OBJECTIVE: The purpose was to explore whether poor parental psychological well-being is associated with behavioral problems of very low birth weight (VLBW, ≤1500 g) infants at 3 years of age.
METHODS: In this prospective cohort study, 189 VLBW preterm infants born between January 2001 and December 2006 at the Turku University Hospital, Finland, were followed. Validated questionnaires (Beck Depression Inventory, Parenting Stress Index, and Sense of Coherence Scale) were mailed to the parents when their children were 2 years corrected age. A total of 140 parents evaluated the behavior of the child at 3 years by filling out the Child Behavior Checklist.
RESULTS: There were significant associations between most of the measures of parental symptoms of depression, parenting stress, and sense of coherence and the behavioral outcome of the VLBW infants. The concomitant symptoms of both parents were associated with more problematic child behavior.
CONCLUSIONS: Parents report more behavioral and emotional problems in VLBW children at age 3 if they themselves have had symptoms of depression, parenting stress, or weak sense of coherence 1 year earlier. The new finding of this study was to show the significance of the father’s psychological well-being on the behavioral development of a preterm child.
- BDI —
- Beck Depression Inventory
- CBCL —
- Child Behavior Checklist
- MDI —
- Mental Development Index
- PSI —
- Parenting Stress Index
- SOC —
- sense of coherence
- SOC-13 —
- Sense of Coherence Scale
- VLBW —
- very low birth weight, ≤1500 g
What's Known on This Subject:
Preterm infants are at increased risk of behavioral problems, which has been associated with maternal distress. Paternal psychological well-being is less studied. Parents’ concerns may affect their perceptions or attitudes and have negative effects on the behavior of the child.
What This Study Adds:
Parents report more behavior problems in VLBW children at age 3 if they themselves have had symptoms of depression, parenting stress, or weak sense of coherence. Also, the paternal psychological well-being contributes to the behavioral development of a preterm child.
Preterm infants are at increased risk of behavioral, emotional, and neurologic problems.1–7 Undoubtedly, these problems are multietiological. It has been assumed that behavioral disorders occur together with neurologic and cognitive deficits attributable to prematurity.8,9 Some authors propose that biological and environmental factors together modify behavior through neuromotor and intellectual functioning.9–12 There is previous evidence that the biological effects of prematurity on psychoneurologic development gradually decrease while the role of social environment grows.13–16 For the parents, a premature birth interrupts the normal psychological process of becoming a parent and parents of preterm infants continue to experience distress after the discharge of the child.17,18 Ongoing concerns about the health, development, abilities, and well-being of the child19,20 may affect parental perceptions, attitudes, parent-child interaction, and parenting styles,10,21–23 which may have negative effects on the behavior of the child.7,24,25
Maternal anxiety and depression are associated with behavioral problems of children, according to population-based surveys,26,27 and this also holds true for preterm children.13,28 Maternal depression and parenting stress may modify the mother’s adjustment to her preterm infant, and also to the social and behavioral development of the child.7,13 Children of depressive fathers are at risk for emotional and behavioral problems, as well.29,30 To our knowledge, no previous studies have described how paternal psychological well-being affects the behavioral outcome of children born preterm. In addition, there is a paucity of studies about parental psychological well-being by using the construct of sense of coherence (SOC), which is a measure of how resourceful and capable parents feel in managing their own lives.31
In this study, we examined the association between psychological well-being of both parents (depressive symptoms, parenting stress, and SOC) and behavioral problems of very low birth weight (VLBW) children. We hypothesized that poor psychological well-being of the mothers and the fathers is associated with behavioral problems of VLBW children.
The current study was a part of a prospective cohort study, the PIPARI study, which is a regional multidisciplinary follow-up study on the development and functioning of VLBW preterm infants from infancy to school age. A total of 246 VLBW preterm (birth weight ≤1500 g and gestational age <37 weeks) infants born at the Turku University Hospital, Finland, between May 2001 and December 2006 were invited to participate in the PIPARI study. Eleven families declined participation and 39 infants died during the neonatal period. Of the enrolled 196, 7 were excluded from the current study: 3 infants lived outside the catchment area of the hospital, 1 had multiple anomalies, 1 had osteogenesis imperfecta, and the parents of 2 infants were not able to fill out the questionnaires because of language problems. Thus, the final study population consisted of 189 eligible children and their parents. The characteristics of the study participants are presented in Table 1.
Background data on the neonatal period were prospectively collected from the medical records. Parental background data were obtained from the parents when the child was born, and complementary questions were asked from the parents when the child was 2 years corrected age (Table 1). Parents were asked to fill out validated self-report questionnaires independently concerning parental psychological well-being (depressive symptoms or parenting stress and SOC). At 2 years corrected age, the children were invited to the outpatient visit for medical and neurologic examination and standardized developmental testing (Bayley Scales of Infant Development, II32). Neurodevelopmental impairment of the child was considered to be present if the child exhibited 1 or more of the following factors: cognitive disability (Mental Development Index [MDI] <70 according to Bayley Scales of Infant Development, II32), cerebral palsy (determined during systematic follow-up to 2 years corrected age), severe visual impairment (blindness or visual acuity below 0.3), or severe hearing deficit (hearing loss requiring amplification in at least 1 ear or a hearing impairment with a cutoff of 40 dB). To assess behavioral problems of the VLBW children, the parents were asked to complete and return the questionnaire that was sent home by mail 4 weeks before the child’s 3-year chronological age.
The PIPARI study protocol was approved by the Ethical Committee of the Hospital District of Southwest Finland. Parents gave informed consent after receiving written and oral information.
Parental Psychological Well-Being, 2-Year Questionnaire
Parental Depressive Symptoms
The depressive symptoms of the parents were assessed by using the modified 13-item Finnish translation33 of the original 21-item Beck Depression Inventory (BDI).34 Both parents were asked to rate symptoms or attitudes common among patients with depression (eg, negative self-concept, sadness, and loss of appetite). The items were rated on a 5-point scale (1–5). The scores of the individual items were added together and rescaled to range from 0 to 39. The assessment was used as a continuous measure because we did not want to lose any information or establish any clinical diagnoses.
We used the first 101 items of 120 of the Finnish translation of the Parenting Stress Index (PSI)35 to measure the stress related to parenting of children. The items were rated on a 5-point scale (1–5) and used as a continuous measure. The PSI comprises a child domain and a parent domain. The 6 subscales of the child domain relate to parenting stress attributable to such characteristics of the child that make it difficult for parents to fulfill their role as parents (eg, distractibility/hyperactivity, adaptability, demandingness, and acceptability). The 7 subscales of the parent domain relate to the parents’ own sources of stress (eg, feelings of isolation, role restriction, attachment to the child, or problems with spouse). The combined score for the child and parent domain yields a total stress score; higher scores represent a higher amount of parenting stress. The scoring of the PSI was performed according to the instructions in the PSI manual.35
Parental Sense of Coherence
The Finnish translation of the short Sense of Coherence Scale (SOC-13)31 consists of 13 items rated on a 7-point scale (1–7). The theoretical range of the scale is from 13 to 91 points; lower scores represent a weaker SOC. We used the inventory as a continuous variable, as the SOC concept has been recommended to be examined without cutoff points.31 According to the theory of SOC by Aaron Antonovsky, the SOC stabilizes during adolescence and is considered to be more stable than depression.36 SOC is a readiness to successfully coordinate and take advantage of personal resources. The core concepts of SOC are the experience of comprehensibility, meaningfulness, and manageability of one’s own life.
Behavioral Evaluation of the VLBW Children at Age 3
The parents rated their children when they were 3 years old with the Child Behavior Checklist for Ages 1.5–5 (CBCL/1.5–5).37 This rating focuses on behavioral and emotional problems of children and has been found to reflect similar patterns of psychiatric problems in 23 countries including Finland.38 The parents completed the questionnaire, which resulted in a description of the child’s behavior within the past 2 months. The questionnaire comprises 100 problem items: 99 closed and 1 open-ended problem item scored as 0 = not true, 1 = somewhat or sometimes true, and 3 = very true or often true. Missing items were scored as 0. Questionnaires missing more than 8 items were excluded. The 100 items comprise the total problem score, which ranges from 0 to 200. For detailed information, 36 items are scored on internalizing problems (depression/anxiety, emotional reactivity, somatic complaints with no medical cause, and withdrawal from social contacts) and 24 items on externalizing problems (attention problems and aggressive behavior). We used the assessment as a continuous measure.
The data were analyzed by using SAS for Windows, version 9.2 (SAS Institute, Cary, NC). P values <.05 were considered statistically significant.
The CBCL scores of singletons and multiples were compared by using mixed model with family as the random effect. One-way analysis of variance was used for examining univariate associations between other categorical independent variables (eg, maternal education and child neurodevelopmental impairment) and continuous outcome variables. Linear regression analysis was used to examine univariate associations between parental psychological well-being variables and CBCL scores. The association between continuous predictors (eg, child MDI and gestational age) and CBCL total score was studied by using the Pearson correlation coefficient. The associations between measures of parental psychological well-being and CBCL scores were further studied by using the CBCL total score as an outcome variable in multiple regression models controlling for child MDI and maternal education. In addition, the scores defining the psychological well-being of mothers and fathers were used simultaneously as independent variables of CBCL to assess the total proportion of variance explained by psychological well-being.
For analysis of nonparticipation, 2 nominal variables listed in Table 1 were compared with the χ2 test or Fisher’s exact test, as appropriate. Associations between parental education and dropout status were studied using the χ2 test for trend. The t test for independent samples was used for comparison of continuous variables between responders and nonresponders.
A total of 140 parents (74.1%) completed the CBCL questionnaire: 62.9% by mothers, 5.7% by fathers, and 30.0% by both parents; in 1.4% there was no information about the responder. All items in the questionnaire were completed in 88% of the questionnaires, 8% had 1 missing item, and 4% had from 2 to 6 missing items. Missing items were scored as 0. No questionnaires were rejected because of missing items.
The background characteristics (see Table 1) of the responders and nonresponders were compared. The 49 parents who did not return the CBCL questionnaire were more often divorced than those who returned the questionnaire (16% vs 4%, P = .01) and smoked more often (mothers 34% vs 16%, P = .01; fathers 45% vs 28%, P = .04).
Behavioral Outcome of VLBW Infants
The CBCL scores of the VLBW children are shown in Table 2. The differences between the genders in terms of their CBCL total problem score and internalizing or externalizing problem mean scores were not significant (P = .47, P = .10, P = .47, respectively). For subsequent analyses, the genders were considered as 1 group. The behavioral problems did not differ between the multiples and the singletons (total problems P = .67, internalizing problems P = .97, externalizing problems P = .73).
Parental Psychological Well-Being and Behavioral Outcome of VLBW Infants
All associations between parental variables and CBCL scores were statistically significant, except the association between paternal depressive symptoms and the child’s externalizing problems (Table 3). The evidence of poorer psychological well-being (higher scores in BDI and PSI, and lower scores in SOC-13) was associated with more frequent behavioral problems in their VLBW child. To examine the possibility that the associations shown in Table 3 were influenced by any of the background characteristics, the data were reanalyzed to control for neonatal and parental background data. The mean CBCL total problem score was significantly higher for children with lower MDI (r = −0.22, P = .01) or who had a less educated mother (12.8 points higher in the children of the least educated mothers than in the most educated mothers [95% confidence interval 2.7–22.9, P = .02]). There were no other significant associations between the background data and the CBCL scores.
Table 3 shows that after adjustment for child MDI and maternal education, all the associations between parental variables and CBCL total score remained statistically significant, whereas the association between SOC of the father and externalizing problems of the child became nonsignificant.
Maternal and paternal depression, parenting stress, and SOC were also used concomitantly as predictors of child behavior problems. The total proportions of variance explained by parental psychological well-being were statistically significant (Table 4). Maternal and paternal mean scores for BDI, PSI, and SOC-13 have been published earlier.39 The parents of multiples or children who were small for gestational age did not report poorer psychological well-being than the other parents (data not shown).
Poor parental psychological well-being included depressive symptoms, parenting stress, and weak SOC. The poorer these measures were for the mother, the father, or both parents, the more the VLBW infant developed behavioral and emotional problems at age 3 as reported by their parents. Regarding mothers, our findings are supported by a previous study of Miceli et al,13 who showed that maternal depressive symptoms and parenting stress are related to more internalizing and externalizing behaviors in 3-year-old children born preterm. The current study adds to this observation: the significance of the father’s psychological well-being on the behavioral development of children born preterm; however, externalizing behavioral problems seemed to be more strongly associated with the psychological well-being of the mothers than of the fathers. As the father-infant interaction involves more playing and stimulation than other forms of caretaking,40,41 fathers might find it easier than mothers to tolerate the externalizing behavior of their preterm children.
There were no differences in the mean CBCL scores between the genders of the VLBW children. This agrees with the findings of Kaaresen et al.18 The mean scores for total, internalizing, and externalizing problems (28.5, 5.8, and 12.5, respectively) were similar to the scores in a nonselected Finnish population sample of 3-year-olds (30.4, 6.0, and 12.5, respectively).42 It is interesting that the present cohort of VLBW infants does not express more behavioral problems than their full-term peers. This is an encouraging finding for the parents of preterm infants, and it is in line with other findings of the PIPARI study, which shows positive developmental outcomes, including cognitive development, in VLBW infants born in the 2000s.43,44
The Finnish VLBW population may not be representative of other VLBW populations. Even though there have been only a few studies reporting outcomes of preterm infants born since 2000,25,45,46 the number of children scoring MDI <70 in the current study is lower than in other preterm populations. The inclusion criteria of these studies are not fully comparable, however. Other possible reasons for better cognitive outcome in our population may be variability in treatment strategies or sociodemographic and environmental factors. A Norwegian study18 reported CBCL scores similar to our study for infants born from 1999 to 2002 with a birth weight <2000 g at 2 years corrected age.
A limitation of this study is that we relied only on parental ratings of the behavioral and emotional problems of their children. Although in studies of preschool children the agreement between parents and kindergarten teachers has been shown to be moderate,47,48 parents and teachers do not differ significantly in their ratings of internalizing or externalizing behavior problems of VLBW and/or very preterm infants.49 Maternal psychopathology affects the ratings of the behavior of preschool children,50 and even subclinical depressive symptoms of parents are linked to increased levels of internalizing and externalizing behavior problems in children aged 3 to 5 years.51 Therefore, we cannot exclude that parents with poorer psychological well-being may, in fact, perceive and report their child’s behavior more negatively than parents with better well-being . Nevertheless, the distress of the parents may also influence their childrearing, and the domestic environment may interfere with the parents’ ability to remain sensitive and accepting of the child. Irrespective of these potentially conflicting views, parents are experts in evaluating the behavior of their child.
A strength of this study is that our data were based on a cohort study of virtually all VLBW infants from a given geographical area and time period. The study methods were standardized, internationally used, and validated. Furthermore, we analyzed maternal and paternal associations on child behavior separately and collected paternal data directly from the fathers. The response rate was satisfactory and analysis of the nonresponders was performed accurately. The nonresponders were more often divorced than the responders, a condition known to be associated with more problematic behavior in the children.52,53 Including the responses from divorced parents may have strengthened our findings even further. Also, many other factors (eg, poor physical health of the child) may contribute to the behavioral outcome of preterm children.54 In this study, we were able to take into consideration cerebral palsy and severe hearing and visual impairments documented at 2 years corrected age, but there was no association between these impairments and the behavioral problems.
Parents report more behavior and emotional problems in VLBW children at age 3 if they themselves have had symptoms of depression, parenting stress, or weak sense of coherence 1 year earlier. This study shows that not only maternal but also paternal psychological well-being is a significant contributor of the behavioral development of preterm children. Thus, adequate and timely psychosocial support to both parents might prevent, or at least reduce, behavioral problems in at-risk infants.18
The PIPARI study group consists of Mikael Ekblad, BM; Satu Ekblad, RN; Eeva Ekholm, MD, PhD; Leena Haataja, MD, Prof; Mira Huhtala, MD; Pentti Kero, MD, PhD; Riikka Korja, PhD; Harry Kujari, MD; Helena Lapinleimu, MD, PhD; Liisa Lehtonen, MD, PhD; Marika Leppänen, MD; Annika Lind, PhD; Hanna Manninen, MD; Jonna Maunu, MD, PhD; Jaakko Matomäki, MSc; Petriina Munck, MA; Pekka Niemi, Prof; Pertti Palo, MD, PhD; Riitta Parkkola, MD, PhD; Jorma Piha, MD, Prof; Liisi Rautava, MD, PhD; Päivi Rautava, MD, Prof; Hellevi Rikalainen, MD, PhD; Katriina Saarinen, Physiotherapist; Elina Savonlahti, MD; Matti Sillanpää, MD, Prof emer; Suvi Stolt, PhD; Päivi Tuomikoski-Koiranen, RN; Milla Ylijoki, MD, PhD; Tuula Äärimaa, MD, Prof emerita.
Special thanks to Jaakko Matomäki for help and advice in data analyses.
- Accepted November 30, 2011.
- Address correspondence to Mira Huhtala, MD, Department of Pediatrics/PIPARI, Turku University Hospital, Kiinamyllynkatu 4-8, 20520 Turku, Finland. E-mail:
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: Supported by a grant from the South-West Finnish Fund of Neonatal Research. The funding source had no role in study design; the collection, analysis, or interpretation of data; the writing of the article; or the decision to submit it for publication.
- Hack M,
- Youngstrom EA,
- Cartar L,
- et al
- Halpern LF,
- Brand KL,
- Malone AF
- Breslau N
- Levy-Shiff R,
- Einat G,
- Mogilner MB,
- Lerman M,
- Krikler R
- Miceli PJ,
- Goeke-Morey MC,
- Whitman TL,
- Kolberg KS,
- Miller-Loncar C,
- White RD
- Elgen I,
- Sommerfelt K,
- Markestad T
- McCormick MC,
- Workman-Daniels K,
- Brooks-Gunn J
- Treyvaud K,
- Anderson VA,
- Howard K,
- et al
- Avan B,
- Richter LM,
- Ramchandani PG,
- Norris SA,
- Stein A
- Antonovsky A
- Bayley N
- Abidin RR
- Achenbach T
- Wilson-Costello D,
- Friedman H,
- Minich N,
- et al
- Maguire CM,
- Walther FJ,
- van Zwieten PHT,
- Le Cessie S,
- Wit JM,
- Veen S
- Aarnoudse-Moens CS,
- Weisglas-Kuperus N,
- van Goudoever JB,
- Oosterlaan J
- Delobel-Ayoub M,
- Kaminski M,
- Marret S,
- et al.,
- EPIPAGE Study Group
- Copyright © 2012 by the American Academy of Pediatrics