ReviewNeurobehavioural correlates of body mass index and eating behaviours in adults: A systematic review
Highlights
► Individual neurobehavioral differences may predispose to obesity. ► The evidence relating specific behavioral measures and eating or BMI is reviewed. ► Certain executive function and food motivation tasks are reliable and predictive. ► Many relevant questionnaire measures tap similar personality factors. ► We identify measures indexing key behavioral differences predisposing to higher BMI.
Introduction
Health worsens as body-mass index (BMI, weight in kg/height in m2) increases (James, 2008), and throughout the world BMI continues to rise (Finucane et al., 2011). This alarming increase is likely due to many interacting factors, ranging from neurobiological mechanisms regulating our behaviour (Speliotes et al., 2010) to public policy, agricultural innovation, and business practices that have significantly lowered the cost and increased the availability of calorie dense food (Chandon and Wansink, 2010, Drewnowski, 2009, Lakdawalla and Philipson, 2002).
While the interaction between individual tendencies and a rapidly changing food environment seems to be critical in the increasing prevalence of obesity (Levitsky, 2005), not every individual is equally susceptible to these environmental pressures. How do individual differences in the ability to regulate food choices protect against weight gain in the modern environment of cheaper food and increased food consumption? While a variety of biological individual differences can conceivably be at play, those that relate to behaviour are likely to be of high interest and high impact: indeed, many of the current interventions to address the obesity epidemic are aimed at changing individual behaviour, notably through education and public health messages that exhort healthy choices and self-control.
We propose that a better understanding of the regulation of food choice and eating behaviour is crucial to explaining existing variability in BMI and increases in BMI, and may also be helpful in developing rational, tailored interventions to prevent or reverse weight gain, or at least in predicting who might benefit from a given intervention, whether that intervention relies on pharmacological, educational or social mechanisms. We argue that a brain-based view of these behaviours will allow mechanistic links between the growing body of knowledge about genetic and other biological determinants of BMI and the individual behaviours that lead to weight gain. There are several methods available to study the brain mechanisms underlying eating behaviours in humans. The tools of cognitive neuroscience are now being brought to bear on this question, with provocative results emerging from functional neuroimaging (Batterink et al., 2010, Killgore and Yurgelun-Todd, 2005, Martin et al., 2010, Stice et al., 2008, Stice et al., 2010, Stice et al., 2011a, Wang et al., 2001), electrophysiology (Nijs et al., 2010b, Silva et al., 2002), non-invasive brain stimulation (Camus et al., 2009, Fregni et al., 2008, Uher et al., 2005) hormonal manipulations (Batterham et al., 2007, Farooqi et al., 2007, Malik et al., 2008), and genetics (Stice et al., 2011b). While these approaches are useful for understanding the neural basis of food choice and other eating related behaviours in tightly focused experiments, they are unwieldy for use on the scale of the population level studies that are increasingly seen as necessary to fully understand the multivariate, multi-level determinants of the complex problem of obesity (Dubé et al., 2008).
Neurobehavioural measures offer a potentially valuable intermediate tool: Such measures quantify a particular behaviour (i.e. psychological construct) in a way that can be linked to the brain, and are feasible for large-scale studies. There are two main types of neurobehavioural measures: neurocognitive measures, which are tasks, many with their origins in neuropsychology, that aim to measure specific cognitive-behavioural abilities, and personality questionnaires or scales that capture participants’ typical behaviour through (mainly self-reported) responses to behaviour-related questions. Commonly applied behavioural constructs in research on eating include self-control, impulsivity, executive control and sensitivity to reward, among others. Cognitive neuroscience research has begun to identify the neurobiological substrates of these constructs in general, and, more helpfully for our purposes, of specific measures of these constructs. As an example, self-control can be indexed by both a neurocognitive stop-signal task and a questionnaire measure of Conscientiousness. Both of these measures have been linked to maladaptive eating behaviours (Bogg and Roberts, 2004, Nederkoorn et al., 2010), and have been related to prefrontal structures (Aron and Poldrack, 2006, DeYoung et al., 2010). Thus, one or both of these measures, suitable for use in large-scale studies, might shed light on the role of prefrontal cortex in eating behaviours.
This paper aims to systematically review current knowledge regarding neurobehavioural measures in relation to obesity and eating behaviours. We set out to answer a practical question: Is there sufficient evidence to allow the confident selection of neurobehavioural measures, whether personality questionnaires or neurocognitive tasks, to characterize individual differences in BMI or ecologically-relevant eating behaviours in humans? Appropriate neurobehavioural measures must be both ecologically and conceptually valid: that is, (1) there must be a valid link between the measure and BMI or eating behaviours, and (2) the measure must be reliable i.e., reproducible and accurate. The existing literature will be reviewed in regards to these points, as well as considering whether the measures can be related to specific brain systems.
Assuming suitable neurobehavioural measures are available, what eating behaviours should they be expected to predict? The most common correlative research designs involve measuring BMI concurrent with the administration of neurobehavioural measures and then comparing performance either between different weight groups or along a continuous BMI scale (e.g., Davis and Fox, 2008, Gunstad et al., 2007). Other related indexes of healthy weight have been used, such as waist-to-hip ratio or high waist circumference. Prospective studies involve measuring BMI at two time points and determining if a neurobehavioural measure is able to predict the change (e.g., Gunstad et al., 2010, Sutin et al., 2011). However, what BMI and other similar measures gain in public health relevance, convenience and reliability, they lose in behavioural specificity: Changes in BMI are the result of various factors accumulating over a period of time (Blundell and Cooling, 2000). Relevant neurobiological factors may not be discernible amongst the many other contributing variables. On the other hand, if a neurobiological factor is identified as relevant to a ‘big picture’ real world outcome such as BMI, particularly in a prospective longitudinal study, this would seem good evidence that it may be a high yield point of leverage in predicting risk or personalizing interventions.
Beyond BMI, a variety of more specific eating-related behaviours have also been examined. These offer more behavioural focus, making it more likely that a mechanistic link will be identified between neurobehavioural tasks and eating behaviours. On the other hand, their relevance to real world outcomes is uncertain. Examples include asking participants how much they plan to eat/avoid certain food products over a given time period and then measuring any discrepancy with actual recorded behaviour (planned-conducted behaviour; e.g., Hall et al., 2008; based on Thompson et al., 2004) or measuring how much participants eat in a particular context, such as in a bogus food tasting test (laboratory intake of food; e.g., Herman and Mack, 1975, Schachter et al., 1968). While more specific than BMI, these measures have not been positioned within a brain-based conceptual framework. In addition, studies using these approaches tend to use idiosyncratic tasks, making cross-study comparison more difficult.
Similarly to eating behaviours, there is great variety in the neurobehavioural measures that have been studied as correlates or predictors of eating behaviours. While constructs such as impulsivity or executive functions are considered important in obesity (e.g., Guerrieri et al., 2008, Smith et al., 2011), these are such broad categories that they provide little insight into the underlying mechanisms. More specific constructs would seem more informative, but the lack of an accepted, common set of measures and lack of communication between different research traditions has hindered progress. The nature of fragmentation is somewhat different in the two main types of neurobehavioural measures, and thus they will be addressed in different sections of this review.
Neurocognitive tasks rarely correlate with each other which makes drawing conclusions about more specific constructs difficult. For instance, Hofmann et al. (2009a) found that stop-signal, affective inhibition and working memory tasks all relate to candy consumption but do not relate to each other. In a similar vein, neurocognitive self-control measures in general have an average correlation of 0.15 (Duckworth and Kern, 2011). A possible way to overcome this issue is to apply a domain-based approach, where tasks meant to capture a given construct, such as memory, language, or executive function, are clustered together to evaluate the domain's feasibility in obesity research. Smith et al. (2011) applied a similar approach to highlight the general role of executive functions in obesity. Here, we attempt to evaluate all cognitive domains that have been tested in the context of maladaptive eating behaviours.
We conducted a systematic search combining terms referring to related neuropsychological domains and to BMI or laboratory measures of food intake. We then categorized the tasks by the primary cognitive domain that each is purported to measure. Whenever necessary, the tasks within a cognitive domain were further classified based on existing, empirically supported conceptual frameworks (Miyake et al., 2000, Oberauer et al., 2000) and neuropsychological expertise. This classification allowed identification of the domains most related to eating behaviours, enabling preliminary conclusions about domains where no single task has been frequently studied.
In contrast to the heterogeneous variety of neurocognitive tests, questionnaire measures are more homogenous. The focus has been on a smaller number of constructs that are perceived to be relevant in obesity: Five-Factor Model personality dimensions, self-control, sensitivity to reward, impulsivity, and a handful of food-related personality constructs. Most of these constructs have established relationships with obesity and maladaptive eating behaviours (Bogg and Roberts, 2004, Bryant et al., 2008, Chalmers et al., 1990, de Ridder et al., 2011, Guerrieri et al., 2008, Herman and Polivy, 2008, Johnson et al., 2011, Lowe and Thomas, 2009, Macht, 2008; see later in this paper). While often each of these constructs has several measures, different measures of a single construct tend to correlate well (self-control questionnaires’ average r = 0.50, Duckworth and Kern, 2011), which has enabled fruitful attempts to clarify which measures within a construct provide the best reliability and validity. This paper does not seek to double this work–rather we highlight the best measures identified so far in relation to BMI or eating behaviours.
Around a dozen personality constructs related to obesity and eating behaviours are actively applied in contemporary research. At this point one might ask: How different are these measures from each other? It seems highly unlikely that obesity would be so multifaceted that each of the highlighted constructs would represent an independent mechanism. Rather, it is quite probable that different measures ultimately rely on a common set of underlying processes that are named differently in different research traditions. For example, a recent review of self-control questionnaires showed that various self-control measures correlate with Neuroticism and Conscientiousness of the Five-Factor Model (McCrae and Löckenhoff, 2010). A similar analysis was conducted here to explore the possible overlap between measures deemed important for obesity. The Five-Factor Model (McCrae and Costa, 1987) was used as the baseline measure given its characterisation of personality as a whole, its established relationship with obesity (e.g., Sutin et al., 2011) and other eating-related behaviours (e.g., Mõttus et al., 2011, Mõttus et al., 2012), and the fact that most of the measures mentioned above have been correlated at least once with the Five-Factor Model.
Reliability is an integral part of a measure's validity, as it sets the upper limit to the potential correlation between a measure and an outcome. Despite its obvious importance, reliability has been of little concern for neurocognitive research; personality questionnaires have done much better in this regard. In this review, reliability will be reported for key measures.
Finally, neurocognitive measures and personality questionnaires capturing the same construct are still often considered as separate entities. For instance, in the domain of impulsivity recent reviews tend to focus on evidence from one of the approaches and are critical of the other, with neurocognitive research claiming that personality questionnaires are vulnerable to subjective bias (de Wit, 2009) and personality research claiming that neurocognitive research has problems with reliability (DeYoung, 2010a). While both of these criticisms are valid, both research traditions are taking steps to refine their measures (e.g., Parrott, 1991, Soto et al., 2008). Thus, evidence from sound measures from both measurement traditions is likely to offer relevant perspectives on the neurobiological underpinnings of a particular behaviour. Note that, while neurocognitive tests are often favoured in neurobehavioural research, the links between personality constructs and the brain are also becoming clearer (for overviews see Carnell et al., 2011, Davis and Panksepp, 2011, DeYoung and Gray, 2009, DeYoung, 2010b). Thus, obesity research is likely to benefit from evidence from both neurocognitive measures and personality questionnaires, perhaps by developing a more elaborated neurobiological framework through which to link neurobehavioural measures and outcomes of interest (e.g., Berthoud and Morrison, 2008, Carnell et al., 2011).
In sum, the current review seeks to unite the scattered knowledge from studies using both broad and narrow approaches to study the relationship between neurobehavioural measures and BMI or eating behaviours. The review will serve two goals: First, it will serve as a guide for researchers choosing measures for a comprehensive, reliable and informative test battery with the potential to predict obesity risk, suitable for use in adults. Second, we will review what these measures may reveal about the putative brain mechanisms contributing to obesity.
Section snippets
Search strategy
On 18.11.11 a topic search in all databases was conducted at ISI Web of Knowledge pairing neurocognitive or psychological constructs with obesity and food words. The search included articles dating from 1985 to 2011 (see supplementary material for a list of keywords). The search was refined to exclude studies involving animals, children (<18 years) and the elderly (>60 years), and also to exclude work not related to psychology or obesity, resulting in total of 7069 papers. Based on titles and
Personality questionnaires
Personality research in relation to eating behaviours can be divided into three major approaches: relating various eating behaviours with personality scales that capture (a) general, (b) specific, or (c) eating-related aspects of human personality. The use of general personality scales provides a framework for understanding eating behaviour in a wider context, as this enables comparing personality profiles related to eating behaviours to personality profiles related to other behaviours of
Brain mechanisms
Several brain-based models of eating behaviour have been proposed, which vary in their level of detail and main focus (e.g., Berthoud and Morrison, 2008, Carnell et al., 2011). Most suggest at least three central mechanisms in the control of eating: (1) a hypothalamic system sensitive to homeostatic signals, through which the organism matches food intake with energy requirements, (2) a striatal and limbic emotion/memory system sensitive to current and past reward experiences, through which the
General discussion
This review has summarised the current literature relating BMI, change in BMI or eating behaviours with neurobehavioural measures in otherwise healthy adults. Amongst neurocognitive measures, those sensitive to executive function and food motivation provide the most robust and reliable associations. Of the 66 tasks reviewed, only a few dependent measures from particular tasks both provide consistent results and have reliable psychometrics. The most robust and reliable executive function tasks
Conflict of interest
This study was supported by the Social Sciences and Humanities Research Council of Canada and the Canadian Foundation for Innovation. The authors have no conflicts of interest to disclose.
Acknowledgements
The authors would like to thank Ami Tsuchida for help with categorising neurocognitive tasks, Matthias Doucerain and Margit Kõiv for their help with designing Fig. 1, and Jüri Allik and two anonymous reviewers for extremely for helpful comments on earlier versions of this manuscript.
References (218)
- et al.
Unintentional eating. What determines goal-incongruent chocolate consumption?
Appetite
(2010) - et al.
Body mass correlates inversely with inhibitory control in response to food among adolescent girls: An fMRI study
Neuroimage
(2010) - et al.
Reliability of verbal–numerical working memory tasks
Personality and Individual Differences
(2007) - et al.
Neuropsychological performance of a clinical sample of extremely obese individuals
Archives of Clinical Neuropsychology
(2008) - et al.
Time discounting and the body mass index: Evidence from the Netherlands
Economics and Human Biology
(2006) - et al.
NEO personality domains and gender predict levels and trends in body mass index over 14 years during midlife
Journal of Research in Personality
(2006) - et al.
The measurement of individual differences in behavioural inhibition and behavioural activation systems: A comparison of personality scales
Personality and Individual Differences
(2003) - et al.
A psychometric evaluation of behavioral inhibition and approach self-report measures
Personality and Individual Differences
(2006) - et al.
Cognitive processing differences between obese and nonobese subjects
Addictive Behaviors
(1985) - et al.
Neuroticism modulates amygdala–prefrontal connectivity in response to negative emotional facial expressions
Neuroimage
(2010)
Altered executive function in obesity. Exploration of the role of affective states on cognitive abilities
Appetite
Implicit and explicit attitudes to high-and low-calorie food in females with different BMI status
Eating Behaviors
Functional brain imaging of appetite
Trends in Endocrinology and Metabolism
Sensitivity to reward and body mass index (BMI): Evidence for a non-linear relationship
Appetite
Immediate pleasures and future consequences. A neuropsychological study of binge eating and obesity
Appetite
From motivation to behaviour: A model of reward sensitivity, overeating, and food preferences in the risk profile for obesity
Appetite
Sensitivity to reward: Implications for overeating and overweight
Appetite
A meta-analysis of the convergent validity of self-control measures
Journal of Research in Personality
Personality traits and eating behavior in the obese: poor self-control in emotional and external eating but personality assets in restrained eating
Eating Behaviours
Food reinforcement, energy intake, and macronutrient choice
American Journal of Clinical Nutrition
A questionnaire approach to measuring the relative reinforcing efficacy of snack foods
Eating Behaviors
Characteristics of psychomotor performance and time cognition in moderately obese patients
Physiology & Behavior
National, regional, and global trends in body-mass index since 1980: Systematic analysis of health examination surveys and epidemiological studies with 960 country-years and 9.1 million participants
Lancet
Individual differences in reward sensitivity are related to food craving and relative body weight in healthy women
Appetite
Transcranial direct current stimulation of the prefrontal cortex modulates the desire for specific foods
Appetite
The international personality item pool and the future of public-domain personality measures
Journal of Research in Personality
Personality traits and eating habits: The assessment of food preferences in a large community sample
Personality and Individual Differences
Validation of a questionnaire measure of the relative reinforcing value of food
Eating Behaviors
Body mass index moderates gaze orienting biases and pupil diameter to high and low calorie food images
Appetite
How impulsiveness and variety influence food intake in a sample of healthy women
Appetite
The influence of trait and induced state impulsivity on food intake in normal-weight healthy women
Appetite
Elevated body mass index is associated with executive dysfunction in otherwise healthy adults
Comprehensive Psychiatry
External cues in the control of food intake in humans: the sensory-normative distinction
Physiology & Behavior
Three ways to resist temptation: The independent contributions of executive attention, inhibitory control, and affect regulation to the impulse control of eating behavior
Journal of Experimental Social Psychology
Overcoming the urge to splurge: Influencing eating behavior by manipulating inhibitory control
Journal of Behavior Therapy and Experimental Psychiatry
Personality is reflected in the brain's intrinsic functional architecture
PLoS ONE
Multiple Regression: Testing and Interpreting Interactions
Missed by an inch or a mile? Predicting the size of intention–behaviour gap from measures of executive control
Psychology & Health
Neurobehavioral inhibition of reward-driven feeding: Implications for dieting and obesity
Obesity
Inhibiting food reward: Delay discounting, food reward sensitivity, and palatable food intake in overweight and obese women
Obesity
Cortical and subcortical contributions to stop signal response inhibition: Role of the subthalamic nucleus
Journal of Neuroscience
Emotional eating moderates the relationship between implicit measures of attitudes and chocolate consumption
European Journal of Personality
The development of implicit attitudes. Evidence of race evaluations from ages 6 and 10 and adulthood
Psychological Science
PYY modulation of cortical and hypothalamic brain areas predicts feeding behaviour in humans
Nature
Individual differences in reward drive predict neural responses to images of food
Journal of Neuroscience
The development of selective inhibitory control across the life span
Developmental Neuropsychology
The brain, appetite, and obesity
Annual Review of Psychology
Routes to obesity: phenotypes, food choices and activity
British Journal of Nutrition
Conscientiousness and health-related behaviors: A meta-analysis of the leading behavioral contributors to mortality
Psychological Bulletin
Female emotional eaters show abnormalities in consummatory and anticipatory food reward: A functional magnetic resonance imaging study
International Journal of Eating Disorders
Cited by (183)
Acute high-intensity interval exercise improves food-related cognition in young adults with obesity: An ERP study
2024, International Journal of Clinical and Health PsychologyObesity and cognitive impairment in the adult population: A case-control study
2022, Obesity MedicinePopulation-based research in obesity – An overview of neuroimaging studies using big data approach
2022, Current Opinion in Endocrine and Metabolic Research