Background and Objectives Hypothermia is likely to develop faster during spinal anesthesia than epidural anesthesia. A natural consequence of the rapid temperature decrease during spinal anesthesia is that the shivering threshold will be reached sooner and that more shivering will be required to prevent further hypothermia. We tested the hypotheses that the onset of hypothermia is more rapid and the onset and intensity of shivering earlier during spinal than epidural anesthesia.

Methods Patients undergoing cesarean delivery were randomly assigned to spinal anesthesia or epidural anesthesia. Spinal anesthesia was induced by injecting 2 mL 0.5% dibucaine into the L4-L5 interspace. Epidural anesthesia was induced with 20 mL 2% mepivacaine injected into the L2-L3 interspace. Thermal comfort and shivering were scored by a blinded observer.

Results Fifteen patients given each type of anesthesia had upper sensory levels ≥T4 dermatome. Sensation was entirely absent from the leg during spinal anesthesia, but lower block levels were near S5 during epidural anesthesia. Tympanic membrane temperatures initially decreased faster during spinal anesthesia, but subsequently decreased at a rate of 0.5°C/h in both groups. The onset and incidence of shivering (detected qualitatively) did not differ significantly between the two groups, but shivering intensity was significantly reduced during spinal anesthesia. Furthermore, the shivering thresholds were 36.4 ± 0.3°C (mean ± SD) during spinal anesthesia versus 37.1 ± 0.4°C in those given epidural anesthesia (P = .006). There were no clinically important differences in thermal comfort with the two kinds of neuraxial anesthesia.

Conclusions We failed to confirm our hypothesis, but for an unexpected reason: Thermoregulation was impaired more by spinal anesthesia than epidural anesthesia. It seems likely that in our patients spinal anesthesia inhibited thermoregulatory control more than epidural anesthesia because it better blocked sensory input from the legs.