| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH COMMUNICATION |
1 From the Institute of Human Nutrition, School of Medicine, University of Southampton, Southampton, United Kingdom (RJ Stratton and ME), and the Rowett Research Institute, Aberdeen, United Kingdom (RJ Stubbs)
2 Supported by the Medical Research Council, United Kingdom. 3 Reprints not available. Address correspondence to RJ Stratton, Institute of Human Nutrition, School of Medicine, University of Southampton, Level F (MP 113) Centre Block, Southampton General Hospital, Tremona Road, Southampton, SO16 6YD, United Kingdom. E-mail: r.j.stratton{at}soton.ac.uk.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objective: We tested the hypothesis that 1) bolus TF does not suppress appetite and food intake and 2) there is no interrelation between food intake and appetite mediators (including ghrelin).
Design: A single-blind, placebo-controlled trial within which 6 healthy men [body mass index (in kg/m2): 21.1 ± 1.61] received 3 d of bolus TF (6.93 ± 0.38 MJ/d of 4.18 kJ/mL multinutrient feed). For 2 d before and after TF, placebo boluses (<0.4 MJ/d) were given by tube. Hourly tracking of appetite, weighed measurements of daily ad libitum food intake, and metabolic and hormonal (including ghrelin) measurements were undertaken.
Results: Total energy intake was significantly increased with bolus TF (18.2 ± 1.86 MJ; P = 0.0005) despite a partial reduction in food intake compared with placebo periods (P = 0.013) and during the TF period (by 15%; P = 0.007). There was little change in hunger and fullness with bolus TF, and within-day temporal patterns did not differ whether TF or placebo was given. Changes in fasting concentrations of ghrelin (1003.6–756.0 pmol/L; P = 0.013) and other mediators (including leptin, insulin, and glucose) were significantly related to subsequent daily food intake (eg, ghrelin: r2 = 0.81, P = 0.022).
Conclusions: In this short-term study, subjects maintained appetite ratings during bolus TF by a significant reduction in food intake and changes in ghrelin and some appetite mediators related to subsequent daily food intake. Longer-term studies are required to fully ascertain the effect of TF on appetite, food intake, and appetite mediators.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
50% of patients, it is uncertain whether TF suppresses appetite and replaces energy taken orally or whether TF provides additional energy to aid in the treatment of the undernourished patient. Controlled study of the effect of TF on appetite control in healthy subjects has been infrequently used because of difficulties recruiting volunteers who may be reluctant to be exposed to the inconvenience of nasogastric tube insertion, of keeping the tube in situ for several days or weeks, and for the restriction of physical activity during pump-assisted feeding. However, with the recent advent of portable lightweight infusion pumps this last issue is much less of a problem. We have previously shown that 3-d administration of liquid feeds providing a mixture of macro- and micronutrients (continuously over 12–24 h) is well tolerated (as are fine-bore nasogastric tubes kept in situ for longer than a week). Importantly, this study also suggested that continuous TF did not suppress appetite and food intake (5). However, the effects of intermittent tube feeding with boluses of feed (which are more physiologic than slow, continuous infusions over 12–24 h) has not been addressed. It is possible that bolus TF may also fail to induce satiety as liquid feeds are administered and the upper gastrointestinal tract and the associated cephalic phase response [with a role in preabsorptive control of appetite (6-8)] are bypassed. Alternatively, it could be that the intermittent delivery of nutrients as boluses, reflecting a more physiologic meal-like pattern, suppresses appetite and food intake to a much greater extent than does a slow continuous liquid infusion of nutrients over time. Although a few studies have examined the effect of liquid bolus TF on appetite or food intake (9-11), these have typically involved short infusions (<3.5 h) of single macronutrients (as opposed to multinutrient liquid feeds) or were poorly controlled (confounded by changes in physical activity or disease) and have failed to consider appetite and food intake, together with changes in putative appetite mediators [including circulating concentrations of hormones (eg, ghrelin, leptin, insulin), metabolites (eg, glucose) etc]. Therefore, this single-blind, placebo-controlled trial tested the hypothesis that liquid TF delivered intermittently as boluses by nasogastric tube over days suppresses appetite and ad libitum food intake (unlike continuous TF schedules) by effects on a range of metabolic and hormonal mediators of appetite and satiety. |
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
± SD) age: 33 ± 8.4 y; weight: 67.7 ± 7.2 kg; body mass index (in kg/m2): 21.1 ± 1.61; with stable weight for 1 y (±2 kg); body fat: 19.1 ± 3.7%). The entire study was conducted in a metabolic suite to control for the effect of physical activity on appetite and food intake. Ethical approval was obtained for this study from the Local Research Ethics Committee, and subjects gave their informed consent before taking part. All subjects completed the study.
Protocol
This study was a single-blind, placebo-controlled design. A fine-bore nasogastric tube (Freka 8F, 120 cm; Fresenius, Warrington, United Kingdom) was inserted at the start of the 9-d study, to give subjects 2 d to become accustomed to it (days 1–2). During these 2 d (days 1–2), subjects were given a fixed "maintenance" diet to consume orally [energy content equivalent to 1.5 times predicted basal energy expenditure (BEE) (12), 10.67 ± 0.68 MJ] to facilitate energy balance with nothing given by nasogastric tube. For the remainder of the study (days 3–9), boluses were given by the nasogastric tube of a placebo for 2 d before (days 3–4) and after (days 8–9) 3 d of a feed (days 5–7). During days 5–7, a liquid, fiber-free tube feed was given (6.93 ± 0.38 MJ; 15% energy from protein, 49% carbohydrate, 35% fat; 4.18 kJ/mL; vitamins, minerals, and trace elements; Nutrison, Nutricia, Trowbridge, United Kingdom), the daily energy content of which was equivalent to 1 times predicted BEE (12). The tube feed was administered intermittently as 4 boluses, each during a 20-min period (average volume: 415 ± 23.5 mL; average rate: 0.09 MJ/min) at the following times during the day: 0900, 1230, 1600, and 1930. For the 2 d before (days 3–4) and after (days 8–9) TF, a placebo (colored water; energy provision <0.4 MJ) was given. This had a similar appearance to the tube feed and was given as boluses in identical volumes, at the same rate, and at the same times of day as the tube feed. On questioning subjects at the end of the study, they were unaware of the use of a placebo feed. Throughout days 3–9, subjects consumed ad libitum covertly manipulated food items that each were isoenergetically dense (550 kJ/100 g) and had the same macronutrient composition (40% energy from fat, 47% from carbohydrate, and 13% from protein). There were 11 different food items available daily as part of a 3-d rotating menu cycle (menu A, B, and C, started on day 3 in random order for each subject) [see Stubbs et al (13) for more details]. All food was preprepared in excess and kept in a refrigerator designated to each subject, and a microwave oven was available for heating food when required. Subjects had free access to caffeine-free beverages throughout the study period.
Measurements
The following measurements were undertaken.
Food intake
Subjects were asked to record the timing of food and drink consumption in a food diary, and a weighed food inventory was performed to assess daily (24 h) oral energy intake and its distribution through the day on days 3–9. Assessment of the energy intake orally from covertly manipulated foods was undertaken with the use of the RONA computer package, created by the Rowett Research Institute, Aberdeen, United Kingdom, and based on the Royal Society of Chemistry's Composition of Foods database (14). Assessment was made of the pleasantness and satisfying nature of the food after each meal with the use of a visual analogue scale (VAS) (scored from 0 mm to 100 mm for the least to the most pleasant and satisfying) (15). The Dutch Eating Behavior Questionnaire to assess dietary restraint was completed by all subjects at the end of the study (16).
Appetite sensations
Subjects rated their appetite sensations each waking hour of every day of the study with the use of paper VAS questionnaires. The questionnaire consisted of 6 VASs to rate "hunger," "fullness," "desire to eat," "how much can you eat now?," "urge to eat," and "preoccupation with thoughts of food" (15, 17). Each subject was given a booklet of questionnaires (containing one questionnaire for each hour) and a small hourly timer to remind them to complete it.
Anthropometry
Body weight was measured (digital platform scale accurate to within 0.001 kg; Sauter, Ebingen, Germany) at the same time on each morning of the study after voiding, and height was measured with the use of a stadiometer (Karimetre; Raven Equipment Limited, Dunmow, Essex, United Kingdom) at the start of the study. Body mass index (in kg/m2) was calculated, and measurements of skinfold thickness at 4 sites (triceps, biceps, suprailiac, subscapular) were undertaken, and the percentage of fat was calculated (18) at the start of the study (day 1) and at the end of the following study periods: after the maintenance diet (before TF started; day 3), after 2 d of placebo bolus TF (day 5), and after 3 d of bolus TF (day 8), in line with previous study methods (5). Results were not divulged to the subjects.
Indirect calorimetry: respiratory quotient and resting energy expenditure (days 3, 5, and 8)
Resting energy expenditure (REE) was measured at the same time of the morning (0815) on day 3 (after the maintenance period before TF started), day 5 (after 2 d of placebo TF), and day 8 (after bolus TF period), in line with previous study methods (5). The ventilated hood technique was used with a Deltatrac Metabolic Monitor (MBM-100; Datex Instrumentarium Corp, Helsinki, Finland), and all measurements were conducted at room temperature (23 ± 1 °C). After a 30-min warm-up period, the Deltatrac was calibrated with standard gases, and pressure and basal measurements were begun (without the subject) until the machine was equilibrated. Subjects were recumbent for measurement of their metabolic rate, having just woken in the morning and were fasting. They were asked to relax, but not sleep, and lie still with their arms by the side of their body with legs uncrossed during the measurement. The ventilated hood (made of clear plastic and ventilated with a constant flow of air 40 L/min) was placed over the subject's head to measure inspired and expired gases. Oxygen consumption, carbon dioxide production, and respiratory exchange rates (RQ) were obtained every minute for 30 min, and REE was calculated (19). On completion of the measurement period (30 min), the hood was removed from over the subject's head, and further basal measurements (without the subject) were collected. Finally, calibration of gases and pressure were rechecked. The accuracy of the Deltatrac (98–100% of predicted values) was checked periodically with the use of nitrogen (80%) and carbon dioxide (20%) infusions (measured with the use of an oil-filled gas meter type DM3A; Alexander Wright and Co, London, United Kingdom) (20).
Blood sampling for the measurement of metabolites and hormones in the fasted state (days 3, 5, and 8)
After measurement of REE, early morning on day 3 (after completion of the 2-d maintenance period), day 5 (after the placebo period), and day 8 (after the 3-d TF period), venous fasting blood samples were taken for the measurement of plasma concentrations of the following hormones: ghrelin (radioimmunoassay; Linco Research Inc, St Louis, MO), leptin (radioimmunoassay; Linco Research Inc), insulin (enzyme amplified sensitivity immunoassay; Biosource Europe SA, Nivelles, Belgium), cholecystokinin (radioimmunoassay; Euro-Diagnostica, Malmo, Sweden), and glucagon (radioimmunoassay; Euro-Diagnostica); and the following metabolites: glucose (Gluc HK, Unimate 5; Roche Diagnostica Instruments, Basel, Switzerland), glycerol (Boehringer Mannheim GmbH, Mannheim, Germany), lactate (21), β-hydroxybutyrate (21), nonesterified fatty acids (NEFAs; Wako Alpha Laboratories, Eastleigh, United Kingdom), and triacylglycerol (Unimate 5 TRIG kit; Roche Diagnostica Instruments).
Statistical methods
Repeated-measures analysis of variance (ANOVA) was undertaken on data that were normally distributed. Time (day) was the within-subject factor. Post hoc analysis with the use of polynomial contrasts (eg, linear, quadratic) enabled statistical analysis of 1) linear effects across time (days 3–9, days 3–4, days 5–7, days 8–9) and 2) the quadratic effect (comparison of the TF period with the colored water periods collectively). Two-factor repeated-measures ANOVA with deviation contrasts was used to compare the individual, within-day, hourly ratings of hunger and fullness with the daily grand mean. Parametric data are presented as mean ± SD. To calculate the correlation coefficients for repeated metabolic measurements, energy intake, and appetite, analysis of covariance was used. These analyses were undertaken on normally distributed data, and data were presented as mean ± SD. Leptin concentrations, which required log transformation to normalize the positively skewed distribution, were presented as geometric mean + SD (the antilog of the mean of logged data + 1 SD of logged data). For nonnormally distributed data, analyses were performed with the use of Friedman's k-related samples for repeated measurements and Wilcoxon's signed rank test (paired comparisons), and data were presented as median (range). Statistical analysis was performed with the use of SPSS, version 11.5 (SPSS Inc, Chicago, IL). Sample size calculations based on the intraindividual variability in food intake for 2-d periods (SD: 12%; 1.67 MJ) suggested that 6 subjects were required to detect a 20% change in food intake (2.8 MJ) with 80% power and a significance of P < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Food (oral) and total (oral + tube) energy intake
Bolus TF for 3 d universally increased total (oral + tube) energy intake to 18.2 ± 1.86 MJ compared with the placebo periods before and after TF (Figure 1; P = 0.0005). Overall, food (oral) intake was significantly reduced with bolus TF compared with the placebo periods (Figure 1; P = 0.013). This suppression of food intake with bolus feeding was equivalent to 39.7% (2.75 MJ) of the tube feed energy infused, leaving 60.3% (4.18 MJ) additive to oral energy intake. During the 3-d bolus TF period, food energy intake significantly declined (by 15%) to 9.94 ± 1.99 MJ, so that only 40% of the tube feed energy was additive (Figure 1). During the 2 placebo periods (days 3–4, days 8–9), no significant differences were observed in food (oral) or total (oral + tube) intake.
|
Results from the Dutch Eating Behavior Questionnaire suggested that 2 subjects were restrained eaters (dietary restraint score > 20) (16). When the 2 restrained eaters were excluded from the analysis (dataset n = 4), there remained a significant increase in total energy intake with bolus TF (17.7 ± 1.90 MJ) compared with the placebo periods before and after TF (P = 0.01). There remained a reduction in food (oral) intake (P = 0.06) compared with placebo periods, and during the 3-d bolus feeding period there remained a reduction in food intake (12%).
Food pleasantness and satisfaction ratings
No significant difference was observed during the bolus TF study (days 3–9) in the daily pleasantness (mean daily scores: 63–71 mm) or satisfaction ratings (69–75 mm) for food. The pleasantness and satisfaction ratings of the food consumed with the 3 different menus (A, B, and C) did not differ significantly (eg, pleasantness scores: menu A, 68 ± 13 mm; menu B, 72 ± 11 mm; menu C, 66 ± 13 mm; satisfaction scores: menu A, 72 ± 9 mm; menu B, 72 ± 13 mm; menu C, 72 ± 12 mm), and no significant differences were observed in oral energy intake according to menu (A, B, and C).
Appetite sensations
No significant differences were observed in the first hour VAS ratings of appetite sensations across the study period as a whole or during the bolus TF period specifically. The only exception was the preoccupation with thoughts of food ratings, which varied significantly during the bolus TF period (P < 0.0, repeated-measures ANOVA). Median daily scores for each of the appetite sensations did not differ significantly during the bolus TF period (days 5–7) or throughout the whole study period, except for preoccupation with thoughts of food scores (P < 0.001, Friedman's k-related samples test). For simplicity, the median results for each of the appetite sensations for the 3 study periods are presented in Table 1
. During 3 d of bolus TF, within-day hunger and fullness ratings varied significantly, but no differently to when only placebo boluses were given as shown in Figure 2.
|
|
Relations between the metabolic and hormonal changes associated with bolus TF, hunger, and food intake shortly after the time of measurement
Changes in the following measured mediators of appetite during bolus TF (Table 2) related positively (ghrelin: r2 = 0.81, P = 0.022; NEFAs: r2 = 0.87, P = 0.002) or negatively (leptin: r2 = 0.87, P = 0.002; insulin: r2 = 0.81, P = 0.014; glucose: r2 = 0.77, P = 0.049; glucagon: r2 = 0.79, P = 0.03; RQ: r2 = 0.86, P = 0.004) to the daily food (oral) energy intake (on day 3: 14.56 ± 3.09 MJ; day 5: 11.68 ± 2.02 MJ; and day 8: 11.91 ± 2.87 MJ, analysis of covariance). No such relation was found between cholecystokinin, triacylglycerol, lactate, β-hydroxybutyrate, or REE and daily energy intake. None of the changes in mediators related to or hence predicted appetite (eg, first hour hunger) or food intake (morning energy intake) shortly after the measurements were made.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The design of the study (with placebo periods before and after TF) also allowed us to confirm that there was no suggestion of volume effects or effects of the method of feeding per se (eg, subjects being more conscious of having boluses during the day) causing the reduced food intake. Similarly, the subjects were given time to become accustomed to the nasogastric tube before TF commenced. Therefore, a possible explanation, as hypothesized, is that liquid boluses given by tube are more satiating than continuous TF infusions, reducing appetite sensations. In this short study we found no differences in appetite sensations (24 h, first hour of the day, hourly within day), whether subjects were bolus fed or given a placebo. As indicated in Figure 2, even the hourly tracked hunger and fullness sensations were no different whether the bolus given was a placebo or feed. This suggests that persons maintain the same degree of hunger and most other appetite sensations during bolus feeding by eating significantly less (only ratings of "preoccupation with thoughts of food" varied significantly with no clear pattern during the study period).
The differences in food intake observed with TF may partly be explained by changes in appetite mediators, a variety of which were investigated. In the present study, the marked increase in total energy intake with bolus TF (Figure 1) was associated with a significant increase in some of the putative satiety mediators measured (leptin, insulin, glucagon, RQ, REE) (22) and a reduction in the appetite stimulant ghrelin (see Table 2
). Studies have suggested that ghrelin, a 28 amino acid growth hormone secretagogue, is suppressed by meals, and intravenous infusions of ghrelin increase food intake and appetite (23, 24). For the first time, this study has longitudinally assessed ghrelin, and other appetite mediators, in relation to changes in food intake and appetite sensations in those receiving bolus TF. Fasting ghrelin concentrations during the study significantly related to subsequent (24 h) food intake. Similarly, changes in fasting circulating concentrations of leptin, NEFAs, insulin, glucose, and glucagon and changes in RQ significantly related to the change in 24-h food intake, at least within the time frame of this study. This contrasts with previous investigations of continuous TF regimens in which changes in a variety of appetite mediators did not relate to changes in daily food intake (5, 25). Notably, with all schedules (bolus and continuous), no relation was found between any of the measured appetite mediators and measures of appetite or food intake shortly after the measurements. To more fully understand the effect of changes in putative mediators of appetite during different TF schedules and subsequent appetite sensations and food intake, more frequent measurements of these metabolic or hormonal variables during the day and night are required and for longer study periods.
Although bolus TF elicited a variety of metabolic changes associated with the significant reduction in food intake observed, total energy intake was still markedly increased. The reduction in food intake with bolus TF was only equivalent to 40% of TF energy during this 3-d period, producing a small increase in body weight but no detectable change in percentage of body fat. Substantial increments in total energy intake also occurred with continuous TF of liquids, because neither appetite nor food intake was reduced (5). In both cases the increment in total energy intake may be facilitated by delivering nutrients by tube, thereby bypassing the cephalic phase response and upper gastrointestinal tract. The short duration of feeding (3 d) might also be a factor, and studies of a longer duration are warranted. The delivery of nutrients as a liquid is also likely to be important. Although intervention studies do not conclusively show that liquids given orally are less satiating than solids (26), the rise in consumption of energy-rich drinks has been linked to the obesity epidemic (27). Although the liquid feed used in this study was a multinutrient, milk-based feed, studies of similar feeds given orally also suggest little effect on appetite or on voluntary food intake (2).
Therefore, it appears that liquids given by tube (with no oral or upper gastrointestinal contact or sensory influences) are a particularly effective way of increasing total energy and nutrient intake, at least in the short term. Not only is this an important observation about liquids and appetite control, but it also has implications for the clinical use of liquids and TF as a means of nutritional support (2). In particular, the study provides important insights for those persons who receive supplementary enteral feeding. For such persons who receive TF as a supplement to food intake, this study suggests bolus schedules are less effective than continuous schedules at increasing total energy or nutrient intake, with greater suppression of food intake. Consequently, bolus TF may be more useful in those persons in whom food intake is contraindicated for long periods of time. However, before definitive conclusions can be made, the limitations of the present study need to be acknowledged. These include the subject group (healthy men, including 2 restrained eaters), its sample size (n = 6), the controlled environment of the metabolic suite, and the short duration of bolus TF (3 d). Therefore, there is a need for further investigation in larger groups of healthy subjects and in patients (both men and women) in controlled and clinical settings to address more fully the effects of both bolus and continuous TF of liquids on appetite and food intake. Longer term trials are also indicated, because an increasing number of patients are fed artificially by tube for long periods of time (1, 28) and to gain greater insights into liquids and appetite control. Furthermore, the current trial specifically investigated supplementary TF because food intake was one of the key outcome variables. It did not address the effects of TF as a sole source of nutrition on appetite. Although our earlier studies indicate distressing appetite sensations in patients receiving TF as a sole source of nutrition (3, 4), a more controlled study that also assesses appetite mediators is required.
In summary, this is the first placebo-controlled investigation of the effect of short-term bolus TF on a variety of appetite mediators, appetite sensations, and food intake in the absence of confounding factors such as disease (29). Bolus TF universally and significantly increased total (oral + tube) energy intake, with >60% of tube feed energy additional to that taken orally. Appetite sensations remained largely unaffected by the additional energy from liquid bolus TF because subjects ate significantly less compared with placebo periods and during the 3-d course of feeding. Accompanying the reductions in food intake was a significant decline in circulating ghrelin concentrations and changes in a range of other putative appetite signals. Although this pilot study in healthy men suggests that bolus TF is well tolerated and may be more satiating than a continuous TF schedule, longer term studies, including investigations in the clinical setting, are required to fully ascertain the implications for the effect of liquids on appetite control and for the use of this method of feeding in the clinical setting.
|
ACKNOWLEDGMENTS |
|---|
There were no conflicts of interest at the time the study was undertaken and data were analyzed. There was no commercial input into any aspect of this study. Since completion of the study, RJ Stratton has joined Nutricia, United Kingdom, while maintaining a position at the University of Southampton, and RJ Stubbs now works for Slimming World, United Kingdom.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
