Sun et al Epidemiology. Volume 17, Number 3, May 2006 consistent with those of the 2 previous studies. 4. 2Our study similar to those of previous studies from other industrialized further showed that changes in Apgar scores from 1 to 5 countries. Some children in our study arbitrarily received an minutes played a role, and we also found an association that Apgar score of zero if they were transferred to another depart persisted into early adulthood. ment immediately after birth(0. 4% for 1-and 0. 2% for 5-minute Apgar score is a nonspecific assessment of the neonates Apgar score) and some children had missing Apgar scores for hortly after birth, and low Apgar scores are not specific to unknown reasons(0.6% for 1- and 1. 1% for 5-minute Apgar birth asphyxia. 24,25Congenital malformations, infections, and score). These children had a slightly higher rate of epilepsy than administration of drugs to the mother can also lead to low those with an Apgar score of 10 scores.-We know of only few studies on the effect of early Our findings call for more research focusing on a fetal perinatal environmental factors and the risk of epilepsy.- and perinatal origin of epilepsy. Potential causal candidates Most of these studies did not find strong associations, -0 but could be infections, maternal lifestyle factors, maternal com- some studies indicate that neonatal encephalopathy, neonatal plications during pregnancy, and factors related to the deliv- seizures, and epilepsy may be related to brain lesions occur- ery process ring the antepartum pe The combination of low Apgar scores and symptoms of neonatal encephalopathy REFERENCES carried an increased risk of a variety of later minor disabili- 1. Hauser WA, AnnegersJF,Kurland LT.Incidence ties, including epilepsy, among children with normal birth I xkms -eiw-res in Rochester, Minesota: 1935-I weights, no congenital malformation, and no major neuro- 2. Sander JW. The epidemiology of epilepsy revisited. Curr Opin Neurol gic abnormalities. Unfortunately, we have no information 2003:16:165-170 bout neonatal encephalopathy to add to these findin 3. Kjeldsen M, Kyvik Ko, Christensen K, et al. Genetic and environmen- The major strengths of our study are the large sample size, tal factors in epilepsy: a population-based study of 11900 Danish twin airs. Epilepsy Res. 2001: 44: 167-178 information on hospitalization, and the fact that data came from 4. been a, ecrdsyinehrodreg, 2 z. study uniform healthcare system that is free of charge for all patients. 5. Lilienfeld AM, Pasamanick B. Association of maternal and fetal factors Children with low Apgar scores may be followed more closely with the development of epilepsy. I. Abnormalities in the prenatal and for adverse outcomes than those with a normal score for a short paranatal periods. JAMA. 1954: 155: 719-724 time after birth, but probably not for months or years. Apgar 6. Nelson kB, Elenberg JH Predisposing and causative factors in child- cores were recorded prospectively and the diagnosis of epilepsy 7. Rocca WA, Sharbrough FW, Hauser WA, et al. Risk factors for gener- was made independently of these scores, making differential alized tonic-clonic seizures: a population-based case-control study in misclassification an unlikely explanation for the association. It is Rochester, Minnesota. Neurology. 1987; 37: 1315-1322 more likely that nondifferential misclassification attenuates the 8. Rocca WA. Sharbrough FW, Hauser WA, et al. Risk factors for absence eizures: a population-based case-control study in Rochester, Minne- true association sota. Neurology. 1987: 37: 1309-1314 The main weakness of the study is that we had limited 9. Rocca WA, Sharbrough FW, Hauser WA, et al. Risk factors for clinical data. Epilepsy is a heterogeneous disorder, and dif- partial seizures: a population-based case-control study. Ann ogies. We were not able to evaluate whether that was the case 101987:21: 22-31 ferent seizure types and syndromes may have different etiol- 10. Leone M, Bottacchi E, Beghi E, et al. Risk factors for a first generalized nic-clonic seizure in adult life. Neurol Sci. 2002: 23: 99-106. for the association with Apgar scores. The information of 11. Apgar V. A proposal for a new method of evaluation of the newborn epilepsy in our study came from ICD codes. The validity of nfant Curr Res Anesth Analg. 1953: 32: 260-267 the epilepsy diagnosis has been assessed according to Inter 12. Drage J, Kenndy C, Schwarz B. The Apgar score as an index of neonatal national League Against Epilepsy criteria in 188 randomly rtality. A report from the collaborative study of cerebral palsy. Obster elected inpatients and outpatients registered with epilepsy in 13. Moster D, Lie RT, Irgens LM, et al. The association of Apgar score with the National Hospital Register. The epilepsy diagnoses met subsequent death and cerebral palsy: A population-based study in term the definition as recurrent, unprovoked epileptic seizures in infants. J Pediatr. 2001: 138: 798-803 153 patients corresponding to a positive predictive value of 14. Thormgren-Jerneck K, Herbst A. Low 5-minute Apgar score: a popula- 81%(95%CI=75-87%), which is comparable with other tion-based register study of 1 million term births. Obstet Gynecol. population-based register studies. Among the 35 persons 15.Casey BM, McIntire DD, Leveno KJ. The continuing value of the Apgar who did not fulfill the criteria, 14 had had a single episode of score for the assessment of newborn infants. N Engl JMed. 2001344 seizures.Thus, our estimate of the positive predictive value is 16. Nelson KB, Ellenberg JH. Apgar scores as predictors of chronic neuro- conservative, because some of these persons will probably logic disability. Pediatrics. 1981: 68: 36-44 develop epilepsy. There were 2 major changes in the hospital 17. Malig C vil registration system in Denmark. Technical Papers register system during the time of follow up. ICD-10 replaced RS.1996;66:1-6 CD-8 in 1994, and outpatients were included in the register 18. Knudsen LB, Olsen J. The Danish Medical Birth Registry. Dan Med system from 1995. The association between Apgar score and 19. Andersen TE, Madsen M, Jorgensen J, et al. The Danish National darde: The Apgar scoring is made by midwives following stan- 20. Breslow NE, Day NE. Statistical methods in cancer research Volume ized procedures, but unfortunately, the interrater reliability II-the design and analysis of cohort studies. IARC Scientific Publica- nons.1987:82. of Apgar score has not been estimated in Denmark. The distri- 21. Clayton D, Hills M. Statistical Models in Epidemiology.Oxford,New bution of 5-minute Apgar score in our study was, however, York, Tokyo: Oxford University Press: 1993 c 2006 Lippincott Williams c wilkins 第61页 Copyright c Lippincott Williams& Wilkins. Unauthorized reproduction of this article is prohibitedconsistent with those of the 2 previous studies.14,23 Our study further showed that changes in Apgar scores from 1 to 5 minutes played a role, and we also found an association that persisted into early adulthood. Apgar score is a nonspecific assessment of the neonates shortly after birth, and low Apgar scores are not specific to birth asphyxia.24,25 Congenital malformations, infections, and administration of drugs to the mother can also lead to low scores.26 –29 We know of only few studies on the effect of early perinatal environmental factors and the risk of epilepsy.4 –10 Most of these studies did not find strong associations,6 –10 but some studies indicate that neonatal encephalopathy, neonatal seizures, and epilepsy may be related to brain lesions occur￾ring in the antepartum period.30 –32 The combination of low Apgar scores and symptoms of neonatal encephalopathy carried an increased risk of a variety of later minor disabili￾ties, including epilepsy, among children with normal birth weights, no congenital malformation, and no major neuro￾logic abnormalities.33 Unfortunately, we have no information about neonatal encephalopathy to add to these findings. The major strengths of our study are the large sample size, the long follow-up time, the low loss to follow up, the reliable information on hospitalization, and the fact that data came from a uniform healthcare system that is free of charge for all patients. Children with low Apgar scores may be followed more closely for adverse outcomes than those with a normal score for a short time after birth, but probably not for months or years. Apgar scores were recorded prospectively and the diagnosis of epilepsy was made independently of these scores, making differential misclassification an unlikely explanation for the association. It is more likely that nondifferential misclassification attenuates the true association. The main weakness of the study is that we had limited clinical data. Epilepsy is a heterogeneous disorder, and dif￾ferent seizure types and syndromes may have different etiol￾ogies. We were not able to evaluate whether that was the case for the association with Apgar scores. The information of epilepsy in our study came from ICD codes. The validity of the epilepsy diagnosis has been assessed according to Inter￾national League Against Epilepsy criteria34 in 188 randomly selected inpatients and outpatients registered with epilepsy in the National Hospital Register.35 The epilepsy diagnoses met the definition as recurrent, unprovoked epileptic seizures in 153 patients corresponding to a positive predictive value of 81% (95% CI
75– 87%), which is comparable with other population-based register studies.36 Among the 35 persons who did not fulfill the criteria, 14 had had a single episode of seizures. Thus, our estimate of the positive predictive value is conservative, because some of these persons will probably develop epilepsy. There were 2 major changes in the hospital register system during the time of follow up. ICD-10 replaced ICD-8 in 1994, and outpatients were included in the register system from 1995. The association between Apgar score and epilepsy decreased slightly when outpatients were included. The Apgar scoring is made by midwives following stan￾dardized procedures, but unfortunately, the interrater reliability of Apgar score has not been estimated in Denmark. The distri￾bution of 5-minute Apgar score in our study was, however, similar to those of previous studies from other industrialized countries.14,33 Some children in our study arbitrarily received an Apgar score of zero if they were transferred to another depart￾ment immediately after birth (0.4% for 1- and 0.2% for 5-minute Apgar score) and some children had missing Apgar scores for unknown reasons (0.6% for 1- and 1.1% for 5-minute Apgar score). These children had a slightly higher rate of epilepsy than those with an Apgar score of 10. Our findings call for more research focusing on a fetal and perinatal origin of epilepsy. Potential causal candidates could be infections, maternal lifestyle factors, maternal com￾plications during pregnancy, and factors related to the deliv￾ery process. REFERENCES 1. Hauser WA, Annegers JF, Kurland LT. Incidence of epilepsy and unprovoked seizures in Rochester, Minnesota: 1935–1984. Epilepsia. 1993;34:453– 468. 2. Sander JW. The epidemiology of epilepsy revisited. Curr Opin Neurol. 2003;16:165–170. 3. Kjeldsen MJ, Kyvik KO, Christensen K, et al. Genetic and environmen￾tal factors in epilepsy: a population-based study of 11900 Danish twin pairs. Epilepsy Res. 2001;44:167–178. 4. Degen R. Epilepsy in children. An etiological study based on their obstetrical records. J Neurol. 1978;217:145–158. 5. Lilienfeld AM, Pasamanick B. Association of maternal and fetal factors with the development of epilepsy. I. Abnormalities in the prenatal and paranatal periods. JAMA. 1954;155:719 –724. 6. Nelson KB, Ellenberg JH. Predisposing and causative factors in child￾hood epilepsy. Epilepsia. 1987;28(suppl 1):S16 –24. 7. Rocca WA, Sharbrough FW, Hauser WA, et al. Risk factors for gener￾alized tonic– clonic seizures: a population-based case– control study in Rochester, Minnesota. Neurology. 1987;37:1315–1322. 8. Rocca WA, Sharbrough FW, Hauser WA, et al. Risk factors for absence seizures: a population-based case– control study in Rochester, Minne￾sota. Neurology. 1987;37:1309 –1314. 9. Rocca WA, Sharbrough FW, Hauser WA, et al. Risk factors for complex partial seizures: a population-based case– control study. Ann Neurol. 1987;21:22–31. 10. Leone M, Bottacchi E, Beghi E, et al. Risk factors for a first generalized tonic– clonic seizure in adult life. Neurol Sci. 2002;23:99 –106. 11. Apgar V. A proposal for a new method of evaluation of the newborn infant. Curr Res Anesth Analg. 1953;32:260 –267. 12. Drage J, Kenndy C, Schwarz B. The Apgar score as an index of neonatal mortality. A report from the collaborative study of cerebral palsy. Obstet Gynecol. 1964;24:222–230. 13. Moster D, Lie RT, Irgens LM, et al. The association of Apgar score with subsequent death and cerebral palsy: A population-based study in term infants. J Pediatr. 2001;138:798 – 803. 14. Thorngren-Jerneck K, Herbst A. Low 5-minute Apgar score: a popula￾tion-based register study of 1 million term births. Obstet Gynecol. 2001;98:65–70. 15. Casey BM, McIntire DD, Leveno KJ. The continuing value of the Apgar score for the assessment of newborn infants. N Engl J Med. 2001;344: 467– 471. 16. Nelson KB, Ellenberg JH. Apgar scores as predictors of chronic neuro￾logic disability. Pediatrics. 1981;68:36 – 44. 17. Malig C. The civil registration system in Denmark. Technical Papers IIVRS. 1996;66:1– 6. 18. Knudsen LB, Olsen J. The Danish Medical Birth Registry. Dan Med Bull. 1998;45:320 –323. 19. Andersen TF, Madsen M, Jorgensen J, et al. The Danish National Hospital Register. Dan Med Bull. 1999;46:263–268. 20. Breslow NE, Day NE. Statistical methods in cancer research Volume II—the design and analysis of cohort studies. IARC Scientific Publica￾tions. 1987;82. 21. Clayton D, Hills M. Statistical Models in Epidemiology. Oxford, New York, Tokyo: Oxford University Press; 1993. Sun et al Epidemiology • Volume 17, Number 3, May 2006 300 © 2006 Lippincott Williams & Wilkins 第 61 页