No calorie comfort: Viewing and drawing “comfort foods” similarly augment positive mood for those with depression

Abstract

Based on behavioral and neurobiological data, we tested the hypothesis that viewing/drawing visual images of comfort foods in the absence of eating will increase positive mood and that this effect is augmented for those with clinical symptoms of depression. A counterbalanced design was used for 60 participants with and without clinical symptoms in two variations: food image and food art. In each variation, participants viewed/drew foods high or low in fat/sugar; pre-post mood was recorded. Results show a consistent pattern: viewing/drawing comfort foods [food image (95% confidence interval): 2.72–4.85; food art (95% confidence interval): 2.65–4.62] and fruits [food image (95% confidence interval): 1.20–2.23; food art (95% confidence interval): 1.51–2.56] enhanced mood. For comfort foods, mood was augmented for those with clinical symptoms of depression [food image (95% confidence interval): 0.95–3.59; food art (95% confidence interval): 0.97–3.46]. Findings corroborate previous data and reveal a novel finding of augmented mood increases for those with clinical symptoms.

Keywords

depression emotion food obesity treatment

Introduction

The prevalence of obesity and depression has shown a marked rise in recent decades (Blazer, 2003; Djernes, 2006; Zamboni et al., 2005). Explanations for the bidirectional link between obesity and depression reveal environmental (Privitera et al., 2013b), neurobiological (Nestler, 2012), cognitive-behavioral (Privitera et al., 2015b), social-perceptual (Saules et al., 2009), and multi-dimensional homeostatic (Marks, 2015) mechanisms. At present, clinical diagnostic criteria for depression are linked to increased intake of comfort foods in general (Privitera et al., 2013b, 2016; Smith and Ditschun, 2009), which presents a problem for many treatments of depression that often increase appetite (Chapman et al., 2005).

Emotional eating, which is observed among people with and without psychopathology, reflects the tendency for individuals to eat “comfort foods” in response to sad or negative emotions (Dubé et al., 2005; Macht, 2008; Privitera et al., 2016). Experimental investigations have shown increased energy intake in response to negative emotions, largely due to increased intake of high-calorie “comfort” foods (not simply eating more food per se), observed in participants where mood was manipulated (Privitera et al., 2016). Eating comfort foods, which are typically high fat (HF), high sugar foods, promotes positive mood likely by way of increased activity in brain reward regions in response to eating these types of foods (Macht, 2008). However, a growing body of evidence shows that displaying visual images of comfort foods—in the absence of eating—is sufficient to observe increased neural activity in brain reward regions (Frank et al., 2010; Tindell et al., 2009) and to promote positive mood in nonclinical groups (Privitera et al., 2013a, 2013c). The extent to which visual images of comfort foods would similarly promote positive mood in a clinical population with depression is not known—but multiple theories predict that such a finding should be expected to include theories based on clinical neurobiological data (see Nestler, 2012) and based on the assumption that such an intervention, that is, presenting food images (FIs) without eating, can restore affective homeostatic imbalances (Marks, 2015).

The key appetitive characteristic in initial assessments for clinical depression was weight loss (Hamilton, 1960), with reduced appetite considered a key feature of mild/moderate depression, even among individuals designated as the significant “other” (Beck, 1967; Zung et al., 1974); force-feeding was often necessary for those with severe depression (Schuyler, 1974). In these early reports, overeating or weight gain linked to depression was identified as “atypical” and thus treatments for depression that also increased appetite were regarded as beneficial to patients. Today, however, weight gain and increased appetite are “typical,” with the prevalence of depression linked to weight gain and increased appetite being almost 40 percent higher than that of depression without these features, based on a representative US national survey of 43,093 adults (Blanco et al., 2012). This shift has, in part, resulted in a bidirectional link between obesity and depression in recent decades and highlights the need to identify strategies that can promote positive mood in the absence of increased food intake, especially for those with clinical symptoms of depression.

Concomitantly, based on the behavioral and neurobiological evidence showing the positive impact of FIs on mood (Macht, 2008; Privitera et al., 2013a, 2013c), we hypothesized that viewing/drawing visual images of comfort foods would similarly increase positive mood, even in the absence of eating, and that this effect would be greatest or augmented for those with clinical symptoms of depression. We therefore implemented two strategies (called “variations” in this study) aimed at increasing positive short-term mood in the absence of eating/changes in hunger in a clinical and nonclinical sample. In each variation, mood was recorded before and after FIs were viewed/drawn, with comparisons made between clinical and nonclinical groups to evaluate possible implications.

Methods

Participants

A total of 60 university undergraduate students (20 men and 40 women) were recruited through university classroom visits and sign-up sheets at a local university in the Western New York region. All participants completed both variations of the study and were adults who gave written informed consent. Prior to conducting any procedures, in an initial screening phase, participants were weighed and height measures taken from which body mass index (BMI) scores were computed. Participants reported that they were in general good health with no physical or doctor diagnosed food allergies, pregnancy, or dietary restrictions. Participants with and without current depressive symptoms were permitted to enter the study; only 5 percent (N = 3) of those sampled had seen a physician for depression. None of the participants sampled were on any type of psychiatric medications, and no participants had been diagnosed at any time with anorexia nervosa or bulimia nervosa. Participants were told not to eat within 2 hours of each variation because hunger states can influence food choice and intake (Fedorchak and Bolles, 1987; Yeomans, 2006); all participants followed these instructions, as identified in an exit survey following each variation. All participants identified that they were familiar with, had consumed, and generally liked the test foods used in this study. All procedures were approved by the institutional review board at the university in the region where the study was conducted.

General procedures

In two variations, participants were observed during lunch hours between 1100 and 1300 hours in groups of 5–10 participants at a time. Participants were given and signed an informed consent prior to each variation with instructions given for the procedures specific to the variation that day. All participants were observed in each variation, and each variation was a within-subjects design. In each variation, participants rated their hunger on a 7-point scale from 1 (very full) to 7 (very hungry) before and after viewing/drawing foods. Participants also reported changes in mood as described for each variation using an adapted version of the Affect Grid (Russell et al., 1989); this is a valid and reliable single-item scale used to assess mood (pleasant vs unpleasant feelings). On this scale, negative difference scores indicate a decrease in mood; positive difference scores indicate an increase in mood. The order that participants were assigned to each variation was counterbalanced using a Latin square procedure. At the start of the first variation only, participants completed a demographic survey, and the Hamilton Depression Rating Scale (HAMD-17; Hamilton, 1960) was administered by an experienced rater in which scores over 7 on this scale met clinical criterion for depression. After the first variation, all participants agreed to return either 1 or 2 days later, depending on their availability for the next variation. All participants returned for both days. After the final variation was complete, participants were debriefed and thanked for their time.

Variation FI

Procedures used for the FI study were the same as those reported in Privitera et al. (2013a) and are briefly described here. Participants were instructed to orient toward the front of the room at all times and not to speak to one another during the study. All participants followed these instructions. Participants were shown four sets of FI slideshows using Microsoft PowerPoint^® software with each slide image depicting a single food on a plain background that was timed every 9 seconds to transition to a new slide. Slides were formatted to be the same size with no other identifying features on the slide, other than the food itself, and the FIs matched in complexity, valence, and arousal, same as criteria used in prior FI studies (Frank et al., 2010; Privitera et al., 2013a, 2013c). Each slideshow had 15 food pictures/slides for a total of 135 seconds per slideshow. The four sets of food pictures (samples are shown in Figure 1) were foods that were high fat-high sugar (HFHS; i.e. desserts), high fat-low sugar (HFLS; i.e. fried foods), low fat-high sugar (LFHS; i.e. fruits), or low fat-low sugar (LFLS; i.e. vegetables).

Figure 1.

A sample of two food pictures depicted in each of four sets of slideshows for foods shown in Variation 1.

Before each slideshow, participants rated their mood after a brief exercise to prime a neutral emotional state (Privitera et al., 2013a, 2013c), then observed four sets of food picture PowerPoint^® slideshows. At the end of each slideshow, participants were asked to rate their post level of mood. A 20-second inter-slide interval was used with the same procedures repeated starting with the exercise to prime a neutral emotional state until all four slideshows were shown and participants had rated their pre-post mood for each slideshow. A Latin square procedure was used to counterbalance the order that participants observed the slideshows.

Variation food art

Procedures used for the food art (FA) study were the same as those reported in Privitera et al. (2013c) and are briefly described here. Participants were randomly assigned to one of four FA groups: HFHS (stimulus food: cupcake), HFLS (stimulus food: pizza), LFHS (stimulus food: strawberry), or LFLS (stimulus food: pepper). Before each drawing, participants rated their mood after a brief exercise to prime a neutral emotional state (Privitera et al., 2013a, 2013c). They were then given 5 minutes to draw a picture of the stimulus food for the group assigned. Because image color can influence emotional responsiveness (Frank et al., 2010; Seong-in, 2010), participants used only three colors (red, green, and black), were required to use all three colors, and told that the colors they used in their art must reflect actual colors that are natural for the food depicted. All participants followed these instructions. After 5 minutes, participants again rated their mood and arousal on the affect grid. A sample participant drawing of each stimulus food is shown in Figure 2. A 3-minute inter-drawing interval was used to allow time to change canvases and colored pencils; the same procedures were then repeated starting with the exercise to prime a neutral emotional state until all drawings were complete and participants had rated their pre-post mood for each drawing. A Latin square procedure was used to counterbalance the order of the food type that participants drew.

Figure 2.

A sample participant drawing of each stimulus food for each group in Variation 2.

Statistical analysis

Variation FI

A 4 × 2 mixed analysis of variance (ANOVA) was computed with depression (normal, depression) as the between-subjects factor and condition (HFHS, HFLS, LFHS, and LFLS) as the within-subjects factor. Sex (male and female) was initially included as a factor and BMI as a covariate but were removed when both showed no significance with the results reported here. The dependent variable was pre-post difference in mood ratings for each slideshow/condition. A Tukey’s honest significant difference (HSD) was computed for each post hoc analysis, and planned comparison two-independent-sample t tests were computed to analyze significant interactions for the HAMD factor at each level of condition, using a Bonferroni procedure to control for experimentwise alpha. In addition, planned one-sample t tests were computed at each level of the within-subjects factor to check which levels showed significant changes in mood. The null hypothesis for each test was 0 difference/change. A regression analysis was not computed on raw scores of depression due to the restriction of range resulting from scores on the depression scale not being observed across the full range of clinical categories. Hypothesis tests were computed at a .05 experimentwise level of significance.

Variation FA

The same analysis was computed for Variation FA as described for Variation FI, except that the levels of the condition reflect the HFHS, HFLS, LFHS, and LFLS art/drawing groups (not slideshows).

Pre-condition mood ratings and hunger ratings were used as a baseline measure on each variation day. These measures were compared between depression groups across each variation using a mixed multivariate analysis of variance (MANOVA) with depression groups as the between-subjects factor and variation (FA and FI) as the within-subjects factor. Baseline mood ratings and baseline hunger ratings were the dependent variables to check that baseline mood and hunger was similar between groups at the start of each variation day. Within-session changes in hunger were also evaluated.

Results

The 60 participants observed in each variation had the following characteristics: age (20.2 ± 1.3 years), weight (73.8 ± 11.0 kg), and height (168.1 ± 6.8 cm). HAMD-17 scores ranged from 3 to 18 with M ± standard deviation (SD) score of 8.4 ± 4.0. BMI scores ranged from 20.0 to 34.8 with M ± SD score of 24.8 ± 3.9. The analysis to check baseline measures showed that baseline mood did not differ across variations (F(1, 58) = 1.07, p = .31) and did not vary between groups on each variation day (F(1, 58) = 2.41, p = .13). Likewise, self-reported baseline hunger ratings did not differ across variations (F < 1.0, p = .97) and did not vary between groups on each variation day (F < 1.0, p = .53). Within-session changes in hunger ratings also did not differ pre-post and did not differ between groups on each variation day (p > .32 for all tests). Pre-post mood ratings for each food type in each variation by depression category showed a similar pattern of statistical significance in each variation, as summarized in Table 1.

Table 1.

Participant pre-post mood for each food type in each variation by depression category.

Groups	N	Variation food image (FI)				Variation food art (FA)
Groups	N	HFHS	HFLS	LFHS	LFLS	HFHS	HFLS	LFHS	LFLS
Normal	36	2.5 (2.0)	3.2 (2.1)	1.6 (1.8)	−0.4 (2.0)	3.3 (1.7)	2.4 (1.8)	2.1 (2.0)	−0.3 (1.7)
Depression	24	4.7 (2.7)	5.8 (1.5)	2.0 (2.3)	−0.8 (2.6)	5.3 (1.9)	4.7 (2.7)	1.9 (2.2)	−0.1 (2.0)
	60	3.6 (2.3)	4.5 (1.8)	1.8 (2.0)	−0.6 (2.3)	4.3 (1.8)	3.5 (2.2)	2.0 (2.1)	−0.2 (1.9)

SD: standard deviation; HFHS: high fat-high sugar; HFLS: high fat-low sugar; LFHS: low fat-high sugar; LFLS: low fat-low sugar.

Scores given as M (SD).

Variation FI

The analyses showed that pre-post mood varied by condition, F(3, 174) = 64.30, p < .001, R² = .52. One-sample t tests showed that conditions HFHS, t(59) = 10.36, p < .001, d = 1.33 (95% confidence interval (CI): 2.72–4.02), HFLS, t(59) = 14.71, p < .001, d = 1.90 (95% CI: 3.69–4.85), and LFHS, t(59) = 6.65, p < .001, d = 0.86 (95% CI: 1.20–2.23) all effectively increased participant pre-post mood; condition LFLS showed no significance, t(59) = −1.81, p = .07, d = −0.23 (95% CI: −1.12 to 0.06). The analyses further showed a significant main effect of depression, F(1, 58) = 18.44, p < .001, R² = .24; participants with depression (vs without) showed greater increases in mood overall (95% CI: 0.63–1.74), and this effect varied by condition as evident by the interaction, F(3, 174) = 6.59, p < .001, R² = .10, which is elucidated in Figure 3. Planned comparison independent-sample t tests for depression at each condition showed that pre-post mood was effectively augmented for participants with depression (vs without) in the HF conditions—HFHS: t(58) = 3.58, p = .001, d = 0.94 (95% CI: 0.95–3.38); HFLS: t(58) = 5.34, p < .001, d = 1.54 (95% CI: 1.63–3.59). Pre-post mood did not differ by depression category in the low fat (LF) conditions (p > .53 for both LF tests).

Figure 3.

Difference in pre-post mood ratings by depression groups for each condition.

Variation FA

The analyses showed that pre-post mood varied by condition, F(3, 174) = 56.21, p < .001, R² = .49. Same as in Variation FI, one-sample t tests showed that conditions HFHS (cupcake), t(59) = 15.77, p < .001, d = 2.05 (95% CI: 3.58–4.62), HFLS (pizza), t(59) = 10.35, p < .001, d = 1.34 (95% CI: 2.65–3.92), and LFHS (strawberry), t(59) = 7.75, p < .001, d = 1.00 (95% CI: 1.51–2.56) all effectively increased participant pre-post mood; condition LFLS (pepper) showed no significance, t(59) = −0.90, p = .37, d = −0.11 (95% CI: −0.70 to 0.26). The analyses further showed a significant main effect of depression, F(1, 58) = 16.82, p < .001, R² = .23; participants with depression (vs without) showed greater increases in mood overall (95% CI: 0.53–1.56), and this effect varied by condition as evident by the interaction, F(3, 174) = 5.69, p = .001, R² = .09, which is elucidated in Figure 4. Planned comparison independent-sample t tests for depression at each condition showed that pre-post mood was effectively augmented for participants with depression (vs without) in the HF conditions—HFHS: t(58) = 4.06, p < .001, d = 1.06 (95% CI: 0.97–2.86); HFLS: t(58) = 3.98, p < .001, d = 1.04 (95% CI: 1.15–3.46). Pre-post mood did not differ by depression category in the LF conditions (p > .62 for both LF tests).

Figure 4.

Difference in pre-post mood ratings by depression groups for each condition.

Discussion

The results in this study show a consistent pattern across both variations that, when taken together, shows evidence to support the prediction that FIs in the absence of intake can enhance positive mood and that this effect is most prevalent among those with clinical symptoms of depression. Whether participants viewed FIs or drew foods as art, participants consistently showed post-session increases in mood after viewing/drawing HF, high sugar “comfort foods” (i.e. desserts, fried foods) but not when viewing/drawing LF, low sugar vegetables—this behavioral pattern is consistent with neurobiological data showing greater responsiveness in limbic and neural reward circuits linked to both depression and obesity (Nestler, 2012) when viewing high-calorie but not low-calorie FIs (Frank et al., 2010). Post-session increases in mood were augmented for participants with clinical symptoms of depression; these clinical groups showed the greatest mood increases after viewing/drawing comfort foods.

In this study, self-reported hunger ratings showed no differences from pre- to post-session for each food type and showed no differences between groups on each variation day. Thus, positive mood increases were observed in the absence of changes in hunger, with mood increases augmented for those with clinical symptoms of depression. Neural regulation of hunger and “reward” (i.e. enhanced motivational salience for “comfort” foods) are distinct. Homeostatic regulation of hunger and satiety is located primarily in the hypothalamus and brainstem; neuronal circuits in the limbic region largely mediate the motivational salience of food rewards (Ahima and Antwi, 2008). Given that food was not consumed, only viewed, in this study, we would not expect homeostatic regulation of hunger to change over the brief duration of each variation, as was observed. Interestingly, a homeostatic regulatory view of the behavioral pattern in this study is largely consistent with Marks’ Homeostatic Model of Obesity, in that these findings demonstrate regulation of short-term mood enhanced most for those with the greatest homeostatic need (i.e. participants with depression), while changes in hunger were absent, as expected because food was not consumed over the brief duration of each variation (i.e. no homeostatic “correction” was needed).

Evidence suggests that depression is a risk factor for obesity, and obesity confers greater risk of clinical depression or elevated symptoms of depression (Kloiber et al., 2007; Luppino et al., 2010; Novick et al., 2005; Sharma and Fulton, 2013; Simon et al., 2006). From neurobiological data, depression and obesity are known to having overlapping circuitries, such as circuitries that act via the hypothalamic–pituitary–adrenal (HPA) axis (Dallman, 2010; Kyrou and Tsigos, 2009; Schellekens et al., 2013; Ulrich-Lai and Herman, 2009), and via pathways within limbic and neural reward circuits (Nestler, 2012). An important challenge to address the bidirectional relation between obesity and depression is to identify strategies that can promote positive mood without increasing appetite despite these overlapping circuitries. This study presents promising evidence demonstrating two such strategies using two variations with FIs, resulting in increased short-term positive mood in the absence of increased food intake, with positive mood increases augmented for participants with clinical symptoms of depression.

A few important limitations and constraints can be further identified here. First, in this study, food was not available to eat. When an edible food is used as a visual cue, short-term mood increases are evident (Privitera et al., 2015a) although such conclusions cannot be corroborated here. Second, for the food types used, differences cannot be compared between the food types. The foods varied in many characteristics; therefore, it is important that future work look at which specific characteristics of the food themselves (e.g. calories, portion size) may lead to additional effects on mood. Likewise, it is important to note that food intake was not measured in this study and each variation was a single session in each food type category. Therefore, any conclusions regarding food intake or conditioning would be largely inappropriate based on the actual design and measures observed in this study (Privitera, 2008). It is also not possible to generalize beyond the food cues used in this study, especially for Variation FA in which only one staple food in each category was drawn—although this was a necessary control to ensure that the foods could be matched on key characteristics across food types. Studies using a variety of foods varying by culture (e.g. various cuisines) and diet type (e.g. vegetarian) and studies using foods specifically matched to the pre-reported dietary histories of participants can lend further insights into possible limitations and constraints based on the individual differences of participants when using visual food cues to promote positive mood.

In all, these results corroborate a growing body of evidence demonstrating the positive impact of FIs on short-term mood (Frank et al., 2010; Privitera et al., 2013a, 2013c) and extend such findings to identify two strategies that can further augment positive effects on short-term mood for those with clinical symptoms of depression. Given that increased appetite is “typical” among those with depression (Blanco et al., 2012), the results presented here show two possible intervention options to enhance short-term mood without affecting hunger. Future studies will evaluate both the short- and long-term effects of implementing these strategies on mood among clinical and nonclinical groups.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Ahima

Antwi

(2008) Brain regulation of appetite and satiety. Endocrinology and Metabolism Clinics of North America 37: 811–823.

Beck

(1967) Depression: Causes and Treatment. Philadelphia, PA: University of Pennsylvania Press.

Blanco

Vesga-López

Stewart

. (2012) Epidemiology of major depression with atypical features. Journal of Clinical Psychiatry 73: 224–232.

Blazer

(2003) Depression in late life: Review and commentary. Journals of Gerontology Series A: Biological Sciences and Medical Sciences 58: 249–265.

Chapman

Perry

Strine

(2005) The vital link between chronic disease and depressive disorders. Preventing Chronic Disease 2: 1–10.

Dallman

(2010) Stress-induced obesity and the emotional nervous system. Trends in Endocrinology and Metabolism 21: 159–165.

Djernes

(2006) Prevalence and predictions of depression in populations of elderly: A review. Acta Psychiatrica Scandinavica 113: 372–387.

Dubé

LeBel

(2005) Affect asymmetry and comfort food consumption. Physiology & Behavior 86: 559–567.

Fedorchak

Bolles

(1987) Hunger enhances the expression of calorie- but not taste-mediated conditioned flavor preferences. Journal of Experimental Psychology: Animal Behavior Processes 13: 73–79.

10.

Frank

Laharnar

Kullmann

. (2010) Processing of food pictures: Influence of hunger, gender and calorie content. Brain Research 1350: 159–166.

11.

Hamilton

(1960) A rating scale for depression. Journal of Neurology, Neurosurgery & Psychiatry 23: 56–62.

12.

Kloiber

Ising

Reppermund

. (2007) Overweight and obesity affect treatment response in major depression. Biological Psychiatry 62: 321–326.

13.

Kyrou

Tsigos

(2009) Stress hormones: Physiological stress and regulation of metabolism. Current Opinion in Pharmacology 9: 787–793.

14.

Luppino

De Wit

Bouvy

. (2010) Overweight, obesity, and depression: A systematic review and meta-analysis of longitudinal studies. Archives of General Psychiatry 67: 220–229.

15.

Macht

(2008) How emotions affect eating: A five-way model. Appetite 50: 1–11.

16.

Marks

(2015) Homeostatic theory of obesity. Health Psychology Open 2: 1–30.

17.

Nestler

(2012) Epigenetics: Stress makes its molecular mark. Nature 490: 171–172.

18.

Novick

Stewart

Wisniewski

. (2005) Clinical and demographic features of atypical depression in outpatients with major depressive disorder: Preliminary findings from STAR*D. Journal of Clinical Psychiatry 66: 1002–1011.

19.

Privitera

(2008) The Psychological Dieter: It’s Not All About the Calories. Lanham, MD: University Press of America.

20.

Privitera

Antonelli

Creary

(2013a) The effect of food images on mood and arousal depends on dietary histories and the fat and sugar content of foods depicted. Journal of Behavioral and Brain Science 3: 1–6.

21.

Privitera

Cuifolo

King-Shepard

(2015a) Wait, bring it back! “Expectancy to eat” moderates the effectiveness of food cues to improve mood. International Journal of Applied Psychology 5: 21–25.

22.

Privitera

King-Shepard

Cuifolo

. (2016) Differential food intake and food choice by depression and body mass index levels following a mood manipulation in a buffet-style setting. Journal of Health Psychology. Epub ahead of print 20 May. DOI: 10.1177/1359105316650508.

23.

Privitera

McGrath

Windus

. (2015b) Eat now or later: Self-control as an overlapping cognitive mechanism of depression and obesity. PLoS ONE 10: e0123136.

24.

Privitera

Misenheimer

Doraiswamy

(2013b) From weight loss to weight gain: Appetite changes in major depressive disorder as a mirror into brain-environment interactions. Frontiers in Psychology 4: 873.

25.

Privitera

Moshaty

Marzullo

. (2013c) Expressing food through art: Evidence for a nutrient-specific effect on mood. Journal of Behavioral and Brain Science 3: 163–167.

26.

Russell

Weiss

Mendelsohn

(1989) Affect grid: A single-item scale of pleasure and arousal. Journal of Personality and Social Psychology 57: 493–502.

27.

Saules

Collings

Hoodin

. (2009) The contributions of weight problem perception, BMI, gender, mood, and smoking status, to binge eating among college students. Eating Behaviors 10: 1–9.

28.

Schellekens

Dinan

Cryan

(2013) Taking two to tango: A role for ghrelin receptor heterodimerization in stress and reward. Frontiers in Neuroscience 7: 148.

29.

Schuyler

(1974) The Depression Spectrum. New York: Jason Aranson.

30.

Seong-in

(2010) A computer system for the analysis of color-related elements in art therapy assessment: Computer_Color-Related Elements Art Therapy Evaluation System (C_CREATES). The Arts & Psychotherapy 37: 378–386.

31.

Sharma

Fulton

(2013) Diet-induced obesity promotes depressive-like behaviour that is associated with neural adaptations in brain reward circuitry. International Journal of Obesity 37: 382–389.

32.

Simon

Von Korff

Saunders

. (2006) Association between obesity and psychiatric disorders in the US adult population. Archives of General Psychiatry 63: 824–830.

33.

Smith

Ditschun

(2009) Controlling satiety: How environmental factors influence food intake. Trends in Food Science & Technology 20: 271–277.

34.

Tindell

Smith

Berridge

. (2009) Dynamic computation of incentive salience: “Wanting” what was never “liked.” Journal of Neuroscience 29: 12220–12228.

35.

Ulrich-Lai

Herman

(2009) Neural regulation of endocrine and autonomic stress responses. Nature Reviews Neuroscience 10: 397–409.

36.

Yeomans

(2006) The role of learning in development of food preferences. In: Shepherd

Raats

(eds) Psychology of Food Choice. Oxford: CABI, pp. 93–106.

37.

Zamboni

Mazzali

Zoico

. (2005) Health consequences of obesity in the elderly: A review of four unresolved questions. International Journal of Obesity 29: 1011–1029.

38.

Zung

Coppedge

Green

(1974) The evaluation of depression symptoms: A triadic approach. Psychotherapy and Psychosomatics 24: 170–174.