Elsevier

Appetite

Volume 133, 1 February 2019, Pages 47-60
Appetite

Drinking microstructure in humans: A proof of concept study of a novel drinkometer in healthy adults

https://doi.org/10.1016/j.appet.2018.08.012Get rights and content

Abstract

Microstructural analysis of ingestion provides valuable insight into the roles of chemosensory signals, nutritional content, postingestive events, and physiological state. Our aim was to develop a novel drinkometer for humans to measure detailed aspects of ingestion of an entire liquid meal or drinking session. The drinkometer records, in high definition (1 kHz), the weight of a fluid reservoir from which participants drink via a tube. An ultrasonic sensor measures the height of the fluid to derive density. Drinking speed over time can be displayed as a waveform. The smallest units of ingestion are sucks, which are organized in bursts. By applying probability density functions (PDF) on loge-transformed inter-suck intervals (ISI), an optimal burst-pause criterion (PC) can be identified. Information on ingestive volumes, rates, and durations can be then computed for the entire session, as well as for sucks and bursts. We performed a validation study on 12 healthy adults in overnight-fasted and in non-fasted states in 16 drinking sessions with 8 concentrations of sucrose (0–280 mM) presented in a blinded and random fashion. PDF determined PC = 2.9 s as optimal. Two-way RM-ANOVA revealed that total caloric intake during a drinking session depended on sucrose concentration (P < .001) and fasted state (P = .006); total drinking time (P < .001), total consumed volume (P = .003), number of sucks in total (P < .001), number of sucks per burst (P = .03), and burst duration (P = .02) were significantly influenced by fasting. In contrast, volume per suck (P = .002), suck speed (P < .001), and maximal speed per suck (P < .001) depended on sucrose concentration. We conclude that the novel drinkometer is able to detect differences in microstructural parameters of drinking behavior dependent on different motivational states, thus, adds to the technological toolbox used to explore human ingestive behavior.

Graphical abstract

Direct measurement of human ingestive microstructure in high definition with the new drinkometer.

Image 1
  1. Download : Download high-res image (229KB)
  2. Download : Download full-size image

Introduction

In humans, initial insights on food intake and appetite can be provided by indirect measures such as verbal report of energy intake, food diaries, and dietary recall questionnaires (Adolphus et al., 2017; Gero, Steinert, le Roux, & Bueter, 2017; Paul, Rhodes, Kramer, Baer, & Rumpler, 2005). However, indirect measures are vulnerable to inaccuracy and, at best, only offer an estimate of the target behavior (Archer, Hand, & Blair, 2013; Dhurandhar et al., 2015; Hill & Davies, 2001; Livingstone, 1995; Livingstone & Black, 2003; Wehling & Lusher, 2017). Such methodological limitations can be improved by complementing existing findings with direct measures of feeding and drinking. With some notable exceptions discussed in subsequent pages, most studies to date that have applied direct measures of ingestive behavior in humans have focused mainly on what kind of food and how much of it was consumed, rather than investigating how the food was consumed. This might be due to the fact that the assessment of the temporal organization of ingestive behavior within a meal in humans poses significant methodological and conceptual challenges to researchers and study design (Adolphus et al., 2017).

In rodent studies, however, the temporal analysis of ingestive behavior has been very effectively exploited for decades (Blundell, 1986; Davis, 1989; Davis & Smith, 1988; Davis, Smith, Singh, & McCann, 2001; Kissileff, 2000). Such assessments with solid food can be obtained by direct visual observation, measurements of the weight of the nutrient container at regular intervals, or by using computerized feeders that record meal intake (Paulus, Geyer, & Sternberg, 1998; Tolkamp & Kyriazakis, 1999). The temporal structure of fluid intake in rodents can be conveniently assessed by measuring licking with electrical contact circuits, as first done by Hill et al. (Hill & Stellar, 1951) or with photobeams (Gril, 1987; Spector AC, 2004; Spector & Smith, 1984; Weijnen, 1998). In these cases continuous recording of drinking over time can occur without disturbing the animal (Smith, 2000).

Rats and mice structure their feeding and drinking behavior in clusters of events separated by pauses of various lengths. Very long pauses on the order of minutes or hours define meals, whereas shorter pauses define bursts of ingestion within a meal (Davis, 1989; Spector, Klumpp, & Kaplan, 1998; Tolkamp et al., 2011). Commonly, liquid diets are used so that licking can serve as an effective dependent measure of ingestion across time. For example, within meals rats lick in bursts separated by pauses. The interval between licks (the so-called inter-lick interval, ILI) is relatively uniform and is thought to be controlled by a central pattern generator (Corbit & Luschei, 1969; Davis, 1996; Halpern, 1977; Johnson, 2018; Lin, Pierce, Light, & Hayar, 2013; Wiesenfeld, Halpern, & Tapper, 1977). Indeed, from a neural standpoint, the brain circuits that turn this pattern generator on and off are the way that the central nervous system effects intake regulation (Travers, 1991; Travers, Dinardo, & Karimnamazi, 1997). The quantitative nature of these bursts, their number, size, duration, and temporal distribution, comprise what is referred to as the microstructure of drinking (Davis & Smith, 1992; Davis, Smith, & Singh, 1999; Spector et al., 1998; Spector & St John, 1998). Any intervention that affects total liquid diet intake can be entirely viewed as a function of its influence on licking patterns, which, through microstructural analysis, then can provide relevant information on the distinct motivational modulators of ingestive behavior, such as orosensory input, postoral events, physiological state, and prior experience (Daniels, 2010; Davis, 1989, 1996; Mathes, Bohnenkamp, le Roux, & Spector, 2015; Smith, 2001; Spector et al., 1998; Spector & St John, 1998). Because, in rodents, burst size (the number of licks in a burst) appears to increase monotonically as the concentration of a sugar solution is raised, it is thought to reflect the palatability of the stimulus. In contrast, burst number is non-monotonically related to sugar concentration and is heavily influenced by deprivation state (Davis & Perez, 1993). These microstructural parameters can reveal the impact of a given drug or neural manipulation on the controls of ingestive behavior (Schneider, Davis, Watson, & Smith, 1990; Wirtshafter, Davis, & Stratford, 2011). For example, when water-deprived rats are presented with a moderate concentration of a quinine solution (bitter to humans), they strikingly decrease their burst size and increase their burst number; the taste appears to be driving them away from the water hole and the physiological state appears to be driving them toward it. If the lingual gustatory nerves are transected, however, the licking microstructure returns to that seen with water (Spector & St John, 1998).

Two other microstructural parameters that appear to be heavily influenced by the orosensory properties of the stimulus in rodents are the size of the very first burst and the initial lick rate (in first min). The interpretive power of these two parameters is provided by an example from Mathes et al. (Mathes, Letourneau, Blonde, le Roux, & Spector, 2016). After Roux-en-Y gastric bypass (RYGB), nondeprived rats drank similar amounts of a fat emulsion as did sham-operated animals on the first 60-min test, but subsequently decreased their intake across the next few days until it was roughly half of that consumed by the control group. One possible interpretation is that the palatability of the stimulus is decreasing, perhaps as a result of the postingestive consequences of the fat ingestion. However, the size of the first burst as well as the first min licking rate increased across the same days in both groups, which had virtually identical values. This outcome is inconsistent with RYGB causing any decrease in the positive motivational potency of orosensory properties of the stimulus and thus the basis for the decreased intake in the RYGB rats relative to their control group lies somewhere else.

Given the context provided by the literature briefly summarized above, the aim of this study was to develop an instrument that enables the high definition assessment of the full spectrum of microstructural parameters of liquid ingestion over an entire drinking session in humans. We herein describe a novel “drinkometer”, which fulfills this aim by continuous recording the weight of a fluid reservoir, while participants drink via a tube. Drinking speed is derived from changes in weight over time and individual sucks and bursts including their respective volumes, durations and rates, as well as the pauses between them can be quantified. In the present methodological paper we report the technical features of the device, and details of the data processing algorithm, as well as results of a pilot study in normal weight human subjects to demonstrate feasibility.

Section snippets

General principles of the novel drinkometer

Fig. 1 illustrates the basic technical components of the drinkometer during an experiment. The measurement of drinking microstructure in humans involves the following steps: 1) the density of the test drink is computed before the start of each test-session by an ultrasonic distance sensor and an electronic weight sensor; 2) a polyethylene tube is immersed in the fluid reservoir and is filled by vacuum aspiration with a syringe; 3) the participant drinks via the tube while the weight of the

Study characteristics

The baseline characteristics of participants and their averaged reported night-sleep hours prior to drinking sessions are presented in Table 1. In the pooled analysis of all sessions, mean ± SD temperature of the fluid reservoir was 9.16 ± 1.26 °C. Further, the mean of Δ beginning session – end of session temperature within sessions was 0.63 ± 0.4 °C, indicating a high temperature stability of the test liquids throughout the sessions.

Identification of an optimal burst pause criterion

Results of the PDF with Kernel estimation and Gaussian

Discussion

This study demonstrates that the drinkometer is suitable for the detection of differences in the microstructural parameters of ingestive behavior within and between human subjects who consume fluids of different sucrose concentrations in fasted and non-fasted states. Characterization of ingestive behavior in this precise and quantitative manner has the potential to reveal fundamental principles of the control of food intake behavior in different experimental settings. Microstructural analysis

Final remarks

As mentioned at the start, the early meal measures, which precede the significant accumulation of the stimulus in postoral body compartments, are thought to be influenced by the orosensory properties of the stimulus and, in the context of a given physiological state, reflect their influence on the motivation to ingest. Microstructural patterns occurring later in the meal appear to be more influenced by post-ingestive/post-absorptive events and reflect the onset of satiation (Blundell et al.,

Authors' contribution

Study design: MB, AS, DG, RES.

Development of the “drinkometer” device (hardware + data acquisition software): JJ.

Development of data-processing algorithm: BF.

Data acquisition: DG.

Data analysis: DG, BF, LF, AS, MB.

Interpretation of data: MB, AS, DG, RES.

Drafting the manuscript: DG, MB, JJ, FB, AS.

Critically revision and final approval of the manuscript: AS, MB, DG, JJ, FB, RES, LF.

Disclosures

The authors have no commercial associations that might be a conflict of interest in relation to this article. R.E. Steinert is employed by DSM Nutritional Products.

Acknowledgments

We are grateful to all of the volunteers for having participated in this study. We acknowledge Daniel Kaufmann, MSc, Dominik Brügger, MSc, Richard Monteil, BSc and Jan Segessenmann, BSc from the Human-Centered Engineering Institute of Applied Sciences, Biel, Switzerland, for having contributed to the development of the drinkometer. We also thank Marcel André Schneider, MD for his help in statistical data-analysis in R software and Nikolaus Wick-Mazzarda for the photography.

References (65)

  • A.W. Johnson

    Characterizing ingestive behavior through licking microstructure: Underlying neurobiology and its use in the study of obesity in animal models

    International Journal of Developmental Neuroscience

    (2018)
  • H.R. Kissileff

    Ingestive behavior microstructure, basic mechanisms and clinical applications

    Neuroscience & Biobehavioral Reviews

    (2000)
  • H.R. Kissileff et al.

    Progressive ratio (PR) schedules and the sipometer: Do they measure wanting, liking, and/or reward? A tribute to Anthony Sclafani and Karen Ackroff

    Appetite

    (2018)
  • H.R. Kissileff et al.

    A quadratic equation adequately describes the cumulative food intake curve in man

    Appetite

    (1982)
  • A.M. Lett et al.

    Emulsion oil droplet size significantly affects satiety: A pre-ingestive approach

    Appetite

    (2016)
  • M.B. Livingstone et al.

    Markers of the validity of reported energy intake

    Journal of Nutrition

    (2003)
  • M.P. Paulus et al.

    Differential movement patterns but not amount of activity in unconditioned motor behavior of Fischer, Lewis, and Sprague-Dawley rats

    Physiology & Behavior

    (1998)
  • T. Rudholm et al.

    Release of regulatory gut peptides somatostatin, neurotensin and vasoactive intestinal peptide by acid and hyperosmolal solutions in the intestine in conscious rats

    Regulatory Peptides

    (2009)
  • A.A. Santa Cecilia Silva et al.

    The association between anxiety, hunger, the enjoyment of eating foods and the satiety after food intake in individuals working a night shift compared with after taking a nocturnal sleep: A prospective and observational study

    Appetite

    (2017)
  • L.H. Schneider et al.

    Similar effect of raclopride and reduced sucrose concentration on the microstructure of sucrose sham feeding

    European Journal of Pharmacology

    (1990)
  • J.C. Smith

    Microstructure of the rat's intake of food, sucrose and saccharin in 24-hour tests

    Neuroscience & Biobehavioral Reviews

    (2000)
  • G.P. Smith

    John Davis and the meanings of licking

    Appetite

    (2001)
  • A.C. Spector et al.

    A detailed analysis of sucrose drinking in the rat

    Physiology & Behavior

    (1984)
  • B.J. Tolkamp et al.

    To split behaviour into bouts, log-transform the intervals

    Animal Behaviour

    (1999)
  • J.B. Travers et al.

    Motor and premotor mechanisms of licking

    Neuroscience & Biobehavioral Reviews

    (1997)
  • J.A. Weijnen

    Licking behavior in the rat: Measurement and situational control of licking frequency

    Neuroscience & Biobehavioral Reviews

    (1998)
  • M.S. Westerterp-Plantenga

    Eating behavior in humans, characterized by cumulative food intake curves--a review

    Neuroscience & Biobehavioral Reviews

    (2000)
  • E. Almiron-Roig et al.

    Factors that determine energy compensation: A systematic review of preload studies

    Nutrition Reviews

    (2013)
  • E. Archer et al.

    Validity of U.S. nutritional surveillance:National Health and Nutrition Examination Survey caloric energy intake data, 1971-2010

    PLoS One

    (2013)
  • Blundell et al.

    Appetite control: Methodological aspects of the evaluation of foods

    Obesity Reviews

    (2010)
  • J. van den Boer et al.

    The availability of slow and fast calories in the Dutch diet: The current situation and opportunities for interventions

    Foods

    (2017)
  • J.D. Corbit et al.

    Invariance of the rat's rate of drinking

    Journal of Comparative & Physiological Psychology

    (1969)
  • Cited by (0)

    Communication: The study was presented as oral presentation at the 26th Annual Meeting of the Society for the Study of Ingestive Behavior in Bonita Springs, Florida, USA, on 20 July 2018

    View full text