Abstract
Objectives
To study the relation of intra‐uterine growth and gestational age with child protection registration in a 20 year whole population birth cohort.
Setting
West Sussex area of England.
Study design
Retrospective whole population birth cohort.
Outcomes
Child protection registration; individual categories of registration—sexual abuse, physical abuse, emotional abuse, and neglect.
Population and participants
119 771 infants born in West Sussex between January 1983 and December 2001 with complete data including birth weight, gestational age, maternal age, and postcode.
Results
In all categories of registration a linear trend was noted such that the lower the birth weight z score the higher the likelihood of child protection registration. Similar trends were noted for gestational age. All these trends were robust to adjustment for maternal age and socioeconomic status.
Conclusions
The results of this study suggest that lower levels of fetal growth and shorter gestational duration are associated with increased likelihood of child protection registration in all categories including sexual abuse independent of maternal age or socioeconomic status. This study does not permit comment on whether poor fetal growth or preterm birth predispose to child abuse and neglect or the association arises because they share a common pathway.
Keywords: fetal growth, gestational age, child abuse and neglect, population based study
The association of low birth weight and preterm birth with child abuse and neglect has been the subject of debate for at least 30 years. Studies from the USA1 and the UK2 published in the 1970s and early 1980s reported an association. However, Leventhal,3 reviewing case‐control studies investigating the association, argued that many were subject to potential bias because of methodological weaknesses particularly in the choice of control group and adjustment for confounding factors and those studies with better methodologies showed no association of low birth weight or preterm birth with child abuse.
Subsequent studies have reported conflicting results.4,5,6,7,8,9,10 The conflicting findings are likely to relate to the different research methodologies used and the different populations studied. Most studies have a case‐control design4,5,6,7,8 with variation in the population from which the cases and controls are derived and the degree to which potential confounding variables are accounted for in the analyses. For example, among the case‐control studies, four4,6,7,8 were based on hospital populations and matching was used in two6,7 to account for confounding but others did not adjust for confounding. The more methodologically robust studies,9,10 based on whole populations, reported associations of child abuse with low birth weight after adjustment for potential confounding variables.
Published studies have examined the relation of pregnancy outcomes to child abuse registration for all types of abuse combined1,2,4,5,7,9,10 or specific forms of abuse such as fractures8 and physical abuse.6 Studies have tended to follow the usual convention of dichotomising birth weight into low (<2500 g) and normal birth weight2,5,9,10 and gestational duration into preterm (<37 weeks) and full term.1,4 This precludes exploration of a possible “dose‐response” relation between the pregnancy outcomes and child abuse. The use of low birth weight as the main predictor variable is also likely to obscure the distinction between low birth weight as a result of impaired fetal growth or short gestation.
This study is based on a 19 year whole population birth cohort (1983–2001), in which data on birth weight and gestation were recorded, linked with data from the child abuse register covering the same population from 1986 onwards. This is the first study to examine the association between risk of registration for child abuse in all major categories and intra‐uterine growth, assessed by birth weight z score, and gestational age in a whole population birth cohort.
Method
This study is a retrospective whole population birth cohort based on linkage of data from the West Sussex Child Health Computer including a Special Conditions File11 with the West Sussex Social Services Child Protection Register on children born between January, 1983 and end of December, 2001. The study was approved by the local research and ethics committee.
Data collection
West Sussex Child Health Computer collects data on all children born with addresses in the West Sussex area including those born outside the area, for example, in tertiary units. Children's files are initiated on the computer system by the birth notification that includes data on maternal age, birth weight, gestational age, and postcode of the address at the time of birth. Records on all children are regularly updated throughout childhood.
Children are entered onto the West Sussex Child Protection Register after a child protection investigation including a child protection conference. The criteria for registration are laid out in the West Sussex Area Child Protection Committee, Child Protection Procedures and state:
“The child can be shown to have suffered ill treatment or impairment of health or development as a result of physical, emotional or sexual abuse or neglect, and the professional judgement is that further ill treatment or impairment are likely”
Children are only registered under physical and emotional abuse and neglect if abuse has actually occurred. A child who has not yet suffered sexual abuse may be registered for sexual abuse if there is a known offender in the household. The entries on the electronic register are coded by category of registration with the main category first: physical abuse; sexual abuse; emotional abuse; neglect and non‐organic failure to thrive. For the purposes of this study, only data on the main category of abuse are available and the numerator used is individual children not abuse incidents.
Data linkage
The following data files were linked to form a single anonymised data file:
West Sussex Child Health Computer data file with Special Conditions Files from 1983–2001
West Sussex Social Services Child Protection Register—data on all children born between 1983 and 2001 with entries on the register during the period 1986–2003
Linkage was undertaken in accordance with the Data Protection Act and with permission of the local research and ethics committee.
Data extraction
For the purposes of this study, a single anonymised data file was created containing the following variables:
Outcomes of interest
Registration on the child protection register in any of the four categories listed above and extracted from the West Sussex Social Services Child Protection Register
Main independent variables of interest
(1) Birth weight standard deviation scores (z scores) were calculated from the birth weight in kilograms, gestational age in weeks and infant sex by use of software provided by the Child Growth Foundation derived from a representative sample of British births in 1990 (LMS software using 1990 British height and weight reference data, London, Child Growth Foundation) and categorised into five groups: <−2.00; −2.00 to −1.01; −1.00 to +1.00; +1.01 to +2.00; >+2.00
(2) Gestational age categorised into three groups: <34 weeks; 34–36 weeks; 37 weeks+
Confounding variables
(1) Socioeconomic status based on area of residence at birth created by conversion of postcode into enumeration district (the lowest census unit) and ranking enumeration districts into quintiles by their score on the Townsend deprivation index12 calculated from the 1991 census.
(2) Maternal age at infant's birth: two dummy variables were created to represent maternal age: <20 v the rest and 40+ v the rest—to account for a possible J shaped relation of the outcome variables with maternal age
Data analysis
Only children with complete data for all variables of interest were included in the analysis. Rates/1000 for registration in any child abuse category and for each category separately by birth weight z score and gestational age categories and χ2 for linear trend with p values were calculated. Binary logistic regression models fitted on the outcomes were used to adjust for maternal age and socioeconomic status. All analyses were carried out in SPSS v10 (SPSS, Chicago, 1999).
Results
Of 158 229 children entered onto the West Sussex Child Health Computer in the 19 year period, 1983–2001, 119 729 (76%) had complete data and were included in this study. Missing postcode data accounted for 33 128 children with missing data and of the remainder (5406), maternal age data were missing in 5187, gestational age in 168 and birth weight in 51. Rates of child abuse registration among the children without complete data did not differ from those included in the study (tables 1 and 2).
Table 1 Rates of child abuse registration.
| Number | Rate/1000 | |
|---|---|---|
| Child abuse registration | ||
| All categories | 1853 | 15.5 |
| Physical abuse | 616 | 5.1 |
| Sexual abuse | 246 | 2.1 |
| Emotional abuse | 635 | 5.3 |
| Neglect | 509 | 4.3 |
Table 2 Pregnancy outcome and demographic characteristics.
| Number | Percentage | |
|---|---|---|
| Birthweight z score group | 2347 | |
| <−2.00 | 14847 | 2.0 |
| −2.00 to −1.00 | 84250 | 12.4 |
| −0.99 to +1.00 | 15068 | 70.4 |
| +1.01 to 2.00 | 3217 | 12.6 |
| >+2.00 | 2.6 | |
| Gestational age group | 1947 | |
| <34 weeks | 5319 | 1.6 |
| 34–36 weeks | 112463 | 4.5 |
| 37+ weeks | 4618 | 93.9 |
| Maternal age group | 60817 | 3.9 |
| <20 | 51419 | 50.8 |
| 20–29 | 2875 | 42.9 |
| 30–39 | 23949 | 2.4 |
| 40+ | 23906 | 20.0 |
| Deprivation quintiles | 23996 | 20.0 |
| Quintile 1 (least deprived) | 23904 | 20.0 |
| Quintile 2 | 23974 | 20.0 |
| Quintile 3 | 20.0 | |
| Quintile 4 | ||
| Quintile 5 (most deprived) |
Statistically significant linear trends across birth weight z score group such that rates decreased as fetal growth increased were noted for each child abuse registration category (table 3). The relation of the combined categories of child abuse registration with birth weight z score showed a slight reverse J shape with a slight rate increase among the largest babies. A similar pattern was noted for physical and for emotional abuse. Registration for both sexual abuse and neglect showed a linear relation with decreasing birth weight z score. These relations persisted after adjustment for maternal age and socioeconomic status in logistic regression models.
Table 3 Number and rate/1000 in different categories of abuse registration by birth weight z score group.
| Abuse category | Birth weight z score group | χ2 for linear trend (p value) | ||||
|---|---|---|---|---|---|---|
| <−2 (n = 2347) | −2 to −1.00 (n = 14846) | −0.99 to +1.00 (n = 84249) | +1.01 to +2 (n = 15067) | >+2 (n = 3217) | ||
| All categories combined | 75 | 314 | 1255 | 166 | 43 | 89.9 |
| 32.0/1000 | 21.2/1000 | 14.9/1000 | 11.0/1000 | 13.4/1000 | (p<0.0001) | |
| Physical abuse | 23 | 110 | 403 | 61 | 19 | 19.5 |
| 9.8/1000 | 7.4/1000 | 4.8/1000 | 4.0/1000 | 5.9/1000 | (p<0.0001) | |
| Emotional abuse | 20 | 108 | 435 | 57 | 15 | 22.6 |
| 8.5/1000 | 7.3/1000 | 5.2/1000 | 3.8/1000 | 4.7/1000 | (p<0.0001) | |
| Sexual abuse | 10 | 40 | 171 | 21 | 4 | 14.9 |
| 4.3/1000 | 2.7/1000 | 2.0/1000 | 1.4/1000 | 1.2/1000 | (p<0.0001) | |
| Neglect | 24 | 94 | 342 | 41 | 8 | 43.1 |
| 10.2/1000 | 6.3/1000 | 4.1/1000 | 2.7/1000 | 2.5/1000 | (p<0.0001) | |
Linear trends by gestational age group in risk of child abuse registration by all categories combined and each separate category were statistically significant (table 4) such that the shorter the gestation the higher the risk of child abuse registration. Physical abuse, emotional abuse, and neglect all showed consistent trends across the gestational age groups. Registration for sexual abuse was more likely in children born at less than 34 weeks gestation but there was no difference in risks between those born at 34–36 weeks and those born at 37 weeks or more. All the trends noted in table 4 remained significant in logistic regression models including maternal age and socioeconomic status.
Table 4 Rate/1000 different categories of abuse registration by gestational age group.
| Abuse category | Gestational age<34 weeks (n = 1947) | Gestational age 34–36 weeks (n = 5319) | Gestational age 37+ weeks (n = 112463) | χ2 for linear trend (p value) |
|---|---|---|---|---|
| All categories combined | 69 | 124 | 1660 | 74.6 |
| 35.4/1000 | 23.3/1000 | 14.8/1000 | (p<0.0001) | |
| Physical abuse | 21 | 47 | 548 | 26.9 |
| 10.8/1000 | 8.8/1000 | 4.9/1000 | (p<0.0001) | |
| Emotional abuse | 18 | 45 | 572 | 15.5 |
| 9.2/1000 | 8.5/1000 | 5.1/1000 | (p<0.0001) | |
| Sexual abuse | 9 | 12 | 225 | 4.8 |
| 4.6/1000 | 2.3/1000 | 2.0/1000 | (p = 0.029) | |
| Neglect | 21 | 38 | 450 | 31.5 |
| 10.8/1000 | 7.1/1000 | 4.0/1000 | (p<0.0001) |
Discussion
The findings of this population based study show that infants experiencing poorer fetal growth or preterm birth are at increased risk of registration for physical, emotional, or sexual abuse, or neglect independent of maternal age and socioeconomic status.
To the best of our knowledge, ours is the first study to examine the relation between the main categories of child abuse registration and fetal growth and to examine the relation across the range of fetal growth. A number of studies have examined the association between low birth weight (defined as <2500 g) and child abuse and neglect,1,5,6,9,10 but unlike this study have been unable to distinguish the effects of fetal growth from those of preterm birth. This distinction is important when considering possible explanations for the association of child abuse registration with poor fetal growth and preterm birth as the determinants of fetal growth and gestation differ.13
It is worthy of note that, in contrast with the conflicting findings of case‐control studies,7 the findings of this study are broadly consistent with previous population based studies.9,10 Needell and Barth9 linked administrative birth data with foster care placement data for California between 1989 and 1994. They reported that infants admitted into foster care as a result of maltreatment were more than twice as likely as those not in care to have been born low birth weight after adjustment for single parenthood, family size, and ethnicity. Sidebotham et al,10 based on the Avon longitudinal study of parents and children with linked data from local child protection registers, showed a similar twofold increase in the likelihood of child abuse registration before the age of 6 years among children born low birth weight after adjustment for unintended pregnancy, hospital admissions, feeding difficulties, and other behavioural attributes of the child. Neither of these studies was able to adjust the findings for socioeconomic status or maternal age.
There are a number of possible explanations for the associations reported here. Preterm infants or those with poor fetal growth may have characteristics that make them more vulnerable to all forms of abuse. It is possible that such infants may be more likely to provoke hostile parental feelings leading to increased risk of abuse. Early separation, more commonly experienced by preterm and small for gestational age infants, may interfere with parent‐infant bonding, although this is unlikely to be an important factor except at the extremes. Alternatively, preterm birth and poor fetal growth may share a common pathway with abuse, for example, through maternal characteristics that predict increased risk of both poor pregnancy outcomes and child abuse. It is also possible that an unidentified confounding variable explains the apparent association. The design of this study does not permit definitive comment on these explanations. However, this study does suggest that any explanation must be consistent with the findings that all main categories of abuse broadly show the same association with both fetal growth and preterm birth and the association is not confined to infants born very early or very small but shows a trend across the range of fetal growth and gestational duration.
Strengths and limitations
The main strengths of this study are that it is population based, allowing us to examine the relation between fetal growth, gestational age and abuse in a whole population rather than in highly selected subgroups, and the large sample size. In addition, the availability of population based data on potential confounding variables including maternal age and socioeconomic status has allowed us to examine the effects of these factors on the reported associations. Finally, linkage with the West Sussex Social Services Child Protection Register allowed us to study these associations with registration for each of the major categories of abuse separately.
Key points
This is the first study to examine the association between risk of registration for child abuse in all major categories and intra‐uterine growth, assessed by birth weight z score, and gestational age in a whole population birth cohort
Lower levels of fetal growth and shorter gestational duration are associated with an increased likelihood of child protection registration in all categories including child sexual abuse
These associations are independent of maternal age and socioeconomic status
Child abuse registration categories and thresholds for registration are likely to have changed over the 19 year study period. These changes are likely to have led to misclassification bias affecting the classification of abuse. However, unless they can be shown to differentially increase or decrease the estimate of abuse rates among children experiencing different levels of fetal growth and gestational duration, they are unlikely to systematically bias the relations studied. The length of the study period means that the period of risk exposure varies from 18 years to one year but, as this is true for all children, independent of pregnancy outcome, it is again unlikely to bias the relation of fetal growth or preterm birth with abuse.
Conclusions
This study confirms the findings of previous population based studies that there is an association of low birth weight with child abuse registration and extends these findings by showing that the association holds for both fetal growth and preterm birth. This is the first study to report that this association exists for all categories of abuse and that the association is seen across the whole range of fetal growth and gestational age, not simply an association with the extremes. These findings do not permit an explanation of the associations noted and clarification must await further research. However, plausible explanations must take account of these findings.
Policy implications
Strategies and interventions aimed at preventing child abuse need to take account of the association with poor fetal growth and short gestational duration
Further study is needed to explore whether poor fetal growth and preterm birth predispose to child protection registration or are markers of maternal characteristics that predict increased risk of both poor pregnancy outcomes and child abuse
Acknowledgements
We wish to acknowledge the work of the paediatricians, child health professionals, and social workers in West Sussex whose dedication and attention to detail has made possible the consistent and reliable collection of the data reported in this study.
Footnotes
Funding: the study was conceived and undertaken by the authors working as a team without financial assistance.
Conflicts of interest: none.
References
- 1.Goldson E, Fitch M J, Wendell T A.et al Child abuse. Its relationship to birthweight, apgar score and developmental testing. Am J Dis Child 1978132790–793. [PubMed] [Google Scholar]
- 2.Murphy J F, Jenkins J, Newcombe R G.et al Objective birth data and the prediction of child abuse. Arch Dis Child 198156295–297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Leventhal J M. Risk factors for child abuse: methodologic standards in case‐ control studies. Pediatrics 198168684–690. [PubMed] [Google Scholar]
- 4.Shearman J K, Evans C E, Boyle M H.et al Maternal and infant characteristics in abuse: a case control study. J Fam Pract 198316289–293. [PubMed] [Google Scholar]
- 5.Starbuck G W, Krantzler N, Forbes K.et al Child abuse and neglect on Oahu, Hawaii: description and analysis of four purported risk factors. J Dev Behav Pediatr 1984555–59. [PubMed] [Google Scholar]
- 6.Benedict M I, White R B. Selected perinatal factors and child abuse. Am J Public Health 198575780–781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Leventhal J M, Berg A, Egerter S A. Is intrauterine growth retardation a risk factor for child abuse? Pediatrics 198779515–519. [PubMed] [Google Scholar]
- 8.Dahlenburg S L, Bishop N J, Lucas A. Are preterm infants at risk for subsequent fractures? Arch Dis Child 1989641384–1385. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Needell B, Barth R P. Infants entering foster care compared with other infants using birth status indicators. Child Abuse Negl 1998221179–1187. [DOI] [PubMed] [Google Scholar]
- 10.Sidebotham P, Heron J and the ALSPAC Study Team Child maltreatment in the “children of the nineties”: the role of the child. Child Abuse Negl 200327337–352. [DOI] [PubMed] [Google Scholar]
- 11.Abra A, Woodroffe C. A special conditions register. Arch Dis Child 199166927–930. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Townsend P, Phillimore P, Beattie A.Health and deprivation: inequality and the North. Beckenham: Croom Helm, 1988
- 13.Kramer M S. Determinants of low birth weight: methodological assessment and meta‐analysis. Bull World Health Organ 198765663–737. [PMC free article] [PubMed] [Google Scholar]