- Email: [email protected]
Term infants with fetal growth restriction are not at increased risk for low intelligence scores at age 17 years
Term infants with fetal growth restriction are not at increased risk for low intelligence scores at age 17 years Ido Paz, MD, Arie Laor, MD, Rena Gale, MD, Susan Harlap MB,BS, David K. Stevenson, MD, and Daniel S. Seidman, MD Objective: To assess the long-term cognitive outcome of small for gestational age (SGA) compared with appropriate for gestational age (AGA) infants. Design: Data from the Jerusalem Perinatal Study was matched with information from the army draft medical board. SGA and severe SGA were defined as birth weight below the 10th and 3rd percentiles for gestational age, respectively. A multiple linear regression analysis was performed to control for clinical, perinatal, and socio-demographic confounding variables. Subjects: A cohort of 13,454 consecutive singleton term infants born between 1974 and 1976. Main outcome measure: IQ at age 17 years. Results: SGA infants had lower adjusted mean ± SE IQ scores compared with their AGA peers: 102.2 ± 0.9 versus 105.1 ± 0.7 (P < .0001) for males and 102.5 ± 0.9 versus 103.9 ± 0.7 (P < .015) for females. SGA was not associated with lower academic achievements compared with AGA. Conclusion: After controlling for multiple confounders, being born SGA at term is associated with slightly lower intelligence test scores at age 17 years. However, the clinical significance of the small difference is not evident in academic achievements. (J Pediatr 2001;138:87-91)
The “Barker hypothesis,” which suggests that deleterious intra-uterine conditions and impaired fetal growth influence adult health status, is now being extensively studied.1-4 Growth-
restricted infants were found to have an increased risk of ischemic heart disease, cerebrovascular disease, and hypertension.1-4 However, only limited data are available regarding the long-
From the Department of Pediatrics, Maimonides Medical Center, Brooklyn, New York; Medical Statistics Unit, Department of Internal Medicine, Carmel Hospital, Haifa, Israel; Department of Neonatology, Bikur Cholim Hospital, Jerusalem, Israel; Department of Obstetrics and Gynecology, Kaplan Cancer Center, New York University School of Medicine, New York, New York; Division of Neonatology, Department of Pediatrics, Stanford University School of Medicine, Stanford, California; Department of Obstetrics and Gynecology, Chaim Sheba Medical Center, Tel Hashomer, Israel; and Sackler Faculty of Medicine, Tel-Aviv University, Tel Aviv, Israel.
Presented in part at the annual meeting of the Society of Pediatric Research-Academic Pediatric Society, New Orleans, La, May 3, 1998. Submitted for publication Dec 9, 1999; revision received June 13, 2000; accepted July 13, 2000. Reprint requests: Ido Paz, MD, Department of Pediatrics, Maimonides Medical Center, 4802 Tenth Ave, Brooklyn, NY 11219. Copyright © 2001 by Mosby, Inc. 0022-3476/2001/$35.00 + 0 9/21/110131 doi:10.1067/mpd.2001.110131
term cognitive outcome of these newborns.2,5,6 Previous studies examined the intellectual performance up to early school years.5,7-12 Evaluation at this age may not represent the final cognitive outcome of these children.13 Other reports did not adequately control for the effect of potential confounding variables, which may have an important role in determining the cognitive outcome.6-12,14 We previously reported that males born small for gestational age at term had lower educational achievements, and SGA females were found to have significantly lower intelligence test scores.15 However, we were not able to control for the effect of some important covariates such as parental smoking, maternal height, body mass index, and weight gain during pregnancy. Therefore the aim of this study was to reassess, in a larger cohort born 5 years later, the association between fetal growth restriction and cognitive outcome while controlling for multiple covariates. We compared the intelligence test scores at age 17 years of SGA with those of appropriate for gestational age infants. AGA BMI SGA
Appropriate for gestational age Body mass index Small for gestational age
METHODS We performed a cohort study that included all 13,454 consecutive singleton term infants born at the 3 major hospi87
PAZ ET AL
THE JOURNAL OF PEDIATRICS JANUARY 2001
Table I. Percent distribution and mean ± SD characteristics of study population by sex, gestational age, and size at birth
Males SGA n Paternal origin Israel Asia Africa Europe and North America Smoking Paternal Maternal Maternal education (y) 0-8 9-12 >12 Maternal age (y) Birth order 1 2-4 5-6 7-9 ≥10 Maternal BMI Instrument delivery Cesarean delivery Gestational age (wk) Birth weight (g) Fetal distress
585
Females AGA
5928
SGA 359
AGA 3664
14.1 33.6 16.1 36.2
15.6 26.6 19.5 38.3
16.9 31.7 17.2 34.2
15.5 28.2 19.3 37.0
40.5 16.9*
42.3 11.7
46.7 17.9*
43.2 12.1
21.2 41.3 37.5 26.5 ± 4.8*
20.3 42.4 37.3 27.1 ± 4.8
22.2 45.3 32.5 26.1 ± 5.0†
19.1 43.6 37.3 27.0 ± 4.9
41.8* 45.9 8.1 3.3 0.9 21.5 ± 3.0† 7.4 8.6 40.0 ± 1.5 2616 ± 287* 13.0‡
30.6 54.6 9.1 4.3 1.4 22.1 ± 3.1 9.4 6.4 40.0 ± 1.5 3434 ± 390 10.4
43.8† 44.9 7.3 2.7 1.3 21.4 ± 2.9† 6.4 6.7 40.1 ± 1.4 2548 ± 240† 11.4‡
30.2 55.4 9.1 3.9 1.3 22.2 ± 3.1 6.5 5.2 40.0 ± 1.5 3301 ± 376 8.6
Birth order was determined according to the number of older living siblings. SGA, Small for gestational age; AGA, appropriate for gestational age; BMI, body mass index. *P < .001 (SGA vs AGA). †P < .0001 (SGA vs AGA). ‡P < .04 (SGA vs AGA).
tals in Jerusalem between January 1975 and February 1976. Infants with major congenital malformations were excluded. Because the treatment of premature infants has changed dramatically in the last 2 decades, we analyzed data on term (gestational age ≥37 weeks) infants only. From the prospectively collected records of the Jerusalem Perinatal Study, we extracted information regarding maternal age, maternal height and weight, socioeconomic status (based on paternal occupation), ma88
ternal years of education, parental smoking, ethnic origin, birth order (determined according to the number of older living siblings), maternal diseases and weight gain during pregnancy, mode of delivery, fetal distress, birth weight, gestational age, and perinatal outcome.16,17 Gestational age was determined prospectively for each newborn based on the date of the last menstrual period. Great efforts were invested in the post-partum interview to determine the date of the last menstrual period with high accuracy, cor-
rected for the woman’s menstrual cycle regularity. For each gestational age, we delineated the 3rd and 10th percentiles in our study population. SGA was defined for each gestational age as birth weight below the 10th percentile in our entire study population of 13,454 subjects. To examine the strength of the association between SGA and intelligence test scores, we divided the SGA group into severe SGA, defined as birth weight below the 3rd percentile, and moderate SGA, which included subjects whose birth weight was between the 3rd and 10th percentiles on a similar growth curve. The subjects’ medical history, the findings on physical examination, and information about intelligence test scores at age 17 years were available from the military draft records.18 In Israel military service is mandatory, and virtually all Jewish teenagers (95%) are evaluated by the army medical board.18 Intelligence tests were conducted with a translated version of the Otis test and a nonverbal matrices test.19 The test scores were transformed into values that correlated with the Wechsler Adult Intelligence Scale, the most widely used intelligence test for adults.20 To provide an additional clinically meaningful outcome measure, we assessed the academic achievements of our subjects. Low educational achievements were defined as less than 11 years of schooling or attendance at a vocational school. In Israel everyone must attend an educational institution at least up to the 10th grade, and most teenagers join the army at age 18. Therefore years of schooling could not be used as a continuous variable. The 2 databases were matched by using a personal identification number assigned to each individual at birth. The completeness and accuracy of the match were verified by comparing maternal identification number and date of birth. Data were analyzed separately by sex because women who declare
PAZ ET AL
THE JOURNAL OF PEDIATRICS VOLUME 138, NUMBER 1 themselves to be orthodox religious are exempted from military service. Thus grouping men and women together might introduce a bias. Statistical analysis was performed with the non-paired Student t test for continuous variables and with the χ2 or Fisher exact test for dichotomous data. Multiple linear regression analysis (SAS software; SAS Institute, Cary, NC) was used to study the effect of SGA versus AGA on intelligence test scores, while controlling for the effect of potential confounding variables. These include maternal and paternal education, maternal age, parental smoking, ethnic origin, socioeconomic status, birth order, maternal weight gain during pregnancy, diabetes and hypertension during pregnancy, operative delivery, and birth asphyxia. In addition, we adjusted for the possible interaction of maternal height and weight. The overall level of adjustment across confounders was automatically chosen by the software package used (SAS software). We used stepwise logistic regression analysis to examine the effect of SGA on the risk for low IQ (IQ <85) and on the risk for low educational achievement, while controlling for the effect of the same independent variables used in the multiple linear regression analysis. Statistical significance was assumed at P < .05.
RESULTS In the cohort of 13,454 newborns, IQ test results at age 17 were not available for 409 (5.9%) males and 2509 (38.4%) females (Table I). SGA babies were more likely to be born to primiparous women, to mothers with lower BMI, and to mothers who smoked. These infants had a higher incidence of fetal distress during delivery. SGA infants had lower mean ± SD IQ scores compared with their AGA peers: 101.3 ± 13.9 versus 103.4 ± 14.6 (P < .002) for males and 103.6 ± 11.4 versus 105.0 ± 12.0 (P < .01) for fe-
Table II. Intelligence test scores by sex, gestational age, and size at birth
Males n Mean ± SD Adjusted mean ± SE* 95% CI Females n Mean ± SD Adjusted mean ± SE† 95% CI
Severe SGA
Moderate SGA
AGA
154 98.7 ± 13.8 100.7 ± 1.3 98.2-103.1
431 102.3 ± 13.9 102.8 ± 0.9 100.9-104.6
5928 103.4 ± 14.6 105.1 ± 0.7 103.8-106.5
86 104.5 ± 11.7 102.6 ± 1.4 100.0-105.3
273 103.3 ± 11.3 102.4 ± 1.0 100.5-104.3
3664 105.0 ± 12.0 103.9 ± 0.7 102.5-105.3
Severe SGA, Birth weight below the 3rd percentile for gestational age; moderate SGA, birth weight between the 3rd and 10th percentiles for gestational age; AGA, appropriate for gestational age. *P < .0001. †P < .026.
males. The adjusted mean IQ scores ± SE, after controlling for the effect of possible confounders by means of a linear regression model, were 102.2 ± 0.9 versus 105.1 ± 0.7 (P < .0001) for males and 102.5 ± 0.9 versus 103.9 ± 0.7 (P < .015) for females, for SGA versus AGA, respectively. Additional variables that were included in the final regression model were ethnic origin, maternal and parental education, maternal age, birth order, social class, and paternal smoking. Maternal BMI was found to have an independent effect only for males. The variables entered into the regression model accounted for 23% and 28% of the variability of the distribution of intelligence test scores in our population, for males and females, respectively. Babies who were included in the severe SGA group had lower intelligence test scores compared with those who were moderately SGA or AGA. A stepwise logistic regression model did not demonstrate an increase in the risk for low IQ (<85) in those who were born SGA: the odds ratios for IQ <85 were 1.46 (95% CI 0.95-2.26) for males and 1.26 (95% CI 0.82-1.93) for females (Table II). In addition, children who were born SGA were not found to be at increased risk for low academic achievements. The odds ratios were 0.976 (95% CI 0.83-1.15)
and 0.82 (95% CI 0.70-1.12) for males and females, respectively.
DISCUSSION We found that SGA infants have a small, but statistically significant, decrement in intelligence test scores in late adolescence compared with their AGA peers. This association remained statistically significant even after correction for the effect of multiple demographic and clinical confounding factors. The clinical significance of this difference is not clear, especially because being born SGA was not associated with IQ <85 or with low academic achievements. The results of this study are in accordance with our previous findings in an earlier, smaller, birth cohort.15 Sorensen et al6 demonstrated, in a large population-based study, a significant association between birth weight and IQ scores in men during early adult life. However, they could not rule out the influence of potential confounders such as fetal distress, maternal education, or parental smoking. In addition, in our study we were able to determine the appropriateness of birth weight for gestational age, which may serve as a better marker of fetal growth than birth weight alone.4 Martyn et al14 did not find a relationship between 89
PAZ ET AL
body size or proportions at birth and adult intelligence test scores. Despite the impressive duration of follow-up, a bias may have been introduced by the limited (47%) response rate and the effect of confounding variables. There are major advantages to the present study over our previous report. First, the sample size is larger and allows for the detection of subtle differences between the 2 groups. More importantly, we were able to control for multiple covariates including maternal size, nutritional status, maternal diseases, and socio-demographic variables that could influence the cognitive outcome. For example, low socioeconomic status is known to be associated with increased risk for both poor perinatal outcome and low intelligence test scores. Moreover, socioeconomic factors and maternal education were found to have an important influence on the achievements of SGA children.2,5,7,8,12,15 Controlling for maternal body size is important, because this is an important determinant of birth weight, even more than the newborn’s genetic makeup.21 By controlling for maternal height and BMI, we reduced the effect of babies who were constitutionally small and not growth impaired. Male babies whose birth weight was below the 3rd percentile were found to be affected to a greater degree than male infants whose birth weight was below the 10th percentile. This doseresponse–like effect supports the validity of our results, although our study could not establish causality. It should be emphasized that regardless of statistical power, epidemiologic observations are not experiments and should be interpreted with due caution. The pathogenetic basis of these findings is not clear. The in utero malnutrition, which resulted in the growth retardation, may have also affected the developing brain.22 Data from animal models demonstrated damage to the brain of growth-retarded fetuses: diminished brain weight and DNA content and reduced neuropil growth23 90
THE JOURNAL OF PEDIATRICS JANUARY 2001 were found, as well as impaired development and behavior disorders.24 In addition, SGA children are more likely to have behavioral problems.2,10,25 These in turn may have a detrimental effect on the child’s postnatal care, stimulation, and environment—crucial factors for favorable mental development.2,26 Such a process may also contribute to the lower achievements of this group. Premature infants who were born SGA are more likely to have lower cognitive achievements and a higher rate of neurologic abnormalities and learning deficits compared with their gestational age-matched AGA peers.7,11,12 However, the outcome of SGA preterm infants is affected by both their prematurity and their growth retardation.5 Therefore one may speculate that term infants are more likely to compensate fully for early life insult. The results of some studies may have been biased by the inclusion of both term and preterm infants.6,7,9,14,25 Our findings are important because they demonstrate that growth-restricted infants, even of 37 weeks’ gestation or more, may have a long-term deficit. In addition, these findings are valuable, because it has been suggested that term SGA infants are not at greater risk for perinatal morbidity.27,28 Our results may be of importance when obstetricians consider induction of labor near term in a growth-restricted fetus. Being born SGA at term has a small adverse effect on cognitive performance at age 17 years. However, this difference was not associated with low academic achievements or IQ scores <85. Further studies are needed to better define the subgroup of infants with intrauterine growth restriction who may benefit from prenatal intervention or postnatal educational support.5
REFERENCES 1. Barker DJP, Bull AR, Osmond C, Simmonds SJ. Fetal and placental size and the risk of hypertension in adult life. BMJ 1990;301:259-62.
2. Pryor J. The identification of long-term effect of fetal growth restriction. Br J Obstet Gynaecol 1997;104:1116-22. 3. Martyn CN, Gale CR, Jespersen S, Sherriff SB. Impaired fetal growth and atherosclerosis of carotid and peripheral arteries. Lancet 1998;352:173-8. 4. Leon DA, Lithell HO, Vagero D, Koupilova I, Mohsen R, Berglund L, et al. Reduced fetal growth rate and increased risk of death from ischaemic heart disease: cohort study of 15,000 Swedish men and women born 191529. BMJ 1998;317:241-5. 5. Goldenberg RL, DuBard MB, Cliver SP, Cathleen GN, Blankson K, Ramey SL, et al. Pregnancy outcome and intelligence at age five years. Am J Obstet Gynecol 1996;175:1511-5. 6. Sorensen HT, Sabroe S, Olsen J, Rothman KJ, Gillman MW, Fischer P. Birth weight and cognitive function in young adult life: historical cohort study. BMJ 1997;315:401-3. 7. Low JA, Handley-Derry MH, Burke SO, Peters RD, Pater EA, Killen HL, et al. Association of intrauterine fetal growth retardation and learning deficits at age 9 to 11 years. Am J Obstet Gynecol 1992;167:1499-505. 8. Westwood M, Kramer MS, Munz D, Lovett JM, Watters GV. Growth and development of full term nonasphyxiated small for gestational age newborns: follow-up through adolescence. Pediatrics 1983;71:376-82. 9. Tenovuo A, Kero P, Korvenranta H, Piekkala P, Sillanpaa M, Erkkola R. Developmental outcome of 519 small for gestational age children at the age of two years. Neuropediatrics 1988;19:41-5. 10. Parkinson CE, Scrivener R, Graves L, Bunton J, Harvey D. Behavioral differences of school age children who were small for date babies. Dev Med Child Neurol 1986;28:498-505. 11. McCarton CM, Wallace IF, Divon M, Vaughan HG. Cognitive and neurologic development of the premature, small for gestational age infant through age 6: comparison by birth weight and gestational age. Pediatrics 1996;98:1167-78. 12. Hutton JL, Pharoah POD, Cook RWI, Stevenson RC. Differential effects of preterm birth and small for gestational age on cognitive and motor development. Arch Dis Child 1997;76:F75-F81. 13. Eysenck HJ. Measurement of intelligence. Baltimore (MD): Williams and Wilkins; 1973. p. 80. 14. Martyn CN, Gale CR, Sayer AA, Fall C. Growth in utero and cognitive function in adult life: follow up study of
PAZ ET AL
THE JOURNAL OF PEDIATRICS VOLUME 138, NUMBER 1
15.
16.
17.
18.
people born between 1920 and 1943. BMJ 1996;312:1393-6. Paz I, Gale R, Laor A, Danon YL, Stevenson DK, Seidman DS. The cognitive outcome of full term small for gestational age infants at late adolescence. Obstet Gynecol 1995;85:452-6. Harlap S, Davies AM. The pills and births: The Jerusalem Study. Bethesda (MD): National Institute of Child Health and Human Development; 1978. Harlap S, Davies AM, Grover NB, Prywes R. The Jerusalem Perinatal Study: the first decade 1964-1973. Isr J Med Sci 1977;13:1073-82. Kark JD, Kedem R, Revach M. Medical examination of Israeli 17-year-olds before military service as a national resource for health information. Isr J Med Sci 1986;12:318-25.
19. Otis AS. Otis Group Intelligence Scale. New York: Worldbook Publishers; 1919. 20. Seidman DS, Laor A, Gale R, Stevenson DK, Mashiach S, Danon YL. Long term effects of vacuum and forceps deliveries. Lancet 1991;337:1583-5. 21. Brooks AA, Johnson MR, Steer PJ, Pawson ME, Abdalla HI. Birth weight: nature or nurture? Early Hum Dev 1995;42:29-35. 22. Morgane PJ, Austin-LaFrance R, Bronzino J, Tonkiss J, Diaz-Cintra S, Cintra L, et al. Prenatal malnutrition and development of the brain. Neurosci Biobehav Rev 1993;17:91-128. 23. Rees S, Bocking AD, Harding R. Structure of the fetal sheep brain in experimental growth retardation. J Dev Physiol 1988;10:211-24.
24. Smart JL, Dobbing J. Vulnerability of developing brain. II. Effects of early nutritional deprivation on reflex ontology and development of behavior in the rat. Brain Res 1971;28:85-95. 25. Pharoah POD, Stevenson CJ, Cook RWI, Stevenson RC. Prevalence of behavior disorders in low birth weight infants. Arch Dis Child 1994;70:271-4. 26. Watt J. Interaction and development in the first year. II. The effects of intrauterine growth retardation. Early Hum Dev 1986;13:211-23. 27. Jones RA, Robertson NR. Small for dates babies: are they really a problem? Arch Dis Child 1986;61:877-80. 28. Chard T, Yoong A, McIntosh M. The myth of fetal growth retardation at term. Br J Obstet Gynaecol 1993;100: 1076-81.
AVAILABILITY OF JOURNAL BACK ISSUES As a service to our subscribers, copies of back issues of The Journal of Pediatrics for the preceding 5 years are maintained and are available for purchase from Mosby until inventory is depleted. The following quantity discounts are available: 25% off on quantities of 12 to 23, and one third off on quantities of 24 or more. Please write to Mosby, Subscription Customer Service, 6277 Sea Harbor Dr, Orlando, FL 32887, or call 800-654-2452 or 407-345-4000 for information on availability of particular issues and prices. If unavailable from the publisher, photocopies of complete issues may be purchased from Bell & Howell Information and Learning, 300 N Zeeb Rd, Ann Arbor, MI 48106-1346, (734) 761-4700 or (800) 521-0600.
91
The “Barker hypothesis,” which suggests that deleterious intra-uterine conditions and impaired fetal growth influence adult health status, is now being extensively studied.1-4 Growth-
restricted infants were found to have an increased risk of ischemic heart disease, cerebrovascular disease, and hypertension.1-4 However, only limited data are available regarding the long-
From the Department of Pediatrics, Maimonides Medical Center, Brooklyn, New York; Medical Statistics Unit, Department of Internal Medicine, Carmel Hospital, Haifa, Israel; Department of Neonatology, Bikur Cholim Hospital, Jerusalem, Israel; Department of Obstetrics and Gynecology, Kaplan Cancer Center, New York University School of Medicine, New York, New York; Division of Neonatology, Department of Pediatrics, Stanford University School of Medicine, Stanford, California; Department of Obstetrics and Gynecology, Chaim Sheba Medical Center, Tel Hashomer, Israel; and Sackler Faculty of Medicine, Tel-Aviv University, Tel Aviv, Israel.
Presented in part at the annual meeting of the Society of Pediatric Research-Academic Pediatric Society, New Orleans, La, May 3, 1998. Submitted for publication Dec 9, 1999; revision received June 13, 2000; accepted July 13, 2000. Reprint requests: Ido Paz, MD, Department of Pediatrics, Maimonides Medical Center, 4802 Tenth Ave, Brooklyn, NY 11219. Copyright © 2001 by Mosby, Inc. 0022-3476/2001/$35.00 + 0 9/21/110131 doi:10.1067/mpd.2001.110131
term cognitive outcome of these newborns.2,5,6 Previous studies examined the intellectual performance up to early school years.5,7-12 Evaluation at this age may not represent the final cognitive outcome of these children.13 Other reports did not adequately control for the effect of potential confounding variables, which may have an important role in determining the cognitive outcome.6-12,14 We previously reported that males born small for gestational age at term had lower educational achievements, and SGA females were found to have significantly lower intelligence test scores.15 However, we were not able to control for the effect of some important covariates such as parental smoking, maternal height, body mass index, and weight gain during pregnancy. Therefore the aim of this study was to reassess, in a larger cohort born 5 years later, the association between fetal growth restriction and cognitive outcome while controlling for multiple covariates. We compared the intelligence test scores at age 17 years of SGA with those of appropriate for gestational age infants. AGA BMI SGA
Appropriate for gestational age Body mass index Small for gestational age
METHODS We performed a cohort study that included all 13,454 consecutive singleton term infants born at the 3 major hospi87
PAZ ET AL
THE JOURNAL OF PEDIATRICS JANUARY 2001
Table I. Percent distribution and mean ± SD characteristics of study population by sex, gestational age, and size at birth
Males SGA n Paternal origin Israel Asia Africa Europe and North America Smoking Paternal Maternal Maternal education (y) 0-8 9-12 >12 Maternal age (y) Birth order 1 2-4 5-6 7-9 ≥10 Maternal BMI Instrument delivery Cesarean delivery Gestational age (wk) Birth weight (g) Fetal distress
585
Females AGA
5928
SGA 359
AGA 3664
14.1 33.6 16.1 36.2
15.6 26.6 19.5 38.3
16.9 31.7 17.2 34.2
15.5 28.2 19.3 37.0
40.5 16.9*
42.3 11.7
46.7 17.9*
43.2 12.1
21.2 41.3 37.5 26.5 ± 4.8*
20.3 42.4 37.3 27.1 ± 4.8
22.2 45.3 32.5 26.1 ± 5.0†
19.1 43.6 37.3 27.0 ± 4.9
41.8* 45.9 8.1 3.3 0.9 21.5 ± 3.0† 7.4 8.6 40.0 ± 1.5 2616 ± 287* 13.0‡
30.6 54.6 9.1 4.3 1.4 22.1 ± 3.1 9.4 6.4 40.0 ± 1.5 3434 ± 390 10.4
43.8† 44.9 7.3 2.7 1.3 21.4 ± 2.9† 6.4 6.7 40.1 ± 1.4 2548 ± 240† 11.4‡
30.2 55.4 9.1 3.9 1.3 22.2 ± 3.1 6.5 5.2 40.0 ± 1.5 3301 ± 376 8.6
Birth order was determined according to the number of older living siblings. SGA, Small for gestational age; AGA, appropriate for gestational age; BMI, body mass index. *P < .001 (SGA vs AGA). †P < .0001 (SGA vs AGA). ‡P < .04 (SGA vs AGA).
tals in Jerusalem between January 1975 and February 1976. Infants with major congenital malformations were excluded. Because the treatment of premature infants has changed dramatically in the last 2 decades, we analyzed data on term (gestational age ≥37 weeks) infants only. From the prospectively collected records of the Jerusalem Perinatal Study, we extracted information regarding maternal age, maternal height and weight, socioeconomic status (based on paternal occupation), ma88
ternal years of education, parental smoking, ethnic origin, birth order (determined according to the number of older living siblings), maternal diseases and weight gain during pregnancy, mode of delivery, fetal distress, birth weight, gestational age, and perinatal outcome.16,17 Gestational age was determined prospectively for each newborn based on the date of the last menstrual period. Great efforts were invested in the post-partum interview to determine the date of the last menstrual period with high accuracy, cor-
rected for the woman’s menstrual cycle regularity. For each gestational age, we delineated the 3rd and 10th percentiles in our study population. SGA was defined for each gestational age as birth weight below the 10th percentile in our entire study population of 13,454 subjects. To examine the strength of the association between SGA and intelligence test scores, we divided the SGA group into severe SGA, defined as birth weight below the 3rd percentile, and moderate SGA, which included subjects whose birth weight was between the 3rd and 10th percentiles on a similar growth curve. The subjects’ medical history, the findings on physical examination, and information about intelligence test scores at age 17 years were available from the military draft records.18 In Israel military service is mandatory, and virtually all Jewish teenagers (95%) are evaluated by the army medical board.18 Intelligence tests were conducted with a translated version of the Otis test and a nonverbal matrices test.19 The test scores were transformed into values that correlated with the Wechsler Adult Intelligence Scale, the most widely used intelligence test for adults.20 To provide an additional clinically meaningful outcome measure, we assessed the academic achievements of our subjects. Low educational achievements were defined as less than 11 years of schooling or attendance at a vocational school. In Israel everyone must attend an educational institution at least up to the 10th grade, and most teenagers join the army at age 18. Therefore years of schooling could not be used as a continuous variable. The 2 databases were matched by using a personal identification number assigned to each individual at birth. The completeness and accuracy of the match were verified by comparing maternal identification number and date of birth. Data were analyzed separately by sex because women who declare
PAZ ET AL
THE JOURNAL OF PEDIATRICS VOLUME 138, NUMBER 1 themselves to be orthodox religious are exempted from military service. Thus grouping men and women together might introduce a bias. Statistical analysis was performed with the non-paired Student t test for continuous variables and with the χ2 or Fisher exact test for dichotomous data. Multiple linear regression analysis (SAS software; SAS Institute, Cary, NC) was used to study the effect of SGA versus AGA on intelligence test scores, while controlling for the effect of potential confounding variables. These include maternal and paternal education, maternal age, parental smoking, ethnic origin, socioeconomic status, birth order, maternal weight gain during pregnancy, diabetes and hypertension during pregnancy, operative delivery, and birth asphyxia. In addition, we adjusted for the possible interaction of maternal height and weight. The overall level of adjustment across confounders was automatically chosen by the software package used (SAS software). We used stepwise logistic regression analysis to examine the effect of SGA on the risk for low IQ (IQ <85) and on the risk for low educational achievement, while controlling for the effect of the same independent variables used in the multiple linear regression analysis. Statistical significance was assumed at P < .05.
RESULTS In the cohort of 13,454 newborns, IQ test results at age 17 were not available for 409 (5.9%) males and 2509 (38.4%) females (Table I). SGA babies were more likely to be born to primiparous women, to mothers with lower BMI, and to mothers who smoked. These infants had a higher incidence of fetal distress during delivery. SGA infants had lower mean ± SD IQ scores compared with their AGA peers: 101.3 ± 13.9 versus 103.4 ± 14.6 (P < .002) for males and 103.6 ± 11.4 versus 105.0 ± 12.0 (P < .01) for fe-
Table II. Intelligence test scores by sex, gestational age, and size at birth
Males n Mean ± SD Adjusted mean ± SE* 95% CI Females n Mean ± SD Adjusted mean ± SE† 95% CI
Severe SGA
Moderate SGA
AGA
154 98.7 ± 13.8 100.7 ± 1.3 98.2-103.1
431 102.3 ± 13.9 102.8 ± 0.9 100.9-104.6
5928 103.4 ± 14.6 105.1 ± 0.7 103.8-106.5
86 104.5 ± 11.7 102.6 ± 1.4 100.0-105.3
273 103.3 ± 11.3 102.4 ± 1.0 100.5-104.3
3664 105.0 ± 12.0 103.9 ± 0.7 102.5-105.3
Severe SGA, Birth weight below the 3rd percentile for gestational age; moderate SGA, birth weight between the 3rd and 10th percentiles for gestational age; AGA, appropriate for gestational age. *P < .0001. †P < .026.
males. The adjusted mean IQ scores ± SE, after controlling for the effect of possible confounders by means of a linear regression model, were 102.2 ± 0.9 versus 105.1 ± 0.7 (P < .0001) for males and 102.5 ± 0.9 versus 103.9 ± 0.7 (P < .015) for females, for SGA versus AGA, respectively. Additional variables that were included in the final regression model were ethnic origin, maternal and parental education, maternal age, birth order, social class, and paternal smoking. Maternal BMI was found to have an independent effect only for males. The variables entered into the regression model accounted for 23% and 28% of the variability of the distribution of intelligence test scores in our population, for males and females, respectively. Babies who were included in the severe SGA group had lower intelligence test scores compared with those who were moderately SGA or AGA. A stepwise logistic regression model did not demonstrate an increase in the risk for low IQ (<85) in those who were born SGA: the odds ratios for IQ <85 were 1.46 (95% CI 0.95-2.26) for males and 1.26 (95% CI 0.82-1.93) for females (Table II). In addition, children who were born SGA were not found to be at increased risk for low academic achievements. The odds ratios were 0.976 (95% CI 0.83-1.15)
and 0.82 (95% CI 0.70-1.12) for males and females, respectively.
DISCUSSION We found that SGA infants have a small, but statistically significant, decrement in intelligence test scores in late adolescence compared with their AGA peers. This association remained statistically significant even after correction for the effect of multiple demographic and clinical confounding factors. The clinical significance of this difference is not clear, especially because being born SGA was not associated with IQ <85 or with low academic achievements. The results of this study are in accordance with our previous findings in an earlier, smaller, birth cohort.15 Sorensen et al6 demonstrated, in a large population-based study, a significant association between birth weight and IQ scores in men during early adult life. However, they could not rule out the influence of potential confounders such as fetal distress, maternal education, or parental smoking. In addition, in our study we were able to determine the appropriateness of birth weight for gestational age, which may serve as a better marker of fetal growth than birth weight alone.4 Martyn et al14 did not find a relationship between 89
PAZ ET AL
body size or proportions at birth and adult intelligence test scores. Despite the impressive duration of follow-up, a bias may have been introduced by the limited (47%) response rate and the effect of confounding variables. There are major advantages to the present study over our previous report. First, the sample size is larger and allows for the detection of subtle differences between the 2 groups. More importantly, we were able to control for multiple covariates including maternal size, nutritional status, maternal diseases, and socio-demographic variables that could influence the cognitive outcome. For example, low socioeconomic status is known to be associated with increased risk for both poor perinatal outcome and low intelligence test scores. Moreover, socioeconomic factors and maternal education were found to have an important influence on the achievements of SGA children.2,5,7,8,12,15 Controlling for maternal body size is important, because this is an important determinant of birth weight, even more than the newborn’s genetic makeup.21 By controlling for maternal height and BMI, we reduced the effect of babies who were constitutionally small and not growth impaired. Male babies whose birth weight was below the 3rd percentile were found to be affected to a greater degree than male infants whose birth weight was below the 10th percentile. This doseresponse–like effect supports the validity of our results, although our study could not establish causality. It should be emphasized that regardless of statistical power, epidemiologic observations are not experiments and should be interpreted with due caution. The pathogenetic basis of these findings is not clear. The in utero malnutrition, which resulted in the growth retardation, may have also affected the developing brain.22 Data from animal models demonstrated damage to the brain of growth-retarded fetuses: diminished brain weight and DNA content and reduced neuropil growth23 90
THE JOURNAL OF PEDIATRICS JANUARY 2001 were found, as well as impaired development and behavior disorders.24 In addition, SGA children are more likely to have behavioral problems.2,10,25 These in turn may have a detrimental effect on the child’s postnatal care, stimulation, and environment—crucial factors for favorable mental development.2,26 Such a process may also contribute to the lower achievements of this group. Premature infants who were born SGA are more likely to have lower cognitive achievements and a higher rate of neurologic abnormalities and learning deficits compared with their gestational age-matched AGA peers.7,11,12 However, the outcome of SGA preterm infants is affected by both their prematurity and their growth retardation.5 Therefore one may speculate that term infants are more likely to compensate fully for early life insult. The results of some studies may have been biased by the inclusion of both term and preterm infants.6,7,9,14,25 Our findings are important because they demonstrate that growth-restricted infants, even of 37 weeks’ gestation or more, may have a long-term deficit. In addition, these findings are valuable, because it has been suggested that term SGA infants are not at greater risk for perinatal morbidity.27,28 Our results may be of importance when obstetricians consider induction of labor near term in a growth-restricted fetus. Being born SGA at term has a small adverse effect on cognitive performance at age 17 years. However, this difference was not associated with low academic achievements or IQ scores <85. Further studies are needed to better define the subgroup of infants with intrauterine growth restriction who may benefit from prenatal intervention or postnatal educational support.5
REFERENCES 1. Barker DJP, Bull AR, Osmond C, Simmonds SJ. Fetal and placental size and the risk of hypertension in adult life. BMJ 1990;301:259-62.
2. Pryor J. The identification of long-term effect of fetal growth restriction. Br J Obstet Gynaecol 1997;104:1116-22. 3. Martyn CN, Gale CR, Jespersen S, Sherriff SB. Impaired fetal growth and atherosclerosis of carotid and peripheral arteries. Lancet 1998;352:173-8. 4. Leon DA, Lithell HO, Vagero D, Koupilova I, Mohsen R, Berglund L, et al. Reduced fetal growth rate and increased risk of death from ischaemic heart disease: cohort study of 15,000 Swedish men and women born 191529. BMJ 1998;317:241-5. 5. Goldenberg RL, DuBard MB, Cliver SP, Cathleen GN, Blankson K, Ramey SL, et al. Pregnancy outcome and intelligence at age five years. Am J Obstet Gynecol 1996;175:1511-5. 6. Sorensen HT, Sabroe S, Olsen J, Rothman KJ, Gillman MW, Fischer P. Birth weight and cognitive function in young adult life: historical cohort study. BMJ 1997;315:401-3. 7. Low JA, Handley-Derry MH, Burke SO, Peters RD, Pater EA, Killen HL, et al. Association of intrauterine fetal growth retardation and learning deficits at age 9 to 11 years. Am J Obstet Gynecol 1992;167:1499-505. 8. Westwood M, Kramer MS, Munz D, Lovett JM, Watters GV. Growth and development of full term nonasphyxiated small for gestational age newborns: follow-up through adolescence. Pediatrics 1983;71:376-82. 9. Tenovuo A, Kero P, Korvenranta H, Piekkala P, Sillanpaa M, Erkkola R. Developmental outcome of 519 small for gestational age children at the age of two years. Neuropediatrics 1988;19:41-5. 10. Parkinson CE, Scrivener R, Graves L, Bunton J, Harvey D. Behavioral differences of school age children who were small for date babies. Dev Med Child Neurol 1986;28:498-505. 11. McCarton CM, Wallace IF, Divon M, Vaughan HG. Cognitive and neurologic development of the premature, small for gestational age infant through age 6: comparison by birth weight and gestational age. Pediatrics 1996;98:1167-78. 12. Hutton JL, Pharoah POD, Cook RWI, Stevenson RC. Differential effects of preterm birth and small for gestational age on cognitive and motor development. Arch Dis Child 1997;76:F75-F81. 13. Eysenck HJ. Measurement of intelligence. Baltimore (MD): Williams and Wilkins; 1973. p. 80. 14. Martyn CN, Gale CR, Sayer AA, Fall C. Growth in utero and cognitive function in adult life: follow up study of
PAZ ET AL
THE JOURNAL OF PEDIATRICS VOLUME 138, NUMBER 1
15.
16.
17.
18.
people born between 1920 and 1943. BMJ 1996;312:1393-6. Paz I, Gale R, Laor A, Danon YL, Stevenson DK, Seidman DS. The cognitive outcome of full term small for gestational age infants at late adolescence. Obstet Gynecol 1995;85:452-6. Harlap S, Davies AM. The pills and births: The Jerusalem Study. Bethesda (MD): National Institute of Child Health and Human Development; 1978. Harlap S, Davies AM, Grover NB, Prywes R. The Jerusalem Perinatal Study: the first decade 1964-1973. Isr J Med Sci 1977;13:1073-82. Kark JD, Kedem R, Revach M. Medical examination of Israeli 17-year-olds before military service as a national resource for health information. Isr J Med Sci 1986;12:318-25.
19. Otis AS. Otis Group Intelligence Scale. New York: Worldbook Publishers; 1919. 20. Seidman DS, Laor A, Gale R, Stevenson DK, Mashiach S, Danon YL. Long term effects of vacuum and forceps deliveries. Lancet 1991;337:1583-5. 21. Brooks AA, Johnson MR, Steer PJ, Pawson ME, Abdalla HI. Birth weight: nature or nurture? Early Hum Dev 1995;42:29-35. 22. Morgane PJ, Austin-LaFrance R, Bronzino J, Tonkiss J, Diaz-Cintra S, Cintra L, et al. Prenatal malnutrition and development of the brain. Neurosci Biobehav Rev 1993;17:91-128. 23. Rees S, Bocking AD, Harding R. Structure of the fetal sheep brain in experimental growth retardation. J Dev Physiol 1988;10:211-24.
24. Smart JL, Dobbing J. Vulnerability of developing brain. II. Effects of early nutritional deprivation on reflex ontology and development of behavior in the rat. Brain Res 1971;28:85-95. 25. Pharoah POD, Stevenson CJ, Cook RWI, Stevenson RC. Prevalence of behavior disorders in low birth weight infants. Arch Dis Child 1994;70:271-4. 26. Watt J. Interaction and development in the first year. II. The effects of intrauterine growth retardation. Early Hum Dev 1986;13:211-23. 27. Jones RA, Robertson NR. Small for dates babies: are they really a problem? Arch Dis Child 1986;61:877-80. 28. Chard T, Yoong A, McIntosh M. The myth of fetal growth retardation at term. Br J Obstet Gynaecol 1993;100: 1076-81.
AVAILABILITY OF JOURNAL BACK ISSUES As a service to our subscribers, copies of back issues of The Journal of Pediatrics for the preceding 5 years are maintained and are available for purchase from Mosby until inventory is depleted. The following quantity discounts are available: 25% off on quantities of 12 to 23, and one third off on quantities of 24 or more. Please write to Mosby, Subscription Customer Service, 6277 Sea Harbor Dr, Orlando, FL 32887, or call 800-654-2452 or 407-345-4000 for information on availability of particular issues and prices. If unavailable from the publisher, photocopies of complete issues may be purchased from Bell & Howell Information and Learning, 300 N Zeeb Rd, Ann Arbor, MI 48106-1346, (734) 761-4700 or (800) 521-0600.
91