Elsevier

Appetite

Volume 107, 1 December 2016, Pages 171-179
Appetite

Explicit and implicit approach vs. avoidance tendencies towards high vs. low calorie food cues in patients with anorexia nervosa and healthy controls

https://doi.org/10.1016/j.appet.2016.08.001Get rights and content

Abstract

Patients with anorexia nervosa (AN) have a strong ability to limit food intake. Thus, dysfunctional approach vs. avoidance behaviors towards food are evident in AN.

We applied an approach-avoidance task (AAT), in which n = 41 AN patients and n = 42 controls either approached (“pull”) or avoided (“push”) high (HC) vs. low calorie (LC) food pictures based solely on the presented picture format (landscape vs. portrait). We tested the hypothesis that -in opposition to controls displaying an approach bias towards HC food cues- AN patients would show an avoidance bias (measured as different response times) towards HC food. Explicit ratings of food cues were also performed.

We found a significant interaction “group” x “direction” (p = 0.03). rm-ANOVAs performed for each of the two groups separately showed a main effect for “direction” of motion in controls (p = 0.02), but not in AN patients (p = 0.40). The two groups did not differ in their reaction times (RTs) with regard to “push” (p = 0.27). However, RTs with regard to “pull” were significantly different between the groups (p = 0.04).

Controls show a clear approach bias, expressed by significantly faster RTs for “pull” compared to “push”, independent of “calorie” content of the food stimuli. This approach bias is absent in the group of AN patients. This is indicative of a global loss of incentive value of food in AN. Implicit trainings as add-on to psychotherapy in AN patients are asked for.

Introduction

Patients suffering from anorexia nervosa (AN) have an extraordinary strong ability to limit food intake. They display a high degree of self-control that helps them keep a dieting goal even in the presence of appetite or a strong desire to eat (Seibt, Häfner, & Deutsch, 2007). Contrary to that, patients suffering from bulimia nervosa or binge-eating disorder are characterized by frequent disruptions in inhibitory control leading to food binges. In these cases, self-regulation abilities are hampered in situations of high-intensity emotion (e.g. great sadness or anger) or when experienced impulses are particularly strong (e.g. due to the availability of high palatable and rapidly available food) and disable counter-regulatory mechanisms (Heatherton & Wagner, 2011).

The principle of reward and punishment is the basis of goal-oriented behavior; an outcome has to be rewarding or be perceived as such, in order to actually motivate behavior. It is the striving for reward and/or escape from punishment that regulates behaviors of approach or avoidance. Affective cues trigger compatible behaviors; aversive cues trigger avoidance and positive cues trigger approach. A successful regulation both on an explicit and implicit level implies always the approach of reward and avoidance of punishment. On the psychological level, the mechanisms of food-intake stand under the influence of both conscious (explicit) and automatic (implicit) regulatory processes. Bargh (1994)'s automotive theory of non-conscious goal pursuit postulates that specific stimuli are able to induce certain approach or avoidance behaviors automatically and in the absence of consciousness, provided that a strong interconnection between the stimulus and the target behavior has occurred in the past (e.g. high calorie food equals threat and must therefore be avoided). Approach-avoidance behaviors contain all responses ranging from those that are entirely automatic to those being entirely under control and a clear distinction is not easy. According to Corr (2010), behavioral responses are not subject of conscious reflection as long as things go according to plan. Conscious attentional processes are involved in cases of goal-conflicts and perceived discrepancies between what is and what should be.

Food is a rewarding stimulus, as it activates central structures of the brain's reward system like the striatum. The visual presentation of food stimuli vs. pictures of neutral objects (e.g. pieces of furniture) distinctly activates brain structures involved in reward (DelParigi et al., 2007, Passamonti et al., 2009). Such activation patterns are significantly more pronounced for high calorie compared to low-calorie food (Schur et al. 2009). The response to reward is underpinned by the incentive valence of cues that is often referred to as “wanting”. Usually, the higher the incentive valence of a certain food is, the more pronounced is the approach bias towards it (Brignell, Griffiths, Bradley, & Mogg, 2009) and the higher is the likelihood of its consumption (Hofmann et al., 2008a, Hofmann et al., 2008b, Nederkoorn et al., 2010).

Only approach and avoidance behaviors to which the individual has conscious access can be assessed by means of clinical interviews. Likewise, subjective ratings in questionnaires may mirror explicit motives, but they are liable to a number of influencing factors (e.g. social desirability) and are therefore not capable of reliably assessing altered (e.g. desensitised) motivational placements towards food. Self-reports that are supplemented by indirect measures assessing automatic, uncontrolled implicit processes are therefore at an advantage (Fazio & Olson, 2003). Reaction times e.g. constitute such indirect (implicit) measures and are therefore necessary add-ons to an explicit approach. Systematic differences in reaction times during experimental measures of implicit preferences may for instance allow for inferences upon cognitions and attitudes that have an influence on behaviors, but are not accessible to the conscious self. This kind of associative/implicit cognitive processes might be relevant for the onset and perpetuation of eating disorders.

In the so-called motor-process hypothesis (Cacioppo, Priester, & Berntson, 1993) the interrelation between approach, avoidance and motor reactions is pointed out. Thus, flexion and extension of the forearm may evoke approach or avoidance motivation, respectively. Based on this, the approach-avoidance task (AAT) has been developed as an experimental paradigm that tests associative processes that are not influenced by attention, memory, interpretation or strategic control (De Houwer, 2006). The AAT is a behavioral reaction time task (Rinck & Becker, 2007) that assesses approach and avoidance motivational processes by requiring participants to respond to cues (pictures) by either “pulling” these towards themselves or “pushing” them away, while responding to an irrelevant feature of the stimulus (e.g. the presented picture format) and not according to the picture content itself. The amount of time required to execute these actions is the dependent variable.

The rationale behind the AAT is that emotionally loaden stimuli evoke spontaneous (automatic, implicit) behavioral responses: unpleasant (negatively connoted) stimuli cause spontaneous avoidance, while pleasant (positively connoted) cues cause spontaneous approach. Such responses may be captured by means of motor movements of the forearm; unpleasant cues are pushed away from oneself faster; pleasant cues are pulled faster -interfering with the explicit instruction to respond to an irrelevant cue feature. Thus, a close connection between the valence of stimuli and the herewith associated motor reaction exists. The AAT uses this connection, in order to assess the association between stimuli and behavioral responses. The different combination of cue valence and direction of the motor reaction leads to compatible (pleasant/pull and unpleasant/push) and incompatible (pleasant/push and unpleasant/pull) combinations. It is expected that compatible combinations are reflected in faster response times than incompatible combinations.

The literature attests validity in measuring approach-avoidance motivational processes to the AAT (e.g. Wiers et al., 2010, Wiers et al., 2011). In the context of anxiety and addiction research it has been shown that positively connoted cues were associated with an approach and negative cues with an avoidance bias (Wiers, Gladwin, & Rinck, 2013). Patients addicted to alcohol displayed an approach bias for alcohol-associated cues (faster response times in the “pull” compared to the “push” condition), whereas patients suffering from arachnophobia had an avoidance bias (faster response times in the “push” compared to the “pull” condition) for spider stimuli (Rinck and Becker, 2007, Wiers et al., 2014). In a study in social phobia, patients were instructed to push or pull pictures of laughing, angry or neutral faces. As expected, there was an avoidance bias for angry faces. However, there was also an implicit avoidance bias for laughing faces, although laughing faces were explicitly evaluated as being pleasant. This result points towards an automatic avoidance process that was not in accordance with the explicit evaluation (Heuer, Rinck, & Becker, 2007).

Approach and avoidance tendencies in clinical cohorts with anorexia nervosa have been investigated before. Both Veenstra and de Jong (2011) and Neimeijer, de Jong, and Roefs (2015) examined adolescent AN patients by means of the affective Simon task, an implicit paradigm in which an animated manikin is moved towards or away from a picture shown on screen. Veenstra and de Jong (2011) found that patients showed less implicit motivational orientation towards food than healthy controls. Neimeijer et al. (2015) found an approach bias for low calorie, but not for high calorie food. In a non-clinical cohort, Brockmeyer, Hahn, Reetz, Schmidt, and Friederich (2015) applied a food approach-avoidance task and could show that high food cravers displayed a stronger automatic approach bias towards food than low food cravers. Thus, the present study is the first to implement an approach-avoidance task for the assessment of implicit bias towards food in a sample of AN patients and healthy controls. In order to investigate approach and avoidance in eating disorders using food stimuli, we applied the approach-avoidance task (AAT), in which AN patients were instructed to either approach or avoid high vs. low calorie food pictures based solely on the presented picture format (landscape vs. portrait). By doing so, we tested the hypothesis that an approach bias for high calorie stimuli would be found in healthy controls, while patients would display an avoidance bias (measured as different response times). With regard to low calorie stimuli, no differences in response times between patients and controls were expected. Explicit measures by means of self-ratings of the presented cues were also carried out.

Section snippets

Participants

N = 41 AN inpatients recruited at the Department of Psychosomatic Medicine and Psychotherapy (University Hospital Erlangen and Schön Klinik Bad Staffelstein, Germany) were included in the present study. The patients fulfilled the DSM-V criteria of AN (APA, 2013). Furthermore, n = 42 healthy control participants were included; the group of healthy controls consisted of medical students at the University Clinic of Erlangen, Germany. Healthy controls were deemed healthy according to their BMI and

Results

Anthropometric variables (age, BMI etc.), scores in the self-report questionnaires and subjective ratings of food pictures are reported as mean ± standard deviation. Response times (RT) are reported as median ± standard deviation. Median response times were included in the analysis, due to lower sensitivity to outliers compared with mean scores (Cousijn et al., 2011, Rinck and Becker, 2007, Wiers et al., 2009, Wiers et al., 2010).

Discussion

The AAT has been implemented for the first time within the context of eating disorders in the present investigation. However, while previous studies have applied disorder-specific cues (e.g. alcohol or spiders) against neutral cues (e.g. non-alcoholic beverages or neutral objects), in the present investigation, a more subtle cue differentiation into high calorie and low calorie food pictures has been tested. Imaging studies have shown that high and low calorie food cues are differently

Conflicts of interest

None.

Source of funding

This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 677804).

References (40)

  • N. Siep et al.

    Hunger is the best spice: An fMRI study of the effects of attention, hunger and calorie content on food reward processing in the amygdala and orbitofrontal cortex

    Behavioural Brain Research

    (2009)
  • American Psychiatric Association

    Diagnostic and statistical manual of mental disorders

    (2013)
  • J.A. Bargh

    The four horsemen of automaticity: Awareness, efficiency, intention, and control in social cognition

  • J. Blechert et al.

    Food-pics: An image database for experimental research on eating and appetite

    Frontiers in Psychology

    (2014)
  • T. Brockmeyer et al.

    Approach bias modification in food craving-a proof-of-concept study

    European Eating Disorders Review

    (2015)
  • J.T. Cacioppo et al.

    Rudimentary determinants of attitudes. II: Arm flexion and extension have differential effects on attitudes

    Journal of personality and social psychology

    (1993)
  • P.J. Corr

    Automatic and controlled processes in behavioural control: Implications for personality psychology

    European Journal Personality

    (2010)
  • J. Cousijn et al.

    Reaching out towards cannabis: Approach-bias in heavy cannabis users predicts changes in cannabis use

    Addiction

    (2011)
  • J. De Houwer

    What are implicit measures and why are we using them

  • J. De Houwer et al.

    On the generality of the affective Simon effect

    Cognition and Emotion

    (2001)
  • Cited by (45)

    • Duration of an acute moderate-intensity exercise session affects approach bias toward high-calorie food among individuals with obesity

      2022, Appetite
      Citation Excerpt :

      A positive value indicates an approach bias toward high-calorie food, whereas a negative value denotes an avoidance bias toward high-calorie food. Explicit ratings were assessed using the food rating task, which has been validated in previous studies and has been proven to be a reliable measure of explicit ratings of food cues (Ha et al., 2016; Lim et al., 2018; Paslakis et al., 2016). Each of the 80 food images (e.g., 40 high-calorie food images and 40 low-calorie food images) was presented one at a time in the center of a computer screen.

    View all citing articles on Scopus
    View full text