نقش حساسیت به پاداش و تکانشگری در واکنش پذیری نشانه غذایی

Abstract Exposure to the visual and olfactory characteristics of food can elicit a desire to eat and can stimulate food intake. This study sought to determine the extent to which sensitivity to reward and impulsivity are associated with this ‘food-cue reactivity’ in two motivational states (food deprived and non-food deprived). Female participants (N = 120) were exposed to a pizza cue for three minutes whilst food deprived and non-food deprived. Before and after this period, three measures of food-cue reactivity were taken (ratings of desire to eat, craving, and desired portion size of the cued food). Two important findings emerged from the study. Firstly, individuals with higher levels of trait impulsivity experienced greater changes in appetite ratings in both motivational states (food deprived and non-food deprived). They also reported greater changes in desired portion size of a cued food when food deprived. Secondly, individuals with a high sensitivity to reward experienced a greater change in their desired portion of the cued food, but only when non-food deprived. These results indicate that individual differences in food-cue reactivity could be related to variation in sensitivity to stimuli that predict the occurrence of a reward, and to an inability to exercise sufficient self control in the presence of tempting environmental stimuli.

. Introduction In humans, exposure to a food cue has been found to stimulate appetite. For example, exposure to food cues has been found to stimulate a physiological preparedness to consume food (see Mattes, 1997), and to increase hunger, desire to eat, and craving for a cued food (Fedoroff et al., 1997 and Nederkoorn et al., 2006). Likewise, studies have also suggested that food intake increases after exposure to the sensory characteristics of food (Cornell et al., 1989 and Fedoroff et al., 1997). To date, relatively few studies have considered the extent to which individuals differ in their sensitivity to the stimulatory effects of food cues. Those that have considered this issue have focused almost exclusively on the association between dietary restraint and food-cue reactivity. Generally, these studies find that restrained eaters consume larger amounts of a cued food than unrestrained eaters after food-cue exposure ( Fedoroff et al., 1997, Fedoroff et al., 2003, Jansen et al., 2003 and Rogers and Hill, 1989), and experience a relatively greater desire to eat, elevated levels of craving ( Fedoroff et al., 1997 and Fedoroff et al., 2003), greater increases in salivation ( Brunstrom et al., 2004 and Klajner et al., 1981), and a relatively greater increase in blood pressure ( Nederkoorn et al., 2000). More recent studies have also suggested that food-cue reactivity might differ in overweight, relative to non-overweight, individuals. For example, Jansen et al. (2003) found that overweight children are more responsive to the sensory characteristics of food cues than non-overweight children. Likewise, Tetley, Brunstrom, and Griffiths (2009) report that overweight adult females are more likely to select larger portion sizes of a cued food than their non-overweight counterparts. Taken together, these findings suggest that food-cue reactivity might present a risk factor for becoming overweight and/or maintaining an overweight body shape. Given the possible association between heightened food-cue reactivity and being overweight, it is important to consider how individual variation in this dietary phenomenon might account for this association. One possibility is that overweight individuals' heightened sensitivity to the stimulatory effects of food cues results from a tendency to be impulsive (Nederkoorn, Smulders, Havermans, Roefs, & Jansen, 2006) and to experience a greater sensitivity to reward (Franken and Muris, 2005 and Tetley et al., 2009). Partly consistent with this possibility, Guerrieri, Nederkoorn, and Jansen (2007) found that high levels of impulsivity are associated with overeating in the laboratory. However, despite this, it remains unclear whether impulsivity and sensitivity to reward are in fact associated with heightened food-cue reactivity. The personality characteristic of sensitivity to reward is assumed to reflect the sensitivity of a neurological system referred to as the Behavioural Approach System (BAS) (Gray, 1976, Gray, 1981, Gray, 1987a and Gray, 1987b). The BAS is believed to be a ‘reward’ or ‘approach’ system which responds to positive incentives in the environment by activating behaviour to obtain specific rewards. It has been suggested that this system continuously monitors the environment for signals of reward. When a cue associated with the receipt of a reward is encountered, the BAS is initiated through activation of the dopaminergic system (Gray, 1987b), and motor output is then increased towards the reward, further activating the BAS and promoting behaviour to obtain the reward (‘approach’ behaviour) (Kane, Loxton, Staiger, & Dawe, 2004). One possibility is that greater reactivity of the BAS, as evidenced by high trait sensitivity to reward, will predict the occurrence of greater appetitive motivation after brief exposure to the visual and olfactory characteristics of a food cue. Previous studies have already established an association between sensitivity to alcohol cues and heightened reactivity of the BAS. For example, Franken (2002) reported that sensitivity to reward is associated with a stronger desire for alcohol after exposure to photographs depicting alcoholic drinks. Likewise, Kambouropoulos and Staiger, 2001 and Kambouropoulos and Staiger, 2004 have reported that individuals with a high sensitivity to reward experience a greater urge to drink in the presence of an alcoholic drink. Importantly, sensitivity to reward has been theoretically associated with heightened ‘impulsivity’ (Dawe et al., 2004 and Dolan and Fullam, 2004). Gray, 1976, Gray, 1981, Gray, 1987a and Gray, 1987b model suggests that one pathway to poor impulse control might be via hyper-responsiveness to reward (i.e., a strong BAS). According to this model, impulsivity results from the failure of cues to inhibit reward seeking behaviour (Dolan & Fullam, 2004). Thus, it is possible that heightened food-cue reactivity might also be associated with a measure of impulsivity. The current study aimed to assess the extent to which higher scores on a Sensitivity to Reward Scale (SR) (Sensitivity to Reward and Sensitivity to Punishment Questionnaire; SRSPQ; Torrubia, Avila, Molto, & Caseras, 2001), and Impulsivity scale (EPQ; Eysenck & Esyenck, 1975), are associated with greater changes in appetite after exposure to the visual and olfactory characteristics of a food cue (pizza). Previously, it has been suggested that exposure to a food cue can induce a motivation to eat even in the absence of a physiological need (Cornell et al., 1989 and Weingarten, 1985). In this study, we sought to explore differences in food-cue reactivity across the personality scales when participants were both food deprived and non-food deprived. This allowed for the consideration of individual variability in food-cue reactivity across different levels of motivation to eat. In previous food-cue reactivity studies (Fedoroff et al., 1997 and Fedoroff et al., 2003), food-cue reactivity has been assessed using a measure of intake of a cued food. Typically, in these studies, participants are randomly assigned to a ‘cued’ or a ‘non-cued’ condition, and subsequent food consumption is compared in an ad-lib intake phase. However, this design has limitations 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 when food is presented for consumption in the ad-lib intake phase. To address this concern, we used a novel measure of desired portion size, previously developed by Tetley et al. (Tetley et al, 2009). Specifically, participants were shown a ‘representation’ of the cued food. By using a three-dimensional rectangular model (green and made of cardboard) we conveyed the important dimensions of the cued food (pizza) without presenting its visual characteristics. In summary, this study sought to investigate the association between food-cue reactivity and a measure of Sensitivity to Reward Scale (SR) (SRSPQ; Torrubia et al., 2001), and Impulsivity scale (EPQ; Eysenck & Esyenck, 1975) when individuals were both food deprived and non-food deprived. It was hypothesised that higher scores on the Impulsivity scale of the EPQ and the Sensitivity to Reward Scale of the SRSPQ would be associated with greater changes in appetite after exposure to the visual and olfactory characteristics of a food cue (pizza) in both the food deprived and non-food deprived states.

. Results 3.1. Outliers One participant was removed from the data set because she had a BMI of 13 which indicated that she was severely undernourished. Another participant experienced a change in desired pizza size when hungry which was 5.57 standard deviations above the mean change in this measure. Therefore, her data were also removed from the data set. 3.2. Manipulation of motivational state This experiment was designed to test the effects of cue exposure in two motivational states (food deprived and non-food deprived). At the outset, the mean hunger rating was 64.8 mm (SD = 20.8 mm) and the mean fullness rating was 13.9 mm (SD = 17.0 mm). After lunch, the mean hunger rating was 11.9 mm (SD = 11.8 mm) and the mean fullness rating was 73.0 mm (SD = 16.8 mm). These final ratings represented a statistically significant change from before lunch (both p < 0.05, within-subject t-tests). Pearson correlation analyses revealed that pre- and post-lunch hunger and fullness ratings were not significantly associated with scores on the SR scale or the EPQ-Impulsivity scale (both p > 0.05). 3.3. Effects of pizza-cue exposure Within-subject t-tests revealed that pizza-cue exposure significantly increased participants' craving for pizza, desire to eat pizza, and their desired portion size of this food, whilst the participants were both food deprived and non-food deprived ( Table 1). Table 1. Within-subject t-tests, means, and standard deviations, for pre-exposure and post-exposure appetite ratings and portion size selections, separately whilst participants were food deprived and non-food deprived. Pre-exposure Post-exposure n Mean SD Mean SD t p Food deprived Desire to eat pizza (mm) 118 57.8 29.7 69.7 28.0 6.45 < 0.001⁎ Craving for pizza (mm) 118 51.6 31.5 64.6 29.6 6.5 < 0.001⁎ Desired portion size of pizza (mm2) 118 30,582.9 19,073.6 34,235.2 23,316.8 3.03 0.003⁎ Non-food deprived Desire-to-eat pizza (mm) 118 16.0 20.0 26.7 25.4 6.1 < 0.001⁎ Craving for pizza (mm) 118 14.8 20.7 24.4 26.6 5.3 < 0.001⁎ Desired portion size of pizza (mm2) 118 6509.4 8328.5 10,784.2 12,521.3 5.6 < 0.001⁎ ⁎ Denotes significance. Table options 3.4. Correlations between changes in reactivity measures Changes in ratings of desire to eat pizza from pre- to post-cue exposure were significantly correlated with changes in desired portion size when participants were both food deprived (r = 0.367, p < 0.001), and non-food deprived (r = 0.465, p < 0.001). Likewise, changes in craving for pizza were significantly correlated with changes in desired portion size when participants were both food deprived (r = 0.414, p < 0.001), and non-food deprived (r = 0.401 p < 0.001). Finally, change in desire to eat pizza and change in craving for pizza were highly correlated, both before (r = 0.846, p < 0.001) and after lunch (r = 0.759, p < 0.001). 3.5. Food-cue reactivity and the SR scale The mean score on the SR scale was 11.4 (SD = 4.0). Individuals obtaining higher scores also scored significantly higher on the EPQ-Impulsivity scale (r = 0.39, p < 0.001). Before lunch, after controlling for each participant's liking for pizza and their pre-exposure desired portion size in the regression analysis, the results of the regression analyses provided little evidence to suggest that the SR scale was associated with a change in any of the measures of cue reactivity ( Table 2). Similarly, after lunch, this scale was not significantly associated with change in desire to eat pizza, or change in craving for this food ( Table 2). However, participants with higher scores on the SR scale were found to experience a greater increase in their desired pizza portion size following exposure to pizza after lunch ( Table 2). Using the parameter estimates from the regression model for change in desired portion size of the cued food after lunch, we predicted the change in desired portion size after pizza-cue exposure in kcal for an individual obtaining low, medium, and high, scores on the SR scale. In this analysis, we held liking for pizza and pre-exposure portion sizes at their mean values for the sample (73 mm and 6509 mm2 [175.7 kcal] respectively) ( Fig. 1). This revealed that individuals in the lowest tertile of the SR scale changed their desired pizza size by approximately 60 cal compared to approximately 160 kcal for those in the highest tertile of the SR scale. Table 2. Adjusted parameter estimatesa from linear regression models of associations between the SR scale and the three measures of food-cue reactivity before, and after, lunch. SR scale EPQ-Impulsivity scale n B SE p n B SE p Before lunch Change in desire-to-eat pizza (mm) 118 − 0.01 0.41 0.98 118 0.59 0.37 0.11 Change in craving for pizza (mm) 118 − 0.08 0.48 0.87 118 0.88 0.42 0.04⁎ Change in desired portion of pizza (mm2) 118 502.95 307. 24 0.10 118 647.72 271.39 0.02⁎ After lunch Change in desire-to-eat pizza (mm) 118 0.34 0.44 0.43 118 1.05 0.39 0.01⁎ Change in craving for pizza (mm) 118 0.19 0.44 0.28 118 1.16 0.40 < 0.01⁎ Change in desired portion of pizza (mm2) 118 441.61 188.49 0.02⁎ 118 2333.51b 1537.72b 0.13b ⁎ Denotes significance. a Adjusted for liking for pizza and the relevant pre-exposure rating. b Impulsivity scores were split at their median value in the sample to explore the association with change in desired portion size after cue exposure because this association was not linear. Table options Predicted change in pizza size (in kcal) from pre- to post-cue exposure after ... Fig. 1. Predicted change in pizza size (in kcal) from pre- to post-cue exposure after lunch for individuals with low (7.23), medium (11.28), and high (15.95) SR scale scores (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 = 1161.7). 4 Figure options 3.6. Cue reactivity and the EPQ-Impulsivity scale The mean score on the EPQ-Impulsivity scale in the sample was 7.60 (SD = 4.34). Before lunch, findings from the regression model suggested that this scale was not significantly associated with a change in desire for pizza (Table 2). However, after controlling for each participant's liking for pizza and their pre-exposure desired portion size in a regression analysis, higher scores on this scale were associated with a significantly greater change in craving for pizza and desired pizza size (Table 2). Using the parameter estimates from the regression model for change in desired portion size, we predicted the change in desired portion size of pizza in calories for an individual obtaining low, medium, and high, scores on the EPQ-Impulsivity scale. Fig. 2 reveals that those in the lowest tertile for the EPQ-Impulsivity scale changed their desired pizza size by approximately 45 kcal compared to approximately 190 kcal in the highest tertile. Predicted change in pizza size (in kcal) before lunch for individuals with low ... Fig. 2. Predicted change in pizza size (in kcal) before lunch for individuals with low (2.82), medium (7.49), and high (12.65), EPQ-Impulsivity scores (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 = 647.72). 4 Figure options After lunch, change in craving for pizza, and change in desire to eat this food, were both greater in individuals obtaining higher scores on the EPQ-Impulsivity scale (Table 2). Visual inspection of the association between impulsivity scores and change in desired portion size of pizza after lunch revealed a curved rather than linear association. For this reason, impulsivity scores were split at their median value in the sample and entered into the regression model predicting change in desired portion size as a discrete variable. This model provided little evidence to suggest that impulsive individuals experienced a greater change in desired portion of the cued food than less impulsive individuals (Table 2).