4 8 0 Varvarigou et al. The Journal of Pediatrics March 1994 Increased levels and positive correlation between erythropoietin and hemoglobin concentrations in newborn children of mothers who are smokers Anastasia Varvarigou, MD, Nicholas G. Beratis, MD, Maria Makri, MSc, and Apostolos G. Vagenakis, M~) From the Departments of Pediatrics and Internal Medicine, University of Patras Medical School, General University Hospital, Patras, Greece The mean erythropoietin concentration in the cord blood of neonates whose mothers were smokers was greater than in neonates whose mothers were not smokers. There was a significant positive correlation between erythropoietin and hemoglobin concentrations. The findings suggest that approximately one of five fetuses who are exposed to tobacco smoke are in a state of chronic hy- poxia. (J PEDIATR1994;124:480"2) Tissue hypoxia is known to cause production of erythropoi- etin.1 Chronic fetal hypoxia is associated with elevated Epo concentrations, as evidenced by increased cord blood Epo levels in infants born to mothers with hypertension, 2 diabe- tes, 2 Rh immunization, 3 and severe preeclampsia.4 Con- versely, increased arginine vasopressin levels have been found in neonates with acute asphyxia, s Decreased placental intervillous blood flow6, 7 and in- creased fetal heart rate and blood flow of both the aortic and umbilical vein have been associated with maternal cigarette smoking,s Because maternal smoking affects fetal hemodynamics and increases the levels of fetal carboxyhemoglobin, 9 we measured the concentrations of Epo and AVP in the cord blood of newborn infants whose mothers were smokers. METHODS Subjects. We studied 62 neonates. Thirty-one were infants of consecutive mothers who admitted smoking 10 or more cigarettes daily throughout pregnancy. The mean (+SD) number of cigarettes smoked per day was 24.5 -2-- 7.8, with a range from 10 to 40 cigarettes per day. The remaining 31 were infants of mothers who stated that they were not smokers. The term control neonates were se- lected randomly after the delivery of a neonate by a mother who was a smoker, provided that the criteria of the study were fulfilled. The preterm control neonates were selected only by the gestational age to match with that of the pre- term infants of the study group. Submitted for publication July 14, 1993; accepted Sept. 30, 1993. Reprint requests: Nicholas G. Beratis, MD, Department of Pedi- atrics, Universityof Patras Medical School, P.O. Box 1045, 26110 Patras, Greece. Copyright © 1994 by Mosby-Year Book, Inc. 0022-3476/94 $3.00 + 0 9/24/51808 Neonates included in the study fulfilled the following criteria: (1) uneventful pregnancy, with no preeclampsia or medications, (2) no history of maternal diabetes or hyper- tension, (3) vaginal delivery, (4) Apgar score ~>8 at 1 and 5 minutes, (5) clear amniotic fuid, (6) gestational age be- tween 32 and 40 weeks, and (7) size appropriate for gesta- tional age. The neonates who were small for gestational age were excluded because we have found abnormalities of the endocrine status of such neonates. 1° AVP Arginine vasopressin CI Confidence interval Epo Erythropoietin The gestational age and the birth weight of the infants included in the study are listed in the Table. The gestational age was determined according to the Ballard scoring system. 11 The deviation among the neurologic and physi- cal assessments, the estimated gestational age from the last menstrual period, and the ultrasound measure- ments obtained before week 20 of gestation was less than 2 weeks. Cases -with a greater discrepancy were excluded. Procedures. Samples were stored at -60 ° C and were as- sayed in the same assay for each of the two hormones. All measurements were obtained in a blind fashion and were made in duplicate. The Epo concentration was measured by a radioimmunoassay technique (Incstar Corp., Stillwater, Minn.). The AVP concentration was determined in ethanol- extracted plasma specimens. A radioimmunoassay kit from Euro-Diagnostics (Apeldoorn, The Netherlands) was used. The sensitivity of the Epo and AVP assays was 4.4 U/L and 0.4 mU/L, and the coefficient of variation was 5% and 7.6%, respectively. We analyzed the data by applying the Student t test and