401
ARTICLE Chewing gum increases energy expenditure before and after controlled breakfasts Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 04/12/15 For personal use only.
Daniel L. Kresge and Kathleen Melanson
Abstract: Chewing has been associated with improved satiation and satiety, but little is known about the metabolic impact of gum chewing. We tested the hypothesis that gum chewing would increase energy expenditure (EE) and reduce respiratory exchange ratio (RER) before and after a controlled test meal. Seventeen males and 13 females (age 21.5 ± 6.6 years, body mass index 23.9 ± 2.8 kg/m2) participated in a randomized crossover study in which subjects chewed sugar-free gum for a total of 1 h (3 sessions of 20 min) on the test day (GC) and did not chew gum on a control day (NG). EE and RER were measured by indirect calorimetry after an overnight fast. Subjects consumed a breakfast shake containing 30% of their measured energy needs, and then postprandial EE and RER were measured for 3 h. Blood glucose (GLC) was measured in the fasting and postprandial states at regular intervals. Fasting EE was higher during GC (1.23 ± 0.04 kcal/min; 1 kcal = 4.2 kJ) than during NG (1.17 ± 0.04 kcal/min; p = 0.016). Postprandial EE was also higher during GC (1.46 ± 0.05 kcal/min) than during NG (1.42 ± 0.05 kcal/min; p = 0.037). Fasting and postprandial RER and GLC did not differ between GC and NG. The findings demonstrate that GC is associated with higher fasting and postprandial EE without altering blood glucose or substrate oxidation as measured by RER. These data suggest that gum chewing potentially could influence short-term energy balance in this population; however, longer-term research is needed. Key words: mastication, resting metabolic rate, thermic effect of food, respiratory exchange ratio, indirect calorimetry, energy balance, substrate oxidation, postprandial. Résumé : Mâcher de la gomme est associé, selon des études, a` une meilleure sensation de rassasiement et de satiété, mais il y a peu d’études sur l’impact métabolique de la gomme a` mâcher. Nous vérifions l’hypothèse selon laquelle mâcher de la gomme suscite une augmentation de la dépense énergétique (« EE ») et une diminution du ratio d’échanges gazeux (« RER ») avant et après un repas test contrôlé. Treize femmes et dix-sept hommes âgés de 21,5 ± 6,6 ans et présentant un indice de masse corporelle de 23,9 ± 2,8 kg/m2 participent a` une étude organisée selon un devis croisé aléatoire dans laquelle les sujets mâchent de la gomme sans sucre au jour test (« GC ») durant 60 min au total (3 séances de 20 min) et ne mâchent pas de la gomme au jour de contrôle (« NG »). On mesure EE et RER par calorimétrie indirecte après un jeûne d’une nuit. Les sujets consomment une boisson pour déjeuner contenant 30% de leurs besoins énergétiques mesurés et on mesure durant 3 h les EE et RER postprandiaux. À intervalles réguliers, on mesure la concentration de glucose sanguin (« GLC ») dans les conditions de jeûne et postprandiale. EE a` jeun est plus élevée durant GC (1,23 ± 0,04 kcal/min) comparativement a` NG (1,17 ± 0,04 kcal/min; p = 0,016). EE postprandiale est aussi plus élevée durant GC (1,46 ± 0,05 kcal/min) comparativement a` NG (1,42 ± 0,05 kcal/min; p = 0,037). La valeur a` jeun et postprandiale de RER et de GLC ne varie pas entre GC et NG. D’après ces observations, GC est associé a` une EE plus élevée a` jeun et après un repas, et ce, sans modification de la concentration sanguine de glucose et de l’oxydation des substrats évaluée par le RER. D’après ces observations, mâcher de la gomme pourrait avoir un impact sur l’équilibre énergétique a` court terme dans cette population; néanmoins, il faut réaliser des études a` long terme. [Traduit par le Rédaction] Mots-clés : mastication, métabolisme de repos, thermogenèse alimentaire, ratio d’échanges gazeux, calorimétrie indirecte, équilibre énergétique, oxydation des substrats, postprandial.
Introduction Gum chewing has been promoted as beneficial in terms of oral health, cognitive performance, alertness, stress relief, and weight management (Leveille et al. 2008). Proposed weight management benefits of gum chewing are based mainly on energy intake research. For example, cravings for and intake of sweet snacks may be decreased by gum chewing (Hetherington and Boyland 2007), although results are not entirely consistent (Julis and Mattes 2007). If gum chewing is to be promoted for the purposes of weight management (Leveille et al. 2008), then its effects on energy balance, substrate oxidation, and other potential metabolic outcomes should be explored. To our knowledge, only 2 studies have examined gum chewing’s effect on energy expenditure. In one of these studies, 7 subjects
chewed the equivalent of 4 sticks of gum for 12 min at a very rapid cadence, and energy expenditure was measured immediately before, during, and after (Levine et al. 1999). Although an increase in energy expenditure of 19% was reported, the paradigm was not typical of free-living gum chewing, and the protocol was limited to 7 short tests. In a study designed to determine energy expenditure while chewing gums with various doses of caffeine and (or) nicotine, energy expenditure of chewing the control gum at a natural pace was estimated as 5% higher than that of no gum chewing (Jessen et al. 2003). This gum chewing was measured in the fasting state. The effects on substrate oxidation within the control condition were not clear, and postprandial measures were not taken.
Received 23 June 2014. Accepted 7 November 2014. D.L. Kresge and K. Melanson. Department of Nutrition and Food Sciences, University of Rhode Island, Kingston, RI 02881, USA. Corresponding author: Kathleen J. Melanson (e-mail: [email protected]). Appl. Physiol. Nutr. Metab. 40: 401–406 (2015) dx.doi.org/10.1139/apnm-2014-0232
Published at www.nrcresearchpress.com/apnm on 18 December 2014.
402
More work is needed to understand the potential role gum chewing may have in avoiding a positive energy balance on a regular basis. Gum chewing under natural situations should be simulated, and multiple chewing bouts should be examined. Thus, the purpose of this study was to compare energy expenditure, substrate oxidation, and blood glucose under conditions of gum chewing at a selfdirected pace versus no gum chewing in both the fasted and postprandial states during 2 mornings within the same individuals.
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 04/12/15 For personal use only.
Materials and methods Study design This study used a randomized crossover design, in which volunteers served as their own controls, with 2 conditions and 9 time points. Each subject chewed gum (GC) on one test day and did not chew gum (NG) on the other test day. We report here on energy expenditure (EE), respiratory exchange ratio (RER), and capillary glucose (GLC), which were measured as part of a broader study examining the relationship between gum chewing and energy balance. Appetite results are reported elsewhere. Volunteers provided written informed consent prior to participating, and they received a $60.00 stipend for completing the study. The Institutional Review Board of the University of Rhode Island approved this study. Subjects Thirty nonsmoking men and women were recruited by advertisements and class announcements. A preliminary screening by telephone was used to verify inclusion criteria of 18–48 years and body mass index (BMI) of 18.5–29.9 kg/m2 and to apply exclusion criteria, which included factors that could influence metabolism or appetite: pregnancy or lactation, diabetes mellitus or other diseases affecting energy balance, currently dieting or exercising to lose weight, or weight change of greater than 3 pounds during the past 6 months. Preliminary data collection The protocol required 3 visits. During the first visit, volunteers completed general health questionnaires and the International Physical Activity Questionnaire (Craig et al. 2003), which were used for descriptive purposes only. Waist circumference was measured using a physician’s tape measure at the level of the umbilicus after normal exhalation. Body composition was determined at least 4 h after subjects had eaten, using a validated air displacement plethysmography system (BODPOD, Life Measurements Inc., Concord, Calif., USA) (Fields et al. 2002; McCrory et al. 1998). Body weight was measured to the nearest 0.1 kg on a calibrated digital scale that was part of the BODPOD system. Height was measured to the nearest millimeter using a wall-mounted stadiometer (Seca, Hamburg, Germany). BMI was calculated, and body composition was determined by the BODPOD software using the Siri equation (Siri 1961). Subjects completed the 3-min submaximal Queen’s College Step Test to estimate maximal aerobic capacity (McArdle et al. 1972). Next, each volunteer was familiarized with test-day procedures by undergoing a 10-min “practice” metabolic measurement. Finally, subjects chose 1 of 6 flavors of commercially available sugarless gum (Orbit, Wm. Wrigley Co., Chicago, Ill., USA) to chew during the GC condition. Preparation days The next 2 visits were test visits. Each test day was preceded by a preparation day, during which volunteers maintained normal diet and activities of daily living, abstained from alcohol and strenuous exercise for at least 24 h prior to testing, and restricted caffeine for at least 18 h prior to testing. Subjects began fasting 10 h prior to their scheduled test visit, except for water. Since most subjects arrived at 0700, their fast began by 2100 the previous night. Subjects wore a validated physical activity monitor (Actical, Mini Mitter, Bend, Ore., USA) on both preparation days to confirm
Appl. Physiol. Nutr. Metab. Vol. 40, 2015
compliance with activity instructions (Esliger and Tremblay 2006; Trost et al. 2005). Subjects followed these restrictions the day before each test visit, so that the 2 preparation days were matched for diet and physical activity. Test days were separated by at least 48 h. Preprandial measurements Only one subject was measured per test day. For women, testing occurred during the mid-follicular phase of the menstrual cycle (days 5–13) to control for potential metabolic shifts across the cycle (Brennan et al. 2009; Bryant et al. 2006; Dye and Blundell 1997). Volunteers reported to our laboratory after a 10-h overnight fast, engaging in the least amount of activity possible to arrive in the laboratory. After voiding, subjects relaxed in an elevated supine position on the laboratory bed. The first capillary blood sample was drawn by lancet from the fingertip for analysis of GLC by autoanalyzer (Hemocue Glucose 201, Hemocue, Mission Viejo, Calif., USA). During the next 30–45 min, volunteers rested while 24-h diet and physical activity recalls were administered to confirm protocol compliance. Diet recalls were subsequently analyzed by Food Processor SQL 9.0 (ESHA Research, Salem, Ore., USA) and physical activity recordings were analyzed by Actical software, version 2.04. At the end of the rest period, subjects underwent a 45-min measurement of resting EE and RER (carbon dioxide output/oxygen uptake) by ventilated hood indirect calorimetry (Vmax, Sensormedics, Yorba, Calif., USA). Flow rate and gas calibrations were performed daily using 2 calibration tanks containing different CO2 and O2 concentrations, according to the manufacturer’s recommendations. During testing, volunteers remained in an elevated supine position and were permitted to read, study, or listen to music. They were not permitted to sleep, write, or hold any materials in their hands. The laboratory was kept quiet, and volunteers were provided with blankets for their comfort. Subjects were monitored closely to assure compliance with the protocol. Ten minutes into the resting EE measurement during the GC condition, data collection was temporarily halted using the pause feature of the calorimeter’s software. The transparent ventilated hood was lifted briefly and a single piece of sugarless gum (1.9 g, 3 kcal; 1 kcal = 4.2 kJ) was placed in the volunteer’s mouth. The hood was immediately replaced, and calorimetry resumed. Subjects chewed gum at their own natural pace for 20 min. They were not permitted to blow bubbles. At the end of the 20-min chewing period, calorimetry was again paused briefly, the gum was collected, and calorimetry resumed. During the NG condition, passing and collection of gum was simulated using the same procedures, except that no gum was given. Data from 1 min after passing, or simulation, were excluded from EE and RER calculations. At the conclusion of the resting EE measurement, the ventilated hood was removed, 45-min resting EE and RER were calculated, and another capillary blood sample was drawn for determination of GLC. Test meals and postprandial procedures Volunteers were given a nutritional shake for breakfast (Boost, Novartis) standardized to 30% of their measured 45-min resting fasting EE during each test day (518 ± 12 mL; 518 ± 12 kcal; 68.3% carbohydrate, 16.7% protein, 15% fat). Subjects consumed the shake within 15 min. Postprandial GLC was measured at 15, 30, 45, 60, 90, 120, 150, and 180 min. After the first postprandial blood draw had been completed (at 15 min), postprandial measurements of EE and RER began, using the same procedures previously described. The hood of the indirect calorimeter remained in place throughout the 3-h postprandial measurement period, except for brief intervals during the scheduled blood draws. During GC, volunteers chewed gum at 2 additional times for 20 min each time (60–80 min and 150–170 min postprandial), for a total of 60 min of chewing time over the morning of testing. In the postprandial state, gum was passed to volunteers after blood sampling, immePublished by NRC Research Press
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 04/12/15 For personal use only.
Kresge and Melanson
diately prior to application of the ventilated hood. The gum was collected (or collection was simulated) after 20 min of chewing using the same procedures as previously described, and EE measurement continued until the next scheduled blood sampling. At the end of the 3-h postprandial measurement period, a final blood sample was drawn, and then volunteers were offered an ad libitum pasta lunch with water. At the conclusion of lunch on the first metabolic test day, volunteers were given copies of their diet and activity recalls and the physical activity monitor, along with instructions to repeat their activity and dietary intake (including portion sizes) as closely as possible during the next preparation day. On their next test day, volunteers repeated the protocol under the other randomized test condition. Data analysis Data are presented as means ± standard error of the mean (SE), unless otherwise noted. Repeated-measures analysis of variance (ANOVA) was used to compare EE, RER, and GLC responses over the time course of the GC and NG conditions. Violation of sphericity was corrected using the Box correction, and corrected degrees of freedom are reported. Paired t tests were used for planned comparisons between test conditions, such as fasting EE and EE during discrete periods of chewing and no chewing. Average postprandial values for EE, RER, and GLC were calculated from the area under the curve (determined by the trapezoid method) divided by 180 (the duration of the postprandial measurement period). Significance tests were 2-tailed and were accepted at p < 0.05. Effect sizes are reported as partial eta squared (2) for ANOVA and as Cohen’s d for paired t tests (Cohen 1988). Since the broader study measured appetite, statistical power was based on anticipated appetite suppression after gum chewing (Hetherington and Boyland 2007). Statistical analyses were conducted using PASW Statistics 18.0 (SPSS Inc., Chicago, Ill., USA).
Results Subject characteristics are shown in Table 1. Fasting and postprandial EE, RER, and GLC values for the 2 test conditions are compared in Table 2. There was no significant effect of treatment order on EE, RER, or GLC in either the fasting or the postprandial state (data not shown). Fasting EE was significantly higher during GC than during NG, as shown in Fig. 1. Since energy intake at breakfast was based on 45-min resting EE, subjects consumed slightly, though significantly, more energy during the GC condition (523.0 ± 18.9 kcal) than during the NG condition (498.8 ± 19.0 kcal; t = 2.458, p = 0.021). However, this difference did not correlate with any of the outcome variables and did not change the overall results (data not shown). Over the course of the mornings, repeated-measures ANOVA showed a marginally significant main effect of gum chewing (F[1,29] = 4.02, p = 0.054, 2 = 0.122) and a significant effect of time (F[8,232] = 72.50, p < 0.001, 2 = 0.714) on EE, and there was a marginally significant time by gum chewing interaction effect (F[8,232] = 1.96, p = 0.052, 2 = 0.063) on EE. As shown in Fig. 1, EE was significantly higher in the fasting state (p = 0.016), at 90 min (p = 0.006), and at 180 min (p = 0.015) during GC compared with NG. These times correspond to the times that subjects chewed. Average postprandial EE was higher in the GC condition than in the NG condition (see Table 2). The thermic effect of food, as determined by the area under the curve, did not differ significantly between GC (41.1 ± 2.5 kcal) and NG (45.0 ± 2.8 kcal; t = 1.819, p = 0.079) owing to the significantly higher fasting EE during GC. Within each of the gum chewing sessions, there were discrete time periods of chewing and no chewing, which are examined in greater detail in Figs. 2 and 3. During fasting EE measurements in the GC condition, subjects chewed gum at 10–30 min of the 45-min measurement period. During 0–10 min and 30–45 min, they did not chew. As shown in Fig. 2, during 0–10 min (prior to the first chewing), there was no significant difference in EE be-
403
Table 1. Demographic, anthropometric, and physiologic values for study participants (n = 30; 17 males, 13 females). Characteristic
Mean±SD
Age (y) BMI (kg/m2) Percent body fat (%) Estimated V˙O2max (mL O2/(min·kg)) Total METs (MET-min/day)
21.4±6.6 23.9±2.8 20±10 47±8 4330±3910
Note: Percent body fat was determined by air displacement plethysmography (BODPOD). Maximal oxygen uptake (V˙O2max) was estimated by the 3-min Queen's College Step Test. Total METs (metabolic equivalents) was calculated using the 7-day International Physical Activity Questionnaire. BMI, body mass index.
Table 2. Resting energy expenditure (EE), resting respiratory exchange ratio (RER), and capillary glucose (GLC). Gum chewing No gum Paired (mean±SE) (mean±SE) t score p GLC (mmol/L) Precalorimetery, fasting Postcalorimetry, fasting Postprandial average Resting EE (kcal/min) Fasting (45 min) Postprandial average Resting RER (V˙CO2/V˙O2) Fasting (45 min) Postprandial average
5.1±0.1 5.2±0.4 6.3±0.6
5.2±0.1 5.2±0.4 6.3±0.6
0.549 0.175 0.158
0.587 0.862 0.875
1.23±0.04 1.46±0.05
1.17±0.04 1.42±0.05
2.56 2.181
0.016* 0.037*
0.81±0.04 0.88±0.03
0.80±0.3 0.88±0.03
0.146 0.160
0.885 0.874
Note: GLC was measured prior to and immediately after fasting REE measurements and 8 times during the postprandial state. Postprandial averages for EE and GLC were determined by calculating the area under the curve for each variable (by the trapezoid method) and dividing by 180 (duration of the postprandial measurement period). Conversions: 1 mg/dL GLC = 0.055 mmol/L GLC; 1 kcal = 4.2 kJ. V˙CO2/V˙O2, carbon dioxide output/oxygen uptake. *p < 0.05 between gum chewing and no gum conditions.
Fig. 1. Energy expenditure (EE) under gum chewing (GC) and no gum (NG) conditions. *, p < 0.05; **, p < 0.01 between conditions. See Results section for p values.
tween GC (1.22 ± 0.04 kcal/min) and NG (1.18 ± 0.04 kcal/min; t = 1.371, p = 0.181). In the NG condition, subjects showed a significant decrease in EE from the 0–10 min period to the 10–30 min period (1.15 ± 0.04 kcal/min; t = 3.175, p = 0.004) that was maintained during the 30–45 min period (1.15 ± 0.04 kcal). Conversely, in the GC condition, there was a nonsignificant increase in EE between the 0–10 min (1.22 ± 0.04 kcal/min) and 10–30 min periods (1.25 ± 0.04 kcal/min; t = 1.697, p = 0.100). Thus, during the chewing period (10–30 min), EE was significantly higher during the GC condition than during the NG condition (t = 3.652, p = 0.001); this finding was Published by NRC Research Press
404
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 04/12/15 For personal use only.
Fig. 2. Fasting energy expenditure before chewing (0–10 min), during chewing (10–30 min), and after chewing (30–45 min) under gum chewing and no gum conditions (1 kcal = 4.2 kJ). Different letters indicate values that are significantly different from one another. See Results section for p values.
Fig. 3. Postprandial energy expenditure during chewing (60–80 and 150–170 min) and after chewing (80–90 and 170–180 min) under gum chewing and no gum conditions (1 kcal = 4.2 kJ). Different letters indicate values that are significantly different from one another. Capital letters represent the first postprandial chewing period and lowercase letters represent the second postprandial chewing period. See Results section for p values.
associated with a small to medium effect size of 0.46, according to Cohen’s standards (Cohen 1988). EE decreased significantly after chewing ceased (30–45 min) in the GC condition (1.17 ± 0.04 kcal/min; t = 6.025, p < 0.001). EE was not significantly different between GC and NG during the 30–45 min period (t = 0.599, p = 0.554). As shown in Fig. 3, during 60–80 min (i.e., during chewing), EE was significantly higher in the GC condition (1.55 ± 0.06 kcal/min) than in the NG condition (1.47 ± 0.04 kcal/min; t = 3.41, p = 0.002); this finding was associated with a small to medium effect size of 0.37, according to Cohen’s standards (Cohen 1988). During 80– 90 min (i.e., after chewing ceased), EE decreased significantly in the GC condition (1.49 ± 0.06 kcal/min; t = 4.049, p < 0.001). In the NG condition, EE also decreased significantly from the 60–80 min to the 80–90 min period (1.43 ± 0.05 kcal/min; t = 3.811, p = 0.001). However, EE at 80–90 min remained significantly higher in GC than in NG (t = 2.32, p = 0.028). During 150–170 min (i.e., during chewing), EE was significantly higher in the GC condition (1.46 ± 0.05 kcal/min) than in the NG condition (1.39 ± 0.05 kcal/min; t = 3.326, p = 0.002); this finding was associated with a small effect size of 0.26, according to Cohen’s standards (Cohen 1988). During 170–180 min (i.e., after chewing), EE decreased significantly in the GC condition (1.32 ± 0.05 kcal/min; t = 6.185, p < 0.001). In the NG condition, EE also decreased significantly from the 150–170 min to the 170–180 min
Appl. Physiol. Nutr. Metab. Vol. 40, 2015
Fig. 4. Respiratory exchange ratio under gum chewing and no gum conditions. V˙CO2/V˙O2, carbon dioxide output/oxygen uptake. *, p < 0.05 between conditions. See Results section for p values.
Fig. 5. Capillary glucose (GLC) under gum chewing and no gum conditions (1 mg/dL GLC = 0.055 mmol/L GLC).
period (1.34 ± 0.05 kcal/min; t = 5.250, p < 0.001). At 170–180 min, there was no significant difference in EE between GC and NG (t = 0.143, p = 0.887). As shown in Table 2, fasting RER did not differ significantly between GC and NG. The RER response is presented in Fig. 4. There was a significant effect of time on RER (F[8,232] = 34.673, p < 0.001, 2 = 0.545), but there was no significant effect of chewing gum and no significant time by chewing interaction effect. Average postprandial RER did not differ significantly between GC and NG, as shown in Table 2. Also shown in Table 2, before and after the measurement of resting EE, there was no significant difference in GLC between the GC and NG conditions. GLC did not change significantly across the resting EE measurement period in either condition (F[1,29] = 0.399, p = 0.533, 2 = 0.014). The postprandial GLC response is presented in Fig. 5. There was a significant effect of time on GLC (F[3.259,110.81] = 61.843, p < 0.001, 2 = 0.645), but there was no effect of gum chewing and no time by chewing interaction effect on GLC. Average postprandial GLC did not differ between GC and NG, as shown in Table 2.
Discussion In these 30 healthy males and females, gum chewing was associated with small yet significantly higher EE than no chewing in both the fasting and postprandial states. This effect of approximately 5%–8% was not associated with a difference in substrate oxidation between conditions, as reflected by RER, nor did blood GLC differ between conditions. In the fasting state, there was no difference in EE between GC and NG prior to chewing (see Fig. 2). In the NG (control) condition, EE decreased significantly from the 0–10 min period to the 10– Published by NRC Research Press
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 04/12/15 For personal use only.
Kresge and Melanson
30 min period and then stayed at the same level throughout the 30–45 min period. In the GC condition, EE did not decline, tending instead to increase from the 0–10 min period to the 10–30 min period. Thus, gum chewing ameliorated the decline in fasting EE observed in the NG (control) condition. Upon cessation of chewing, EE decreased to the same level as during NG, indicating that the impact of chewing on EE is likely to be limited to the duration of chewing, with little carryover effect in the fasting state. In the postprandial state, similar patterns occurred at both the second and third chewing periods. During 60–80 min (i.e., during chewing), EE in the GC condition was significantly higher than that in the NG condition (see Fig. 3). Upon cessation of chewing, EE decreased significantly, yet remained significantly higher than that during NG. During 150–170 min (i.e., during chewing), EE in the GC condition was again significantly higher than that in the NG condition (see Fig. 3). Upon cessation of chewing, EE decreased significantly, this time to a level not significantly different than that during NG, similar to the pattern seen during the fasting state. The elevated EE at 80–90 min (after chewing) might suggest a carryover effect of chewing. However, since this effect was not present in the fasting or late postprandial periods, it is likely that postprandial thermogenesis or other factors may have impacted EE at 80–90 min, so a substantial carryover effect after chewing is unlikely. Fasting and postprandial RER did not differ significantly between GC and NG (see Fig. 4), indicating that neither the nominal energy content of the gum nor the act of chewing influenced substrate oxidation. This is not surprising given that subjects chewed sugar-free gum containing sugar alcohols and that there were no differences in blood glucose between GC and NG (see Table 2 and Fig. 5). Animal data have suggested that chewing may activate histamine pathways (Sakata et al. 2003) that could stimulate lipolysis (Masaki et al. 2001). However, human data are lacking, and it is probable that the amount of chewing in this study was insufficient to induce measureable changes. Gum chewing is associated with significant intrinsic tongue muscle activity, along with activity of the masticatory muscles (Rikimaru et al. 2001). This muscular movement is likely responsible for the majority of increased energy expenditure observed. Very few published papers have examined EE during gum chewing. One brief letter reported on EE measured in 7 nonobese subjects before chewing gum (58 ± 11 kcal/h), while chewing gum for 12 min (70 ± 14 kcal/h), and for 12 min after chewing gum (59 ± 12 kcal/h) (Levine et al. 1999). The authors stated that an average increase of 11 ± 3 kcal/h (19% ± 4%) associated with chewing was greater than increases in EE associated with standing (11% ± 11%) measured in the same subjects. The increase they reported (approximately 0.18 ± 0.05 kcal/min) is substantially larger than the increases we are reporting, in both fasting (0.094 kcal/min, 8.1% ± 3.6%) and postprandial states (0.078 kcal/min, 5.3% ± 3.5% at 60– 80 min; 0.071 kcal/min, 5.2% ± 3.9% at 150–170 min) (see Figs. 2 and 3). Some of that difference might be explained by subjects chewing 8.4 g of gum compared with 1.9 g in our study. In addition, chewing rate might also have affected EE differences, as discussed below. Only limited comparisons are possible with our study, however, because the experimental conditions in the report of Levine et al. (1999) were not revealed (e.g., fasting vs. fed; anthropometric data), and there does not appear to have been a control condition. In the brief letter described above (Levine et al. 1999), subjects chewed at a frequency of 100 bites/min. A subsequent paper examined the impact of various doses of caffeine and nicotine in gums on EE in a fasting study population similar to ours (Jessen et al. 2003). During the placebo gum condition, subjects in that study had an increase in EE of approximately 4 kcal/h while chewing at a cadence of 60 bites/min for 25 min. In our study, during the fasting state, the average difference in EE during chewing was 5.6 kcal/h while subjects chewed at a self-selected rate. We used
405
the self-selected chewing rate to assess the effects of chewing under conditions that mimic the real world. In addition, there is evidence that subjects may not comply with an artificial cadence while chewing (Nemeth-Coslett et al. 1988). In a dose–response study of nicotine gum, subjects were asked to match their chewing to cadence tones of one bite per second, per 2 s, per 4 s, and per 8 s. Despite the instructions to follow the cadence, subjects’ measured chewing rates were one bite per 1.9 s, 2.1 s, 3.7 s, and 4.2 s under each condition, respectively (Nemeth-Coslett et al. 1988). Thus, it appears possible that the use of cadence might introduce variance into chewing studies, rather than controlling it, owing to noncompliance. It remains plausible that chewing rate could have influenced EE. We did not measure subjects’ chewing rates, so this is a potential area for future research. Varying the duration of chewing might also impact the total energy expended in gum chewing. The amount of energy expended in each 20-min bout of chewing a sugar-free gum was approximately equal to the energy content of the sugar-free gum (⬃2–3 kcal). This would suggest an advantage of chewing sugar-free gum over gums containing sugar and more energy, since it is possible that the energy contained in sugar-sweetened gums (⬃20–25 kcal) might not be expended in a normal chewing session of 20 min or less. If people use sugar-free gum as a substitute for snacks containing more energy, then they may not have to account for the energy in the sugar-free gum if they chew it for a long enough time. Thus, this study offers empirical evidence that can be used for patients or clients who tend to chew a lot of gum and (or) seek to replace snacks with gum. While it may be tempting to calculate the potential impact of gum chewing on overall energy balance, such extrapolations should be made conservatively. At all 3 chewing time periods in the current study, EE declined rapidly and significantly upon the cessation of chewing. To impact whole-day energy expenditure, individuals would have to chew gum throughout the day. Strengths of this study include testing 3 sessions of gum chewing in fasting and fed states in both sexes under carefully standardized conditions. However, our subject population was limited to young normal-weight and overweight individuals in a controlled laboratory setting over the course of one morning. The impact on other populations in free-living conditions has not been explored. Without a carryover effect, the contribution of gum chewing to daily EE will likely be related to the rate and duration of chewing and is likely to be relatively small, as suggested by the effect sizes. The impact of gum chewing on energy balance, which is crucial to understanding its potential role in weight management, needs to be considered in conjunction with data on energy intake. If gum chewing is associated with reduced energy intake in addition to increased EE, as demonstrated here, then its value as part of a comprehensive weight control program would be enhanced. An 8-week randomized control study in 201 adults undergoing a diet or a diet-plus-gum chewing (90 min/day) intervention showed no between-group differences in body weight (Shikany et al. 2012). However, the dietary information provided to both groups was aimed at healthy eating rather than weight loss. Within the dietplus-gum chewing group only, decreases in waist circumference and blood pressure were seen. Given these considerations, this area of research may warrant more investigation. In summary, chewing gum at one’s natural pace increases energy expenditure in the fasting and fed states without impacting blood glucose or substrate oxidation. Longer-term studies should be conducted to fully discern whether gum chewing would improve energy balance in weight management programs.
Disclosure Wrigley Science Institute donated the chewing gum used in this study. The authors were responsible for all aspects of the research design, data collection, and analysis and manuscript preparation. Published by NRC Research Press
406
Acknowledgements This study was funded by a research award from The Obesity Society (TOS) presented in October 2007. The award was made possible by a donation to TOS from Wrigley Science Institute.
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 04/12/15 For personal use only.
References Brennan, I.M., Feltrin, K.L., Nair, N.S., Hausken, T., Little, T.J., Gentilcore, D., et al. 2009. Effects of the phases of the menstrual cycle on gastric emptying, glycemia, plasma GLP-1 and insulin, and energy intake in healthy lean women. Am. J. Physiol. Gastrointest. Liver Physiol. 297(3): G602–G610. doi:10. 1152/ajpgi.00051.2009. PMID:19556358. Bryant, M., Truesdale, K.P., and Dye, L. 2006. Modest changes in dietary intake across the menstrual cycle: implications for food intake research. Br. J. Nutr. 96(5): 888–894. doi:10.1017/BJN20061931. PMID:17092378. Cohen, J. 1988. Statistical power analysis for the behavioral sciences. 2nd ed. L. Erlbaum Associates, Hillsdale, N.J. Craig, C.L., Marshall, A.L., Sjostrom, M., Bauman, A.E., Booth, M.L., Ainsworth, B.E., et al. 2003. International physical activity questionnaire: 12-country reliability and validity. Med. Sci. Sports Exerc. 35(8): 1381–1395. doi:10.1249/01.MSS. 0000078924.61453.FB. PMID:12900694. Dye, L., and Blundell, J.E. 1997. Menstrual cycle and appetite control: implications for weight regulation. Hum. Reprod. 12(6), 1142–1151. PMID:9221991. Esliger, D.W., and Tremblay, M.S. 2006. Technical reliability assessment of three accelerometer models in a mechanical setup. Med. Sci. Sports Exerc. 38(12): 2173–2181. doi:10.1249/01.mss.0000239394.55461.08. PMID:17146326. Fields, D.A., Goran, M.I., and McCrory, M.A. 2002. Body-composition assessment via air-displacement plethysmography in adults and children: a review. Am. J. Clin. Nutr. 75(3): 453–467. PMID:11864850. Hetherington, M.M., and Boyland, E. 2007. Short-term effects of chewing gum on snack intake and appetite. Appetite, 48(3): 397–401. doi:10.1016/j.appet.2006. 10.001. PMID:17118491. Jessen, A.B., Toubro, S., and Astrup, A. 2003. Effect of chewing gum containing nicotine and caffeine on energy expenditure and substrate utilization in men. Am. J. Clin. Nutr. 77(6): 1442–1447. PMID:12791621. Julis, R.A., and Mattes, R.D. 2007. Influence of sweetened chewing gum on appetite, meal patterning and energy intake. Appetite, 48(2): 167–175. doi:10.1016/ j.appet.2006.08.003. PMID:17050036. Leveille, G., McMahon, K., Alcantara, E., and Zibell, S. 2008. Benefits of chewing
Appl. Physiol. Nutr. Metab. Vol. 40, 2015
gum: oral health and beyond. Nutr. Today, 43(2), 75–81. doi:10.1097/01.NT. 0000303316.67157.c5. Levine, J., Baukol, P., and Pavlidis, I. 1999. The energy expended in chewing gum. N. Engl. J. Med. 341(27): 2100. doi:10.1056/NEJM199912303412718. PMID: 10627208. Masaki, T., Yoshimatsu, H., Chiba, S., Watanabe, T., and Sakata, T. 2001. Central infusion of histamine reduces fat accumulation and upregulates UCP family in leptin-resistant obese mice. Diabetes, 50(2), 376–384. doi:10.2337/diabetes. 50.2.376. PMID:11272150. McArdle, W.D., Katch, F.I., Pechar, G.S., Jacobson, L., and Ruck, S. 1972. Reliability and interrelationships between maximal oxygen intake, physical work capacity and step-test scores in college women. Med. Sci. Sports, 4(4): 182–186. PMID:4648576. McCrory, M.A., Mole, P.A., Gomez, T.D., Dewey, K.G., and Bernauer, E.M. 1998. Body composition by air-displacement plethysmography by using predicted and measured thoracic gas volumes. J. Appl. Physiol. 84(4): 1475–1479. PMID: 9516218. Nemeth-Coslett, R., Benowitz, N.L., Robinson, N., and Henningfield, J.E. 1988. Nicotine gum: chew rate, subjective effects and plasma nicotine. Pharmacol., Biochem. Behav. 29(4): 747–751. doi:10.1016/0091-3057(88)90197-9. PMID: 3413200. Rikimaru, H., Kikuchi, M., Itoh, M., Tashiro, M., and Watanabe, M. 2001. Mapping energy metabolism in jaw and tongue muscles during chewing. J. Dent. Res. 80(9): 1849–1853. doi:10.1177/00220345010800091501. PMID:11926246. Sakata, T., Yoshimatsu, H., Masaki, T., and Tsuda, K. 2003. Anti-obesity actions of mastication driven by histamine neurons in rats. Exp. Biol. Med. (Maywood), 228(10): 1106–1110. PMID:14610247. Shikany, J.M., Thomas, A.S., McCubrey, R.O., Beasley, T.M., and Allison, D.B. 2012. Randomized controlled trial of chewing gum for weight loss. Obesity (Silver Spring), 20(3): 547–552. PMID:22076595. Siri, W. 1961. Body composition from fluid spaces and density: analysis of methods. In Techniques for measuring body composition. Edited by J. Broézek and A. Henschel. National Academy of Sciences-National Research Council, Washington, D.C. pp. 113–224. PMID:8286893.. Trost, S.G., McIver, K.L., and Pate, R.R. 2005. Conducting accelerometer-based activity assessments in field-based research. Med. Sci. Sports Exerc. 37(11): S531–S543. doi:10.1249/01.mss.0000185657.86065.98. PMID:16294116.
Published by NRC Research Press
ARTICLE Chewing gum increases energy expenditure before and after controlled breakfasts Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 04/12/15 For personal use only.
Daniel L. Kresge and Kathleen Melanson
Abstract: Chewing has been associated with improved satiation and satiety, but little is known about the metabolic impact of gum chewing. We tested the hypothesis that gum chewing would increase energy expenditure (EE) and reduce respiratory exchange ratio (RER) before and after a controlled test meal. Seventeen males and 13 females (age 21.5 ± 6.6 years, body mass index 23.9 ± 2.8 kg/m2) participated in a randomized crossover study in which subjects chewed sugar-free gum for a total of 1 h (3 sessions of 20 min) on the test day (GC) and did not chew gum on a control day (NG). EE and RER were measured by indirect calorimetry after an overnight fast. Subjects consumed a breakfast shake containing 30% of their measured energy needs, and then postprandial EE and RER were measured for 3 h. Blood glucose (GLC) was measured in the fasting and postprandial states at regular intervals. Fasting EE was higher during GC (1.23 ± 0.04 kcal/min; 1 kcal = 4.2 kJ) than during NG (1.17 ± 0.04 kcal/min; p = 0.016). Postprandial EE was also higher during GC (1.46 ± 0.05 kcal/min) than during NG (1.42 ± 0.05 kcal/min; p = 0.037). Fasting and postprandial RER and GLC did not differ between GC and NG. The findings demonstrate that GC is associated with higher fasting and postprandial EE without altering blood glucose or substrate oxidation as measured by RER. These data suggest that gum chewing potentially could influence short-term energy balance in this population; however, longer-term research is needed. Key words: mastication, resting metabolic rate, thermic effect of food, respiratory exchange ratio, indirect calorimetry, energy balance, substrate oxidation, postprandial. Résumé : Mâcher de la gomme est associé, selon des études, a` une meilleure sensation de rassasiement et de satiété, mais il y a peu d’études sur l’impact métabolique de la gomme a` mâcher. Nous vérifions l’hypothèse selon laquelle mâcher de la gomme suscite une augmentation de la dépense énergétique (« EE ») et une diminution du ratio d’échanges gazeux (« RER ») avant et après un repas test contrôlé. Treize femmes et dix-sept hommes âgés de 21,5 ± 6,6 ans et présentant un indice de masse corporelle de 23,9 ± 2,8 kg/m2 participent a` une étude organisée selon un devis croisé aléatoire dans laquelle les sujets mâchent de la gomme sans sucre au jour test (« GC ») durant 60 min au total (3 séances de 20 min) et ne mâchent pas de la gomme au jour de contrôle (« NG »). On mesure EE et RER par calorimétrie indirecte après un jeûne d’une nuit. Les sujets consomment une boisson pour déjeuner contenant 30% de leurs besoins énergétiques mesurés et on mesure durant 3 h les EE et RER postprandiaux. À intervalles réguliers, on mesure la concentration de glucose sanguin (« GLC ») dans les conditions de jeûne et postprandiale. EE a` jeun est plus élevée durant GC (1,23 ± 0,04 kcal/min) comparativement a` NG (1,17 ± 0,04 kcal/min; p = 0,016). EE postprandiale est aussi plus élevée durant GC (1,46 ± 0,05 kcal/min) comparativement a` NG (1,42 ± 0,05 kcal/min; p = 0,037). La valeur a` jeun et postprandiale de RER et de GLC ne varie pas entre GC et NG. D’après ces observations, GC est associé a` une EE plus élevée a` jeun et après un repas, et ce, sans modification de la concentration sanguine de glucose et de l’oxydation des substrats évaluée par le RER. D’après ces observations, mâcher de la gomme pourrait avoir un impact sur l’équilibre énergétique a` court terme dans cette population; néanmoins, il faut réaliser des études a` long terme. [Traduit par le Rédaction] Mots-clés : mastication, métabolisme de repos, thermogenèse alimentaire, ratio d’échanges gazeux, calorimétrie indirecte, équilibre énergétique, oxydation des substrats, postprandial.
Introduction Gum chewing has been promoted as beneficial in terms of oral health, cognitive performance, alertness, stress relief, and weight management (Leveille et al. 2008). Proposed weight management benefits of gum chewing are based mainly on energy intake research. For example, cravings for and intake of sweet snacks may be decreased by gum chewing (Hetherington and Boyland 2007), although results are not entirely consistent (Julis and Mattes 2007). If gum chewing is to be promoted for the purposes of weight management (Leveille et al. 2008), then its effects on energy balance, substrate oxidation, and other potential metabolic outcomes should be explored. To our knowledge, only 2 studies have examined gum chewing’s effect on energy expenditure. In one of these studies, 7 subjects
chewed the equivalent of 4 sticks of gum for 12 min at a very rapid cadence, and energy expenditure was measured immediately before, during, and after (Levine et al. 1999). Although an increase in energy expenditure of 19% was reported, the paradigm was not typical of free-living gum chewing, and the protocol was limited to 7 short tests. In a study designed to determine energy expenditure while chewing gums with various doses of caffeine and (or) nicotine, energy expenditure of chewing the control gum at a natural pace was estimated as 5% higher than that of no gum chewing (Jessen et al. 2003). This gum chewing was measured in the fasting state. The effects on substrate oxidation within the control condition were not clear, and postprandial measures were not taken.
Received 23 June 2014. Accepted 7 November 2014. D.L. Kresge and K. Melanson. Department of Nutrition and Food Sciences, University of Rhode Island, Kingston, RI 02881, USA. Corresponding author: Kathleen J. Melanson (e-mail: [email protected]). Appl. Physiol. Nutr. Metab. 40: 401–406 (2015) dx.doi.org/10.1139/apnm-2014-0232
Published at www.nrcresearchpress.com/apnm on 18 December 2014.
402
More work is needed to understand the potential role gum chewing may have in avoiding a positive energy balance on a regular basis. Gum chewing under natural situations should be simulated, and multiple chewing bouts should be examined. Thus, the purpose of this study was to compare energy expenditure, substrate oxidation, and blood glucose under conditions of gum chewing at a selfdirected pace versus no gum chewing in both the fasted and postprandial states during 2 mornings within the same individuals.
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 04/12/15 For personal use only.
Materials and methods Study design This study used a randomized crossover design, in which volunteers served as their own controls, with 2 conditions and 9 time points. Each subject chewed gum (GC) on one test day and did not chew gum (NG) on the other test day. We report here on energy expenditure (EE), respiratory exchange ratio (RER), and capillary glucose (GLC), which were measured as part of a broader study examining the relationship between gum chewing and energy balance. Appetite results are reported elsewhere. Volunteers provided written informed consent prior to participating, and they received a $60.00 stipend for completing the study. The Institutional Review Board of the University of Rhode Island approved this study. Subjects Thirty nonsmoking men and women were recruited by advertisements and class announcements. A preliminary screening by telephone was used to verify inclusion criteria of 18–48 years and body mass index (BMI) of 18.5–29.9 kg/m2 and to apply exclusion criteria, which included factors that could influence metabolism or appetite: pregnancy or lactation, diabetes mellitus or other diseases affecting energy balance, currently dieting or exercising to lose weight, or weight change of greater than 3 pounds during the past 6 months. Preliminary data collection The protocol required 3 visits. During the first visit, volunteers completed general health questionnaires and the International Physical Activity Questionnaire (Craig et al. 2003), which were used for descriptive purposes only. Waist circumference was measured using a physician’s tape measure at the level of the umbilicus after normal exhalation. Body composition was determined at least 4 h after subjects had eaten, using a validated air displacement plethysmography system (BODPOD, Life Measurements Inc., Concord, Calif., USA) (Fields et al. 2002; McCrory et al. 1998). Body weight was measured to the nearest 0.1 kg on a calibrated digital scale that was part of the BODPOD system. Height was measured to the nearest millimeter using a wall-mounted stadiometer (Seca, Hamburg, Germany). BMI was calculated, and body composition was determined by the BODPOD software using the Siri equation (Siri 1961). Subjects completed the 3-min submaximal Queen’s College Step Test to estimate maximal aerobic capacity (McArdle et al. 1972). Next, each volunteer was familiarized with test-day procedures by undergoing a 10-min “practice” metabolic measurement. Finally, subjects chose 1 of 6 flavors of commercially available sugarless gum (Orbit, Wm. Wrigley Co., Chicago, Ill., USA) to chew during the GC condition. Preparation days The next 2 visits were test visits. Each test day was preceded by a preparation day, during which volunteers maintained normal diet and activities of daily living, abstained from alcohol and strenuous exercise for at least 24 h prior to testing, and restricted caffeine for at least 18 h prior to testing. Subjects began fasting 10 h prior to their scheduled test visit, except for water. Since most subjects arrived at 0700, their fast began by 2100 the previous night. Subjects wore a validated physical activity monitor (Actical, Mini Mitter, Bend, Ore., USA) on both preparation days to confirm
Appl. Physiol. Nutr. Metab. Vol. 40, 2015
compliance with activity instructions (Esliger and Tremblay 2006; Trost et al. 2005). Subjects followed these restrictions the day before each test visit, so that the 2 preparation days were matched for diet and physical activity. Test days were separated by at least 48 h. Preprandial measurements Only one subject was measured per test day. For women, testing occurred during the mid-follicular phase of the menstrual cycle (days 5–13) to control for potential metabolic shifts across the cycle (Brennan et al. 2009; Bryant et al. 2006; Dye and Blundell 1997). Volunteers reported to our laboratory after a 10-h overnight fast, engaging in the least amount of activity possible to arrive in the laboratory. After voiding, subjects relaxed in an elevated supine position on the laboratory bed. The first capillary blood sample was drawn by lancet from the fingertip for analysis of GLC by autoanalyzer (Hemocue Glucose 201, Hemocue, Mission Viejo, Calif., USA). During the next 30–45 min, volunteers rested while 24-h diet and physical activity recalls were administered to confirm protocol compliance. Diet recalls were subsequently analyzed by Food Processor SQL 9.0 (ESHA Research, Salem, Ore., USA) and physical activity recordings were analyzed by Actical software, version 2.04. At the end of the rest period, subjects underwent a 45-min measurement of resting EE and RER (carbon dioxide output/oxygen uptake) by ventilated hood indirect calorimetry (Vmax, Sensormedics, Yorba, Calif., USA). Flow rate and gas calibrations were performed daily using 2 calibration tanks containing different CO2 and O2 concentrations, according to the manufacturer’s recommendations. During testing, volunteers remained in an elevated supine position and were permitted to read, study, or listen to music. They were not permitted to sleep, write, or hold any materials in their hands. The laboratory was kept quiet, and volunteers were provided with blankets for their comfort. Subjects were monitored closely to assure compliance with the protocol. Ten minutes into the resting EE measurement during the GC condition, data collection was temporarily halted using the pause feature of the calorimeter’s software. The transparent ventilated hood was lifted briefly and a single piece of sugarless gum (1.9 g, 3 kcal; 1 kcal = 4.2 kJ) was placed in the volunteer’s mouth. The hood was immediately replaced, and calorimetry resumed. Subjects chewed gum at their own natural pace for 20 min. They were not permitted to blow bubbles. At the end of the 20-min chewing period, calorimetry was again paused briefly, the gum was collected, and calorimetry resumed. During the NG condition, passing and collection of gum was simulated using the same procedures, except that no gum was given. Data from 1 min after passing, or simulation, were excluded from EE and RER calculations. At the conclusion of the resting EE measurement, the ventilated hood was removed, 45-min resting EE and RER were calculated, and another capillary blood sample was drawn for determination of GLC. Test meals and postprandial procedures Volunteers were given a nutritional shake for breakfast (Boost, Novartis) standardized to 30% of their measured 45-min resting fasting EE during each test day (518 ± 12 mL; 518 ± 12 kcal; 68.3% carbohydrate, 16.7% protein, 15% fat). Subjects consumed the shake within 15 min. Postprandial GLC was measured at 15, 30, 45, 60, 90, 120, 150, and 180 min. After the first postprandial blood draw had been completed (at 15 min), postprandial measurements of EE and RER began, using the same procedures previously described. The hood of the indirect calorimeter remained in place throughout the 3-h postprandial measurement period, except for brief intervals during the scheduled blood draws. During GC, volunteers chewed gum at 2 additional times for 20 min each time (60–80 min and 150–170 min postprandial), for a total of 60 min of chewing time over the morning of testing. In the postprandial state, gum was passed to volunteers after blood sampling, immePublished by NRC Research Press
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 04/12/15 For personal use only.
Kresge and Melanson
diately prior to application of the ventilated hood. The gum was collected (or collection was simulated) after 20 min of chewing using the same procedures as previously described, and EE measurement continued until the next scheduled blood sampling. At the end of the 3-h postprandial measurement period, a final blood sample was drawn, and then volunteers were offered an ad libitum pasta lunch with water. At the conclusion of lunch on the first metabolic test day, volunteers were given copies of their diet and activity recalls and the physical activity monitor, along with instructions to repeat their activity and dietary intake (including portion sizes) as closely as possible during the next preparation day. On their next test day, volunteers repeated the protocol under the other randomized test condition. Data analysis Data are presented as means ± standard error of the mean (SE), unless otherwise noted. Repeated-measures analysis of variance (ANOVA) was used to compare EE, RER, and GLC responses over the time course of the GC and NG conditions. Violation of sphericity was corrected using the Box correction, and corrected degrees of freedom are reported. Paired t tests were used for planned comparisons between test conditions, such as fasting EE and EE during discrete periods of chewing and no chewing. Average postprandial values for EE, RER, and GLC were calculated from the area under the curve (determined by the trapezoid method) divided by 180 (the duration of the postprandial measurement period). Significance tests were 2-tailed and were accepted at p < 0.05. Effect sizes are reported as partial eta squared (2) for ANOVA and as Cohen’s d for paired t tests (Cohen 1988). Since the broader study measured appetite, statistical power was based on anticipated appetite suppression after gum chewing (Hetherington and Boyland 2007). Statistical analyses were conducted using PASW Statistics 18.0 (SPSS Inc., Chicago, Ill., USA).
Results Subject characteristics are shown in Table 1. Fasting and postprandial EE, RER, and GLC values for the 2 test conditions are compared in Table 2. There was no significant effect of treatment order on EE, RER, or GLC in either the fasting or the postprandial state (data not shown). Fasting EE was significantly higher during GC than during NG, as shown in Fig. 1. Since energy intake at breakfast was based on 45-min resting EE, subjects consumed slightly, though significantly, more energy during the GC condition (523.0 ± 18.9 kcal) than during the NG condition (498.8 ± 19.0 kcal; t = 2.458, p = 0.021). However, this difference did not correlate with any of the outcome variables and did not change the overall results (data not shown). Over the course of the mornings, repeated-measures ANOVA showed a marginally significant main effect of gum chewing (F[1,29] = 4.02, p = 0.054, 2 = 0.122) and a significant effect of time (F[8,232] = 72.50, p < 0.001, 2 = 0.714) on EE, and there was a marginally significant time by gum chewing interaction effect (F[8,232] = 1.96, p = 0.052, 2 = 0.063) on EE. As shown in Fig. 1, EE was significantly higher in the fasting state (p = 0.016), at 90 min (p = 0.006), and at 180 min (p = 0.015) during GC compared with NG. These times correspond to the times that subjects chewed. Average postprandial EE was higher in the GC condition than in the NG condition (see Table 2). The thermic effect of food, as determined by the area under the curve, did not differ significantly between GC (41.1 ± 2.5 kcal) and NG (45.0 ± 2.8 kcal; t = 1.819, p = 0.079) owing to the significantly higher fasting EE during GC. Within each of the gum chewing sessions, there were discrete time periods of chewing and no chewing, which are examined in greater detail in Figs. 2 and 3. During fasting EE measurements in the GC condition, subjects chewed gum at 10–30 min of the 45-min measurement period. During 0–10 min and 30–45 min, they did not chew. As shown in Fig. 2, during 0–10 min (prior to the first chewing), there was no significant difference in EE be-
403
Table 1. Demographic, anthropometric, and physiologic values for study participants (n = 30; 17 males, 13 females). Characteristic
Mean±SD
Age (y) BMI (kg/m2) Percent body fat (%) Estimated V˙O2max (mL O2/(min·kg)) Total METs (MET-min/day)
21.4±6.6 23.9±2.8 20±10 47±8 4330±3910
Note: Percent body fat was determined by air displacement plethysmography (BODPOD). Maximal oxygen uptake (V˙O2max) was estimated by the 3-min Queen's College Step Test. Total METs (metabolic equivalents) was calculated using the 7-day International Physical Activity Questionnaire. BMI, body mass index.
Table 2. Resting energy expenditure (EE), resting respiratory exchange ratio (RER), and capillary glucose (GLC). Gum chewing No gum Paired (mean±SE) (mean±SE) t score p GLC (mmol/L) Precalorimetery, fasting Postcalorimetry, fasting Postprandial average Resting EE (kcal/min) Fasting (45 min) Postprandial average Resting RER (V˙CO2/V˙O2) Fasting (45 min) Postprandial average
5.1±0.1 5.2±0.4 6.3±0.6
5.2±0.1 5.2±0.4 6.3±0.6
0.549 0.175 0.158
0.587 0.862 0.875
1.23±0.04 1.46±0.05
1.17±0.04 1.42±0.05
2.56 2.181
0.016* 0.037*
0.81±0.04 0.88±0.03
0.80±0.3 0.88±0.03
0.146 0.160
0.885 0.874
Note: GLC was measured prior to and immediately after fasting REE measurements and 8 times during the postprandial state. Postprandial averages for EE and GLC were determined by calculating the area under the curve for each variable (by the trapezoid method) and dividing by 180 (duration of the postprandial measurement period). Conversions: 1 mg/dL GLC = 0.055 mmol/L GLC; 1 kcal = 4.2 kJ. V˙CO2/V˙O2, carbon dioxide output/oxygen uptake. *p < 0.05 between gum chewing and no gum conditions.
Fig. 1. Energy expenditure (EE) under gum chewing (GC) and no gum (NG) conditions. *, p < 0.05; **, p < 0.01 between conditions. See Results section for p values.
tween GC (1.22 ± 0.04 kcal/min) and NG (1.18 ± 0.04 kcal/min; t = 1.371, p = 0.181). In the NG condition, subjects showed a significant decrease in EE from the 0–10 min period to the 10–30 min period (1.15 ± 0.04 kcal/min; t = 3.175, p = 0.004) that was maintained during the 30–45 min period (1.15 ± 0.04 kcal). Conversely, in the GC condition, there was a nonsignificant increase in EE between the 0–10 min (1.22 ± 0.04 kcal/min) and 10–30 min periods (1.25 ± 0.04 kcal/min; t = 1.697, p = 0.100). Thus, during the chewing period (10–30 min), EE was significantly higher during the GC condition than during the NG condition (t = 3.652, p = 0.001); this finding was Published by NRC Research Press
404
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 04/12/15 For personal use only.
Fig. 2. Fasting energy expenditure before chewing (0–10 min), during chewing (10–30 min), and after chewing (30–45 min) under gum chewing and no gum conditions (1 kcal = 4.2 kJ). Different letters indicate values that are significantly different from one another. See Results section for p values.
Fig. 3. Postprandial energy expenditure during chewing (60–80 and 150–170 min) and after chewing (80–90 and 170–180 min) under gum chewing and no gum conditions (1 kcal = 4.2 kJ). Different letters indicate values that are significantly different from one another. Capital letters represent the first postprandial chewing period and lowercase letters represent the second postprandial chewing period. See Results section for p values.
associated with a small to medium effect size of 0.46, according to Cohen’s standards (Cohen 1988). EE decreased significantly after chewing ceased (30–45 min) in the GC condition (1.17 ± 0.04 kcal/min; t = 6.025, p < 0.001). EE was not significantly different between GC and NG during the 30–45 min period (t = 0.599, p = 0.554). As shown in Fig. 3, during 60–80 min (i.e., during chewing), EE was significantly higher in the GC condition (1.55 ± 0.06 kcal/min) than in the NG condition (1.47 ± 0.04 kcal/min; t = 3.41, p = 0.002); this finding was associated with a small to medium effect size of 0.37, according to Cohen’s standards (Cohen 1988). During 80– 90 min (i.e., after chewing ceased), EE decreased significantly in the GC condition (1.49 ± 0.06 kcal/min; t = 4.049, p < 0.001). In the NG condition, EE also decreased significantly from the 60–80 min to the 80–90 min period (1.43 ± 0.05 kcal/min; t = 3.811, p = 0.001). However, EE at 80–90 min remained significantly higher in GC than in NG (t = 2.32, p = 0.028). During 150–170 min (i.e., during chewing), EE was significantly higher in the GC condition (1.46 ± 0.05 kcal/min) than in the NG condition (1.39 ± 0.05 kcal/min; t = 3.326, p = 0.002); this finding was associated with a small effect size of 0.26, according to Cohen’s standards (Cohen 1988). During 170–180 min (i.e., after chewing), EE decreased significantly in the GC condition (1.32 ± 0.05 kcal/min; t = 6.185, p < 0.001). In the NG condition, EE also decreased significantly from the 150–170 min to the 170–180 min
Appl. Physiol. Nutr. Metab. Vol. 40, 2015
Fig. 4. Respiratory exchange ratio under gum chewing and no gum conditions. V˙CO2/V˙O2, carbon dioxide output/oxygen uptake. *, p < 0.05 between conditions. See Results section for p values.
Fig. 5. Capillary glucose (GLC) under gum chewing and no gum conditions (1 mg/dL GLC = 0.055 mmol/L GLC).
period (1.34 ± 0.05 kcal/min; t = 5.250, p < 0.001). At 170–180 min, there was no significant difference in EE between GC and NG (t = 0.143, p = 0.887). As shown in Table 2, fasting RER did not differ significantly between GC and NG. The RER response is presented in Fig. 4. There was a significant effect of time on RER (F[8,232] = 34.673, p < 0.001, 2 = 0.545), but there was no significant effect of chewing gum and no significant time by chewing interaction effect. Average postprandial RER did not differ significantly between GC and NG, as shown in Table 2. Also shown in Table 2, before and after the measurement of resting EE, there was no significant difference in GLC between the GC and NG conditions. GLC did not change significantly across the resting EE measurement period in either condition (F[1,29] = 0.399, p = 0.533, 2 = 0.014). The postprandial GLC response is presented in Fig. 5. There was a significant effect of time on GLC (F[3.259,110.81] = 61.843, p < 0.001, 2 = 0.645), but there was no effect of gum chewing and no time by chewing interaction effect on GLC. Average postprandial GLC did not differ between GC and NG, as shown in Table 2.
Discussion In these 30 healthy males and females, gum chewing was associated with small yet significantly higher EE than no chewing in both the fasting and postprandial states. This effect of approximately 5%–8% was not associated with a difference in substrate oxidation between conditions, as reflected by RER, nor did blood GLC differ between conditions. In the fasting state, there was no difference in EE between GC and NG prior to chewing (see Fig. 2). In the NG (control) condition, EE decreased significantly from the 0–10 min period to the 10– Published by NRC Research Press
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 04/12/15 For personal use only.
Kresge and Melanson
30 min period and then stayed at the same level throughout the 30–45 min period. In the GC condition, EE did not decline, tending instead to increase from the 0–10 min period to the 10–30 min period. Thus, gum chewing ameliorated the decline in fasting EE observed in the NG (control) condition. Upon cessation of chewing, EE decreased to the same level as during NG, indicating that the impact of chewing on EE is likely to be limited to the duration of chewing, with little carryover effect in the fasting state. In the postprandial state, similar patterns occurred at both the second and third chewing periods. During 60–80 min (i.e., during chewing), EE in the GC condition was significantly higher than that in the NG condition (see Fig. 3). Upon cessation of chewing, EE decreased significantly, yet remained significantly higher than that during NG. During 150–170 min (i.e., during chewing), EE in the GC condition was again significantly higher than that in the NG condition (see Fig. 3). Upon cessation of chewing, EE decreased significantly, this time to a level not significantly different than that during NG, similar to the pattern seen during the fasting state. The elevated EE at 80–90 min (after chewing) might suggest a carryover effect of chewing. However, since this effect was not present in the fasting or late postprandial periods, it is likely that postprandial thermogenesis or other factors may have impacted EE at 80–90 min, so a substantial carryover effect after chewing is unlikely. Fasting and postprandial RER did not differ significantly between GC and NG (see Fig. 4), indicating that neither the nominal energy content of the gum nor the act of chewing influenced substrate oxidation. This is not surprising given that subjects chewed sugar-free gum containing sugar alcohols and that there were no differences in blood glucose between GC and NG (see Table 2 and Fig. 5). Animal data have suggested that chewing may activate histamine pathways (Sakata et al. 2003) that could stimulate lipolysis (Masaki et al. 2001). However, human data are lacking, and it is probable that the amount of chewing in this study was insufficient to induce measureable changes. Gum chewing is associated with significant intrinsic tongue muscle activity, along with activity of the masticatory muscles (Rikimaru et al. 2001). This muscular movement is likely responsible for the majority of increased energy expenditure observed. Very few published papers have examined EE during gum chewing. One brief letter reported on EE measured in 7 nonobese subjects before chewing gum (58 ± 11 kcal/h), while chewing gum for 12 min (70 ± 14 kcal/h), and for 12 min after chewing gum (59 ± 12 kcal/h) (Levine et al. 1999). The authors stated that an average increase of 11 ± 3 kcal/h (19% ± 4%) associated with chewing was greater than increases in EE associated with standing (11% ± 11%) measured in the same subjects. The increase they reported (approximately 0.18 ± 0.05 kcal/min) is substantially larger than the increases we are reporting, in both fasting (0.094 kcal/min, 8.1% ± 3.6%) and postprandial states (0.078 kcal/min, 5.3% ± 3.5% at 60– 80 min; 0.071 kcal/min, 5.2% ± 3.9% at 150–170 min) (see Figs. 2 and 3). Some of that difference might be explained by subjects chewing 8.4 g of gum compared with 1.9 g in our study. In addition, chewing rate might also have affected EE differences, as discussed below. Only limited comparisons are possible with our study, however, because the experimental conditions in the report of Levine et al. (1999) were not revealed (e.g., fasting vs. fed; anthropometric data), and there does not appear to have been a control condition. In the brief letter described above (Levine et al. 1999), subjects chewed at a frequency of 100 bites/min. A subsequent paper examined the impact of various doses of caffeine and nicotine in gums on EE in a fasting study population similar to ours (Jessen et al. 2003). During the placebo gum condition, subjects in that study had an increase in EE of approximately 4 kcal/h while chewing at a cadence of 60 bites/min for 25 min. In our study, during the fasting state, the average difference in EE during chewing was 5.6 kcal/h while subjects chewed at a self-selected rate. We used
405
the self-selected chewing rate to assess the effects of chewing under conditions that mimic the real world. In addition, there is evidence that subjects may not comply with an artificial cadence while chewing (Nemeth-Coslett et al. 1988). In a dose–response study of nicotine gum, subjects were asked to match their chewing to cadence tones of one bite per second, per 2 s, per 4 s, and per 8 s. Despite the instructions to follow the cadence, subjects’ measured chewing rates were one bite per 1.9 s, 2.1 s, 3.7 s, and 4.2 s under each condition, respectively (Nemeth-Coslett et al. 1988). Thus, it appears possible that the use of cadence might introduce variance into chewing studies, rather than controlling it, owing to noncompliance. It remains plausible that chewing rate could have influenced EE. We did not measure subjects’ chewing rates, so this is a potential area for future research. Varying the duration of chewing might also impact the total energy expended in gum chewing. The amount of energy expended in each 20-min bout of chewing a sugar-free gum was approximately equal to the energy content of the sugar-free gum (⬃2–3 kcal). This would suggest an advantage of chewing sugar-free gum over gums containing sugar and more energy, since it is possible that the energy contained in sugar-sweetened gums (⬃20–25 kcal) might not be expended in a normal chewing session of 20 min or less. If people use sugar-free gum as a substitute for snacks containing more energy, then they may not have to account for the energy in the sugar-free gum if they chew it for a long enough time. Thus, this study offers empirical evidence that can be used for patients or clients who tend to chew a lot of gum and (or) seek to replace snacks with gum. While it may be tempting to calculate the potential impact of gum chewing on overall energy balance, such extrapolations should be made conservatively. At all 3 chewing time periods in the current study, EE declined rapidly and significantly upon the cessation of chewing. To impact whole-day energy expenditure, individuals would have to chew gum throughout the day. Strengths of this study include testing 3 sessions of gum chewing in fasting and fed states in both sexes under carefully standardized conditions. However, our subject population was limited to young normal-weight and overweight individuals in a controlled laboratory setting over the course of one morning. The impact on other populations in free-living conditions has not been explored. Without a carryover effect, the contribution of gum chewing to daily EE will likely be related to the rate and duration of chewing and is likely to be relatively small, as suggested by the effect sizes. The impact of gum chewing on energy balance, which is crucial to understanding its potential role in weight management, needs to be considered in conjunction with data on energy intake. If gum chewing is associated with reduced energy intake in addition to increased EE, as demonstrated here, then its value as part of a comprehensive weight control program would be enhanced. An 8-week randomized control study in 201 adults undergoing a diet or a diet-plus-gum chewing (90 min/day) intervention showed no between-group differences in body weight (Shikany et al. 2012). However, the dietary information provided to both groups was aimed at healthy eating rather than weight loss. Within the dietplus-gum chewing group only, decreases in waist circumference and blood pressure were seen. Given these considerations, this area of research may warrant more investigation. In summary, chewing gum at one’s natural pace increases energy expenditure in the fasting and fed states without impacting blood glucose or substrate oxidation. Longer-term studies should be conducted to fully discern whether gum chewing would improve energy balance in weight management programs.
Disclosure Wrigley Science Institute donated the chewing gum used in this study. The authors were responsible for all aspects of the research design, data collection, and analysis and manuscript preparation. Published by NRC Research Press
406
Acknowledgements This study was funded by a research award from The Obesity Society (TOS) presented in October 2007. The award was made possible by a donation to TOS from Wrigley Science Institute.
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by San Francisco (UCSF) on 04/12/15 For personal use only.
References Brennan, I.M., Feltrin, K.L., Nair, N.S., Hausken, T., Little, T.J., Gentilcore, D., et al. 2009. Effects of the phases of the menstrual cycle on gastric emptying, glycemia, plasma GLP-1 and insulin, and energy intake in healthy lean women. Am. J. Physiol. Gastrointest. Liver Physiol. 297(3): G602–G610. doi:10. 1152/ajpgi.00051.2009. PMID:19556358. Bryant, M., Truesdale, K.P., and Dye, L. 2006. Modest changes in dietary intake across the menstrual cycle: implications for food intake research. Br. J. Nutr. 96(5): 888–894. doi:10.1017/BJN20061931. PMID:17092378. Cohen, J. 1988. Statistical power analysis for the behavioral sciences. 2nd ed. L. Erlbaum Associates, Hillsdale, N.J. Craig, C.L., Marshall, A.L., Sjostrom, M., Bauman, A.E., Booth, M.L., Ainsworth, B.E., et al. 2003. International physical activity questionnaire: 12-country reliability and validity. Med. Sci. Sports Exerc. 35(8): 1381–1395. doi:10.1249/01.MSS. 0000078924.61453.FB. PMID:12900694. Dye, L., and Blundell, J.E. 1997. Menstrual cycle and appetite control: implications for weight regulation. Hum. Reprod. 12(6), 1142–1151. PMID:9221991. Esliger, D.W., and Tremblay, M.S. 2006. Technical reliability assessment of three accelerometer models in a mechanical setup. Med. Sci. Sports Exerc. 38(12): 2173–2181. doi:10.1249/01.mss.0000239394.55461.08. PMID:17146326. Fields, D.A., Goran, M.I., and McCrory, M.A. 2002. Body-composition assessment via air-displacement plethysmography in adults and children: a review. Am. J. Clin. Nutr. 75(3): 453–467. PMID:11864850. Hetherington, M.M., and Boyland, E. 2007. Short-term effects of chewing gum on snack intake and appetite. Appetite, 48(3): 397–401. doi:10.1016/j.appet.2006. 10.001. PMID:17118491. Jessen, A.B., Toubro, S., and Astrup, A. 2003. Effect of chewing gum containing nicotine and caffeine on energy expenditure and substrate utilization in men. Am. J. Clin. Nutr. 77(6): 1442–1447. PMID:12791621. Julis, R.A., and Mattes, R.D. 2007. Influence of sweetened chewing gum on appetite, meal patterning and energy intake. Appetite, 48(2): 167–175. doi:10.1016/ j.appet.2006.08.003. PMID:17050036. Leveille, G., McMahon, K., Alcantara, E., and Zibell, S. 2008. Benefits of chewing
Appl. Physiol. Nutr. Metab. Vol. 40, 2015
gum: oral health and beyond. Nutr. Today, 43(2), 75–81. doi:10.1097/01.NT. 0000303316.67157.c5. Levine, J., Baukol, P., and Pavlidis, I. 1999. The energy expended in chewing gum. N. Engl. J. Med. 341(27): 2100. doi:10.1056/NEJM199912303412718. PMID: 10627208. Masaki, T., Yoshimatsu, H., Chiba, S., Watanabe, T., and Sakata, T. 2001. Central infusion of histamine reduces fat accumulation and upregulates UCP family in leptin-resistant obese mice. Diabetes, 50(2), 376–384. doi:10.2337/diabetes. 50.2.376. PMID:11272150. McArdle, W.D., Katch, F.I., Pechar, G.S., Jacobson, L., and Ruck, S. 1972. Reliability and interrelationships between maximal oxygen intake, physical work capacity and step-test scores in college women. Med. Sci. Sports, 4(4): 182–186. PMID:4648576. McCrory, M.A., Mole, P.A., Gomez, T.D., Dewey, K.G., and Bernauer, E.M. 1998. Body composition by air-displacement plethysmography by using predicted and measured thoracic gas volumes. J. Appl. Physiol. 84(4): 1475–1479. PMID: 9516218. Nemeth-Coslett, R., Benowitz, N.L., Robinson, N., and Henningfield, J.E. 1988. Nicotine gum: chew rate, subjective effects and plasma nicotine. Pharmacol., Biochem. Behav. 29(4): 747–751. doi:10.1016/0091-3057(88)90197-9. PMID: 3413200. Rikimaru, H., Kikuchi, M., Itoh, M., Tashiro, M., and Watanabe, M. 2001. Mapping energy metabolism in jaw and tongue muscles during chewing. J. Dent. Res. 80(9): 1849–1853. doi:10.1177/00220345010800091501. PMID:11926246. Sakata, T., Yoshimatsu, H., Masaki, T., and Tsuda, K. 2003. Anti-obesity actions of mastication driven by histamine neurons in rats. Exp. Biol. Med. (Maywood), 228(10): 1106–1110. PMID:14610247. Shikany, J.M., Thomas, A.S., McCubrey, R.O., Beasley, T.M., and Allison, D.B. 2012. Randomized controlled trial of chewing gum for weight loss. Obesity (Silver Spring), 20(3): 547–552. PMID:22076595. Siri, W. 1961. Body composition from fluid spaces and density: analysis of methods. In Techniques for measuring body composition. Edited by J. Broézek and A. Henschel. National Academy of Sciences-National Research Council, Washington, D.C. pp. 113–224. PMID:8286893.. Trost, S.G., McIver, K.L., and Pate, R.R. 2005. Conducting accelerometer-based activity assessments in field-based research. Med. Sci. Sports Exerc. 37(11): S531–S543. doi:10.1249/01.mss.0000185657.86065.98. PMID:16294116.
Published by NRC Research Press
