I; in" v'. .‘I: ." ”I?“h " n|ufir+‘ m I" ...—: .I' , I‘IP‘- |' T - \ ‘I 3%.] > I'; ‘ 1",". .‘T{“ —J. 3» 62;? .33"; ":.~.\"-\- ‘ H .2. '("'"5I :“ —- | 'J r“! W .- d (.J‘Kfi, Tr}: \f‘ ," . _' .' ‘ " 1.‘ .l‘ " . A . ‘ I. " ‘ ' . :o‘4:,4~. ’_ ~_ x" 3.} M n ' ‘ ‘ H 7 _ ’. . ‘. . ‘SL‘. ”KW: :.: - Afi'.‘ IV .I (Q r. .4 J _J. .. . H2433. :3 '5‘th! 4: 95;" 3.13;: H1, “* é“? IM'~\ it?!“ L ”SJ. F 5‘ ~ n’t. -"( '. , '. . 9%., n ' is» ma}. -' “ ”6‘"; 9'16. ‘Ooflfi'fidqfifid‘g . :0.“ . ans-Ht ., ,. .. “i Edam-“s IIIIIIIIIIIIIIIIIIIIIIIIIIIIII WWWW“\IWHHIWWNW“NW 3 1293 10702 3768 I fi—mctnwus.“ i-‘g. ”new”. 3 y. l ’.‘ . 4' ":9 x J 1. ’I' ‘7' .. ‘ a m . n n a I p I - . H. Gut-mar: ' ~33 '.-.-¢ n " ' .. 1'? 'o D. n - 3'“ -U' L . .A ‘ ~12 Jpn-f .4 A, £‘Iflfl'\-IN'KUI This is to certify that the dissertation entitled ETIOLOGY AND PROGNOSIS FOR FIVE YEAR OLD CHILDREN WITH HEMORRHAGIC ENDOVASCULITIS OF THE PLACENTA presented by Lori Eliicott has been accepted towards fulfillment of the requirements for Ph.D. degree in Counseh’ng Psycho]ogy WWW [YWF MS U is an Afl’trnmtt'vc Action/Equal Opportunity Institution 0-12771 __.___—______, MSU LIBRARIES :— RETURNING MATERIALS: P1ace in book drop to remove this checkout from your record. ‘FINES wiII be charged if book is returned after the date stamped be10w. 4““ ETIOLOGY AND PROGNOSIS FOR FIVE YEAR OLD CHILDREN WITH HEMORRHAGIC ENDOVASCULITIS OF THE PLACENTA By Lori Ann Ellicott A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling and Educational Psychology 1985 O (/0 5‘0 9 LORI ANN ELLICOTT 1985 ABSTRACT ETIOLOGY AND PROGNOSIS FOR FIVE YEAR OLD CHILDREN WITH HEMORRHAGIC ENDOVASCULITIS OF THE PLACENTA by 4 Lori Ellicott Five year old children with hemorrhagic endovasculitis (REV) of the placenta were examined on various outcome measures with the objective of determining the nature and extent of sequelae for liveborn children with this placental lesion. Prior studies had found significant associations with stillbirth, small for gestational age, and maternal hypertension or pre-eclampsia. In light of these risk factors, it was hypothesized that HEV children might be at risk for later problems. Thus, the study set out to assess the relationship of various pre-, peri-, and postnatal factors with sequelae. The study consisted of examining medical records to assess the health status of mother and infant, and assessing the children on the following measures: The McCarthy Scales of Children’s Abilities, Bender Visual Motor Gestalt Test, Personality Inventory for Children-Revised, and Dimensions of Temperament Survey-Revised. Maternal interviews elicited demographic and health and pregnancy information. Two control groups, matched on birth date and location, were used in the design. The Registry control group consisted of children with a variety of pregnancy and/or birth complications but not HEV. The community control group was comprised of healthy, term Lori Ellicott infants. Significant confounding variables included chronic villitis and maternal hypertension. Presence of chronic villitis was significantly associated with increased difficulties on the personality measure across groups. Presence of maternal hypertension significantly correlated to increased approach behavior. Additionally, for the HEV group, presence of maternal hypertension correlated to several dimensions of more difficult temperament. Increased rates of abnormal scores for the HEV group were noted when chronic villitis, hypertension, or both were present. It appears that children with HEV affected placentas are at risk for poorer prognosis at five years of age, particularly in the face of chronic villitis or hypertension. However, children with chronic villitis of the placenta and/or presence of maternal hypertension regardless of group membership are at increased risk for personality and temperament difficulties at five years of age. To my husband Vern and daughter Christine whose love, support, and laughter have warmed me. iii TABLE OF CONTENTS List of Tables......... ............. ...... ...... ...... ...... .....vi Chapter I............................. ......................... .. 1 Importance of the Study.... ..... ..... .......... .................1 Statement of Purpose............................................2 Background of Theory and Research.......... ..... ................3 Limitations of the Study. ..... ..... .................. . .......... 5 Chapter Summary ......... .......... .......................... ....6 Chapter II........................................................7 Review of Related Literature....................................7 Placental Pathology and Viral Infections...... ....... . ..... .....7 Chronic Villitis.................... ...................... ....9 Villitis of Unknown Etiology..................................9 Congenital CMV Infection ...... ..... ............ ..............lO Summary......................................................12 Etiology of Intrauterine Growth Retardation....................13 Placental Conditions.........................................14 Maternal Factors....... .................................. ....16 Socio-economic Factors.......................................19 Summary......................................................19 Classification Systems for Fetal Growth Retardation............20 Prognosis for Small for Gestational Age Infants................23 Birth Weight by Gestational Age..............................24 Physical Growth........................ ..... .................24 Neurological Growth.................... ..... ... ..... .........28 Intellectual and Academic Growth....... ..... .................34 Psycho/Social Growth............ ........... ..................37 Summary................. ............................... ......38 The Ponderal Index (PI)......................................40 Physical Growth...... ................. . .................... ..40 Neurological Growth... ................... ....................41 Intellectual Growth.... ..................................... .43 Psycho/Social Growth.................... ......... ............43 Summary..................... .............................. ...45 Timing of Onset of Fetal Growth Retardation. ........ .........45 Physical Growth.......................... ....... ..... ..... ...46 Neurological Growth. ......................................... 47 Intellectual and Academic Growth...... ..... . .............. ...47 Psycho/Social Growth............. ...... . ....... ..............49 Summary............................... ............ . ....... ...49 A Review and Synthesis. ........................................ SO Hypotheses............. ........................................ 51 Chapter III ....................................... . .............. 54 Methodology......................... ...................... .....5 Design ..... ..................... ...... . .................... ....5 Description of the Pilot Study ....... . ...... .................5 Population...................... ........ ................. ..... .5 iv Selection Procedures ......................................... 55 Contact Procedures... ........... . ............................ 57 Instruments......... ............................. . .......... ...58 Procedures ........ . ..... ... ............ .... .................. 59 Methods of Analysis ............. . ......... . ................... .60 Chapter Summary... ..... ....................:. ................ ..62 Chapter IV............. ...................... ... .............. ...63 Results............................ ...... . ........ . ...... ......63 Description of the Population...... .......................... ..63 Statistical Analysis and Findings.. .......... .... .......... ....68 Chapter V ......... ...... ........................................ 115 Discussion......................................... ........... 115 Future Directions for Research ..... . ................... .......126 Conclusions........................................ ....... ....128 Glossary............................. ..... ...... ............... .132 Appendix A The Michigan Placental Tissue Registry .................. . ..... 134 Appendix B The Colorado Intrauterine Growth Chart ........................ 135 Appendix C Maternal and Child Questionnaires ........... ... .............. .136 Bibliography...... ............. . ............................... .146 Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table 1. 2. 3 A 13. 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 33. 34. 35. LIST OF TABLES Mean Values for Various Continuous Variables by Group Membership Marital Status of Parents by Group Membership Sex and Race by Group Incidences of Pre- and Perinatal Factors by Group Membership A List of Continuous Variables of Importance to the Study A List of Discrete Variables of Importance to the Study Outcome Measures Mean MSCA Scores by Group Membership and Sex Univariate F Tests for Group Effects on the MSCA Multivariate Effects for the MSCA Univariate F Tests for the Three-Way Analysis on the MSCA Multivariate Effects for the Three-Way Analysis on the MSCA - _ Mean MSCA Scores for the Three-Way Analysis Group Size for the Three-Nay Analysis on the MSCA Univariate F Test for Group Effects on the PIC-R Psychosis Scale Mean PIC-R Scale Scores by Group Membership and Sex Multivariate Effects for the PIC-R Univariate F Tests for Chronic Villitis on the PIC-R Univariate F Tests for Group Effects on the PIC-R Univariate F Tests for Maternal Hypertension on the PIC-R Mean PIC-R Scores for the Three-Way Analysis Group Size for the Three-Way Analysis on the PIC-R Multivariate F Tesrs for the Three-Nay Analysis on the PIC-R Mean DOTS-R Dimensions by Group Membership and Sex Correlation Coefficients of Sex to the DOTS-R Dimensions Mean DOTS-R Dimensions for the Three-Way Analysis Group Size for the Three-Way Analysis on the DOTS-R Univariate F Tests for Maternal Hypertension Effects on the DOTS-R Univariate F Tests for Group by Maternal Hypertension Interaction Effects on the DOTS-R Mean Bender Gestalt Developmental Scores by Group and Sex Tests of Significance for the Bender Gestalt Using Sequential Sums of Squares for Group and Sex Tests of Significance for the Sender Gestalt Using Sequential Sums of Squares for the Three-Way Analysis Mean Bender Gestalt Scores for the Three Way Analysis Degree of Handicap by Group Membership Mean Handicap Scores by Group and Sex vi Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. SO. 51. 52. 53. 54. 55. 56. S7. 58. 59. 60. Tests of Significance for Handicap Using Sequential Sums of Squares for Group and Sex Tests of Significance for Handicap Using Sequential Sums of Squares for the Three-Way Analysis Mean Handicap Score for the Three-Way Analysis Normal and Abnormal Outcomes by Group Membership Normal and Abnormal Group Outcomes by Chronic Villitis, Maternal Hypertension, and Other Associations Normal and Abnormal Outcome by HEV, Chronic Villitis, and Maternal Hypertension Correlation Coefficients of Pre-, Peri-, and Postnatal Factors to the MSCA Correlation Coefficients of Pre-, Peri-, and Postnatal Factors to the PIC-R Scales Correlation Coefficients of Pre-, Peri-, and Postnatal Factors to the DOTS-R Dimensions Correlation Coefficients of Pre-, Peri-, and Postnatal Factors to the Bender Gestalt and Handicap Mean Socio-Economic Variables by Group Membership Correlation Coefficients for Socio-Economic Variables the MSCA Scales Correlation Coefficients for Socio-Economic Variables the PIC-R Scales Correlation Coefficients for Socio-Economic Variables the DOTS-R Dimensions Correlation Coefficients for Socio-Economic Variables the Bender Gestalt and Handicap Correlation of Small for Gestational Age Infants with Outcome Measures Characteristics of Children with Early and Late Onset Delayed Fetal Growth Kruskal-Wallis One Way ANOVA on Kruskal-Wallis One Way ANOVA on Kruskal-Wallis One Way ANOVA on Kruskal-Wallis One Way ANOVA on Handicap Correlation Correlation Correlation Dimensions Correlation Gestalt and and and and and of MSCA PIC-R Scales DOTS-R Dimensions Bender Gestalt and the the the the the MSCA the PIC-R Scales the DOTS-R and and and Between Viral Infection Between Viral Infection Between Viral Infection the Bender Between Viral Infection and Handicap vii CHAPTER I Importance of the Study Hemorrhagic Endovasculitis (HEV) is a recently identified abnormality of human placentas (Sander, 1980). This lesion focuses on placental blood vessels, destroying fetal red blood cells and causing hemorrhage within the placenta. Since its initial discovery in 1977 at the Michigan Placental Tissue Registry (see Appendix A) nearly 202 of the Registry’s placentas show some degree of HEV. In a review of 218 HEV affected placentas, Stevens (1984) found 51% (112/218) associated with stillborn infants in contrast to 22% (89/400) of the Registry control placentas. The Registry control placentas had various pregnancy, delivery, and/or placental abnormalities, but not HEV. Additionally, liveborns with HEV affected placentas were significantly smaller for gestational age (332 to 19%, using a criterion of weight less than the 10th percentile for gestational age) and were more likely to be female (57% to 48.4%) than were Registry controls. (Infants who are small for gestational age, or sga, are inadequately grown for their gestational age. Both preterm and fullterm infants comprise this group.) Other significant positive associations with HEV included later gestational age, smaller placentas, and maternal hypertension pd or pre-eclampsia. No association was found for maternal age, race, parity, infant Apgar score or fetal anomalies. The prevalence of RSV in the general population is not known at this time. The focus of research in this area has been largely on stillborn cases and placentalybathology in both liveborn and stillborn infants. Thus there is a strong need to investigate liveborn children with HEV affected placentas particularly in light of later sequelae stemming from the early risk factors of maternal hypertension or pre-eclampsia and small for gestational age. A pilot study to begin assessing children with HEV affected placentas was initiated in 1983. The original cases of children with HEV affected placentas identified in 1977 were located and assessed on cognitive and personality scales at approximately five years of age. Fourteen of the original 17 children were available for follow up. They were matched on birthdate and location to a Registry control group of children. Results of a paired t test found the General Cognitive Index on the McCarthy Scales of Children’s Abilities to be significant (p < .01). This indicated that children with HEV affected placentas preformed poorer on this summary measure of cognition than their matched controls. Additionally, the children with HEV affected placentas exhibited more extreme scores on personality indices. However, the differences were not statistically significant. Statement of Purpose It is the purpose of this study to follow up children with a specific placental abnormality identified as hemorrhagic endovasculitis (HEV) on various measures of perception, cognition, (\J temperament, personality, and health and physical status. The children will be assessed at approximately five years of age. The outcomes of the children with HEV affected placentas will be compared to those of a community control group and a Registry control group. The outcome variables will be statistically examined so as to identify the nature, incidence, and severity of sequelae for the HEV children. Particular attention will be given to factors in the pre-, peri-, and postnatal environments associated with poorer outcomes. Background of Theory and Research Several theories pertain to this study. Foremost are the contrasting continuums of reproductive casualty and caretaking casualty posited by Pasamanick and Knobloch (1961) and Sameroff and Chandler (1975) respectively. These theories address the issue of a linear (predictive) relationship between perinatal history and outcome. The continuum of reproductive casualty suggests a relationship between certain harmful events of pregnancy and delivery which result in damage to the fetus or neonate primarily localized to the central nervous system. The lethal events on the continuum are evidenced in abortions, stillbirths, and neonatal death, whereas the sublethal components result in cerebral palsy, epilepsy, mental retardation, and behavioral and'learning disorders. Evidence for the continuum of reproductive casualty largely has come from retrospective studies which have implicated anoxia (also referred to as asphyxia or hypoxia in the literature; its presence indicates a reduction in oxygen level below the requirements of the organism), prematurity, delivery complications and social factors in eventual outcome (Knobloch & Pasamanick, 1966; Pasamanick & Knobloch, 1961, 1966; Pasamanick & Lilienfeld, 1955). In contrast to this is the continuum of caretaking casualty noted by Sameroff and Chandler (1975) with its largely prospective focus. The authors argue that certain pregnancy and/or birth complications are recognized as affecting newborn behavior and intellectual functioning through the preschool period, but these deficits are greatly reduced as the child enters school. Moreover, social and economic factors tend to either reduce or amplify these deficits, leading to the conclusion that social and economic factors in the child’s environment exert stronger influences on the course of deve10pment than does perinatal history. Substantiating this point of view is the observation that many children suffering perinatal problems have normal developmental outcomes. Thus, a transactional model where the child and environment mutually influence each other was felt to be necessary to understand outcomes. To the extent that the child elicited or was provided with a positive environment, enhanced outcomes resulted. Conversely, when a child elicited negative responses from the environment, the child was at greater risk for the development of later problems (Sameroff, 1975; Sameroff & Chandler, 1975; Sameroff & Seifer, 1983). A third theory pertaining to this study centers around the etiology of HEV and its implications for developmental sequelae. Sander (1980), notes that many features of HEV simulate the placental response to some viral infections that may occur during pregnancy. The most suspect infectious agents are of the TORCH group (which includes toxoplasmosis, rubella, cytomegalovirus, syphilis,and other infectious agents). The literature is suggestive of a wide range of effects from perinatal viral infections, including fetal death, intrauterine growth retardation, stillbirth, death in infancy, congenital anomalies (referring to any abnormality present at or detectable immediately after birth resulting from such perinatal environmental insults as infectious agents or teratogens), late onset of congenital disease or defects, infections without evidence of damage, and no apparent damage (Hanshaw & Dudgeon, 1978; Sever, 1970). The precise nature of the fetal damage is suggested to vary depending on the type of infectious agent, the gestational age at which the infection occurs, and the natureof the transmission from mother to fetus. The point of departure for this paper will be to examine the relationship of HEV’s infectious association with the continuums of reproductive and caretaking casualty. Thus the factors in the pre-, peri-, and postnatal environments will be examined in light of outcome measures, with the goal of delineating those factors that are associated with performance at five years of age. Limitations of the Study There are several sources of bias affecting this study. Among them are the effects of migrator bias. It may well be that children unavailable for follow up due to moving outside the geographic scope of this study may have differed systematically from those who remained within the parameters of the study. Recall bias is another source of potential confound. Mothers may have recalled only certain aspects of their pre-, peri-, and postnatal health status while omitting some relevant details. However, this is minimized somewhat by the use of medical records to supplement maternal reports. Further confounding this was the maternal interviewer’s knowledge regarding the status of the subjects. The interviewer for the mothers may have held diagnostic suspicions or expectations which influenced the intensity with which a question was asked of the mother, the number of repetitions of a question, depth of pursuit or inquiry, or systematic errors in recording results. This problem was lessened by utilizing interview schedules (see Appendix C). One pathologist is responsible for identifying presence of HEV of the placenta at the Registry. This is a strength in consistency but examiner bias becomes an issue. However, this is somewhat minimized due to a series of slides sent to other pathologists who have concurred with the original findings. Chapter Summary This chapter has attempted to delineate the scope of the study to follow. Specifically, it has identified children with HEV affected placentas as being the target population. It was noted that HEV is a recently discovered placental abnormality in need of investigation regarding sequelae in liveborn children. An association with viral infection has been noted by Sander (1980). Sequelae at five years of age with various pre-, peri-, and postnatal conditions will be the subject of inquiry in the ensuing study. CHAPTER II Review of Related Literature The ensuing review of literature will progress from postulated mechanisms of HEV to a review of related outcomes. Since chronic villitis (a chronic inflammation within placental villi due to a variety of infectious agents) and sga infants are both significantly associated with HEV (602 and 33% respectively, Stevens, 1984), these areas will be examined in some depth. The chapter will begin with an examination of placental pathology and viral infections since the latter have been suspect in the etiology of HEV. A discussion of the etiology of intrauterine growth retardation will follow. Classification systems for fetal growth retardation will be surveyed, concluding with a review of prognosis for sga infants. Placental Pathology and Viral Infections The well being of the newborn infant depends not only on its genetic endowment, but also on its mother’s health and management through pregnancy and birth. Thus, the role of the placenta in maternal and fetal welfare is attracting increased attention. Abnormalities of the placenta have been associated with an increased incidence of abortion, prematurity, stillbirth, bleeding in pregnancy, and intrauterine growth retardation, the latter of which is relevant to the study at hand (Altshuler, Russell, & Ermocilla, 1974; Scott & Usher, 1966). Despite the high incidences of mortality and morbidity associated with placental pathology, the placenta has been a largely neglected organ for investigation when compared to other organs of the human body. Part of this neglect stems from the economic and time constraints faced by obstetricians, pediatricians, and pathologists. Since it is impractical to examine all delivered placentas, those submitted tend to be of the stillborn or critically ill newborn. Thus there tends to be very little information about the placenta in relation to many conditions, including two fairly common viral infections; rubella and cytomegalovirus disease (CMV) since infants tend to survive the neonatal period in the case of the former, and in the latter instance most infections are not apparent. As a diagnostic intrument representing fetal and maternal tissues, the placenta can be useful in identifying the presence of infectious agents. Such was the case with rubella, where the first clue that the fetus could become infected from maternal rubella came from joint examination of the placenta and fetus (Hanshaw & Dudgeon, 1978). The placenta is similarly useful in evidencing CMV infection in the fetus or neonate, however, unlike rubella, the placenta is one of several modes of transmission of this disease. CMV is currently acknowledged as being the leading cause of congenital infection, surpassing even the rubella virus (Griffiths, Campbell- Benzie & Heath, 1980). Viral infections have been implicated in poor fetal and infant outcome for some time. Damage may take the form of intrauterine growth retardation, stillbirth, infant death, or mild to severe forms of congenital defects (Jones 8 Roberton, 1984). There may also be no incidence of damage. Outcome is suggested to depend on the type of infectious agent, the gestational age at which infection occurs, and the route by which the agent is transmitted from mother to fetus or neonate (e.g. breast milk, placenta, respiratory route, sperm, etc.). Chronic Villitis Approximately 602 of HEV affected placentas show chronic inflammation within villi, or chronic villitis compared to a 20% overall incidence in Registry specimens (Sander & Stevens, 1984; Stevens, 1984). This is a non-specific response to a variety of infectious agents, in particular those of the TORCH group. Viral type intranuclear inclusion bodies have been found in 10.23% (71/694) of HEV affected placentas from 1978 to 1980. Excluded were nine placentas with CMV inclusions associated with known intrauterine CMV infection. The majority of HEV placentas evidencing chronic villitis have no known etiology. Villitis of Unknown Etiology (VUE). In most instances chronic villitis is of unknown etiology. It is suggested that this lesion is many times more common than placental toxoplasmosis, CMV, rubella, syphilis and herpes infections combined (Russell, 1980). Studies have failed to reveal any etiologic agent, Specifically excluding the TORCH agents (Altshuler et al., 1974). It is implicated in the etiology of intrauterine growth retardation, as shall be evidenced in a subsequent section. There are several studies investigating VUE. The various studies come to slightly different conclusions. For example, Altshuler et a1. (1974) believe that placental infection is the cause of VUE; a conclusion drawn in part from evidence of intrauterine infection in some of the cases. This differs somewhat from the view of Labarrere, Althabe, and Talenta (1982) who postulate an immunological origin (which often is in response to a viral agent) similar to that postulated in placentas of pre- eclamptic pregnancies wherein the maternal immune system attacks the placental tissue. Russell (1980) notes that VUE is probably not a single, circumscribed, and easily defined entity that responds to only one organism. He notes that there are no obvious signs of intrauterine infection in the majority of neonates and so argues for a local source of infection in most instances. The absence of infection in mothers and infants are seen as attesting to this conclusion. Knox and Fox (1984) suggest that ethnic, environmental and socio-economic factors affect the incidence of villitis. They caution against assuming that all villitis are of infective origin, despite the documentation that many specific maternal infections (e.g. rubella, toxoplasmosis, etc.) can cause a villitis. In short, it seems that the infectious origin (whether it be local, infecting only the placenta; evidenced in the mother or neonate; or an abnormal immune response) is the predominant view for the etiology of VUE. However, in the absence of identification and isolation of an infectious agent, this view remains inconclusive. Congenital CMV Infection. Several studies have investigated the sequelae of CMV infection. It will be noted here due to its association with chronic villitis and its incidence in a small 10 percentage of HEV affected placentas identified between 1979 and 1982. Nine examples of chronic villitis associated with CMV have had coexistent HEV (Sander & Stevens, 1984). CMV infection has been shown to cross the placenta and infect and damage the developing fetus. An incidence of 0.5% to 2.4% has been noted, with an increased incidence being noted in the lower socio-economic classes (Griffiths et al., 1980; Hanshaw, Scheiner, Moxley, Gaev, Abel & Scheiner, 1976; Keirse, 1984; Marx, 1975). Spread of the virus is facilitated by crowding and lack of sanitation facilities (Marx, 1975). In a study of neonates with clinical evidence of congenital CMV infection, 35-50% are sga (Allen, 1984). Pass, Stagno, Myers and Alford (1980) point out that 10% of infants in this country acquire CMV during the first year of life, however Marx (1975) states that acquiring the infection after birth has minimal effects. At present, there is no treatment for CMV infection nor a vaccine to prevent it. The sequelae for congenital CMV infection is a wide range of clinical manifestations. Mortality rate for children with symptomatic congenital CMV infection varies from six to 30% (Allen, 1984). It is noted to be the most common viral cause of mental retardation, surpassing even the rubella virus (Marx, 1975; Pass et al., 1980). Significant hearing impairments have been noted in these children (Griffiths et al., 1980; Hanshaw et al., 1976; Melish & Hanshaw, 1973; Pass et al., 1980) as have higher than expected incidences of rare opthalmological abnormalities (Frankel, Keys, Hefferen, Rola-Pleszczysnki & Bellanti, 1980; Pass et al., 1980). Microencephaly, neuromuscular disorders, minimal brain dysfunction, 11 lower 10 scores and school failure are also among the clinical manifestations of congenital CMV infection (Griffiths et al., 1980; Hanshaw et al., 1976; Marx, 1975; Melish & Hanshaw, 1973; Pass et al., 1980). Keirse (1984) notes that about 90% of infants with CMV infection have no clinical manifestations at birth but some may reveal some developmental abnormalities by two years of age. Some degree of growth retardation occurs in approximately 40% of infants who reveal clinical manifestations of infection at birth. Many factors may be associated with the clinical outcome of the virus, including gestational age at the time of infection, sex of the fetus, and individual differences in host susceptibility (Hanshaw, 1973; Kumar, Gold, Jacobs, Ernhart & Nankervis, 1984). However, the general conclusion in the reviewed studies appears to be that children with congenital CMV infection are at increased risk for handicaps that will significantly impair normal development. Summary It was pointed out that the placenta is useful as a biopsy specimen for identifying the presence of infectious agents in fetal and maternal tissues. Potential outcomes of viral infections were noted and the role of chronic villitis of both known and unknown etiology were reviewed as they pertain to a presumed infectious origin. It was found that the TORCH group is implicated in chronic villitis where infection can be identified, but there is presently a lack of evidence to implicate any viral agents (TORCH or others) as being causative in VUE with any certainty. However, due to the known response of chronic villitis to the TORCH agents, viral infections continue to be suspect in the etiology of VUE by a number 12 of pathologists despite the aforementioned problem. Thus the implication is that if VUE is a response to viral infections, then the clinical manifestations of children with VUE might be expected to resemble such sequelae as is noted for rubella or congenital CMV infections and other agents. This has implications for 60% of the HEV children whose placentas had signs of chronic villitis. The reviewed literature suggests the importance of examining the association of chronic villitis and congenital viral infection to sequelae in our study. The increased incidence of chronic villitis and congenital infection among children with HEV affected placentas leads us to expect poorer results on various outcome measures for this group. Etiology of Intrauterine Growth Retardation Thirty three percent of HEV affected placentas are associated with sga births. SGA children are a distinct group born either preterm or fullterm who are markedly underweight for their gestational age and thus intrauterine malnutrition is presumed (Caputo & Mandell, 1970). Therefore it may be helpful to survey some of the known factors associated with delayed fetal growth. We should note at the onset that the definition of IUGR is far from uniform. Various criteria have been used, including the third or fifth or tenth percentiles, or two standard deviations from the mean on differing growth curves. (This study has used the Colorado Intrauterine Growth Chart to determine the presence of IUGR in the study subjects - see Appendix B.) Different labels are used interchangeably to denote this group of children, including small for date (sfd), small for gestational age (sga), intrauterine growth 13 retardation (IUGR), fetally malnourished, low ponderal index (low PI), and growth retarded infants. An incidence of two to ten percent of IUGR (below the tenth percentile) is noted in the general population (Douglas, 1982; Norman, 1982). In addition, there is evidence suggesting that mothers whose first baby was sga have a three times higher risk of their second infant being sga than the population as a whole. This seems to indicate that a number of factors which retard fetal growth have a high recurrence rate (Keirse, 1984). Placental Conditions The placenta has been implicated in various ways in IUGR. For example, Altshuler et al. (1974) note an incidence of 92% (N=63) of gross and/or microscopic placental lesions in their sga subjects. Forty percent of the placentas evidenced circulatory disturbances, and 64% of this group (or 16 of 25) had pre-eclampsia of pregnancy. The role of maternal vascular disease (which includes hypertension, pre-eclampsia, antepartum hemorrhage, diabetes mellitus and chronic nephritis) has been further investigated in studies by Sheppard and Bonnar (1980) and DeWolf, Brosens, and Renaer (1980). The former study investigated 25 normal pregnancies and 25 pregnancies complicated by IUGR (birth weight below the fifth percentile for gestational age) and found evidence suggesting that lesions in the maternal uterine vasculature may be the major cause for impairment of placental function, marked by degenerative changes in placental villi. The conclusion of maternal origin is arrived at since fetal capillaries and blood cells appear structurally normal. These findings are similar to the study of five IUGR infants (below the tenth percentile for gestational age) by DeWolf et al. (1980) who found that maternal vascular lesions in the placenta impaired blood flow and resulted in infarction and IUGR. The cause of the vascular lesions are unclear. However the possibility that underlying latent hypertensive or renal vascular disease may be the cause of the lesions in the placenta, for which fetal growth retardation is the first clinical manifestation is posited. The severity of maternal vascular disease, and hypertension in particular, was not seen as being correlated to the degree of IUGR. Chronic villitus of unknown etiology (VUE) has been the subject of several studies investigating IUGR. In a study by Labarrere et al. (1982) the authors examined seven or eight sections of the placenta and found a very high incidence of chronic villitis in both the sga group (less than the tenth percentile on the ponderal curve) and in the normal control group (86% and 26% for the sga group and controls respectively). Ten percent of villi within each placenta were inflamed in the sga group compared to 1.2% in the control group. In contrast to these findings, Altshuler et al. (1974) found an incidence of 24% of VUE with sga infants (less than the tenth percentile for gestational age). Russell (1980) found that VUE was associated with IUGR and that the severity of the villitis process correlated positively with degree of IUGR and perinatal mortality. Some additional placental characteristics have been associated with IUGR. For example, the absence of one umbilical artery has been implicated in fetal growth retardation as has smaller size of the placenta. However in the latter case where smaller fetuses are 15 observed to have smaller placentas, it is not certain what is cause and what is effect (Miller, 1983). Additionally, microscopic studies suggest that placentas of sga infants may receive insufficient oxygen, suggesting that the cause of fetal growth retardation is not placental damage, but restricted blood flow from the mother, which has implications for a restriction of nutrient transfer as well. Placenta previa has been shown to retard all parameters of fetal growth depending on the frequency and duration of bleeding (Crawford, 1982). Placental metabolism is also critical in supporting a pregnancy. Animal studies point out that the placenta consumes a major portion of the nutrient supply provided by the mother; up to 50% of the oxygen and 75% of the glucose put into uterine circulation by the mother in fetal lambs (Miller, 1983). Maternal Factors Various maternal characteristics have been associated with IUGR. Some of these have been previously mentioned which affect the placenta. These include the various vascular diseases (pre- eclampsia, antepartum hemorrhage, hypertension, diabetes mellitus, chronic nephritis). In one way or another, these conditions are thought to interfere with the nutrition of the infant (Chamberlain, 1984; Dawkins, 1965; Galbraith, Karchmar, Piercy, & Low, 1979; Keirse, 1984; McDonald, 1965; Schutt, 1965; Tejani, 1982). Intrauterine infections of the TORCH group are also implicated in IUGR, as has been previously noted. Maternal habits, such as smoking, alcohol, and drug abuse are well documented correlates to fetal growth retardation. Smoking impairs both fetal length and weight and has been reported to cause 16 an increased rate of spontaneous abortion, prematurity, abruptio placenta, placenta previa, and premature rupture of the membranes. All growth parameters are affected when the mother is a heavy smoker (greater than 15 cigarettes a day) and the number of cigarettes per day has been shown to relate directly to the degree of growth retardation and to the incidence of prematurity. When smoking occurs in the presence of other known growth retarding factors, the result is more than additive (Chamberlain, 1984; Crawford, 1982; Keirse, 1984; Norman, 1982). Excessive ingestion of alcohol by pregnant women has lead to the identification of a pattern of physical malformations in the offspring known as fetal alcohol syndrome or FAS (Abel, 1980). FAS is noted by a distinct pattern of congenital malformations which includes microcephaly, shortened palpebral fissures (the size of the eye opening) and joint, limb, and cardiac anomalies. Additionally, most of the affected infants are growth retarded in both length and weight (Abel, 1980; Clarren & Smith, 1978; Norman, 1982). Fetal weight may be reduced as much as 1200 grams. The influence on growth begins in the first half of pregnancy and continues into postnatal life (Keirse, 1984). Whether these effects are the direct result of alcohol or due to conditions secondary to alcohol intake (e.g. altered nutrition) is inconclusive at this time (Robertson, 1983). Alcohol consumption does not necessarily lead to the FAS. Studies have linked heavy drinkers (greater than 45 drinks per month) to weight, length, and head circumference below the tenth percentile, despite not evidencing PAS (Keirse, 1984). 17 A high incidence of IUGR occurs with chronic heroin addiction as well. However, since other growth retarding factors frequently accompany drug abuse, it is uncertain whether fetal growth retardation results from the drug itself, from secondary factors or from a combination of these (Crawford, 1982). Chronic maternal disease is frequently noted to result in IUGR. Such conditions as cyanotic congenital heart disease, advanced diabetes, and renal and gastrointestinal disease have been implicated (Crawford, 1982). Multiple gestation is similarly noted to frequently result in fetal growth retardation and is most common in monozygotic twins. Intrauterine crowding and limited nutrient supply are among the possible causes. This may be compounded by such conditions as a small uterus or pelvis, or presence of tumors or ovarian cysts which further inhibit fetal growth, whether it is a multiple or single gestation (Crawford, 1982; Galbraith et al. 1979; Kopp & Parmelee, 1979; North, 1966). Various chromosomal and congenital anomalies have been associated with IUGR (Collins & Turner, 1971; Crawford, 1982; Schutt, 1965). Among these are Trisomy 18 and 4p (short arm deletion). Trisomy 21 (Down’s syndrome) infants are affected to a lesser degree, as are infants with sex chromosome abnormalities. Congenital anomalies have been noted to be of higher incidence among very low birth weight infants (less than 1500 grams) and among the premature sga infants. 18 Socio-economic Factors Socio-economic factors have long been associated with IUGR and poorer fetal outcome in general (Sameroff, 1975; Sameroff 8 Chandler, 1975). For example, the lower classes are less likely to receive prenatal care or adequate nutrition during their pregnancy (Drillien, 1970). Women of low socio-economic status are more likely to adapt practices that adversely affect growth, such as drug or alcohol use or smoking. However if they did not adopt such practices, these women were no more likely to have infants with lower birth weights than women in other socio-economic groups (Keirse, 1984). Amongst mothers of low socio-economic status, there tends to be a higher incidence of various complications of pregnancy, including antepartum hemorrhage and toxemia (Caputo & Mandell, 1970; McDonald, 1965). Increased rates of family instability and unemployment as well as minority group membership are often found in lower socio- economic classes as well. Thus, a multiplicity of stressors in the prenatal environment may have their effect in poorer prognosis at birth and in subsequent development, as shall be discussed later. Summary The primary mechanisms postulated to produce fetal growth retardation include impairment of oxygen and nutrient transport which are activated in a variety of ways, including conditions affecting: 1) the placenta (e.g. vascular disease, chronic villitis, placenta previa, placental metabolism); 2) maternal factors (e.g. habits such as smoking, alcohol and drug use; intrauterine infection, chronic disease, multiple gestation); and 19 3) socio-economic factors (e.g. inadequate nutrition, increased incidences of pregnancy complications, minority group membership, family and environmental stressors). A number of potential confounds have been pointed out in the preceding section which will be measured in this study in order to understand group differences on the various outcome measures. Factors of interest include the presence of maternal disease (such as hypertension, diabetes, and pre-eclampsia or toxemia), smoking, alcohol consumption, drug use, multiple gestation, race, and income. All of these factors are associated with fetal growth retardation and poorer prognosis for the child. Since the HEV group has a higher incidence of IUGR (Stevens, 1984), this study expects to find increased incidences of growth retarding factors and poorer outcomes for this target group of children. Classification Systems for Fetal Growth Retardation Various means of classifying the fetally malnourished infant have been devised. The most common involves an examination of birth weight by gestational age and comparing this to a particular growth curve (Lubchenco, Hansman, Dressler & Boyd, 1963; Yerushalmy, 1967). However, disparty is still common even amongst these methods. For example, some curves include all births, others exclude stillbirths, multiple births or both. Some include only gestations known with certainty, while others include approximations or estimations of gestational age. In some, gestational age means completed weeks, or zero to six days, while others consider it to mean plus or minus three days (Keirse, 1984). Various criteria have been used to designate the sga infant; typically the third, fifth, or tenth 20 percentile, and most commonly the latter. The ensuing discussion will highlight other means for denoting these inadequately grown infants. Keirse (1984) points out the existence of important differences between birth weight standards which relate to racial, genetic, socio-economic and environmental factors in the various growth curves. This points to the strong need to select standards which were obtained on a population that is similar to the one under study. The ponderal index (PI) is perhaps the next most commonly used method for classifying inadequately grown infants after birth weight by gestational age methods. The PI yields an index to indicate if an infant is underweight for its length (weight in grams times 100 devided by the cube of body length in centimeters). Low PI newborns are typically less than or equal to the third or fifth percentile on a scale plotting PI as related to gestational age. High PI infants are greater than or equal to the 95th or 97th percentile. For example, a weight-length ratio of less than 2.17 for the 37 week gestation infant and 2.20 for the baby with 38 or more weeks gestation has been used in the diagnosis of fetal malnutrition. Unlike weight for gestational age norms, this index of fetal malnutrition is relatively independent of sex, race, and parity of the newborn and the physical size of the mother. Infants with low Pl’s have little or no subcutaneous fat and visually appear malnourished. Some infants who were diagnosed as sga according to birth weight by gestational age (less than the third percentile) were considered well grown because their PI’s were well above the 21 third percentile and their crown-heel lengths were normal (Miller & Hassanein, 1971). Crane and Kopta (1980) describe two types of IUGR, symmetric and asymmetric. The former refers to the subsequent and pr0portionate reduction in size of all body organs whereas the latter affects some body organs more than others. Etiology is distinct for these two types of growth retardation. Factors responsible for symmetric growth retardation include intrauterine infections (e.g. TORCH agents), chromosome abnormalities, congenital anomalies, maternal malnutrition, and smoking. Ultrasonically, the symmetric sga fetus demonstrates a proportionally reduced growth potential prior to the third trimester. Etiologic factors in asymmetric IUGR include uteroplacental insufficiency and is commonly seen in pregnancies complicated by chronic hypertension, pre- eclampsia, and advanced diabetes. Ultrasonically, growth is essentially normal until some point in the third trimester following which fetal head growth is relatively spared but other structures evidence reduced growth (Norman, 1982). Differential diagnosis is made ultrasonically and/or in light of the existence of factors which predispose to one of the forms of IUGR. All 33 infants in the Crane and Kopta study (1980) plotted below the tenth percentile for gestational age, however, only 50% of the asymmetrically growth retarded infants had low PI’s and 54% of the symmetrically growth retarded had low PI’s (less than the tenth percentile for gestational age), thus indicating that the IUGR sample may vary considerably according to the use of a particular diagnostic classification system. 22 Jones and Roberton (1984) do well to point out that there exists a group of infants who show symmetrical reduction in size at birth, but who are small for genetic reasons. Ethnicity, maternal size, and sonic-economic factors influence the size of some of these small but normal babies. It seems clear that the absence of uniformity in methods for determining which infants are sga can produce a heterogenous group of growth retarded infants. This has implications for the next section which surveys prognosis. Specifically, it suggests the liklihood that different classification methods will select out different subgroups of sga children, thus raising the possibility of differential outcomes. Accordingly, the ensuing review will be organized by method of classification. Birth weight by gestational age studies will be reviewed first, followed by studies employing the PI, and lastly, those making reference to the timing of onset of fetal growth retardation. Prognosis for Small for Gestational Age Infants A wealth of material exists regarding outcome for infants labeled sga. It will be the purpose of this secion to review the findings and to draw some conclusions regarding this heterogenous group of children. Before the review commences, it seems noteworthy to point out that medical management of newborns has changed drastically in the past 30 years. Thus, the deleterious outcomes so long associated with infants of low birth weight (weight less than five and one half pounds at birth), of whom some are sga, may have been due in part to such former neonatal practices as prolonged fasting and administration of lethal and sublethal concentrations of 23 oxygen to the immature neonate. Thus, a more optimistic outlook in recent studies might be expected to reflect recently improved standards of care (Hack & Fanaroff, 1984; Nelson & Ellenberg, 1984; Sokol, 1984). Birth Weight by Gestational Age By far the majority of studies have classified the fetally malnourished infant according to various weight by gestational age curves. The most frequently used criterion for determining the sga infant is less than or equal to the tenth percentile. Note will be made in the ensuing studies wherever this standard is deviated from. Within this method of classification, we will examine in sequence, physical, neurological, intellectual and academic, and psycho/social growth patterns. These distinctions are somewhat arbitrary and will overlap to a certain extent, yet nonetheless seem necessary in order to assimilate the studies to follow. Physical Growth. In a study investigating the later growth patterns of sfd infants (eliminating cases complicated by prematurity, defined as gestation less than 37 completed weeks; multiple gestation; chromosomal defects and congenital anomalies; and using a criteria of two standard deviations below the mean for weight by gestational age), Fitzhardinge and Steven (1972) found most aspects of growth still retarded at four to six years. Later growth could not be predicted by the degree of weight retardation at birth, but was found to relate instead to the rate of growth in the first six months. Of those sfd infants attaining normal height, a significantly greater than normal growth velocity was noted to occur during the first year. Among the sfd children who remained shorter 24 than normal, a higher proportion were in the lower income group, thus implicating poor home environment as affecting growth. Assessment of these same children in adolescence while controling for SES and maternal size (excluding those who had moderate or severe asphyxia as defined by five minute Apgar scores and/or occurrence of first breath) revealed significant deficits in height, weight, and head circumference at 13 to 19 years of age (Westwood, Kramer, Munz, Lovett & Watters, 1983). Similar results were obtained in the studies by Low, Galbraith, Muir, Killen, Karchmar and Campbell (1978) and Low, Galbraith, Muir, Killen, Pater, and Karchmar (1982). They noted small but significant differences in the growth measures between the mature IUGR and matched control children from birth to 60 months. Additionally, the IUGR group demonstrated accelerated growth during the three months after delivery, but this failed to result in catch up effects in later childhood. An increased prevalence of maternal smoking and a lower male-female sex ratio of children were noted in the IUGR group; either of which may have functioned so as to lower mean growth scores. Neligan, Kolvin, Scott and Garside (1976) investigated growth parameters of four groups of children; normal controls, premature, sfd, and vsfd (very small for dates which were less than the fifth percentile) children and found that the vsfd group measured consistently and significantly shorter and lighter than the premature children up to seven and a half years of age. The sample subjects had a history of known growth retarding perinatal factors. Social class was not associated with increased growth in the sfd group. Kumar, Anday, Sacks, Ting and Delivoria-Papadopoulos (1980) 25 examined preterm and fullterm lbw infants. Infants who were appropriately grown for their gestational age (aga) and sga infants (greater than two standard deviations and less than two standard deviations respectively) who had various perinatal risk factors and neonatal complications were included. At one year of age, weight less than the third percentile occurred in eight percent of the age group and in 46% of the sga group. Height less than the third percentile occurred in 13.5% and 38% of the aga and sga infants respectively. None of the sga infants had heights or weights greater than the 50th percentile at one year. These findings compare to the results of Commey and Fitzhardinge (1979) who assessed preterm sga (less than two standard deviations) infants (excluding congenital and chromosomal anomalies) during the first two years of life. Most children continued to show growth retardation. Thirty four percent had heights and weights below the third percentile at two years of age. Accelerated growth velocity beginning at or slightly before the term date and for six months thereafter was noted. However, because of the early postnatal delay in growth, the cumulative linear growth from birth to six months postterm was virtually identical to the standard for normal term infants. Evidence for prolonged growth deficits despite accelerations in growth velocity in the first year of life are reflected in other studies. In their study of preterm sga (less than two standard deviations) and aga infants, Hack and Fanaroff (1984) found catch up growth through eight months in both groups. Some of the aga infants continued to catch up during the second year. By the third year, 26 46% of the sga and 17% of the aga infants continued to be subnormal in weight. Fewer sga infants with subnormal intrauterine head growth caught up in weight by age three than those with normal intrauterine head growth. This seems to suggest that poor catch up growth occurs in sga infants who suffered from early, prolonged, and severe intrauterine growth failure. The studies of Vohr, Oh, Rosenfield, Cowett and Berstein (1979) and Vohr and on (1983) followed a sample of preterm sga and aga infants over a five year period. Excluded from the study were children with congenital anomalies and genetic defects. Catch up effects were noted, resulting in comparable weight and height measurements between the two groups at one year. At two through five years, age infants were significantly heavier. Significantly greater height was noted at two, three, and five years in the age group. SES was significantly correlated to outcome. Similarly, Tudehope, Burns, O’Callaghan, Mohay and Silcock (1983) note comparable physical growth parameters on their 12 month old preterm vlbw (very low birth weight) aga and sga children. However, both groups continued to evidence subnormal growth at this corrected age, with 20% of age and 30% of sga children with weights still less than two standard deviations below the mean. This difference was not significant and did not predict later developmental outcome. In their study of Indian preterm aga and sga children, Bhargava, Kumari and Choudhury (1984) note catch up effects in the age group at three to four years, so that at five to six years, this group was comparable to normal children on measures of height and weight. In contrast the sga group showed no catch up effects from birth to six years of age. Neurological Growth. Several studies have examined the incidence of neurological sequelae in sga infants. One of the problems of these studies is that oftentimes the neurological measures are poorly identified in the study or not identified at all, making comparison of findings rather difficult. Neurological measures can include hard and/or soft signs, encompassing seizures, cerebral palsy, speech defects, hyperactivity, learning difficulties, etc. Fitzhardinge and Steven (1972) followed mature sfd infants over a minimum of five years on a variety of measures and compared the results to a sibling group. The sfd group had more speech defects and a higher incidence of minimal brain dysfunction, characterized by hyperactivity, short attention span, learning difficulties, poor fine coordination, and hyper-reflexia than did the control group. No differences in hearing or vision were observed between groups. Some form of speech abberation was noted in 33% of the sfd boys and 26% of the sfd girls. Neligan et al. (1976) similarly found significantly increased incidences of both soft and hard neurological deficits in vsfd infants as compared to normal controls or children with a shorter gestation at six and seven years. Moreover, this was significant for boys and not for girls, suggesting a greater vulnerability for boys. Social class was related to neurological abnormality for both soft and hard signs for the sfd (including vsfd) and control groups. Additionally, language skills were significantly poorer at five years of age in the sfd and vsfd groups compared to controls, and at seven years between the vsfd group and the control group. 28 Rubin, Rosenblatt and Balow (1973) found significantly higher incidences of neurological abnormalities for both their preterm and mature lbw groups as compared to their higher birth weight preterm and term infants during the neonatal period, at 12 months, and at seven years. The fullterm, lbw infants correspond to a criterion of sga infants less than the tenth percentile. The preterm lbw group is most probably comprised of a heterogenous group of sga and age infants; though the authors are rather vague on this. At four months the preterm lbw group had the most neurological abnormalities (as assessed by 73-123 items administered by a physician specially trained in pediatrics or neurology). Additionally, motor development at eight months and language skills at five years were inferior in both lbw groups. Overall, the premature lbw group tended to score worse than the mature lbw (sga) group. Drillien (1970) found an increased incidence of major and minor congenital anomalies and neurological defects (defined as presence of cerebral palsy, epilepsy, or both) in the sfd group compared to other lbw groups. The total incidence of all congenital anomalies was 47% in the sfd group (less than the tenth percentile) compared to 36% and 25% respectively in the ten to 25 percentile group and the group of infants with weights greater than the 25th percentile. In a later study (1972) the author found a higher incidence of transient abnormal neurologic signs suggestive of cerebral palsy during infancy but who were neurologically normal at one year among the sfd group. Of those infants showing these transient signs, there were increased incidences of mental retardation or borderline retardation by two to three years of age. 29 In considering Drillien’s findings, it is important to note that these sfd infants were heterogenous in regards to gestational age, thus blurring the distinction between outcomes for premature and term fetally malnourished infants and later sequelae. Commey and Fitzhardinge (1979) similarly found an increased incidence of neurological handicaps in their sga group. Twenty one percent had major neurological defects (e.g. hydrocephaly, cerebral palsy, recurrent nonfebrile seizures) and 42% had Bayley scores less than or equal to 80. These children were premature and evidenced a variety of pregnancy and neonatal complications. Presence of cerebral depression upon admission to the Neonatal Intensive Care Unit was significantly associated with later handicap. Birth asphyxia was an earlier complication in 21 of the 24 surviving infants admitted with cerebral depression, 15 of whom were later handicapped. No relationship was shown between neurological sequelae and SES, degree of IUGR, or degree of prematurity. Incidence of neurological defects have been further investigated by Ellenberg and Nelson (1979). These authors found that mature sfd infants composed a small nonsignificant subgroup who later had cerebral palsy or seizure disorders. In the case of the former, lbw and short gestation were more important risk factors. Similarly, Collins and Turner (1971) found a subgroup of sfd term infants in the patient records of aproximately 14% of their mentally retarded sample. IUGR comprised the second highest risk category in a study by Fitzhardinge, Kalman, Ashby and Pape (1978) investigating handicapping defects. Fifty three percent (21/40) of the sga children had major handicaps at two years of age (defined as the 30 presence of a major neurological defect such as epilepsy, cerebral palsy, or hydrocephalus and/or a mean Bayley score of less than 80), four presented with neurological defects only, seven with neurological defects and low Bayley scores, and ten with low Bayley scores in the absence of major neurological defects. In order to assess the effects of IUGR apart from premature birth and subsequent postnatal complications, the sga infants were paired with an aga infant of the same birth weight and sex and a similar neonatal course. The sga infants had significantly more major neurological defects and lower Bayley scores than the age infants, thus implicating poor fetal growth in poorer prognosis. In their study of lbw infants in India, Bhargava et al. (1984) note that developmental lags in motor and language skills were apparent in the extreme lbw (less than 1500 grams) and sga infants compared to preterm aga and normal babies. In particular, the sga group manifested developmental arrest in the acquisition of the skill of abstraction. This became more pronounced with age. Unfortunately, neither measures, methods, nor statistical findings were reported in the study. Kitchen, Ford, Orgill, Rickards, Astbury, Lissenden, Bajuk, Yu, Drew and Campbell (1984) in their study of very lbw infants (500 to 999 grams) found a 30% (three of ten) incidence of severe functional handicap in their two year old sga children. Severe functional handicap was considered present in children who had one or more of the following: moderate or severe cerebral palsy, a Bayley Mental DeveIOpmental Index less than 69 (less than two standard deviations below the mean), sensorineural deafness 31 requiring the use of hearing aids or bilateral blindness. No statistically significant differences in social or perinatal variables were noted between the age, sga or multiple birth groups. Several studies have examined the motor development of sga infants in an attempt to determine the incidence of neurological defect. Michaelis, Schulte and Nolte (1970) examined mature sfd infants presenting various birth and pregnancy complications (e.g. multiple gestation, forceps delivery) during the first week of life. They found significantly more abnormalities suggestive of greater neurological immaturity in this group as compared to a mature aga group of infants. However, at nine months of age, the sfd group had normal developmental quotients on the Gesell, suggesting that the previously noted immaturity is not persistent in its effects (Parmelee & Schulte, 1970). Similarly, Paine and Pasquali (1982) found no evidence of neurological impairment as assessed by motor development in their group of Brazilian term sfd infants at four, eight, twelve and eighteen months. They excluded from their sample conditions known to be associated with poor intrauterine growth (e.g. congenital anomalies, chromosomal abnormalities, prenatal infection, maternal chronic disease). At 18 months, lower SES and increasing maternal parity were the most important sources of variation in explaining lower psychomotor scores. Several studies reflect this more optimistic note regarding neurological sequelae. Tudehope et al. (1983) found nonsignificant differences on neurological and sensory handicap status between their premature vlbw aga and sga children. In their study of preterm sga and age infants, Hack and Fanaroff (1984) note a greater 32 incidence of neurological sequelae in the age group, including increased rates of transient hypotonia, cerebral palsy, hydrocephalus and blindness. This was felt to be attributable to the shorter mean gestation for the aga group compared to the sga group. Vohr and Oh (1983) note comparable rates of major neurologic abnormalities (seizures, spastic diplegia, hemiplegia) between their groups of five year olds who were preterm age and sga at birth. However, a 26% incidence of minor neurologic abnormalities (language delay, fine motor inefficiency and attention deficit) were noted in the sga group compared to 12% in the age group. SES was significantly correlated to developmental outcome. Developmental test scores were significantly lower for the sga group from nine months to three years. At five years of age, the two groups were comparable, suggesting the importance of following these children over time. Parmelee and Schulte (1970) examined term sfd children with various known growth retarding factors and found they performed similarly to the normal controls at nine months of age. They also compared favorably regarding neurological maturity as determined by nerve conduction velocities to the normal control group. The Low et al. study (1982) of term IUGR children from one to six years of age similarly included various known factors affecting fetal growth. Despite this, no significant differences on motor or language development, or in visual or hearing defects were noted in the IUGR group compared to the aga control group. However, the IUGR group had a higher proportion of motor and language handicaps. Slightly elevated WPPSI full scale scores at 60 months were noted for males in the IUGR group compared to the male controls (106 and 103 33 respectively). The Westwood et al. study (1933) found no major organic neurologic signs in the nonasphyxiated subgroup of sga infants in adolescence. Similarly Vohr et al. (1979) found comparable outcomes at one year with sga and age children. At 18 months, the sga group had lower Bayley scores but this difference was not present at 24 months. The more favorable findings in these last few studies may reflect in part different birth cohorts and changes in neonatal care in more recent years. Another possible explanation is that it may reflect the less stringent criterion of defining the sga samples as less than or equal to the tenth percentile on growth curves or to the exclusion of certain subgroups of sga infants, as in the Westwood et al. study (1983). In other words, the inclusion of more adequately grown healthy infants may have the effect of raising the mean group scores on various outcome measures. Intellectual and Academic Growth. A pattern of normal IQ scores but academic problems are suggested by various studies of sga children. Hack and Fanaroff (1984) note nonsignificant differences between their preterm sga and aga children. IQ scores on the Stanford Binet at three years were slightly lower for the sga group, but mean scores were still in the normal range. Babson and Kangas (1969) note 10 scores in the normal range at four years of age for their term sfd group and a nonsignificant three point difference with the control group. Fitzhardinge and Steven (1972) similarly note normal IQ scores in their mature sfd group at four, six, and eight years of age, with boys scores averaging six points less than the girls scores (95 compared to 101). However, a high incidence of 34 school failure was reported in the sfd group, with 50% of boys and 36% of girls functioning poorly in school. Seven of the 16 unsatisfactory scholars had been assigned to either a special school or a special class. The remainder were consistently failing subjects in the regular classroom. Evidence of academic problems were not apparent in the sibling control group. No relation between degree of poor fetal growth and either intelligence or school performance were noted. Normal 10 scores were again noted in the Rubin et al. study (1973), however significant differences were found in the two lbw groups (a preterm and a fullterm group) compared to the higher birth weight groups. The proportion of retentions, special class placements and use of special services were more prevalent in the two lbw groups, with more special educational problems being noted in the term lbw group (or sfd) than in the preterm lbw group. A 50% incidence of special educational problems were noted in the sfd group. Two thirds of boys in both lbw groups accounted for a higher proportion of school identified educational problems than did the girls of similar birth weight. Francis-Williams and Davies (1974) note the same pattern of normal IQ scores which were significantly lower for the preterm sfd group compared to the aga group. This group of sfd children had various pre- and perinatal complications and were over represented in the lower classes, thus confounding these findings. Neligan et al. (1976) found normal IQ scores at five years of age for sfd and vsfd groups, however these scores were significantly lower than those for normal controls. A battery of cognitive and sensory motor functions were administered at five to seven years of age with the general 35 finding that the vsfd group most often performed the least maturely of all the groups. However, sex and social class were significantly associated with outcome. More recently, Drillien. Thomson, and Burgoyne (1980) investigated sequelae of sfd infants with variable term status at six years of age. They found that the presence of transient abnormal neurological signs in the first year of life was predictive of schooling problems in the sfd and lbw groups. This finding may be due to an increased incidence of postnatal complications as compared to the neurologically normal infants in the first year. Intelligence scores were normal but slightly lower in the sfd group. Correlated to low scores on measures of intelligence and educational achievement were low social class, male sex and early intrauterine insult. Cases with known growth retarding factors were not excluded from this study. This is similar to the nonsignificant differences on multiple measures of cognitive development at two to five years of age which were noted in the Low et al. study (1982). In this study, more sfd children failed senior kindergarten (15% compared to seven percent in the age group), however the difference was not significant. Socio-economic status correlated to outcome in this study. Tudehope et al. (1983) note lower, but nonsignificant differences between their groups of preterm sga and aga children at three to four years of age on the McCarthy Scales of Children’s Abilities. Westwood et al. (1983) further elucidate these findings in their assessment of adolescents who were unasphyxiated sga newborns. They found nonsignificant trends towards lower scores on the WAIS and WISC when compared to the normal birth weight control group, while controlling for SES and maternal size. However, most scores in both groups were well in the normal range. In an attempt to reconcile these relatively mild findings with the severe cognitive deficits reported by Fitzhardinge and Steven (1972) on the same subjects at a younger age, the authors compared the 11 excluded asphyxiated sga infants to the nonasphyxiated sga infants. The mean full-scale IQ was 87.1 in the asphyxiated subgroup compared to 101.1 in the nonasphyxiated subgroup. Thus, the poorer outcomes for sga children may result from perinatal asphyxia or other complications in the neonatal period, rather than from IUGR. Psycho/Social Growth. One might expect to find an increased incidence of psycho/social problems (either at home or at school) in children evidencing significantly more school related problems. However, the studies cited regarding school problems failed to investigate what may be concomitant problems to academic difficulty. The Neligan et al. study (1976) however, does examine psycho/social sequelae at five to seven years of age. In the area of behavior, significantly more abnormalities were noted in the vsfd group as compared to the premature group according to mother, teacher, psychologist, and psychiatrist reports. Boys in both groups were significantly more likely to exhibit behavioral abnormalities than were girls. Additionally, social class particularly in the sfd group, was related to behavioral abnormality. A hyperactive pattern was especially pronounced in the sfd group. This resembles the higher incidence of hyperactive behaviors found in the Fitzhardinge and Steven study (1972). Measures of temperament showed few 37 differences at age five between groups, however of the eight items that were worse for the vsfd children, five were related to a pattern of high activity. This pattern of high activity continued to be apparent at seven years in both the vsfd and short gestation groups. Personality scores at seven years of age on an extraversion and neuroticism indice revealed nonsignificant differences between groups (Neligan et al. 1976). Bhargava et al. (1984) note poorer outcome in adaptive and personal social skills in their sga group of Indian children. However, as noted previously, measures, methods, and statistical analyses are not reported. Summary. In studies which have examined sga children using a criterion of various growth curves, evidence indicates a normal pattern of physical maturation. However the children remain consistently shorter and lighter throughout childhood and adolescence, particularly in studies using stricter criteria for sample selection or those excluding certain unhealthy subgroups of sga infants. (An exception to this may be a subgroup of preterm sga infants whose early delivery facilitates normal growth and prevents sequelae associated with fullterm pregnancy complications.) Normal intelligence is consistently found, but higher rates of neurological abnormalities, including motor, speech and language delays, presence of cerebral palsy, epilepsy, mental retardation, sensory loss, hyperactivity, and academic problems have been reported in several studies, although the differences are not always significant. The majority of studies contained subjects with known growth retarding factors (e.g. maternal chronic disease, multiple gestation, etc.) and various neonatal complications including asphyxia and/or preterm 38 status which may tend to obscure the findings. It may well be that certain subgroups of sga children do have an increased incidence of neurological defects and academic difficulties which tend to elevate the mean incidence of problems for the sga group as a whole (Ounsted, 1970). Sex and social class tend to be the most frequently cited covariates to outcome in the studies reviewed thus far, with parity and neonatal complications being implicated as well. Those children with the best overall prognosis tend to be mature female children of higher SES with an absence of known growth retarding factors in their pre- or perinatal history. There is some evidence to suggest that more recent birth cohorts may fare better too; but this is not consistent. The preceding discussion suggests a number of significant factors to measure in this study. Presence of IUGR will be important to note since this correlates to later neurological, academic and behavioral difficulties. This leads us to expect poorer outcomes for children with HEV affected placentas due to the increased rate of IUGR in this group. Male sex and increased parity are also considered to be risk factors in some of the studies reviewed and so will be of concern to us here. Various complications of birth, such as cesarean delivery, fetal distress, lbw, and length of gestation are also linked in the literature to poorer outcome and will therefore be examined here. Additionally, SES has correlated to prognosis, therefore various components of SES will be measured in the study at hand, including parental education, income level, marital status, and maternal age at time of delivery. 39 The Ponderal Index (PI) It has previously been pointed out that the various classifi- cation systems are not mutually exclusive, nor reliable in selecting sga subjects. For example, fetal growth retardation in low PI infants is noted to begin a few weeks before birth, thus overlapping with the asymmetric form of fetal growth retardation to be discussed in the next section (Walther & Ramaekers, 1982a, 1982b; Zeskind & Ramey, 1978). In spite of this, the ensuing discussion will look towards the expanding knowledge base regarding sequelae for sga children as contributed to by this method. Physical Growth. Healthy mature infants were examined in a study by Davies, Platts, Pritchard and Wilkinson (1979) which compared disproportionally grown sga infants (less than the third percentile) to proportionally grown sga infants (greater than the tenth and less than the 50th percentile). They found the , disproportionately grown (or late onset of IUGR) infants grew more rapidly during the first month than did the proportionately grown (or early onset) infants. Thereafter, growth rates were similar. Both groups had greater growth velocities in the first three months after birth than did normal term infants. This period of catch up effects was also noted in the Walther and Ramaekers study (19823) which assessed healthy term sga and age children. The sga group evidenced a greater growth velocity than the controls during the first six months. Despite the above average rate of growth, 13 out of 25 sga infants (compared to two out of 25 controls) had weights below the tenth percentile at three years of age and nine (compared to three controls) were below the tenth percentile for length. 40 Neurological Growth. Walther and Ramaekers (1982b) note that language development is dependent on the biological maturation of the brain, and as such is useful in assessing higher cerebral function in preschool children. Thus, term sfd infants (who were free from fetal disease and neonatal morbidity) and matched normally grown controls were assessed at the age of three years on the Reynell Developmental Language Scales. The mean raw scores and corresponding mean developmental ages were worse for the sfd children than for the controls. The differences were significant and amounted to four months for verbal comprehension and five months for expressive language. Language retardation (scores less than or equal to two standard deviations below the mean) was observed in eight percent of the controls compared to 32% of the sfd group. Additionally, the incidence of behavior problems in language delayed sfd children was high as was neurological dysfunction in sfd children with either delayed or adequate language. These findings compare to the results in the Fitzhardinge and Steven study (1972) previously reviewed who used a different measure of IUGR but found an elevated incidence of speech problems and minimal brain dysfunction in their three and four year old subjects. Zeskind and Ramey (1978) investigated motor development in low PI and normal PI infants. Fullterm full birth weight black infants from low SES families, with no serious prenatal or perinatal complications were studied. They were assigned to either an instructional day care program or to a nonintervention control group. At three months of age, both low PI groups (i.e. the day care program and control conditions) showed significantly lower 41 motor develOpment scores than the normal PI group. A significant difference between scores in low PI and normal PI infants in the home control group was still present at 18 months, however the difference between low and normal PI infants in the day care program had disappeared. Additionally, normal PI infants in the day care program had higher motor development scores than normal PI infants in the control group at 18 months of age. Thus the differences that characterized the fetally malnourished group at three months were ameliorated in the supportive day care environment. In another study examining motor development during the first year of life, Goggin, Holmes, Hassanein and Lansky (1978) found no significant differences between healthy fullterm low, normal, and high PI infants. However, this study used subjects with diverse socio- economic status which may account for discrepant findings from the Zeskind and Ramey study. . The newborn period has been the subject of investigation in several studies examining neurological sequelae in low PI infants. For example, Zeskind and Ramey (1981) analyzed the cry features of low, normal, and high PI fullterm full birth weight infants who were without evidence of pre- or perinatal complications. They found that infants at both extremes of the PI were characterized by cry features which have been shown to distinguish infants whose central nervous system functioning has been impaired (e.g. threshold, latency, activity, and fundamental frequency of cry). Als, Tronick, Adamson and Brazelton (1976) examined caucasian healthy term infants on the Brazelton Neonatal Behavioral Assessment Scale during the first ten days after birth and found significant differences between 42 the low and normal PI groups on motor behavior and interactive processes. Organization of state and physiological stability were comparable between groups. All newborns were informally assessed later prior to nine months on the Denver Developmental Screening Test and were found to perform within normal limits, thus offering similar results to the Goggin et al. study (1978). Intellectual Growth. The Zeskind and Ramey studies (1978, 1981) investigated cognitive development in inadequately grown children. They found at 18 and 24 month assessments, no significant differences between low and normal PI children in the day care setting. However, low PI children showed significantly lower scores at 36 months than did normal PI children within the control environment. Thus the detrimental effects of fetal malnutrition in lower class black children were observed at three years in an intellectually nonsupportive environment, but were ameliorated with the provision of an intellectually supportive atmosphere. Psycho/Social Growth. Several studies have investigated different aspects of the parent-child relationship in low PI children. Receiving particular research focus have been various interactional variables, of which the investigations of Zeskind and Ramey have made especially noteworthy contributions (1978, 1981). The authors implicate the simultaneous action of various agents, which taken together have a total effect that is greater than the sum of the parts. For example, it is argued that the removal of one factor in the fetally malnourished infant’s environment may prevent part of the cyclical occurrence of nonoptimal development. This can be seen in the parent-child relationship that develops with the low 43 PI infant who had qualitatively distinct cry features in infancy. At 18 and 30 months of age, mothers of low PI infants in the nonintervention group were significantly less involved with their children than mothers in any other group. The low PI infants in the nonsupportive environment also showed poor intellectual development and fewer positive social behaviors than other groups. Thus it is posited that the malnourished infant may put off the caregiver in a nonsupportive environment in which the infant behaviors persist, thus exacerbating a cycle of nonsupportive infant and maternal interactive behaviors. However, in a supportive environment this cycle is not established. Instead, the low PI child develops normally as a function of the supportive day care environment, which effects the quality of the infant-maternal pattern of interaction. This transactional process is further elucidated by the A13 et al. study (1976) wherein the low PI infants were significantly less responsive to social stimuli and deficient in interaction processes at ten days. Additionally, eight of ten mothers reported problems of temperament, rhythmicity, and activity level at one and a half to nine months of age. The Goggin et al. study (1978) found a significant increase of maternal negative feedback in social interactions with their infants in the low PI group during the first year of life. Thus, distinctive cry patterns, social unresponsiveness and deficient interactional processes in infancy appear as part of the cycle that eventuates in decreased maternal involvement, increased maternal negative feedback and decreased positive social behaviors by the child at three years of age. Summary. Studies utilizing the PI as a means of classifying the fetally malnourished infant have tended to concur with and substantiate results arrived at using weight by gestational age growth curves. Deficient growth, and speech and behavior problems were again noted in the IUGR group. However, findings of abnormal neurological and cognitive development may relate more to the nonsupportive nature of the caretaking environment, particularly for healthy sga infants in lower SES groups, than to the status of being fetally malnourished; a finding not readily apparent in the birth weight by gestational age criterion studies. Additionally, parent- child interaction variables and social development may be transactionally related. Findings in this section of our review point to the importance of assessing the home environment and aspects of the parent-child relationship in order to understand sequelae for our target group of children. Use of the Family Relations scale on the Personality Inventory for Children-Revised (PIC-R), Dimensions of Temperament Survey-Revised (DOTS-R), and maternal interviews will yield information in these important areas. Timing of Onset of Fetal Growth Retardation The distinction has previously been made between the onset of IUGR and its associated maternal, placental, and fetal factors (Crane & Kopta, 1980). Diagnosis of onset of IUGR in the majority of studies reviewed here was by serial ultrasonic cephalometry. Slow intrauterine growth was defined as a weekly measurement of biparietal diameter (reflecting head and brain size) which is below the fifth centile for a period of two weeks or more. The time of 45 onset of growth failure was determined on the graph of the normal range by establishing the point at which the fetal growth line deviated from normal. When slow growth was of short duration, the onset was usually clearly defined. In cases of prolonged growth failure, slowing often occurred prior to the first ultrasonic measurement. This method was used in all but the Winer, Tejani, Atluru, DiGiuseppe and Borofsky study (1982). The method used by Winer et al. (1982) to determine the timing of onset of IUGR will be noted in the ensuing discussion. Physical Growth. Fancourt, Campbell, Harvey and Norman (1976) used a criterion of birth weight by gestational age (less than the tenth percentile) to select their four groups of term sga subjects. The groups were as follows: 1) growth failure less than or equal to 26 weeks; 2) greater than 26 but less than or equal to 34 weeks; 3) greater than 34 weeks; and 4) no evidence of growth failure. Excluded were children with known growth retarding factors associated with onset prior to the third trimester, thus there was an over preponderance of asymmetric growth retardation (onset in the third trimester) in this sample. They found that children whose head growth began to slow in utero prior to 34 weeks were more likely to have a height and weight less than the tenth centile at four years of age. This was significantly different than children whose growth slowed after 34 weeks or not at all, even when sex and social class were matched. These matched pairs differed significantly on head circumference as well. Winer et a1. (1982) examined two groups of four to seven year olds, one group whose mothers had hypertensive disorders during pregnancy and were sfd 46 (less than the tenth percentile) and another group of sfd children whose mothers had no hypertensive disorder. The former group corresponds to the asymmetric form of late onset growth retardation whereas the latter group corresponds to the early growth failure associated with the symmetric form of growth retardation. Head circumference and weight measures were greater in the hypertensive (late onset) group but this difference did not reach statistical significance, even when SES was corrected. Both groups showed values around the 50th percentile for head circumference, weight, and height. Neurological Growth. The four growth patterns examined in the Fancourt et al. study (1976) found significant differences when followed up at four years of age on the Griffiths scale which assesses six areas: locomotion, personal-social, hearing and speech, eye and hand coordination, performance, and practical reasoning. Those children whose growth failure began very early (prior to 26 weeks) had a significantly lower mean development quotient than the other three groups. Only the two subscales of performance and hearing and speech failed to reach statistical significance. Intellectual and Academic Growth. Harvey, Prince, Bunton and Campbell (1976) and Harvey, Prince, Bunton, Parkinson, and Campbell (1982) followed up term sfd children at five years of age with the McCarthy Scales of Children’s Abilities. The four groups of differential onset of growth retardation were the same as described in the Fancourt et al. study (1976). Prolonged slow growth in utero (less than or equal to 26 weeks) was significantly associated with lower abilities as assessed by the General Cognitive Index, and four 47 of the five subscales (perceptual-performance, quantitative, motor, and memory) compared to the other three combined groups. There were no significant differences on verbal scale scores. When the children were grouped by birth-weight percentiles (rather than by timing of IUGR onset), nonsignificant differences were noted on intellectual outcome, suggesting a more deleterious outcome for the subgroup of sfd children suffering prolonged intrauterine growth delay. Parkinson, Wallis and Harvey (1981), using the same four groups plus a matched control group for the two early onset groups, found that children with slowed fetal growth prior to 34 weeks gestation were sigdificantly more likely to have problems in achievement at seven years of age than were children whose growth retardation started later in pregnancy or was not apparent by ultrasonic measurement. Sex and social class were also significantly correlated to outcome. The two early onset groups also had significantly poorer concentration than other groups. Additionally, the very early onset group had significantly lower scores in a variety of subject areas, including reading, writing, and reasoning ability. Sex was again significantly correlated to outcome. These general findings of early growth failure and statistically significant lower scores are further corroborated in the Winer et al. study (1982). The later onset group (with maternal hypertensive disorder) performed better in all phases of intellectual testing than the early onset group, but this reached significance only on the Verbal IQ (on the WPPSI) after SES was corrected. In the late onset group (maternal hypertensive disorders) there was a negative correlation between both birth 48 weight and gestational age and test scores, indicating that, in mothers who had vascular disease, earlier delivery and the resulting lighter infants did better. It is suggested that this may reflect the late-flattening growth curve pattern, which produces maximum compromise of head growth in late gestation, resulting in a favorable outcome in newborns delivered earlier. In contrast, the early onset group appeared to benefit by prolonged gestation and greater weight at birth. Psycho/Social Growth. Psycho/social sequelae were investigated in the Parkinson et al. study (1981). The composition of the four groups and a control group have previously been described. No significant differences were noted in the total number of types of atypical behavior among the four sfd groups. However, boys with onset of growth failure prior to 34 weeks gestation had significantly more atypical markings than their normal birthweight controls. These boys were considered by their teachers to be significantly more clumsy, worried, fidgety, unhappy, and upset by new situations. Girls with IUGR onset prior to 34 weeks evidenced significantly more crying and irritability than other sfd girls. Summary. The reviewed studies suggest that early onset of growth failure is more frequently associated with poorer physical growth, and more immature neurological, intellectual, academic and psycho/social functioning than is late onset of IUGR. However, the Winer et al. study (1982) suggests that shorter gestation in the presence of maternal hypertensive disease (IUGR of late onset) is correlated to improved cognitive outcome. Sex and SES variables continue to be associated with prognosis. 49 These readings raise implications that have an impact on the current study. First, we might expect poorer outcomes where the pregnancy history is suggestive of early growth failure. Second, shorter gestations and lighter weights may be associated with better outcomes where maternal hypertensive disorders have complicated the pregnancy and birth. A Review and Synthesis Several areas having implications for a follow up study of children with HEV affected placentas have been reviewed. The chapter began with a brief survey of the placenta as a biopsy specimen for determining pathology, particularly regarding its usefulness in assessing the presence of viral infections. A discussion of chronic villitis and villitis of unknown etiology followed, noting than an infectious origin is documented in the former case and suspected in the latter. (Eleven percent of HEV affected placentas between 1978 and 1980 suggest the possibility of some type of viral infection, although this isn’t conclusive at this time.) Sixty percent of the placentas evidenced chronic villitis. Additionally, VUE and TORCH agents by implication are associated with an increased incidence and severity of intrauterine growth retardation. A brief review of sequelae of congenital CMV infection (an etiologic agent in chronic villitis and a small percentage of HEV affected placentas) indicated significantly higher incidences of neurological handicap in these children. Fetal growth retardation is associated with 33% of HEV affected placentas. A review of outcome studies indicates that sga children are a heterogenous group. It is possible that certain subgroups 50 have more deleterious outcomes than others. Boys, lower socio- economic classes, infants with neonatal complications, samples using more stringent criterion for defining sga (e.g. less than the third percentile), increased parity, or prolonged intrauterine growth retardation and/or associated factors (e.g. TORCH infections, congenital abnormalities, chromosomal anomalies, smoking, etc.) may be among those subgroups of sga children appearing to be at greater risk for later sequelae. Additionally, prolonged gestation with maternal hypertensive disease may be likely to result in poorer prognosis at follow up. Despite these many risk factors, there continues to be evidence suggesting the mediating influence of a more optimal environment on prognosis. Thus, a variety of factors will be important to make note of in the ensuing study. In particular, documentation of TORCH or other viral agents, chronic villitis, fetal growth retardation, maternal and fetal factors associated with IUGR, sex, and socio-economic variables. It will be critical to the purpose of this study to determine which of the preceeding factors may be associated with later sequelae in children with HEV affected placentas. Hypotheses The present study assessed children in three groups: those with HEV affected placentas, a Registry control group, and a community control group. Outcome measures include the McCarthy Scales of Children’s Abilities (MSCA), Personality Inventory for Children-Revised (PIC-R), Dimensions of Temperament Survey-Revised (DOTS-R), Bender Gestalt (BG), and presence or degree of handicapping conditions. 51 The following hypotheses were generated: 1. Those children with HEV affected placentas will do the poorest on outcome measures. The Registry control group will have lower scores than those children in the community control group but not as low as the children in the HEV group. 2. Male sex will be associated with poorer outcome. 3. Certain factors in the pre-, peri-, and/or postnatal environments will correlate to poorer prognosis. These factors include the presence of chronic maternal disease, multiple gestation, fetal distress, increased parity and gravida, cesarean section, maternal hypertensive disease, pre-eclampsia smoking in pregnancy, abnormal Apgar scores, male sex, low birth weight, earlier gestation, increased maternal age, greater ordinal position, minority group membership, and disrupted marital status (e.g. single, seperated, or divorced). 4. Socio-economic variables will be positively correlated to outcome. Parental education level and income at time of birth and time of assessment will be examined. 5. Presence of fetal growth retardation will be associated with lower scores. Additionally, early onset of delayed fetal growth will correlate to poorer outcome than will late onset of fetal 52 growth retardation. 6. Presence of congenital viral infection will be associated with poorer prognosis. 53 CHAPTER III Methodology This chapter outlines the methodology employed in the study. Discussion will focus on the design of the research, including a description of the pilot study, sample selection, contact procedures, instrumentation, and methods for analysis. All procedures described were conducted either by the author (LE) or the field epidemiologist (LK). The former was responsible for child assessments, while the latter was in charge of securing subject participation, doing maternal interviews, and abstracting data from medical records. This arrangement allowed the author and study participants to be blind to group status. The field epidemiologist was not blind to group status. Design The follow up of children with HEV affected placentas consists of examining the birth records for both mother and infant. Information will be abstracted in a systematic manner regarding health status and medical history. Additionally, those children with HEV affected placentas, born between November 1, 1977 and November 1, 1979, will be assessed at approximately five years of age. Comparison will be made with a Registry control group and a 54 community control group of children of similar ages. Mothers of all subjects will be interviewed as well. Description of the Pilot Study A pilot study assessing 28 children was initiated in I983; 14 of these children were in the HEV group and 14 were in the Registry control group. Comparison was made with Registry control children, matched on date of birth and location. Assessment consisted of the McCarthy Scales of Children’s Abilities (MSCA), the Personality Inventory for Children (PIC), and the Bender Visual Motor Gestalt test. Maternal interviews were conducted simultaneously with the child assessments. A paired t test on the MSCA General Cognitive Index (GCI) found significant group differences (p < .01). The GCI significantly correlated with gestational age, birth weight, and ordinal position (p < .05) for the HEV group, whereas gestational age (p < .05) and maternal education (p < .01) correlated with the Registry control group. Additionally, the HEV children exhibited more extreme PIC scores and greater immaturity on the Bender. Differences were not significant on either test. It was felt that further investigation was needed following the pilot study to see if the differences between groups were maintained and to further clarify the nature of these differences. Thus, the study to be described is a modification and extension of the original pilot study. Population Selection Procedures All 56 cases of HEV associated with live births, referred to the Registry from Michigan’s lower peninsula between November 1, 55 1977 and November 1, 1979 were identified. The objective was to secure 40 participants for this study. Each was matched on location and birthdate to Registry controls having a variety of pregnancy or birth complications, including placental abnormalities, but not HEV. Additionally, a community control group was identified, with the group size of 20 selected. The following selection methods were used. The log for the Registry was consulted to identify children for the HEV group, excluding stillbirths, neonatal deaths and infancy and childhood deaths. Registry controls were selected by choosing a liveborn infant from the same hospital as each HEV case, while matching as closely as possible on birthdate (with the calendar year plus or minus a month, or a 14 month span, being the parameters). If more than one Registry control was available (as in the instance where a hospital had two infants born on the same day), gestational age was also matched. Where Registry controls were unavailable from the hospital supplying the HEV case (e.g. where a hospital may have sent only one placenta to the Registry during the year and this one had HEV), then a hospital in the same town or a nearby town was used. In this manner, 40 children with HEV affected placentas were identified. Of the 56 potential cases, 16 were lost to follow up for the following reasons: child deceased-four, moved out of area- four, unable to locate-four, refused-three, child missing-one. Thirty seven Registry control children participated in the study. Of the 48 potential cases, 11 were lost to follow up for the following reasons: moved out of area-seven, unable to locate-one, 56 refused-three. The community controls were selected in the following manner. The numbers for each of the 40 children with HEV affected placentas were placed in a hat. Five numbers were randomly drawn. The physicians of these five HEV cases were contacted by letter and then phone, and requested to identify four children by means of their personal delivery log or hospital log; two infants delivered by the physician immediately before and two immediately after the HEV case. These were to be greater than 2500 grams at birth, more than 37 weeks gestation, requiring no special care at the hospital, possessing no congenital anomalies and whose mothers were free of hypertensive disease during the pregnancy. Since none of these children had their placentas examined at the Registry, it is not known if they had HEV. In this manner, five physicians were located and contacted. One of these was uncooperative. The remaining four doctors supplied names of suitable children; nine of which we were able to locate and secure their participation. The geographic parameters of the study were identified as being Michigan’s lower peninsula or immediately across the Michigan/Ohio border. All participants were born in Michigan hospitals and were assessed at approximately five years of age. Contact Procedures The following sequence of events describes the process for locating and contacting all study participants. Initally, letters were sent to the delivering physicians to inform them of the study. Telephone contact with physicians followed, with a request for 57 assistance in locating the subjects, including information about the child’s present physician. Letters and telephone contact was then made to these referred physicians. The procedure was slightly different for the community controls. The physicians were contacted and their assistance sought regarding entering their own hospital or personal delivery log to elicit names for the study subjects. Where hospital logs were consulted, permission to enter these logs was obtained from the hospital. Parents for all three groups were then contacted by letter and telephone to secure their participation. The study was explained in such a way so as to leave the subjects blind to their group status. Letters of informed consent were signed by the parent(s) in each case, with the stipulation that withdrawal from the study at any time was permitted. Testing and interviewing sites and times were established in the subjects home communities. Following the conclusion of the study, provision was made to inform participants as to their group status and general findings upon assessment. Instruments The following data collection methods were used: maternal interviews, retrospective examination of hospital records, McCarthy Scales of Children’s Abilities (MSCA - a measure of cognitive abilities), Bender Visual Motor Gestalt Test (a measure of visual- motor abilities), Personality Inventory for Children (used in the pilot study) and Personality Inventory for Children-Revised (PIC and PIC-R - a parent-report questionnaire assessing their child’s personality), and Dimensions of Temperament Survey-Revised (DOTS-R - a parent-report questionnaire assessing aspects of their child’s 58 temperament). Procedures Mothers were interviewed in a private room, all by the same person (LK) who was not blind to their group status. The interview schedule assessed general health, obstetrical history, toxic substance exposure, and health of the child and siblings. Demographic information was also obtained at this time. Mothers were sent the PIC or PIC-R and the DOTS-R prior to the interview, with the instructions to bring the completed forms with them. The PIC and PIC-R yield 12 clinical scales plus an overall adjustment scale: achievement, intellectual screening, development, somatic concern, depression, family relations, delinquency, withdrawal, anxiety, psychosis, hyperactivity and social skills. The DOTS-R includes scales of activity level-general, activity level-sleep, approach-withdrawal, flexibility-rigidity, mood, rhythmicity-sleep, rhythmicity-eating, rhythmicity-daily habits, and task orientation. Simultaneous to the maternal interviews were the child assessments on the MSCA and the Bender, administered by the author. The MSCA yields scales in five areas: verbal, perceptual- performance, quantitative, memory, and motor skills. A general cognitive index (GCI) is comprised of the first three scales. The Bender was administered to children five years of age or older, using the scoring system devised by Koppitz which yields a develOpmental score based on number of errors. Both tests are normed at six month intervals. All testing was done by the author using a double blind procedure. Children were assessed in health care or academic settings whenever possible using standardized procedures. 59 In some instances (approximately ten) the child was tested in their home in a quiet room as free from distractions as possible, when other arrangements could not be made. Children who became anxious during the testing situation were allowed to visit their mothers, following which testing was resumed. Birth records of mother and child were examined by two individuals, one who was blind to the group status of the participants (a medical student) and one who was not (LK). An abstraction form was devised to elicit and record the information in an objective manner. Methods of Analysis Statistical procedures were used to test the research hypotheses. The statistics employed included the ANOVA and chi square procedures, Pearson product and Spearman rho rank correlation coefficients, and the Bartlett test for homogeneity of variance. H1 and H2: The chi square and ANOVA statistics were used to determine group differences on the various outcome measures; including the MSCA, PIC-R, DOTS-R, Bender Gestalt, and handicapping conditions. The chi square statistic was used where both variables were discrete, whereas the ANOVA was used when one variable was continuous and one was discrete. A 2 X 3 ANOVA was used to determine group by sex differences on the various outcome measures. Where statistically significant confounding variables were noted to relate to both outcome measures and group status, a multi-factor ANOVA was used to control for the confounding variables. 60 H3: The Pearson product moment correlation coefficient was used to assess relationships between factors in the pre-, peri-, and postnatal environments and outcome measures: such as chronic maternal disease, multiple gestation, fetal distress, C section, maternal hypertension, pre- eclampsia, marital status, race, sex, smoking and drinking in pregnancy, birth weight, gestational age, maternal age, Apgar scores, chronic villitis, sga, and ordinal position. The Spearman rho rank correlation coefficient was used where group size was very small. In this study, this included analyses for the variables parity and gravida. H4: The Pearson product moment correlation coefficient was used to test for association of socio-economic variables to later sequelae, examining parental education and income levels at birth and at time of assessment. H5 The Pearson product moment correlation coefficient was used to assess the degree of relationship between sequelae and sga. Early and late onset of IUGR was analyzed using the Kruskal-Wallis One-Way ANOVA. This nonparametric test is apprOpriate where group size is small. H6: The association of congenital viral infection and outcome was assessed by calculating the Pearson product moment correlation coefficient. 61 Chapter Summary This chapter has attempted to outline the research methodology. It began with a description of the design, the population, instrumentation, and methods for statistical analysis. In the next chapter, we will view the results of these methods. 62 CHAPTER IV Results This chapter will review the results of the research project following up five year old children with HEV affected placentas. We will begin by describing this target population of children. An analysis of the statistical procedures and findings will follow. Comparison to a Registry control group and a community control group will be made. The analyses to be discussed herein result from the use of the Statistical Package for the Social Sciences. Description of the Population This section will delineate the characteristics of five year old children with HEV affected placentas. This data was compiled from maternal interviews and abstractions from birth records. As described in Chapter III, the maternal interviews were conducted in the home communities by LK, the field epidemiologist for the project, who was aware of the participants group status. Birth records were abstracted by LK and a medical student, the latter of which was blind to group status. Children with HEV affected placentas had a mean age of 63.2 months at time of assessment (range 55 to 74 months). They were born to mothers who were approximately 27 years old at time of this particular birth. Average years of education for mothers was slightly lower than for fathers. However, both parents tended to have graduated from high school (or its equivalent) and to have 63 received less than one year of school beyond the high school level. Average income level at birth of the child with HEV was in the $15,000 to $20,000 range. By the time of assessment, approximately five years later, the average family was earning in the $20,000 to $25,000 range. Average amount of smoking during pregnancy was less than half a pack a day, with 53% of HEV mothers not smoking at all during this pregnancy. Most of the families were married at the time of assessment. Only 13% were in the category of "not married" (single, divorced, seperated). See Tables 1 and 2 for a summary of these findings. 64 TABLE 1. Mean Values for Various Continuous Variables by Group Membership HEV Registry Community Variables mean n mean n mean n Age in Months 63.2 40 64.37 37 67.11 9 Maternal Age 27.15 40 25.32 37 25.0 7 Maternal Education/Yrs. 12.58 40 12.81 37 13.14 7 Paternal Education/Yrs. 12.97 39 12.87 37 13.71 7 Income at Birth* 4.48 40 4.46 37 4.29 7 Present Income* 5.2 40 4.76 37 5.14 7 Gestational Age/Weeks 38.67 39 36.83 36 40 5 Birth Weight/lbs. 5.99 40 6.3 37 7.52 7 Ordinal Position 2.45 40 1.95 37 3.43 7 Smoking in Pregnancy** .83 40 .78 37 .28 7 Gravida 3.71 38 2.84 37 3.29 7 Apgar: 1 Minute 7.75 20 7.13 16 8.0 3 Apgar: 5 Minutes 9.2 20 8.27 15 9 3 3 *180'55000 2=$5000-$10,000 3=$10,000-$15,000 4=$15,000-$20,000 5=$20,000-$25,000 6=$25,000-$30,000 7=$30,000-$35,000 8=$35,000+ **0-no smoking l=less than 1/2 pack a day 2= 1/2 to 1 pack a day 381 to 1 1/2 packs a day 4, 5, 6, 78 increase by 1/2 pack increments TABLE 2. Marital Status of Parents by Group Membership* HEV Registry Community Yes% No% Yes% No% Yes% No% Married 87 13 69 31 71 29 *married at time of assessment 65 Seventy percent of the children with HEV affected placentas were female. Ninety five percent of the HEV group were white (only two were nonwhite, one black and one hispanic; see Table 3). TABLE 3. Sex and Race by Group HEV Registry Community Sex: Male% 30 46 56 Female% 70 54 44 Race: White% 95 89 100 Other% 5 11 0 As a group, the children with HEV affected placentas tended to be neither premature nor of low birth weight. Average gestational age was 38.67 weeks (range 29 to 45 weeks) and mean birth weight was just under six pounds (range two lbs. one half oz. to eight lbs. fifteen 02.). These children tended to be second born children in their families, born to a gravida four. Apgars at one and five minutes were normal, with average ratings of 7.75 and 9.20 respectively. Of particular interest to this study was investigation of pre- and perinatal factors in this target group of children. (See Table 4 for summary information.) TABLE 4. Incidences of Pre- and Perinatal Factors by Group Membership* HEV Registry Community Variables No% Yes% No% Yes% No% Yes% Pre-eclampsia 80 20 92 8 100 0 Chronic Villitis*** 58 43 78 22 100 0 Small for Gestational Age 85 15 92 8 100 0 Cesarean Section 58 43 68 32 88 13 Viral Infection 100 O 95 5 100 0 Fetal Distress 78 23 78 22 88 13 Multiple Gestation 93 8 89 11 100 O Maternal Hypertension** 70 30 86 14 100 0 Maternal Diabetes 95 5 95 5 100 0 * (Percentages may not add up to 100% due to rounding error) **p < .10 ***p < .05 A 43% incidence of chronic villitis was noted in this group of children. As a group, these children had a high incidence of delivery by cesarean section (43%). Fifty percent of the HEV group had some form of hypertensive disease. Maternal hypertension was noted in 30% of the HEV cases and pre-eclampsia was observed in 20%. Both conditions appeared simultaneously in 15% of the group. Chronic maternal disease was noted in a small percentage (five percent) of children with HEV affected placentas. Of these (n=2), maternal diabetes was the sole category. Multiple gestation was noted in eight percent of the cases and fetal distress in 23%. Fifteen percent were sga. No incidences of documented viral infection were noted in the children with HEV affected placentas. Statistical Analysis and Findings The data resulting from the research project was analyzed using the Statistical Package for the Social Sciences. Contrast of the HEV group was made with two control groups; a Registry control group and a community control group, which have elsewhere been described. The relationship between factors in the pre-, peri-, and postnatal environments and group outcome were the focus of this research. The first step in the analysis of the data was to obtain descriptive statistics. These have been reported for the HEV group in the previous section. This was followed by obtaining correlation coefficients on each variable. A list of these variables can be seen in Tables 5 and 6. 68 TABLE 5. A List of Continuous Variables of Importance to the Study childs age in months maternal age in years maternal and paternal education in years income at birth and at present by category gestational age in weeks birth weight ordinal position smoking in pregnancy Apgars at one and five minutes gravida parity TABLE 6. A List of Discrete Variables of Importance to the Study sex race (white and other) pre-eclampsia (yes or no) chronic villitis (yes or no) small for gestational age (yes or no) cesarean section (yes or no) viral infection (yes or no) fetal distress (yes or no) multiple gestation (yes or no) maternal hypertension (yes or no) maternal diabetes (yes or no) marital status (married or not married at time of assessment) 69 All outcome variables were continuous in nature. These included six scales on the MSCA, 13 scales on the PIC-R, nine scales on the DOTS-R, one score on the Bender Gestalt, and one score denoting degree of handicapping conditions (Oanone, 1=mild, 2=moderate, 38severe; to be described more fully in the ensuing discussion). See Table 7 for a complete list of outcome measures and subscales. 7O TABLE 7. Outcome Measures McCarthy Scales of Children’s Abilities Verbal Perceptual-Performance Quantitative General Cognitive Index (GCI) Memory Motor Personality Inventory for Children-Revised Adjustment Achievement Intellectual Screening Development Somatic Concerns Depression Family Relations Delinquency Withdrawal Anxiety Psychosis Hyperactivity Social Skills Dimensions of Temperament Survey-Revised Activity Level-General Activity Level-Sleep Approach-Withdrawal Flexibility-Rigidity Mood (Positivie) Rhythmicity-Sleep Rhythmicity-Eating Rhythmicity-Daily Habits Task Orientation Bender Gestalt Developmental Score (Koppitz) Degree of Handicapping Conditions Degree of severity, coded 0 to 3 71 The chi square statistic and ANOVA were used to examine group differences on the variables. The chi square statistic was used on the discrete variables (see Table 6). The .05 level of significance was established. When assumptions for the chi square statistic were violated (i.e. when minimum expected frequencies were less than five in each cell), the two control groups were pooled to form one group (a Registry/Community control group) which then allowed the chi square statistic to be computed. This had a minimal effect on the results. Chronic villitis was significant at the .05 level of significance (p-.0267 and p=.0281 in the two control group and the pooled control group analysis respectively). Maternal hypertension neared significance (p=.0648 using two control groups and p=.0572 using the pooled control group). ANOVA was used to examine group status with the continuous variables (see Table 5). The Bartlett test for homogeneity of variances was used. The .05 level of significance was set for these analyses. Variances exceeded the level of significance on the Bartlett test for three variables; gestational age, ordinal position, and five minute Apgar scores. Taking the log adequately transformed the variances so they no longer differed significantly for ordinal position and five minute Apgar scores. A variety of transformations failed to equalize the variances on the variable gestational age. This variable had a negative skew that was nearer normal than was true of the untransformed variables of ordinal position and five minute Apgar scores. It appears that the two variables which had a greater degree of skewness and thus differed more from normal were more readily transformed than was gestational age. With the completion of these computations, the results were examined for possible confounding variables which could be statistically controlled in the next phase of the analysis. If significant confounding variables of a continuous nature were found, the ANCOVA procedure would be used, which is appropriate when all variables are continuous. If the significant confounds were discrete in nature, then a multi-factor ANOVA would be used since it is appropriate for the analysis of continuous and discrete variables. Two variables which were significantly related to (or nearly so) group status and outcome measures at the .05 level of significance were noted (i.e. chronic villitis, p=.0267 and maternal hypertension, p=.0572; both using the pooled control group). Both were discrete variables so a three factor ANOVA design was employed to analyze the results. The next section will attempt to relate the statistical findings to the research in a systematic way. Discussion of each hypothesis in turn and the pertinent data will ensue. The first two hypotheses will be examined together due to the nature of their statistical analysis which employed a 2 X 3 ANOVA on the variables group and sex. H1: Those children with HEV affected placentas will do the poorest on outcome measures. The Registry control group will have lower scores than those children in the community control group but not as low as the children in the HEV group. 73 H2: Male sex will be associated with poorer outcome. In order to address these hypotheses, it will be most orderly to examine each outcome measure in turn. Focus will begin with the MSCA scales, followed by the PIC-R, DOTS-R, Bender Gestalt, and degree of handicapping conditions. An examination of Table 8 reveals that on all six scales, the HEV group preformed the poorest of all the groups. The Registry control group did better than the HEV group and the community control group outperformed both groups on all six measures. 74 TABLE 8. Verbal Perc.** Quant. Memory** Motor** Group n *mean for GCIleO, all others **p < .05 Mean MSCA Scores by Group Membership and Sex* Total Male Female Total Male Female Total Male Female Total Male Female Total Male Female Total Male Female Total Male Female HEV 44.26 44.0 44.37 44.03 42.09 44.81 44.87 45.55 44.59 88.79 87.63 89.26 43.55 41.55 44.37 44.21 44.82 43.96 38 11 27 Registry 47.68 47.94 47.45 48.53 50.12 47.00 46.05 45.53 46.5 94.73 95.71 93.90 47.95 48.24 47.70 47.62 45.18 49.70 37 17 20 mean=50 75 Community 52.0 53.0 50.75 52.44 55.6 48.5 49.78 51.2 48.0 102.44 105.4 98.75 52.67 54.0 51.0 53.89 57.6 49.25 9 5 4 Differences were statistically significant on the perceptual- performance, memory, and motor scales at the .05 level of significance on the univariate F test for group effects (see Table 9). All other univariate F tests were nonsignificant. No statistically significant differences by sex were noted. TABLE 9. Univariate F Tests For Group Effects On The MSCA ss df as F Sign Verbal 512.76 2 256.38 1.48 .233 Perc-Perf 677.28 2 338.64 3.14 " .049* Quant. 176.63 2 88.31 1.11 .332 GCI 1585.72 2 792.86 2.24 .113 Memory 749.61 2 374.80 3.23 .045* Motor 731.84 2 365.92 3.41 " .038* *p < .05 All multivariate effects for the MSCA are nonsignificant as seen in Table 10. TABLE 10. Multivariate Effects for the MSCA Wilks lambda Value Approx F Hypoth df Error df Sign. Constant .025 477.29 6.0 73.0 0.0 Group .823 1.242 12.0 146.0 .26 Sex .969 .389 6.0 73.0 .884 Group X Sex .787 1.55 12.0 146.0 .113 In the three factor ANOVA, all main effects and interaction effects are nonsignificant. However, some of the scales of the MSCA approach significance in the three-way interaction (group status by chronic villitis by maternal hypertension). The verbal scale notes a p value of .06 and the quantitative scale notes a p value of .073. These nonsignificant findings indicate that the two confounding variables interacting with group membership account for the significant differences noted in the one factor analysis (see Table 11). TABLE 11. Univariate F Tests for the Three-Way Analysis on the MSCA Variable SS df MS F Sign. Verbal 608.24 1 608.24 3.65 .06 Perc.-Perf. 102.66 1 102.66 .94 .336 Quant. 253.68 1 253.68 3.30 .073 GCI 1031.81 1 1031.81 2.94 .091 Memory 332.96 1 332.96 2.88 .094 Motor 12.79 1 12.79 .11 .738 All multivariate F tests for the three-way analysis on the MSCA are nonsignificant, as can be seen in Table 12. Means and group size for the three-way analysis can be seen in Tables 13-14. 77 TABLE 12. MSCA Source Grp CV HTN Grp X CV Grp X HTN CV X HTN Grp x cv x HTN TABLE 13. Verb HTN: Perc Quant GCI Memory Motor TABLE 14. HTN: N Y Multivariate Effects for the Three-Way Analysis On The Value .862 .949 .905 .934 .921 .910 .917 .87 .59 Approx F 5 9 1.186 .801 .97 1.11 1.02 0 Hypoth df 12. O\O\O\O\O\O\ OOOOOOO Mean MSCA Scores for the Three-Way Analysis HEV CV: N Y KZNZNZNZNZMZ 44.5 46.3 42.6 42.8 44.3 42.9 87.3 88.6 44.3 45.1 43.3 42.9 42.4 45.3 45.6 48.3 47.7 41.7 90.1 91.3 42.3 41.0 45.9 45.7 Registry N Y 47.6 43.8 41.0 70.0. 48.8 49.0 39.7 55.5 46.2 46.2 38.7 54.5 95.4 92.0 79.7 117.5 47.8 45.5 44.0 63.0 47.2 48.8 45.7 52.0 Error df Sign 136.0 .574 68.0 .730 68.0 .324 68.0 .572 68.0 .452 68.0 .365 68.0 .418 Community N Y 51.9 -- 50.7 -- 50.43 -- 101.71 -- 51.3 -- 50.7 ~- Group Size For The Three-Way Analysis On The MSCA HEV CV: N 15 8 Y 12 3 Registry 2 N 6 3 78 Community N Y 7 O 0 O An examination of the PIC-R results reveals a pattern of more extreme scores for the HEV children on several scales, including overall adjustment, achievement, intellectual screening, development, withdrawal, social skills, and psychosis. Only the latter neared statistical significance (p=.0567) using the one factor ANOVA (see Table 15). Elevations greater than two standard deviations were noted on the adjustment and delinquency scales for the HEV group. Both males and females contributed to these elevated scores. Interestingly, the community controls also peaked at over two standard deviations on these same two scales. An analysis of Table 16 reveals that the female community controls contributed to these elevelations more so than did the males. Female community controls also peaked at over two standard deviations above the mean on the anxiety and hyperactivity scales while the males did not. The Registry control group was within the normal range on all but the adjustment scale, where the males were observed to have scores greater than two standard deviations above the mean. TABLE 15. Univariate F Test For Group Effects on the PIC-R Psychosis Scale Source SS df MS F Sign. Between Groups 1588.19 2 794.09 3.0 .0567 Within Groups 16923.59 64 264.43 Total 18511.79 66 79 TABLE 16. Variable Adj Ach IS Dvl Som Dep Fam qu Wdl Anx Psy Hyp SSK Total Male Female Total Male Female Total Male Female Total Male Female Total Male Female Total Male Female Total Male Female Total Male Female Total Male Female Total Male Female Total Male Female Total Male Female Total Male Female Grp n Total *Scores are T scores, with a normal range of 30T to 70T Male Female HEV 74.0** 80.1** 71.0** 61.6 61.0 61.9 65.9 66.1 65.8 59.0 59.1 59.0 59.7 66.1 56.6 59.7 61.7 58.7 59.0 66.2 55.5 71.2** 74.8** 69.4 61.5 63.6 60.5 62.4 66.0 60.7 64.7 65. 64.3 59.9 63.6 58.1 57.0 56.0 57.5 30 10 20 0‘ Registry 68.7 73.1** 64.5 56.1 54.9 57.3 64.7 64.4 64.9 56.4 54.9 57.8 60.4 60.4 60.3 59.6 56.2 62.7 57.6 59.9 55.3 63.9 63.5 64.2 57.3 56.1 58.4 59.2 58.0 60.4 58.3 55.1 61.2 59.5 64.4 54.9 56.1 52.7" 59.2 29 14 15 Mean PIC-R Scale Scores by Group Membership and Sex* Community 71.3** 65.0 81.7** 54.5 55.4 53.0 59.4 57.4 62.7 52.1 53.8 49.3 56.0 54.8 58.0 59.8 56.0 66.0 60.6 63.2 _56.3 73.1** 65.6 85.7** 57.1 55.2 60.3 65.0 58.2 76.3** 49.8 54.0 42.7 67.9 64.6 73.3** 55.0 53.4 57.7 8 5 3 **greater than two standard deviations above the mean 80 All of the multivariate analyses were nonsignificant (see Table 17). TABLE 17. Multivariate Effects For The PIC-R Wilks lambda Value Approx F Hypoth df Error df Sign. Constant .004 701.19 16.0 46.0 0.0- Grp .567 .942 32.0 92.0 .562 Sex .738 1.02 16.0 46.0 .453 Grp X Sex .643 .708 32.0 92.0 .865 The three factor ANOVA revealed group membership and maternal hypertension to be relatively less important in accounting for significant differences on outcome measures than was chronic villitis. An examination of Tables 18-20 indicates that presence of chronic villitis, regardless of group membership and across both levels of maternal hypertension, significantly correlated to higher scores on the adjustment and delinquency scales of the PIC-R at the .05 level of statistical significance. Higher scores on the PIC-R indicate heightened concern and greater problems in these areas according to parental report. Nearing significance, and correlating in the same direction were the scales of somatic concern and anxiety (p=.062 and p=.063 respectively). 81 TABLE 18. Univariate F Tests For Chronic Villitis On The PIC-R Variable SS df MS F Sign. Adj 1885.32 1 1885.32 6.11 .017* Ach 430.16 1 430.16 2.03 .159 IS 96.12 1 96.12 .43 .516 Dvl 61.71 1 61.71 .29 .588 Som 899.48 1 899.48 3.64 .062 Dep 418.86 1 418.86 1.73 .193 Fan 164.66 1 164.66 .56 .457 qu 1433.62 1 1433.62 7.36 .009* Wdl 271.78 1 271.78 1.42 .238 Anx 865.46 1 865.46 3.58 .063 Psy 446.71 1 446.71 1.58 .213 Hpr 492.84 1 492.84 2.78 .101 SSK 46.99 1 46.99 .35 .556 *p < .05 TABLE 19. Univariate F Tests For Group Effects On The PIC-R Variable SS df MS F Sign. Adj 847.78 2 423.89 1.37 .262 Ach 265.61 2 132.80 .63 .537 IS 106.66 2 53.33 .24 .789 Dvl 314.85 2 157.43 .76 .474 Sam 113.54 2 56.77 .23 .795 Dep 95.91 2 47.95 .19 .821 Fam 205.74 2 102.85 .35 .706 qu 1646.19 2 823.10 4.23 .019* Wdl 149.28 2 74.64 .39 .679 Anx 351.67 2 175.84 .73 .487 Psy 1524.57 2 762.29 2.70 .076 Hpr 1416.91 2 708.46 4.00 .024* SSK 6.35 2 3.17 .02 .977 *p < .05 82 TABLE 20. Variable Adj Ach IS Dvl Som Dep Fam qu Wdl Anx Psy Hpr SSK *p < .05 $8 124.70 780.91 1.48 1305.03 2.48 17.51 54.24 417.59 107.81 203.38 21.26 168.62 191.00 d HHHHHHHHHHt—HHM MS 124.70 780.91 1.48 1305.03 2.48 17.51 54.24 417.59 107.81 203.38 21.26 168.62 191.00 83 Univariate F Tests For Maternal Hypertension On The PIC-R F Sign. .40 .528 3.70 .060 .006 .936 6.28 .015* .01 .921 .07 .789 .18 .669 2.15 .149 .56 .456 .84 .363 .08 .785 .95 .333 1.42 .238 Group membership was significantly associated with the delinquency and hyperactivity scales of the PIC-R (however, it must be remembered that since this is an unbalanced design, group effects, which are the second variable in the design, are not independent of chronic villitis effects, which are the first variable in the design). Children in the HEV group tended to have more parental concerns in the delinquency dimension whereas the Registry control group had greater problems with hyperactivity. Approaching significance (p=.076) was the psychosis scale, with HEV children possessing the higher scores. Interestingly, absence of maternal hypertension during pregnancy was significantly associated with poorer outcome (or higher scores) on the deveIOpment scale of the PIC-R and approached significance (p-.06) on the achievement scale. The average scores and group size for this three-way analysis can be viewed in Tables 21-22. 84 TABLE 21. CV: Adj HTN: N Y Ach N Y IS N Y Dvl N Y Som N Y Dep N Y Fam N Y qu N Y Wdl N Y Anx N Y Psy N Y Hpr N Y SSK N Y HEV N 64.6 71.8* 63.6 54.0 65.8 68.0 63.9 50.4 53.3 56.4 55.9 54.2 58.6 61.8 62.9 75.0* 57.2 55.6 54.0 62.2 66.0 57.6 53.4 63.6 54.9 55.0 Y 80. 83. 63. 58. 68. 53. 60. 50. 64. 66. 65. 59. 60. 52. 76. 73. 68. 56. 68. 67. 66. 64. 62. 65. 59. 56. om » a Registry N Y 65.4 77.8* 71.0* 73.5* 55.7 66.2 47.3 48.5 62.0 70.4* 71.0* 66.5 57.1 63.2 48.7 45.0 58.4 70.0* 63.3 51.0 57.4 65.0 68.3 53.5 56.6 54.8 74.3* 48.0 58.7 75.8* 70.7* 72.5* 57.7 54.4 65.7 50.5 55.5 67.6 75.0* 50.5 56.9 59.0 66.3 57.5 56.8 68.2 55.0 69.0 57.5 57.0 49.7 50.0 Mean PIC-R Scores For The Three-Way Analysis Community N 78.83* 56.0 60.2 50.7 56.8 61.0 79.3* 58.2 65.7 46.0 73.0* 56.7 * greater than two standard deviations above the mean TABLE 22. CV: HTN: N Y HEV N 10 5 Registry N Y 19 5 3 2 85 Community N 6 0 Y Group Size For The Three-Way Analysis On The PIC-R All of the multivariate tests of significance failed to reach statistical significance (see Table 23). TABLE 23. Multivariate F Tests For The Three-Way Analysis On The PIC-R Source Value Approx. F Hypoth. df Error df Sign. CV .662 1.31 16.0 41.0 .239 Grp .569 .835 32.0 82.0 .711 HTN .647 1.40 16.0 41.0 .189 Grp X CV .785 .701 16.0 41.0 .776 Grp X HTN .674 1.24 16.0 41.0 .281 CV X HTN .678 1.22 16.0 41.0 .298 Grp X CV X HTN .805 .62 16.0 41.0 .848 No clear pattern of group or sex differences were apparent on the DOTS-R. The HEV group scored between the two control groups in terms of general activity level, with the Registry control group being the most active. The HEV group exhibited the most approach behavior of the three groups, the least regularity in sleeping and less persistence in task orientation. Quality of mood, flexibility/rigidity, and rhythmicity daily habits were comparable for all groups (see Table 24). None of these differences were statistically significant. Table 25 reveals the one significant correlation (rs-.31, p < .05) of temperament with sex on the rhythmicity-sleep dimension, which favored males. 86 TABLE 24. Mean DOTS-R Dimensions By Group Membership And Sex HEV Registry Community Act-Gen Total 19.0 20.4 16.1 Male 21.1 20.6 15.6 Female 17.7 19.9 17.5 Act-Sl Total 9.0 10.1 8.4 Male 9.1 9.8 9.4 Female 8.9 10.6 6.0 App/Wdl Total 21.2 20.6 18.9 Male 21.4 19.9 20.2 Female 21.0 21.9 15.5 Flex/Rig Total 16.3 16.6 16.4 Male 15.3 16.5 16.8 Female 16.9 16.7 15.5 Mood Total 25.9 25.1 25.29 Male 25.6 25.8 26.8 Female 26.1 23.7 21.5 Rhy-Sl Total 16.9 18.3 18.6 Male 18.4 18.5 20.0 Female 15.9 17.9 15.0 Rhy-Eat Total 15.3 14.8 17.0 Male 15.9 14.5 18.8 Female 14.9 15.4 12.5 Rhy-Hab Total 13.5 13.9 14.1 Male 13.9 14.2 14.2 Female 13.2 13.3 14.0 Task Total 45.6 46.9 49.7 Male 48.2 47.1 53.0 Female 43.9 46.6 41.5 Grp n Total 23 20 7 Male 9 13 5 Female 14 7 2 TABLE 25. Correlation Coefficients of Sex To The DOTS-R Dimensions Act-Gen Act-Sl App Flex Mood Rhy-Sl Rhy-Eat Rhy-Hab Task -.15 -.06 .03 .12 -.14 -.31* -.13 -.13 -.23 * p < .05 87 Analysis using the three factor ANOVA pointed to slightly different results. For mean scores and group size for the three-way analysis, see Tables 26-27. The main effects of maternal hypertension were statistically significant for one of the dimensions; approach-withdrawal. This indicates that regardless of group membership, or presence or absence of chronic villitis, those children with maternal hypertension present during gestation exhibit greater approach behavior than those without maternal hypertension. Of the nine DOTS-R dimensions, five of these reached statistical significance in the group by maternal hypertension interaction at the .05 level of significance. This indicates that on a certain level of maternal hypertension, there was a difference by group membership on certain outcome variables. Specifically, the HEV group with maternal hypertension present during pregnancy had lower scores (reflecting less of the dimension according to parental report) on the following scales than those without maternal hypertension: (positive) mood, rhythmicity-sleep, rhythmicity-daily habits, and task-orientation. The Registry control group had higher scores on these same scales in the presence of maternal hypertension compared to those without maternal hypertension. Higher scores on the activity level-general dimension were received by the HEV children with maternal hypertension as compared to those without maternal hypertension. The Registry control group scored in the opposite direction on this dimension (see Tables 28-29). 88 TABLE 26. CV: Act-G HTN: N Act-Sl APP Flex Mood Rhy-Sl Rhy-Eat Rhy-Hab Task mzmz»m; Ashby, S. Is intensive care justified for infants weighing less than 801 gm at birth? The Journal of Pediataraics, 1981, 99, 937-943. Brood, C. Consequences of intrauterine growth retardation. British Medical Journal, 1983, 286, 164-165. Caputo, 0.; Mandell, W. Consequences of low birth weight. Developmental Psychology, 1970, 3, 363-383. Chamberlain, G. Epidemiology and aetiology of the preterm baby. Clinics in Obstetrics and Gynaecology, 1984, 11, 297-314. Clarren, S.K.; Smith, D.W. The fetal alcohol syndrome. The New England Journal of Medicine, 1978, 298, 1063-1067. 146 Cockburn, J.; Ounsted, M.; Moar, V.A.; Redman, C. Final report on study on hypertension during pregnancy: the effects of specific treatment on the growth and development of the children. The Lancet, 1982, 647-649. -—_' Cohen, S.E.; Beckwith, L. Preterm infant interaction with the caregiver in the first year of life and competence at age two. Child Development, 1979, 59, 767-776, Colle, E.; Schiff, 0.; Andrew, C.; Bauer, C.; Fitzhardinge, P.M. Insulin response during catch-up growth of infants who were small for gestational age. Pediatrics, 1976, 61, 363-371. Collins, E.; Turner, G. The importance of the small for dates baby to the problem of mental retardation. The Medical Journal of Australia, 1971, 313-315. Commey, J.O.; Fitzhardinge, P.M. Handicap in the preterm small for gestational age infant. The Journal of Pediatrics, 1979, 94, 779- 786. —- _— Crane, J.P.; Kopta, M.M. Comparative newborn anthropometric data in symmetric versus asymmetric intrauterine growth retardation. American Journal of Obstetrics and Gynecology, 1980, 138, 518-522. Crawford, C.S. Maternal conditions that affect the fetus/neonate. In R.J. Bolognesse; R.H. Scwarz; J. Schneider (Eds.) Perinatal Medicine. Baltimore: Williams & Wilkins, 1982. Crawford, C.S. The growth retarded newborn. In R.J. Bolognesse; R.H. Scwarz; J. Schneider (Eds.) Perinatal Medicine. Baltimore: Williams & Wilksins, 1982. Dann, M.; Levine, 8.2.; New, E.V. A long term follow up study of small premature infants. Pediatrics, 1964, 233 945-955. Davies, D.P.; Platts, P.; Pritchard, J.M.; Wilkinson, P.W. Nutritional status of light for date infants at birth and its influence on early postnatal growth. Archives of Diseases in. Children, 1979, 54, 703-706. Dawkins, M. The small for dates baby. In M. Dawkins; W.G. MacGregor (Eds.) Gestational Age, Size, and Maturity. London: Spastics Society Medical Education and Information Unit, 1965. DeWolf, F.; Brosens, I.; Renaer, M. Fetal growth retardation and the maternal arterial supply of the human placenta in the absence of sustained hypertension. British Journal of Obstetrics and Gynecology, 1980, 81,678-685. ... Douglas, C.P. Low birth weight babies; preterm and small: the obstetric challenge. British Journal of Hospital Medical, 1982, 28, 455-461. '-— "— Drillien, C.M. The small for date infant: etiology and prognosis. Pediatric Clinics of North America, 1970, 11, 9-24. Drillien, C.M. Abnormal neurologic signs in the first year of life in low birthweight infantss: possible prognostic significance. Developmental Medicine and Child Neurology, 1972, 14, 575-584. Drillien, C.M. Aetiology and outcome in low birthweight infants. Developmental Medicine and Child Neurology, 1972, .14, 563-574. Drillien, C.M.; Thomson, A.J.M.; Burgoyne, K. Low birthweight children at early school age: a longitudinal study. Developmental Medicine and Child Neurology, 1980, 223 26-47. Drorbaugh, J.E.; Moore, D.M.; Warram, J.H. Association between gestational and environmental events and central nervous system function in 7 year old children. Pediatrics, 1975, 26, 529-537. Ellenberg, J.H.; Nelson, K.B. Birth weight and gestational age in children with cerebral palsy or seizure disorders. American Journal of Diseases of Children, 1979, 133, 1044-1048.. Everett, M.J. Cytomegalovirus. Unpublished manuscript, Michigan State University, 1982. Fancourt, R.; Campbell, 3.; Harvey, 0.; Norman, A.P. Follow-up study of small-for dates babies. British Medical Journal, 1976, 1435-1437. Fisher, K. Hypertension in pregnancy. Primary Care, 1983, 10, 115- 124. "" Fitzhardinge, P.M.; Kalman, E.; Ashby, 8.; Pape, K.E. Present status of the infant of very low birth weight treated in a referral neonatal intensive care unit in 1974. In Major Mental Handicap: Methods and Costs of Prevention. Ciba Foundation Symposium 59, New York: 1978. Fitzhardinge, P.M.; Ramsay, M. The improving outlook for the small prematurely born infant. Developmental Medicine and Child Neurology, 1973, 15, 447-459. Fitzhardinge, P.M.; Steven, E.M. The small for date infant. I. Later growth patterns. Pediatrics, 19722.32’ 671-681. Fitzhardinge, P.M.; Steven, E.M. The small for date infant. II. Neurological and intellectual sequelae. Pediatrics, 1972, 50, 50- 57. ‘_" 148 Francis-Williams, J.; Davies, P.A. Very low birthweight and later intelligence. Developmental Medicine and Child Neurology, 1974,‘16, 709-728. Frenkel, L.D.; Keys, M.P.; Hefferen, S.J.; Rola-Pleszczysnki, M.; Bellanti, J.A. Unusual eye abnormalities associated with congenital cytomegalovirus infection. Pediatrics, 1980, 66, 763-765. Galbraith, R.S.; Karchmar, E.J.; Piercy, W.N.; Low, J.A. The clinical prediction of intrauterine growth retardation. American Journal of Obstetrics and Gynecology, 1979, 281-286. Goggin, J.B.; Holmes, G.E.; Hassanein, R.; Lansky, S.B. Observations of postnatal developmental aactivity in infants with fetal malnutrition. The Journal of Genetic Psychology, 1978, 132, 247-253. Goldenberg, R.L.; Nelson, R.S.; Davis, R.O.; Koski, J. Delay in delivery: influence of gestational age and the duration of delay in perinatal outcome. Obstetrics and Gynecology, 1984, 64, 480-484. Griffiths, P.D.; Campbell-Benzie, A.; Heath, R.B. A prospective study of primary cytomegalovirus infection in pregnant women. British Journal_g£ Obstetrics and Gynecology, 1980, 81, 308-314. Gruenwald, P. Fetal malnutrition. In H.A. Waisman & G.R. Kerr (Eds.) Fetal Growth and Development. New York: McGraw-Hill Book Company, 1970. Hack, M.; Fanaroff, A.A. The outcome of growth failure associated with preterm birth. Clinical Obstetrics and Gynecology, 1984,_21, 647-663. Hack, M.; Fanaroff, A.A.; Merkatz, I.R. The low-birth-weight infant: evolution of a changing outlook. The New England Journal of Medicine, 1979, 301, 1162-1165. Hanshaw, J.B. Congenital cytomegalovirus infection. The New England Journal of dedicine, 1973, 1405-1406. Hanshaw, J.B.; Dudgeon, J.A. Viral Diseases of the Fetus and Newborn. Philadelphia: W.B. Saunders Company, 1978. Hanshaw, J.B.; Scheiner, A.P.; Moxley, A.W.; Gaev, L.; Abel, V.; Scheiner, B. School failure and deafness after "silent" congenital cytomegalovirus infection. The New England Journal of Medicine, 1976, 468-470. '— Harvey, D.R.; Prince, J.; Bunton, W.J.; Campbell, 8. Abilities of children who were small for dates at birth and whose heaad growth in utero was measured by ultrasonic cephalometry. Pediatric Research, 1976, 10, 891. 149 Harvey, 0.; Prince, J.; Bunton, J.; Parkinson, C.; Campbell, 8. Abilities of children who were small-for-gestational-age babies. Pediatrics, 1982, 69, 296-300. Hepner, R.; Bowen, M. The placenta and the fetus. The Journal of the American Medical Association, 1960, 172, 427-432. Jones, R.A.K.; Roberton, N.R.C. Problems for the small-for-dates baby. Clinics in Obstetrics and Gynaecology, 1984, 11, 499-523. Kaufman, A.S.; Kaufman, N.L. Clinical Evaluation of Young Children with the McCarthy Scales. New York: Grune & Stratton, Inc., 1977. Keirse, M. Epidemiology and aetiology of the growth retarded infant. Clinics in Obstetrics and Gynaecology, 1984, 11, 415-436. Kilbride, H.W.; Johnson, D.L.; Streissgath, A.P. Social class, birth order and newborn experience. Child Development, 1977, 48, 1686-1688. ' 'fi- Kimble, K.J.; Ariagno, R.L.; Stevenson,, D.K.; Sunshine, P. Growth to age 3 years among very low-birth-weight sequelae-free survivors of modern neonatal intensive care. The Journal of Pediatrics, 1982, 1%, 622-624. ‘— Kinnane, L.; Sander, C.H.; Ellicott, L.A.; Tauscher, J.W.; Wright, L.M. Developmental abnormalities in live births with hemorrhagic endovasculitis of the placenta: follow-up at five years. Laboratory Investigations, 1985, 52, 35A. Kitchen, W.; Ford, G.; Orgill, A.; Rickards, A.; Astbury, J.; Lissenden, J.; Bajuk, B.; Yu, V.; Drew, J.; Campbell, N. Outcome in infants with birth weight 500-999 gm: 3 regional study of 1979 and 1980 births. The Journal of Pediatrics, 1984, 194, 921-927. Kitchen, W.; Ryan, M.M.; Rickards, A.; McDougall, A.B.; Billson, P.A.; Keir, E.H.; Naylor, F.D. A longitudinal study of very low- birthweight infants. IV: an overview of performance at eight years of age. Developmental Medicine and Child Neurolog_, 1980,_22, 172- 188. Knobloch, H.; Pasamanick, B. Prospective studies on the epidemiology of reproductive casualty: methods, findings, and some implications. Merrill-Palmer Quarterly, 1966, .12, 27-43. Knox, W.F.; Fox, H. Villitis of unknown aetiology: its incidence and significance in lacentae from a British population. Placaenta, 1984, 5, 395-402. """"’ Kopp, C.B.; Parmelee, A.H. Prenatal and perinatal influences on infant behavior. in J.D. Osofsky (Ed.) Handbook of Infant Deve10pment. New York: John Wiley & Sons, 1979. 150 Kumar, N.L.; Gold, E.; Jacobs, T.B.; Ernhart, C.B.; Nankervis, G.A. Primary cytomegalovirus infection in adolescent pregnancy. Pediatrics, 1984, 14, 493-500. Kumar, S.P.;; Anday, E.K.; Sacks, L.M.; Ting, R.Y.; Delivoria- Papadopoulos, M. Follow-up studies of very low birth weight infants (1,250 grams or less) born and treated within a perinatal center. Pediatrics, 1980,_§6, 438-444. Labarrere, C.; Althabe, 0.; Talenta, M. Chronic villitis of unknown aetiology in placentae of idiopathic small for gestational age infants. Placenta, 1982, 3, 309-318. Lerner, R.M.; Palermo, M.; Spiro, A.; Nesselroade, J.R. Assessing the dimensions of temperamental individuality across the life span: The Dimensions of Temperament Survey (DOTS). Child Development, 1982, 23, 149-159. Levene/Lilly, M.I.; Dubowitz, M.S. Low birth weight babies: long- term follow up. British Journal of Hospital Medicine, 1982, 28. 487-493. ‘-— '_— Low, J.A.; Galbraith, R.S.; Muir, D.; Killen, H.; Karchmar, J.; Campbell, D. Intrauterine growth retardation: ,a preliminary report of long-term morbidity. American Journal of Obstetrics and Gynecology, 1978, 130, 534-545. Low, J.A.; Galbraith, R.S.; Muir, 0.; Killen, H.; Pater, 8.; Karachmar, J. Intrauterine growth retardation: a study of long- term morbidity. American Journal of Obstetrics and Gynecology, 1982, 142, 670-677. ‘__ Lubchenco, L.O.; Delivoriai-Papadoupoulos, M.; Searls, D. Long-term follow-up studies of prematurely born infants. II. Influence of birth weight and gestational age in sequelae. The Journal of Pediatriacs, 1972, 89, 509-512. Lubchenco, L.O.; Hansman, C.; Dressler, M.; Boyd, E. Intrauterine growth as estimated from liveborn birth-weight data at 24 to 42 weeks of gestation. Pediatrics, 1963, 793-800. Marx, J. Cytomegalovirus: a major cause of birth defects. Science, 1975, 190, 1184-1186. McCarthy, Dorothea. Manual for the McCarthy Scales of Children’s Abilities. New York: The Psychological Corporation, 1970. McDonald, A. Retarded foetal growth. In M. Dawkins & W.G. MacGregor (Eds.) Gestational Age, Size and Maturity. London: Spastics Society Medical Education and dInformation Unit, 1965. 151 Melish, M.E.; Hanshaw, J.B. Congenital cytomegalovirus infection: developmental progress of infants detected by routine screening. American Journal of Diseases of Children, 1973, 126, 190-194. Michaelis, R.; Schulte, F.J.; Nolte, R. Motor behavior of small for gestational age newborn infants. The Journal of Pediatrics, 1970, ‘16, 208-213. Miller, H.C.; Hassanein, K. Diagnosis of impaired fetal growth in newborn infants. Pediatrics, 1971, 48, 511-522. Miller, J.A. Small-baby biology. Science News, 1983, 124, 266-267. Neligan, G. A community study of the relationship between birth weight and gestational age. In M. Dawkins & W.G. MacGregor (Eds.) Gestational Age, Size, and Maturity. London: Spastics Society Medical Education and Information Unit, 1965. Neligan, G.A.; Kolvin, I.; Scott, D.M.; Garside, R.F. Born Too Soon or Born Too Small. London: Spastics International Medical Publications, 1976. Nelson, K.B.; Ellenberg, J.H.. Obstetric complications as risk factors for cerebral palsy or seizure disorders. The Journal of the American Medical Association, 1984, 251, 1843-1848. '-_ Nesbitt, R.B.L. Perinatal development. In F. Falkner (Ed.) Human Development. Philadelphia: W.B. Saunders Company, 1966. Norman, L.A. Intrauterine growth retardation. American Family Physician, 1982, 26, 171-176. North, A.F. Small-for-dates neonates. Pediatrics, 1966, 33, 1013- 1019. “'— Ounsted, M. Fetal growth and mental ability. Developmental Medicine and Child Neurology, 1970, 2, 222-224. Ounsted, M.; Cockburn, J.; Moar, V.A.; Redman, C. Maternal hypertension with superimposed pre-eclampsia: effects on child development at 7 1/2 years. British Journal of Obstetrics and Gynaecology, 1933, _9_9_, 644-649. " Ounsted, M.; Moar, V.; Good, F.; Redman, C. Hypertension during pregnancy with and without specific treatment; the development of the children at the age of four years. British Journal of Obstetrics and Gynaecology, 1980, 81, 19-24. ‘_— Ounsted, M.; Moar, V.; Cockburn, J.; Redman, C. Factors associated with intellectual ability of children born to women with high risk pregnancies. British Medical Journal, 1984, 238, 1038-1041. Ounsted, M.; Moar, V.; Redman, C. Infant growth and development following treatment of maternal hypertension. The Lancet, 1980, 705. Paine, P.A.; Pasquali, L. Effects of intrauterine growth and gestational age upon infants’ early psychomotor development in Brazil. Perceptual §_Motor Skills, 1982, 55, 871-880. Pape, K.B.; Buncic, R.J.; Ashby, S.; Fitzhardinge, P.M. The status at two years of low-birth-weight infants born in 1974 with birth weights of less than 1,001 gm. The Journal of Pediatrics, 1978, 22, 253-260. Parkinson, C.E.; Wallis, 8.; Harvey, D. School achievement and behavior of children who were small-for-dates at birth. Developmental Medicine and Child Neurology. 1981, 23, 41-50. Parmelee, H.H.; Schulte, F.J. Developmental testing of pre-term and small-for-date infants. Pediatrics, 1970,‘45, 21-28. Pasamanick, B.; Knobloch, H. Epidemiologic studies on the complications of pregnancy and the birth process. In G. Caplan (Ed.) Prevention of Mental Disorders in_Childhood. New York: Basic Books, 1961. Pasamanick, 8.; Knobloch, H. Retrospective studies on the epidemiology of reproductive casualty: old and new. Merrill-Palmer Quarterly, 1966, 12, 7-26. Pasamanick, B.; Lilienfeld, A.M. Association of maternal and fetal factors with development of mental deficiency. The Journal o£_the American Medical Association, 1955, 159, 155-159. Pass, R.F.; Stagno, 8.; Myers, G.J.; Alford, G.A. Outcome of symptomatic congenital cytomegalovirus infection: results of long- term longitudinal follow-up. Pediatrics, 1980, 66, 758-762. Pick, W. The malnourished newborn. Pediatrics, 1964, 33, 1000- 1001. Quaranta, P.; Currell, R. Prediction of small-for-dates infants by measurement of symphysial-fundal-height. British Journal of Obstetrics and Gynecology, 1981, 88, 115-119. Ramey, C.T.; Farrah, D.C.; Campbell, F.A. Predicting I0 from mother-infant interactions. Child Development, 1979, 59, 804-814. Redman, C. The definition of pre-eclampsia. Scandinavian Journal ‘of Clinical and Laboratory Investigation, 1984, 169, 7-14. Reynolds, 3.; Butters, L.; Evans, J.; Adams, T.; Rubin, P. First year of life after the use of atenolol in pregnancy associated with hypertension. Archives of Disease in_Childhood, 1984, 22, 1061- 1063. 153 Robertson, J.S. Low birth weight, drink, smoking, and diet. public Health, 1983, 27, 307-308. Rubin, R.A.; Rosenblatt, C.; Balow, B. Psychological and educational sequelae of prematurity. Pediatrics, 1973, 22, 352-363. Russell, P. Inflammatory lesions of the human placenta. III. The histopathology of villitis of unknown aetiology. Placenta, 1980,_1, 227-244. Saigal, S.; Rosenbaum, P.; Stoskepf, B.; Milner, R. Follow-up of infants 501 to 1,500 gm. birth weight delivered to residents of a geographically defined region with perinatal intensive care facilities. The Journal o£_Pediatrics, 1982, 199, 606-613. Salonen, J.T.; Heinonen, O.P. Mental retardation and mother’s hypertension during pregnancy. Journal o£_Mental Deficiency Research, 1984, 28, 53-56. Sameroff, A.J. Early influences on development: fact or fancy? Merrill-Palmer Quarterly, 1975, 4, 267-294. Sameroff, A.J.; Chandler, M.J. Reproductive risk and the continuum of caretaking casualty. Review of Child DevelOpment Research, 1975, ‘4, 187-244. Sameroff, A.J.; Seifer, R. Familial risk and child competence. Child Development, 1983, 54, 1254-1268. Sander, C.H. Placental tissue registry important to health of future mothers and infants. Michigan Medicine, 1978, 362. Sander, C.H. Hemorrhagic endovasculitis and hemorrhagic villitis of the placenta. Archives of Pathology and Laboratory Medicine, 1980, 104, 371-373. Sander, C.H.; Stevens, N.G. Hemorrhagic endovasculitis of the placenta: an indepth morphologic appraisal with initial clinical and epidemiologic observations. Pathology Annual, 1984, 12, 37-79. Schutt, W. Foetal factors in intrauterine growth retardation. In M. Dawkins & W.G. MacGregor (Eds.) Gestational Age, Size and Maturity. London: Spastics Society Medical Education and Information Unit, 1965. Scott, K.B.; Usher, R. Fetal malnutrition: its incidence, causes, and effects. American Journal o£_Obstetrics and Gynecology, 1966, 24, 951-963. Sever, J.L. Infectious agents and fetal disease. In H.A. Waisman & G.R. Kerr (Eds.) Fetal Growth and Development. New York: McGraw- Hill Book Company, 1970. 154 Shennan, A.T.; Milligan, J.E. The growth and development of infants weighing 1000 to 2000 grams at birth and delivered in a perinatal unit. American Journal of Obstetrics and Gynecology. 1980, 136. 273-275. Sheppard, B.L.; Bonnar, J. Ultrastructural abnormalities of placental villi in placentae from pregnancies complicated by intrauterine fetal growth retardation: their relationship to decidual spiral arterial lesions. Placenta, 1980, I, 145-156. Silverstone, A.; Trudinger, E.J.; Lewis, P.J.; Bulpitt, C.J. Maternal hypertension and intrauterine fetal death in mid-pregnancy. British Journal of Obstetrics and Gynaecology, 1980, 81, 457-461. Sokol, R.J. Perinatal risk and cerebral palsly. Journal o£_the American Medical Association, 1984, 251, 1868-1869. Stevens, N.G. Placental hemorrhagic endovasculitis: risk factors and impact on pregnancy outcome. Laboratory Investigation, 1984, 50, 57A. Stevens, N.G.; Sander, C.H. Placental hemorrhagic endovasculitis: risk factors and impact on pregnancy outcome. International Journal of_§ynaecology and Obstetrics, 1984, 22, 393-397. Taylor, D.J. Low birthweight and neurodevelopmental handicap. Clinics in Obstetrics and Gynaecology, 1984, 11, 525-542. Tejani, N.A. Recurrence of intrauterine growth retardation. Obstetrics and Gynecology, 1982, 29, 329-331. Tejani, N.A.; Mann, L.I. Diagnosis and management of the small-for- gestational-age fetus. Clinical Obstetrics and Gynecology, 1977, 29, 943-955. . Tudehope, D.; Burns, Y.; O’Callaghan, M.; Mohay, H.; Silcock, A. The relationship between intrauterine and postnatal growth on the subsequent psychomotor development of very low birthweight (vlbw) infants. Australian Paediatric Journal, 1983, 19, 3-8. van den Berg, B.J.; Yerushalmy, J. The relationship of the rate of intrauterine growth of infants of low birth weight to mortality, morbidity, and congenital anomalies. The Journal of Pediatrics, 1966, _6__9_, 531-545. ‘— Vohr, B.R.; Oh, W. Growth and development in preterm infants small for gestational age. The Journal of Pediatrics, 1983, 103, 941-945. Vohr, B.R.; Oh, W.; Rosenfield, A.G.; Cowett, R.M.; Berstein, J. The preterm small-for-gestational age infant: a two year follow-u study. American Journal of Obstetrics and Gynecology, 1979, 133, 425-431. 155 Walther, F.J.; Ramaekers, L.H.J. Growth in early childhood of newborns affected by disproportionate intrauterine growth retardation. Acta Paediatrica Scandinavica, 1982, 11, 651-656. (a) Walther, F.J.; Ramaekers, L.H.J. Language development at the age of 3 years of infants malnourished in utero. Neuropediatrics, 1982, _1_3_, 77-81. (b) Warkany, J.; Monroe, B.B.; Sutherland, B.S. Intrauterine growth retardation. Journal of Diseases of Children, 1961, 192, 127-157. Westwood, M.; Kramer, M.S.; Munz, D.; Lovett, J.M.; Watters, G. Growth and development of full-term nonasphyxiated small-for- gestational-age newborns: follow-up through adolescence. Pediatrics, 1983, 11, 376-382. Winer, E.K.; Tejani, N.A.; Atluru, V.L.; DiGiuseppe, R.; Borofsky, L.G. Four- to seven-year evaluation on two groups of small-for- gestational age infants. American Journal of Obstetrics and Gynecology, 1982, 425-429. "_ Yerushalmy, J. The classification of newborn infants by birth weight and gestational age. The Journal of Pediatrics, 1967, 71, 164-172. ~ “"‘ Zeskind, P.S.; Lester, B.M. Acoustic features and auditory perceptions of the cries of newborns with prenatal and perinatal complications. Child Development, 1978, 49, 580-589. Zeskind, P.S.; Lester, B.M. Analysis of cry feataures in newborns with differential fetal growth. Child Development, 1981, 52, 207- 212. “ Zeskind, P.S.; Ramey, C.T. Fetal malnutrition: an experimental study of its consequences on infant development in two caregiving environments. Child Development, 1978, 423 1155-1162. Zeskind, P.S.; Ramey, C.T. Preventing intellectual and interactional sequelae of fetal malnutrition: a longitudinal, transactional and synergistic approach to development. Child Development, 1981, 52, 213-218. 156