Appetite 52 (2009) Contents lists available at ScienceDirect. Appetite. journal homepage:

Transcription

1 Appetite 52 (2009) Contents lists available at ScienceDirect Appetite journal homepage: Research report Individual differences in food-cue reactivity. The role of BMI and everyday portion-size selections Amanda Tetley a, *, Jeffrey Brunstrom b, Paula Griffiths a a Department of Human Sciences, Loughborough University, Loughborough, Leicestershire LE11 3TU, UK b Department of Experimental Psychology, 12a Priory Road, Bristol BS8 1TU, UK ARTICLE INFO ABSTRACT Article history: Received 15 August 2008 Received in revised form 12 January 2009 Accepted 17 February 2009 Keywords: Overweight Everyday portion size Dietary restraint Food-cue reactivity This study sought to explore the association between sensitivity to food cues (sight and smell of food) and (i) body weight (overweight or non-overweight), (ii) typical everyday portion sizes, and (iii) the DEBQ-restraint scale. One hundred and twenty female participants (aged between 18 and 30) were exposed to the sight and smell of pizza for three minutes. Both before and after this period, three measures of food-cue reactivity were taken (ratings of desire to eat, craving, and a measure of desired portion size of the cued food). Separate regression analyses were used to explore the associations between the change in measures of cue reactivity from pre- to post-cue exposure, and (i) an overweight/ non-overweight classification, (ii) a measure of participants everyday portion-size selection, and (iii) scores on the DEBQ-restraint scale. This analysis revealed that scores on the DEBQ-restraint scale were not significantly associated with changes in food-cue reactivity. However, individuals identified as overweight, and participants who reported consuming the largest everyday portion sizes, experienced a significantly greater change in their desired portion size of the cued food. The findings suggest that heightened food-cue reactivity might present an under-explored risk factor for overeating, and becoming overweight, and/or maintaining an overweight body shape. ß 2009 Elsevier Ltd. All rights reserved. Introduction The effects of cue reactivity in drug-taking behaviour have been explored extensively. Numerous studies show that exposure to a cue such as the sight of a cigarette, or other drug paraphernalia, can elicit both a physiological response (e.g., Breteler, Schippers, De Jong, van der Stark, 2000; Rikard-Figuero & Zeichner, 1985), and a specific craving (e.g., Burton & Tiffany, 1997; Field & Duka, 2005; Pomerlau, Fertig, Baker, & Conney, 1983) that can lead to drugtaking behaviour (e.g., Wigmore & Hinson, 1991). By contrast, relatively less is known about reactivity to food cues. In animals, exposure to a familiar food is found to elicit feeding behaviour, and this occurs even in the absence of nutritional need (Weingarten, 1985). Similar reactivity to food cues has also been observed in humans. For example, exposure to food cues has been found to stimulate a physiological preparedness to consume food (see Mattes, 1997), evidenced by increases in salivation, heart rate, heart rate variability, blood pressure, skin conductance, and gastric activity (Nederkoorn, Smulders, Havermans, & Jansen, 2004; Nederkoorn, Smulders, & Jansen, 2000). * Corresponding author. address: Amanda.Tetley@nottshc.nhs.uk (A. Tetley). Likewise, exposure to the sensory characteristics of food has been found to stimulate food intake (Cornell, Rodin, & Weingarten, 1989). Despite basic research investigating food-cue reactivity, relatively few studies have explored individual differences in this dietary phenomenon, and very little is known about the extent to which elevated food-cue reactivity represents a risk factor for behaviours leading to overeating and obesity. In a recent study, Jansen et al. (2003) suggested that sensitivity to food cues might differ in overweight, and non-overweight, children. Yet, to date, this finding has not been confirmed in adults. Rather, previous authors have considered differences in food-cue reactivity across restrained, and unrestrained, eaters. These studies have revealed that restrained eaters consume larger amounts of a cued food than unrestrained eaters (Fedoroff, Polivy, & Herman, 1997; Fedoroff, Herman, & Polivy, 2003; Jansen & van den Hout, 1991; Rogers & Hill, 1989) after food-cue exposure, and experience a greater desire to eat, elevated levels of craving (Fedoroff et al., 1997, 2003), increases in salivation (Brunstrom, Yates, & Witcomb, 2004; Klajner, Herman, Polivy, & Chhabra, 1981) and increased blood pressure (Nederkoorn et al., 2000). There is reason to suspect that heightened food-cue reactivity in adults might be associated with a more general tendency towards being overweight and daily overeating. Put simply, food-cue /$ see front matter ß 2009 Elsevier Ltd. All rights reserved. doi: /j.appet

2 A. Tetley et al. / Appetite 52 (2009) reactivity is assumed to represent a form of conditioned response. Thus, after repeated pairing of an external cue with the ingestion of food, mere exposure to the cue is found to initiate appetitive behaviour, even in the absence of prior hunger (Birch, McPhee, Sulivan, & Johnson, 1989; Weingarten, 1984). Given this, one possibility is that individuals who are especially cue reactive may be prone to overeating and being overweight. This is because they are more likely to engage in dietary behaviours that promote a positive energy balance. Here, our primary objective was to examine whether cue reactivity is associated with being overweight and consuming larger everyday portion sizes. Specifically, we hypothesised (i) that overweight individuals will experience greater reactivity to a food cue than non-overweight individuals and (ii) that individuals who report consuming larger everyday portion sizes will experience greater food-cue reactivity than individuals who typically consume smaller everyday portion sizes. A secondary aim was to determine the extent to which food-cue reactivity is associated with a tendency to restrict ones dietary intake. As discussed, previous studies have suggested that restrained eaters experience elevated levels of food-cue reactivity (Fedoroff et al., 1997, 2003). However, these studies have assessed dietary restraint using a measure which identifies restrained eaters who engage in dietary restriction interspersed with bouts of disinhibited eating. Accordingly, in this study we sought to determine the extent to which food-cue reactivity is associated with a tendency to restrict ones dietary intake, independently of a tendency to engage in disinhibited eating by using the Dutch Eating Behaviour Questionnaire Restraint Scale (DEBQ-R; van Strien, Frijiter, Begers, & Defares, 1986) as a measure of food restriction. In the present study, we used a version of a paradigm that was introduced by Cornell et al. (1989). Following Weingarten s (1985) notion of cue-induced hunger, Cornell et al. (1989) sought to determine the extent to which food-cue exposure induces a motivation to eat in the absence of physiological need. Accordingly, participants were offered a buffet lunch prior to food-cue exposure. Likewise, in the present study, participants were asked to consume items from a buffet lunch until they felt satiated prior to exposure to a food cue. In previous food-cue reactivity studies (Fedoroff et al., 1997, 2003), food-cue reactivity has been assessed using a measure of intake of a cued food. Typically, in these studies, intake in a control (non-cued) condition is compared with intake in a foodcued condition. However, this design is flawed given that even the presentation of food immediately prior to an eating episode acts as a cue to stimulate intake (Weingarten, 1985). Thus, a non-cued condition effectively becomes a cued condition. Consequently, in the present study, a novel measure of desired portion size of the cued food was employed to minimise undesirable cueing effects in the non-cued condition. Participants were asked to indicate their desired portion of the test food (pizza) before, and after, exposure to this food using a representation of the food, rather than the actual food. This allowed us to detect changes in prospective intake brought about by exposure to the food cue, and most importantly it minimised cueing effects in the pre-exposure phase. Method Participants One hundred and twenty participants were recruited via from the population of female undergraduate students at Loughborough University (UK) (mean age = 20.9, SD = 2.6). Ethics approval was obtained from the local Departmental Ethics Committee. All participants gave written consent to participate in the study and were informed that they could withdraw at any time during the experiment. Measures Cue reactivity Craving and desire to eat were measured using 100-mm visualanalogue rating scales. Respectively, these were headed How strong is your desire to eat pizza right now? and How strong is your craving for pizza right now? Both were anchored with the phrases Not at all and Extremely. Desired portion size was assessed using a cardboard rectangular model of pizza (150 mm 130 mm 32 mm). This was identical in size to the slice presented during cue exposure and provided participants with an indication of the depth of the pizza. Participants were told that this model represented a cheese and tomato pizza. Using this model as a reference, the participants were asked to select the amount of pizza they would want to eat at that moment in time. They indicated this selection using a 420 mm 594 mm sheet of card. The card had a diagonal line running from the bottom left to the top right hand corner, and participants were told that the corner of their selected portion size should intersect this line (examples were provided). The area of the selected portion size was then calculated. Everyday portion size To obtain a measure of everyday portion size, participants were shown sets of eight photographs. Each set depicted a commonly consumed food (Nelson, Atkinson, & Meyer, 1997). The series of pictures contained portion sizes ranging from the 5th to 95th centile on a distribution of portion sizes observed in The Dietary and Nutritional Survey of British Adults (Gregory, Foster, Tyler, & Wiseman, 1990). Participants were asked to use these photographs to indicate the amount of food that they typically consume. They were told that if the amount of a food corresponded exactly with the amount depicted in one of the pictures, then they should put a cross through the corresponding number on a scale. However, if the amount was slightly larger or smaller then they should indicate this by placing a cross to the left or right of the corresponding number. This response was recorded using a 9-point visual analogue scale (80 mm in length) anchored with the numbers 0 and 8. The points on the scale were numbered consecutively at 10-mm intervals. Participants may place their cross between 0 and 1 on the scale if their everyday portion was smaller than the first picture presented. In total 15 foods were presented in this way (pasta, cornflakes, chocolate cake, sponge pudding, potatoes, baked beans, lasagne, spaghetti bolognaise, chips, cheesecake, roast beef, battered fish, carrots, fruit salad, and quiche). Pizza was not included in the assessment of everyday portion size to exclude any cuing effects that may be brought about by exposure to a pictorial cue of this food. Dietary restraint Dietary restraint was assessed using the restraint scale of the Dutch Eating Behaviour Questionnaire (DEBQ-R; van Strien et al., 1986). This scale is described as a measure of successful dietary restriction (Gorman & Alison, 1995). 1 Procedure Before arriving to be tested the participants were told that the aim of the experiment was to explore the relationship between 1 Participants were not asked to complete the external, and emotional, scales of the DEBQ to avoid overburdening the participants with further assessments in a study already incorporating a number of assessments. Furthermore, these scales did not address the specified study aims.

3 616 A. Tetley et al. / Appetite 52 (2009) Table 1 Within-subject t-tests, means, and standard deviations, for pre-exposure and post-exposure appetite ratings and portion-size selections. n Pre-exposure Post-exposure t p Mean SD Mean SD Desire to eat pizza (mm) <.001 * Craving for pizza (mm) <.001 * Desired portion size of pizza (mm 2 ) * p < appetite and mood. They were also told that they would have to rate their mood throughout the experiment, that they would be asked to offer an opinion on various foods, and that they would receive a buffet-lunch. Participants were tested individually between 11 a.m. and 3 p.m. All were instructed to refrain from eating for three hours prior to the onset of the experiment. On arrival, participants rated their hunger and fullness using 100-mm visual-analogue ratings scales. Hunger and fullness rating scales were headed with How hungry/ full are you right now? Both were anchored with the phrases Not at all hungry/full and Extremely hungry/full. They were then presented with a buffet-lunch and were asked to eat until they felt comfortably full. The buffet lunch comprised three sandwiches (ham, cheese, and chicken), one and half sausage rolls, six small scotch eggs (egg filling covered in sausage meat), three handfuls of original flavoured crisps, two large oranges, six Jaffa Cakes (sponge and orange filling covered in chocolate), and a glass of water. To determine the efficacy of the lunch, participants provided a second set of hunger and fullness ratings after lunch. They were then asked to complete the everyday portion-size assessment. This was followed by the first set of cue reactivity measures. Consistent with the cover story, this initial set of reactivity measures also included a number of 100-mm visual-analogue ratings scales relating to participants current mood. The participants were then exposed to the sight and smell of pizza for three minutes. The pizza was presented in a rectangular slice and weighed 300 g (810 kcal). It was placed on the table directly in front of the participant and the participants were told that the experimenter would return shortly. The duration of the exposure phase was set at three minutes because similar cue exposure periods have been reported to induce changes in motivation to eat (Cornell et al., 1989). After pizza-cue exposure, the participants provided a second set of reactivity measures, after which they rated their liking for pizza and completed the DEBQ-restraint scale. Finally, height was measured using a stadiometer (Bodycare, Warwickshire, UK) and weight was assessed using a set of weighing scales (Soehnle, Germany). BMI was then calculated from this data. Data analysis In this experiment we were interested to assess the effect of foodcue exposure on an individual s motivation to eat. We reasoned that if food-cue exposure is able to motivate food intake, desire to eat, craving for, and desired portion size of the cued food should be elevated after, relative to before, food-cue exposure. In our initial analyses, we sought to compare how the pre- and post-cue exposure ratings of desire to eat and craving, and desired portion size differed by applying a series of within-subjects t-tests. For the remaining analyses, we calculated a change score for each of our three measures of reactivity (desire to eat, craving, desired portion size), derived from the difference between the measure of reactivity before and after cue exposure. To determine the extent to which these changes are associated with overweight/normal weight, average everyday portion size, and the DEBQ-restraint scale, we used separate regression analyses. An average measure of everyday portion size was calculated using the scores for the 15 foods. BMI scores were used to dichotomise individuals into a normal weight (BMI < 25.0), and an overweight (BMI 25.0), group. In each regression model, we controlled statistically for liking for pizza by entering this simultaneously into the regression model as a covariate. For each measure of change in reactivity, we also included the corresponding baseline measure of reactivity by entering this simultaneously into the regression model as a covariate. Where one of our independent variables (overweight/non-overweight classification, everyday portion-size selection and scores on the DEBQ-restraint scale) shared a statistically significant association with a measure of food-cue reactivity, the parameter estimates from the regression equation were used to predict the change in the measure of food-cue reactivity across the values of the independent measure. 2 All analyses in the present study were conducted using SPSS Version 11. The significance level tested was set at p <.05 and all tests were two-tailed to ensure that associations between variables (regression analyses), or changes across variables (within-subject t-tests), were detected irrespective of whether they were in the hypothesised direction. Results Efficacy of the buffet lunch The buffet lunch was designed to ensure that participants were satiated prior to pizza-cue exposure. After the buffet lunch, mean hunger rating was 11.2 mm (SD = 11.8 mm) and mean fullness rating was 73.4 mm (SD = 17.8 mm), suggesting that participants were satiated by this phase in the experiment. These ratings represented a statistically significant change from baseline (p <.05). Post-lunch hunger and fullness ratings did not differ across overweight and nonoverweight individuals, across our measure of everyday portion size, or across DEBQ-restraint scores (all p <.05). Effects of pizza-cue exposure Pizza-cue exposure significantly increased participants craving for pizza, and their desire to eat this food. However, it did not significantly stimulate the selection of a larger portion of pizza (Table 1). Correlations between changes in reactivity measures Change in ratings of desire to eat pizza from pre- to post-cue exposure were not significantly correlated with changes in desired 2 These were calculated using the following formula; y = b 1 x + b 2 z + b 3 w + b 0, where y is the change in the measure of reactivity, x is the independent variable, z is the liking for pizza, and y is the relevant pre-exposure reactivity measure, b 1,b 2,b 3 is the relevant parameter estimate from the regression model (the gradient of the regression line), b 0 is the constant coefficient from the model (i.e., the change in the measure of reactivity when the value of the independent variable is zero). In all predictions, the relevant pre-exposure measure and liking for pizza were held at their average in the sample. Thus, all predictions were made for a participant with an average liking for pizza and average pre-exposure rating for the particular measure of cue reactivity.

4 A. Tetley et al. / Appetite 52 (2009) Table 2 Between-subject t-tests, means, and standard deviations for age, dietary restraint, and everyday portion-size selection for overweight, and non-overweight, individuals. Table 4 Predicted values from the linear regression model for overweight, and normal weight, individuals for pre-exposure portion-size selection, and the total amounts these individuals would be expected to consume a. Non-overweight Overweight t p Pre-exposure Post-exposure Mean SD Mean SD Age DEBQ-restraint scores Everyday portion-size selection Non-overweight Overweight a Holding liking for pizza, and pre-exposure pizza size, at their mean values for the sample (70.83 mm, mm 2 [ kcal], respectively). portion size of pizza after cue exposure (r =.062, p =.503). Although there was a statistically significant association between change in craving for pizza from pre- to post-exposure and change in desired portion of this food after pizza-cue exposure (r =.233, p =.010), this association was modest. Changes in the ratings of desire to eat and craving after cue exposure were significantly correlated (r =.462, p <.001). Cue reactivity and overweight Mean BMI was 22.9 (SD = 2.6). Twenty-six participants were classified as overweight (BMI > 25.0, BMI range ), and 94 were classified as not overweight (BMI < 25.0, BMI range ). There was no evidence of a statistically significant difference in the age of the overweight, relevant to the non-overweight, group or in the scores they obtained on the DEBQ-restraint scale (both p >.05) (Table 2). However, overweight individuals did report selecting larger everyday portion sizes that non-overweight individuals (p =.014) (Table 2). After cue exposure, overweight participants experienced a greater change in both their desire to eat and their craving for pizza. However, in each case, the difference between overweight and normal weight participants was not statistically significant (desire, p =.122; craving, p =.128) (Table 3). After controlling for each participant s liking for pizza and their pre-exposure desired portion size in a regression analysis, overweight participants were found to experience a greater increase in their desired pizza portion size following exposure to pizza (p =.031) (Table 3) and the model explained over 60% of the variance in change in desired portion size. Using the parameter estimates from the regression model for change in desired portion size of the cued food, we predicted the change in the number of kilocalories an overweight individual with an average liking for pizza (average liking for pizza = 70.8 mm) and average pre-exposure desired portion size ( mm 2 [174.2 kcal]) wanted to eat after cue exposure compared to a non-overweight individual with an average liking for pizza and average pre-exposure portion size. This prediction suggested that their desired portion size increased on average by 46.1 kcal. By contrast, the desired portion-size selection of normalweight participants decreased by 17.2 kcal. We also calculated the total amount of pizza in kilocalories that overweight, and nonoverweight, individuals were likely to consume after cue exposure by using a regression model to predict the amount of pizza (kcal) that would be ingested prior to cue exposure, and then adding onto this the change in intake after cue exposure (Table 4). Cue reactivity and everyday portion size A further aim of the present study was to explore the possibility that high levels of cue-reactivity are associated with a general tendency to consume large portion sizes. The findings from the regression models suggest that everyday-portion size was not significantly associated with either change in craving for pizza (p =.378) or change in desire for pizza (p =.328) (Table 5). However, everyday portion size was associated with a significant change in desired pizza size (p =.036) (Table 5) after controlling statistically for participants liking for pizza and their pre-exposure desired portion size. In fact, this model accounted for over 60% of the variance in change in desired portion size. Using the parameter estimates from the regression model for change in desired portion size of pizza, we predicted the change in desired portion size after pizza-cue exposure in kilocalories for an individual who reported consuming small, medium, and large, everyday portion sizes (calculated as the mean value in each tertile of the data). In this analysis, we held liking for pizza and pre-exposure portion sizes at their mean values for the sample (70.8 mm and mm 2 [174.2 kcal] respectively) (Fig. 1). This suggested that individuals who typically consume the largest everyday portion size experi- Table 3 Parameter estimates from linear regression models of associations between overweight and three measures of food-cue reactivity (n = 120), and the corresponding covariates a. Dependent variable Independent variables and covariates R square B SE p Change in desired portion size.636 Overweight b * Pre-exposure desire-to-eat.7.05 <.001 * Liking for pizza <.001 * Change in craving.114 Overweight b Pre-exposure craving Liking for pizza <.001 * Change in desire-to-eat.109 Overweight b Pre-exposure desire-to-eat * Liking for pizza * a Covariates are liking for pizza and the relevant pre-exposure rating. b Reference category is non-overweight. * p < 0.05.

5 618 A. Tetley et al. / Appetite 52 (2009) Table 5 Parameter estimates from linear regression models of associations between everyday portion-size selection (small, medium, and large) and three measures of food-cue reactivity and corresponding covariates (n = 120) a. Dependent measure Independent variables and covariates R square B SE p Change in desired portion size.636 Average everyday portion size b * Pre-exposure desired portion size <.001 * Change in craving Liking for pizza * Average everyday portion size b Pre-exposure craving Liking for pizza * Change in desire-to-eat.097 Average everyday portion size b Pre-exposure desire-to-eat * Liking for pizza * * p < a Covariates were liking for pizza and the relevant pre-exposure rating. b Comparison across the continuous measure of average everyday portion-size selection enced the greatest change in desired portion size after cue exposure. Again, we also calculated the total number of kilocalories that would be ingested after pizza-cue exposure for an individual reporting consuming small, medium, and large, everyday portion sizes. To do this, parameter estimates from a regression model predicting pre-exposure portion-size selection were added to the predicted change in desired pizza sizes (Table 6). This revealed that individuals who consume larger everyday portion sizes report a desire to eat a portion of a cued food comprising a greater number of kilocalories (approximately 100 kcal) relative to individuals who consume smaller everyday portion sizes. To summarise, in this experiment, both being overweight and selecting larger everyday portion sizes was associated with desiring larger portion sizes of a cued food. To determine the extent to which one of these predictor variables have greater predictive validity in relation to food-cue reactivity, weight status (overweight/normal weight) and average everyday portion size were entered simultaneously into the regression model predicting change in desired portion size of the cued food. In this model, being overweight and selecting larger everyday portion sizes failed to significantly predict change in desired portion size (p >.05). This shows that neither everyday portion-size selection, nor being overweight, has greater predictive validity in relation to food-cue reactivity, and consequently that shared variance between the two variables accounts for changes in desired portion of a cued food observed in this experiment. Cue reactivity and dietary restraint The final aim of this study was to determine the extent to which food-cue reactivity is associated with participants scores on the DEBQ-restraint scale. The results of the regression analyses provided little evidence to suggest that this scale was associated with a change in any of the measures of cue reactivity (all p >.05, results not shown here). Discussion The present study was designed to determine the extent to which exposure to the visual and olfactory characteristics of food would stimulate a greater motivation to eat in (i) overweight, relative to non-overweight, individuals, (ii) individuals who typically consume larger everyday portion sizes, and (iii) individuals who obtain higher scores on the DEBQ-restraint scale. To test these relationships, we compared participants ratings of desirability, craving for, and desired portion size of pizza, both before and after brief exposure to the olfactory and visual characteristics of this food. The results of the study revealed little evidence to suggest that when dietary restraint is assessed independently of disinhibited eating it shares any association with food-cue reactivity. Further, the results suggested that whilst an individual s weight status Table 6 Predicted values from the linear regression model for pre-exposure pizza size, and the total amounts individuals would be expected to consume in kilocalories for small, medium, and large, portion sizes (calculated as the average value in each tertile of the data) a. Fig. 1. Predicted change in pizza size in kilocalories (kcal) for individuals with small (2.8), medium (3.7) and large (4.5) average everyday portion size (calculated as the average score in each tertile of the data) estimated using the parameter estimates from a linear regression model for change in pizza size (B = ). In this model pre-exposure pizza size and liking for pizza are held at their mean values in the sample ( mm 2 [ kcal] and 70.8 mm respectively) and their respective parameter estimates (B = 0.7, B = 64.5, respectively) were used to predict change in desired pizza size in kilocalories. Pre-exposure Everyday portion size Small Medium Large Post-exposure a Holding liking for pizza, and pre-exposure pizza size, at their mean values for the sample (70.83 mm, mm 2 [ kcal], respectively).

6 A. Tetley et al. / Appetite 52 (2009) and their everyday portion-size selections were not significantly associated with changes in ratings of appetite (desire to eat and craving) following food-cue exposure, they were associated with a relatively greater change in desired portion size of the cued food. This latter finding is unsurprising given that changes in desire to eat and craving for pizza were poorly correlated with changes in desired pizza portion size. One possibility is that that the rating scales for desire to eat and craving lacked sufficient sensitivity to reveal weight-related differences in food-cue reactivity. A further finding from this study was that neither everyday portion-size selection, nor being overweight, significantly predicted change in desired portion size when entered simultaneously into a regression model. This is likely to result from the fact that overweight individuals were more likely to report selecting larger everyday portion sizes than their normal-weight counterparts. This suggests that everyday portion size and BMI were acting as proxy measures for each other. We also note several limitations to this study. The most important of these is that we did not include a control condition in which participants are exposed to three minutes of non-foodexposure. In the absence of this condition, it is difficult to conclude that differences in pre- and post-exposure desired portion sizes resulted exclusively from pizza-cue exposure. For example, they may result from differences in food digestion in overweight, relative to non-overweight, individuals following the buffet lunch, or differences in intake at the buffet lunch between these two groups. Both differences could have caused overweight individuals to desire larger portion sizes of the cued food. The second limitation of this study was that the assessment of participants everyday portion sizes relied on a self-report measure. This may not provide a valid measure of everyday portion size as individuals may attempt to portray themselves in particular ways. Final limitations were that the study sample comprised a non-random sample of the same gender (female) from a specific age range (18 30), of a particular educational level (university students), and most likely from a particular social class. Further to this, the study was only able to test the effects of exposure to one food cue (pizza) and did not account for any effects of individual differences in sense of olfaction on sensitivity to an olfactory, and visual, cue. If the findings of this study are replicated in future studies addressing the limitations of this experiment, one possibility is that selecting larger everyday portion sizes may be a precursor to elevated food-cue reactivity (Jansen, 1998). Specifically, individuals who overeat may come to habitually overeat in the presence of food cues. Given this, one possibility might be that an overweight individual s tendency to overeat causes them to desire relatively larger portions of a cued food. There could be other potential explanations for an overweight individual s tendency to desire larger portions of a cued food. One possibility is that they are more cue reactive than their nonoverweight counterparts because they have a tendency to be more impulsive (Nederkoorn, Smulders, Havermans, Roefs, & Jansen, 2006), and to have a heightened sensitivity to reward (Franken & Muris, 2005). This possibility has not been considered empirically and would be an interesting avenue for further research. Another possibility, identified by Jansen et al. (2003), is that overweight individuals experience delayed satiety signals when consuming a cued food. According to Jansen et al. (2003), this might be because the decline in the pleasantness of food, which occurs during intake and leads to the termination of an eating episode (sensory specific satiety), takes longer in overweight individuals. The authors suggest that this decline in the pleasantness of a food s sensory characteristic may occur during exposure to a food cue for nonoverweight individuals and consequently causes these individuals to consume smaller amounts of a cued food. However, the exact length of cueing required to initiate these effects in overweight, and non-overweight, individuals is unclear from Jansen et al. s (2003) account. One implication of observing greater sensitivity to food cues in overweight individuals is that levels of food-cue reactivity could be manipulated and reduced in these individuals. Given that food-cue reactivity is assumed to result from learned associations between the sensory characteristics of a food (visual and olfactory) and food ingestion (Wardle, 1990; Weingarten, 1985), it may be possible that this can be achieved by extinguishing these learned associations. Previously, authors have reported utilising a technique referred to as response prevention to reduce cue reactivity in bulimics (Jansen, 1998) and alcoholics (Drummond & Glautier, 1994; Monti et al., 1993). This involves presenting individuals with a cue associated with food or alcohol and preventing them from eating, or from drinking alcohol, in the presence of this stimulus. Given this, future studies may wish to explore the prospect that levels of food-cue reactivity can be manipulated in overweight and obese individuals. References Birch, L. L., McPhee, L., Sulivan, S., & Johnson, S. (1989). Conditioned meal initiation in young children. Appetite, 13(2), Breteler, J. A., Schippers, C. M., De Jong, C. A. J., & van der Stark, C. P. F. (2000). No effect of negative mood on the alcohol cue reactivity of in-patient alcoholics. Addictive Behaviors, 25(4), Brunstrom, J. M., Yates, H. M., & Witcomb, G. L. (2004). Dietary restraint and heightened reactivity to food. Physiology & Behaviour, 81(1), Burton, S. M., & Tiffany, S. T. (1997). The effect of alcohol consumption on craving to smoke. Addiction, 92(1), Cornell, C. E., Rodin, J., & Weingarten, H. P. (1989). Stimulus-induced eating when satiated. Physiology & Behaviour, 45, Drummond, D. C., & Glautier, S. T. (1994). A controlled trial of cue exposure treatment in alcohol dependence. Journal of Consulting and Clinical Psychology, 62, Fedoroff, I., Polivy, J., & Herman, C. P. (1997). The effects of pre-exposure to food cues on the eating behaviour of restrained and unrestrained eaters. Appetite, 28, Fedoroff, I., Polivy, J., & Herman, C. P. (2003). The specificity of restrained versus unrestrained eaters responses to food cues: general desire to eat, or craving for the cued food? Appetite, 41, Field, M., & Duka, T. (2005). Cue reactivity in smokers: the effects of perceived availability and gender. Pharmacology, Biochemistry, and Behaviour, 78(3), Franken, I. H. A., & Muris, P. (2005). Individual differences in reward sensitivity are related to food craving and relative body weight in healthy women. Appetite, 45, Gorman, B. S., & Alison, D. B. (1995). Measures of restrained eating. In D. B. Allison (Ed.), Handbook of assessment methods for eating behaviors and weight-related problems (pp ). Thousand Oaks: Sage. Gregory, J., Foster, K., Tyler, H., & Wiseman, M. (1990). The dietary and nutritional survey of British adults. London: HMSO. Jansen, A. (1998). A learning model of binge eating: cue reactivity and cue exposure. Behaviour Research and Therapy, 36(3), Jansen, A., Theunissen, N., Slechten, K., Nederkoorn, C., Boon, B., Mulkens, S., et al. (2003). Overweight children overeat after exposure to food cues. Eating Behaviours, 4, Jansen, A., & van den Hout, M. (1991). On being led into temptation counterregulation of dieters after smelling a preload. Addictive Behaviors, 16, Klajner, F., Herman, C. P., Polivy, J., & Chhabra, R. (1981). Human obesity, dieting, and anticipatory salivation to food. Physiology & Behaviour, 27(2), Mattes, R. D. (1997). Physiologic responses to sensory stimulation by food nutritional implications. Journal of American Dietetic Association, 97, Monti, P. M., Rohsenow, D. J., Rubonis, A. V., Niaura, R. S., Sirota, A. D., Colby, S. M., et al. (1993). Cue exposure with coping skills treatment for male alcoholics. A preliminary investigation. Journal of Consulting and Clinical Psychology, 61, Nederkoorn, C., Smulders, F., Havermans, R., & Jansen, A. (2004). Exposure to bingefood in bulimia nervosa: finger pulse amplitude as a potential measure of urge to eat and predictor of food intake. Appetite, 42(2), Nederkoorn, C., Smulders, F., Havermans, R., Roefs, A., & Jansen, A. (2006). Impulsivity in obese women. Appetite, 47, Nederkoorn, C., Smulders, F. T. Y., & Jansen, A. (2000). Cephalic phase responses, craving and food intake in normal subjects. Appetite, 35(1), Nelson, M., Atkinson, M., & Meyer, J. (1997). A photographic atlas of food portion sizes. London: MAFF Publications. Pomerlau, O. F., Fertig, J., Baker, L., & Conney, N. (1983). Reactivity to alcohol cues in alcoholics and non-alcoholics: implications for a stimulus control analysis of drinking. Addictive Behaviors, 8, 1 10.

7 620 A. Tetley et al. / Appetite 52 (2009) Rikard-Figuero, K., & Zeichner, A. (1985). Assessment of smoking urge and its concomitants under an environmental smoking cue manipulation. Addictive Behaviors, 10, Rogers, P. J., & Hill, A. (1989). Breakdown of dietary restraint following mere exposure to food stimuli: interrelationships between restraint, hunger, salivation and food intake. Addictive Behaviors, 14, van Strien, T., Frijiters, J. E., Bergers, G. P. A., & Defares, P. B. (1986). The Dutch Eating Behaviour Questionnaire (DEBQ-R) for assessment of restrained, emotional and external eating behaviour. International Journal of Eating Disorders, 5, Wardle, J. (1990). Conditioning processes and cue exposure in the modification of excessive eating. Addictive Behaviors, 15(4), Weingarten, H. P. (1984). Meal initiation controlled by learned cues: basic Behavioural Properties. Appetite, 5, Weingarten, H. P. (1985). Stimulus control of eating: implications for a two-factor theory of hunger. Appetite, 6, Wigmore, S. W., & Hinson, R. E. (1991). The influence of setting on consumption in the balanced placebo design. British Journal of Addiction, 86(2),