点击显示 收起
1 From the Department of Medicine, University of Adelaide, Royal Adelaide Hospital, Adelaide, Australia
2 Supported by a project grant from the National Health and Medical Research Council (NH&MRC) of Australia, a Florey Research Fellowship from the Royal Adelaide Hospital (2001-2003) and a Career Development Award from the NH&MRC (to CF-B), and a fellowship jointly from the NH&MRC and Diabetes Australia (to KLJ). 3 Reprints not available. Address correspondence to M Horowitz, Department of Medicine, Royal Adelaide Hospital, North Terrace, Adelaide, SA 5000, Australia. E-mail: michael.horowitz{at}adelaide.edu.au.
ABSTRACT
Background: Gastric distension reduces food intake, and antral, rather than proximal, gastric distension may be the dominant mechanism in the induction of appetite-related sensations. Healthy aging is associated with reduced appetite.
Objective: We examined the effects of different energy preloads on appetite, plasma cholecystokinin, antral area, and subsequent energy intake in healthy older and young subjects.
Design: On 3 separate days, 12 young and 12 older subjects consumed 400 mL of a drink containing either 0 kcal (water), 250 kcal, or 750 kcal 70 min before a buffet-style meal.
Results: Hunger was less in the older than in the young subjects (P < 0.001). Both nutrient preloads reduced hunger and increased fullness more than did water (P < 0.02), and older subjects were more full than were the young (P < 0.05). Antral area was greater after the nutrient preloads than after water (P = 0.001) and greater in the older than in the young subjects (P = 0.005). In both groups, food intake was suppressed in an energy-dependent manner (P = 0.008). Plasma cholecystokinin was greater in the older than in the young subjects (P = 0.003). Immediately before the meal, hunger (r = –0.59) and energy intake (r = –0.90) were inversely related and fullness (r = 0.66) was directly related to antral area (all: P < 0.001). Antral area, but not plasma cholecystokinin, was a predictor of subsequent energy intake.
Conclusion: In healthy young and older subjects, the suppression of subsequent energy intake by a liquid preload is nutrient dependent and comparable, and both satiation and satiety are related to antral area and (presumably) antral distension.
Key Words: Antral area • energy intake • older subjects • young subjects • cholecystokinin
INTRODUCTION
Aging is associated with a progressive decrease in hunger and energy intake, termed the anorexia of aging (1–3). This physiologic anorexia probably predisposes the elderly to pathological anorexia and malnutrition (1); the latter represents a major cause of morbidity and mortality in the elderly (4, 5). The cause or causes of the anorexia of aging are poorly defined.
Gastrointestinal factors that may decrease energy intake include gastric distension, exposure of small-intestine receptors to nutrient, and hormones, such as cholecystokinin (CCK) (6–8). Gastric distension reduces food intake in young and obese subjects (9, 10) and may mediate the effects of gut hormones on food intake (11). Mechanical properties and neural innervation vary in different regions of the stomach (12), and it is uncertain whether the site of gastric distension is important in mediating appetite-related sensations. In some studies using gastric balloon distension, the site of distension was not precisely defined (9, 10). In young subjects, distension of the proximal stomach with the use of a barostat increases the perception of fullness (13, 14), but effects on food intake have not been evaluated. There is also evidence, albeit inconsistent, in both healthy subjects and patients with functional dyspepsia, that energy intake is influenced by proximal gastric accommodation (15, 16). In contrast, observations in young subjects established that the perception of postprandial fullness is closely related to antral content or area and not to the content of the proximal stomach (17–19). Hence antral, rather than proximal gastric, distension may be the dominant intragastric mechanism in the induction of appetite-related sensations. The relation between energy intake and antral area has not been evaluated.
When potential gastrointestinal factors that may contribute to the anorexia of aging are considered, aging may be associated with an increased sensitivity to the satiating effects of CCK (20, 21). Moreover, plasma CCK concentrations are greater in older subjects, both healthy (21, 22) and malnourished (23). Other gastrointestinal peptides, perhaps particularly peptide YY (24), may suppress food intake; however, there does not appear to be any effect of aging on glucagon-like peptide-1, gastric inhibitory polypeptide, or peptide YY (20, 21). Fasting plasma ghrelin concentrations are higher in malnourished older subjects than in healthy older and healthy young subjects, but there is no difference between the 2 healthy groups (25). Whereas ghrelin may stimulate food intake (26), these observations suggest that a reduction in ghrelin activity is not a cause of the anorexia of aging. The sensitivity of the small intestine to nutrients does not appear to be affected by healthy aging (7, 27). The perception of proximal gastric distension is reduced, rather than increased, in healthy older subjects (28). It is unclear whether the effects of oral nutrients on perceptions of appetite or energy intake are modified by aging (5, 29–31), nor is it known whether these responses can be accounted for by antral area, the overall rate of gastric emptying, plasma CCK, or all of those factors.
The aims of this study were to evaluate in healthy young and older subjects the effects of different energy preloads on appetite, gastric antral area, gastric emptying, plasma CCK concentrations, and the relations between these factors and subsequent energy intake. The primary hypothesis was that energy intake is related to antral area.
SUBJECTS AND METHODS
Subjects
We studied 24 healthy volunteers, all of whom were living independently and were recruited by advertisement. The cohort was divided into 2 groups of 12 each (6 F, 6 M). The older subjects had a mean (±SD) age of 74.4 ± 1.2 y (range: 67-83 y) and a body mass index (BMI; in kg/m2) of 24.1 ± 0.5 (range: 21.1-27.1). The young subjects had a mean age of 23.9 ± 1.4 y (range: 18-33 y) and a BMI of 23.2 ± 0.6 (range: 21.0-27.0). No subject was >5% below or above his or her ideal body weight (7). Each older subject was selected so his or her BMI was matched to within 1 unit of that of a young subject—accordingly, the BMIs of the 2 groups did not differ significantly. Exclusion criteria were an alcohol intake >20 g/d, significant gastrointestinal symptoms, previous abdominal surgery (apart from uncomplicated appendectomy), abnormalities in concentrations of either serum electrolytes or thyroid-stimulating hormones, significant chronic illness (heart disease, malignant neoplasm, diabetes mellitus), impaired cognitive function (score <25 on Mini-Mental State Examination; 32), depression [score >11 on the Geriatric Depression Questionnaire (older subjects only; 33) and absence of a history of depression (young subjects only)], and the use of medications that may affect gastrointestinal motor function or appetite (eg, prokinetics, anticholinergics, calcium antagonists, appetite suppressants, and antidepressants). Subjects were also required to be unrestrained eaters, with a score <11 for Factor I (cognitive restraint) on the Three-Factor Eating Questionnaire (34). All subjects completed a detailed 3-d diet diary before participation (22) and were asked to maintain their usual diet and activity levels throughout the study. The study protocol was approved by the Ethics Committee of the Royal Adelaide Hospital, and each subject gave written informed consent.
Protocol
Subjects underwent 3 studies in randomized order, and each study day was separated by at least a week. Subjects arrived at the laboratory at 0900 after a 12-h overnight fast from food and fluid, except water. Subjects were allowed one standard (200-mL) glass of water on the morning of the test, but not until 2 h after arrival at the laboratory. Ultrasound measurements were used to confirm that the stomach was empty at baseline. On arrival, subjects were seated comfortably on a bed (upright and in a relaxed position, so that the angle between the upper and the lower part of the body was 90°; a pillow was placed under the knees). An intravenous cannula was placed in an antecubital vein for blood sampling. Subjects then rested for 20 min. At t = 0 min, each subject drank a 400-mL preload (composition given below) within 10 min. The subjects were blinded to the composition of the energy preloads; blinding to the water was not possible. At t = 70 min (ie, 60 min after completion of ingestion of the preload), subjects were offered a standard buffet-style meal containing food in excess of what they would normally eat and were invited to eat as much as they wished over 30 min (t = 70-100 min) until they felt comfortably full. The interval of 60 min between the preload and subsequent meal was chosen on the basis of previous studies (35). Visual analogue scales (VAS) were administered, and venous blood samples were taken at t = –10 min, immediately before and after ingestion of the preload, and then every 10 min from t = 10 min to t = 160 min, excluding the period of the buffet meal (t = 70-100 min). Measurements of antral area were made by using ultrasound at t = –10 min, immediately before and after ingestion of the preload, and then every 5 min from t = 10 min to t = 160 min, excluding the period of the buffet meal (t = 70-100 min). Subjects were monitored over the period t = 110-160 min, but the data were not included in the statistical analysis.
Preloads
On the control day, the preload was 400 mL spring water (0 kcal, at room temperature). On the other 2 d, the preloads differed in energy content (250 kcal on the low-dose day and 750 kcal on the high-dose day) but had the same macronutrient distribution (13% of energy as protein, 33% as fat, and 54% as carbohydrate). The nutrient preloads were a mixture of yogurt, cream, milk, corn flour, golden syrup, oil (sunflower), and protein supplement (ProMod; Ross Products Division, Abbott Laboratories, Columbus, OH). Both nutrient preloads were yogurt-type drinks of similar odor, taste, palatability, consistency, and sweetness (Table 1). Gelatin and an artificial sweetener (aspartame 951, Equal; Merisant Australia Pty Ltd, Crow’s Nest, Australia) were added to the low-energy preload to match its consistency and sweetness, respectively, to those of the high-energy preload. The buffet meal included various food items: bread, cheese, chicken, ham, yogurt, fruit, juices, biscuits, and crackers of predetermined caloric content (22). At the end of the buffet meal, food remnants were weighed by a single observer (22).
View this table:
TABLE 1. Composition of low- (250 kcal) and high- (750 kcal) energy preloads1
Antral area and gastric emptying
Measurements of antral area were performed with an ultrasound machine (Aloka SSD-650 CL; ALOKA Co, LTD, Tokyo) by using either a 3.5- or a 5-MHz sector transducer. Antral area (cm2) was measured with the use of a caliper and calculation program built in to the ultrasound machine. To optimize precision, the transducer was positioned vertically to obtain a parasagittal image of the antrum with the superior mesenteric vein and the abdominal aorta in a longitudinal section, and measurements were performed at the end of inspiration, as described previously (17, 18). Antral area at baseline and antral area at t = 70 min (immediately before the start of the buffet meal) were used to evaluate relations with baseline ratings of appetite and energy intake at the buffet meal, respectively. The time at which the antral area had decreased to % of maximum (T75%) was used as an index of gastric emptying (17).
Appetite and energy intake
Sensations of hunger and fullness were rated by each subject with the use of 100-mm VAS (by using anchors such as "not hungry" to "hungry" and "empty" to "full") as described previously (36). Energy intake, as assessed by the 3-d food diary maintained before entry into the study and the amount of food consumed at the buffet meal (including its macronutrient distribution), was quantified by using FOODWORKS software (version 2.10; Xyris Software Pty Ltd, Highgate Hill, Australia; 7).
Blood glucose, plasma insulin, and plasma CCK concentrations
Blood glucose concentrations were measured by using a portable blood glucose meter (MediSense Precision Q · I · D System; Abbott Laboratories, MediSense Products Inc, Bedford, MA). The accuracy of this method has been confirmed by using the hexokinase technique (37).
Blood was collected in ice-chilled EDTA-containing tubes also containing 1000 kIU aprotinin/mL blood (Trasylol; Bayer PLC, Newbury, United Kingdom) for measurement of insulin and CCK. Plasma was separated by centrifugation at 3000 rpm for 15 min at 4 °C (Labofuge 400 R; Heraeus, Osterode, Germany) within 10 min of collection and stored at –70 °C until assayed.
Plasma insulin concentrations (mU/mL) were measured by using an enzyme immunoassay (Imx Microparticle Enzyme Immunoassay; Abbott Laboratories, Diagnostic Division, Tokyo; 38). The detection limit of the assay was 1.0 mU/L; the interassay CVs were 4.5% at 8.3 mU/L and 3.4% at 40.4 mU/L.
Plasma CCK concentrations (pmol/L) were measured by using a radioimmunoassay as described previously (20). A commercially available antibody (C2581, Lot 105H4852; Sigma Chemical, St Louis) raised in rabbits against synthetic sulfated CCK-8 was employed. This antibody binds all CCK peptides containing the sulfated tyrosine residue in position 7. It shows <2% cross-reactivity with human gastrin I and does not bind to structurally unrelated peptides. The detection limit was 1 pmol/L, and the intraassay CV at 50 pmol/L was 9.5%.
Statistical analyses
For VAS, blood glucose and gastrointestinal hormone concentrations, and antral area, baseline was calculated as the mean of t = –10 and 0 min values. Comparisons between single characteristics of the 2 age groups (eg, BMI, baseline symptom scores, and fasting blood hormone concentrations) were performed by using one-way analysis of variance software (SuperANOVA Version 1.11; Abacus Concepts Inc, Berkeley, CA). Repeated-measures analysis of covariance software (version 8.2; SAS Institute Inc, Cary, NC), with time (t = 10-70 min) and treatment (preloads) as within-subject factors, age group (older or young) as a between-subject factor, and baseline as a covariate, was performed for ratings of appetite; blood glucose, plasma insulin, and gastrointestinal hormone concentrations; and antral area. Differences in mean energy intake and macronutrient content of the buffet meal were analyzed by repeated-measures two-way analysis of variance, with treatment as the within-subject factor and age group as the between-subject factor. When significant effects were observed, pairwise comparisons of adjusted means were performed by using Student’s t test and adjusted for multiple comparisons by using the Bonferroni (Holms adjustment, analysis of covariance) and stepdown Bonferroni (analysis of variance) adjustments. Data from the period after the buffet meal (t = 110-160 min) were not analyzed, because the amount of food eaten was variable. Within-subject correlations accounting for repeated measures across treatments were calculated for each combination of energy intake (kcal) and antral area, CCK, gastric emptying (T75%), and sensations of hunger and fullness with the use of a method specified by Bland and Altman (39). When relations between the above variables were found, multiple regression analysis was performed to establish determinants of hunger, fullness, and energy intake by taking into account effects of other possible determinants (STATVIEW software, version 5.0; SAS Institute Inc). Where no significant relation was observed, results were omitted. Results are shown as means ±SEM. A P value < 0.05 was considered statistically significant.
RESULTS
All of the studies were well tolerated without untoward events. Measurements of antral area were technically adequate in all cases. Energy intake immediately before the study (3-d food diary) was 23% lower in the older than in the young subjects (1860 ± 123 kcal/d and 2430 ± 204 kcal/d, respectively; P = 0.007). There were no significant differences between the 2 age groups in the proportions of carbohydrate, fat, or protein ingested. There was no significant effect of sex on the results (data not shown).
Hunger ratings
Effect of preload
In both age groups, there was a significant reduction in hunger (t = baseline to 10 min: treatment effect, P < 0.001) after both the low-dose (compared with control, P < 0.001) and high-dose (compared with control, P = 0.019) preloads without a significant difference between the 2 nutrient preloads in the magnitude of this reduction (P = 0.10; Figure 1A). Scores for hunger were significantly lower (t = 10 – 70 min) on both the low-dose (P = 0.003) and high-dose (P < 0.001) days than on the control day (treatment effect, P < 0.001) in both age groups; there was no significant difference between the 2 nutrient preloads.
FIGURE 1.. Mean ± SEM absolute ratings of hunger (A) and fullness (B) in young (n = 12, ) and older (n = 12, •) subjects who received preloads of either water (0 kcal; control), 250 kcal (low-dose), or 750 kcal (high-dose). Analyzed by repeated-measures analysis of covariance with baseline (bl) as a covariate. A: At baseline, hunger ratings were significantly lower in the older than in the young subjects (P < 0.001). In both age groups, there was a significant reduction in hunger after the nutrient preloads as compared with the control (P < 0.001) but no significant difference between the 2 nutrient preloads. Ingestion of the buffet meal was associated with a significant reduction in hunger under all 3 conditions in both age groups (P < 0.001). The magnitude of the decrease (t = 100 – 70 min) was less after both nutrient preloads than after the control, and there was no significant difference between the 2 nutrient preloads. B: There was no significant difference in fullness ratings between the 2 age groups at baseline. There was a significant difference in fullness ratings between the age groups after the preloads (P = 0.049); the older subjects had greater fullness ratings. In both age groups, ratings for fullness were significantly greater after the nutrient preloads than after the control (treatment effect, P < 0.001; low-dose compared with control, P = 0.003; high-dose compared with control, P < 0.001), and there was a trend for the magnitude of this increase to be greater with the high-dose preload than with the low-dose preload (P = 0.08). Ingestion of the buffet meal was associated with a significant (P < 0.001) increase in fullness ratings under all 3 conditions in both age groups. The magnitude of the increase (t = 100 – 70 min) was less after both nutrient preloads than after the control; there was no significant difference between the 2 nutrient preloads.
Effect of buffet meal
Ingestion of the buffet meal was associated with a reduction in hunger under all 3 conditions in both groups (P < 0.001). The magnitude of the decrease (t = 100 – 70 min) was greater on the control day than on the low-dose (P < 0.001) and high-dose (P < 0.001) days; there was no significant difference between the 2 nutrient preloads.
Effect of age group
Hunger ratings were significantly (P < 0.001) lower in the older than in the young subjects both at baseline (47.3 ± 4.8 mm and 68.6 ± 3.8 mm, respectively) and after ingestion of the preloads. The magnitude of the decrease in hunger after the low-dose and high-dose preloads did not differ significantly between the 2 groups.
Fullness ratings
Effect of preload
Fullness increased significantly in both age groups after ingestion of the preloads (t = baseline – 10 min; treatment effect, P < 0.001); the magnitude of this increase was significantly (P < 0.001) greater with the low-dose and high-dose preloads than with the control preload and significantly (P = 0.03) greater with the high-dose preload than with the low-dose preload (Figure 1B). Scores for fullness were significantly greater on both the low-dose (P = 0.003) and high-dose (P < 0.001) days than on the control day (treatment effect, P < 0.001), and there was a nonsignificant trend (P = 0.08) for this increase to be greater with the high-dose preload than with the low-dose preload.
Effect of buffet meal
Meal ingestion was associated with an increase in fullness in both groups for all 3 treatments (P < 0.001); the magnitude of this increase (t = 100 – 70 min) was significantly greater on the control day than on the low-dose (P = 0.005) and high-dose (P < 0.001) days; there was no significant difference between the 2 nutrient preloads.
Effect of age group
There was no significant difference in fullness between the 2 age groups at baseline (older: 16.5 ± 2.6; young: 15.1 ± 1.8 mm) or after preload ingestion. The magnitude of the increase in fullness differed significantly between the age groups (P = 0.049): the older subjects were more full after the preloads. There was no treatment x age group interaction.
Energy intake
As shown in Figure 2, there was an energy-dependent suppression of energy intake by the preloads (treatment effect, P = 0.008) in both age groups. The magnitude of this suppression did not differ significantly between the age groups, irrespective of whether energy intake was expressed in absolute terms or as a percentage of that on the control day: in the young subjects, the values were 80.4 ± 3.1% on the low-dose day and 61.1 ± 4.8% on the high-dose day; in the older subjects, the values were 84.8 ± 3.2% on the low-dose day and 66.8 ± 3.7% on the high-dose day. On the control day, energy intake was less, albeit not significantly so, in the older than in the young subjects (1077 ± 121 and 1274 ± 102 kcal, respectively). The macronutrient distribution did not differ significantly between the 2 groups, nor was there a significant difference between the 3 preloads (data not shown).
FIGURE 2.. Mean (± SEM) differences in energy intake at the buffet meal between young (n = 12) and older (n = 12) subjects who received preloads of water (0 kcal; control), 250 kcal (low-dose), or 750 kcal (high-dose). Analyzed by repeated-measures two-way ANOVA with stepdown Bonferroni correction to adjust for multiple comparisons. There was no age x treatment effect. In both age groups, there was an energy-dependent suppression of energy intake by both the nutrient preloads in comparison with the control (treatment effect: P = 0.008). The magnitude of this suppression did not differ significantly between the age groups.
Antral area and gastric emptying
Effect of preload
As shown in Figure 3, there was a significant increase in both age groups in antral area after the preloads (t = baseline – 10 min; treatment effect, P = 0.04). The magnitude of this increase was significantly greater on the low-dose (P = 0.001) and high-dose (P = 0.002) days than on the control day; there was no significant difference between the 2 nutrient preloads. Antral area was greater after ingestion of both the low-dose and high-dose (P < 0.001 for both age groups) preloads than than after that of the control, and it tended to be greater after ingestion of the high-dose preload than after that of the low-dose preload (P = 0.095).
FIGURE 3.. Mean (± SEM) antral area in young (n = 12, ) and older (n = 12, •) subjects who received preloads of water (0 kcal; control), 250 kcal (low-dose), or 750 kcal (high-dose). Analyzed by repeated-measures analysis of covariance with baseline (bl) as a covariate. There was no significant difference in antral area at baseline between the older and young subjects. After nutrient preload ingestion, antral area was significantly greater in both age groups (P < 0.001) than after ingestion of control. Antral area was significantly greater in the older than in the young subjects after preload ingestion (age group x time interaction, P = 0.0006), irrespective of the preload.
Effect of age group
There was no difference between the older and young subjects in antral area at baseline (3.8 ± 0.1 and 3.6 ± 0.1 cm2, respectively). After the preload, antral area was significantly greater in the older than in the young subjects (P = 0.002; age group x time interaction, P = 0.006), irrespective of the preload. Antral area at t = 70 min did not differ significantly between the 2 age groups after the control (3.8 ± 0.3 cm2 in the older subjects and 3.2 ± 0.2 cm2 in the young subjects) or after the high-dose preload (7.8 ± 0.5 cm2 in the older subjects and 7.2 ± 0.5 cm2 in the young subjects). However, it was significantly (P = 0.03) greater in the older subjects than in the young subjects after the low-dose preload (6.14 ± 0.26 and 5.1 ± 0.3 cm2, respectively; treatment effect, P < 0.001; effect of age group, P = 0.005), even though there was no age group x treatment interaction.
As shown in Figure 4, there was an energy-dependent slowing of gastric emptying (T75%) in both groups (treatment effect, P < 0.001), but there was no significant difference between the 2 age groups.
FIGURE 4.. Mean (± SEM) gastric emptying (T75%: time for antral area to decrease to 75% of maximum) between young (n = 12) and older (n = 12) subjects who received preloads of water (0 kcal; control), 250 kcal (low-dose), or 750 kcal (high-dose). Analyzed by repeated-measures two-way ANOVA with stepdown Bonferroni correction to adjust for multiple comparisons. There was no age x treatment effect. There was an energy-dependent slowing of T75% in both groups after the low- and high-dose preloads as compared with T75% after the control (treatment effect: P < 0.001); there was no significant difference between the age groups.
Blood glucose and plasma insulin concentrations
As shown in Figure 5, there was no significant difference between older and young subjects in baseline blood glucose concentrations (6.0 ± 0.1 and 5.9 ± 0.1 mmol/L, respectively). In both age groups, blood glucose concentrations increased significantly (P < 0.001) after ingestion of the 2 nutrient preloads as compared with the control; there was no significant difference between the 2 nutrient preloads. Whereas the magnitude of the increase in blood glucose concentrations after ingestion of the nutrient preloads did not differ significantly between the 2 age groups, this increase was more sustained in the older subjects than in the young subjects (low-dose and high-dose compared with control, P < 0.001; age group x time x treatment interaction: P = 0.03).
FIGURE 5.. Mean (± SEM) blood glucose concentrations (A) and plasma insulin concentrations (B) in young (n = 12, ) and older (n = 12, •) subjects who received preloads of either water (0 kcal; control), 250 kcal (low-dose), or 750 kcal (high-dose). Analyzed by repeated-measures analysis of covariance with baseline (bl) as a covariate. A: There was no significant difference in baseline blood glucose concentrations between older and young subjects. In both age groups, blood glucose concentrations increased after ingestion of the 2 nutrient preloads. The magnitude of this increase did not differ significantly between the 2 age groups, but the increase was more sustained in the older than in the younger subjects (treatment effect, P < 0.001; low-dose and high-dose compared with control, P < 0.001; age group x time x treatment interaction, P = 0.03). B: Plasma insulin concentrations increased in both age groups after ingestion of the 2 nutrient preloads as compared with those after ingestion of the control (P < 0.001), and there was no significant difference between the 2 age groups in the magnitude of this increase.
Plasma insulin concentrations increased significantly (P < 0.001) in both age groups after ingestion of the 2 nutrient preloads as compared with the control, and there was no significant difference in the magnitude of the increase with the 2 preloads. There also was no significant difference in the magnitude of this increase between the 2 age groups.
Plasma CCK concentrations
Effect of preload
As shown in Figure 6, there was a significant (P < 0.001) increase in both age groups in plasma CCK after ingestion of the nutrient preloads as compared with the control. There was no significant difference in the magnitude of this increase between the 2 nutrient preloads (P = 0.09).
FIGURE 6.. Mean (± SEM) plasma cholecystokinin (CCK) concentrations in young (n = 12, ) and older (n = 12, •) subjects who received preloads of water (0 kcal; control), 250 kcal (low-dose), or 750 kcal (high-dose). Analyzed by repeated-measures analysis of covariance with baseline (bl) as a covariate. Baseline CCK concentrations were significantly higher in the older than in the young subjects (P = 0.003). After preload ingestion, CCK concentrations were significantly higher in the older than in the young subjects for all 3 treatments (P = 0.001). In both age groups, there was a significant (P < 0.001) increase in plasma CCK concentrations after the nutrient preloads as compared with the control, and there was no significant difference in the magnitude of this increase between the 2 nutrient preloads (P = 0.09).
Effect of age group
CCK concentrations were significantly higher in the older than in the young subjects at baseline (4.7 ± 0.7 and 2.2 ± 0.3 pmol/L, respectively; P = 0.003) and after preload ingestion (P = 0.001). The magnitude of the increase in plasma CCK immediately after ingestion of the nutrient preloads did not differ significantly between the 2 age groups.
Relations of appetite, energy intake, and plasma CCK to antral area
Relation of hunger and fullness to antral area
There was no relation between baseline scores of hunger or fullness with baseline antral area (data not shown). As shown in Figure 7 and Table 2, there were significant inverse relations between the score for hunger at t = 70 min and the mean score for hunger (t = 10 – 70 min) with both antral area at t = 70 min (ie, immediately before the buffet meal) and the mean antral area (t = 10 – 70 min). As also shown in Figure 7 and Table 2, there were significant positive relations between the score for fullness at t = 70 min and the mean score for fullness (t = 10 – 70 min) with both antral area at t = 70 min and mean antral area (t = 10 – 70 min).
FIGURE 7.. Relation between hunger (A) and fullness (B) at t = 70 min (immediately before a buffet meal) and antral area (cm2) at t = 70 min in young (n = 12, ) and older (n = 12, •) subjects who received preloads of water (0 kcal; control), 250 kcal (low-dose), or 750 kcal (high-dose). Within-subject correlations accounting for repeated measures across treatments were calculated for each combination of antral area and sensations of hunger and fullness by using the method of Bland and Altman (39). A: There were significant inverse relations between hunger ratings and antral area (r = –0.59, P < 0.0001). B: There were significant relations between ratings of fullness and antral area (r = 0.66, P < 0.0001).
View this table:
TABLE 2. Relations between appetite scores, antral area (AA), gastric emptying, and plasma cholecystokinin (CCK) in older (n = 12) and young (n = 12) subjects combined1
Relation of energy intake to antral area
As shown in Figure 8, there were significant inverse relations between energy intake at the buffet meal and antral area at t = 70 min (Table 2) and between energy intake and antral area at t = 100 min (ie, at the end of the buffet meal). Similarly, there were significant relations between energy intake at the buffet meal and the magnitude of increase (t = 100 – 70 min) in antral area (data not shown). There was also a significant inverse relation between energy intake at the buffet meal and gastric emptying (T75%) (Table 2).
FIGURE 8.. Relation between energy intake at a buffet meal and antral area (cm2) at t = 70 min (immediately before the buffet meal) in young (n = 12, ) and older (n = 12, •) subjects who received preloads of water (0 kcal; control), 250 kcal (low-dose), or 750 kcal (high-dose). Within-subject correlations accounting for repeated measures across treatments were calculated for each combination of energy intake (kcal) and antral area by using the method of Bland and Altman (39). There were significant inverse relations between energy intake at the buffet meal and the premeal antral area (r = –0.90, P < 0.0001).
Relation of hunger and fullness to CCK
After ingestion of the preloads, there was an inverse relation between plasma CCK concentrations and scores for hunger, and there was a positive relation between plasma CCK concentrations and fullness (Table 2). There was also an inverse relation between energy intake at the buffet meal and plasma CCK at t = 70 min (Table 2).
Predictors of hunger, fullness, and energy intake
Multiple regression analysis of the combined data showed that both antral area and plasma CCK concentrations at 70 min were predictors of hunger (Table 3). In addition, it showed that antral area at t = 70 min was the only predictor of fullness, and that both antral area at t = 70 min and mean antral area (between t = 10-70 min) were predictors of energy intake (Table 3).
View this table:
TABLE 3. Predictors of energy intake (kcal), hunger, and fullness derived from multiple regression analysis in older (n = 12) and young (n = 12) subjects combined1
DISCUSSION
Our study evaluated the effects of liquid preloads of varied energy content on perceptions of hunger and fullness (satiation), antral area, plasma CCK, and subsequent energy intake (satiety) in healthy young and older adults. We have confirmed (1) that in young adults, antral area, gastric emptying, and energy intake are dependent on the nutrient content of a preload, and satiation is inversely related to antral area (18, 19, 40) and (2) that aging is associated with a reduction in preprandial hunger (2, 3, 7) and an increase in plasma CCK (20–22). New observations after ingestion of a nutrient-containing liquid are that antral area is greater in older subjects than in young subjects who had ingested the same nutrient-containing liquid, hunger is inversely and fullness is directly related to antral area, and the suppression of hunger and energy intake is comparable to that observed in the young. However, arguably the most important observation is that, in both age groups, satiety was related to antral area (and presumably, antral distension), so that preprandial antral area accounted for 34% of the variance in subsequent energy intake.
In considering the gastrointestinal mechanisms involved in the regulation of appetite, it is important to make a distinction between satiation and satiety. Satiation refers to the process that controls the size of a meal by terminating the period of eating, whereas satiety can be described as the state after a meal during which hunger is dampened and the urge to consume food is inhibited (41). The effects of oral nutrients on both postprandial fullness (18, 42–44) and subsequent energy intake (29) correlate relatively poorly with the overall rate of gastric emptying (which of necessity is associated with distension of both the proximal stomach and the distal stomach), which supports the concept that the site of gastric distension may be important in triggering both satiation and satiety. Gastric distension with a balloon suppresses food intake, but the positioning of the balloon in those studies was not precise (9, 10). More recent studies of the relation between upper gastrointestinal sensations (including those of appetite) and gastric distension focused on proximal gastric motor and sensory functions, which were assessed with the use of a barostat technique, and particularly emphasized the pathophysiology of functional dyspepsia (13–16). A limitation of these studies is the recognition that the barostat balloon, when positioned in the proximal stomach, may affect the motility of the distal stomach (45, 46). An alternative possibility—that the distal stomach has an important role in the etiology of appetite related sensations, dyspeptic symptoms, and energy intake—has hitherto received little attention. We initially reported in healthy young subjects that, after ingestion of a glucose drink, the perception of fullness is closely related to the antral area as measured ultrasonographically, as well as to the content of the distal stomach as measured by scintigraphy, but not to the content of the proximal stomach (18). Similar observations were made subsequently by others with the use of solid meals (19). Moreover, in a recent study reported in abstract form, in which gastric volume was quantified with 3-dimensional ultrasound in healthy young subjects, the perception of fullness after a drink was closely related to antral area relative to total gastric volume (46). It is of interest that functional dyspepsia is associated with a wide antrum and increased antral filling (47–49). Accordingly, the current study confirms that satiation is related to antral distension in young subjects and establishes that this is also the case in healthy older subjects. The relation between subsequent energy intake and antral area or content has not previously been evaluated.
We (17) and others (50, 51) established that there is a close concordance between scintigraphic and ultrasound measurements of liquid gastric emptying, although the latter is derived from antral area. In normal subjects after ingestion of solid (52) and nutrient liquid (17) meals, antral content is relatively stable for 30-60 min, whereas there is a progressive reduction in the content of the proximal stomach (53). Although the correlations shown do not necessarily imply causality, our observations suggest that antral distension, rather than the overall rate of gastric emptying or the content of the proximal stomach, is a major determinant of satiety as well as of satiation. The mechanisms mediating the effects of antral distension on satiation and satiety also remain to be defined. The perception of fullness is likely, at least in part, to reflect the activation of gastric stretch receptors by gastric distension (54); substantial variations in patterns of mechanoreceptor activity within the stomach have been observed (55). A number of hormones, which may affect energy intake, are released from the antrum.
In healthy young subjects, oral nutrients suppress subsequent energy intake (8, 31), and the rate of gastric emptying is dependent on the energy content of a meal (12, 40). Healthy aging is associated with abnormalities in a number of gastrointestinal mechanisms that are potentially relevant to appetite regulation, including diminished perception of proximal gastric distension (28), delayed gastric accommodation (28), a more satiating effect of small-intestine glucose than of lipid (7), and greater stimulation of phasic pyloric pressure waves by intraduodenal lipid (27). Studies by ourselves and others (2, 56) suggest that aging may also be associated with slowing of gastric emptying, but the magnitude of any difference is small, and observations have been inconsistent (57). It is therefore not surprising that no difference in gastric emptying between the 2 age groups was evident in this study. We have, however, shown that, after a nutrient preload, antral area is greater in older subjects, and this finding may potentially account for the increase in postprandial fullness. The mechanism or mechanisms underlying the increase in antral area are unknown.
The study of the effect of oral nutrients on energy intake has the potential to offer insights into the pathophysiology of the anorexia of aging, as well as potential approaches to its treatment. As in previous studies (20), hunger was less in the older than in the young subjects; the absence of a significant difference in energy intake is likely to represent a type 2 error—immediately before the study, energy intake was 23% lower in the older subjects than in the young subjects. We showed that there is a strong correlation in healthy older subjects between hunger, as assessed by VAS questionnaires, and subsequent food intake (58). Rolls et al (31) reported that energy compensation in response to different yogurt preloads is less precise in older than in young men, so that the suppression of energy intake is relatively greater in the young. Other studies also suggest that the precision of energy intake regulation declines with age (30). In contrast, Zandstra et al (59) reported that the capacity to regulate energy intake after a nutrient preload does not differ between children, young adults, and older subjects, which is consistent with our observations. These discrepant results may be attributable to differences in study design, particularly with respect to the volume and composition of the preloads and the interval between the preload and the meal. For example, in the study by Rolls et al (31), the period between the preload and the test meal was 30 min, whereas it was 90 min in the study by Zandstra et al (59) and 60 min in our study.
CCK, which is released from the small intestine after ingestion of nutrients, appears to stimulate both satiation and satiety (5, 6, 11, 20, 60). For example, in young adults, intravenous infusion of CCK in physiologic concentrations results in a dose-dependent suppression of energy intake (20), which may be mediated by stimulation of vagal afferent activity, slowing of gastric emptying, an increase in gastric sensitivity, or all of those factors (6, 11, 20). In young subjects, the suppression of food intake by exogenous CCK is enhanced by gastric distension (11). We have confirmed that both fasting and postprandial CCK is increased in healthy, older subjects (20–22, 61). It is clear that the effect of exogenous CCK on appetite is maintained (and may be increased) in older subjects (20), which is consistent with a role for CCK in the anorexia of aging. The greater increase in blood glucose and plasma insulin concentrations after ingestion of oral nutrients observed in older than in young subjects is indicative of the impaired glucose tolerance and insulin resistance associated with aging. Whereas a role for either glucose or insulin in the anorexia of aging cannot be discounted, that possibility appears unlikely (38, 62).
ACKNOWLEDGMENTS
We thank Justin Lokhorst and Kristyn Willson for assistance with the statistical analysis.
KS and BP were involved in designing the experiment, collecting and analyzing the data, and writing the manuscript, JW was involved in analyzing the data and performing the plasma hormone assays, KLJ provided advice and training on the ultrasound equipment, CF-B, IC, and MH were involved in designing the experiment, interpreting the data, and writing the manuscript. None of the authors had any conflict of interest to declare in relation to this study.
REFERENCES