Little is known about pregnancy course and the health of infants born to women with eating disorders. The findings in the studies published to date suggest that both anorectic and bulimic behavior may affect fetal outcome negatively. Thus, previous eating disorder or current eating disorder behaviors during pregnancy have been associated with preterm delivery, intrauterine growth restriction, low birth weight and low Apgar scores.1,2 However, some case reports have shown infants with normal birth weight and absence of medical complications.3,4
In a retrospective study, Bulik et al5 reported significantly more miscarriages, cesarean deliveries, preterm deliveries, and lower birth weight in offspring of women with a history of anorexia nervosa as compared with controls. Conti et al6 demonstrated a higher prevalence of eating disorders in women delivering small for gestational age (SGA) infants in a case–control study. Prospective follow-up studies in the literature lack control groups.7–12 Lacey et al7 reported a high incidence of obstetric complications and fetal abnormalities in a prospective study of 20 women with bulimia nervosa. In a follow-up study of women formerly diagnosed with anorexia nervosa, the rate of prematurity among the offspring was twice the expected rate and perinatal lethality 6 times the expected rate.10 As part of a longitudinal study, Franko et al12 concluded that pregnant women with active anorexia nervosa or bulimia nervosa seem to be at greater risk of cesarean delivery and of postpartum depression. Recently, Petersen Sollid et al,13 in a register-based study, reported that the risk of low birth weight in children of women with eating disorders was twice the corresponding risk in children of control women.
Limited data exist on the impact of eating disorders on pregnancy. The purpose of this study was to investigate pregnancy and neonatal outcomes in women with past or current eating disorders of either anorectic or bulimic type in comparison with a control group in a prospective study.
MATERIALS AND METHODS
The study population consisted of nulliparous nonsmoking women who were recruited from 13 prenatal clinics in the northwest area of Stockholm in early pregnancy (gestational week 10). As a part of routine clinical investigation, the women were interviewed by a trained midwife about previous and present eating disorder symptoms. These screening interviews were conducted between August 1997 and June 2001. Once a patient with a history of probable eating disorder was identified, she received oral and written information about the study and was invited to participate. The screening proceeded until 50 patients with a history of eating disorders had been recruited. Another 6 screening-positive patients refused to participate. Given that patients with eating disorders are often reluctant to disclose their symptoms, a number of cases is likely to have been missed in the screening procedure. Apart from that, our sample of nulliparous nonsmoking patients consisted of consecutive cases.
The diagnoses of anorexia nervosa, bulimia nervosa and unspecific eating disorder (eating disorder not otherwise specified) were obtained according to the diagnostic criteria of American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders, 4th edition.14 The screening-positive participants were interviewed again to confirm the Diagnostic and Statistical Manual of Mental Disorders diagnoses. The preliminary diagnoses were confirmed in each case. Information about the course and duration of the disease and about past or current therapy for eating disorder was also obtained and confirmed from medical records when available.
During the same period a control cohort of 68 nulliparous nonsmoking pregnant women without a history of eating disorders were recruited from the same prenatal clinics. Their entry into the study was evenly distributed over the period. None of those approached refused to participate.
The research protocol was approved by the local Committee of Medical Ethics, and informed consent was obtained from all subjects. Once enrolled, subjects were assessed at weeks 10, 20, 25, 28, 31, 34, 36, and 40 of gestation. Maternal measurements during pregnancy included weight, blood pressure, hemoglobin, and screening for gestational diabetes and preeclampsia. A screening ultrasound examination was performed during weeks 16–18 of gestation for determination of gestational age, simple or multiple parity, and detection of malformations. An extra ultrasound examination was performed during weeks 33–36 of gestation for measurement of fetal growth.
Maternal antenatal complications, including hyperemesis, gestational hypertension, preeclampsia, and gestational diabetes, were evaluated based on reports from the medical records. Gestational hypertension was defined as 2 or more antenatal systolic blood pressure readings of 140 mm Hg or more, or 2 or more antenatal diastolic readings of 90 mm Hg or more. Preeclampsia was diagnosed as blood pressure of 140/90 mm Hg or more in combination with proteinuria 0.3 g or more per 24 hours. Obstetric complications including intrauterine growth restriction and preterm delivery (before 37 completed weeks of gestation) were also evaluated. Intrauterine growth restriction was based on the second ultrasound examination and defined as fetal growth less than 22% of fetal growth distribution of a Swedish reference population.15
Data on mode of delivery and on complications such as induction of labor, vacuum extraction, use of forceps, caesarian delivery, and postpartum bleeding were obtained from medical records. Neonatal outcome variables, including Apgar score, birth weight, length, head circumference, SGA, and neonatal care (defined as any intervention by a pediatrician), were also evaluated. Small for gestational age was defined as a birth weight below 2 standard deviations of birth weight distribution of a Swedish reference population stratified by gender and gestational age at delivery.16 Microcephaly was defined as smaller than 2 standard deviations of head circumference distribution of the same Swedish reference population.
The primary outcome variable was birth weight. We included 50 patients and 68 controls to detect a difference between groups of approximately 10% with 80% power. Data are presented as means ± standard deviation. For comparisons between patients and controls, 1-way analysis of variance was used. For blood pressure variables an analysis of covariance was performed to control for the effect of body mass index (BMI). Discrete variables were analyzed with χ2 test of independence. Patients were categorized into 2 groups: anorexia nervosa and bulimia nervosa, excluding patients diagnosed with eating disorder not otherwise specified. For comparisons between the 2 different types of eating disorders and controls 1-way analysis of variance was used followed by Fisher's least significant difference test.17P < .05 was considered significant (2-tailed test).
In each group of patients and controls there was 1 twin pregnancy diagnosed at the first ultrasound. These 2 pregnancies were excluded from the study. Twenty-four women were found to have a history of anorexia nervosa, 20 of bulimia nervosa, and 5 of eating disorder not otherwise specified. The distribution of the specific types of eating disorder is shown in Table 1. Patients with history of different episodes of both anorexia nervosa and bulimia nervosa were counted as having had anorexia nervosa only. When the patients entered the study, they all claimed to have been free from previous eating disorders. The mean duration of eating disorders was 9 years (range 3–15 years).
Sixteen (33%) of the patients with a history of eating disorders had received treatment. Nine had been treated at an eating disorder clinic, 6 at a general psychiatry or child and adolescent psychiatry out-patient clinic, and 1 in primary care. All medical records from these treatment episodes were scrutinized to obtain a clearer picture of the patients’ eating disorders.
In 11 of the patients (22%), a relapse of eating disorder occurred during pregnancy that led to contact with a psychologist or psychiatrist. Eight patients diagnosed with previous anorexia nervosa and 3 patients with past bulimia nervosa relapsed to similar symptoms. None became hospitalized due to eating disorders during pregnancy. Only 2 patients received psychotropic medication (citalopram in one case and fluoxetine in the other).
Physical characteristics in patients and controls in week 10 of gestation are shown in Table 2. Weight and BMI (body weight in kilograms divided by squared height in meters) were significantly lower in patients than in controls. The anorectic subgroup had significantly lower BMI as compared with controls (19.3 ± 2.9 compared with 22.3 ± 2.8, P < .001) and bulimics (19.3 ± 2.9 compared with 21.6 ± 2.9, P < .01), whereas there was no difference in BMI between bulimics and controls (P = .30). Systolic and diastolic blood pressures were not significantly different between groups when statistical analyses were adjusted for BMI.
The mean duration of pregnancy and maternal weight gain did not differ between patients and controls (Table 3). However, the anorectic subgroup showed significantly lower weight gain than the control group (10.4 ± 3.9 compared with 12.1 ± 2.6 kg, P < .05). Anemia (hemoglobin < 110 g/L) evaluated at gestational week 25 was significantly increased in the patient group as compared with controls (Table 3). Treatment and care of hyperemesis were significantly more common among patients than controls (Table 3). Diagnosis of gestational hypertension and preeclampsia, as well as intrauterine growth restriction, showed an insignificant increase in patients compared with controls (P = .07) (Table 3). The increase in diastolic blood pressure during pregnancy was significantly higher in the anorectic subgroup as compared with controls (11.7 ± 12.4 compared with 3.2 ± 10.6, P < .05). There was no case of gestational diabetes in any group. Intrauterine growth restriction was diagnosed in 4 patients, but not in any control woman (Table 3).
Delivery events are shown in Table 4. There was no significant difference between the patients and controls with respect to epidural anesthesia, induction of labor, vacuum extraction, use of forceps, breech presentation, preterm delivery, caesarean delivery, or postpartum bleeding.
Mean birth weight was significantly lower in the patient group than in the control group (Table 5). The patients’ infants also had highly significant smaller head circumference than those of controls (Table 5). When comparing the different types of eating disorders, birth weight was only significantly lower in the infants of women with previous anorexia nervosa (3,210 ± 533 compared with 3,516 ± 515, P < .05), whereas head circumference was significantly smaller in infants of women with both anorexia nervosa and bulimia nervosa as compared with controls (33.7 ± 1.6 and 33.7 ± 1.0 compared with 35.2 ± 1.6, P < .001, respectively). Apgar score of 6 or less at 5 minutes and neonatal care of the infants due to respiratory distress and prematurity did not differ between the patient group and the control group (Table 5). However, the occurrence of SGA and microcephaly was significantly increased in infants of patients compared with controls (Table 5). There were 8 patients delivering infants with SGA or microcephaly or both, and 2 of the infants had both SGA and microcephaly. In those patients, 4 mothers had anemia, 2 had verified relapse in eating disorder, and in 1 mother group B streptococcal infection was diagnosed. Furthermore, there was 1 heart malformation among the SGA infants, the only malformation in either group.
This is a prospective study in which women with past or current eating disorder have been followed up during pregnancy and compared with controls with regard to pregnancy course and neonatal outcome. Further methodologic strengths are the inclusion of only nulliparous and nonsmoking women. The main findings in this study are significantly lower birth weight and smaller head circumference in infants of mothers with a history of eating disorders. Furthermore, there was an increased occurrence of SGA and microcephaly in children of mothers with eating disorders.
Lower birth weight and increased rate of SGA in infants of women with eating disorder has been found in some5,6,8,13 but not all3,4 previous studies. Our finding of an increased occurrence of hyperemesis gravidarum is also reported by others.18,19 However, we were not able to corroborate earlier reports of a higher rate of preterm and cesarean delivery.5,10,12,13 It could be due to relatively small sample size in our study. However, only 2 of the previous studies contained a control group.5,13 Furthermore, fetal abnormalities were not increased in the eating disorder group, as previously reported.7,10
A new finding in the present study is a diminished mean head circumference in children of mothers with a history of eating disorder. Furthermore, the prevalence of microcephaly was 8% in the same group of children. This is high in comparison with 2.5% in the general population, using the same definition of microcephaly as in our study.20 It is noteworthy that infants of both the anorectic and the bulimic subgroup had significantly smaller head circumference, whereas only the infants of the anorectic subgroup had significantly lower birth weight than the offspring of controls.
The precise mechanisms behind our findings of lower birth weight, smaller head circumference, and higher rate of microcephaly and SGA in children of mothers with past or active eating disorder are not known. However, it is likely that inadequate eating behavior during pregnancy might result in a deficient nutrient supply to the fetus. Eleven patients had a verified relapse of eating disorders during pregnancy, and of those, 2 patients delivered infants with SGA or microcephaly. It could be assumed that the real rate of recurrence was even higher, because the majority of the women with eating disorders before pregnancy had not consulted any medical service.
It is also probable that eating disordered women, with their often disturbed self-image, are at risk of stress reactions during pregnancy. Lou et al21 showed that maternal stress was a significant determinant of small head circumference, indicating a specific effect on brain development. Patients with anorexia nervosa have elevated levels of plasma cortisol and show less cortisol suppression after dexamethasone.22 Glover23 reported a strong correlation between plasma concentration of cortisol in the mother and in the fetus and concluded that raised cortisol in the mother may have a direct effect on the development of the brain. Thus, hypercortisolemia in anorexia nervosa might cause microcephaly. However, cortisol levels were not investigated in this study. Not much is known about cortisol levels in pregnant bulimics who, however, are certain to be exposed to high stress, because they have nearly as disturbed self-image as anorectics.24
Although maternal mean weight gain among the patients was not significantly different from that of the controls, the patient group did not reach the recommended weight gain of 11.5–16.0 kg during pregnancy.25 Furthermore, the anorectic subgroup had significantly lower weight gain than the controls, indicating restricted eating behavior. In the whole patient group there was also an increased frequency of anemia. Occurrence of nutritional deficiency of the fetus is supported by the tendency for an increased rate of intrauterine growth restriction among the patients. Gestational hypertension and preeclampsia tended to be increased in the patient group, and there was a significantly higher increase of the diastolic blood pressure in the subgroup of anorectics as compared with controls. Whether this finding has a clinical significance remains to be elucidated.
The placenta also produces growth hormones that are important for fetal growth, but such hormones were not investigated in this study. Furthermore, fetal growth is dependent on genetic disposition. The anorectic subgroup had significantly lower BMI in week 10 of gestation than controls and bulimics. It can therefore not be excluded that genetic factors contributed to the lower birth weight in the anorectic group. Many other factors that affect the mother, such as emotional and physical stress and alcohol and substance abuse, are important for fetal growth. No abuse was known among patients and controls. Maternal smoking, the most well-known contributor to intrauterine growth restriction, was excluded in this study. Further studies are needed to clarify the mechanisms for lower birth weight and smaller head circumference in infants of mothers with eating disorders.
In conclusion, this prospective study demonstrates that pregnant women with past or active eating disorders are at greater risk of delivering infants with lower birth weight, smaller head circumference, microcephaly, and SGA. Therefore, these women should be recognized as at-risk patients during pregnancy. The long-term effects of our findings, and particularly of decreased head circumference, are very important to address in future studies.
1. Katz MG, Vollenhoven B. The reproductive endocrine consequences of anorexia nervosa. BJOG 2000;107:707–13.
2. Morrill ES, Nickols-Richardson HM. Bulimia nervosa during pregnancy: a review. J Am Diet Assoc 2001;101:448–54.
3. Willis DC, Rand CS. Pregnancy in bulimic women. Obstet Gynecol 1988;71:708–10.
4. Lemberg R, Phillips J. The impact of pregnancy on anorexia nervosa and bulimia. Int J Eat Disord 1989;8:285–95.
5. Bulik CM, Sullivan PF, Fear JL, Pickering A, Dawn A, McCullin M. Fertility and reproduction in women with anorexia nervosa: a controlled study. J Clin Psychiatry 1999;60:130–7.
6. Conti J, Abraham S, Taylor A. Eating behavior and pregnancy outcome. J Psychosom Res 1998;44:465–77.
7. Lacey JH, Smith G. Bulimia nervosa: the impact of pregnancy on mother and baby. Br J Psychiatry 1987;150:777–81.
8. Stewart DE, Raskin J, Garfinkel PE, MacDonald OL, Robinson GE. Anorexia nervosa, bulimia, and pregnancy. Am J Obstet Gynecol 1987;157:1194–8.
9. Treasure JL, Russel GF. Intrauterine growth and neonatal weight gain in babies of women with anorexia nervosa. BMJ 1988;296:1038.
10. Brinch M, Isager T, Tolstrup K. Anorexia nervosa and motherhood: reproductional patterns and mothering behavior of 50 women. Acta Psychiatr Scand 1988;77:98–104.
11. Blais MA, Becker AE, Burwell RA, Flores AT, Nussbaum KM, Greenwood DN, et al. Pregnancy: outcome and impact on symptomatology in a cohort of eating-disordered women. Int J Eat Disord 2000;27:140–9.
12. Franko DL, Blais MA, Becker AE, Delinsky SS, Greenwood DN, Flores AT, et al. Pregnancy complications and neonatal outcomes in women with eating disorders. Am J Psychiatry 2001;158:1461–6.
13. Petersen Sollid C, Wisborg K, Hjort J, Secher NJ. Eating disorder that was diagnosed before pregnancy and pregnancy outcome. Am J Obstet Gynecol 2004;190:206–10.
14. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th ed. Washington (DC): American Psychiatric Association; 1994.
15. Marsal K, Persson PH, Larsen T, Lilja H, Selbing A, Sultan B. Intrauterine growth curves based on ultrasonically estimated foetal weights. Acta Paediatr 1996;85:843–8.
16. Persson PH, Weldner BM. Normal range growth curves for fetal biparietal diameter, occipito frontal diameter, mean abdominal diameters and femur length. Acta Obstet Gynecol Scand 1986;65:759–61.
17. Roger E. Kirk. Experimental design: procedures for the behavioral sciences. 3rd ed. Pacific Grove (CA): Brooks/Cole; 1995.
18. Lingam R, McCluskey S. Eating disorders associated with hyperemesis gravidarum. J Psychosom Res 1996;40:231–4.
19. Abraham S. Sexuality and reproduction in bulimia nervosa patients over 10 years. J Psychosom Res 1998;44:491–502.
20. Vargas JE, Allred EN, Leviton A, Holmes LB. Congenital microcephaly: phenotypic features in a consecutive sample of newborn infants. J Pediatr 2001;139:210–4.
21. Lou HC, Hansen D, Nordentoft M, Pryds O, Jensen F, Nim J, et al. Prenatal stressors of human life affect fetal brain development. Dev Med Child Neurol 1994;36:826–32.
22. Walsh BT, Roose SP, Katz JL, Dyrenfurth I, Wright L, Van de Wiele R, et al. Hypothalamic-pituitary-adrenal-cortical activity in anorexia nervosa and bulimia. Psychoneuroendocrinology 1987;12:131–40.
23. Glover V. Maternal stress or anxiety during pregnancy and the development of the baby. Pract Midwife 1999;2:20–2.
24. Björck C, Clinton D, Sohlberg S, Hällström T, Norring C. Interpersonal profiles in eating disorders: ratings of SASB self-image. Psychol Psychother 2003;76:337–49.
25. Institute of Medicine. Nutrition during pregnancy, Part I. Weight gain. National Academy of Sciences, Washington (DC): National Academy Press; 1990. p. 1-233.