American Journal of Diseases of Children, Volume 139, May 1985, 456-459.
Childhood Ingestions as Symptoms of Family Distress
William G. Bithoney, MD; Jane Snyder, PhD: Joanne Michalek, RN, MS;
Eli H. Newberger, MD
• Familial, child developmental, and demographic concomitants of serious Ingestions In preschool children were measured In 23 hospitalized children and controls matched on age, race, and socioeconomic status. A precoded maternal Interview focused on family stress, parental discipline, parental emotional history, as well as childhood temperament and social maturity. A regression analysis was performed on variables that discriminated between Ingestion cases and controls. Sixteen variables from the regression equation were entered into a stepwise discriminant function analysis. Significant descriptors of ingestion victims Included the following: lack of extended family, low Vineland Social Maturity quotient, few maternal opportunities to escape caregiving, good health, a high frequency of physical punishment In the mother’s childhood, and Increased current advocacy needs. Using these six variables as a screening device, the discriminant function correctly classified 87% of the subjects as either cases or controls. These data suggest that ingestions are symptoms of familial and, especially of maternal, distress. These healthy, active, but delayed children appear to overwhelm their caregivers. (AJDC 1985;139:456-459)
The number of childhood poisonings reported each year continues to rise but the reported mortality from them seems to be decreasing.1 Current estimates suggest that over 85,000 non-lethal poisonings involving children under 5 years of age occur each year in the United States.2 Improvements in acute medical care are believed to be associated with favorable survival statistics, but many children are left after ingestions with permanent disabilities. Poisonings are now the sixth most common cause of death in preadolescent children.3
Notwithstanding the relative importance of poisoning in pediatrics, there are few systematic studies of origin, The medical literature focuses on diagnosis and treatment of individual intoxications; the relative absence of systematic study of the psychological and social concomitants of childhood poisonings in the literature may reflect the widely held assumption among child health practitioners that ingestions, like other childhood “accidents,” are chance occurrences, which result from the random intersection of children’s oral explorations, the availability of toxic agents, and temporary lapses in caretaker supervision. Given the relative importance of ingestion in pediatric practice, the present study was undertaken. The investigators decided to examine the hypothesis that ingestions are merely random or chance occurrences by determining which, if any, familial, social, and child developmental variables are associated with ingestion.
Availability of Toxic Agents and Childhood Ingestions
The literature on the family context in which child poisonings occur suggests that the commonsense formulation of accidental cause may not be accurate. Several studies note no differences in the accessibility of environmental hazards and poisons in the homes of children who were victims of poisoning and children in comparison groups.4-6
A study of cases reported to a poison control center observed that 31% of parents of poison victims did not believe that their children could climb, open drawers and other hiding places, or unscrew safety caps.7 The risk of poisoning is described as greatest between 24 and 36 months of age.8-10 Developmentally, children in this age group are described as mobile, with interests in exploring their environment and new substances through action, taste, and gaining independence from adults. Children who ingest poisons have been found to demonstrate impulsive and fearless behavioral qualities and exaggerated oral tendencies.5,9,10 Such children have been sometimes described as hyperactive, and may suffer from marked behavior problems, stubbornness, and temper tantrums.188.8.131.52
Pica has been found to be more frequent in victims of ingestion than in comparison populations. 5,10 Baltrop13 has defined pica as putting nonedible objects into the mouth beyond the age of 19 months, Children were found in one study to sort into two groups: (1) toddlers with normal exploratory behavior, with ingestion of poisons who would taste an attractive liquid or tablet but who, in most cases, would reject it before swallowing the substance; and (2) a second group of about half the children under study who appeared to demonstrate exaggerated oral traits.
A greater prevalence of family problems has been observed in homes of ingestion victims in comparison to other children. Sobel8 found evidence of increased parental, familial, and environmental stress in the homes of poisoning victims. A longitudinal study by Margolis12 found an increased prevalence of marital conflict, marital separation, adult physical and mental illness, and personal losses in the families of victims of ingestions. Over time, there continued to be greater numbers of stressful events in the homes of ingestion victims in comparison to the control group. Families of children with more than one ingestion appeared to live in more stressful environments than children who were reported to have only a single ingestion. Families of children with ingestions were found in another study to have greater levels of psychopathologic disorders in mothers, accompanying conflicts among family members, and more frequent physical illnesses than a comparison group.8
The finding of a greater prevalence of physical and psychological illnesses in the families of poisoned children explains the propensity of children to ingest adult medication.8,14,15 Children in the age group 2 to 3 years are quite imitative, and parental pill-taking may provide an interesting model. Since the introduction of “childproof” containers for aspirin, tranquilizers, and psychoactive drugs are responsible for a greater proportion of childhood ingestions, their prevalence in the homes of troubled individuals and stressed families is not surprising: that their children take them need not be seen as wholly “accidental.”
Many of the studies in the literature suffer from methodologic difficulties that the present report endeavored to correct. Previous flaws include the following: (1) use of subjective measures of child and adult functioning: (2) insufficient matching of cases and controls, leading to possible confounding on such important variables as age of the child and social class of the family: (3) focus on a single variable, for example, the developmental status of the child in isolation from other variables that are highly associated with the variable under study, such as parental relationship and the life context of the family: (4) the life context of the family: and (5) a dilution of the sample with varying levels of clinical severity, leading to limited generalizability of the clinical findings. All of our ingestion victims were hospitalized and our findings therefore should be related to severe ingestions only. This uniformity of clinical severity may explain, in part, the striking statistical significance of our findings.
SUBJECTS AND METHODS
After obtaining appropriate informed consent, 23 hospitalized ingestion victims and their families were matched with 23 control subjects on three variables: age of child, race of child, and socioeconomic status of family (using the Hollingshead Two-Factor Index of Social Position).16 All children were inpatients in a pediatric hospital and were under 4 years of age. Control subjects had been hospitalized for acute illnesses such as pneumonia or meningitis. None of the control subjects had a history of ingestions as determined by review of the medical records. Also, none of the ingestion cases were ingestion repeaters. The sample was ascertained as part of a larger descriptive epidemiologic study of pediatric social illness that included four case groups and individually matched controls. (The case groups included victims of child abuse or neglect, failure to thrive, and accidents, in addition to ingestions.)17
The main instrument used in the study was an indepth precoded maternal interview conducted during the child’s hospitalization. Child development was assessed with a standardized maternal report measure. Included in the interview schedule were questions about the household composition, demographic data, questions about past and current stressful events (eg, recent moves and personal losses), child care arrangements, parental discipline practices, maternal and paternal histories, and parental emotional status. Included in the interview were a standard adaptation of the Carey scale of child temperament,18 and several questions addressing the respondent’s understanding of the child’s developmental progress in relation to other children and the child’s age. The Vineland Social Maturity Scale,19 which measures children’s skills in communication, self-care, locomotion, and independent activity, was administered. Data from the routine physical examination was also ascertained from the hospital record.
Included in the 23 ingestion cases were 12 boys and II girls with a mean age of 26.5 months (range, 14 to 47 months). Sixteen of the subjects were white, six were black, and one was Hispanic.
The mean socioeconomic status of the families in both the ingestion and control groups was 4.3 on the Hollingshead scale. Mean per capita monthly income for the ingestion group was $162.70 (1975 US dollars).
This subsample had the third lowest socioeconomic status of the four case categories in the larger study: only the child abuse and neglect sample had a lower mean rating and a lower per capita income. This was true notwithstanding the presence of fathers in the home in 16 of the 23 ingestion cases: all the fathers present were employed.
The data were analyzed in several steps. First, a stepwise regression analysis was performed on all variables felt to be potentially associated with the ingestion diagnosis. Ten variables explained 78% of the between-groups variance on F tests within the regression analysis. These ten variables along with their statistical significance levels are presented in Table 1.
|Table 1. Significant Variables in Ingestion Regression Analysis*
|Direction of Relationship†
|Significance of ß Coefficient
|Vineland social quotient
|Child’s usual health
|Mother and father violently disagree
|Times a month mother gets away by herself
|Mother has telephone
|Total family advocacy needs
|Physical Punishment in mother’s childhood
|Times a week mother spanks child
|Mother sees relatives enough
*Multiple correlation squared=0.782; multiple correlation=0.884.
±Minus sign indicates one case group lower; plus sign, one case group higher.
†All relations significant below P<.004 by F test.
§All relations significant at below P<.001 by F test.
||All relations significant below p<.012
The first six variables in Table I alone explained 69% of the variance between cases and controls (Table 2). The additional four variables explained only an additional 9% of the variance. Next, a stepwise discriminant function analysis using these variables was performed. The discriminant function analysis uses the variables entered to mathematically predict or classify a subject’s case or control status: the prediction will work only when cases differ significantly from controls on the variables entered, and it gives a useful indication of the most salient attributes that differentiate cases from controls.
|Table 2. Results of Discriminant Function Analysis
|No. of Cases
*Percent of grouped cases correctly classified, 87%
The first six variables above were entered into a stepwise discriminant function and correctly classified 87% of the subjects either as cases or controls. The first five variables alone correctly classified 85% of the subjects (Table 2).
It is important to note that families of children who have suffered an ingestion may be a priori more likely to report chronic distress than families of other hospitalized children based on the fact that an ingestion occurred. The very nature of the injury, ie, self-inflicted, may result in overreporting of stress by the parents of hospitalized children. For example, guilt feelings triggered by a self-perceived lapse of parental vigilance may result in increased parental distress and self-castigation. Given this, these data must be interpreted with some caution in spite of their striking statistical significance.
However, the findings suggest that a combination of family stress, maternal psychologic, and child developmental factors are implicated in childhood ingestions. The study adds further support to the literature that suggests that family distress, not chance, may determine which children are likely to become victims of serious ingestions.
The mothers interviewed in the study reported fewer opportunities to get away and be alone and they noted they had fewer numbers of their extended families available to them than did mothers in the comparison groups. Yet, they claimed to see their relatives enough. They reported having fewer resources available to them and they had a sharply increased need for social services such as psychologic counseling. They reported more frequent physical punishment in their own childhood, a factor that may be associated with lower self-esteem and a sense of powerlessness in their adult lives.20,21 In addition, mothers of victims of ingestions appeared to be coping with children who were physically healthy, but who were less socially mature than their peers. Such healthy, yet more dependent, immature children may overwhelm their caregivers who may, in turn, be unable to attend adequately to their explorations.
Other variables that significantly differentiated ingestion cases from controls but that did not emerge as independent predictors in the regression equation flesh out the portrait of a stressed and depleted mother: these women reported feelings of being depressed and being unable to get going in the morning more frequently than the mothers in the comparison group. Hence, these mothers may have found it more difficult to adequately monitor their children, Initially, psychologic depression is associated with an inability to attend to the needs around one; the child, frustrated with not being given attention and nurturance, also may seek stimulation through oral activity. Indeed, in this study, there was the suggestion that children who were victims of ingestion, were less easily distracted from eating than were matched controls, According to their mothers, 14 of the 23 ingestion cases, as opposed to eight of 23 control children, were “not easily” to “never” distracted from eating.
This study suggests that family factors, hitherto regarded in practice as epiphenomena, are central to an understanding of childhood poisonings that are severe enough to require hospitalization. One must be cautious in generalizing to less severe ingestions from these cases. Failure to acknowledge distress in a child’s family will lead to a symptom-bound therapeutic approach.
Similarly, attempts to prevent ingestions that focus only on securing containers or educating parents about environmental hazards will not be sufficient. Parents who are in distress and feeling depleted themselves may not necessarily be psychologically able to avail themselves of advice and guidance.
These data suggest that severe ingestions requiring hospitalization may be childhood symptoms of familial, and, especially of maternal, distress. These active, healthy, but socially delayed children appear to overwhelm their caregivers. A more enlightened practice must focus on the family context in which the ingestion symptom presents. When confronted with a child who has ingested a poison, pediatricians must inquire into the child’s social and psychologic environment rather than simply focusing on whether his or her home is “childproofed.”
It is the authors’ experience born out by these data that the parents of ingestion victims should be asked about current family functioning. The clinician should empathetically inquire into any recent losses. He should also inquire about the availability of extended family and friends to assist with caregiving. Also, when possible, the child should have formal developmental testing. Based on the social/family history and developmental testing, concrete individualized recommendations can then be made that may ameliorate family functioning and caretaker-child interaction. This may be of great assistance in preventing further ingestion episodes.
This work was supported by grant OCD-CB.1 from the Office of Child Development and grant 1 TO MN 15517 from the National Institute of Mental Health, Department of Health and Human Services, Washington, DC.
Jessica H. Daniel, PhD, Carolyn Moore Newberger, EdD, Milton Kotelchuck, PhD, Thomas J. Marx, EdD, and Timothy Schuettge, MSW, contributed to this work.
1. Rudolph AM, Barnett HL, Einhorn AH (eds): Childhood Poisonings in Pediatrics. New York, Appleton-Century-Crofts, 1977, pp 778-787.
2. Chafee-Bahamon C, Lovejoy FH Jr: The effectiveness of a regional poison center in reducing excess emergency room visits for children’s poisonings. Pediatrics 1983;72:164-169.
3. Accident Facts. Chicago, National Safety Council, 1982.
4. Sobel R: Traditional safety measures and accidental poisoning in childhood. Pediatrics 1969;44:811-816.
5. Baltimore CL, Meyer RJ: A study of storage, child behavioral traits, and mother’s knowledge in 52 poisoned families and 52 comparison families. Pediatrics 1969;44:816-820.
6. Sobel R, Margolis JA: Repetitive poisoning in children: A psychological study. Pediatrics 1966;35:641-651.
7. Jensen GO, Wilson W: Preventive implications of a study of 100 poisonings of children. Pediatrics 1980;65:490-496.
8. Sobel R: The psychiatric implications of accidental poisoning in childhood. Pediatr Clin North Am 1970;17:653-685.
9. Craig JO: Oral factors in accidental poisoning. Arch Dis Child 1965;30:419.
10. Sibert JR, Newcombe RG: Accidental ingestion of poisons and child personality. Postgrad Med J 1977;53:254-256.
11. Stewart MA, Thach B, Freidman CM: Accidental poisoning and the hyperactive child syndrome. Dis Nerv Sys 1970;31:403-407.
12. Margolis J: Psychosocial study of childhood poisonings: A five-year follow-up. Pediatrics 1971;47:439-444.
13. Baltrop D: The prevalence of pica. AJDC 1966;112:116-123.
14. Sibert JR: Stress in families of children who have ingested poisons. Br Med J 1975;3: 87-92.
15. Illingworth C: Childhood poisonings: Who is to blame? Practitioner 1974;212:73-76.
16. Hollingshead AB: The Two-Factor Index of Social Position. New Haven, Conn, Yale Press, 1965.
17. Newberger EH, Reed RB, Daniel JH, et al: Pediatric social illness: Toward an etiologic classification. Pediatrics 1977;60:178-185.
18. Carey WB, McDevitt SC: A revision of the infant temperament scale. Pediatrics 1978;61: 735-739.
19. Doll EA: Vineland Social Maturity Scale. Circle Pines, Minn, American Guidance Service, 1965.
20. Herzberger SD: A social cognitive approach to the cross-generational transmission of abuse. Read before the National Conference for Family Violence Researchers, Durham, NH, July 1981.
21. Martin HP, Beezley P: Personality of abused children, in Martin HP, Kempe CH (eds): The Abused Child. Cambridge, Mass, Ballinger Publishing Co, 1976, pp 105-111.
From the Departments of Pediatrics (Drs Bithoney and Newberger) and Psychology (Dr Snyder), Harvard Medical School, Boston, and the Department of Nursing, Children’s Hospital, Boston (Ms Michalek).