Objective. To study the effects of a critical illness and hospitalization of a newborn on family functioning and child behavior during the subsequent 12 years.
Methods. With the use of a randomized stratified cluster sampling, a follow-up of 1443 pregnant women was started at early pregnancy. These pregnancies resulted in 1294 deliveries. A total of 170 infants were admitted to neonatal units and were classified according to their medical risk, and 1112 healthy-born infants were eligible for the control group. After excluding the children with disabilities at 3 years of age, 134 remained in the risk groups and 952 children remained in the control group. The follow-up examinations were performed at the ages of 3 and 9 months and at 3 and 12 years. The main outcome measures were parents’ subjective well-being and adjustment to the child, family functioning, and child’s behavioral problems.
Results. The families with a critically ill newborn experienced more need for support and maladaptation during the first year after delivery. They reported more child behavior problems at 3 years, but no differences were found at 12 years of age. The families with a hospitalized, low-risk infant coped as well as the controls.
Conclusions. A critical illness of a newborn had long-lasting effects on the family and child behavior. The hospitalization of a newborn with a low medical risk did not have any negative consequences on family or child behavior. This is encouraging for a large group of families that experience early separation from their newborn infant as a result of hospitalization.
A large proportion of all newborns—9.8% of all newborn infants in Finland1—are admitted to a neonatal unit during the newborn period for medical reasons. The hospitalization of a newborn is highly stressful for the parents.2,3 The classic work by Klaus and Kennell4 suggesting harm to the mother-infant relationship by early separation guided the hospital practice toward the current family-centered practice in neonatal care. Although the practice has changed dramatically, the hospitalization is still disruptive to the close proximity between the mother and the infant. In addition to the separation, a medical risk for the infant is inherent to an admission. A life-threatening situation of a newborn, per se, can cause a long-lasting parental reaction known as “vulnerable child syndrome,” introduced by Green and Solnit.5 The results of recent studies support these theories. Postpartum thoughts and behaviors of mothers were different in a group of very low birth weight infants with a danger of death and prolonged hospitalization compared with a group of preterm infants who did not require intensive care.6 Many studies of preterm infants have reported increased amounts of later behavioral problems and psychosocial complaints,7 as well as disturbances in parent-infant relationships8,9 and family functioning.10,11
Our prospective longitudinal study was designed to examine the long-term effects of high medical risk and hospitalization of a newborn on the family functioning and child behavior during family-centered care in the neonatal units. Our hypothesis was that a critical illness of a newborn but not hospitalization, per se, has harmful effects on subsequent family dynamics and later behavior of the child.
We followed a large representative sample of families with their first-born children for a period of 12 years. Hospitalized newborns were a part of the Finnish Family Competence Study. Using a stratified randomized cluster sampling procedure, we collected a representative sample of the population in the provinces of Turku and Pori in southwestern Finland with the population totaling 700 000. Eleven health authority areas of the total 35 were weighted according to their degree of urbanization and randomly chosen for the study.12–14 The study covered virtually all women with who had not previously had a delivery and who visited a maternity health care clinic at the beginning of the pregnancy during 1986. In Finland, >98% of pregnant women visit the public maternity health care clinics with an average of 17 visits per pregnancy starting on an average of the 10th week of gestation.1,15 Only first-time mothers were included in the study to eliminate confounding factors of experiences of pregnancy, delivery, and child rearing.
Midwives and primary care nurses at 67 public maternity health care clinics suggested study participation to 1582 successive mothers. Of these, 1443 (91.2%) gave their informed consent and 139 (8.8%) refused to participate. There were 1294 deliveries in this study group. Of these newborns, 170 were admitted to neonatal units for medical reasons. They were divided into low-risk and high-risk infants according to their medical risk when they were discharged from the hospital. The control group consisted of a total of 1112 families with a healthy-born child. As the purpose was to study the effect of early hospitalization on the family and the child behavior but not the effect of a long-term illness of the child, only children who were judged to be healthy at the age of 3 years were included in the final analysis. “Healthy at 3 years” was defined as a child who had no major long-term illnesses and was considered healthy and developmentally normal at the 3-year visit at the well-infant clinic. The 3-year visit included both a physical and a neurologic examination by a physician and a psychological evaluation of the child by a nurse. Infections and other acute illnesses were recorded and analyzed as possible confounding factors. In addition, sociodemographic variables were obtained as potential confounding factors. Thus, we could focus on 1) the effects of early hospitalization of the newborn on later adjustment of the family and child behavior and 2) the effect of a critical illness of the newborn on the family and the child behavior.
The medical risk categories were assigned to the 170 infants at discharge from the hospital as used at the University Hospital of Turku since 1981 (Table 1). Infants who needed observation as a result of minor problems and were at low risk for later consequences (eg, mild breathing problems, hyperbilirubinemia) were classified to the low-risk group (n = 93). The high-risk group (n = 77) were more severely ill newborns (eg, birth weight <2000 g, grade 3 or 4 intraventricular hemorrhage, severe breathing problems).
|At Least Once a Week||At Least Once a Month||Less Often||Hardly Ever|
|Pain in wrist, elbow joints, or Shoulder joints||1||2||3||4|
|Back pain, backache||1||2||3||4|
|Swelling in legs or feet||1||2||3||4|
|Abdominal pains (excluding menstrual pains)||1||2||3||4|
Parents completed questionnaires when the child was 3 and 9 months of age and again when the child was 3 and 12 years of age. The Finnish Family Competence Study questionnaires are shown in detail in the Appendix. The validation process of the questionnaires was performed in 2 steps. First, the questionnaires were tested in a pilot group and after that, semistructured interviews were performed with comparable questions in another group of families. The questions focused on parents’ subjective well-being and adaptation to the child, parents’ social activity, father’s role in child care, problems in child rearing, and children’s behavioral problems. The Achenbach Child Behavior Checklist16,17 was used at 3 and 12 years of age. McMaster’s Family Assessment Scale18 was completed by the parents when the child was 12 years of age.
The differences between the 3 risk groups were statistically tested with Pearson’s χ2 test for categorical variables and with Kruskal-Wallis and Mann-Whitney U tests for continuous variables except for the Achenbach Behavior Checklist scores and McMaster’s family Assessment scores, which were tested using the analysis of variance. P < .05 was interpreted as statistically significant. Statistical analysis was performed with the BMDP program package (Dixon WJ, ed. BMDP Statistical Software Manual. Berkeley, CA: University of California Press; 1992). The study design was approved by the Ethics Committee of the University of Turku Medical School.
Group Characteristics and Background Factors
A total of 36 (21%) of the 170 hospitalized children were excluded because of congenital malformations, a major long-term disease, or a developmental delay at the age of 3 years, and 30 children (18%) were lost to follow-up. A total of 104 children remained in the risk groups: 40 children in the high-risk group and 64 children in the low-risk group. A total of 160 (14%) of the 1112 healthy-born children were excluded because of any long-term disease or a neurologic delay at the age of 3 years, and 176 (16%) were lost to follow-up. A total of 776 children who were healthy both at birth and at the age of 3 years remained in the control group at 3 years of age. The number of families (McMaster’s Family Assessment Scale) studied at 12 years was 27 in the high-risk group, 44 in the low-risk group, and 575 in the control group. The numbers of children were 27 (68%), 46 (72%), and 599 (77%), respectively, at 12 years of age.
The mean length of stay in a neonatal intensive care unit was 11.7 days in the high-risk group (standard deviation: 8.1; range: 1–36 days) and 4.5 days in the low-risk group (standard deviation: 4.5; range: 1–23 days). No statistically significant differences were found in sociodemographic factors in the high-risk group, in the low-risk group, and in the control group (Table 2). The sociodemographic background of the study groups corresponds to the national statistics.14
Upper respiratory infections and other minor illnesses occurred with equal frequency in all groups. The median frequency of otitis media before the age of 3 years was 4 in the high-risk group, 2 in the low-risk group, and 2 in the control group (P = .052). The children in the high- and low-risk groups had received antibiotics more frequently (a median of 5 courses) than in the control group (median: 3; P = .014). A diagnosis of asthma was made in 2.5% of the children in the high-risk group, 1.6% in the low-risk group, and 1.3% of the control children.
Parents’ Subjective Well-Being and Adaptation to the Child
Three months after the delivery, 29% of the fathers in the high-risk group had adapted well, 40% of the fathers in the low-risk group had adapted well, and 45% of the fathers in the control group had adapted well (P = .043). According to the fathers’ opinion, 75% of the families in the high-risk group needed extra help compared with 68% in the low-risk group and 57% in the control group (P = .039). Of the mothers, 28% in the high-risk group reported that they had adapted well to the birth of the child, compared with 47% of the mothers in the low-risk group and 33% of the mothers in the control group. In the mothers’ opinion, no group differences were found in the need of extra support and reassurance (26%, 20%, and 20% in the high-risk, low-risk, and control groups, respectively) in questions of feelings of uncertainty (36%, 44%, and 40%, respectively), mood (variable in 31%, 18%, and 17%, respectively), or the amount of energy (lack of energy in 18%, 15%, and 16%, respectively).
When the infant was 9 months of age, no differences were found in the stress symptoms perceived by either parent. A total of 11% of the mothers in the high-risk group belonged to the highest stress category compared with 18% of the mothers in the low-risk and control groups. Respectively, 29%, 13%, and 12% of the fathers belonged to the highest stress category. It is interesting that the fathers in the high-risk group reported more frequently (17%) that they were “energetic” compared with the fathers in the low-risk group (4%) and in the control group (6%; P = .021).
Social Contacts of the Family
The mothers in the high-risk group were more often willing (18%) to ask for help from their parents-in-law compared with the mothers in the low-risk group (10%) and in the control group (6%) when the infant was 3 months of age (P = .012). No differences between the groups were found in the fathers’ responses. The mothers in the high-risk group reported more frequently (49%) a need for a support group for mothers of newborn infants than the mothers of the low-risk group (22%) and in the control group (37%; P = .016). Thus, these more detailed questions showed differences in the mothers’ need of support, which was not seen in the more general question on need of support and reassurance (see above). One third of the mothers and fathers in all groups were willing to ask for help from their own parents. Offered help from friends was accepted by one half of the mothers, help from a child health care nurse by one third, and help from some other professional person by one fourth of the families. A need for an occasional infant sitter was reported by one third of all families.
When the infant was 3 months of age, the fathers expressed that they had too few contacts with their friends in 26% of the high-risk group, 9% of the low-risk group, and 19% of the control group. When the infant was 9 months of age, 27%, 22%, and 25% of the fathers, respectively, were wishing for more frequent contacts with their friends. Fathers reported that they wished their friends would visit them more often at home in 15%, 9%, and 9% of the families in the high-risk, low-risk, and control groups, respectively, when the child was 3 months of age. When the child was 9 months of age, more visits of the friends were wished by 25%, 10%, and 15% of the fathers, respectively (P = .008). The mothers wanted more visits by their friends in 22%, 13%, and 9% in the high-risk, low-risk, and control groups at 3 months and in 10%, 14%, and 14% at 9 months, respectively.
When the child was 3 years of age, the mothers of the children in the high-risk group less often had a hobby of their own (57%) compared with the mothers in the control group (73%; P = .046). The parents had a shared hobby in 21%, 32%, and 31%, respectively. Only 7% of the mothers of the high-risk children reported that they had enough leisure time compared with 21% of the mothers both in the low-risk group and in the control group.
Father’s Role in Child Care
There were no differences in the amount the fathers participated in the caregiving of the child or in the frequency they used their possibility to take a paternity leave. One fourth of the fathers used >4 hours daily in child caregiving when the child was 3 months and 9 months of age; no differences were found between the groups. When the child was 3 months of age, the mothers of the high-risk group more often reported that daily tasks in the family were shared as they were before the birth of the child (70% in the high-risk group vs 52% in the control group; P = .029).
At 9 months, no differences between the groups were found in the frequency the child had been left with a infant sitter. Parents reported no differences in child behavior at separation situations.
At 3 years of age, the children in the high-risk group were more often reported to have problems in falling asleep compared with the children in the low-risk group or in the control group. The mothers reported problems in falling asleep in 80% of the children in the high-risk group compared with 42% in the low-risk group and 54% in the control group (P = .004). The mothers also reported daily temper tantrums in 67% of the children in the high-risk group, in 40% in the low-risk group, and in 47% in the control group (P = .057). When the high-risk group was tested against the control group, the difference became statistically significant (P = .034; Table 3). The mothers used discussions to solve behavioral problems in 64% of the high-risk group families, in 72% of the low-risk group families, and in 74% of the control families. The fathers reported daily problems in the behavior of the 3-year-old child in 42% in the high-risk group, in 32% in the low-risk group, and in 29% in the control group.
Achenbach Child Behavior Checklist at 3 Years of Age
The total score of the Achenbach Child Behavior Checklist was comparable between the groups at 3 years of age. However, there were significantly higher scores for sleep problems in the high-risk group compared with the other groups (P < .001). Although the scores were lower in the low-risk group compared with the control group, the difference between these groups was not statistically significant. The amount of externalizing symptoms was slightly higher in the high-risk group, but this difference did not reach statistical significance either. The proportion of children above the 90th percentile in the externalizing scores was 9.9%, 14.3%, and 21.6% in the control group, low-risk group, and high-risk group, respectively (P = .054).
Achenbach Behavior Checklist at 12 Years of Age
The differences in the Achenbach Behavior Checklist scores found at 3 years of age had disappeared by 12 years of age. There were no group differences either in the total score or in subscores of the questionnaire.
When the child was 3 years of age, 10.0% of the parents in the high-risk group had divorced compared with 9.4% in the low-risk group and 7.5% in the control group (not significant). When the child was 12 years of age, there were no group differences in the total score of McMaster’s Family Assessment Scale. The proportion of families with the total score above the 90th percentile was comparable between the groups, too.
Our study showed that a critical illness of a newborn affected the family and child behavior for several years even when the medical outcome was good. The finding of later behavioral problems after a life-threatening illness of a newborn is consistent with the concept of “vulnerable child syndrome” introduced by Green and Solnit.5 Although a potential loss of a newborn caused long-lasting consequences, the hospitalization itself seemed not to be harmful during current practice, allowing free access for the families to neonatal units and promoting parent-infant interaction. On the contrary, the families of hospitalized infants with low medical risk coped slightly better than the control families.
Our study was based on a large, representative population sample recruited during early pregnancy. It also extends the time span for the follow-up compared with earlier studies. In the literature, few follow-up studies have examined parental adjustment, family functioning, and child behavior after hospitalization of the newborn.7–11 All of the children were chosen to be first-born to avoid confounding effects as a result of experience in parenting. The sociodemographic background factors were comparable in the study groups. Only the children without major health problems at 3 years of age were included because our study question focused on the impact of early hospitalization.
The fathers seemed to be affected more than the mothers by the critical illness of the newborn. This finding, to our knowledge, has not been reported in others studies. The fathers of critically ill newborns reported maladaptation to the child, and they needed more help and support during the first year of the infant as compared with the low-risk or control groups. After 3 years, the parents of critically ill newborns were more intensively involved in the child care and had less time for themselves compared with the other groups. They reported more sleeping problems and temper tantrums in their children at 3 years of age.
In 1 follow-up study of very low birth weight infants, the psychological distress and parenting stress was shown to be higher in the group with major pulmonary problems compared with those with fewer pulmonary problems and with term infants.7 The researchers, however, did not exclude children with later neurologic disabilities as we did in the current study. Thus, it cannot be concluded whether the amount of distress was increased by current developmental problems or was an impact of more severe medical condition during the neonatal period.
Consistent with our data of the impact of a critical illness of a newborn, Minde et al9 showed that the degree of illness in preterm infants correlated with mothers’ interactional skills up to 2 months after the due date. A postpartum interview study with 3 risk groups consistent with ours6 showed that a potential loss of the infant affected the mother’s attachment during the first postpartum weeks. In a retrospective study about family functioning after preterm birth,10 a longer hospitalization was found to be associated with poorer family role functioning for the mothers still 2 to 4 years after the delivery. However, the study was compromised by a low return rate (34%) of the questionnaires, potentially biasing the results. In a follow-up study by Stjernqvist,11 a group of 20 extremely low birth weight (ELBW) infants, aspects of family functioning were assessed at the end of the first year of life. The mothers reported more physical symptoms in the ELBW group compared with the control group, but no group differences were found in the fathers. More problems were also reported in the marital relationship in the ELBW group. No significant difference were found in parental variables or number of divorces between families of ELBW infants and control families at 4 years.19
That the families in the low-risk group coped well, even better, than the control group is encouraging for a large group of families who experience this separation during early infancy. This information may suggest that hospitalization can provide extra support for the families and help them to adapt to the child. Consistent with our findings in the low-risk group, Wolke et al20 reported fewer night awakenings at the age of 5 months in infants who were born between 32 and 36 weeks of gestation compared with their full-term controls. Similarly, it was found in the studies by Singer et al7 and Trause and Kramer21 that mothers of low-risk preterm infants had less distress than the mothers of healthy control infants. It is noteworthy that the parents of healthy newborns had a significant amount of adaptation problems. The mothers of healthy infants experienced the same amount of problems as the mothers of high-risk infants. The frequency of complaints was consistent with the literature. The birth of the first child gives the new parents a demanding task to adjust to their new roles. In addition, postpartum depression, which occurs in 12% to 17% of women after their first delivery, may contribute to the adaptation problems reported by the mothers. However, our findings raise the question of whether a normal newborn nursery can provide enough support for first-time mothers.
Large multicultural studies have shown that behavioral problems are more commonly found in ELBW infants at school age.22,23 It remains speculative how many of these differences are attributed to differences in parenting as a result of the effects of a critical illness in a newborn period. In our study, the behavioral and family problems had disappeared by 12 years of age. It is interesting that a recent study showed that very low birth weight infants were less often engaged in risk-taking behavior in young adulthood.24 The alteration in parenting as a result of critical illness in a newborn may turn out to be a benefit for the child.
A critical illness of a newborn infant affects the family and the child such that the effects will be seen for years after discharge from the hospital. The differences in family and child behavior, however, disappear by 12 years of life. Hospitalization, per se, seems not to be harmful but, instead, supportive for first-time mothers during current family-centered practice in neonatal units. Close to 10% of all newborns are admitted to a neonatal unit soon after birth. That the separation of the newborn from the close contact to the mother has no apparent consequences on the family and child behavior, is comforting and encouraging for the thousands of parents who experience this distressing situation.
The questions about sociodemographic background, parents’ subjective well-being, adaptation to the child and social contacts, father’s role in child care, and problems in child rearing.
The sociodemographic factors (asked both the mother and the father):
What is your birthday?
What is your school education?
Name your occupation
Are you married/living in a marriage-like relationship/single/divorced (asked the mother only during early pregnancy and when the child was 3 years old)
Parents’ subjective well-being and adaptation to the child
After the delivery at home, the most difficult thing was my own uncertainty as a mother yes/no
Do you feel that you would need general support and encouragement? yes/no
Do you feel that you would need some outside help now when you are at home with the baby? no/some/a lot
Now when your baby is 9 months old and your maternity leave is ending, how would you like to arrange your baby’s day care, if it could be freely selected?
I would take care of her/him by myself
I would like to take a babysitter to our home
Family child care
Child care center
Asked both the mother and the father:
How would you estimate your own perceived condition and mood when the baby is 3/9 months old?
active 1–2–3–4–5 tired
balanced 1–2–3–4–5 changing
satisfied 1–2–3–4–5 feeling down
*How have you adapted to the changes that the baby has caused?
badly 1–2–3–4–5 very well
For every parent, stress points (by selecting 1 = at least once a week, a parent was included) were calculated.
*How many cigarettes do you smoke daily? ____ cigarettes
Social contacts of the family (asked both parents)
Who would you turn to should you need extra help?
to my parents yes/no
to my parents-in-law yes/no
to my own friends yes/no
to the public health nurse yes/no
to another professional yes/no
to a personal support person yes/no
to a group of mothers with newborn babies at the well-baby clinic yes/no
I would not seek any extra help yes/no
Do you feel that you would need occasional help in caring for the baby now when your baby is 3 months old? yes/no
Do you feel that you would need a babysitter for some evenings? yes/no
How many times a month do you have guests at your home now when your baby is 3 months 9 months old?
Do you feel that you have guests more than before the baby/as much as earlier/less than before the baby?
Do you feel that you have guests too often/just right/too seldom
Have you already visited somewhere with your baby when she/he is now 3 months old? not yet/sometimes/several times
Have you been out (outside home) together with your spouse? not yet/sometimes/several times
Do you see your friends nowadays? too frequently/just right/too seldom
Have you already returned to the leisure time activities you had before the baby? not yet/partly/my leisure time activities are as before the baby
Have you the same friends as before the baby? friends are the same as before the baby/some friends have changed/I have other friends now compared to time before the baby
*Do you have your own leisure time activities when your child is 9 months/3 years old? yes/no
*Do you have shared leisure time activities with your spouse? yes/no
*How do you estimate the adequacy of your own leisure time? I have no leisure time/I have leisure time, but not enough/I have enough leisure time
Father’s role in child care
After the delivery, have you changed the way you share the work or tasks within the family? (asked when the infant was 3 and 9 months old) practically not at all/yes, a little/yes, quite a lot
Will you use your right to have paternity leave? (asked fathers) yes/no
Do you participate 1 = daily, 2 = at weekends, 3 = weekly, 4 = seldom (asked fathers)
taking the child out
bathing the child
feeding the child
changing her/his diapers
playing with the child
For every father, participating points (the fathers selecting 1 = daily were included) were calculated.
How many hours a day do you spend caring for your child now when she/he is 3/9 months old?
Children’s behavioral problems
Has your infant been with a babysitter (who is not a family member) more than 1 hour? not yet/once or twice/frequently
How does your child behave in a separation situation? He/she does not cry practically at all/cries a little/cries a lot
How would you describe your own behavior when your child has his/her temper tantrums? (questions to both the mothers and the fathers):
I try to discuss with him/her
I punish him/her physically
Has your 3-year-old child difficulties with falling asleep? no/sometimes/frequently
Has she/he temper tantrums? no/sometimes/practically daily
How do you find the upbringing of a 3-year-old child? (questions to the father)
1 reasoning is well understood and obeyed
2 sometimes there are situations when talking is not obeyed
3 daily there are situations when talking is not obeyed
1 consistency has succeeded in practice
2 consistency mostly works, but sometimes you have to give in
3 consistency does not succeed in practice
1 bringing up a 3-year-old child has been quite easy
2 sometimes there are problems that you are not able to deal with at once
3 there are often situations that puzzle you
We thank the pediatricians who helped review the hospital charts of the patients: Minna Aromaa, MD; Maarit Haapalehto, MD; Olli Honkinen, MD; Martti Janas, MD; Ritva Sihto, MD; and Elina Vähä-Eskeli, MD. We also thank Olli Kaleva, BSc, for the statistical analysis and Inger Vaihinen for secretarial help.
- ↵Stakes/Stakes Information, Welfare and Health Care Statistics, The Medical Birth Register year 2000, Finland. Available at: www.stakes.fi/stakestieto/pdf/2001/tp16.pdf
- ↵Miles MS, Funk SG, Kasper MA. The neonatal intensive care unit environment: sources of stress for parents. AACN Clin Iss Crit Care Nurs.1991;2 :346– 354
- ↵Green M, Solnit AJ. Reactions to the threatened loss of a child: a vulnerable child syndrome. Pediatrics.1964;34 :8– 66
- ↵Stjernqvist KM. Extremely low birth weight infants less than 901 g. Impact on the family during the first year. Scand Soc Med.1992;20 :226– 233
- ↵Rautava P, Sillanpää M. The Finnish Family Competence Study: knowledge of childbirth women seen at maternity health care clinics. J Epidemiol Community Health.1989;43 :253– 260
- ↵Central Statistical Office. Statistical Yearbook of Finland, Volume 82. Helsinki, Finland: Central Statistical Office of Finland; 1987
- ↵Koskinen R, Meriläinen J, Gissler M, Virtanen M. Finnish Perinatal Statistics 1996: Statistical Report. Helsinki, Finland: Stakes Helsinki; 1998
- ↵Achenbach TM. Manual for the Child Behavior Checklist/2–3 and 1992 Profile. Burlington, VT: University of Vermont, Department of Psychiatry; 1992
- ↵Achenbach TM. Manual for the Child Behavior Checklist 4–18 and 1991 Profile. Burlington, VT: University of Vermont, Department of Psychiatry; 1991
- ↵Sternqvist K. The birth of an extremely low birth weight infant (ELBW) <901 g: impact on the family after 1 and 4 years. J Reprod Infant Psychol.1996;14 :243– 264
- Copyright © 2003 by the American Academy of Pediatrics