CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2009;7:317–322

Relationship Between Symptoms and Dietary Patterns in Patients With

Functional Dyspepsia

AMELIA N. PILICHIEWICZ,*,‡ MICHAEL HOROWITZ,*,‡ GERALD J. HOLTMANN,‡,§ NICHOLAS J. TALLEY,储,¶ and

CHRISTINE FEINLE–BISSET*,‡
*University of Adelaide Discipline of Medicine, §Department of Gastroenterology and Hepatology, Royal Adelaide Hospital, and ‡National Health and Medical Research
Council, Centre of Clinical Research Excellence in Nutritional Physiology, Outcomes and Interventions, Adelaide, Australia; 储Mayo Clinic College of Medicine,
Jacksonville, Florida; and the ¶Department of Medicine, University of Sydney and Nepean Hospital, Sydney, Australia

Background & Aims: Patients with functional dyspepsia ported to be greater (by about 9%), and the number of larger
(FD) often report that their symptoms are related to food meals to be lower (by about 24%), in FD patients compared with
ingestion. However, there is a lack of information about healthy subjects, but these differences were not statistically
differences in eating patterns and nutrient intake between significant.2 Furthermore, in that study only 55% of FD patients
these patients and healthy individuals or the association consumed 3 meals per day compared with 80% of healthy
with specific symptoms. We performed a prospective trial to subjects,2 although interpretation of these data is difficult be-
evaluate the relationship between FD symptoms and dietary cause definitions of “meals” and “snacks” were not provided.2 In
factors. Methods: Twenty patients with FD (17 women) contrast, a study using food-frequency diaries reported no dif-
and 21 healthy subjects (18 women) completed detailed diet ferences in eating patterns, including the number of regular
diaries, recording all foods eaten, drinks consumed, and meals and total number of eating episodes between patients
times of consumption, as well as the occurrence, timing, with FD and healthy subjects;4 however, food intake was eval-
and severity of dyspeptic symptoms (ie, nausea, discomfort, uated only in food categories. A further study, which used a
fullness, bloating, upper-abdominal/epigastric pain) for 7 7-day diet history to assess nutrient intake, found no significant
days. Data from the diet diaries were analyzed for the differences, except that subjects with FD had a significantly
number of meals, light meals, snacks and drinks, energy lower intake of dietary fiber.1 Hence, previous studies have
intake, and macronutrient distribution. Results: Patients yielded inconsistent observations with regard to eating patterns
with FD ate fewer meals (P < .01) and consumed less total in FD. Furthermore, none of these studies evaluated the rela-
energy (P ⴝ .1) and fat (P ⴝ .1) than healthy subjects. Their tionship between symptoms with eating patterns concurrently.
symptoms were modest in severity (score out of 10; 5 [range, Many FD patients report that they can eat only small meals
3– 8]) and occurred within 31 minutes (range, 8 – 64 min) of and do not tolerate fat, consistent with the outcome of labo-
eating. Fullness was related directly to the amount of fat ratory-based studies showing diminished tolerance of volume/
ingested (z, 1.91; P < .05) and overall energy intake (z, 2.12; pressure, as indicated by increased sensitivity to distension of
P < .05) and related inversely to the amount of carbohy- the proximal5–7 and distal8 stomach in approximately 35% of
drate ingested (z, ⴚ1.9; P ⴝ .05). Similarly, bloating was patients, reduced ingestion capacity during an oral nutrient
related to the amount of fat ingested (z, 1.68; P ⴝ .09). drink test in approximately 40% of patients,9,10 and increased
There was no significant relationship between symptom sensitivity to oral and duodenal fat administration in approxi-
severity and any of the dietary variables measured. mately 60% to 70% of patients;7,11,12 and our recent study
Conclusions: Management of patients with FD might be showed that in FD patients, a high-fat test-meal induces more
improved by instructing them to consume smaller meals nausea and pain than a high-carbohydrate meal.13
with reduced fat content. Patients with FD as a group have more life and psychologic
distress, as well as a lower quality of life than healthy individ-

T he pathophysiology and etiology of functional dyspepsia

(FD) remain poorly defined. Many patients report their
symptoms are related to the ingestion of food,1,2 however, few
uals.14,15 It would not be surprising if eating attitudes in FD
were abnormal and impacted on eating behavior and/or symp-
toms and were related to food intake.
studies have investigated this putative association. Two studies The aims of this study were to quantify eating habits and
reported that dyspeptic symptoms were associated with inges- dyspeptic symptoms, as well as relationships between them, in
tion of onions, peppers, fried and fatty foods, alcohol, citrus FD. We hypothesized the following: (1) that patients with FD
fruits, and spicy foods,1,2 or rich cakes and carbonated bever- would consume smaller meals and experience more meal-asso-
ages.2 Another study, using a questionnaire about 39 different ciated symptoms, but eat more frequently when compared with
foods, reported that FD patients identified 22 of these as ag- healthy subjects, and (2) that symptoms would be related di-
gravating their symptoms, and, therefore, avoided them.3 Of
these foods, the highest rates of intolerance were for mayon-
Abbreviations used in this paper: BMI, body mass index; CCK, cho-
naise (80%), nuts (70%), fish (66%), and chocolate (62%)—3 out lecystokinin; FD, functional dyspepsia; NDI, Nepean Dyspepsia Index.
of these 4 foods have a high fat content. © 2009 by the AGA Institute
It is uncertain whether eating patterns and nutrient intake 1542-3565/09/$36.00
are different in FD. The prevalence of snacking has been re- doi:10.1016/j.cgh.2008.09.007
318 PILICHIEWICZ ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 7, No. 3

rectly to the amount eaten, as well as the amount of fat in the (with 0 representing symptom not present and 10 representing
diet. symptom most severe), and the time at which these symptoms
occurred.
Patients and Methods Measurements
Subjects Diet diaries. The total number of consecutive eating/
Twenty FD patients (17 women, 3 men; age, 45 ⫾ 3 y; drinking episodes for the week was determined. These episodes
range, 23–73 y; body mass index [BMI], 24.0 ⫾ 0.9 kg/m2; range, were classified into 5 categories: (1) meals, (2) light meals, (3)
19.3–35.7 kg/m2) were recruited through advertisements in a snacks, (4) caloric drinks, and (5) noncaloric drinks. Breakfast,
local newspaper (n ⫽ 3) and from the endoscopy list and lunch, and dinner generally were classified as either meals or
outpatients of the Department of Gastroenterology and Hepa- light meals. Meals were defined as the main eating occasion(s) of
tology at the Royal Adelaide Hospital (n ⫽ 17). The patients the day and comprised foods traditionally eaten as a main meal
had to have experienced postprandial fullness, bloating, epigas- (eg, continental breakfast, pasta, meat and vegetables, 4 slices of
tric pain, nausea, or vomiting for more than 3 months of at pizza, and so forth). Light meals were defined as episodes in
least a moderate severity, according to the Rome II criteria which the amount ingested was less than expected at a main
(Rome III criteria were not available at the time the study was meal (by ⬃30%) (eg, cereals, sandwiches, 2 slices of pizza, salads
initiated).16 Severity was scored on a 0 to 3 scale, with 0 repre- containing meat). Snacks were defined as eating episodes in
senting symptom not experienced, 1 representing slight symp- which consumption was less and which were easier to prepare
tom, 2 representing moderate symptom, bothering, but not than meals or light meals, usually consisting of one food item
impairing daily activities, and 3 representing severe symptom, (eg, biscuits, chocolate, crisps, fruit, yogurt, slice of pizza) and
impairing daily activities. Subjects were eligible if the total score consumed at times between meals. A caloric drink was defined as
was 3 or greater, or one symptom scored 2 or greater. Exclusion any nutrient-containing beverage (eg, cola, coffee with milk
criteria included any organic cause of dyspepsia, positive ultra- and/or sugar, juice, milkshakes, alcohol). A noncaloric drink was
sound and upper-gastrointestinal endoscopy performed during defined as any beverage that contained no calories (eg, water,
the previous 12 months, heartburn or lower-bowel symptoms as diet drinks, black coffee/tea).
predominant symptoms, positive Helicobacter pylori status, and Foods consumed within a time interval of less than 15
clinically significant positive results on laboratory tests, includ- minutes were considered to be part of one eating episode.
ing biochemistry, hematology, and lactose breath tests. Intervals between eating episodes during the day and night also
Twenty-one healthy subjects (18 women, 3 men; age, 40 ⫾ were calculated.
4 y; range, 20 –74 y; BMI, 22.9 ⫾ 0.5 kg/m2; range, 18.6 –26.5 Symptom diaries. Symptoms were divided into 3 cat-
kg/m2) were included as controls. Inclusion criteria were age, egories: meal-associated symptoms (bloating, nausea, upper-
sex, and BMI matching those of a patient (FD patients were abdominal pain, belching, epigastric pain, fullness, vomiting, dis-
recruited first), although 6 FD patients, but only 3 healthy comfort), which occurred within a 2-hour period after completion
subjects, were overweight (BMI ⬎ 25), and 1 FD patient was of eating;8 other gastrointestinal symptoms, which was any
obese (BMI ⬎ 30). Exclusion criteria included dyspeptic symp- other symptom(s) (heartburn, diarrhea, constipation, lower-
toms or any significant illnesses. abdominal pain) that occurred within a 2-hour period after
Patients and healthy subjects also were excluded if they were eating; or meal-unrelated, which was any symptom that oc-
taking medication known to affect gastrointestinal motility or curred more than 2 hours after, or before, an eating episode (eg,
appetite, smoked, or habitually consumed more than 20 g of when the subject awoke). Symptom severity was classified as
alcohol per day. Although 9 of 20 patients reported perceived follows on a scale of 1 to 10: mild (score, 1–3), not influencing
intolerances of certain foods, including caffeine (n ⫽ 2), fatty usual activities; modest (score, 4 –7), diverting from, but not
foods (n ⫽ 6), dairy products (n ⫽ 3), and others, including requiring modification of, daily activities; or strong (score,
citrus fruits, cabbage, and onions, none of the subjects reported 8 –10), impairing daily activities. The times at which these
any food allergies or gluten intolerance. symptoms occurred were evaluated.
Behavioral questionnaires. To obtain behavioral
Protocol information on each patient and healthy subject, a number of
Each patient and healthy subject completed a diet diary, questionnaires were completed. These included the following:
in which they recorded all foods and drinks consumed, and the (1) Three-Factor Eating questionnaire, which measures 3 di-
time of each eating or drinking episode, over an entire 7-day mensions of eating behavior: cognitive restraint, disinhibition
period (ie, 5 weekdays and 2 weekend days),17 while maintaining of eating, and hunger,18 (2) Eating Attitudes Test, which iden-
their usual eating habits. For this purpose, subjects were asked tifies individuals with disordered eating,19 (3) Northwest Lipid
to weigh as many foods as practical, or, alternatively, to use cup Research Clinical Fat Intake Scale, which estimates fat intake,20
or spoon measures or common serves (eg, slice of bread), and to (4) Nepean Dyspepsia Index (NDI), which assesses changes in
be as specific as possible (eg, type of bread: white/wholemeal: symptoms and quality of life,21 (5) Eysenck Personality Ques-
degree of fat trimming of meat: type of margarine or oil; type of tionnaire,22 and (6) Hospital Anxiety and Depression scale,23
milk; whole-fat or skim); and type of cooking method; fried, which measures personality and psychologic distress.
boiled, roasted. If a recipe was followed the subject was in-
structed to record it in the food diary. The diary also evaluated Data and Statistical Analysis
symptoms, where all subjects recorded any symptoms experi- Questionnaire and eating behavior data were compared
enced (abdominal pain, cramps, bloating, nausea, uncomfort- between FD patients and healthy subjects using the Mann–
able fullness after meals), their severity on a scale out of 10 Whitney U test. To evaluate relationships between symptoms
March 2009 DIET AND SYMPTOMS IN FUNCTIONAL DYSPEPSIA 319

Table 1. Number of Eating and Drinking Episodes, Meals, Table 3. Occurrence and Severity of Individual
Light Meals, Snacks, and Caloric and Noncaloric Meal-Associated Symptoms and the Timing of
Drinks During a 7-Day Period in FD Patients and Their Occurrence After Meals
Healthy Subjects
Occurrence, % Severity (out of 10) Timing, min
FD patients Healthy subjects P value
Bloating 28 (7–57) 5 (2–9) 30 (1–94)
Total number of 46 (25–96) 44 (21–84) .8 Nausea 11 (4–51) 5 (3–8) 37 (10–95)
episodes Upper-abdominal 11 (2–72) 6 (3–8) 25 (6–60)
Eating episodes 32 (18–40) 29 (21–46) .9 pain
Drinking episodes 14 (1–57) 10 (0–39) .6 Belching 37 (3–41) 4 (3–8) 40 (12–59)
Meals 5 (1–10) 7 (2–16) ⬍.01 Epigastric pain 28 (5–97) 4 (3–6) 31 (2–105)
Light meals 13 (4–19) 10 (4–17) .3 Fullness 20 (3–86) 6 (2–7) 14 (1–68)
Snacks 15 (3–25) 11 (0–29) .8 Vomiting 3 (3–3) 8 (8–8) 1 (1–1)
Caloric drinks 4 (1–36) 3 (0–30) .2 Discomfort 22 (14–50) 5 (4–7) 45 (16–60)
Noncaloric drinks 9 (0–21) 5 (0–29) .7
% meals 12 (2–26) 15 (6–52) .06 NOTE. Data are shown as median (range) for 20 FD patients.
% light meals 28 (6–57) 21 (10–57) .4
% snacks 30 (8–44) 27 (0–54) .8
% caloric drinks 10 (2–46) 7 (0–46) .2 interval between meals (FD: 162 min; range, 100 –320 min;
% noncaloric drinks 17 (0–36) 10 (0–48) .7 healthy subjects: 172 min; range, 120 –347 min), or the duration
NOTE. Data are shown as median (range) for 20 FD patients and 21 of the overnight fast (FD: 733 min; range, 450 –905 min; healthy
healthy subjects. subjects: 669 min; range, 527–910 min) between FD patients
and healthy subjects.

and variables of food intake, because of repeated measures over Symptom Assessment
time and day, food intake and occurrence/severity were mod- No healthy subject experienced any symptoms during
eled with a generalized estimating question, with multinomial the 7-day assessment. FD patients reported a total of 612
distribution and a cumulative log link. Relationships between symptoms, or 26 (range, 1–92) per patient; 64% (range, 0%–
NDI subscales, Hospital Anxiety and Depression scores, and 100%) of these were meal-associated with a severity of 5 (range,
eating attitudes with food intake variables were analyzed using 3– 8) and occurring 31 minutes (range, 8 – 64 min) after eating,
the Spearman rho correlations. Statistical significance was ac- 9% (range, 0%– 43%) were other gastrointestinal symptoms with
cepted at a P value of less than .05, and data are presented as a severity of 5 (range, 3– 8) and occurring 25 minutes (range,
medians (ranges). 5–120 min) after eating, whereas 14% (range, 0%–36%) were
meal-unrelated with a severity of 5 (range, 2–7) and occurring
135 minutes (range, ⫺180 – 0 and 120 –200 min) before, or
Results
after, eating. Occurrence, severity, and timing of individual
Eating Behavior meal-associated symptoms are summarized in Table 3.
The number of meals was less in FD patients when com-
pared with healthy subjects (P ⬍ .01), and in healthy subjects Eating Attitudes, Gastrointestinal
reflected average Australian eating habits of one main meal per Symptoms, Quality of Life, Personality, and
day. There were no other significant differences (Table 1). There Psychologic Distress
was also no difference in the total weight, carbohydrate, pro- There was a significant difference in eating attitudes
tein, or alcohol content of foods consumed, or the percentage between FD patients and healthy subjects (P ⬍ .01). Three of 20
of macronutrient distribution, although fat and energy intake FD patients (15%), but only 1 healthy subject (5%), scored
tended to be less in the patients (P ⫽ .1) (Table 2), the time greater than 30 on this scale, indicating disordered eating

Table 2. Energy Intake, Amount Eaten, and Macronutrient Distribution of Fat, Carbohydrate, Protein, and Alcohol During a
7-Day Period in FD Patients and Healthy Subjects
FD patients Healthy subjects P value

Weight, g 16,404 (7827–31,454) 18,825 (6683–22,491) .6

Energy, kJ 48,131 (17,844–97,548) 56,570 (34,780–78,997) .1
Fat, g 479 (89–798) 564 (301–703) .1
Carbohydrate, g 1337 (629–2459) 1578 (856–2496) .2
Protein, g 511 (237–1051) 531 (335–846) 1.0
Alcohol, g 14 (0–49) 19 (0–28) .6
% fat 29.5 (15.3–45.0) 28.3 (16.0–38.3) .9
% carbohydrate 49.6 (25.8–60.4) 51.2 (37.0–62.7) .2
% protein 17.8 (10.3–25.1) 16.0 (12.5–22.7) .2
% alcohol 0.8 (0–2.6) 1.3 (0–1.9) 1.0

NOTE. Data are shown as median (range) for 20 FD patients and 21 healthy subjects.
320 PILICHIEWICZ ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 7, No. 3

Table 4. Scores for Eating Attitudes, Quality of Life, and Personality/Psychologic Distress in FD Patients and
Healthy Subjects
FD patients Healthy subjects P

Eating attitudes
Three-Factor eating questionnaire
(1) Cognitive restraint of eating (cut off, 12; maximum, 21) 7 (1–18) 6 (1–14) .2
(2) Disinhibition of eating (maximum, 16) 4 (0–13) 5 (1–10) .4
(3) Hunger (maximum, 14) 4 (0–11) 4 (1–10) .3
Eating attitudes test (cut off, 30) 15 (5–47) 9 (1–31) ⬍.01
Fat intake scale (maximum, 44) 26 (15–35) 27 (21–40) .5
Upper abdominal symptoms and quality of life (NDI)
Symptoms (maximum, 195) 56 (6–132) 4 (0–31) ⬍.0001
Interference/difficulty with activities (maximum, 65) 28 (17–46) 13 (13–13) ⬍.0001
Lack of knowledge or control (maximum, 35) 16 (10–28) 7 (7–10) ⬍.0001
Eating/drinking disturbances (maximum, 15) 8 (4–14) 3 (3–5) ⬍.0001
Sleep disturbances (maximum, 10) 4 (0–10) 2 (2–2) ⬍.001
Personality/psychologic distress
Eysenck Personality Questionnaire
Neuroticism (maximum score, 12) 4 (0–12) 4 (0–9) .8
Extroversion (maximum score, 12) 8 (0–11) 7 (2–12) .9
Hospital Anxiety and Depression scale
Anxiety (cut off, 11) 14 (7–23) 12 (8–17) .3
Depression (cut off, 11) 11 (7–20) 8 (7–16) .08

NOTE. Data are shown as median (range) for 20 FD patients and 21 healthy subjects.

behavior (Table 4). There were no differences in scores for the take, or the amount and macronutrient content of foods con-
Three-Factor Eating questionnaire or the fat intake scale be- sumed.
tween the 2 groups. Upper-abdominal symptoms, as assessed by
the NDI, were greater in FD patients when compared with Relationships Between Quality of Life,
healthy subjects (P ⬍ .0001). There was a significant difference Eating Attitudes, Anxiety, and Depression
in quality of life between FD patients and healthy subjects. With Food Intake
Scores for (1) interference, or difficulty with activities of daily There was a direct relationship between fat (r, 0.50; P ⬍
living or work because of dyspepsia, combined with impaired .05) and protein (r, 0.4; P ⫽ .06), and an inverse relationship
enjoyment of life and emotional well-being (P ⬍ .0001), (2) lack between carbohydrate intake (r, ⫺0.60; P ⬍ .01), with scores for
of knowledge of, and control over, the illness (P ⬍ .0001), (3) disturbance in eating and drinking, as measured by the NDI, in
disturbances in eating and drinking (P ⬍ .0001), and (4) sleep FD patients. There were no other significant relationships.
disturbances (P ⬍ .001), were greater in FD patients when
compared with healthy subjects, indicating poorer quality of
life in the patients. There was no difference in personality (ie, Discussion
neuroticism and extroversion) between the 2 groups. Both We formally evaluated eating behavior in FD patients
FD patients and healthy subjects presented with the same and healthy subjects and the relationships between symptoms
degree of anxiety, however, depression tended to be higher in with food intake. The major findings in FD patients are as
FD patients (P ⫽ .08). follows: (1) symptoms are related to food intake, specifically,
fullness with both fat and energy intake and bloating with fat
Relationships Between Symptoms and intake; (2) the number of full meals consumed is less, with no
Food Intake differences in the number of other eating episodes, or the
Only relationships between meal-associated symptoms weight or macronutrient content consumed; (3) fat and energy
with dietary factors were analyzed because the number of other intake tended to be less; (4) symptoms occur within 30 minutes
and meal-unrelated symptoms was low. of eating and usually at a modest severity; and (5) fat and
The occurrence of overall meal-associated symptoms was protein intake apparently affect quality of life adversely.
related positively to energy intake (z, 2.02; P ⬍ .05), and in- This study established a relationship between symptom oc-
versely to carbohydrate ingestion (%: z, ⫺2.08, P ⬍ .05). There currence and food intake. Overall symptoms were related di-
was no relation between overall symptoms with fat, protein, rectly to energy intake and inversely to the ingestion of carbo-
alcohol (g and %), or weight consumed. For individual symp- hydrate, fullness and bloating to the intake of fat, and fullness
toms, fullness was related directly to fat (absolute and %: z, 1.91; to energy intake. These observations are consistent with those
P ⬍ .05), protein (absolute: z, 2.64, P ⬍ .001; %: z, 1.82, P ⫽ .06) of a laboratory study in which duodenal infusion of fat, but not
and energy intake (z, 2.12, P ⬍ .05), and inversely to carbohy- glucose, exacerbated symptoms in FD.6 Moreover, in this pre-
drate consumption (%: z, ⫺1.9; P ⫽ .05). Bloating was related to vious study, duodenal infusion of fat increased the sensitivity to
the ingestion of fat (z, 1.68; P ⫽ .09). There was no significant gastric distension in FD patients, in that discomfort was per-
relationship between the severity of symptoms and energy in- ceived at lower volumes and pressures than during the control
March 2009 DIET AND SYMPTOMS IN FUNCTIONAL DYSPEPSIA 321

infusion, whereas in healthy subjects the lipid infusion reduced ever, in the current study differences were apparent that were
the sensitivity to gastric distension. In contrast, glucose infu- indicative of a higher prevalence of anorectic behaviors in the
sion reduced gastric sensitivity in both healthy subjects and FD FD group. Depression scores also were greater in patients than
patients, underlining the apparently specific role of lipid in in healthy subjects, consistent with previous observations.27,28
symptom induction. However, there was no association in this cohort between anx-
The second observation was that although there was no iety and depression with food intake, although it is possible
overall difference in the total number of eating episodes, FD that depression, anxiety, and poor quality of life explain a
patients consumed fewer full meals, although the difference particular dietary behavior or at least the increased aversion to
was small and there was considerable overlap with healthy certain foods.
subjects. Taken together, with the outcome of a previous study,2 Our study design warrants some comment. Although our
these findings suggest that FD patients may eat smaller meals subject numbers were relatively small, the resulting data, par-
in an attempt to reduce their symptoms. In the present study ticularly those relating symptoms to energy and fat intake,
there was also a trend for decreased fat and energy intake in FD appear clear-cut. Nevertheless, as mentioned, type 2 errors can-
patients. The absence of significant differences may represent a not be excluded, and the applicability of our observations to
type 2 error. Given that the occurrence of symptoms was related men remains uncertain. Furthermore, we cannot exclude that
to fat and energy intake, this may suggest an important first the dietary intake of FD patients was not influenced by their
point-of-call for dietary therapy, that is, reducing fat and energy convictions of associations between food and symptoms, even if
intake (coupled with diminished meal size) may alleviate symp- they were unaware that the end point was the relationship
toms, particularly fullness and bloating. It has been reported between symptom occurrence and food intake, or by the fact
that there are gender-specific differences in nutrient intakes, that some of the patients may have been conscious of their
that is, energy, carbohydrate, fat, and protein were reported to energy intake because they were overweight. Our study focused
be less in female FD patients, compared with healthy subjects, on the relationship between dietary intake (ie, energy and ma-
with no differences in men.2 Because the cohort that comprised cronutrients, including total fat intake) and symptoms. It is
the current study only included 3 men, it was not possible to important to recognize that other nutrient or nonnutrient
evaluate this. factors may be important in deteriorating (eg, coffee) or im-
All FD patients reported some symptoms throughout the proving (omega-3 fats) symptoms, which deserves evaluation in
week—this observation is interesting because, despite small dif- larger studies, although our data did not indicate differences in
ferences in fat and energy intake, total energy intake was not intakes of coffee or fish between healthy subjects and FD
different between FD and healthy subjects. The majority of patients.
these symptoms, as expected, were meal-associated, reflecting In summary, this study showed the following in a cohort of
the cohort recruited and rigorous inclusion criteria. These FD patients: (1) symptoms are related to fat and energy intake;
symptoms were perceived approximately 30 minutes after meal (2) FD patients eat fewer meals than healthy subjects and tend
ingestion and were of moderate severity. The effect of food towards lower fat and energy intake; (3) symptoms usually
intake on these symptoms, within this time frame, may reflect occur within 30 minutes of eating at a moderate severity; and
gastrointestinal factors, such as the release of the hormones (4) the quality of life was a determinant of fat and protein
cholecystokinin (CCK) and peptide YY. Both hormones have intake. These observations provide insights into the poten-
been shown to mediate, at least in part, the effect of meals on tial gastrointestinal mechanisms underlying symptom induc-
hunger and fullness.24,25 In a recent study13 we showed that tion; accordingly, the consumption of smaller meals, com-
fasting and postprandial CCK concentrations after a high-fat bined with a reduced fat intake, may prove beneficial in the
preload are higher in FD patients when compared with healthy management of FD.
subjects. This increase in CCK also was associated with higher
scores for nausea and pain. References
A previous study indicated that patients had the capacity to 1. Kearney J, Kennedy NP, Keeling PW, et al. Dietary intakes and
identify foods that exacerbate their symptoms and, therefore, adipose tissue levels of linoleic acid in peptic ulcer disease. Br J
avoided them;1 however, this association was anecdotal and was Nutr 1989;62:699 –706.
not assessed formally. In the present study, however, there was 2. Mullan A, Kavanagh P, O’Mahony P, et al. Food and nutrient
intakes and eating patterns in functional and organic dyspepsia.
a significant relationship between fat intake and the subscale of
Eur J Clin Nutr 1994;48:97–105.
the NDI relating to disturbances in eating or drinking. These 3. Kaess H, Kellermann M, Castro A. Food intolerance in duodenal
items included statements of “the ability to eat/drink is dis- ulcer patients, non ulcer dyspeptic patients and healthy subjects.
turbed by stomach problems,” “diet changed due to stomach A prospective study. Klin Wochenschr 1988;66:208 –211.
problems,” and “enjoyment of eating/drinking reduced due to 4. Cuperus P, Keeling PW, Gibney MJ. Eating patterns in functional
stomach problems.” This observation is consistent with the dyspepsia: a case control study. Eur J Clin Nutr 1996;50:520 –
concept that patients are aware that eating is the cause of their 523.
symptoms, as suggested by their responses to the questionnaire 5. Mertz H, Fullerton S, Naliboff B, et al. Symptoms and visceral
at baseline. Although fat intake was less (albeit nonsignifi- perception in severe functional and organic dyspepsia. Gut 1998;
42:814 – 822.
cantly) in the FD patients, the differences were small, suggest-
6. Barbera R, Feinle C, Read NW. Nutrient-specific modulation of
ing that FD patients may not be able to identify foods with a gastric mechanosensitivity in patients with functional dyspepsia.
high fat content precisely, which, in turn, may have major Dig Dis Sci 1995;40:1636 –1641.
implications for dietary counseling. It has been reported that 7. Barbera R, Feinle C, Read NW. Abnormal sensitivity to duodenal
eating attitudes, as assessed by the eating attitudes test, did not lipid infusion in patients with functional dyspepsia. Eur J Gastro-
differ between patients with IBS and healthy subjects.26 How- enterol Hepatol 1995;7:1051–1057.
322 PILICHIEWICZ ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 7, No. 3

8. Caldarella MP, Azpiroz F, Malagelada JR. Antro-fundic dysfunc- 22. Eysenck SB, Eysenck HJ. An improved short questionnaire for the
tions in functional dyspepsia. Gastroenterology 2003;124: measurement of extraversion and neuroticism. Life Sci 1964;3:
1220 –1229. 1103–1109.
9. Tack J, Piessevaux H, Coulie B, et al. Role of impaired gastric 23. Zigmond AS, Snaith RP. The hospital anxiety and depression
accommodation to a meal in functional dyspepsia. Gastroenter- scale. Acta Psychiatr Scand 1983;67:361–370.
ology 1998;115:1346 –1352. 24. Gutzwiller JP, Drewe J, Ketterer S, et al. Interaction between CCK
10. Tack J, Caenepeel P, Fischler B, et al. Symptoms associated with and a preload on reduction of food intake is mediated by CCK-A
hypersensitivity to gastric distention in functional dyspepsia. receptors in humans. Am J Physiol 2000;279:R189 –R195.
Gastroenterology 2001;121:526 –535. 25. Degen L, Oesch S, Casanova M, et al. Effect of peptide YY3-36
11. Houghton LA, Mangnall YF, Dwivedi A, et al. Sensitivity to nutri- on food intake in humans. Gastroenterology 2005;129:1430 –
ents in patients with non ulcer dyspepsia. Eur J Gastroenterol 1436.
Hepatol 1993;5:109 –113. 26. Sullivan G, Blewett AE, Jenkins PL, et al. Eating attitudes and the
12. Feinle C, Meier O, Otto B, et al. Role of duodenal lipid and irritable bowel syndrome. Gen Hosp Psychiatry 1997;19:62– 64.
cholecystokinin A receptors in the pathophysiology of functional 27. Van Oudenhove L, Vandenberghe J, Geeraerts B, et al. Relation-
dyspepsia. Gut 2001;48:347–355. ship between anxiety and gastric sensorimotor function in func-
13. Pilichiewicz A, Feltrin KL, Horowitz M, et al. Oral carbohydrate and tional dyspepsia. Psychosom Med 2007;69:455– 463.
fat differentially modulate symptoms, gut hormones and antral 28. Soo S, Moayyedi P, Deeks J, et al. Psychological interventions for
area in functional dyspepsia. Am J Gastroenterol 2008;103: non-ulcer dyspepsia. Cochrane Database Syst Rev 2003;2:
2613–2623. CD002301.
14. Talley NJ, Phillips SF, Bruce B, et al. Relation among personality
and symptoms in nonulcer dyspepsia and the irritable bowel
syndrome. Gastroenterology 1990;99:327–333.
15. Talley NJ, Piper DW. Major life event stress and dyspepsia of Reprint requests
unknown cause: a case control study. Gut 1986;27:127–134. Address correspondence to: Assistant Professor Christine Feinle-
16. Talley NJ, Stanghellini V, Heading RC, et al. Functional gastrodu- Bisset, PhD, National Health and Medical Research Council of Austra-
odenal disorders. Gut 1999;45(Suppl 2):37– 42. lia Senior Research Fellow, University of Adelaide Discipline of Medi-
17. Karvetti RL, Knuts LR. Validity of the estimated food diary: com- cine, Royal Adelaide Hospital, Adelaide SA 5000, Australia. e-mail:
parison of 2-day recorded and observed food and nutrient in- [email protected]; fax: (61) 8-8223-3870.
takes. J Am Diet Assoc 1992;92:580 –584.
18. Stunkard AJ, Messick S. The three-factor eating questionnaire to Acknowledgements
measure dietary restraint, disinhibition and hunger. J Psychosom The authors thank Ms Nancy Briggs for her assistance with the
Res 1985;29:71– 83. statistical analysis.
19. Garner DM, Garfinkel PE. The Eating Attitudes Test: an index of
the symptoms of anorexia nervosa. Psychol Med 1979;9:273– Conflicts of interest
279. The authors disclose the following: Amelia Pilichiewicz was sup-
20. Retzlaff BM, Dowdy AA, Walden CE, et al. The Northwest Lipid ported by a Dawes Postgraduate Scholarship provided by the Royal
Research Clinic Fat Intake Scale: validation and utility. Am J Adelaide Hospital and Christine Feinle-Bisset was supported by a
Public Health 1997;87:181–185. National Health and Medical Research Council of Australia Career
21. Talley NJ, Verlinden M, Jones M. Validity of a new quality of life Development Award. The study was supported by National Health and
scale for functional dyspepsia: a United States multicenter trial Medical Research Council of Australia project grant 242802
of the Nepean Dyspepsia Index. Am J Gastroenterol 1999;94: (2003ⴚ2005) and a Royal Adelaide Hospital Research Committee
2390 –2397. Project Grant in 2002.