Psychology Department

Health Psychology Home Page

Papers written by students providing scientific reviews of topics related to health and well being

Search HomeWeight LossAlternative Therapy | Supplements | Eating Disorders | Fitness | Links | Self-Assessment | About this Page |


Pregnancy and Eating Disorders

Meegan Peery




Concerns about gaining weight and retaining a youthful figure are expressed by many pregnant women. When there has been a history of anorexia nervosa or bulimia nervosa, weight gain and body shape changes accompanying pregnancy can provoke extreme distress (Rand et al., 1987). Very little is known about the impact of pregnancy on women with anorexia nervosa or bulimia nervosa. Despite the fact that amenorrhea, the lack of menstruation, and infertility are common features of these syndromes, some women have been able to conceive, even at below normal body weight (Lemberg & Phillips, 1989). Information is lacking in general on psychological impact on the mother-to-be as well as on the course of pregnancy from a medical and nutritional point of view as it relates to both the mother and the unborn child.




Blinder and Hagman (1984) interviewed six women with anorexia nervosa or bulimia nervosa who had given birth while actively symptomatic. Most of the women reported developing better control over their symptoms during the pregnancy and were motivated by wanting to have healthy children and to be healthy themselves in order to be better caretakers. However, following the deliveries they regressed to dysfunctional eating patters, often with an increase in severity (Lemberg & Phillips, 1989).

Stewart, Raskin, Garfinkel, MacDonald, and Robinson (1987) compared the pregnancies of seven women, four of whom were anorexic and three bulimic at the time of pregnancy, with eight women whose symptoms were in remission. In contrast to the above studies, the authors reported that the women with an active eating disorder did not improve in their symptoms, but continued or deteriorated during the pregnancy compared with the pre-pregnancy level. Moreover, the active eating disorder group was significantly less able to gain weight during the pregnancy than the group in remission. Consequently, their babies had significantly lower birth weights and lower 5-minute Apgar scores than the babies of the women who were in remission (Franko & Walton, 1993).

Lacey and Smith (1987) investigated the pregnancies and fetal outcomes of 20 normal-weight bulimic women in the largest study to date. Consistent with both studies by Blinder and Hagman (1984) and Namir et al. (1986), they reported a significant reduction in the eating disorder symptoms in the majority of women during the course of the pregnancy, with 75% having a complete cessation of bingeing and purging by the third trimester. Also, consistent with previous reports, the majority of women regressed in the postpartum period (Psychological Medicine, 1991). However, the authors note that a full 25% of the sample appeared to be cured of their eating disorder symptoms. Despite a significant reduction in the binge-purge behavior, the authors reported significant fetal abnormality involving a cleft pallet child, another with a cleft lip, and higher than average pregnancy complications, multiple pregnancies, and obstetric complications. Unfortunately, no information was provided on maternal weight gain or infant weights (Lemberg & Phillips, 1989).

Hollifield and Hobdy (1990) reported their experience with three bulimic women who became pregnant while in therapy. Although the women felt a great deal of guilt about their bulimic behaviors, they nevertheless continued to binge and purge throughout pregnancy. None of the women informed their obstetrician or health care professionals about their eating disorder and all minimized or lied about their eating behaviors to family and friends. Although all deliveries and birth weights were normal, one woman began bingeing and purging immediately after delivery, which resulted in damage to the sutures in her episiotomy. Each woman reported considerable fear, guilt, and shame because of her inability to stop the bulimic behaviors. The authors concluded that persons involved with the delivery of health care to the bulimic women must challenge the assumption that bulimia ceases or diminishes once the pregnancy is confirmed and must consider the medical implications of continuing bulimic practices in the pregnant woman (Franko & Walton, 1993).




Namir, Melman, and Yager (1986) interviewed six restricter-type anorexics during pregnancy and reinterviewed four 3-4 months after childbirth. As in the Blinder and Hagman study, it was observed that the women felt obliged to care for themselves better in order to be better mothers. Eating habits improved overall, although the women continued to have difficulties with poor body image and generally were unable to distinguish being pregnant from being "fat". Most feared that their weight gain would become out-of-control. Of the women studied, one-half experiences serious anxiety and depression during the pregnancy. Toward the end of the pregnancy there was less food preoccupation and anorexic thinking. However, of the four studied at follow-up, all become reinvested in the anorexic symptoms, with three of the four women losing weight to lower levels than prior to pregnancy. Although the deliveries were reported as "easy," the babies weighed 6.9 lb., on the average, which appears to be lower than the population average of 7.5 lb. reported by Dohrmann and Lederman in 1986.

This is one of the first studies to examine women with anorexia nervosa during and following pregnancy. However results and findings would be considered to be stronger if a larger sample size was used so as a more representational study of pregnancy in association with eating disorders could be reviewed. Finally, the need for close postpartum monitoring of these patients, their babies and the mother-child interaction should be underscored. As of the conclusion of this study it is not known to what extent the mother-infant patterns are distorted and how much at risk these babies are for the development of psychopathology.




Rand, Willis, & Kuldau (1987) describe two case studies of the pregnancies of two women with a history of anorexia nervosa. In both of the two cases the women experienced emotional distress associated with body shape changes and weight gain throughout their pregnancies. This emotional distress has been reported in other studies including, Stewart et al. (1987). Mrs. A hoped that her baby would be spared from negative consequences derived from her inability to follow dietary guidelines or to eliminate food abuses and vomiting. Mrs. B was able to significantly reduce anorexic symptoms with the help of dietary guidance and constant reassurance. Both women experienced two episodes of acute emotional distress. In the first trimester, early shape changes were interpreted as signs of obesity. Dieting and vomiting were moderated only because of concerns about their baby’s health. Later, both women were able to attribute their "fatness" to the pregnancy. Following delivery, both women equated inability to wear their pre-pregnancy clothes with obesity. Both then dealt with their fears by means of extreme dieting and vomiting. These results are also seen in studies by Lacey & Smith (1987), and Namir et al. (1986).

This review associates the need of patients with a history of anorexia nervosa or bulimia nervosa with far greater emotional support and dietary guidance than other patients. Rand et al. State that when trained office staff are available, supplemental counseling should be provided during each prenatal visit and a liaison with a mental health specialist is recommended to help the mother cope with fears and extreme dieting behaviors both during pregnancy and after delivery.

Although Rand et al. offer sound advice in stating that pregnant women with eating disorders should receive emotional support and guidance from trained medical professionals during and after pregnancy; the fact remains that most women with an eating disorder do not ask for help or try to hide their eating disorder from their doctors. Another improvement to this study would include increasing the sample size and performing a proactive experimental design rather than a retrospective case study. With a true experiment the study would allow itself to be tested for reliability and validity.




Ramchandani and Whedon (1988) report on two case studies where bulimic symptoms remitted during pregnancy and offer an alternative hypothesis regarding remission. While studies by Lacey & Smith (1987), Blinder and Hagman (1984), and Namir et al. (1986), attribute the positive effect to the increased motivation, provided by pregnancy, to the women’s control of their eating disorder for fear of harming their babies. Ramchandani and Whedon believe that the remission of bulimic symptoms during pregnancy can be attributed to a person’s physiological set point. A person’s physiological set point is postulated to exist in every individual around which the body weight is maintained fairly predictably over the life span (Keesey, 1980). This set point determines the level of activity of the hypothalamic hunger and satiety centers. Bulimic craving for food may result in an individual whose physiological set point of body weight is higher than ones psychological body image. In pregnancy, the physiological set point moves up to accommodate for fetal needs. As the difference between the set point and the psychologically determined ideal body weight lessens bulimic symptoms may subside at the cost of weight gain.

This study makes us of Keesey’s work (1980), which focused on obesity and not eating disorders. Therefore, the inclusion of the set point theory as a motivation for decreasing bulimic symptoms has not been fully studied and the conclusions made by Ramchandani and Whedon are not scientifically sound. The authors also focus on two case studies in which they have not manipulated any variable to observe a change. Therefore, I believe that this study would be more effective if a larger sample size was used and if a proactive approach instead of a retrospective approach was taken.




Lemberg and Phillips (1989) surveyed women with active eating disorders during their first pregnancies to gather a larger database concerning several aspects of pregnancy and childbirth. Sixty-one women from across the United States volunteered to participate in the study. These women were described as being well educated, married, with a chronic history of teenage onset anorexia nervosa or bulimia nervosa, who had their first child in their mid-twenties. Of these, 43 women met the self-report criteria of having had an active eating disorder involving anorexia nervosa, bulimia nervosa, or mixed symptoms during the 6 months prior to the first pregnancy. Of these 43 women, 43% were self-identified as having bulimia nervosa, 5% as having anorexia nervosa, and 52% as having a combination of both symptoms. Additionally, as a validity check for inclusion into the study subjects were asked about the presence of specific symptoms, including weight, restrictions of intake, and binge-purge behavior. Based on an analysis of this information, the sample appears to contain 5-7% restricter-type anorexics, 77% bulimics, and the remaining 16% having a combination of both anorexic and bulimic symptoms.

The pregnancy experience appears to have a pronounced beneficial impact on both anorexic and bulimic symptoms, with 70% of the women in the study reporting overall improvement and 56% seeing themselves as largely in remission with their eating disorder symptoms during pregnancy. These trends toward the positive benefit of pregnancy are consistent with the majority of existing literature including the studies by Blinder and Hagman (1984), Lacey & Smith (1987), and Namir et al. (1986).

The larger majority of women were fearful of losing control of weight during pregnancy, yet this proved to be unfounded generally. On the whole, the respondents had normal weight gains. Despite the overwhelming majority of women having fears that their child might be damaged owing to poor nutrition, there appears to be little evidence to support this from this study based on birth weights, birth complications, or health status of the infant at delivery. This contrasts sharply with other cited studies. The mean birth weights of 18 children measured in the sub-sample were normal. Of the 18 children, only one child had a near low birth weight, with the exception of twins. Seventy-four percent of the women in the sample and 82% of the women in the sub-sample of known infant weights, had been able to gain at least 20 lb. during their pregnancies. Perhaps the self-selection aspect of this sample, as contrasted with the other studies of Namir et al (1986), Stewart et al. (1987), Rand et al. (1987), and Lacey & Smith (1987) accounts for the positive aspect of this sample. Women having problem-free pregnancies might be more likely to volunteer for the study than women with complications in their pregnancies.

Disappointingly, about half of the women who made significant improvement in their eating disorder symptoms during pregnancy regressed in the first year after birth and attributed their regression to "feeling fat" and therefore wanting to lose weight. This is similar to findings noted by Lacey & Smith (1987) and Namir et al. (1986). On the positive side, about one-fourth of the sample did quite well following the birth and appeared to maintain improvements in eating disorder symptoms and in greater body-image acceptance. This percentage of "cured" women matches almost exactly with what Lacey & Smith (1987) reported. For those individuals pregnancy proved to have significant psychological impact in terms of increase meaning and purpose of life.

Although the current study provides a larger database of pregnant women with an eating disorder than what has been available to date, the results remain preliminary. It would be in error to generalize these data to the eating disorder population at large, since the survey procedures did not ensure a representative sample. The respondents in this study appeared to have very serious eating disorders preceding their first pregnancy on average by eight years. It is also possible that this study due to self-selection in the sample is biased toward women with better pregnancy experiences and childbirth outcomes or perhaps biased toward women with worse pregnancy experiences and childbirth outcomes. Many other variables that can impact adversely on pregnancy including exercise, smoking, drinking habits, and nutritional status were not examined. A large survey employing a random sample is needed. A follow up study of developmental progress of the infants would also be desirable to assess both psychological progress as well as potential neurological damage that may emerge over time.




Mitchell, Seim, Glotter, Soll, and Pyle (1991) performed a retrospective comparison of the outcome of 38 pregnancies in 20 actively bulimic women and 50 pregnancies in 31 control women. Women being evaluated for bulimia nervosa in the Eating Disorders Clinic at the University of Minnesota who satisfied DSM-III criteria and who indicated that the had been pregnant and had been binge eating at least once a week throughout the first trimester were asked to participate. Subjects completed a questionnaire that asked questions about eating behavior before, during, and after each pregnancy as well as questions about psychiatric and medical history.

Half of the bulimic subjects were binge eating at least daily and 80% were vomiting on a regular basis, whereas 30% were using laxatives for weight control on a regular basis. When compared to controls, the bulimic women became pregnant at an older age. The number of weeks of gestation were essentially identical. The bulimic subjects numerically gained less weight than did the control patients, but this difference did not reach statistical significance. The mean birth weight of the infants born during the bulimic pregnancies was not significantly different from the mean birth weight of the infants born during the control pregnancies, the results parallel the study by Hollifield & Hobdy (1990). However, the bulimic pregnancies were more than twice as likely to result in fetal loss through miscarriage or neonatal death, most attributable to a miscarriage rate of 39%, compared to a miscarriage rate of 17% among controls. The results indicate that there is no statistically significant greater risk of miscarriage in individuals who are actively bulimic during the first trimester of their pregnancy, although the rate of miscarriage was approximately twice as high in bulimic pregnancies.

The results of this investigation must be regarded as preliminary for several reasons. First, the sample sizes were relatively small for this type of study. However, the majority of women seen for treatment with bulimia nervosa have not been pregnant and it will be difficult to enlarge the sample size. Secondly, this study was retrospective, and is open to all the problems inherent in a retrospective design. The study relied primarily on patients self report, although hospital information was obtained whenever possible. Much of this information was collected about situations that had taken place several years prior to the interview, introducing a significant possibility of inaccuracy. Third, individuals who had experienced certain complications may have been more likely to remember them if they were concerned already about the possible adverse consequences of their bulimia on pregnancy outcome and may have exaggerated their reports out of a sense of guilt. The opposite conclusion could be drawn as well; perhaps the self-report may have led to an underestimation of the severity of the bulimic symptoms. Fourth, many other variables that can impact adversely on pregnancy including exercise and nutritional status were not examined. Fifth, the controls were chosen to be similar in age and geographic area, but were not matched on a number of important variables, including smoking, drinking habits, and amount of prenatal care. Therefore, variables not addressed may have accounted for the differences.




Bonne, Rubinoff, and Berry (1996) focused on the delayed detection of pregnancy in two anorexia nervosa binge eating/purging type patients. Early subjective signs suggestive of pregnancy include a feeling of engorgement and tenderness in the breasts, nausea with or without vomiting, lightheadedness, and fatigue are all commonly reported by patients with eating disorders. Therefore, detection of pregnancy in such patients is likely to be delayed. Although cessation of menstruation is considered an indication of pregnancy, gestation may begin without prior menstruation. In the two cases presented here, cessation of menstruation occurred long before the beginning of pregnancy due to amenorrhea. In keeping with their eating disorder, both patients had feelings of bloated bellies, engorged breasts, nausea, weakness, and lassitude. Although anorexia nervosa mainly affects women of childbearing age, pregnancy in symptomatic patients is seldom observed. In the rare instances when conception occurs the authors mention increased risk of complications, both during pregnancy and at birth supported by Namir et al. (1986), Stewart et al. (1987), and Lacey & Smith (1987).

Again this study is a retrospective observation of two anorexic women. In order to increase the reliability and validity of these findings future studies should be proactive longitudinal studies beginning before conception and ending after the birth of the child. The studies should also include random sampling of a larger sample size so that the results will accurately reflect the population of pregnant anorexic women.




Turton, Hughes, Bolton, and Sedgwick (1999) surveyed five hundred thirty women attending prenatal follow-up clinics at a large London district general hospital during a 4-week period. The Eating Attitudes Test (EAT), the Edinburgh Postnatal Depression Scale (EPDS), and a demographic questionnaire were administered. A total of 492 (92.8%) women accepted questionnaires. For the 2 year period prior to conception, 37 women (10%) scored above the threshold on the EAT. In pregnancy, 20 women (4.9%) scored above the threshold on the EAT.

This study found that around 8% of those women surveyed had experienced a number of the symptoms of anorexia nervosa and had had significant anxieties about eating and weight at a level that interferes with psychosocial functioning in the 2 years preceding pregnancy. In agreement with other studies (Lacey & Smith, 1987; and Rand et al., 1988), they found that symptom levels were lower in pregnancy. However, nearly one third of those women who reported eating disorder symptomatology in the past 2 years also reported symptoms in pregnancy, suggesting that as many as 1% of women may suffer from some form of an eating disorder during pregnancy.

These screening procedures using anonymous or self-report questionnaires lack the stringency of a diagnostic interview. However, although the EAT has not been validated for use in pregnancy or retrospectively, the use of anonymous self-report questionnaires does offer some benefits. Eating disorders are almost certainly underreported because of shame about the condition. Anonymity increases the probability of achieving a truer picture of the incidence of symptoms, born out of the high response rate of this survey. This study has indicated that there may be a significant group of women suffering from eating disorder symptomatology in pregnancy. This is a time when women are already in contact with medical services and when they may be more prepared to seek treatment because of concern for the health of their infant.




Blais et al., (2000) identified 54 women self-reporting 82 pregnancies from a sample of 246 women participating in a longitudinal study of anorexia nervosa and bulimia nervosa. These women met the criteria for DSM-III-R for anorexia nervosa or bulimia nervosa. On the 225 subject 51 were classified as anorexia nervosa restricting type, 85 as anorexia nervosa bingeing and purging subtype, and 110 bulimia nervosa. The Eating Disorders Longitudinal Interval Follow-up Evaluation (LIFE-EAT II) was administered to subjects at 6-month intervals. The LIFE-EAT II is a semistructured interview designed to assess eating disorder symptomatology, comorbid psychopathology, treatment participation, and psychosocial functioning. Of the 82 pregnancies reported, 46 (56%) resulted in live births, 25 (30%) resulted in therapeutic abortion, and 11 (13%) resulted in spontaneous abortion.

This is the first study ever to use prospectively collected data and a large cohort of subjects to examine the impact and outcome of pregnancy in women with eating disorders. This study showed a slightly lower live birth rate (56%) compared to the expected rate of (66%), a higher than expected rate of therapeutic abortion (30%) compared with an expected rate of (19%), and a spontaneous abortion rate of (13%) compared to the expected rate of (14%). Furthermore the findings indicate that pregnancy outcome was not related to intake eating disorder diagnosis, age at time of pregnancy, or the presence of an eating disorder diagnosis at conception.

These findings are somewhat inconsistent with those reported in other studies. Although their findings were not statistically significant, Mitchell et al. (1991) found twice the rate of fetal loss among subjects with bulimia nervosa compared to a control group, and even higher the rate of fetal loss among more severely symptomatic women with bulimia nervosa. However Mitchell et al (1991) included prenatal and neonatal deaths, respectively, among their rates, whereas this study examined fetal demise due to spontaneous abortion only. The factor that was found to be most strongly associated with pregnancy outcome was marital status. Married women had a significantly greater number of live births than did single women. The meaning of the finding is somewhat unclear and it can be speculated that the presence of a marriage partner likely indicated the availability of additional social, financial, and emotional support, making the prospect of parenthood more acceptable.

With regard to the effect of pregnancy on eating disorder symptoms, these finding suggest that during pregnancy, women with eating disorder generally experience a reduction in the severity of their symptoms. The anorexic subjects demonstrated a similar patter of symptom reduction. Overall these findings are consistent with the majority of earlier studies that also showed a decrease in symptom severity during pregnancy (Lacey & Smith, 1987; Lemberg & Phillips, 1989). Although these findings suggest that pregnancy has a significant positive effect on some eating disorder symptomatology, a number of variables in this study were unaffected by the occurrence of pregnancy. Variables that did not show significant change include self-induced vomiting, over concern with body image, compulsive exercise, and restrictive eating. In this regard, the findings do not agree with those of other studies that found a significant decrease in self-induced vomiting (Lacey & Smith, 1987; Lemberg & Phillips, 1989).

Despite the large sample and prospective nature of these data, the study’s findings should be considered preliminary due to certain limitations in the design of this study. Included in these limitations are the absence of a control group, the reliance on self-report data, and a relatively small sample of anorexic subjects. Nonetheless, the prospective results indicate a general overall reduction in the severity of eating disorder symptoms for both bulimia and anorexia subjects that are consistent with a majority of studies in this area, lending support to the validity of this finding. Addition data including duration of the gestation period, birth weights, medical complications occurring during the course of pregnancy, and Apgar scores need to be examined in order to obtain a more complete assessment of the outcome of pregnancy in eating disordered women.




Anorexia nervosa and bulimia nervosa are rarely diagnosed during pregnancy, but many cases come to light later, usually after seeking treatment for their eating disorders at specialist clinics. This method of case identification means that most of the work reviewed above is based on a highly selected sample of patients, possibly exaggerating the negative effects of the eating disorders on pregnancy.

The evidence which is available suggest that serious eating disorders are rarely precipitated during pregnancy, bulimic symptoms frequently improve temporarily, but the course of anorexia is less vulnerable to change. The data suggesting an association between eating disorders and a variety of complications of pregnancy, delivery, and raised prenatal morbidity are more persuasive than those linking maternal bulimia with fetal abnormalities. There is a clear need for accurate prevalence rates of eating disorders in pregnancy to be derived in order that this issue can be addressed and so that obstetricians can be advised of the clinical risks and the possible benefits of psychiatric intervention.




"Eating Disorders in Pregnancy." Psychological Medicine. 1991 Aug: Vol. 21(3): 577-580.


Blais, Mark A., et al. "Pregnancy: Outcome and Impact on Symptomatology in a Cohort of Eating-Disordered Women." International

Journal of Eating Disorders. 2000 Jan; Vol. 27(1): 140-149.


Bonne, Omer B., et al. "Delayed Detection of Pregnancy in Patients with Anorexia Nervosa: Two Case Reports." International Journal of

Eating Disorders. 1996 Dec; Vol. 20(4): 423-425.


Franko, Debra L. and Barbara E. Walton. "Pregnancy and Eating Disorders: A Review and Clinical Implications." International Journal of

Eating Disorders. 1993 Jan; Vol. 13(1): 41-48.


Keesey, R. E. "A Set point analysis of the regulation of body weight." Obesity. 1980; 144-165.


Lemberg, Raymond and Jeanne Phillips. "The Impact of Pregnancy on Anorexia Nervosa And Bulimia." International Journal of Eating

Disorders. 1989 May; Vol. 8(3):285-295.


Mitchell, James E., et al. "A Retrospective Study of Pregnancy in Bulimia Nervosa." International Journal of Eating Disorders. 1991

Mar; Vol. 10(2): 209-214.


Namir, Shelia, et al. "Pregnancy in Restricter-Type Anorexia Nervosa: A Study of

Six Women." International Journal of Eating Disorders. 1986 Jul; Vol. 5(5): 837- 845.


Ramchandani, Dilip and Barbara Whedon. "The Effect of Pregnancy on Bulimia."

International Journal of Eating Disorders. 1988 Nov; Vol. 7(6): 845-848.


Rand, Colleen S., et al. "Pregnancy After Anorexia Nervosa." International Journal of

Eating Disorders. 1987 Sep; Vol. 6(5): 671-674.


Turton, Penelope, et al. "Incidence and Demographic Correlates of Eating Disorder

Symptoms in a Pregnant Population." International Journal of Eating Disorders.

1999 Dec; Vol. 26(4): 448-452.



Psychology Department

The Health Psychology Home Page is produced and maintained by David Schlundt, PhD.

Vanderbilt Homepage | Introduction to Vanderbilt | Admissions | Colleges & Schools | Research Centers | News & Media Information | People at Vanderbilt | Libraries | Administrative Departments | Medical 

  Return to the Health Psychology Home Page
  Send E-mail comments or questions to Dr. Schlundt


Search: Vanderbilt University
the Internet
  Help  Advanced
Tip: You can refine your last query by searching only the results by clicking on the tab above the search box
Having Trouble Reading this Page?  Download Microsoft Internet Explorer.