ABSTRACT.An index of adversity is a measureof risk that can be considered independently of individual risk factors. This study examined four areas of adversity in early childhood, namely perinatal complications, family background, child-rearing practices, and the child's phys ical health, and their relationship to developmental out comes. Four indices of adversity in these areas were examined as predictors of cognitive ability and motor
ability for 476girls and 510boysat age5 years.Results of the study indicated that indices of family background
and child-rearing practices were highly related to these developmental outcomes. An index of health problems was found to be significantly related to motor ability. The perinatal complications index was significantly re lated only to specific cognitive ability scores for boys. Previously, developmental outcomes have been assessed in terms of the magnitudeof IndividUalrisk factors,but more effective screening procedures may need to take account of the additive effect of the number of relevant adverse risk factors. Pediatrics 1991;88:954-959; devel
opmental outcomes, perinatal complications, family back
ground, child-rearing practices, health status, cognitive
abilit)@motor ability.
There is a wide range of factors that have been shown to influence the early development of chil then, including perinatal, health, family, parental, and experiential variables. Historically, research has been primarily concerned with the effects of individual risk factors such as low birth weight, low
Received for publication May 9, 1990;accepted Sep 29, 1990. Reprint requests to (W.RS.) Dunedin Multidisciplinary Health and Development Research Unit, Dept of Paediatrics and Child Health, University of Otago Medical School, P0 Box 913, Dij. nedin, New Zealand.
PEDIATRICS (ISSN 0031 4005). Copyright ©1991 by the American Academy of Pediatrics.
socioeconomic status, parental separations, and childhood illnesses. More recently, in addition to the independent associations ofthese variables with measures of child development, the cumulative ef fect of the presence of a number of such risk factors is receiving an increasing amount of attention. An “¿index―approach to developmental and behavioral outcomes is a useful way of dealing with a large array of known risk factors. More importantly, this approach acknowledges that a number of risk fac tors may have more important cumulative effects on child development than any one risk factor considered in isolation.
The concept of a “¿continuumof reproductive casualty― associated with perinatal complications which independently may or may not result in impairment was suggested by Pasamanick and colleagues―2 in the 1950s. Recently Cohen et al3 used an index of pregnancy problems including serious illness, trauma during pregnancy, compli cations of pregnancy or delivery, and birth weight to examine effects on emotional and behavioral problems in late childhood and adolescence. They reported a significant effect for pregnancy problems in relation to conduct disorder and separation anx iety. A similar perinatal complications index was used by Stanton et al4 to study the effect on cog nitive ability. Results of that study showed a sig nificant difference in mean IQ scores for children with none and one or more perinatal complications which persisted from age 5 to age 13 years. Addi tional perinatal problems were not associated with a further significant decrease in IQ score.
A “¿continuumof caretaking adversity―resulting from environmental risk factors has been suggested by Sameroff.5 Rutter6 developed an index of family
Indicesof PerinatalComplications,Family
Background,Child Rearing, and Health as
Predictorsof EarlyCognitiveand Motor
Development
Warren R. Stanton, PhD; Rob McGee, PhD;and Phil A. Silva, PhD
From the Dunedin Multidisciplinary Health and Development Research Unit, Department of Paediatricsand ChildHealth,Universityof OtagoMedicalSchool,
adversity consisting of maternal, paternal, and fam ily background variables and concluded that single isolated sources of chronic stress are unimportant in relation to conduct disorders in childhood. A cumulative risk index containing maternal, family, and cultural variables was used by Sameroff et al7 in their study of the association between levels of family adversity and IQ scores. As different com binations of equal numbers of risk factors resulted in similar effects they concluded that, in support of the index approach, no single factor enhanced or limited intellectual functioning. The family adver sity index used by Stanton et al4 included measures of socioeconomic status, maternal characteristics, and family-related variables. Significant decreases in IQ scores were found to be associated with in creases in the level of family adversity, which per sisted through to adolescence.
Indices of health seem to have received the least amount of attention in the literature. In part, this may be due to the difficulties in combining dispar ate measures of health status. A cumulative index of illness was developed by Carmichael and Williams8 and included measures of acute episodic illnesses, duration of these illnesses, seeking medi cal attention for episodic illnesses, continuing or recurrent illness, and admission to hospital. They found no statistically significant association be tween the measure of level of illness and variables related to family background, including ethnicity, economic level, the fathers' support in household tasks or child rearing, and the mothers' level of education, English ability, or psychological health. An index of illness and injury used by Cohen et al3 included number of accidents, major illness such as infectious diseases and hospitalization, and minor illness including minor communicable diseases, asthma, and allergy. Their results indicated that illness and injury was related to all the measures of psychopathology used in the study.
The present study was designed to test the rela tive cumulative effect of indices of perinatal corn plications, family background, child rearing, and health as predictors of early cognitive development (measured by IQ scores and language ability scores) and motor development.
METHODS
Sample
The children in the sample were enrolled in the Dunedin Multidisciplinary Health and Develop ment Study, a longitudinal investigation of their health, behavior, and development. They are part of a cohort born at Queen Mary Hospital, Dunedin,
between April 1, 1972, and March 31, 1973. When followed up at age 3 years, 1139 children were known to be living in Otago province and 1037 were traced and assessed. There were 991 children as sessed at age 5 years. Most of them were assessed within 2 months of their birthdays. The sample is known to be slightly overrepresentative of children from relatively higher socioeconomic levels in corn parison with the country as a whole, and under representative of Maori and other Polynesian chil dren. The history of the study and the sample are described in more detail by McGee and Silva9 and Silva.'°
Measures
In the development of the indices, consideration was given to weighting individual items. Maintain ing that equally weighted composite variables es sentially provide the same results and conclusions as optimally weighted ones, Cohen et al3 summed relevant items to produce their indices. In addition, the characteristics of weighted composites are more likely to be sample specific. Equal weighting was therefore given to each of the items used in the following indices.
Perinatal Complications Index. The perinatal in dex used in this study consisted of the sum of the following complications which were present for some mothers during pregnancy or birth of the study child diabetes; prediabetes; epilepsy; glyco suria; hypertension, moderate (diastolic blood pres sure 100 to 109 mm Hg) or severe (more than 109 mm Hg); antepartum hemorrhage; accidental hem orrhage; placenta previa; twins; delivery other than spontaneous (forceps and rotation vertex delivery, cesarean section or breech birth); low Apgar score at birth (required resuscitation, regular respiration not established 10 minutes after birth; at 15 mm utes heart rate less than 100 beats per minute, respiration irregular or absent and signs of ventral cyanosis); small for gestational age (10th percentile of birth weight for gestational age or less); preterm (<37 weeks gestational age); idiopathic respiratory distress syndrome; apnea; minor neurological signs of the neonatal period (jitteriness, tenseness, limp ness, hypotonicity) or major neurological signs; nonhemolytic hyperbilirubinemia (serum bilirubi nemia levels >15 mg/i® mL); Rh incompatibility and hyperbilirubinemia; ABO incompatibility and hyperbilirubinemia. The prevalence of these pen natal problems in the Dunedin sample is reported elsewhere.11―2
Health Index. A health index at age 3 years was obtained by summing the presence of the following conditions: accidental poisoning that resulted in
ARTICLES 955
at Viet Nam:AAP Sponsored on September 1, 2020 www.aappublications.org/news
TABLE
Four In1.
Number(Percent)of Girlsin E dices of Adversity (n = 501)ach
Level ofLevelIndexPerinatal
Family Child
Background RearingHealth0349
(69.6) 197 (39.3) 240 (47.9)306 (61.1)1100
(20.0) 176 (35.1) 169 (33.7)139 (27.7)235
(7.0) 82 (16.4) 66 (13.2)36 (7.2)310
(2.0) 36 (7.2) 23 (4.6)18 (3.6)42
(0.4) 7 (1.4) 2 (0.4)1 (0.2)54
(0.8) 3 (0.6) 1 (0.2)0
(0.0)61
(0.2) 0 (0.0) 0 (0.0)1 (0.2)
TABLE
Four In2.
Number(Percent)of Boysin E dices of Adversity (n = 536)ach
Level ofLevelIndexPerinatal
Family Child
Background RearingHealth0370
(69.1) 223 (41.6) 258 (48.1)287 (53.5)1113
(21.1) 162 (30.2) 183 (34.2)164
(30.6)235
(6.5) 101 (18.9) 74 (13.8)62 (11.6)312
(2.2) 36 (6.7) 17 (3.2)19 (3.5)46
(1.1) 13 (2.4) 4 (0.7)3 (0.6)50
(0.0) 1 (0.2) 0 (0.0)1
(0.2)60
(0.0) 0 (0.0) 0 (0.0)0 (0.0) parents' seeking medical advice'3; injury (two or
more types) resulting in attendance to outpatient department of a hospital, including burns, lacera tions, falls, fractures'3; admissions to hospital (two or more episodes); neurological signs (two or more) including assessment of spontaneous motility, range of passive movements, resistance to passive movements, reflexes, planter response, foot grasp, palmo-mental response, fingertip touching, facial asymmetry, strabismus, nystagmus, posture of feet and gait'4; colds (six or more episodes in last year); ear infections (two or more episodes in last year); infectious disease (either English measles, mumps, whooping cough, or chickenpox); under care of a doctor (including, for example, allergies, asthma, heart murmur, hernia, recurrent tonsillitis, ortho paedic correction, strabismus correction).
Family Adversity Index. The index of family ad versity used by Stanton et al4 and included in this study contained the following measures adminis tered at birth or at the age 3 assessments: low socioeconomic status (father with semiskilled or unskilled job or single mother at birth of study child); mother younger than 20 years old at birth of her first child single mother when the study child was 3 years of age; large family size of four or more children; low maternal mental ability (SRA Verbal Form'5 score below the 10th percentile); high maternal neuroticism score (Eysenck Person ality Inventory'6 score above the 90th percentile).
Child-Rearing Index. A child-rearing index was developed from the presence of the following con ditions measured at age 3 years. Previous studies have shown that child-rearing attitudes and prac tices and preschool experience are related to devel opmental outcomes, but they have not previously been used to develop an index. A cutoff of 15% of the scores for each measure was used to define adversity relative to the remainder of the sample.
The following conditions were measured: high score for authoritarianism (based on the subscales
of Excluding Outside Influences, Intrusiveness, and
Acceleration of Development, from the Parental Attitude Research Instrument)'7; low score for egal itanianism (based on the subscales of Encouraging Verbalization, Egalitarianism, and Comradeship and Sharing, from the Parental Attitude Research Instrument)'7; high score for rejection of child (based on a psychornetrist's observation of the mother's expression of affection, critical evaluation of child, physical handling of child, management of child during testing, reactions to child's needs, re action to child's test performance, mother's focus of attention, physical appearance of child'8 (each item was scored 0 to 2 and the scores were summed.); high score for protectiveness of child
(based on the above observations of the mother by a psychometnist)'8; number of experiences (includ ing, for example, rides on a bus, a boat, or a horse, going to a restaurant, zoo, factory, movies, museum, sports event, beach, holiday, or circus, and watching television)'9; child did not attend preschool; parent had no training, information, or experience in child reaning separations from child (more than one separation of more than a week on each occasion).
Developmental Measures. Assessment of cogni tive development at age 5 years included adininis tration of the Stanford-Binet Intelligence Scale@°'21 and the Reynell Receptive and Expressive Lan guage Scales.22'@ The psychornetrists administering the measures were unaware of the children's pen natal histories, early family histories, and previous test results. Motor skills at age 5 were assessed with the McCarthy Motor Scale.24
RESULTS
TABLE3. Spearmanp CorrelationCoefficientsfor
Four Indices of Adversity
PerinatalFamily Child Health BackgroundRearingBoys
(n = 536)Perinatal1.01
.07t.06Family
background.041 [email protected]@Child
rearing.lOt.17@ 1.04Health.06.02
.01 Girls (n = 501)1
scores. The mean trend in IQ scores for groups of increasing numbers of adversities (0 to 7 or more) showed a consistent decrease which totaled 28 IQ points (1.7 standard deviations). The largest de crease of 11 IQ points occurred between the group with 6 adversities and the group with 7 adversities. Receptive language scores showed a steady decrease across consecutively higher levels of adversity and represented a total decrease of 1.4 standard devia tions of the mean score.
DISCUSSION
Four major areas of adversity were examined in this study, namely, peninatal complications, family background, child-rearing practices, and physical health. The aim was to study the relative effects of these sources of adversity on cognitive ability (measured by IQ, receptive language, and expressive language) and motor ability. Development of these indices has to some extent been limited by the choice of measures that were administered in the early stages of this longitudinal study. Thus, it is possible that other measures may provide a more sensitive composite index in any of the four do mains of adversity. Consequently, our indices should be regarded as providing an estimate of the degree of adversity experienced by any particular child.
The results of the present study supported the concept of a continuum of adversity in the domains of perinatal complications,3 family background,6'7 and health3'8 and provided evidence of its applica biity to the area of child rearing. The results mdi cated that girls' general cognitive ability and motor ability in early childhood were not related to the number of perinatal complications present during pregnancy or birth. In addition, girls' language abil ity was not related to the number of health prob lems present at age 3. Their general cognitive ability and motor ability were related to the number of adverse family background conditions present from birth to age 3 and the number of adverse child rearing practices present at age 3. Motor ability was related to the number of health problems pres ent at age 3.
A slightly different pattern of results emerged for boys. Their IQ scores and receptive language scores but not their expressive language and motor ability scores were related to the index of perinatal corn plications. IQ, receptive language, and motor ability scores were related to the family background index and child-rearing index. Motor ability was also related to the health index. The only significant predictor of boys' expressive language scores was the child-rearing index, indicating that different
t P < .05.
:1:P < .005.of peninatal complications and family background. That is, adverse child-rearing practices were asso ciated with a relatively higher level of perinatal complications and a higher level of family adversity. For boys there was also a significant association between an adverse family background and a poor level of health.
To test the relative value of the four indices of adversity, they were entered in a regression analysis in the order in which they could be expected to influence children's development, that is the chron ological order of perinatal complications, family background, child rearing, and then health. The health index could be entered before the family rearing index but doing so made no difference to the overall model. A separate analysis was done for girls and boys and for each of the developmental outcome variables administered at the age of school
entry. There were 476 girls and 510 boys for whom
data were available at all relevant ages.
The results of the regression analysis for girls are shown in Table 4 and the results for boys are shown in Table 5. The peninatal complications index was not associated with any of the cognitive ability measures or the motor ability measure for girls but for boys was predictive of their IQ scores and Rey nell Receptive Language scores at age 5. The family background index and the child-rearing index were predictive of scores for each of the developmental measures for girls but for boys there was no effect of family background on their Reynell Receptive Language scores. The health index was related to girls' IQ scores and motor ability scores and boys' motor ability scores.
The cumulative percentage of the variance ex plained by the four indices is highest for the meas ures of IQ and receptive language. A total score obtained from summing the scores across the four indices produced a range of from 0 to 12 adversities, with 55 children (5.3% of the total sample at age 3 years) scoring 7 or more. However, use of this variable in regression analyses did not increase the variance explained in the IQ and receptive language
ARTICLES 957
at Viet Nam:AAP Sponsored on September 1, 2020 www.aappublications.org/news
TABLE4. Resultsof RegressionAnalysesfor the Effectsof Four mdi on Girls' Cognitive Ability and Motor Ability at Age 5 (n = 476)ces
of AdversityDevelopmental
Outcome Multiple Percentage lncrease inF forIndex
r inanceof Van- VarianceChange
Ex-ExplainedMultipleplainedCorrelationIQPerinatal
complications .08 0.7
0.73.3Family
background .30 8.8
8.141.9*Child
rearing .42 17.3
8.548.8*Health
.43 18.6
1.37@4*Receptive LanguagePerinatal
complications .01 0.0
0.00.1Family
background .17 3.0
3.013.9*Child
rearing .34 11.8
8.845.2*Health
.34 11.8
0.00Expressive LanguagePerinatal
complications .01 0.0
0.00.1Family
background .13 1.6
1.67.2*Child
rearing
.20
4.1
2.511.8*Health
.20 4.1
0.00Motor
abilityPerinatal
complications .06 0.3
0.31.5Family
background .13 1.7
1.46.4*Child
rearing .21 4.4
2.712.4*Health
.23 5.4
1.04.6**
P < .05.
TABLE 5. Resultsof RegressionAnalysesfor the Effectsof Fourmdi on Boys' Cognitive Ability and Motor Ability at Age 5 (n = 510)ces
of AdversityDevelopmental
Outcome Multiple Percentage Increase inF forIndex
r inanceof Van- VarianceChange
Ex-ExplainedMultiplepinnedCorrelationIQPerinatal
complications .10 0.9
0.94.8*Family
background .30 9.1
8.245•7*Child
rearing .40 16.3
7.243.2*Health
.41 16.8
0.53•3*Receptive LanguagePerinatal
complications .14 1.9
1.99.2*Family
background .23 5.4
3.517.4*Child
rearing .33 11.2
5.831.1*Health
.33 11.2
0.00.1Expressive
LanguagePerinatal
complications .03 0.1
0.10.5Family
background .08 0.6
0.52.5Child
rearing .18 3.2
2.612.9*Health
.18 3.3
0.10.2Motor abilityPerinatal
complications .00 0.0
0.00.0Family
background .19 3.5
3.517.9*Child
rearing .22 4.9
1.47.0*Health
.25 5.4
1.575**
@ < .05.
psychological stress than girls.25 However, in this study there were similar proportions of girls and boys at each level of risk for the four indices and in general the effects on cognitive development were similar in size for girls and boys. If a 2% increase in the explained variance is considered as a cutoff mechanisms seem to operate in the development of
8. Carmichael A, Williams HE. Infant health and family func
tioning in a poor socio-economic status multi-ethnic munic
ipality. Aust Paediatr J. 1983;19:61.-67
9. McGee R, Silva PA. A Thousand New Zealand Children
Their Health and Development From Birth to Seven. Auck land, New Zealand. Medical Research Council of New Zea land; 1982. Special Report Series No.8
10. Silva PA. The Dunedin Multidisciplinary Health and De
velopment Study a fifteen year longitudinal study. Paediatr Perinat Epidemiol 1990;4:76-107
11. Buckfield PM. Perinatal events in the Dunedin city popu
lation 1967—73.NZ Med J. 1978;88:244—246
12. Silva PA, McGee R, Williams S. A seven year follow up
study of the cognitive development of children who experi enced common perinatal problems. Aust Paediatr J. 1984;2th23—28
13. Silva PA, Buckfield PM, Spears GF, Williams S. Poisoning, burns, and other accidents experienced by a thousand Du nedin three year olds: a report from the Dunedin Multidis ciplinary Child Development Study. NZ Med J.
1978;87:242—244
14. McGee R, Clarkson JE, Silva PA, Williams S. Neurological
dysfunction in a large sample of three year old children: a report from the Dunedin Multidisciplinary Child Develop.
ment Study. NZ Med J. 1982;95:693-696
15. Thurstone TG, Thurstone LL The SPA Verbal Form. Chi
cago, IL: Science Research Associates; 1973
16. Eysenck HJ, Eysenck SBG. Manual of the Eysenck Person
ality Inventory. London, England. University of London
Press; 1964
17. Schaefer RS, Bell RW. Development of a parental attitude research instrument. Child Dev. 1958;29:339—361
18. The American COllabOrativeStudy of CerebralPalsy, Mental
Retardation and Other Sensory Disorders of Infancy and Childhood—PartiI-D, Manuals: Behavioral Examinations. Bethesda, MD: US Dept of Health, Education and Welfare, National Institutes of Health; 1966
19. Silva PA. Experiences, activities and the pre-school child. a
report from the Dunedin Multidisciplinary Child Study.
Aust J Early Child. 198O@,5:13-19
20. Terman LM, Merrill PA. Stanford-Binet Inteil@genceScale.
London, England. Harrop; 1961
21. Silva PA, Bradahaw J. Some factors contributing to intelli
gence at age of school entry. BrJ Educ PsychoL 1980;50:11-16
22. Reynell J. Reyneil Developmental tinguage Scales. London, England. National Foundation for Educational Research; 1969
23. McKerracher DW, Saklofske DH, Silva PA. An evaluation
andcross-cultural comparison ofthe Reynell Developmental Language Scales. Aust Read Educ J. 1977;2:14—17 24. McCarthy D. McCarthy Scales of Children's Abilities. New
York, NY: the Psychological Corporation; 1972
25. Werner EE, Smith RS. Vulnerable but Invincible: A Study
ofResiient Children. New York, NY: McGraw-Hill; 1982 point, the pattern of results is the same for girls
and boys except for motor ability, which for girls is related to child-rearing practices and for boys is related to family background. Finally, the multiple correlations in Tables 4 and 5 indicate that the size of the effects on the expressive language scores and motor ability scores were relatively smaller than the effects on the other measures, indicating that use of an index approach to screen for these partic ular outcomes may not be as useful as for IQ and receptive language ability.
In summary, the perinatal complications index was not significantly predictive of the majority of the cognitive ability and motor ability outcomes and the health index was predictive mainly of motor ability. The family background index and child rearing index were predictive of a wide range of developmental outcomes, most importantly the IQ and receptive language scores. The indices of family background and child rearing added equally to the variance explained in the IQ scores but child rearing contributed relatively more than family background to the receptive language scores.
REFERENCES
1. Pasamanick B, Rogers ME, Lilienfield AM. Pregnancy ex perience and the development ofbehavior disorder in Amer ican children. Am J Psychiatry. 1956;112:613—618
2. Pasamanick B, Knobloch H. Brain damage and reproductive
casuality. Am J Orthopsychiatry. 1960;30:298—305 3. Cohen P, Velez CN, Brook J, Smith J. Mechanisms of the
relation between perinatal problems, early childhood illness, and psychopathology in late childhood and early adoles cence. Child Dev. 1989;60:701—709
4. Stanton WR, McGee R, Silva PA. A longitudinal study of
the interactive effects of perinatal complications and early
family adversity on cognitive ability. Aust Paediatr J.
1989;25:130—133
5. Sameroff AJ. Caretaking or reproductive casualty? In: Ho.
rowitz FD, ed. Early Developmental Hazards: Predictions
and Precautions. Boulder, CO:Westview Press; 1978
6. Rutter M. Family, area and school influences in the genesis
of conduct disorders. In: Hersov LA, Berger M, Shaffer D,
ode.AggressionandAntisocialBehaviourin Childhoodand Adolescence. Oxford, Englan& Pergamon Press; 1978 7. Sameroff AJ, Seifer R, Barocas R, Zax M, Greenspan S.
Intelligence quotient scores of 4-year-old children: social environmental risk factors. Pediatrics. 1987;79:343—350
ARTICLES 959
at Viet Nam:AAP Sponsored on September 1, 2020 www.aappublications.org/news
1991;88;954
Pediatrics
Warren R. Stanton, Rob McGee and A. Silva
Predictors of Early Cognitive and Motor Development
Indices of Perinatal Complications, Family Background, Child Rearing, and Health as
Services
Updated Information &
http://pediatrics.aappublications.org/content/88/5/954
including high resolution figures, can be found at:
Permissions & Licensing
http://www.aappublications.org/site/misc/Permissions.xhtml
entirety can be found online at:
Information about reproducing this article in parts (figures, tables) or in its
Reprints
http://www.aappublications.org/site/misc/reprints.xhtml
1991;88;954
Pediatrics
Warren R. Stanton, Rob McGee and A. Silva
Predictors of Early Cognitive and Motor Development
Indices of Perinatal Complications, Family Background, Child Rearing, and Health as
http://pediatrics.aappublications.org/content/88/5/954
the World Wide Web at:
The online version of this article, along with updated information and services, is located on
American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.
American Academy of Pediatrics, 345 Park Avenue, Itasca, Illinois, 60143. Copyright © 1991 by the been published continuously since 1948. Pediatrics is owned, published, and trademarked by the Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it has
at Viet Nam:AAP Sponsored on September 1, 2020 www.aappublications.org/news