Reduced nutrient intake is common in cystic fibrosis (CF) and contributes to the malnutrition frequently associated with this disease (1,2). Delayed small intestinal transit, assessed with a breath hydrogen technique, has been reported in CF (3), with the suggestion that this could be due to delayed gastric emptying. Recently, in a series of subjects with CF, delayed gastric emptying has been described with an ultrasound technique (4). Delayed gastric emptying has also been reported in anorexia nervosa (5), and also been shown to correlate with suboptimal energy intakes in malnourished young people with Crohn's disease (6). The mean gastric emptying time of the Crohn's patients with malnutrition was significantly slower compared to both normal control subjects and the Crohn's patients with normal growth and nutritional status.
In normal humans, unabsorbed fat present in the distal small intestine, has been shown to delay gastric emptying (7,8), to reduce the amount of time spent eating, and to induce early satiety (9). Moreover, the presence of nutrients in the duodenum causes feedback inhibition of gastric emptying (10). Thus, fat and protein malabsorption, usual features of CF (11), could contribute to impaired gastrointestinal transit. Anecdotally, early satiety is frequently an area of parental concern.
Additional indirect evidence suggests abnormal gastric emptying in CF may be prevalent (1,2,6,12,13). Gastroesophageal reflux (GER), commonly found in CF (14), has been associated with delayed gastric emptying, although it appears likely that delayed gastric emptying leads to GER rather than the converse (12,13). Hyperacidity or failure of bicarbonate secretion may result in an acid pH in the duodenum that directly feeds back to slow gastric emptying (15). Hyperglycemia is reported to delay gastric emptying even in normal subjects (16), and CF is associated with a high prevalence of impaired glucose tolerance (17,18).
Few studies have examined gastric emptying in CF. This omission has important clinical implications given the effect of impaired gastric function on satiety, energy intake, and ultimately nutritional status. Therefore, the aims of this project were to document solid phase gastric emptying times in people with CF and age- and sex-matched healthy control subjects, and to investigate whether delayed gastric time is a major factor contributing to suboptimal energy intake in people with CF.
MATERIALS AND METHODS
This study, conducted from 1994 to 1996, was a cross-sectional measurement of gastric emptying times, clinical status, energy intake, and fecal fat excretion in 19 patients with CF, 7 of whom were younger than 10 years of age, attending the John Hunter Children's Hospital Cystic Fibrosis clinic, and in 17 age- and sex-matched healthy controls subjects, 6 of whom were younger than 10 years of age. Ethics approval was obtained from the Hunter Area Health Service and the University of Newcastle Ethics Committees. The radiological aspects of the study were approved by the John Hunter Hospital (JHH) Radiation Safety Committee. An introductory letter was sent to families having a member with CF older than 2 years of age inviting them to participate. Normal controls subjects were recruited from advertisements placed on hospital and university notice boards. Subjects were enrolled continuously through the CF outpatient clinic. Written consent was obtained from the patients, parent, or both. Exclusion criteria were: (a) age less than two years, (b) allergy to egg, or (c) acute exacerbation of chest disease. For subjects taking medication known to affect gastric emptying the gastric emptying study was undertaken when the subject had been off the medication for at least 4 days (1 subject). For control subjects exclusion criteria were: (a) relevant gastrointestinal medical or surgical history, (b) egg allergy, or (c) medication affecting gastric emptying.
Subjects with CF were recruited first and a control subject was then sought of the same sex and a similar age. For the females (12 with CF and 9 controls) the mean age difference was 0.54 years (range, 0-2.1 years) with three subjects unmatched. For the males (7 with CF and 8 controls) the mean age difference was 1.28 years (range, 0.03-3.09 years) with 1 control subject unmatched.
Subjects weighed and measured all food, drinks, and medication consumed during 4 consecutive days and recorded the information in a printed food record book. An explanation was given of the methodology with a demonstration of how to operate the scales and how to weigh and measure various food items, using food models. A written checklist was used to standardize the procedure. Battery operated, digital Soehnle 2000-g scales with 1-g increments from 0 to 1000 g and 2-g increments from 1000 to 2000 g and metric cups and spoons were provided for weighing and measuring foods. Four subjects (two controls and two with CF) who were instructed to keep 4-day records only collected for 3 days. Three were because of poor compliance and the other misunderstood the instructions. Daily fat intake and%RDI in the 3-day food records was similar to the 4-day records.
Gastric Emptying Time
The gastric emptying (GE) study was undertaken when the subjects were clinically stable. A single isotope solid-phase GE study was selected because a dual isotope study, incorporating labeled water, would have exceeded the radiation dose limits for volunteers. Gastric emptying time was measured by a solid-phase GE study undertaken in the department of nuclear medicine at JHH. Subjects consumed their usual dose of pancreatic enzymes with the test meal. Solid-phase GE was assessed with a standard test meal of pancakes labeled with 99mTc-macroalbumin aggregates (Pulmolife, Du Pont Merck Pharmaceutical Co., Billerica, MA, USA). The test meal size offered was reduced by half for children younger than 10 years of age. The GE study was performed at 8 AM after an overnight fast from 10 PM the night before. The test meal was made into pancakes from one 50-g egg, 80 g commercial pancake mix, 10 mL polyunsaturated oil, 40 mL water, and 30 g jam. An additional 150 mL of water was given as a drink. To reduce the possibility of subjects not receiving the required dose of technetium one quarter of the ingredients was measured out and the dose mixed with the egg before being cooked as the first pancake. All subjects consumed this labeled pancake first. Total cooked pancake test meal weight and any uneaten portion were weighed and recorded and the test meal was consumed in less than 15 minutes. To ensure standardization of the procedure all the pancakes were prepared and weights recorded by a single investigator (C.E.C.).
All images were acquired on a Elscint SP-6 large field view gamma camera (Elscint, Haifa, Israel) using a low-energy resolution collimator with the acquisition centered on 140 keV photopeak of technetium using a 20% symmetrical window. Sixty-second images were obtained over the anterior and posterior abdomen every 5 min for the first 30 min, and then every 15 min until 90 min, and a final image at 120 min. The image matrix size was 128 × 128 in word mode. A time activity curve was generated using the geometric mean of the anterior and posterior images at each time point.
Subjects remained supine for the first 60 min but were then allowed to walk around for the second 60 min as recommended by Christian et al. (19). The lag phase (the time from finishing the test meal to the moment 5% of activity has left the stomach) and the percent of solids retained in the stomach at 30, 60, and 120 min were determined along with the time at which 50% of the technetium remained in the stomach (T1/2).
Clinical status was measured by the Shwachman score (SS) as part of routine clinical care. Patients were evaluated in the four categories of nutritional status, physical findings, X-ray findings, and general activity by a single, independent observer, blinded to the history of the patients. The closer the total score is to 100 the better is the clinical status of the patient (20). The chest X-ray films were also scored in a blind manner by a single observer (R.L.H.).
Fecal Fat Excretion
Fecal fat excretion (% FFE) was expressed as the percent of the total dietary fat intake that was excreted based on the results of a 3-day stool collection. All bowel motions passed for a defined 72 hours were collected and stored in the refrigerator or freezer before analysis for total fat content. Patients consumed their usual enzyme therapy throughout the collection period. The fecal fat collection started 24 hours after the commencement of the weighed food record so that the collection and the food record finished at the same time. Fecal samples were analyzed for fat content by the method of van der Kamer (21).
The methods of Cameron were used for all the anthropometric measures (22). Heights were measured with a Holtain, Crymych, Dyfed, stadiometer. Weights were recorded to the nearest 0.001 g on GEC/Avery digital scales, model number 824/890, with subjects wearing only minimal underwear. Skinfold measurements were taken as part of the Shwachman score evaluation. Skinfolds were measured in triplicate at the four sites using Holtain (Crymych, UK) skin callipers by a single observer (C.E.C.). The average of the 3 measures were taken on the subjects left-hand side.
Analysis of food records was conducted using the Diet/1 Nutrient Calculation Software, which is based on the 1992 Australian food table database and the composition of Australian manufactured foods (23). Mean daily total kilojoules and grams of protein, fat, and carbohydrate were calculated along with the percentage of total daily energy derived from protein, fat, and carbohydrate. Energy intake was expressed as the percent of the Recommended Daily Intake (% RDI) for energy for Australians for subjects aged 18 years or more, and as the percent of the WHO recommended kilojoules per day for subjects less than 18 years old, using an activity factor of 1.6 times the basal metabolic rate. For people with CF, daily energy intakes less than 100% RDI are considered inadequate, whereas optimal energy intakes are usually those greater than or equal to 120% RDI.
The time taken for GE of a standard amount of energy (420 kJ), which is approximately equivalent to one pancake, was calculated knowing the lag phase and the reported percent of solids remaining 120 min after the start of the GE study. The rate of GE (kJ/min) of energy was also calculated from the energy value of the test meal consumed and the percent of solids retained at 120 min. The assumptions made here are first, that adequate mixing of the labeled pancake and remaining test meal has occurred and second, that gastric emptying of solids is linear.
Statistical analysis was performed using Minitab 10Xtra for Windows. The mean and 95% confidence interval (CI) are reported for normally distributed data with the median and 95% CI reported for non-normally distributed data. Statistical comparisons were performed for normally distributed variables using t-tests and the Mann-Whitney nonparametric test for non-normally distributed variables. Correlation studies were undertaken to assess how closely the variables studied were related within both subject groups (CF and controls) and regression analysis was used to explore how much of the variation in gastric emptying time and energy intake could be explained by the other variables. For our sample of 19 subjects with CF the 5% and 1% points for correlation coefficients are 0.46 and 0.58, respectively. For the 17 control subjects the 5% and 1% points for correlation are 0.48 and 0.61, respectively.
Clinical characteristics of the subjects with CF and the controls are summarized in Table 1. There was no significant difference in age between the subject groups. The mean height and weight for the subjects with CF was lower than the normal controls and the mean% ideal weight was higher in the subjects with CF; however, neither of these differences reached statistical significance.
Dietary and Gastric Emptying Results
Dietary and gastric emptying variables are summarized in Tables 2 and 3. The% RDI for energy was greater in the subjects with CF compared with controls by 26% (95% CI, 10.0 and 41.3), whereas total fat intake was greater by 31 g/day (95% CI, 9.9 and 51.6). The gastric emptying time measured at T1/2 was significantly faster (p = 0.018) in subjects with CF compared to normal controls by 19.8 minutes (95% CI, 3.5 and 34.9) as illustrated in Figure 1. There was no significant difference between groups for the weight of the test meal consumed, lag phase of gastric emptying,% of solids retained at 120 min, the rate of gastric emptying (kJ/min), or the calculated time taken to empty 420 kJ.
As illustrated in Fig. 2, there was a negative correlation between T1/2 and% RDI in CF, (r = -0.50, p<0.05), such that as daily energy intake increased solid-phase gastric emptying time decreased. In CF, the correlation between T1/2 and test meal weight (r = -0.44, p≅0.05) almost reached significance, suggesting that the greater the weight of test meal consumed the faster it was emptied from the stomach. There was a similar finding for the correlation between T1/2 and total fat intake (r = -0.45, p≅0.05), such that the higher the usual daily fat intake in grams, the faster the solid-phase gastric emptying time. These relationships were heightened when T1/2 was replaced with T420 kJ as the measure of solid-phase gastric emptying with the following r values against% RDI (r = -0.51, p<0.05), meal weight (r = -0.82, p<0.01), total fat intake (r = -0.65, p<0.01), and age (r = -0.70, p<0.01). Age was positively correlated with% RDI in CF patients (r = 0.61, p<0.01), meaning that older subjects consumed higher relative energy intakes even though% RDI is already adjusted for age. Test meal weight was significantly correlated with T420 kJ for both subject groups (r = -0.82 CF, p<0.01 and r = -0.67 controls, p<0.01), suggesting that consumption of a greater amount of test meal resulted in faster gastric emptying of an equivalent energy value. Significant correlations were also observed for both groups for the gastric emptying rate (kJ/min) vs. test meal weight (r = 0.94, p<0.01 CF, and r = 0.72 controls, p<0.01), indicating that consumption of a greater amount of test meal resulted in a faster gastric emptying rate of energy per minute.
There was a positive trend for the correlation between% FFE and T1/2 (r = 0.38), such that as% FFE increased the solid-phase gastric emptying time increased, but this was not significant (p>0.05).
Regression was used to investigate the variables associated with the outcomes T1/2 and T420, measures of gastric emptying in the CF subjects. In each case, the best model contained only one variable: T1/2 was best predicted by% RDI. On average, an increase of 10% in% RDI was associated with a 5-min reduction in the time taken to empty half of the radiolabel from the stomach. T420 was best predicted by test meal weight. On average, a 10-g increase in test meal weight was associated with a 3.7-min reduction in the time taken to empty one pancake.
This is the first report of scintigraphic investigation of solid-phase gastric emptying in CF. We have demonstrated that GE is faster in healthy, young people with CF compared to healthy, age- and sex-matched controls.
Subjects with CF were relatively healthy as indicated by their high Shwachman scores, good control of malabsorption, high% ideal body weights, and high energy intakes. They may have been slightly stunted in height as suggested by lower heights and weights. Their relative energy intakes were approximately 25% higher and fat intakes approximately 40% higher compared to the healthy control subjects. The energy intakes in the CF subjects are consistent with reports for other healthy groups with CF (24). Although energy intakes in the controls appear to be relatively low, they are consistent with reports that the current recommendations for normal children are too high (25). When total energy expenditure (TEE), measured in normal children and adolescents by using the doubly labeled water technique, was compared to the results of a 7-day weighed food record, it appears that adolescents may underestimate food intake (26). Black et al. (27) have shown good agreement between mean energy intake and mean energy expenditure when food intake was recorded by an observer or when it was self-reported by normal weight, self-selected, highly motivated volunteer subjects using weighed records. We sought to minimize recording bias in both subjects with CF and controls by recruiting volunteers and having parents help with the weighing of food records for both groups.
Solid-phase gastric emptying time, measured at T1/2, was 30% faster in the subjects with CF compared to normal, healthy, age- and sex-matched controls. However, there was no difference in lag time,% solids at 120 min, the time to empty 420 kJ, or kilojoules emptied per minute. The results of the study are in direct contrast to a recent report of delayed gastric emptying in CF measured by ultrasonography (4). Methodology differences between studies may account for the opposing findings. We were unable to measure liquid-phase emptying in this study because the dose of radiation obtained by labeling water with indium meant that the combined radiation dose would have exceeded the limits set for volunteers at the John Hunter Hospital. Cucchiara et al. (4) did not report the test meal weight consumed by subjects and controls or whether the gastric emptying time reported was that of liquids or solids. Further, the clinical status of their subjects was not described and as malnutrition has been associated with delayed GE in individuals with anorexia nervosa (5) and malnourished subjects with Crohn's disease (6), it may be that differences in subject selection and methodology explain the opposing results.
Scintigraphy is now recognized as the gold standard method for studying GE (28). Vantrappen reported that ultrasonic measures of GE can only measure changes in volume of the gastric antrum as the technique is unable to visualize either the fundus or the corpus. He added that even when a mixed, liquid plus solid meal was used, GE rate measured correlated with liquid emptying rather than solid emptying (28). Differential emptying of solids and liquids does occur and alterations in one does not predict alterations in the other (29). In order to make valid comparisons of GE data it is essential to use a standardized test meal (30) and this can present problems in young children who refuse or simply are unable to consume all the test meal offered. For this reason we chose to reduce the meal size offered to subjects younger than 10 years of age and to record the weight of the test meal consumed. An attempt was made to correct for varying meal weight consumed by using the calculated time taken to empty 420 kJ, thereby focusing on the initial slope of the emptying curve or initial solid-phase gastric emptying time. In our study, substituting T1/2 with T420, as a measure of solid-phase gastric emptying time in both subjects with CF and controls, led to stronger negative correlations between gastric emptying time and test meal weight. Strong positive correlations were also seen for both groups between test meal weight and the gastric emptying rate of energy in kilojoules per minute. As the amount of test meal eaten increased, the solid-phase gastric emptying time (T420) was faster and the GE rate of energy (kJ/min) was also faster. This is in direct contrast to literature reports of slower GE as stomach volume, energy content, and body size increase (31,32).
The GE rates in this study are consistent with the literature reports that the energy emptying rate is similar across studies and that “liquefied” solids appear to empty at a linear rate of 1.5-3 kcal/min (32). The emptying rates of energy for the subjects with CF and controls were similar (2.8 kcal/minute for CF and 2.5 kcal/min for controls, p = 0.92).
Faster GE in CF was positively correlated with higher relative energy intakes. Read et al., in a review of the role of the gut in regulating food intake (10), suggest that the gut can adapt to process larger amounts of food more rapidly and conserve a higher energy intake through altered regulation of nutrient receptors that can modify gastrointestinal transit. He reports that GE has been shown to be accelerated with “overfeeding” of both healthy volunteers and obese subjects (10). Cunningham et al. (33) showed that ingestion of a high-fat, high-energy diet (19.2 MJ, 270 g lipid) for 2 weeks accelerated both GE and small bowel transit of a high-fat meal in adults. The accelerated GE observed in obese subjects (34) and the delayed GE in anorexia nervosa (5) suggest that alterations in GE impact on self-selected energy intake (or vice versa), which in turn impacts on body weight.
Rapid GE has also been documented in newly diagnosed non-insulin-dependent patients with diabetes mellitus (NIDDM) (35,36). Gastric inhibitory protein (GIP), which is a good marker of glucose absorption, is increased in NIDDM patients after meals (36). This is consistent with rapid GE or defective insulin response, which in turn leads to reduced feedback inhibition of GIP secretion by insulin. Phillips et al. hypothesized that there is a loss of feedback control of gastric emptying in NIDDM (36). Hypoinsulinemia associated with impaired glucose tolerance may contribute to the rapid GE observed in this group of subjects with CF given that CF is associated with a high prevalence of impaired glucose tolerance secondary to hypoinsulinemia (17,18). Two of our subjects had CF-related diabetes. In one of these subjects the GE test was repeated because the first result was so rapid that hypoglycemia was queried. Although mean HbA1C was not outside the range for normal subjects, the highest HbA1C (6.9 mM) belonged to one of the subjects with diabetes, whereas the next highest HbA1C (5.3 mM) was recorded in a patient who developed diabetes the following year.
Gastric emptying in CF may be bimodal. We know from studies in anorexia nervosa and Crohn's disease that malnutrition in both these states is associated with delayed GE (5,6). For well subjects with CF, gastric emptying may be rapid secondary to adaptation to a high-fat, high-energy intake and/or alterations in the glucose insulin feedback mechanisms. As the disease progresses and chest inflammation and or malnutrition develops, GE may become delayed as found by Cuccihara et al. (4). Clearly, clinical status of patients would be an important parameter in future studies of gut function in CF.
Solid-phase gastric emptying time is faster in relatively healthy subjects with CF compared to normal, healthy, age- and sex-matched controls. Faster solid-phase GE times were found to be associated with higher relative energy intakes and may represent a survival advantage in CF.
Although further studies are required to assess gastric function under altered physiological conditions, the relationship between energy intake and GE time in individuals with CF should be exploited in order to optimize nutritional status.
1. O'Loughlin E, Forbes D, Parsons H, Scott B, Cooper D, Gall G. Nutritional rehabilitation of malnourished patients with cystic fibrosis. Am J Clin Nutr
2. Durie PR, Pencharz PB. Cystic fibrosis: nutrition. Br Med Bull
3. Bali A, Stableforth DE, Asquith P. Prolonged small-intestinal transit time in cystic fibrosis. BMJ
4. Cucchiara S, Raia V, Minella R, Frezza T, De Vizia B, De Ritis G. Ultrasound measurement of gastric emptying time in patients with cystic fibrosis and effect of ranitidine on delayed gastric emptying. J Pediatr
5. Robinson P, Clarke M, Barrett J. Determinants of delayed gastric emptying in anorexia nervosa and bulimia nervosa. Gut
6. Grill BB, Lange R, Markowitz R, Hillemeier AC, McCallum RW, Gryboski JD. Delayed gastric emptying in children with Crohn's disease. J Clin Gastroenterol
7. Cunningham KM, Read NW. The effect of incorporating fat into different components of a meal on gastric emptying and postprandial blood glucose and insulin responses. Br J Nutr
8. Welch IMcL, Cunningham KM, Read NW. Regulation of gastric emptying by ileal nutrients in humans. Gastroenterology
9. Hyman PE. Absent postprandial duodenal motility in a child with cystic fibrosis. Correction of the symptoms and manometric abnormality with cisapride. Gastroenterology
10. Read NW, French S, Cunningham K. The role of the gut in regulating food intake in man. Nutr Rev
11. Durie PR, Pencharz PB. Cystic fibrosis: nutrition. Br Med Bull
12. Hillemeier AC, Lange R, McCallum R. Delayed gastric emptying in infants with gastroesophageal reflux. J Pediatr
13. Rosen PR, Treves S. The relationship of gastroesophageal reflux and gastric emptying in infants and children. J Nucl Med
14. Scott RB, O'Loughlin EV, Gall DG. Gastroesophageal reflux in patients with cystic fibrosis. J Pediatr
15. Park RW, Grand RJ. Gastrointestinal manifestations of cystic fibrosis: a review. Gastroenterology
16. Horowitz M, Edelbroek M, Fraser R, Maddox A, Wishart J. Disordered gastric motor function in diabetes mellitus. Recent insights into prevalence, pathophysiology, clinical relevance, and treatment. Scand J Gastroenterol
17. Sheehan JP, Ulchaker MM, Doershuk CF, Stern RC, Amini SB. Diet, cystic fibrosis, and diabetes: making friends with the perfect enemy. Lancet
18. Hayes DR, Sheehan JP, Ulchaker MM, Rebar JM. Management dilemmas in the individual with cystic fibrosis and diabetes. J Am Diet Assoc
19. Christian P, Datz F, Sorenson JA, Taylor A. Teaching editorial: technical factors in gastric emptying studies. J Nucl Med
20. Shwachman H, Kulczycki LL. Long-term study of one hundred five patients with cystic fibrosis: studies made over a 5 to 14 year period. Am J Dis Child
21. Van de Kamer JH, Huinick H, Weyers HA. Rapid method for determination of fat in feces. J Biol Chem
22. Cameron N, Falcon F, Tanner JM, eds. Human growth
. New York: Plenum Press, 1978;2:35.
23. Goodhill C. Diet/1 nutrient calculation software. 20 Westborne St., Highgate Hill, Brisbane, Queensland, 4101. © Xyris Software; 1992.
24. Tomezsko JL, Stallings VA, Scanlin TF. Dietary intake of healthy children with cystic fibrosis compared with normal control children. Pediatrics
25. Prentice A, Lucas A, Vasquez-Velasquez L, Davies P, Whitehead R. Are current dietary guidelines for young children a prescription for overfeeding? Lancet
26. Vantrappen G. Methods to study gastric emptying. Dig Dis Sci
27. Horowitz M, Maddox A, Wishart J, Harding P, Chatterton B, Shearman D. Relationships between oesophageal transit and solid and liquid gastric emptying in diabetes mellitus. Eur J Nucl Med
28. Smout A, Horowitz M, Armstrong D. Methods to study gastric emptying. Dig Dis Sci
29. Hunt JN, Stubbs DF. The volume and energy content of meals as determinants of gastric emptying. J Physiol
30. Phillips W, Schwartz J, McMahan C. Rapid gastric emptying of an oral glucose solution in type 2 diabetic patients. J Nucl Med
31. Phillips W, Schwartz J, McMahan C. Reduced postprandial blood glucose levels in recently diagnosed non-insulin-dependent diabetics secondary to pharmacologically induced delayed gastric emptying. Dig Dis Sci