SEARCH

SEARCH BY CITATION

Keywords:

  • Anorexia nervosa;
  • bulimia nervosa;
  • growth development;
  • head circumference;
  • offspring

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Objective

To characterise early growth and neurocognitive development in children of mothers with a history of eating disorders (ED).

Design

A longitudinal cohort study.

Setting

Child-care centres in Stockholm, Sweden.

Population

Children born to mothers with previous ED (n = 47) (24 anorexia nervosa, 20 bulimia nervosa, 3 unspecified ED), and controls (n = 65).

Methods

Mean values and standard deviation scores of weight and height from birth to 5 years of age and head circumference up to 18 months of age were compared between groups. Neurocognitive development was studied at the age of 5 years by the validated parent questionnaire Five to Fifteen.

Main outcome measures

Head growth and neurocognitive development.

Results

We previously reported that mothers with a history of ED conceived infants with lower birthweight and head circumference than controls. At 3 months of age, body mass index (BMI) was no longer reduced but mean head circumferences of the children born to mothers with ED were smaller throughout the observation period. Similarly, the longitudinal results of the standard deviation scores of head circumference showed a significant overall group effect with lower levels in both subgroups of ED (anorexia nervosa and bulimia nervosa). The children of the ED mothers also had significantly higher Five to Fifteen scores than controls, reflecting difficulties in language skills. Head circumference at birth correlated with language skills in the children of mothers with ED.

Conclusion

Children of mothers with previous ED demonstrated an early catch-up in BMI, but the average head circumference continued to be delayed until at least 18 months of age. The reduced head growth was related to delayed neurocognitive development.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

There is little information about the consequences of maternal eating disorders (ED) on growth development in the offspring. Previous studies have demonstrated an association between maternal anorexia nervosa (AN) and lower birthweight infants.[1-4] Higher rates of prematurity and small-for-gestational age in infants of women with ED have also been found in some but not all previous studies.[1, 2] Furthermore, dysfunctional patterns of attachment in mothers with ED and feeding problems in their infants are relatively common.[5-9] How this could affect the growth development of these children is less clear.

In the study by Treasure and Russell,[10] seven infants born to mothers with AN were underweight at birth, but caught up so that their weight was normal at 30 weeks of age. In another investigation by Stein et al.,[11] 33 infants of women who exhibited, bulimia nervosa (BN) or an unspecified eating disorder (EDNOS) in the postnatal year were found to have the same length, but to weigh less than control children at 1 year of age. However, at 10 years of age the average body mass index (BMI) was similar.[12] In a small study by Waugh and Bulik,[7] infants whose mothers had a history of AN or BN were lighter and shorter than control infants at the time of birth, but these differences were no longer significant at 3 months of age. In a large investigation by Micali et al.,[13] infants of 9 months of age whose mothers had a self-reported history of BN (n = 194) were overweight in comparison to control children and to those born to women with other psychiatric disorders.

In a longitudinal cohort study, we previously reported significantly lower birthweight in infants of nulliparous nonsmoking women with a history of ED compared with healthy control women from the same living area.[2] A novel finding was smaller head circumference at birth in infants of women with either AN or BN.[2] We hypothesised that early growth and neurocognitive development may be affected in the offspring of mothers with previous ED. In this follow-up study, we now report data on weight, height and BMI until 5 years of age and of head circumference up to 18 months of age of the same children. Neurocognitive development was assessed at the age of 5 years and related to head growth.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The initial study population comprised 49 women with a history of ED (24 AN, 20 BN, 5 EDNOS) and 67 control women.[2] All women were nulliparous, nonsmokers and conceived spontaneously. They were recruited in early pregnancy (gestational week 10) from 13 prenatal clinics in the northwest area of Stockholm and followed throughout pregnancy and up to 3 months postdelivery.[2, 8] The women with a history of ED were initially diagnosed by interview according to the DSM-IV diagnostic criteria[14] and the diagnosis was further confirmed from medical records when available. The mean duration of eating disorders was 9 years (range 3–15 years) and the duration of recovery before study recruitment was 3.2 ± 3.0 (mean ± SD) years. The controls without a history of ED were recruited consecutively from the same prenatal clinics during the same period (August 1997 to June 2001). None of them refused to participate. Table 1 shows pregnancy and neonatal characteristics of the initial study population.[2]

Table 1. Pregnancy and neonatal characteristics (mean ± SD or percentage) in women with a previous history of ED and controls of the initial study population
CharacteristicsED (n = 49)Controls (n = 67)
  1. Significant differences between groups are indicated: *P < 0.05; **P < 0.01; ***P < 0.001.

  2. Data are adapted from Koubaa et al.[2]

Age of the mothers (years)29.3 ± 4.630.0 ± 3.7
Maternal BMI (kg/m2)20.5 ± 3.0***22.3 ± 2.8
Maternal weight gain (kg)11.3 ± 3.912.1 ± 2.6
Weeks of gestation38.9 ± 1.839.2 ± 1.8
Hyperemesis (%)33**9
Gestational hypertension/pre-eclampsia (%)123
Birthweight (kg)3.2 ± 0.6**3.5 ± 0.5
Birth length (cm)49.6 ± 2.750.2 ± 2.4
Head circumference (cm)33.7 ± 1.4***35.2 ± 1.6
Apgar score <6 (at 5 minutes)24
Small for gestational age (%)12*1
Neonatal care (%)104
Malformations (%)20

Early adaptation to motherhood was evaluated 3 months postdelivery and we previously reported that maternal adjustment was clearly impaired and related to postpartum mental problems in the mothers with ED before pregnancy.[8]

The clinical material of the present follow-up study comprised 47 children born to mothers with previous ED (24 AN, 20 BN, 3 EDNOS) and 65 control children. Two children born to mothers with a history of EDNOS were not included because one was lost to follow up and the other child died shortly after birth due to heart malformation. Furthermore, one of the mothers in the control group was missed at follow up and another mother in the same group declined further participation. The local Committee of Medical Ethics approved the research protocol and informed consent was obtained from all mothers.

The primary outcome of this study was early growth and neurocognitive development of the children. Data on weight (±10 g) and length/height (±1 mm) at 3, 6, 12, 18 months and 3, 4 and 5 years of age and head circumference (maximal fronto-occipital circumference; ±1 mm) from birth up to 18 months of age were collected from medical records at the respective child health centres in Stockholm. The anthropometric measurements were performed by specially trained paediatric nurses working at the child-care centres. The weights, heights, BMIs and head circumferences were expressed as standard deviations below or above the mean of a large reference population of Swedish children used in the national paediatric health care for body growth monitoring (standard deviations scores, SDS).[15]

We also examined the relation of children's growth to maternal adjustment as evaluated 3 months after delivery by a subscale of the Maternal adjustment and maternal attitude questionnaire (MAMA).[8, 16] This questionnaire consists of 12 items, uses a four-point scale from 1 = never/not at all to 4 = very often/very much/during the last month.[16] The scores 3 and 4 were considered to reflect less favourable adjustment to motherhood.[8] In this study, we selected the MAMA question number 1 (Have you been worrying that you might not be a good mother? much, very much) and number 12 (Have you enjoyed feeding your baby? (little, not at all) for further studies. The test–retest reliability for the MAMA questionnaire was previously found to be satisfactory (0.84), as well as split-half reliability (0.73).[16] We calculated Cronbach's α for the MAMA and found internal consistency to be satisfactory high (α = 0.86).[8] Furthermore, comparisons with other interview findings and questionnaire data have provided good evidence for criterion validity.[16]

Neurocognitive development of the children was investigated at the age of 5 years using a validated parent questionnaire, the Five to Fifteen (FTF),[17] which was delivered by mail and completed by the mothers. The complete questionnaire consists of 181 items on neurocognitive development divided into eight domains (motor skills, executive functions, perception, memory, language, learning, social skills and emotional/behavioural problems) and their subdomains, but for children who are 5 years old the domain learning is not used. The response alternatives to each statement were: 0 = ‘Does not apply’, 1 = ‘Applies sometimes/to some extent’ or 2 = ‘Definitely applies’. Higher scores reflect difficulties in neurocognitive function, and scoring above the 90th centile of a reference population was considered significant and clinically relevant.[18] The FTF has relatively high internal consistency (Cronbach's α = 0.69–0.94) and an acceptable to excellent inter-rater and test-retest reliability.[17] Intercorrelations between the FTF and other scales (the Child BehavioUr Checklist and NEPSY, a neuropsychological assessment instrument) show substantial overlap supporting the validity of the FTF scales.[18-20]

Statistics

Descriptive statistics for weight, height, BMI and head circumferences of the children are presented with mean and SD. At each time-point a general linear model was employed to compare the ED and controls regarding the growth after adjusting for the mother's BMI, the child's sex and the true age (Table 2).

Table 2. Weight, height and BMI (mean ± SD, respectively) of the children of mothers with a previous history of ED and control children (Controls) from birth to 5 years of age and head circumference (mean ± SD) from birth to 18 months of age
AgeWeight (kg)Length/height (cm)BMI (kg/m2)Head circumference (cm)
EDControlsEDControlsEDControlsEDControls
  1. Significant differences between groups are indicated: *P < 0.05, **P < 0.01, ***P < 0.001.

  2. P-values are adjusted for the mother's BMI, the child's sex and the true age.

At birth

3.2 ± 0.6*

= 47

3.5 ± 0.5

= 64

49.6 ± 2.7

= 47

50.2 ± 2.4

= 64

13.1 ± 1.7*

= 47

13.9 ± 1.4

= 64

33.6 ± 1.4***

= 47

35.1 ± 1.6

= 64

3.1 ± 0.3 months

6.2 ± 0.8

= 47

6.5 ± 0.9

= 65

61.8 ± 2.2

= 47

62.5 ± 2.7

= 65

16.1 ± 1.5

= 47

16.6 ± 1.5

= 65

40.3 ± 1.3**

= 47

41.2 ± 1.2

= 64

6.2 ± 0.4 months

8.1 ± 1.0

= 47

8.4 ± 1.0

= 65

68.0 ± 2.2

= 47

69.0 ± 2.5

= 65

17.5 ± 1.7

= 47

17.5 ± 1.4

= 65

43.3 ± 1.2***

= 47

44.2 ± 1.3

= 65

12.2 ± 0.5 months

10.2 ± 1.2

= 47

10.4 ± 1.2

= 64

75.8 ± 2.4

= 47

76.9 ± 2.6

= 64

17.6 ± 1.5

= 47

17.5 ± 1.2

= 64

46.3 ± 1.4**

= 47

47.2 ± 1.3

= 64

18.2 ± 0.6 months

11.6 ± 1.4

= 47

11.8 ± 1.4

= 65

82.3 ± 2.9

= 47

83.6 ± 2.9

= 65

17.2 ± 1.6

= 47

16.9 ± 1.3

= 65

47.8 ± 1.4**

= 46

48.6 ± 1.4

= 64

3.0 ± 0.1 years

15.6 ± 2.3

= 34

15.7 ± 2.6

= 58

96.2 ± 3.8

= 33

97.3 ± 4.0

= 57

16.7 ± 1.8

= 33

16.6 ± 2.0

= 57

  
4.0 ± 0.1 years

17.8 ± 2.6

= 38

17.9 ± 3.0

= 58

103 ± 4.5

= 38

105 ± 4.1

= 58

16.7 ± 1.6*

= 38

16.2 ± 1.9

= 58

  
5.0 ± 0.0 years

20.1 ± 3.3

= 39

20.0 ± 3.1

= 45

110 ± 5.0

= 39

112 ± 4.0

= 46

16.4 ± 1.9

= 39

16.0 ± 1.8

= 45

  

The longitudinal trend of the SDS for weight, height, BMI and head circumference during different intervals of time was analysed employing the mixed in sas® system 9.1 (SAS Institute Inc., Cary, NC, USA) with a random coefficients model. The relationship with time was tested for nonlinear trends and therefore models using polynomials of time were used. Interactions were tested between time effects and Group (previous history of AN, BN and controls) after controlling for the mother's BMI and the child's sex. Between-groups comparisons were employed at each time-point with a general linear model in which we adjusted for the mother's BMI, the child's sex and the true age. Results are presented as adjusted means and 95% confidence intervals (CI) (Table 3 and Figures 1 and 2).

Table 3. Standard deviation scores of weight and length/height (mean ± 95% confidence intervals, respectively) of the children of mothers with a previous history of AN, BN and control children (Controls) from birth to 5 years of age
AgeWeight (kg)Length/height (cm)
ANBNControlsANBNControls
  1. Significant differences between ED groups (AN or BN) and control children are indicated: *P < 0.05.

  2. Significant differences in comparison with the National reference population are indicated in bold figures.

  3. Data are adjusted for the mother's BMI, the child's sex and the true age.

At birth

−0.66 [−1.26 to −0.06]

= 24

−0.62 [−1.24 to 0.00]

= 20

+0.01 [−0.34 to 0.37]

= 64

−0.39 [−0.86 to 0.08]

= 24

−0.15 [−0.64 to 0.34]

= 20

−0.14 [−0.42 to 0.13]

= 64

3.1 ± 0.3 months

+0.50 [−0.06 to 1.07]

= 24

+0.63 [0.06 to 1.21]

= 20

+0.78 [0.44 to 1.11]

= 65

+0.31 [−0.19 to 0.80]

= 24

+0.79 [0.28 to 1.30]

= 20

+0.60 [0.30 to 0.89]

= 65

6.2 ± 0.4 months

+0.15 [−0.32 to 0.61]

= 24

+0.33 [−0.15 to 0.82]

= 20

+0.37 [0.10 to 0.65]

= 65

+0.18 [−0.25 to 0.61]

= 24

+0.23 [−0.21 to 0.68]

= 20

+0.46 [0.21 to 0.72]

= 65

12.2 ± 0.5 months

−0.19 [−0.63 to 0.26]

= 24

+0.09 [−0.37 to 0.55]

= 20

−0.11 [−0.37 to 0.16]

= 65

−0.02 [−0.42 to 0.46]

= 24

+0.08 [−0.38 to 0.54]

= 20

+0.30 [0.04 to 0.57]

= 65

18.2 ± 0.6 months

−0.23 [−0.70 to 0.23]

= 24

−0.06 [−0.53 to 0.41]

= 20

−0.26 [−0.54 to 0.01]

= 65

−0.07 [−0.52 to 0.38]

= 24

+0.04 [−0.42 to 0.50]

= 20

+0.31 [0.04 to 0.57]

= 65

3.0 ± 0.1 years

+0.38 [−0.25 to 1.00]

= 19

+0.11 [−0.61 to 0.83]

= 13

+0.22 [−0.14 to 0.58]

= 57

+0.11 [−0.45 to 0.67]

= 19

−0.21 [−0.86 to 0.43]

= 13

+0.28 [−0.04 to 0.60]

= 57

4.0 ± 0.1 years

+0.55 [−0.00 to 1.10]

= 21

+0.27 [−0.32 to 0.86]

= 16

+0.30 [0.02−0.63]

= 58

−0.13 [−0.60 to 0.35]

= 21

−0.24 [−0.75 to 0.27]*

= 16

+0.38 [0.10 to 0.66]

= 58

5.0 ± 0.0 years

+0.27 [−0.31 to 0.85]

= 22

+0.25 [−0.41 to 0.92]

= 15

+0.24 [−0.16 to 0.64]

= 43

−0.05 [−0.50 to 0.41]

= 22

−0.27 [−0.80 to 0.25]

= 15

+0.16 [−0.16 to 0.47]

= 43

image

Figure 1. Standard deviation scores of BMI in subgroups of children born to mothers with AN or BN and control children. The values presented are means [95% CI] adjusted for the mother's BMI, the child's sex and the true age. The SDS of BMI in the AN subgroup was significantly lower at birth (< 0.05) but thereafter comparable to the controls, whereas the SDS of BMI in the BN subgroup was comparable to controls at all time points.

Download figure to PowerPoint

image

Figure 2. Standard deviation scores of head circumference in subgroups of children born to mothers with AN or BN and in control children. The values presented are means [95% CI] adjusted for the mother's BMI, the child's sex and the true age. The SDS of head circumference was significantly lower at birth in both subgroups (< 0.001), and at 3 and 6 months of age (< 0.05–0.01) with similar tendency at 12 and 18 months of age in the AN group.

Download figure to PowerPoint

Forward stepwise multiple linear regression analysis was employed to examine the extent to which different increases in growth development between birth and the last time point (5 years or 18 months for head circumference) could be explained by background parameters such as maternal ED; maternal height and BMI at 10 weeks of pregnancy; the presence or absence of hyperemesis; relapsing ED during pregnancy; gestational age and the presence or absence of small-for-gestational age; sex of the infant; growth data at birth and the responses to questions 1 and 12 of the MAMA. The adjusted R2 value is presented.

The FTF-data were analysed using the Kruskal–Wallis analysis of variance by ranks followed by multiple comparisons between groups (AN, BN control) based on ranks. The P values were then corrected according to the Bonferroni procedure (Table 4). The association between head circumference at birth and FTF-data was evaluated with Spearman rank correlation coefficient within the three groups AN, BN and control. Comparison between women with ED and controls concerning the correlation coefficient was carried out using Fisher's Z-transformation. P < 0.05 was considered to be statistically significant.

Table 4. Raw scores of subdomains of neurocognitive function (mean ± 95% CI, respectively) of the children of mothers with a previous history of AN and BN and control children (Controls) at the age of 5 years
Neurocognitive variablesAN (n = 24)BN (n = 19)Controls (n = 62)
  1. Significant differences between ED groups (AN or BN) and control children are indicated: *P < 0.05, **P < 0.01, ***P < 0.001.

  2. P values are adjusted for multiple comparisons.

Motor skills
Gross motor skills0.21 (0.08–0.35)0.14 (0.03–0.26)0.12 (0.06–0.17)
Fine motor skills0.40 (0.24–0.57)**0.28 (0.14–0.44)0.16 (0.10–0.22)
Executive functions
Attention0.40 (0.20–0.58)0.45 (0.22–0.68)0.24 (0.16–0.32)
Hyperactive/Impulsive0.52 (0.31–0.73)0.42 (0.26–0.57)0.41 (0.30–0.52)
Hypoactive0.26 (0.07–0.45)0.21 (0.05–0.37)0.13 (0.06–0.20)
Planning/Organising0.46 (0.30–0.61)0.47 (0.23–0.72)0.28 (0.20–0.37)
Perception
Relation in space0.27 (0.12–0.41)0.19 (0.08–0.30)0.15 (0.08–0.23)
Memory 0.32 (0.17–0.48)*0.23 (0.10–0.37)0.14 (0.07–0.20)
Language
Comprehension0.27 (0.15–0.39)0.29 (0.09–0.50)0.18 (0.12–0.24)
Expressive language skills0.20 (0.10–0.29)*0.23 (0.07–0.40)*0.08 (0.02–0.13)
Communication0.15 (0.03–0.28)0.47 (0.05–0.90)0.15 (0.07–0.23)
Social skills 0.15 (0.08–0.21)**0.12 (0.06–0.17)0.09 (0.03–0.15)
Emotional/Behavioural problems
Internalising0.08 (0.02–0.15)0.06 (0.01–0.11)0.05 (0.01–0.09)

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Growth development

As shown in Table 2, the mean birth weight of the children of mothers with a previous history of ED was lower than in the control group (P values adjusted for the mother's BMI, the child's sex and the true age) although gestational age was similar between the groups (38.9 ± 1.9 versus 39.2 ± 1.8 weeks of gestation, = 0.39). However, from the age of 3 months there was no longer any difference in weight between the groups. Birth length/height to 5 years of age was similar between groups. The average BMI at birth of infants born to mothers with ED was significantly lower than that of the control infants, but not at older ages. In contrast, the children of mothers with a previous history of ED demonstrated a significantly smaller average head circumference compared with the control children throughout the entire observation period (Table 2).

Standard deviation scores of growth

The SDS for weight, length/height and BMI demonstrated significant trends over time but no interaction and no overall group effect could be demonstrated. Table 3 shows the SDS of weight and length/height (adjusted for the mother's BMI, the child's sex and the true age) in subgroups of children born to mothers with AN or BN and control children. Analysis at each time-point revealed a significantly lower height in the BN group at the age of 4 years compared with controls.

Figure 1 illustrates the SDS of BMI (adjusted for the mother's BMI, the child's sex and the true age) in subgroups of children born to mothers with AN or BN and control children. The SDS of BMI in the AN subgroup was significantly lower at birth but thereafter comparable to the controls, whereas the SDS of BMI in the BN subgroup was comparable to controls at all time-points.

The longitudinal results of the SDS for head circumference displayed a significant trend over time but no significant interactions could be demonstrated. However, the overall group effect was clearly significant, < 0.001 showing lower levels of SDS for head circumference in both subgroups of ED compared with controls (AN versus controls and BN versus controls, < 0.001, respectively) (Figure 2). Analysis at each time-point revealed a significantly lower SDS of head circumference (adjusted for the mother's BMI, the child's sex and the true age) at birth in both subgroups, and at 3 and 6 months of age with similar tendency at 12 and 18 months of age in the AN group in comparison to controls.

Predictors of growth development

Dysfunctional answers to MAMA question number 1 (Have you been worrying that you might not be a good mother? much, very much) (< 0.01), as well as small-for-gestational age (< 0.05), correlated significantly with a lower increase in weight from birth to 5 years of age. These two variables explained 10% of the variance (R2) in weight development (< 0.01). Dysfunctional answers to MAMA question number 1 also correlated with a lower increase in length/height (< 0.01). Together with length at birth (< 0.01), as well as the mother's height (< 0.05), these variables explained 31% of the variance (R2) in height development (< 0.001). Furthermore, dysfunctional answers to the same MAMA question correlated with an attenuated increase in head circumference development (< 0.01). This factor together with head circumference at birth (< 0.001) and sex of the child (< 0.05) explained 53% of the variance (R2) in head circumference development (< 0.001).

Neurocognitive function

Ninety-six percent of the mothers completed the questionnaire of neurocognitive development. One mother with previous BN and three mothers in the control group declined to participate. Children of mothers with ED had significantly higher mean raw scores (lower neurocognitive function) for most domains (motor skills 0.27 ± 0.26 versus 0.14 ± 0.18, < 0.01; memory 0.28 ± 0.33 versus 0.14 ± 0.26, < 0.01; language skills 0.23 ± 0.27 versus 0.11 ± 0.20, < 0.05; and social skills 0.12 ± 0.12 versus 0.09 ± 0.22, < 0.01, respectively), than controls. Nineteen of 46 (41%) children of mothers with ED and 7 of 62 (11%) control children had a mean domain score above the 90th centile in one or more domains (< 0.001). A logistic regression (adjusted for the child's sex) showed an odds ratio (OR) of 5.5 (95% CI 2.2–15.8) (< 0.001) for impaired development in at least one domain in children of ED mothers.

As presented in Table 4, children of mothers with AN had significantly higher raw scores of several subdomains such as fine motor skills, memory, expressive language skills and social skills than controls (P-values adjusted for multiple comparisons). Similarly, children of mothers with BN had significantly higher raw scores of expressive language skills. There were no significant differences in raw scores between the subgroups of AN and BN.

In the children of mothers with ED, head circumference at birth correlated significantly with language skills (r = 0.32, < 0.05). In the subgroup of BN, head circumference at birth correlated with planning/organising (r = 0.64, < 0.01), social skills (r = 0.51, < 0.05) and language skills (r = 0.48, < 0.05). Similar correlations were not found in the AN group (r = 0.05, r = 0.20 and = 0.18, respectively) or the control group (r = 0.10, r = 0.09 and r = 0.04, respectively). The correlation coefficients for head circumference versus planning/organising, language skills and social skills were significantly different between the BN and the control group (< 0.01, < 0.05 and < 0.05, respectively). Furthermore, the correlation coefficient for head circumference versus planning/organising was significantly different between the BN and the AN group (< 0.05).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Main findings

In the present study, we found that children born to mothers with ED had a lower birthweight but displayed an early catch-up in weight. In contrast, the average head circumference was found to be delayed up to at least 18 months of age. The finding of reduced head growth has to our knowledge not been reported previously and was found in both the AN and BN subgroups. The children of mothers, who worried much about not being a good mother (MAMA question number 1) 3 months after delivery, exhibited more pronounced delays in growth development including head growth. Furthermore, children of mothers with AN or BN had significantly higher FTF scores than controls reflecting difficulties in language and social skills.

Interpretation

Our present results generally confirm previous findings of an early catch-up in weight/BMI in children born to mothers with ED.[7, 10] The novel finding of a delayed postnatal development of head circumference in children born to mothers with ED could be associated with delayed brain growth.[21] This retarded head growth is probably not due to postnatal undernutrition, as body weight was normalised after only 3 months. Rather, the lower mean weight and head circumference at the time of birth might be caused by fetal undernutrition during pregnancy.[22] In comparison to control infants, our children born to mothers with a previous history of AN exhibited a mean reduction of 300 g in birthweight and of 1.5 cm in head circumference at birth. Reductions similar to those were observed (300 g and 1.0 cm, respectively) during the famine in urban areas of the western Netherlands in 1945.[23]

In contrast to the AN subgroup with both a low weight and a small head circumference at birth, the infants born to mothers with BN had a normal birthweight, but a decreased head circumference at birth. This discrepancy between the subgroups might reflect different underlying causes behind delayed head growth. Monteleone et al.[24] have reported that the plasma levels of neuroactive steroids, including cortisol, in women with ED are higher than in healthy women. Hypothetically, exposure to elevated levels of neurotoxic steroids in utero as a consequence of higher maternal stress could affect brain development.[25] Moreover, poor maternal nutrition attenuates placental activity of 11β-hydroxysteroid dehydrogenase type 2, thereby enhancing maternal cortisol production.[26] In a prospective observational study of preterm singleton infants, exposure to multiple doses of dexamethasone to prevent complications of prematurity was associated with an increased risk of neurodevelopmental abnormalities at 2 years of age.[27] Furthermore, prenatal treatment with low dose of dexamethasone of a fetus at risk of congenital adrenal hyperplasia has shown impaired working memory associated with low self-perceived scholastic competence in the exposed children.[28] Possibly, both intrauterine undernutrition and maternal stress, associated with elevated levels of cortisol, could influence fetal brain development, and have epigenetic effects later in life.[29, 30]

We have no information on blood levels of cortisol in the mothers during pregnancy. However, the finding that maternal worries about not being a good mother predicted slower development of weight, height and head circumference might support the notion that these effects are related to maternal stress. Altogether, the results presented here are compatible with the recently proposed developmental model of possible mechanisms of risk for adverse long-term biological outcomes in the offspring of women with ED.[1]

The children of mothers with ED had higher mean raw scores for most domains of the FTF than controls, and 41% had a score above the 90th centile of a large national reference population in at least one domain. The profiles suggest a general delay of development in the children of mothers with ED and not any specific child psychiatric disorder.[19] There were no significant differences in FTF scores between the ED subgroups, and both the AN and BN subgroups had scores reflecting difficulties in expressive language skills. Head circumference at birth correlated significantly with language skills in the children of mothers with ED. These results are new for this population and indicate that reduced head growth is of clinical importance for neurocognitive development in children born to mothers with ED.

Strengths and weaknesses

The present study has a number of significant strengths, such as an area-based sampling of the ED and control mothers to enhance generalisability; only nulliparous women and nonsmokers were included; the diagnosis of ED was based on interviews; a longitudinal design; the period of the observation was prolonged and included repeated monitoring; there was no attrition up to 18 months of age and little even up to 5 years of age. However, there was sparse information on breastfeeding and eventual feeding problems, and also a lack of information about a possible paternal influence. Furthermore, we have no data on head circumference of the parents. The FTF profiles were based on parent information of the mothers. Although the test has demonstrated high inter-rater and test–retest reliability and validity,[18-20] we cannot exclude reporter bias due to the possibility of worries and feelings of guilt in the mothers with a history of ED.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

In conclusion, children of mothers with a previous history of ED demonstrate an early catch-up in BMI from a decreased value at birth, but the head growth of children born to mothers with either AN or BN was retarded up to at least 18 months of age. Maternal stress and prenatal undernutrition are possible causes of this delayed head growth. Our results also indicate that reduced head growth may be of importance for neurocognitive development in children born to mothers with ED who perceived difficulties in expressive language skills and social skills. Clinicians should be aware of the potential influence of maternal ED on head growth of the offspring to prevent later negative outcomes. There is also an apparent need for further long-term follow-up of neurocognitive development in these children.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The authors thank Elisabeth Berg for statistical advice.

Disclosure of interests

The authors have no conflict of interest with regard to this study.

Contribution to authorship

SK planned the study, collected and analysed the data, interpreted the results and wrote the paper together with ALH. TH interpreted the results and revised the manuscript. LH provided additional results for the study and revised the manuscript. ALH interpreted the results and wrote the paper together with SK.

Details of ethics approval

This study was approved by the Local Committee of Medical Ethics at Karolinska Institutet (No. 04-199/1) and all the mothers gave informed consent.

Funding

This study was supported financially by the Swedish Research Council (20324: ALH), the Stockholm County Council (ALF-funds: SK, ALH) and Karolinska Institutet, Stockholm, Sweden.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  • 1
    Micali N, Treasure J. Biological effects of a maternal ED on pregnancy and foetal development: a review. Eur Eat Disord Rev 2009;17:44854.
  • 2
    Koubaa S, Lindholm C, Hällström T, Hirschberg AL. Pregnancy and neonatal outcomes in women with eating disorders—a prospective controlled study. Obstet Gynecol 2005;105:25560.
  • 3
    Ekéus C, Lindberg L, Lindblad F, Hjern A. Birth outcomes and pregnancy complications in women with a history of anorexia nervosa. BJOG 2006;113:92529.
  • 4
    Wentz E, Gillberg IC, Anckarsäter H, Gillberg C, Råstam M. Reproduction and offspring status 18 years after teenage-onset anorexia nervosa—a controlled community-based Study. Int J Eat Disord 2009;42:48391.
  • 5
    Stein A, Woolley H, Cooper SD, Fairburn CG. An observational study of mothers with eating disorders and their infants. J Child Psychol Psychiatr 1994;35:73348.
  • 6
    Agras S, Hammer L, McNicholas F. A prospective study of the influence of eating-disordered mothers on their children. Int J Eat Disord 1999;25:25362.
  • 7
    Waugh E, Bulik CM. Offspring of women with eating disorders. Int J Eat Disord 1999;25:12333.
  • 8
    Koubaa S, Hällström T, Hirschberg AL. Early maternal adjustment in women with eating disorders. Int J Eat Disord 2008;41:40510.
  • 9
    Reba-Harreleson L, Von Holle A, Hamer RM, Torgersen L, Reichborn-Kjennerud T, Bulik CM. Patterns of maternal feeding and child eating associated with eating disorders in the Norwegian Mother and Child Cohort Study (MoBa). Eat Behav 2010;11:5461.
  • 10
    Treasure JL, Russell GFM. Intrauterine growth and neonatal weight gain in babies of women with anorexia nervosa. BMJ 1988;296:1038.
  • 11
    Stein A, Murray L, Cooper P, Fairburn CG. Infant growth in the context of maternal eating disorders and maternal depression: a comparative study. Psychol Med 1996;26:56974.
  • 12
    Stein A, Woolley H, Cooper S, Winterbottom J, Fairburn CG, Cortina-Borja M. Eating habits and attitudes among 10-year-old children of mothers with eating disorders: longitudinal study. Br J Psychiatry 2006;189:3249.
  • 13
    Micali N, Simonoff E, Treasure J. Infant feeding and weight in the first year of life in babies of women with eating disorders. J Pediatr 2009;154:5560.
  • 14
    American Psychiatric Association (APA). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. Washington, DC: American Psychiatric Association, 1994.
  • 15
    Albertsson Wikland K, Luo ZC, Niklasson A, Karlberg J. Swedish population-based longitudinal reference values from birth to 18 years of age for height, weight and head circumference. Acta Paediatr 2002;91:73954.
  • 16
    Kumar R, Robson KM, Smith AMR. Development of a self-administered questionnaire to measure maternal adjustment and maternal attitudes during pregnancy and after delivery. J Psychosom Res 1984;28:4351.
  • 17
    Kadesjö B, Janols L-O, Korkamn M, Michelson K, Strand G, Trillingsgaard A, et al. 5–15-formuläret (Five to Fifteen): the development of a parent questionnaire for the assessment of AD/HD and comorbid conditions. Eur Child Adolesc Psychiatry 2004;13(Suppl 3):313.
  • 18
    Korkman M, Jaakola M, Ahlroth A, Personen A-E, Turunen M-M. Screening of developmental disorders in five-year-olds using the FTF (Five to Fifteen) questionnaire. A validation study. Eur Child Adolesc Psychiatry 2004;13(Suppl 3):318.
  • 19
    Trillingsgard A, Damm D, Sommer S, Jepsen JRM, Ostergaard O, Frydenberg M, et al. Developmental profiles on the basis of the FTF (Five to Fifteen) questionnaire. Eur Child Adolesc Psychiatry 2004;13(Suppl 3):3949.
  • 20
    Bohlin G, Janols L-O. Behavioural problems and psychiatric symptoms in 5–13-year-old Swedish children—a comparison of parent ratings on the FTF (Five to Fifteen) with the ratings on CBCL (Child Behavior Checklist). Eur Child Adolesc Psychiatry 2004;13(Suppl 3):1422.
  • 21
    Bartholomeusz HH, Courchesne E, Karns CM. Relationship between head circumference and brain volume in healthy normal toddlers, children, and adults. Neuropediatrics 2002;33:23941.
  • 22
    Baschat AA. Neurodevelopment following fetal growth restriction and its relationship with antepartum parameters of placental dysfunction. Ultrasound Obstet Gynecol 2011;37:50114.
  • 23
    Stein AD, Zybert PA, van de Bor M, Lumey LH. Intrauterine famine exposure and body proportions at birth: the Dutch Hunger Winter. Int J Epidemiol 2004;33:8316.
  • 24
    Monteleone P, Luisi M, Colurcio B, Casaroa E, Monteleone P, Ioime R, et al. Plasma levels of neuroactive steroids are increased in untreated women with anorexia nervosa or bulimia nervosa. Psychosom Med 2001;63:628.
  • 25
    Uno H, Eisele S, Sakai A, Shelton S, Baker E, DeJesus O, et al. Neurotoxicity of glucocorticoids in the primate brain. Horm Behav 1994;28:33648.
  • 26
    Seckl JR, Holmes MC. Mechanisms of disease: glucocorticoids, their placental metabolism and fetal ‘programming’ of adult pathophysiology. Nat Clin Pract Endocrinol Metab 2007;3:47988.
  • 27
    Spinillo A, Viazzo F, Colleoni R, Chiara A, Cerbo RM, Fazzi E. Two-year infant neurodevelopmental outcome after single or multiple antenatal courses of corticosteroids to prevent complications of prematurity. Am J Obstet Gynecol 2004;191:21724.
  • 28
    Lajic S, Nordenström A, Hirvikoski T. Long-term outcome of prenatal dexamethasone treatment of 21-hydroxylase deficiency. Endocr Dev 2011;20:96105.
  • 29
    Wadhwa PD. Psychoneuroendocrine processes in human pregnancy influence fetal development and health. Psychoneuroendocrinology 2005;30:72443.
  • 30
    Belfort MB, Rifas-Shiman SL, Sullivan T, Collins CT, McPhee AJ, Ryan P, et al. Infant growth before and after term: effects on neurodevelopment in preterm infants. Pediatrics 2011;128:e899e906.
Commentary on ‘Retarded head growth and neurocognitive development in infants of mothers with a history of eating disorders: longitudinal cohort study’

The association between maternal undernutrition and fetal growth restriction has been documented in resource-poor settings, where food scarcity and hunger are common, and confounding from poverty, poor access to health care, and maternal infections is difficult to assess. In this study in Stockholm, maternal eating disorders (EDs), in which undernutrition is less driven by socio-economic factors, were associated with fetal growth restriction, providing additional evidence linking maternal nutrition to fetal growth. All women were non-smokers, and their period of recovery from EDs before recruitment into the study averaged 3.2 years.

Anorexia nervosa (AN) and bulimia nervosa (BN) were associated with reductions in birthweight and head circumference, but not length. Although effects on weight had resolved by 3 months post-delivery, head circumference restriction, comparable with that observed during famine, persisted through 18 months, when last assessed. These findings differ markedly from the classic pattern of growth restriction from maternal undernutrition or placental insufficiency, in which weight and length are reduced, but head circumference, reflecting brain growth, is spared. Conditions associated with restriction of both weight and head circumference, as seen in this study, include: congenital infections; genetic abnormalities; low pre-pregnancy weight, which exacerbates fetal alcohol-related growth restriction (Carter et al., ACER 2013;37:452–462); and effects of exposure to teratogens, such as smoking, which typically resolve by 6 months postpartum, and alcohol, which persist. Although mothers with EDs had lower body mass indexes (BMIs) than controls, gestational weight gain was unaffected, suggesting normal prenatal nutrition. Thus, the relation between EDs and reduced brain growth may reflect long-term effects set in motion prior to conception, including gene imprinting or epigenetic factors in maternal germ cells.

Maternal EDs were also associated with 5-year neurocognitive deficits; however, mediation of the relationship between EDs and cognition by ED-related smaller head circumference was not examined. Furthermore, maternal and child undernutrition were not assessed. Because AN is far more likely to entail severe malnutrition than BN, the differential pattern of offspring deficits associated with these disorders may provide clues regarding which deficits are attributable to undernutrition. Both groups demonstrated poorer expressive language and social skills, which could reflect poor mother–child interactions, but the observed relationship with reduced head circumference makes this interpretation unlikely. Attention deficits in the children of mothers with BN, and poorer attention and planning in the children of mothers with AN, which may have fallen short of significance because of the limited sample size, also indicate which effects may be related to EDs.

Of note is the association of elevated stress reported at 3 months postpartum by the ED groups with infant growth and 5-year cognition. Maternal stress and elevated cortisol are known to affect infant developmental outcomes, and in primates, elevated cortisol potentiates adverse prenatal alcohol exposure effects (Schneider et al. Child Dev, 1997;68:747–759). Research on the relationship between prenatal maternal stress and EDs, and subsequent development, is therefore warranted.

This is the first study to report an association of maternal EDs with longterm reductions in head circumference. A prospective cohort study relating maternal nutrition, life stress, and prenatal cortisol with child neurocognitive and behavioural development is needed to clarify the contributions of pre-pregnancy and gestational nutrition and comorbid conditions, such as maternal stress and drug use.

Disclosure of interests

NIH/National Institute on Alcohol Abuse and Alcoholism grants K23 AA020516 (RCC) and R01 AA016781 (SWJ and JLJ). The authors have no conflicts of interest to disclose.

  • R Colin Carter,a JL Jacobsonb,c & SW Jacobsonb,c

  • aDivision of Emergency Medicine, Department of Pediatrics, Columbia University College of Physicians and Surgeons, New York, NY, USA

  • bDepartment of Psychiatry and Behavioral Neurosciences, School of Medicine, Wayne State University Detroit, Detroit, MI, USA

  • cDepartments of Human Biology and of Psychiatry and Mental Health, University of Cape Town, Cape Town, South Africa