- Email: [email protected]
Does basal metabolic rate drive eating rate?
Physiology & Behavior 189 (2018) 74–77
Contents lists available at ScienceDirect
Physiology & Behavior journal homepage: www.elsevier.com/locate/physbeh
Brief communication
Does basal metabolic rate drive eating rate? Christiani Jeyakumar Henry a b c
a,b,⁎
T
a
a
, Shalini Ponnalagu , Xinyan Bi , Ciaran Forde
a,c
Clinical Nutrition Research Centre, A*Star, Singapore Institute for Clinical Sciences, Singapore Department of Biochemistry, National University of Singapore, Singapore Department of Physiology, National University of Singapore, Singapore
A R T I C LE I N FO
A B S T R A C T
Keywords: Basal metabolic rate Eating rate Fat free mass Energy intake Energy requirement
There have been recent advances in our understanding of the drivers of energy intake (EI). However, the biological drivers of differences in eating rate (ER) remain less clear. Studies have reported that the fat-free mass (FFM) and basal metabolic rate (BMR) are both major components that contribute to daily energy expenditure (EE) and drive EI. More recently, a number of observations report that higher ER can lead to greater EI. The current study proposed that adults with a higher BMR and higher energy requirements would also exhibit higher ERs. Data on BMR, FFM, and ER were collected from 272 Chinese adults (91 males and 181 females) in a crosssectional study. Analysis showed significant positive associations between BMR and ER (rs = 0.405, p < 0.001), and between FFM and ER (rs = 0.459, p < 0.001). BMR explained about 15% of the variation in ER which was taken to be metabolically significant. This association provides metabolic explanation that the differences in an individual's BMR (hence energy requirements) may be correlated with ERs. This merits further research.
1. Introduction
calculate the energy requirements [12]. This unique approach focused considerable attention on BMR and its metabolic consequences. Previous research by Lissner et al. in 1989 [13] reported that EI was correlated with fat-free mass (FFM) and not fat mass (FM). This was a major observation as it denoted for the first time that a possible driver of EI may be an individual's energy requirement, contributed mainly by BMR. More recently Blundell and colleagues [14] have reported similar observations about the predictive power of the FFM on EI. Moreover, FFM was implicated as being an orexigenic (increase appetite) or ergostatic driver of meal size and EI [14]. A number of recent acute feeding studies, epidemiological reports and meta-analyses have demonstrated that eating rate (ER) can influence acute EI such that faster eaters tend to consume more energy than slower eaters within a meal. It is also correlated with greater BMI and adiposity among children and adults [15–20]. In addition, it has also been observed that there were large differences in ER between individuals consuming the same foods (with the same food textures) [18,20,21]. However, the association between ER and underlying differences in BMR and EI requirements has yet to be established. In a similar way to BMR [22], a recent study had also demonstrated that an individual's ER is consistent over time across multiple consumptions of the same meal, and predictive of EI [23]. This raises the possibility that there may be an underlying association between the metabolic requirement for energy to support a higher BMR, and the associated behavioural response to have a higher ER that supports increased EI.
Energy balance (EB) is made up of two components namely, energy intake (EI) and energy expenditure (EE), i.e. EB = EI – EE. A considerable amount of research has centered on understanding the mechanisms that regulate both EI and EE and the potential underlying physiological differences that drive differences at an individual level [1]. Previously, glucose, fat and temperature were proposed to be involved in the regulation of food intake and were classified as the glucostatic, lipostatic and thermostatic theories regulating food intake [2–4]. However, studies over the years have shown that EI is not exclusively regulated by these signals [5,6]. In 1955, Edholm et al. [7,8] reported that EI was driven by daily EE, and proposed that EB was regulated through this mechanism. The largest component of EE (representing up to 70% of total EE) is the basal metabolic rate (BMR), which is the amount of energy per unit time that a person needs to keep the body functioning at rest. The total EE in adults can be compartmentalized into four components namely BMR, physical activity (22–30% of total EE), diet induced thermogenesis (10–15% of total EE) and growth component (< 2% of total EE) [9,10]. Indeed even in children, apart from the first year of life and during puberty where growth is a contributor, BMR still remains by far the largest component. Interest in the role of BMR in determining EI emerged with the publication of the document Energy and Protein Requirements in 1985 [11] that proposed the use of EE (i.e. BMR) rather than food intake to
⁎
Corresponding author at: Clinical Nutrition Research Centre, A*Star, Singapore Institute for Clinical Sciences, Singapore. E-mail address: [email protected] (C.J. Henry).
https://doi.org/10.1016/j.physbeh.2018.03.013 Received 28 November 2017; Received in revised form 31 January 2018; Accepted 12 March 2018 Available online 13 March 2018 0031-9384/ © 2018 Published by Elsevier Inc.
Physiology & Behavior 189 (2018) 74–77
C.J. Henry et al.
appetite need state 2 h before an ad libitum buffet lunch was served. The 2-h gap between breakfast and lunch was deemed appropriate to normalize appetite sensations following the standardised test breakfast, and was chosen to ensure all participants were in a similar need state at the beginning of the test meal, in line with previous studies in a similar test population [26]. The ad libitum buffet lunch consisted of 1000 g (189 kcal/100 g) of olive vegetable fried rice (JR Foods, Singapore; main ingredients: rice, egg, olive, garlic, and vegetable oil) and a glass of plain water (250 mL). The olive vegetable fried rice was reheated from frozen according to the manufactures instructions. Olive vegetable fried rice was chosen as it was a familiar, contextually appropriate lunchtime test meal that was acceptable. This test lunch was chosen based on experience in previous studies [27] as it was not highly hedonically appealing, had a medium energy density (189 kcal/100 g), so as not to exaggerate differences in EI and was homogenous with limited variety to minimize the impact of variety on sensory specific satiety. Our previous studies have shown that minor changes in food liking had negligible effect on ER [23]. These factors in combination imply that the ER was measured objectively without being influenced by plausible confounding variables. Participants were given 15 min to consume their lunch and were instructed to eat as little or as much as they wish until they feel comfortably full. The task was recorded on a webcam (Logitech HD c310) mounted on a computer and all participants were told that the session would be recorded, but they were unable to see themselves on the video display. Leftover rice and water were weighed using a Sartorius balance (Goettingen, Germany) and weight was recorded, to two-decimal places, to derive the amount consumed.
The current study seeks to explore whether there is a relationship between energy requirement, using BMR as a proxy estimate, and ER as a measure of an adaptive behaviour to support increased requirements for energy. Since ER has been shown to affect EI within and across meals, the question arises as to whether ER varies across individuals in a similar way to BMR. The goal of our study was to examine whether there is an association between ER and an individual's energy requirement, as explained by differences in their underlying BMR. We set out to (i) confirm the relationship between FFM and BMR in a large cross-sectional Asian cohort, (ii) test whether there is an association between BMR and ER, and (iii) test whether there is a relationship between FFM and ER (as this relationship is yet to be tested). 2. Methods and subjects The data for this paper was obtained from a cross-sectional study conducted at Clinical Nutrition Research Centre (CNRC), Singapore between June 2015 and December 2016. The participants included 272 healthy adults aged 21 to 69 years old: 91 males (33.5%) and 181 females (66.5%). They were recruited through advertisement on newspaper and posters that were placed around the National University of Singapore campus, public area, and on the CNRC website. Eligibility included healthy males and females who were Singaporean or permanent resident, who had resided in Singapore for at least five years and were not diagnosed with any major diseases, taking any regular medication, or pregnant. All procedures involving human subjects were approved by the National Healthcare Group Domain Specific Review Board (Reference number: 2013/00783), Singapore. All participants gave written informed consent before starting.
2.3. Behavioural coding of eating rate 2.1. Body composition and BMR The video recordings of the ad libitum lunch were coded by a trained assessor using a coding scheme to capture the total number of bites, chews and swallows and time in mouth. This coding scheme is described in detail elsewhere [28,29]. Using these measures, it was possible to derive a series of eating behaviours including average bite size, chews per bite, oral-processing time and eating rate (g/min or kcal/ min). Videos were coded using behavioural annotation software ELAN (ELAN 4.9.2, Max Planck Institute for Psycholinguistics). Behavioural video coding was completed by a single trained video-coder and 10% of all coded videos were later blind-validated by the second trained videocoder, and only accepted as data when they surpass an agreement level of 80%. This is in line with the previously published recommendations [30]. Inter-rater reliability was further assessed using two-way mixed, consistency, single-measures to assess the consistency between coders based on 10% of the subjects. This resulted in high levels of agreement between the coders, ICC = 0.98 (95% CI, 0.96–0.99). Hence, duration of food in mouth as measured by the coder was deemed suitable for the current study. Average ER was calculated by dividing the grams consumed by the total oral exposure time (g/min) [28]. All statistical analysis in this study was done using Statistical Package for the Social Sciences (IBM SPSS version 24, IBM Corp, Armnok, NY, USA) and two-tailed statistical significance test was set at α = 0.05. Following measures of association in this study were evaluated using Spearman's correlation as it is robust to deviations from the center. A minimum sample of 84 was necessary to achieve a medium sized correlation of 0.3 at 0.05 significance level with 80% power. The sample size of 272 used in this study was seen to be more than necessary.
Weight (kg) was measured to the nearest 0.1 kg in light clothing without footwear by using an electrical weighing scale (Seca Limited, Birmingham, UK) and height (cm) was measured using a stadiometer (Seca Limited, Birmingham, UK) to the nearest 0.1 cm. BMR was measured using an indirect calorimeter (COSMED, Rome, Italy) in the fasted state (at least 12 h from 10 pm the night before the test day). Compliance to this protocol was checked using their completed diet record and a physical activity questionnaire for the day before the test session and on the day of the test session. The participants were instructed to lie in a supine position for 30 min, in a room maintained at thermo-neutral temperature (22–26 °C), without moving while a plastic canopy with veil was placed over the upper part of their body to prevent external air from entering the hood [24]. During the period of measurement, the participants were instructed to not sleep and to limit their movement. BMR was measured using the last 10 min of the measurement period to ensure stable and interpretable measurements are obtained [25]. The amount of FeO2 was verified to be within 0.7%–1.3% throughout the test duration and was adjusted if necessary. Ergo (reference value for O2: 1598, CO2: 503) and air (reference value for O2: 2093, CO2: 4) calibration were done prior to measurements. FFM was evaluated from body composition measures assessed using dual-energy X-ray absorptiometry (DXA, QDR 4500A, fan-beam densitometer, Hologic, Waltham, USA). It was calculated by using the manufacturer's software (software version 8.21). All anthropometric measurements were taken from the participants on the day they reported for the study. 2.2. Energy intake
3. Results and discussion On the morning of the test session, participants were requested to consume a standardised breakfast comprising a small packet of orange juice (Marigold: 250 mL; 46 kcal/100 mL) and two slices of white bread (Gardenia: 57 g × 2; 263 kcal/100 g) with kaya spread (NTUC FairPrice: 16 g; 300 kcal/100 g). This was consumed to standardize
Table 1 shows characteristic data for the 272 Chinese participants recruited in the study. The average age of the participants was 39 years with BMI ranging from 16.3 kg/m2 to 37.8 kg/m2. The participant's ER varied widely from 8.6 g/min to 68.2 g/min. Table 1 shows that males 75
Physiology & Behavior 189 (2018) 74–77
C.J. Henry et al.
80
Table 1 Characteristics of 272 participants. Males (n = 91)
Mean (SD)
Range
Mean (SD)
Range
38.7 (13.8) 55.4 (9.6) 160.0 (5.6) 21.6 (3.3) 35.4 (4.9)
21.3–68.6 40.0–103.9 144.2–174.8 16.4–37.8 27.2–54.9
40.8 (14.3) 67.7 (8.6) 170.6 (5.9) 23.3 (2.7) 49.5 (4.9)
21.0–68.0 49.0–90.8 157.6–184.3 16.3–28.6 37.6–62.1
0.25 < 0.001 < 0.001 < 0.001 < 0.001
1114.1 (174.5) 26.2 (9.3)
749–1853
1403.7 (166.2) 35.7 (11.1)
988–1896
< 0.001
14.5–68.2
< 0.001
8.6–54.7
60 ER (g/min)
Age (y) Weight (kg) Height (cm) BMI (kg/m2) FFM (DXA) (kg) BMR (kcal/ day) ER (g/min)
70 P valuea
Females (n = 181)
50 40 30 20 10 0
20
30
40
50
60
70
FFM (kg) Abbreviation: SD, standard deviation; BMI, body mass index; FFM, fat free mass; BMR, basal metabolic rate; ER, eating rate; DXA: dual-energy x-ray absorptiometry. a p value: non-parametric t-test testing difference in means between males and females.
Fig. 2. Scatter plot visually representing the association between ER and FFM in 272 participants. Males (n = 91) were represented by the open triangles and females (n = 181) by the open circles. Spearman's correlation coefficient using the entire dataset rs = 0.459, p value < 0.001.
80 70
underlying energy requirements (BMR) and proposes that ER may be an adaptive behavioural response to higher energy requirements. It is plausible that the association between ER and BMR could be in response to the greater energy required to maintain the body's metabolism such that individual differences in overall energy requirement are driving individual differences in the rate of EI. Faster eating rates have been shown to promote greater EI during ad libitum meals, as observed in a number of recent studies [15–20]. These first findings are tentative and should be interpreted with caution, as the current cross-sectional study was not specifically set up to test this relationship. This cannot confirm the direction of the relationship between ER and BMR. The relationship seems to show that ER is at least partially account for by the energy needs. Future studies are necessary to confirm the relationship between ER and BMR to compare their ER and ad-libitum EI across a range of different food items. If an individual's ER is partially governed by energy requirements linked by BMR, it questions the longer term efficacy of short term interventions aimed at slowing down ERs using cues or harder food textures to reduce EI [33]. This further highlights the need to explore adaptive eating behaviours in conjunction with harder food texture to modulate eating rate [33]. Individuals who eat at a faster rate are at a higher risk of increased EI, particularly when these faster ERs are combined with higher energy density foods [23]. Further studies are now needed to confirm the relationship between BMR and ER, and to compare the habitual eating rates of people with low medium and high BMR's. The strengths of this study lie in the objective measurement of BMR (as per standard protocol), ER, FFM (as per standard protocol) and EI in a large and diverse sample size. The current study also has a number of limitations. We were unable to establish a causal relationship or the direction of the association between BMR and ER as the current study is a cross-sectional comparison of measurements obtained from participants at a single time point. Additional studies are necessary within other ethnic groups and over longer periods of time to confirm these associations and for generalizability of the results. We also did not account for the stage of the menstrual cycle of the participants in the current study [31,34]. Future studies should account for these in their test methodology. Whether the results obtained in the current study with adults is applicable to children and puberty remains to be further explored.
ER (g/min)
60 50 40 30 20 10 0
500
700
900
1100
1300
1500
1700
1900
2100
BMR (kcal/day) Fig. 1. Scatter plot visually representing the association between ER and BMR in 272 participants. Males (n = 91) were represented by the open triangles and females (n = 181) by the open circles. Spearman's correlation coefficient using the entire dataset rs = 0.405, p < 0.001.
had significantly higher FFM, ER, and BMR than females (all p < 0.001). As proposed, there was a significant positive association (rs = 0.405, p < 0.001) between BMR and ER (Fig. 1). Moreover, the association between BMR and ER remained significant after controlling for participants' BMI (partial correlation coefficient rs = 0.237, p < 0.001), indicating that the association was independent of BMI. This association between BMR and ER still remained significant even after controlling for participants' EI (partial correlation coefficient rs = 0.298, p < 0.001). Significant positive association (rs = 0.459, p < 0.001) was also observed between ER and FFM (Fig. 2). To the best of our knowledge, this is the first time such an association has been observed. There was no significant interaction between the gender and BMR (p = 0.292) in a linear regression of ER (dependent variable) with BMR and gender (as independent variables). This implies that the change in ER for a unit change in BMR was not significantly different between males and females. Furthermore, the relationship between BMR and ER remained significant among males (rs = 0.274, p = 0.009) and females (rs = 0.180, p = 0.015). The difference in the strength of the correlation between males and females would potentially be due to changes in BMR among females throughout the menstrual cycle [31,32] while being remarkably constant in males [22]. Simple linear regression of ER (dependent variable) and BMR (explanatory variable) showed that BMR explained about 15% of the variation in ER (Table 2). Using the regression equation, this means that for every 100Kcal/d increase in BMR, the ER increases by 2 g/min. The result contributes to our understanding of differences in an individual's
4. Conclusion Our study has identified an association between ER and BMR, suggesting that the faster ERs that support greater EI may be an adaptive behavioural response to higher energy requirements. Significant 76
Physiology & Behavior 189 (2018) 74–77
C.J. Henry et al.
Table 2 Multiple regression on ER by BMR in 272 participants. The regression equation: ER = 6.27 + 0.02 (BMR). Variables
Coefficients (β)
Standardised coefficients
SE
P value
R2 (coefficient of determination)
BMR Constant
0.02 6.27
0.39
0.003 3.43 SEE = 10.1
< 0.001 0.068
0.148
Abbreviations: BMR, Basal Metabolic Rate; SE, standard error; SEE, standard error of estimation.
association between ER and FFM was also shown in this study. The variations in BMR observed across our sample, account for approximately 15% of the variation in ER, which we propose to be a metabolically significant association. We also conclude that ER is significantly associated with the energy requirements of the body and this relationship merits further investigation in the future.
[14] J.E. Blundell, P. Caudwell, C. Gibbons, M. Hopkins, E. Naslund, N.A. King, et al., Body composition and appetite: fat-free mass (but not fat mass or BMI) is positively associated with self-determined meal size and daily energy intake in humans, Br. J. Nutr. 107 (2012) 445–449. [15] E. Robinson, E. Almiron-Roig, F. Rutters, C. de Graaf, C.G. Forde, C. Tudur Smith, et al., A systematic review and meta-analysis examining the effect of eating rate on energy intake and hunger, Am. J. Clin. Nutr. 100 (2014) 123–151. [16] T. Ohkuma, Y. Hirakawa, U. Nakamura, Y. Kiyohara, T. Kitazono, T. Ninomiya, Association between eating rate and obesity: a systematic review and meta-analysis, Int. J. Obes. 39 (2015) 1589. [17] B. Zhu, Y. Haruyama, T. Muto, T. Yamazaki, Association between eating speed and metabolic syndrome in a three-year population-based cohort study, J. Epidemiol. 25 (2015) 332–336. [18] A. Fogel, A.T. Goh, L.R. Fries, S.A. Sadananthan, S.S. Velan, N. Michael, et al., Faster eating rates are associated with higher energy intakes during an ad libitum meal, higher BMI and greater adiposity among 4.5-year-old children: results from the growing up in Singapore towards healthy outcomes (GUSTO) cohort, Br. J. Nutr. 117 (2017) 1042–1051. [19] M. Viskaal-van Dongen, F.J. Kok, C. de Graaf, Eating rate of commonly consumed foods promotes food and energy intake, Appetite 56 (2011) 25–31. [20] J. van den Boer, M. Werts, E. Siebelink, C. de Graaf, M. Mars, The availability of slow and fast calories in the Dutch diet: the current situation and opportunities for interventions, Foods 6 (2017) 87. [21] R.S. Mattfeld, E.R. Muth, A. Hoover, A comparison of bite size and BMI in a cafeteria setting, Physiol. Behav. 181 (2017) 38–42. [22] C. Henry, J. Hayter, D. Rees, The constancy of basal metabolic rate in free-living male subjects, Eur. J. Clin. Nutr. 43 (1989) 727–731. [23] K. McCrickerd, C.G. Forde, Consistency of eating rate, oral processing behaviours and energy intake across meals, Nutrients. 9 (2017). [24] C. Henry, Basal metabolic rate studies in humans: measurement and development of new equations, Public Health Nutr. 8 (2005) 1133–1152. [25] C. Compher, D. Frankenfield, N. Keim, L. Roth-Yousey, Group, E. A. W, Best practice methods to apply to measurement of resting metabolic rate in adults: a systematic review, J. Am. Diet. Assoc. 106 (2006) 881–903. [26] S. Tey, N. Salleh, J. Henry, C. Forde, Effects of aspartame-, monk fruit-, stevia-and sucrose-sweetened beverages on postprandial glucose, insulin and energy intake, Int. J. Obes. 41 (2017) 450–457. [27] K. McCrickerd, N. Salleh, C. Forde, Removing energy from a beverage influences later food intake more than the same energy addition, Appetite 105 (2016) 549–556. [28] C.G. Forde, N. van Kuijk, T. Thaler, C. de Graaf, N. Martin, Oral processing characteristics of solid savoury meal components, and relationship with food composition, sensory attributes and expected satiation, Appetite 60 (2013) 208–219. [29] C. Forde, C. Leong, E. Chia-Ming, K. McCrickerd, Fast or slow-foods? Describing natural variations in oral processing characteristics across a wide range of Asian foods, Food Funct. 8 (2017) 595–606. [30] K.K. Haidet, J. Tate, D. Divirgilio-Thomas, A. Kolanowski, M.B. Happ, Methods to improve reliability of video-recorded behavioral data, Res. Nurs. Health 32 (2009) 465–474. [31] C.J.K. Henry, H.J. Lightowler, J. Marchini, Intra-individual variation in resting metabolic rate during the menstrual cycle, Br. J. Nutr. 89 (2003) 811–817. [32] P. Shetty, C. Henry, A. Black, A. Prentice, Energy requirements of adults: an update on basal metabolic rates (BMRs) and physical activity levels (PALs), Eur. J. Clin. Nutr. 50 (1996) S11–23. [33] D.P. Bolhuis, C.G. Forde, Y. Cheng, H. Xu, N. Martin, C. de Graaf, Slow food: sustained impact of harder foods on the reduction in energy intake over the course of the day, PLoS One 9 (2014) e93370. [34] S.A. Elliott, J. Ng, M.K.-S. Leow, C.J. Henry, The influence of the menstrual cycle on energy balance and taste preference in Asian Chinese women, Eur. J. Nutr. 54 (2015) 1323–1332.
Funding Studies reported in this manuscript were funded by the Agency for Science, Technology and Research (A*STAR) (SPF/2013 - 003), Singapore. Conflict of interest The authors declare no conflict of interest. Acknowledgments We would like to acknowledge Yi Ting Loo, Chia Ming En Edwin, Ai Ting Goh, Siew Ling Tey, Nurhazwani Binte Salleh and Leong Shu-Fen Claudia for their assistance in data collection for the study. References [1] J.O. Hill, H.R. Wyatt, J.C. Peters, Energy balance and obesity, Circulation 126 (2012) 126–132. [2] J. Mayer, Glucostatic mechanism of regulation of food intake, New Engl. J. Med. 249 (1953) 13–16. [3] J. Mayer, Regulation of energy intake and the body weight: the glucostatic theory and the lipostatic hypothesis, Ann. N. Y. Acad. Sci. 63 (1955) 15–43. [4] J.R. Brobeck, Food intake as a mechanism of temperature regulation, Yale J. Biol. Med. 20 (1948) 545. [5] L.M. Bernstein, M.I. Grossman, An experimental test of the glucostatic theory of regulation of food intake, J. Clin. Investig. 35 (1956) 627–633. [6] Food intake and brain temperature, Nutr. Rev. 26 (1968) 215–216. [7] O.G. Edholm, J.G. Fletcher, E.M. Widdowson, R.A. McCance, The energy expenditure and food intake of individual men, Br. J. Nutr. 9 (2007) 286. [8] O. Edholm, Energy balance in man studies carried out by the division of human physiology. National institute for medical research, J. Hum. Nutr. 31 (1977) 413. [9] P. Payne, J. Waterlow, Relative energy requirements for maintenance, growth, and physical activity, Lancet 298 (1971) 210–211. [10] J.E. Blundell, P. Caudwell, C. Gibbons, M. Hopkins, E. Naslund, N. King, et al., Role of resting metabolic rate and energy expenditure in hunger and appetite control: a new formulation, Dis. Model. Mech. 5 (2012) 608–613. [11] FAO/WHO, Energy and protein requirements: report of a joint FAO/WHO Ad Hoc expert committee, FAO Nutrition Meetings Report Series, No. 52, WHO Technical Report Series, 522 1973. [12] C. Henry, Estimates of metabolic adaptation in women living in developing countries: technical limitations, J. Biosoc. Sci. 24 (1992) 347–353. [13] L. Lissner, J.-P. Habicht, B.J. Strupp, D. Levitsky, J.D. Haas, D. Roe, Body composition and energy intake: do overweight women overeat and underreport? Am. J. Clin. Nutr. 49 (1989) 320–325.
77
Contents lists available at ScienceDirect
Physiology & Behavior journal homepage: www.elsevier.com/locate/physbeh
Brief communication
Does basal metabolic rate drive eating rate? Christiani Jeyakumar Henry a b c
a,b,⁎
T
a
a
, Shalini Ponnalagu , Xinyan Bi , Ciaran Forde
a,c
Clinical Nutrition Research Centre, A*Star, Singapore Institute for Clinical Sciences, Singapore Department of Biochemistry, National University of Singapore, Singapore Department of Physiology, National University of Singapore, Singapore
A R T I C LE I N FO
A B S T R A C T
Keywords: Basal metabolic rate Eating rate Fat free mass Energy intake Energy requirement
There have been recent advances in our understanding of the drivers of energy intake (EI). However, the biological drivers of differences in eating rate (ER) remain less clear. Studies have reported that the fat-free mass (FFM) and basal metabolic rate (BMR) are both major components that contribute to daily energy expenditure (EE) and drive EI. More recently, a number of observations report that higher ER can lead to greater EI. The current study proposed that adults with a higher BMR and higher energy requirements would also exhibit higher ERs. Data on BMR, FFM, and ER were collected from 272 Chinese adults (91 males and 181 females) in a crosssectional study. Analysis showed significant positive associations between BMR and ER (rs = 0.405, p < 0.001), and between FFM and ER (rs = 0.459, p < 0.001). BMR explained about 15% of the variation in ER which was taken to be metabolically significant. This association provides metabolic explanation that the differences in an individual's BMR (hence energy requirements) may be correlated with ERs. This merits further research.
1. Introduction
calculate the energy requirements [12]. This unique approach focused considerable attention on BMR and its metabolic consequences. Previous research by Lissner et al. in 1989 [13] reported that EI was correlated with fat-free mass (FFM) and not fat mass (FM). This was a major observation as it denoted for the first time that a possible driver of EI may be an individual's energy requirement, contributed mainly by BMR. More recently Blundell and colleagues [14] have reported similar observations about the predictive power of the FFM on EI. Moreover, FFM was implicated as being an orexigenic (increase appetite) or ergostatic driver of meal size and EI [14]. A number of recent acute feeding studies, epidemiological reports and meta-analyses have demonstrated that eating rate (ER) can influence acute EI such that faster eaters tend to consume more energy than slower eaters within a meal. It is also correlated with greater BMI and adiposity among children and adults [15–20]. In addition, it has also been observed that there were large differences in ER between individuals consuming the same foods (with the same food textures) [18,20,21]. However, the association between ER and underlying differences in BMR and EI requirements has yet to be established. In a similar way to BMR [22], a recent study had also demonstrated that an individual's ER is consistent over time across multiple consumptions of the same meal, and predictive of EI [23]. This raises the possibility that there may be an underlying association between the metabolic requirement for energy to support a higher BMR, and the associated behavioural response to have a higher ER that supports increased EI.
Energy balance (EB) is made up of two components namely, energy intake (EI) and energy expenditure (EE), i.e. EB = EI – EE. A considerable amount of research has centered on understanding the mechanisms that regulate both EI and EE and the potential underlying physiological differences that drive differences at an individual level [1]. Previously, glucose, fat and temperature were proposed to be involved in the regulation of food intake and were classified as the glucostatic, lipostatic and thermostatic theories regulating food intake [2–4]. However, studies over the years have shown that EI is not exclusively regulated by these signals [5,6]. In 1955, Edholm et al. [7,8] reported that EI was driven by daily EE, and proposed that EB was regulated through this mechanism. The largest component of EE (representing up to 70% of total EE) is the basal metabolic rate (BMR), which is the amount of energy per unit time that a person needs to keep the body functioning at rest. The total EE in adults can be compartmentalized into four components namely BMR, physical activity (22–30% of total EE), diet induced thermogenesis (10–15% of total EE) and growth component (< 2% of total EE) [9,10]. Indeed even in children, apart from the first year of life and during puberty where growth is a contributor, BMR still remains by far the largest component. Interest in the role of BMR in determining EI emerged with the publication of the document Energy and Protein Requirements in 1985 [11] that proposed the use of EE (i.e. BMR) rather than food intake to
⁎
Corresponding author at: Clinical Nutrition Research Centre, A*Star, Singapore Institute for Clinical Sciences, Singapore. E-mail address: [email protected] (C.J. Henry).
https://doi.org/10.1016/j.physbeh.2018.03.013 Received 28 November 2017; Received in revised form 31 January 2018; Accepted 12 March 2018 Available online 13 March 2018 0031-9384/ © 2018 Published by Elsevier Inc.
Physiology & Behavior 189 (2018) 74–77
C.J. Henry et al.
appetite need state 2 h before an ad libitum buffet lunch was served. The 2-h gap between breakfast and lunch was deemed appropriate to normalize appetite sensations following the standardised test breakfast, and was chosen to ensure all participants were in a similar need state at the beginning of the test meal, in line with previous studies in a similar test population [26]. The ad libitum buffet lunch consisted of 1000 g (189 kcal/100 g) of olive vegetable fried rice (JR Foods, Singapore; main ingredients: rice, egg, olive, garlic, and vegetable oil) and a glass of plain water (250 mL). The olive vegetable fried rice was reheated from frozen according to the manufactures instructions. Olive vegetable fried rice was chosen as it was a familiar, contextually appropriate lunchtime test meal that was acceptable. This test lunch was chosen based on experience in previous studies [27] as it was not highly hedonically appealing, had a medium energy density (189 kcal/100 g), so as not to exaggerate differences in EI and was homogenous with limited variety to minimize the impact of variety on sensory specific satiety. Our previous studies have shown that minor changes in food liking had negligible effect on ER [23]. These factors in combination imply that the ER was measured objectively without being influenced by plausible confounding variables. Participants were given 15 min to consume their lunch and were instructed to eat as little or as much as they wish until they feel comfortably full. The task was recorded on a webcam (Logitech HD c310) mounted on a computer and all participants were told that the session would be recorded, but they were unable to see themselves on the video display. Leftover rice and water were weighed using a Sartorius balance (Goettingen, Germany) and weight was recorded, to two-decimal places, to derive the amount consumed.
The current study seeks to explore whether there is a relationship between energy requirement, using BMR as a proxy estimate, and ER as a measure of an adaptive behaviour to support increased requirements for energy. Since ER has been shown to affect EI within and across meals, the question arises as to whether ER varies across individuals in a similar way to BMR. The goal of our study was to examine whether there is an association between ER and an individual's energy requirement, as explained by differences in their underlying BMR. We set out to (i) confirm the relationship between FFM and BMR in a large cross-sectional Asian cohort, (ii) test whether there is an association between BMR and ER, and (iii) test whether there is a relationship between FFM and ER (as this relationship is yet to be tested). 2. Methods and subjects The data for this paper was obtained from a cross-sectional study conducted at Clinical Nutrition Research Centre (CNRC), Singapore between June 2015 and December 2016. The participants included 272 healthy adults aged 21 to 69 years old: 91 males (33.5%) and 181 females (66.5%). They were recruited through advertisement on newspaper and posters that were placed around the National University of Singapore campus, public area, and on the CNRC website. Eligibility included healthy males and females who were Singaporean or permanent resident, who had resided in Singapore for at least five years and were not diagnosed with any major diseases, taking any regular medication, or pregnant. All procedures involving human subjects were approved by the National Healthcare Group Domain Specific Review Board (Reference number: 2013/00783), Singapore. All participants gave written informed consent before starting.
2.3. Behavioural coding of eating rate 2.1. Body composition and BMR The video recordings of the ad libitum lunch were coded by a trained assessor using a coding scheme to capture the total number of bites, chews and swallows and time in mouth. This coding scheme is described in detail elsewhere [28,29]. Using these measures, it was possible to derive a series of eating behaviours including average bite size, chews per bite, oral-processing time and eating rate (g/min or kcal/ min). Videos were coded using behavioural annotation software ELAN (ELAN 4.9.2, Max Planck Institute for Psycholinguistics). Behavioural video coding was completed by a single trained video-coder and 10% of all coded videos were later blind-validated by the second trained videocoder, and only accepted as data when they surpass an agreement level of 80%. This is in line with the previously published recommendations [30]. Inter-rater reliability was further assessed using two-way mixed, consistency, single-measures to assess the consistency between coders based on 10% of the subjects. This resulted in high levels of agreement between the coders, ICC = 0.98 (95% CI, 0.96–0.99). Hence, duration of food in mouth as measured by the coder was deemed suitable for the current study. Average ER was calculated by dividing the grams consumed by the total oral exposure time (g/min) [28]. All statistical analysis in this study was done using Statistical Package for the Social Sciences (IBM SPSS version 24, IBM Corp, Armnok, NY, USA) and two-tailed statistical significance test was set at α = 0.05. Following measures of association in this study were evaluated using Spearman's correlation as it is robust to deviations from the center. A minimum sample of 84 was necessary to achieve a medium sized correlation of 0.3 at 0.05 significance level with 80% power. The sample size of 272 used in this study was seen to be more than necessary.
Weight (kg) was measured to the nearest 0.1 kg in light clothing without footwear by using an electrical weighing scale (Seca Limited, Birmingham, UK) and height (cm) was measured using a stadiometer (Seca Limited, Birmingham, UK) to the nearest 0.1 cm. BMR was measured using an indirect calorimeter (COSMED, Rome, Italy) in the fasted state (at least 12 h from 10 pm the night before the test day). Compliance to this protocol was checked using their completed diet record and a physical activity questionnaire for the day before the test session and on the day of the test session. The participants were instructed to lie in a supine position for 30 min, in a room maintained at thermo-neutral temperature (22–26 °C), without moving while a plastic canopy with veil was placed over the upper part of their body to prevent external air from entering the hood [24]. During the period of measurement, the participants were instructed to not sleep and to limit their movement. BMR was measured using the last 10 min of the measurement period to ensure stable and interpretable measurements are obtained [25]. The amount of FeO2 was verified to be within 0.7%–1.3% throughout the test duration and was adjusted if necessary. Ergo (reference value for O2: 1598, CO2: 503) and air (reference value for O2: 2093, CO2: 4) calibration were done prior to measurements. FFM was evaluated from body composition measures assessed using dual-energy X-ray absorptiometry (DXA, QDR 4500A, fan-beam densitometer, Hologic, Waltham, USA). It was calculated by using the manufacturer's software (software version 8.21). All anthropometric measurements were taken from the participants on the day they reported for the study. 2.2. Energy intake
3. Results and discussion On the morning of the test session, participants were requested to consume a standardised breakfast comprising a small packet of orange juice (Marigold: 250 mL; 46 kcal/100 mL) and two slices of white bread (Gardenia: 57 g × 2; 263 kcal/100 g) with kaya spread (NTUC FairPrice: 16 g; 300 kcal/100 g). This was consumed to standardize
Table 1 shows characteristic data for the 272 Chinese participants recruited in the study. The average age of the participants was 39 years with BMI ranging from 16.3 kg/m2 to 37.8 kg/m2. The participant's ER varied widely from 8.6 g/min to 68.2 g/min. Table 1 shows that males 75
Physiology & Behavior 189 (2018) 74–77
C.J. Henry et al.
80
Table 1 Characteristics of 272 participants. Males (n = 91)
Mean (SD)
Range
Mean (SD)
Range
38.7 (13.8) 55.4 (9.6) 160.0 (5.6) 21.6 (3.3) 35.4 (4.9)
21.3–68.6 40.0–103.9 144.2–174.8 16.4–37.8 27.2–54.9
40.8 (14.3) 67.7 (8.6) 170.6 (5.9) 23.3 (2.7) 49.5 (4.9)
21.0–68.0 49.0–90.8 157.6–184.3 16.3–28.6 37.6–62.1
0.25 < 0.001 < 0.001 < 0.001 < 0.001
1114.1 (174.5) 26.2 (9.3)
749–1853
1403.7 (166.2) 35.7 (11.1)
988–1896
< 0.001
14.5–68.2
< 0.001
8.6–54.7
60 ER (g/min)
Age (y) Weight (kg) Height (cm) BMI (kg/m2) FFM (DXA) (kg) BMR (kcal/ day) ER (g/min)
70 P valuea
Females (n = 181)
50 40 30 20 10 0
20
30
40
50
60
70
FFM (kg) Abbreviation: SD, standard deviation; BMI, body mass index; FFM, fat free mass; BMR, basal metabolic rate; ER, eating rate; DXA: dual-energy x-ray absorptiometry. a p value: non-parametric t-test testing difference in means between males and females.
Fig. 2. Scatter plot visually representing the association between ER and FFM in 272 participants. Males (n = 91) were represented by the open triangles and females (n = 181) by the open circles. Spearman's correlation coefficient using the entire dataset rs = 0.459, p value < 0.001.
80 70
underlying energy requirements (BMR) and proposes that ER may be an adaptive behavioural response to higher energy requirements. It is plausible that the association between ER and BMR could be in response to the greater energy required to maintain the body's metabolism such that individual differences in overall energy requirement are driving individual differences in the rate of EI. Faster eating rates have been shown to promote greater EI during ad libitum meals, as observed in a number of recent studies [15–20]. These first findings are tentative and should be interpreted with caution, as the current cross-sectional study was not specifically set up to test this relationship. This cannot confirm the direction of the relationship between ER and BMR. The relationship seems to show that ER is at least partially account for by the energy needs. Future studies are necessary to confirm the relationship between ER and BMR to compare their ER and ad-libitum EI across a range of different food items. If an individual's ER is partially governed by energy requirements linked by BMR, it questions the longer term efficacy of short term interventions aimed at slowing down ERs using cues or harder food textures to reduce EI [33]. This further highlights the need to explore adaptive eating behaviours in conjunction with harder food texture to modulate eating rate [33]. Individuals who eat at a faster rate are at a higher risk of increased EI, particularly when these faster ERs are combined with higher energy density foods [23]. Further studies are now needed to confirm the relationship between BMR and ER, and to compare the habitual eating rates of people with low medium and high BMR's. The strengths of this study lie in the objective measurement of BMR (as per standard protocol), ER, FFM (as per standard protocol) and EI in a large and diverse sample size. The current study also has a number of limitations. We were unable to establish a causal relationship or the direction of the association between BMR and ER as the current study is a cross-sectional comparison of measurements obtained from participants at a single time point. Additional studies are necessary within other ethnic groups and over longer periods of time to confirm these associations and for generalizability of the results. We also did not account for the stage of the menstrual cycle of the participants in the current study [31,34]. Future studies should account for these in their test methodology. Whether the results obtained in the current study with adults is applicable to children and puberty remains to be further explored.
ER (g/min)
60 50 40 30 20 10 0
500
700
900
1100
1300
1500
1700
1900
2100
BMR (kcal/day) Fig. 1. Scatter plot visually representing the association between ER and BMR in 272 participants. Males (n = 91) were represented by the open triangles and females (n = 181) by the open circles. Spearman's correlation coefficient using the entire dataset rs = 0.405, p < 0.001.
had significantly higher FFM, ER, and BMR than females (all p < 0.001). As proposed, there was a significant positive association (rs = 0.405, p < 0.001) between BMR and ER (Fig. 1). Moreover, the association between BMR and ER remained significant after controlling for participants' BMI (partial correlation coefficient rs = 0.237, p < 0.001), indicating that the association was independent of BMI. This association between BMR and ER still remained significant even after controlling for participants' EI (partial correlation coefficient rs = 0.298, p < 0.001). Significant positive association (rs = 0.459, p < 0.001) was also observed between ER and FFM (Fig. 2). To the best of our knowledge, this is the first time such an association has been observed. There was no significant interaction between the gender and BMR (p = 0.292) in a linear regression of ER (dependent variable) with BMR and gender (as independent variables). This implies that the change in ER for a unit change in BMR was not significantly different between males and females. Furthermore, the relationship between BMR and ER remained significant among males (rs = 0.274, p = 0.009) and females (rs = 0.180, p = 0.015). The difference in the strength of the correlation between males and females would potentially be due to changes in BMR among females throughout the menstrual cycle [31,32] while being remarkably constant in males [22]. Simple linear regression of ER (dependent variable) and BMR (explanatory variable) showed that BMR explained about 15% of the variation in ER (Table 2). Using the regression equation, this means that for every 100Kcal/d increase in BMR, the ER increases by 2 g/min. The result contributes to our understanding of differences in an individual's
4. Conclusion Our study has identified an association between ER and BMR, suggesting that the faster ERs that support greater EI may be an adaptive behavioural response to higher energy requirements. Significant 76
Physiology & Behavior 189 (2018) 74–77
C.J. Henry et al.
Table 2 Multiple regression on ER by BMR in 272 participants. The regression equation: ER = 6.27 + 0.02 (BMR). Variables
Coefficients (β)
Standardised coefficients
SE
P value
R2 (coefficient of determination)
BMR Constant
0.02 6.27
0.39
0.003 3.43 SEE = 10.1
< 0.001 0.068
0.148
Abbreviations: BMR, Basal Metabolic Rate; SE, standard error; SEE, standard error of estimation.
association between ER and FFM was also shown in this study. The variations in BMR observed across our sample, account for approximately 15% of the variation in ER, which we propose to be a metabolically significant association. We also conclude that ER is significantly associated with the energy requirements of the body and this relationship merits further investigation in the future.
[14] J.E. Blundell, P. Caudwell, C. Gibbons, M. Hopkins, E. Naslund, N.A. King, et al., Body composition and appetite: fat-free mass (but not fat mass or BMI) is positively associated with self-determined meal size and daily energy intake in humans, Br. J. Nutr. 107 (2012) 445–449. [15] E. Robinson, E. Almiron-Roig, F. Rutters, C. de Graaf, C.G. Forde, C. Tudur Smith, et al., A systematic review and meta-analysis examining the effect of eating rate on energy intake and hunger, Am. J. Clin. Nutr. 100 (2014) 123–151. [16] T. Ohkuma, Y. Hirakawa, U. Nakamura, Y. Kiyohara, T. Kitazono, T. Ninomiya, Association between eating rate and obesity: a systematic review and meta-analysis, Int. J. Obes. 39 (2015) 1589. [17] B. Zhu, Y. Haruyama, T. Muto, T. Yamazaki, Association between eating speed and metabolic syndrome in a three-year population-based cohort study, J. Epidemiol. 25 (2015) 332–336. [18] A. Fogel, A.T. Goh, L.R. Fries, S.A. Sadananthan, S.S. Velan, N. Michael, et al., Faster eating rates are associated with higher energy intakes during an ad libitum meal, higher BMI and greater adiposity among 4.5-year-old children: results from the growing up in Singapore towards healthy outcomes (GUSTO) cohort, Br. J. Nutr. 117 (2017) 1042–1051. [19] M. Viskaal-van Dongen, F.J. Kok, C. de Graaf, Eating rate of commonly consumed foods promotes food and energy intake, Appetite 56 (2011) 25–31. [20] J. van den Boer, M. Werts, E. Siebelink, C. de Graaf, M. Mars, The availability of slow and fast calories in the Dutch diet: the current situation and opportunities for interventions, Foods 6 (2017) 87. [21] R.S. Mattfeld, E.R. Muth, A. Hoover, A comparison of bite size and BMI in a cafeteria setting, Physiol. Behav. 181 (2017) 38–42. [22] C. Henry, J. Hayter, D. Rees, The constancy of basal metabolic rate in free-living male subjects, Eur. J. Clin. Nutr. 43 (1989) 727–731. [23] K. McCrickerd, C.G. Forde, Consistency of eating rate, oral processing behaviours and energy intake across meals, Nutrients. 9 (2017). [24] C. Henry, Basal metabolic rate studies in humans: measurement and development of new equations, Public Health Nutr. 8 (2005) 1133–1152. [25] C. Compher, D. Frankenfield, N. Keim, L. Roth-Yousey, Group, E. A. W, Best practice methods to apply to measurement of resting metabolic rate in adults: a systematic review, J. Am. Diet. Assoc. 106 (2006) 881–903. [26] S. Tey, N. Salleh, J. Henry, C. Forde, Effects of aspartame-, monk fruit-, stevia-and sucrose-sweetened beverages on postprandial glucose, insulin and energy intake, Int. J. Obes. 41 (2017) 450–457. [27] K. McCrickerd, N. Salleh, C. Forde, Removing energy from a beverage influences later food intake more than the same energy addition, Appetite 105 (2016) 549–556. [28] C.G. Forde, N. van Kuijk, T. Thaler, C. de Graaf, N. Martin, Oral processing characteristics of solid savoury meal components, and relationship with food composition, sensory attributes and expected satiation, Appetite 60 (2013) 208–219. [29] C. Forde, C. Leong, E. Chia-Ming, K. McCrickerd, Fast or slow-foods? Describing natural variations in oral processing characteristics across a wide range of Asian foods, Food Funct. 8 (2017) 595–606. [30] K.K. Haidet, J. Tate, D. Divirgilio-Thomas, A. Kolanowski, M.B. Happ, Methods to improve reliability of video-recorded behavioral data, Res. Nurs. Health 32 (2009) 465–474. [31] C.J.K. Henry, H.J. Lightowler, J. Marchini, Intra-individual variation in resting metabolic rate during the menstrual cycle, Br. J. Nutr. 89 (2003) 811–817. [32] P. Shetty, C. Henry, A. Black, A. Prentice, Energy requirements of adults: an update on basal metabolic rates (BMRs) and physical activity levels (PALs), Eur. J. Clin. Nutr. 50 (1996) S11–23. [33] D.P. Bolhuis, C.G. Forde, Y. Cheng, H. Xu, N. Martin, C. de Graaf, Slow food: sustained impact of harder foods on the reduction in energy intake over the course of the day, PLoS One 9 (2014) e93370. [34] S.A. Elliott, J. Ng, M.K.-S. Leow, C.J. Henry, The influence of the menstrual cycle on energy balance and taste preference in Asian Chinese women, Eur. J. Nutr. 54 (2015) 1323–1332.
Funding Studies reported in this manuscript were funded by the Agency for Science, Technology and Research (A*STAR) (SPF/2013 - 003), Singapore. Conflict of interest The authors declare no conflict of interest. Acknowledgments We would like to acknowledge Yi Ting Loo, Chia Ming En Edwin, Ai Ting Goh, Siew Ling Tey, Nurhazwani Binte Salleh and Leong Shu-Fen Claudia for their assistance in data collection for the study. References [1] J.O. Hill, H.R. Wyatt, J.C. Peters, Energy balance and obesity, Circulation 126 (2012) 126–132. [2] J. Mayer, Glucostatic mechanism of regulation of food intake, New Engl. J. Med. 249 (1953) 13–16. [3] J. Mayer, Regulation of energy intake and the body weight: the glucostatic theory and the lipostatic hypothesis, Ann. N. Y. Acad. Sci. 63 (1955) 15–43. [4] J.R. Brobeck, Food intake as a mechanism of temperature regulation, Yale J. Biol. Med. 20 (1948) 545. [5] L.M. Bernstein, M.I. Grossman, An experimental test of the glucostatic theory of regulation of food intake, J. Clin. Investig. 35 (1956) 627–633. [6] Food intake and brain temperature, Nutr. Rev. 26 (1968) 215–216. [7] O.G. Edholm, J.G. Fletcher, E.M. Widdowson, R.A. McCance, The energy expenditure and food intake of individual men, Br. J. Nutr. 9 (2007) 286. [8] O. Edholm, Energy balance in man studies carried out by the division of human physiology. National institute for medical research, J. Hum. Nutr. 31 (1977) 413. [9] P. Payne, J. Waterlow, Relative energy requirements for maintenance, growth, and physical activity, Lancet 298 (1971) 210–211. [10] J.E. Blundell, P. Caudwell, C. Gibbons, M. Hopkins, E. Naslund, N. King, et al., Role of resting metabolic rate and energy expenditure in hunger and appetite control: a new formulation, Dis. Model. Mech. 5 (2012) 608–613. [11] FAO/WHO, Energy and protein requirements: report of a joint FAO/WHO Ad Hoc expert committee, FAO Nutrition Meetings Report Series, No. 52, WHO Technical Report Series, 522 1973. [12] C. Henry, Estimates of metabolic adaptation in women living in developing countries: technical limitations, J. Biosoc. Sci. 24 (1992) 347–353. [13] L. Lissner, J.-P. Habicht, B.J. Strupp, D. Levitsky, J.D. Haas, D. Roe, Body composition and energy intake: do overweight women overeat and underreport? Am. J. Clin. Nutr. 49 (1989) 320–325.
77