Seasonal variation of neonatal transient hyperthyrotropinemia in Tehran province, 1998-2005.

Seasonal aggregation and the monthly rate of neonatal transient hyperthyrotropinemia (THT) were assessed. From November 1998 to April 2005, neonates of gestational age ≥37 wks, birth weight 2500-4000 g, birth length 45-55 cm, and 1st min Apgar score >3, who had thyrotropin (TSH) ≥20 mU/L in their cord dried-blood specimen, but without congenital hypothyroidism, were enrolled in the study. The recall rate equals the rate of THT occurrence in this study. Of 47,945 neonates, 555 had THT (recall rate: 1.2%). The aggregated seasonal recall rate (recall for further assessment to rule out congenital hypothyroidism) was significantly higher in winter (January, February, and March) than the other seasons (p < .0001). Winter had higher recall rate in each year as compared to other seasons, but the overall rate of recalls decreased in 2001 and 2002. Excluding the first 6 months (due to erratic variations), the remaining 72 months revealed a relatively sinusoidal pattern in monthly recall rates; indeed, there was an initial 11-month high recall rates (1.7%), followed by a 33-month decrease (0.7%), a 19-month increase (1.9%), and a final 9-month decrease (0.8%). The recall rate of each of these time intervals was significantly different from that of the next time interval (p < .0001). The monthly recall rates were best fitted to cubic curve estimation and then autoregressive integrated moving average (ARIMA) (0, 1, 1) models. THT occurs significantly more in winter than in other seasons, and this suggests a possible role for time-varying factor(s) contributing to its seasonal preponderance.