Appearance and disappearance of functional

gastrointestinal disorders in patients with eating disorders

CATHERINE BOYD,* SUZANNE ABRAHAM* & JOHN KELLOW�

*Department of Obstetrics and Gynaecology, Royal North Shore Hospital, University of Sydney, Sydney, NSW, Australia

�Department of Medicine, Royal North Shore Hospital, University of Sydney, Sydney, NSW, Australia

Abstract

Background Functional gastrointestinal disorders or

‘functional gastrointestinal disorder-like’ symptoms

(FGIDs) occur commonly in eating disorders (ED), but

it is not known if these disorders are stable over time.

The aims were to evaluate the turnover of FGIDs in

patients with ED, and to relate this turnover to

changes in body mass index (BMI), ED behaviors,

and psychological variables. Methods Patterns and

repeated measures analysis of presence of individual

FGIDs and regional FGID categories (esophageal,

gastroduodenal, bowel, and anorectal) in ED patients

(n = 73) at admission to hospital and at 12-month

follow-up, using change in BMI and ED behaviors as

between patient variables. Key Results Functional

gastrointestinal disorders prevalence was 97% at

admission and 77% at follow-up. The only individual

FGIDs to decrease over time were functional heart-

burn (admission 53%, follow-up 23%) and functional

dysphagia (21%, 7%). There was significant patient

variation in the disappearance, persistence, and

appearance of both individual FGIDs and FGID

regional categories. Twenty-five (34%) of patients

acquired at least one new FGID regional category at

follow-up. There was no relationship between changes

in BMI, self-induced vomiting, laxative use, binge

eating, anxiety, depression, somatization, and the

turnover of individual or regional FGIDs. Conclusions

& Inferences Functional gastrointestinal disorders

remain common after 12 months in patients with an

ED. Considerable turnover of the FGIDs occurs, how-

ever, and the appearance of new FGIDs is not

restricted to the original FGID regional category.

There is no apparent relationship between the turn-

over of the FGIDs and ED behaviors, psychological

variables or body weight change. These findings have

implications for the clinical evaluation and manage-

ment of FGIDs in ED patients.

Keywords disappearance and appearance, eating

disorder behaviors, eating disorders, functional

gastrointestinal disorders.

INTRODUCTION

Functional gastrointestinal disorders or ‘functional

gastrointestinal disorder-like’ symptoms (FGIDs) are a

common feature of eating disorder (ED) patients, with a

prevalence of over 90% among anorexia nervosa,

bulimia nervosa or ED not otherwise specified

patients.1,2 Moreover, in FGID patients, a higher than

expected proportion (16%) report a past history of EDs.3

There have been few other investigations of the

associations between FGIDs and EDs.

It could be postulated that as people recover from an

ED, the prevalence of FGIDs may decrease, as many of

the weight-losing behaviors associated with the

ED—particularly those behaviors that may have pro-

voked changes in gastrointestinal function—improve.

Such behaviors include food and fluid restriction, self-

induced vomiting, laxative use and binge eating,

leading to slow gut transit, constipation, dehydration

and abdominal bloating, respectively. Indeed, Chami

et al.4 studied ED patients both at admission and at

discharge from hospital, and found an improvement in

gastrointestinal symptoms even in this short time

period, suggesting an association between hospital

treatment and symptom improvement. On the other

hand, if the FGIDs in EDs are related more to brain–gut

dysregulation intrinsic to the EDs, with accompanying

alterations in gut motility and sensitivity (as with

FGIDs in non-ED patients), then the FGIDs may be

more persistent over time. A longer term study is

Address for Correspondence

Prof. Suzanne Abraham, Department of Obstetrics andGynaecology, Building 52, Royal North Shore Hospital,St Leonards, Sydney, NSW 2065, Australia.Tel: 61 2 99268608; fax: 61 2 94363719;e-mail: sabraham@med.usyd.edu.auReceived: 14 January 2010Accepted for publication: 16 June 2010

Neurogastroenterol Motil (2010) 22, 1279–1283 doi: 10.1111/j.1365-2982.2010.01576.x

� 2010 Blackwell Publishing Ltd 1279

required to establish the stability of the FGIDs in EDs,

and to determine the relationship between the pres-

ence of FGIDs and body mass index (BMI) and ED

behaviors.

The aims of this study were therefore (i) to examine

the prevalence of individual FGIDs and the main FGID

regional categories (esophageal, gastroduodenal, bowel,

and anorectal disorders) in ED patients at admission to

hospital for specialized treatment and again after

12 months; (ii) to examine the turnover of the FGIDs,

in other words the appearance, persistence and disap-

pearance of the FGIDs and FGID regional categories in

ED patients after 12 months; and (iii) to evaluate the

relationship between the appearance and disappearance

of FGIDs, and changes in BMI, behaviors associated

with the ED, and psychological variables, after

12 months.

MATERIALS AND METHODS

Patients

A total of 108 consecutive female ED patients were contacted12 months after discharge from the Eating Disorder Unit at theNorthside Clinic. All patients fulfilled the DSM-IV criteria for anED.5 Inclusion criteria were no major medical illness, no othermajor psychiatric disorder, and not pregnant. Patients were askedto complete the same questionnaires1 they had completed onadmission to the Unit. The questionnaires evaluated: (i) thepresence of FGIDs, using the validated and reliable Rome IIquestionnaire,6 (ii) demographic details and ED behaviors, usingthe computerized eating and exercise examination (EEE) and theeating attitudes test (EAT),7,8 and (iii) psychological question-naires, namely the Beck Depression Inventory,9 the State-traitAnxiety Inventory10 and the somatization subscale from the BriefSymptom Inventory.11 All patients underwent routine clinicalevaluation including blood tests (hematology, biochemistry, andthyroid function) and specific investigations to exclude organicgastrointestinal disease where appropriate. Pateints did not rou-tinely undergo physiologic testing for a formal diagnosis of thoseFGIDs requiring such testing. The study was approved by theNorthside Clinic Human Ethics Committee.

Of the 108 patients, 16 were unavailable or ineligible forfollow-up (two traveling overseas, two pregnant, four inpatients inanother hospital and eight with no contact details). Nineteenpatients did not complete, or only partially completed, thequestionnaires, despite weekly reminders. The remaining 73patients (age 20 ± 5 years) participated in the study, an overallresponse rate of 73/92 (79%). Eight of the participating patientswere readmitted as inpatients during the time of follow-up; forthese patients, data obtained during that admission were used.The mean (SD) current BMI, lowest ever BMI and highest everBMI were 17.7 (3.1), 16.0 (2.2), and 22.9 (3.7) kg m)2, respectively.The average duration of ED illness was 5.3 (5.4) years. There wereno significant differences between those patients who did and didnot participate in the follow-up study, in terms of age, currentBMI, highest BMI, lowest BMI, or duration of illness at admission.

Eating disorder behaviors were reported as days of binge eating,self-induced vomiting and laxative use in the previous 28 days.7

Objective binge eating was defined as episodes of overeating thatthe individual considered were outside of their control and when

the amount of food eaten during each binge was greater than fiveserves (one serve equivalent to three slices of bread).7 Behaviorswere categorized as present if they occurred on more than 4 daysin the previous 28 days. Low BMI was defined asBMI < 17.5 kg m)2, and EAT cutoff score suggestive of thepossible presence of an ED was >30.8

Data and statistical analysis

Simple comparisons using paired t-tests or Chi-square (withFishers Exact) were undertaken as applicable. Rates of FGIDdisappearance (i.e. FGID present at admission but not follow-up),persistence (FGID present at both admission and follow-up), andappearance (FGID present at follow-up but not admission) weredetermined. Repeated measures analysis of variance (ANOVA) atadmission and follow-up was undertaken, using change in BMI,self-induced vomiting, laxative use, and objective binge eating asbetween subject variables to test differences between patients anddifferences over time. Data were analyzed for both individualFGIDs and the main FGID regional categories (esophageal,gastroduodenal, bowel, and anorectal disorders). Spearman corre-lation was used to compare the number of FGID regionalcategories at both time points.

RESULTS

Changes in BMI, ED behaviors, EAT score, andpsychological variables between admission andfollow-up

Frequency of patients with BMI <17.5 kg m)2, ED

behaviors present, EAT score >30, and psychological

variables at admission and 12-month follow-up are

shown in Table 1. There was a significant improve-

ment in all variables, with the exception of objective

binge eating, at 12 months.

Changes in FGID prevalence between admissionand follow-up

Seventy-one (97%) of patients had at least one FGID

(range 0–7) at admission compared to 56 patients (77%)

at 12-month follow-up (range 0–5). Overall, there was a

significant decrease in the number of FGIDs present

from admission to follow-up (admission 2.8 ± 1.3,

follow-up 1.7 ± 1.4, t = 5.581, P < 0.001). The preva-

lence of the FGID regional categories at admission and

at 12 months were: esophageal disorders 73% and 34%,

gastroduodenal disorders 32% and 18%, bowel disor-

ders 81% and 66%, anorectal disorders 33% and 19%.

Appearance, persistence, and disappearance ofFGIDs between admission and follow-up

Table 2 shows the disappearance, persistence, and

appearance of FGID regional categories and individual

FGIDs between admission and 12-month follow-up.

C. Boyd et al. Neurogastroenterology and Motility

� 2010 Blackwell Publishing Ltd1280

For each of the FGID categories, there were varying

rates of FGID turnover. Repeated measures ANOVA

showed that there were significant differences between

patients (F = 47.504, df = 9, P < 0.001) and from admis-

sion to follow-up (F = 5.494, df = 9, P < 0.001). For

individual FGIDs, however, only two showed signifi-

cant changes over time, namely functional heartburn

(F = 20.504, df = 1, P < 0.001) and functional dysphagia

(F = 7.819, df = 1, P < 0.007). For FGID regional cate-

gories, only esophageal disorders were significantly

different over time (F = 15.357, df = 1, P < 0.001).

Twenty-five (34%) of patients acquired at least one

new FGID regional category at follow-up. The appear-

ance at follow-up of one or more new FGID categories

in patients who had recovered from their index admis-

sion FGID is shown in Fig. 1. In these patients, it was

more common, with the exception of the bowel

disorders, to move to another FGID regional category

than to remain in the index category. Moreover, there

was no predilection for the appearance of a new

individual FGID within a particular FGID regional

category (data not shown). There were no significant

correlations between the FGID regional categories at

admission and follow-up, except for a correlation

between anorectal disorders at admission and follow-

up (Spearman correlation coefficient = 0.275, P < 0.02).

Relationship between changes in FGIDs andchanges in BMI, ED behaviors, and psychologicalvariables

Neither change in BMI nor change in ED behaviors

(self-induced vomiting, laxative use, and objective

binge eating), and psychological variables (anxiety,

Table 2 Changes in occurrence of functional gastrointestinal disorders at admission and 12-month follow-up in 73 eating disorder patients

FGID disappearance* FGID persistence� FGID appearance�

N % N % N %

Esophageal disorders

Functional heartburn 26 36 13 18 4 6

Functional dysphagia 12 16 3 3 2 3

Gastroduodenal disorders

Functional dyspepsia 12 16 1 1 7 10

Aerophagia 8 11 2 3 5 7

Bowel disorders

Irritable bowel 17 23 16 22 11 12

Functional abdominal bloating 12 16 10 14 9 10

Functional constipation 11 15 5 5 7 10

Anorectal disorders

Functional anorectal pain 7 10 4 6 6 8

Proctalgia fugax 9 12 4 6 4 6

FGID, functional gastrointestinal disorders. *FGID present at admission and not follow-up. �FGID present at both admission and follow-up. �FGID

present at follow-up but not admission.

Table 1 Number of patients with low body mass index, eating disorder behaviors, Eating Attitudes Test score > 30, and psychological variables, at

admission to hospital and at 12-month follow-up, in a sample of 73 eating disorder patients

Admission Follow-upChi-square test*

df = 1 P-valueN % N %

BMI < 17.5 kg m)2 40 55 18 25 11.210 <0.001

Binge eating > 4 days� 33 40 11 15 0.409 ns

Self-induced vomiting > 4 days� 34 47 14 19 14.914 <0.001

Laxative use > 4 days� 17 23 3 4 10.306 <0.001

Eating Attitudes Test > 30� 65 89 38 52 9.755 <0.002

Mean SD Mean SD Paired t-test

Beck Depression Inventory 29 11 17 15 7.366 <0.001

Somatization 12 5 6 6 6.927 <0.001

Speilberger trait anxiety 60 11 51 14 5.653 <0.001

Speilberger state anxiety 59 11 47 15 7.573 <0.001

N, number of patients; SD, standard deviation; BMI, body mass index. *Fishers Exact for low numbers. �In past 28 days. �Number of patients who

scored >30.

Volume 22, Number 12, December 2010 Turnover of FGIDs in eating disorders

� 2010 Blackwell Publishing Ltd 1281

depression, and somatization) had a significant inter-

action with change in individual FGIDs or in FGID

regional categories from admission to 12-month

follow-up.

DISCUSSION

This is the first study to examine changes in FGIDs in

ED patients over a 12-month period. We have docu-

mented that the prevalence of FGIDs remains high

(77%), 12 months after patients with an ED received

treatment in hospital. Indeed, there was a significant

decrease in only two FGIDs—functional heartburn and

functional dyspepsia—and, consistent with this, in

only one FGID regional category, namely the esopha-

geal disorders. The only previous study to examine ED

patients more than once examined gastrointestinal

symptoms at admission and discharge from hospital,4

and found gastrointestinal symptoms decreased during

inpatient treatment, suggesting a possible association

with ED behaviors. Our prevalence data for FGIDs in

ED patients are far greater, both on admission and after

12 months, than that found in a group of Australian

females in the general population using the same

instrument, namely 36% for any FGID diagnosis.12

Our second main finding was that, although the

prevalence of FGIDs remains high, individual FGIDs

and FGID regional categories were relatively unstable in

ED patients, in other words there were varying rates of

FGID disappearance, persistence, and appearance,

between admission to hospital for treatment and at

12-month follow-up. There was no relationship

between the individual FGIDs and whether they

appeared, persisted, or disappeared. In a substantial

number of patients (34%), at least one new FGID

regional category appeared at follow-up. It was more

common, in fact, for an individual who had lost an FGID

to acquire a new FGID from a different FGID category

rather than from the same category, except for the bowel

disorders. It is of interest that this phenomenon of

instability has been found previously in non-ED

individuals, with a 9% development of IBS in initially

IBS-free people.13 This latter study found that the

overall prevalence of FGIDs remained the same over

12–20 months, but that the turnover of diagnoses was

substantial. Our findings in ED patients are consistent

with a similar degree of turnover; the prevalence of IBS

acquisition in this latter study is close to the 12% found

in our study. Other studies, however, have shown that,

for many individuals, the symptoms of FGIDs tend to

persist,14,15 while for others there are symptoms fluc-

tuations.16 An association of functional heartburn and

self-induced vomiting may be expected, but there was

no relationship of change in vomiting with FGID

change, including heartburn, over time.

The third main finding of this study was that the

prevalence and turnover of FGIDs was not related to

the substantial improvement in BMI, or to the

improvement in weight-losing behaviors or psycholog-

ical variables that had taken place in the ED patients at

12-month follow-up. Most patients were in fact recov-

ering from their ED at 12 months, according to the

EAT. This lack of FGID relationship with ED improve-

ment suggests that the FGIDs in EDs behave more like

those in non-ED patients and seems unlikely to be

related to the physiologic aspects of active weight-

losing behaviors characteristic of EDs. Thus, it is

feasible that mechanisms similar to the FGIDs in non-

ED patients may be in operation in ED patients.

In conclusion, FGIDs remain common at 12-month

follow-up in patients with an ED. Similar to the

presence of FGIDs in non-ED patients, however, there

is considerable turnover of FGIDs at this time point. In

addition, the appearance of new FGIDs is not restricted

to the original (admission) FGID regional category,

although the bowel disorders category exhibits more

consistency in this respect. Finally, there is no appar-

ent relationship between the turnover of the FGIDs

and the ED behaviors of self-induced vomiting,

laxative use or binge eating, body weight change, or

psychological variables. These findings have implica-

tions for the clinical evaluation and management of

FGIDs in ED patients, and for future research studies.

COMPETING INTERESTS

The authors have no competing interests.

Figure 1 Presence of a new functional gastrointestinal disorder (FGID)

at 12-month follow-up in 73 eating disorder patients, according to

whether the FGID was from the same or a different admission (index)

FGID regional category (i.e esophageal, gastroduodenal, bowel, ano-

rectal). Note it was more common for the new FGID to be acquired

from a different FGID regional category than that present at admission.

C. Boyd et al. Neurogastroenterology and Motility

� 2010 Blackwell Publishing Ltd1282

REFERENCES

1 Boyd C, Abraham S, Kellow J. Psy-chological features are importantpredictors of functional gastrointesti-nal disorders in patients with eatingdisorders. Scand J Gastroenterol

2005; 40: 929–35.2 Perkins SJ, Keville S, Schmidt U,

Chalder T. Eating disorders and irri-table bowel syndrome: is there a link?J Psychosom Res 2005; 59: 57–64.

3 Porcelli P, Leandro G, Carne MD.Functional gastrointestinal disordersand eating disorders: relevance of theassociation in clinical management.Scand J Gastroenterol 1998; 33:577–82.

4 Chami TN, Andersen AE, CrowellMD, Schuster MM, Whitehead WE.Gastrointestinal symptoms in buli-mia nervosa: effects of treatment. Am

J Gastroenterol 1995; 90: 88–92.5 American Psychiatric Association:

Diagnostic and Statistical Manual of

Mental Disorders, 4th edn. Washing-ton, DC: American Psychiatric Asso-ciation, 1994.

6 Drossman DA, ed. Rome II: The

Functional Gastrointestinal Disor-

ders, 2nd edn. McLean, VA: DegnonAssociates, 2000.

7 Abraham S, Lovell N. Research andclinical assessment of eating andexercise behaviour. Hospital Med1999; 60: 481–5.

8 Garner DM, Garfinkel PE. The EatingAttitudes Test: an index of thesymptoms of anorexia nervosa. Psy-chol Med 1979; 9: 273–9.

9 Beck AT, Ward CH, Mendelson M,Mock J, Erbaugh J. An inventory formeasuring depression. Arch Gen

Psychiatry 1961; 4: 61–71.10 Spielberger CD, Gorsuch RL, Lushene

RE. Manual for the State-Trait Anxi-ety Inventory. Palo Alto, CA: Con-sulting Psychologisys Press Inc.,1970.

11 Derogatis LR, Melisaratos N. Thebrief Symptom Inventory: an intro-ductory report. Psychol Bull 1983; 13:595–605.

12 Boyce PM, Talley NJ, Burke C, Kolo-ski NA. Epidemiology of the func-

tional gastrointestinal disordersdiagnosed according to Rome II cri-teria: an Australian population-basedstudy. Int Med J 2006; 36: 28–36.

13 Talley NJ, Weaver AL, ZinsmeisterAR, Melton JL. Onset and disappear-ance of gastrointestinal symptomsand functional gastrointestinaldisorders. Am J Epidemiol 1992; 136:165–77.

14 Koloski NA, Talley NJ, Boyce PM.Does psychological distress modulatefunctional gastrointestinal symptomsand health-care seeking? A prospec-tive, community cohort study.Am J Gastroenterol 2003; 98: 789–97.

15 Owens DM, Nelson DK, Talley NJ.The irritable bowel syndrome: long-term prognosis and the physician-patient interaction. Ann Inter Med

1995; 122: 107–12.16 Agreus L, Svardsudd K, Talley NJ,

Jones MP, Tibblin G. Natural historyof gastroesophageal reflux diseaseand functional abdominal disorders:a population-based study. Am JGastroenerol 2001; 96: 2905–14.

Volume 22, Number 12, December 2010 Turnover of FGIDs in eating disorders

� 2010 Blackwell Publishing Ltd 1283