Elsevier

Behavioural Processes

Volume 83, Issue 1, January 2010, Pages 23-30
Behavioural Processes

Percent body fat is related to delay and probability discounting for food in humans

https://doi.org/10.1016/j.beproc.2009.09.001Get rights and content

Abstract

This study describes delay and probability discounting patterns for hypothetical food and money in relation to percent body fat (PBF). Sixty university students completed four computerized discounting tasks in which they were asked to make a series of hypothetical decisions between (a) 10 dollars after one of several different delays (1, 2, 30, 180, and 365 days) or a smaller amount of money available immediately; (b) 10 bites of food after one of several delays (1, 2, 5, 10, and 20 h) or a smaller number of bites available immediately; (c) $10 at one of several probabilities (0.9, 0.75, 0.5, 0.25, 0.1) or a smaller amount of money to be received for sure; and (d) 10 bites of food at one of several probabilities (0.9, 0.75, 0.5, 0.25, 0.1) or a smaller number of bites to be received for sure. Median indifference points for all participants across each task were well described using the hyperbolic discounting function. Results suggest that percent body fat predicted discounting for hypothetical food, but not money, using regression analyses with the entire sample and when comparing individuals in the high and low quartiles for PBF. None of the other dietary variables (body mass index, subjective hunger, and time since last meal or snack) were related to discounting patterns. This suggests that individuals with high PBF may exhibit heightened sensitivities to delay and probability when making decisions about food.

Introduction

Delay discounting (DD) refers to the degree to which delay to an outcome reduces its value. DD corresponds to the behavioral definition of impulsiveness, which is the tendency to choose small, relatively immediate rewards over larger, more delayed rewards (Ainslie, 1975, Rachlin, 1995, Rachlin et al., 1991). In research concerning discounting-related choice patterns in humans, researchers pose a series of forced-choice options to participants in which they choose between a relatively small reward (e.g., $10) available immediately and a larger delayed reward (e.g., $100 in 1 day). Over the course of the choices, the smaller sooner amount is adjusted incrementally to identify the point at which the individual “switches” from choosing the larger, delayed amount to choosing the smaller, sooner amount. This value, termed the “indifference point” represents the current subjective value of the larger reward. The series of choices is presented repeatedly across several different delay periods (e.g., one week, one month, six months), yielding indifference point values that typically decrease as a function of the delays (larger delays yield smaller indifference point values). The pattern of these indifference point values can be described using a hyperbolic discounting function:V=A1+kDIn this equation, V represents the indifference point (or subjective value) of the delayed reward, A is the amount of the delayed reward, D is the delay of the reward, and k represents a free parameter that quantifies the rate of decay of the reward value as delay increases, or the relative degree of discounting (i.e., higher k values represent higher sensitivity to delay, or greater impulsivity).

Probability discounting (PD) refers to the degree to which the value of a reward decreases as the odds against receiving it increase (Rachlin et al., 1991). The probability discounting task is similar to that used in delay discounting studies, except that an individual makes decisions between a relatively small reward amount (e.g., $10 for sure) and a larger, but less probable reward amount (e.g., 25% chance at $100). In the task, the smaller certain amount is adjusted to determine indifference point values for the larger amount across several probabilities and the patterns of these indifference point values can be characterized with a hyperbolic discounting function:V=A1+hOHere V represents the subjective value (indifference point) of a probabilistic reward, A represents the amount of the larger probabilistic reward, O represents the odds against receiving the larger reward [(1/p)  1], and p represents the probability of receiving the large outcome. The free parameter h indexes the rate of discounting in which higher values represent a preference for more certain outcomes over less certain ones. Probability discounting appears to represent impulsive behavior in a manner similar to delay discounting (e.g., Green et al., 1999), but may also represent more of an index of risk (Green et al., 1999, Green and Myerson, 2004, Holt et al., 2003) since a choice is being made between two outcomes that differ with regard to risk of not receiving them, instead of simply delaying a reinforcer.

The delay discounting task has been used to address various socially-relevant health problem behaviors (see reviews by Critchfield and Kollins, 2001, Perry and Carroll, 2008, Reynolds, 2006). A large literature draws important connections between substance abuse, which has conceptual ties to impulse control problems, and choice patterns using discounting tasks. For example, discounting research consistently reports that heavy and problem drinkers (Vuchinich and Simpson, 1998), alcoholics (Odum and Rainaud, 2003, Petry, 2001), cigarette smokers (Bickel et al., 1999, Odum et al., 2002), and heroin addicts (Kirby et al., 1999, Madden et al., 1999, Madden et al., 1997) discount the value of delayed outcomes more than non-using comparison groups, suggesting that impulsive choice patterns are associated with a variety of substance abuse problems.

With humans, the vast majority of discounting research has been centered on choices with regard to money (see reviews Critchfield and Kollins, 2001, Perry and Carroll, 2008, Reynolds, 2006). Recently, researchers have begun applying discounting to answer questions about food-related outcomes in humans. The hyperbolic discounting function nicely describes and quantifies hypothetical and real food choices in humans (Epstein et al., 2009; Estle et al., 2007, Odum et al., 2006, Odum and Rainaud, 2003) and real food choices in non-humans (Chelonis and Logue, 1997, Green et al., 2004, Mazur, 2000, Mazur, 2007, Mitchell and Rosenthal, 2003; Ostaszewski et al., 2003; Perry et al., 2007, Saeki et al., 2002). Some of the human studies on food discounting have compared food discounting to other hypothetical outcomes and found that humans tend to discount the value of delayed hypothetical food rewards more steeply than, for example, hypothetical monetary rewards (Odum et al., 2006, Odum and Rainaud, 2003) and other secondary rewards (Charlton and Fantino, 2008), a phenomenon referred to as the domain effect. It has been suggested that because food is a primary reinforcer, it may be discounted more steeply than generalized or conditioned secondary reinforcers. However, when probability discounting is considered, food and money appear to be discounted similarly (Estle et al., 2007), though there are very few studies published on probabilistic discounting with food.

Approximately 30–40% of Americans are overweight and more than 20–30% are obese (Kohn and Booth, 2003, Sturm, 2003) and this prevalence has increased in the last three decades (Centers for Disease Control, 2006). A variety of environmental factors that have changed over the past 30 years are associated with this increase in obesity. Consider, for example, the increased prevalence of fast-food restaurants in recent decades which allows quick access to high-calorie, inexpensive meals (Powell et al., 2007). These food alternatives may compete with delayed, possibly healthier meals made at home. Ready access to inexpensive, high-calorie foods and beverages increases the chances of consuming unhealthy (but palatable) food alternatives (e.g., cheeseburgers) over healthier food alternatives (Burdette and Whitaker, 2004, Proctor et al., 2003). Indeed, the increased prevalence of obesity is linked to increased prevalence and patronage of fast-food restaurants (Powell et al., 2007).

One key to understanding the factors that make food more reinforcing for obese individuals may have to do with understanding choices related to the environment. For example, an obese individual may be especially sensitive to delays in outcomes (i.e., may be more likely to make impulsive choices). Individuals who have higher body mass indices, for example, perform more impulsively on the Iowa Gambling Task than those with lower body mass indices (Davis et al., 2004, Pignatti et al., 2006). Moreover, a vast literature describing impulsivity in obese individuals comes from self-report data (e.g., Chalmers et al., 1990, de Zwaan et al., 1994, Nasser et al., 2004) and these studies generally suggest that obese individuals endorse more impulsive statements than individuals who are not obese. Self-report measures of impulsivity are informative, but are vulnerable to various response biases and do not provide opportunities for manipulating environmental factors that may influence impulsive food choices. An understanding of the factors that influence impulsive decision-making is fundamental to obesity treatment and prevention efforts.

Weller et al. (2008) recently reported discounting patterns in relation to hypothetical money among obese and non-obese participants. They found that obese women had higher rates of discounting (i.e., were more impulsive) than were healthy-weight women, suggesting that impulsive decisions are associated with obesity. This effect was specific to women, as no discounting differences in obese and healthy-weight men were found. No research to date has examined food-related discounting decisions in relation to health-related measures to determine whether impulsivity is a general pattern specific to many types of outcomes, or whether there is something unique about food that is associated with steeper discounting. However, such efforts are important in light of the aforementioned research that discounting rates for consumable outcomes (e.g., food) often are different than those for monetary outcomes. Moreover, numerous discounting studies have found stimulus-specific discounting patterns in individuals with experiences with those outcomes—smokers discount cigarettes more than non-smokers (Field et al., 2006), heroin addicts discount heroin-related outcomes more than non-addicts (Madden et al., 1999), and consumers of erotica discount erotic stimuli more than non-consumers of erotica (Lawyer, 2008). Therefore, a clearer understanding of food-related decisions in relation to weight-related health factors may provide important information about the behavioral processes that underlie problematic dietary decisions.

This study attempted to extend the growing discounting literature by (1) describing food- and money-related decisions using delay and probability discounting procedures and analyses, and (2) relating those data to diet-associated factors.

Section snippets

Participants

Sixty participants (n = 43 female) were recruited from undergraduate psychology courses at Idaho State University and received course credit for their participation in the study. The average age of the participants was 23.5 (SD = 6.4) years old; 90.0% (n = 54) reported European-American ethnicity. Participants were not asked to refrain from eating or drinking prior to the experimental session.

Discounting tasks

Discounting choices were delivered via a PC-compatible computer using a modified version of an established

Analysis

Data were analysed using SPSS 14.0 statistical software. Individual and group (median) indifference point data were fit to Eq. (1) (for delay discounting) and Eq. (2) (for probability discounting) using non-linear regression. Two methods were used to estimate rate of discounting. The k and h parameters derived from Eqs. (1), (2) provided one measure of discounting rate, which is tied to the hyperbolic discounting function. (Data were also fit to the exponential model of discounting, but the

Discounting for food and money

Although there were relative differences in the frequency of non-systematic response patterns in the food versus money tasks, two χ2 analyses revealed no statistically significant commodity-based difference in the frequency of non-systematic response patterns when comparing delay discounting for money (n = 11) and delay discounting for food (n = 17; χ2 = 0.84; p = ns) and when comparing probability discounting for money (n = 9) and probability discounting for food (n = 17; χ2 = 1.57; p = ns). For purposes of

Discussion

In this study, delay discounting and probability discounting patterns for food and money were characterized in humans and related to two variables associated with physical health (body mass index and percent body fat) and with variables associated with current hunger status (current subjective hunger and hours since last meal and snack). Consistent with a growing literature, choice patterns for money were well described by the hyperbolic discounting function (Green et al., 1994, Green et al.,

References (66)

  • C.H. Lagorio et al.

    Delay discounting of real and hypothetical rewards III: steady-state assessments, forced-choice trials, and all real rewards

    Behavioural Processes

    (2005)
  • S.R. Lawyer

    Probability and delay discounting of erotic stimuli

    Behavioural Processes

    (2008)
  • G. Loewenstein

    Out of control: visceral influences on behavior

    Organizational Behavior and Human Decision Processes

    (1996)
  • J.E. Mazur

    Tradeoffs among delay, rate, and amount of reinforcement

    Behavioural Processes

    (2000)
  • S. Mitchell et al.

    Effects of multiple delayed rewards on delay discounting in an adjusting amount procedure

    Behavioural Processes

    (2003)
  • J.A. Nasser et al.

    Impulsivity and test meal intake in obese binge eating women

    Appetite

    (2004)
  • A.L. Odum et al.

    Discounting of delayed hypothetical money, alcohol, and food

    Behavioural Processes

    (2003)
  • A.L. Odum et al.

    Discounting of delayed hypothetical money and food: effects of amount

    Behavioural Processes

    (2006)
  • J.L. Perry et al.

    Impulsivity (delay discounting) for food and cocaine in male and female rats selectively bred for high and low saccharin intake

    Pharmacology, Biochemistry and Behavior

    (2007)
  • B. Reynolds et al.

    Measuring state changes in human delay discounting: an experimental discounting task

    Behavioural Processes

    (2004)
  • B. Reynolds et al.

    Acute alcohol effects on the Experiential Discounting Task and a question-based measure of delay discounting

    Pharmacology, Biochemistry and Behavior

    (2006)
  • R.E. Weller et al.

    Obese women show greater delay discounting than healthy-weighted women

    Appetite

    (2008)
  • G. Ainslie

    Specious reward: a behavioral theory of impulsiveness and impulse control

    Psychological Bulletin

    (1975)
  • W.K. Bickel et al.

    Impulsivity and cigarette smoking: delay discounting in current, never, and ex-smokers

    Psychopharmacology

    (1999)
  • Centers for Disease Control

    State-specific prevalence of obesity among United States, 2005

    Journal of the American Medical Association

    (2006)
  • D.K. Chalmers et al.

    Problem drinking and obesity: a comparison in personality patterns and life-style

    International Journal of the Addictions

    (1990)
  • T.S. Critchfield et al.

    Temporal discounting: basic research and the analysis of socially important behavior

    Journal of Applied Behavior Analysis

    (2001)
  • C. Davis et al.

    Decision-making deficits and overeating: a risk model for obesity

    Obesity Research

    (2004)
  • M. de Zwaan et al.

    Eating related and general psychopathology in obese females with binge eating disorder

    International Journal of Eating Disorders

    (1994)
  • S.J. Estle et al.

    Discounting of monetary and directly consumable rewards

    Psychological Science

    (2007)
  • M. Field et al.

    Delay discounting and the behavioural economics of cigarette purchases in smokers: the effects of nicotine deprivation

    Psychopharmacology

    (2006)
  • L.A. Giordano et al.

    Mild opioid deprivation increases the degree that opioid-dependent outpatients discount delayed heroin and money

    Psychopharmacology

    (2002)
  • L. Green et al.

    A discounting framework for choice with delayed and probabilistic rewards

    Psychological Bulletin

    (2004)
  • Cited by (216)

    View all citing articles on Scopus
    View full text