compared among groups. We used a general linear mixed model to adjust for statistically significant covariates. RESULTS: The 230 participants comprised 22, 83, 19, and 106 in Groups I-IV, respectively. In multivariate analyses, the MN action potential amplitude in females was the only variable that was significantly associated with DM (p<0.001). Groups with DM tended to have a longer latency, smaller amplitude, and lower conduction velocity in the NCSs compared to groups without DM. The measured US CSA values did not differ significantly among the groups. CONCLUSIONS: NCS measurements of the MN tended to differ between DM and non-DM patients regardless of the presence or absence of CTS. However, US did not reveal any statistically significant relationship between CTS and DM.