Development of circadian rhythmicity of temperature in full-term normal infants.

Twelve full-term infants (7 girls and 5 boys) with normal neurological, behavioral and somatic development were followed at regular intervals during the first 5 months of life to appreciate the development of circadian rectal temperature rhythmicity. Activity and temperature (oral at birth, rectal thereafter) were monitored for a minimum of 60 hours on seven separate occasions: at birth, 3 weeks, 6 weeks, 8 weeks, 10 weeks, 16 weeks and 20 weeks of age. Activity was measured using an actigraph worn on the infant's wrist, and rectal temperature was measured using a rectal probe attached to a portable microprocessor (Vitalog TM). Data points were collected every 2 minutes. No fewer than ten infants were monitored at each session, and no infant missed more than one session. Missing recordings were due to equipment malfunctions, probe expulsions and minor health problems. Six infants out of 12 were successfully monitored at each of the first four sessions, from birth to 8 weeks of age inclusively, and two subjects were successfully monitored at all seven sessions. Periodic regression analysis was performed by least squares curve fit with secondary analysis of variance. Analysis of covariance was performed on repeated measures. There was no evidence of rectal temperature circadian rhythmicity at 3 weeks. Two infants demonstrated a circadian rhythmicity at 6 weeks, and all infants had a circadian rhythmicity at 10 weeks post-natal age. At the time of the first observance of circadian rhythmicity of rectal temperature, the mean delta in temperature from peak to trough was 0.6 +/- 0.3 degrees C. This delta was greater at the 16th week, with a mean value of 1.2 +/- 0.3 degrees C. The trough was seen during the first part of the long nocturnal inactivity period. Circadian rhythmicity of rectal temperature was always observed in the studied subjects before the establishment of a consolidated, long daytime wake period.