An implicit priming intervention alters brain and behavioral responses to high-calorie foods: a randomized controlled study

Kristina T Legget; Marc-Andre Cornier; Christina Erpelding; Benjamin P Lawful; Joshua J Bear; Eugene Kronberg; Jason R Tregellas

doi:10.1093/ajcn/nqac009

. 2022 Jan 14;115(4):1194–1204. doi: 10.1093/ajcn/nqac009

An implicit priming intervention alters brain and behavioral responses to high-calorie foods: a randomized controlled study

Kristina T Legget ^1,^2,^✉, Marc-Andre Cornier ^3,^4,^5,⁶, Christina Erpelding ⁷, Benjamin P Lawful ⁸, Joshua J Bear ^9,¹⁰, Eugene Kronberg ^11,¹², Jason R Tregellas ^13,¹⁴

PMCID: PMC8970978 PMID: 35030242

ABSTRACT

Background

Conditioned food cues (e.g., smell, sight) can affect intake of foods associated with those cues, regardless of homeostatic need. As such, altering automatic associations with food cues could support weight loss or maintenance efforts by affecting the salience of those cues and the effort required to resist consumption.

Objectives

This study investigated neuronal and behavioral effects of an implicit priming (IP) intervention, in which negatively valenced images were paired with high-calorie foods and positively valenced images with low-calorie foods. Priming images were presented immediately before food images, but below conscious perception (20 ms). We hypothesized that this evaluative conditioning approach could alter food cue responses by modifying affective associations.

Methods

The final sample included 41 adults with BMI ≥25 kg/m² (n = 22, active IP; n = 19, control IP). In control IP, food images were primed with neutral, scrambled images. Participants completed a visual food cue task during fMRI, both before and after IP. To determine the replicability of prior behavioral findings, food image ratings were completed before and after IP as a secondary outcome.

Results

In a whole-brain analysis, reduced dorsolateral prefrontal cortex (dlPFC) response to high-calorie foods was observed after active compared with control IP (t = 4.93, P = 0.033). With a region of interest analysis, reduced response to high-calorie foods in active compared with control IP was also observed in the striatum (t = 2.40, P = 0.009) and insula (t = 2.38, P = 0.010). Active compared with control IP was associated with reduced high-calorie food ratings (F = 4.70, P = 0.038).

Conclusions

Reduced insula and striatum response to high-calorie foods after active compared with control IP suggests effectiveness of IP in altering food cue salience. Reduced dlPFC response to high-calorie foods after active compared with control IP may reflect fewer attentional resources being directed to those images and reduced engagement of inhibitory processes.

This trial was registered at clinicaltrials.gov as NCT02347527.

Keywords: obesity, overweight, disgust, implicit priming, evaluative conditioning, functional magnetic resonance imaging, fMRI

Introduction

Given high overweight/obesity rates (1), developing novel approaches to support weight loss and maintenance is critically important for public health. One contributing factor to obesity may be responsivity to food cues. Conditioned food cues, including visual, olfactory, and situational cues, such as the smell of pizza, the sight of chocolate cake, or entering the coffee shop where one often enjoys a morning pastry, can elicit cravings for foods associated with those cues, which can increase consumption regardless of homeostatic need (2–4). As such, altering automatic associations to food cues may represent a powerful method of altering food preferences and intake.

One potential way to alter these associations may be via evaluative conditioning (EC), which involves altering the valence of a given stimulus [conditioned stimulus (CS)] by pairing it with a positively or negatively valenced stimulus [unconditioned stimulus (US)] (5, 6). EC can be implemented using a number of different approaches. Some strategies involve awareness of the pairings between CS and US (e.g., stimuli are explicitly presented and relations between them are clear) and others use implicit approaches, in which participants are unaware of CS–US relationships. Common methods for the latter involve concurrent or subliminal stimulus presentation, such that CS–US pairings are not obvious (5–7). The aim is to transfer the affective response elicited by the US to the CS, such that the CS comes to elicit that response itself. As per the implicit misattribution model of EC, this is accomplished by the US-elicited response being mistakenly attributed as originating from the CS, which promotes EC in this context (6, 7).

The current study investigated an EC approach that used implicit priming (IP) in an attempt to alter responses to food cues by priming food images with positively or negatively valenced images presented immediately beforehand, but not consciously perceived (i.e., subliminal). Low-calorie foods were paired with positively valenced images, such as playful puppies, whereas high-calorie foods were paired with negatively valenced images. Negative images were selected to elicit disgust, either through contamination (e.g., insects on food; vomit) or through mutilation (e.g., severed limb, burned flesh). Disgust is a particularly potent motivator of food preference, shown to reliably produce conditioned food aversions (8–10).

Our preliminary behavioral study of this intervention found active IP, compared with control IP (using scrambled images as primes), to be associated with reduced ratings of “desire to eat” high-calorie foods (11). The current study expanded upon this work to investigate effects of this novel intervention on the neuronal response to visual food cues in adults with overweight/obesity, using fMRI. Given preliminary behavioral effects, we hypothesized that active IP would be associated with reduced responsivity to high-calorie food images in brain regions associated with salience and reward processing. Food ratings were included as secondary outcomes, hypothesizing reduced high-calorie ratings after active compared with control IP. Although effects on low-calorie ratings were not observed in our preliminary work, these effects were also examined in the present study, as increasing low-calorie food preferences in combination with reducing the desire to eat high-calorie foods may represent an ideal strategy to promote healthy eating.

Methods

Participants

A total of 47 adults 18–65 y old with BMI ≥25 kg/m² completed the study. Of those 47 participants, 25 were randomly assigned to the active IP group and 22 to the control group. Participants were recruited via email and flyer advertisement in Aurora, CO, between May 2015 and August 2019. All study activities were completed at the University of Colorado Anschutz Medical Campus in Aurora, CO. Individuals reporting current dieting for weight loss, following a vegan or vegetarian diet, being a self-described “picky eater,” or having other wide-ranging food restrictions were excluded. Exclusion criteria also included the following: history of bariatric surgery; current use of medications affecting weight, appetite, or metabolism (e.g., oral steroids); history of major medical condition affecting weight/metabolism; self-reported eating disorder or score >20 on the Eating Attitudes Test (EAT-26) (12); current nicotine use; current pregnancy, having given birth in the past 6 mo, or nursing within the past 6 mo; and MRI contraindications (e.g., weight >500 lbs/227 kg, claustrophobia, metal in the body). One participant was unable to complete the MRI scan owing to technical difficulties and 1 was unable to complete the MRI scan owing to anxiety. Four additional participants were excluded from all analyses owing to excess movement (>2 mm) during scanning (n = 2), technical difficulties during scanning (n = 1), or noncompliance with study procedures (n = 1). The final sample consisted of 41 participants [22 in the active group (4 males, 18 females), 19 in the control group (9 males, 10 females); see Supplemental Figure 1 for a participant flowchart]. Of these participants, 1 was excluded from fMRI analyses for the visual food cues task only and 1 was excluded from fMRI analyses for the IP task only (task details follow), both owing to excess movement during those tasks, resulting in a sample size of 40 for each task [21 in the active group (4 males, 17 females), 19 in the control group (9 males, 10 females)]. Participants provided written informed consent and all procedures were in accordance with and approved by the Colorado Multiple Institutional Review Board. The clinicaltrials.gov registration (NCT02347527) encompasses both the current study and the preliminary behavioral study involving this intervention (11). These studies involved different participants, with those in the preliminary study recruited during 2013–2014 (11) and those in the current study during 2015–2019.

Sample size calculations [G*Power 3.1, Heinrich-Heine University Dusseldorf (13)] were based on our preliminary fMRI data (unpublished), demonstrating reduced neuronal response to visual food cues post-IP compared with pre-IP. A reduction in insula response post-priming (compared with pre-priming) to high-calorie foods compared with low-calorie foods was observed (mean ± SD: 0.35 ± 0.38), suggesting a power of 0.80 to detect this interaction with 20 subjects/group. We based power for the active/control comparison on a study on effects of cognitive reappraisal on neuronal responses to food cues (14). Comparing percent signal change in insula response for passive viewing (mean ± SD: 0.05 ± 0.06) with cognitive reappraisal (mean ± SD: 0.11 ± 0.04), similar effects with our intervention would yield a power of 0.90 to detect group differences with 17 subjects/group. As such, we estimated that a sample size of 17–20 subjects/group in the current study would yield ≥80% power (2-tailed α = 0.05) to detect similar group differences.

Study design and randomization

The study utilized a double-blind, parallel, randomized design and was conducted over the course of 3 research visits (see Figure 1). As will be described in detail, participants first completed a eucaloric run-in diet day, followed immediately by the main study day. A follow-up visit was completed 3–6 d after the study day. On the morning of the study day, participants were randomly assigned to either active IP or control IP (details on both interventions follow). Random assignment utilized a 1:1 ratio using a random number sequence generated by JRT, who did not interact with participants. Group assignment was coded as either “A” or “B,” with research staff interacting with participants unaware of which letter was associated with which intervention. The primary outcome measure for this study was the change in brain response to high-calorie visual food cues from baseline to post-intervention (details follow). Secondary outcome measures were 1) change in brain response to low-calorie visual food cues from baseline to post-intervention, 2) brain response to high-calorie food cues during the IP paradigm, and 3) change in behavioral ratings of food images (both high- and low-calorie) from baseline to post-intervention. Finally, exploratory analyses were conducted to assess 1) associations between change in food image ratings and subsequent food intake, and 2) the effects of intervention (active compared with control) and food type (high-calorie or low-calorie) on response time to categorize foods as healthy or unhealthy during the IP paradigm (details follow).

Eucaloric run-in diet day

The first study visit included a 1-d eucaloric run-in diet prepared by the metabolic kitchen of the Anschutz Health and Wellness Center (AHWC) at the University of Colorado Anschutz Medical Campus. The purpose of this controlled eucaloric diet was to ensure some degree of energy and macronutrient balance before the study day. Estimates of daily energy needs were calculated using estimated resting energy expenditure (15), multiplied by an activity factor (based on self-reported physical activity). Meals across the run-in diet day were macronutrient-controlled, with 50% carbohydrates, 30% fat, and 20% protein. Participants arrived at the AHWC after an overnight fast (since 22:00 the night before) and ate breakfast during the visit. They were sent home with food for lunch, dinner, and snacks, with instructions to consume all of the food provided, but not to eat anything additional or consume any additional beverages containing calories. The take-home meals were to be finished no later than 22:00 that evening to begin the overnight fast for the main study day. Participants who were unable to finish any part of the meals were asked to return the remaining food to be weighed out by the kitchen staff. The Three Factor Eating Questionnaire (TFEQ) (16) and Power of Food Scale (PFS) (17) were administered at this visit, to measure eating-related behavioral traits.

Study day

The second visit was completed the day after the eucaloric run-in diet day. Participants arrived at the AHWC the morning after an overnight fast. Participants who regularly consumed caffeinated beverages were asked to refrain from consuming anything with calories (e.g., coffee with cream or sugar added) and to abstain from caffeine consumption within 2 h of their scheduled MRI scan (because participants arrived for the study visit 2 h before the scan, research staff were able to confirm caffeine abstention). Visual analog scale (VAS) measures (from 0 to 100 mm) of fasting hunger (“how hungry are you?” from “not at all hungry” to “extremely hungry”), satiety (“how full do you feel right now?” from “not at all full” to “extremely full”), and prospective food consumption (“how much food do you think you could eat right now?” from “nothing at all” to “a large amount”) were collected. Participants then completed a food image ratings task. For this task, participants rated 96 food images [48 high-calorie (e.g., chocolate cake, cheeseburger, nachos, potato chips, cinnamon roll, pizza) and 48 low-calorie (e.g., mixed berries, omelet, roasted carrots, kiwi, grilled chicken with asparagus, melon balls)] on “desire to eat” [VAS from 0 (“no desire”) to 100 (“strong desire”)] using the ImageRate program (18). “Desire to eat” was defined as “the degree to which a person wants to eat that food at that time.” Of the 48 images in each category, 32 were also included in the IP paradigm (details follow), and 16 were only viewed during the image ratings task (see Supplemental Figure 2). This design allowed us to assess generalizability of effects, i.e., to assess the ratings of images that had been primed as well as those that had not been directly paired with the priming stimuli (termed “novel” images). Food images for both tasks were selected from various websites (copyright-free images), from the International Affective Picture System (IAPS) (19), and from a pre-existing set of food images (18). Because of technical difficulties, image rating data collection was not completed for 5 participants (final sample size for image rating data: active, n = 19; control, n = 17). After completing food image ratings, participants consumed a standardized breakfast meal comprising 25% of their estimated total daily energy requirements (macronutrient composition matched that of the run-in diet day). VAS measures of hunger, satiety, and prospective food consumption were repeated immediately after breakfast.

The MRI session (details follow) was completed 1.5 h after breakfast, with the goal of having participants in a neutral state of hunger (i.e., neither fasted nor acutely fed), in concordance with the satiety state in our preliminary behavioral study assessing this intervention (11). Within 5 min before the scan starting, food cravings were assessed with the Food Cravings Questionnaire-State (FCQ-S) (20) and VAS measures of hunger, satiety, and prospective food consumption were repeated.

Body composition was assessed via BOD POD air displacement plethysmography (COSMED). Participants abstained from food and liquids for ≥2 h before body composition assessment, which was performed after the MRI for all participants except 1 in the active group, who completed the body composition assessment at the start of the day, because of scheduling conflicts. To assess food intake, participants were given lunch in the form of an ad libitum buffet meal, 3.5 h after breakfast (1 h after the MRI scan). The lunch buffet was designed to constitute 40% of estimated total daily energy requirements, plus 30% more than this estimate. VAS measures of hunger, satiety, and prospective food consumption were repeated before and immediately after eating lunch. Before eating lunch, participants repeated the food image ratings task. The same images as in the first image ratings task were used, but presented in a different random order to reduce recall effects. After lunch, sensitivity to disgust was assessed using the Disgust Scale-Revised (DS-R) (21).

Follow-up visit

Participants were asked to return to the AHWC 3–6 d after the study day (mean ± SEM days between study day and follow-up: active group, 4.62 ± 0.22; control group, 4.25 ± 0.27; n = 16 for both) to assess lasting effects of the intervention. Participants arrived after an overnight fast and repeated the food image ratings task again (images were presented in a different random order than the first 2 administrations of the task).

fMRI data acquisition

Imaging was performed using a Siemens Skyra 3.0 T MR system, with a 20-channel head coil. A T1-weighted magnetization-prepared rapid acquisition gradient-echo (MPRAGE) anatomical image was acquired for each participant [repetition time (TR) = 2300 ms, echo time (TE) = 2.45 ms, inversion time (TI) = 900 ms, flip angle = 9°, matrix = 256 × 256, field of view (FOV) = 220 mm, slice thickness = 1.2 mm, 192 slices]. After this, functional images were acquired with a gradient-echo T2*-weighted BOLD imaging contrast technique, with the following parameters: TR = 2000 ms, TE = 30 ms, flip angle = 70°, matrix = 64 × 64, FOV = 220 ms, slice thickness = 2.6 mm thick, 1.4-mm gap, 30 axial slices angled parallel to the planum sphenoidale. Visual stimuli were presented using E-Prime 2.0 (Psychology Software Tools) and a magnetic resonance–compatible goggle system (Resonance Technology, Inc.).

Visual food cues task

Both before and after the IP intervention (description to follow), functional imaging was performed while participants viewed images of foods (high-calorie and low-calorie) and nonfood objects, to assess intervention effects on response to visual food cues. The same images were used in both sessions (pre-IP and post-IP), but presented in a different order to reduce habituation. The food images used during this task (48 high-calorie, 48 low-calorie) were different from those used during the behavioral food image ratings task, also to reduce the potential for habituation. As with the behavioral food image ratings task, 32 of the food images in each category (high-calorie and low-calorie) used in this task were also included in the IP intervention and 16 in each category were unique to this task (i.e., not included in the IP intervention or in the behavioral image rating task), to allow for assessment of effect generalization (i.e., if effects extend to images not directly primed). See Supplemental Figure 2 for a depiction of the number of images in each task and the overlap between them. Copyright-free images were selected from various websites, in addition to images from the IAPS database and a pre-existing set of food images (18). For each session (pre-IP and post-IP), 2 runs were completed, with a brief break (∼5–10 s) between runs. A blocked design was used (fixed random order), with a total of 12 blocks of high-calorie food images, 12 blocks of low-calorie food images, and 12 blocks of nonfood-object images (neutrally rated IAPS object images) across both runs for each session (pre-IP and post-IP). Each block included 4 images, presented for 4 s each (for a total of 16 s per block). Sessions also included 6 blocks of a baseline fixation condition (fixation cross). Four additional volumes were automatically acquired and discarded before all fMRI runs to minimize saturation effects. Participants were asked to lie quietly while viewing the images.

IP intervention

During the IP intervention, participants also viewed food images (80 high-calorie, 80 low-calorie) during functional imaging. During this task, however, all food images were preceded by a 500-ms fixation cross and a 20-ms implicit prime (see Figure 2), as detailed in our previous behavioral study with this paradigm (11). Participants were informed of the presence of the prime, but did not know if they were in the active or control group. Research staff interacting with participants were also blind to group assignment (the randomization scheme instructed them to run either paradigm A or B for a given participant, but research staff did not know which paradigm was which). In the active group, the primes were either negatively valenced images (paired with high-calorie foods) or positively valenced images (paired with low-calorie foods). The images used as primes were all selected from the IAPS database, with “negative” images largely chosen to elicit disgust (e.g., mutilated flesh, a bleeding wound, vomit) and having an IAPS emotion rating between 1.5–3.99, and “positive” images being those with an IAPS emotion rating >6 (e.g., smiling people, puppies, happy babies). Of the 80 food images per category (high-calorie and low-calorie) in the IP paradigm, 32 were included in the behavioral image ratings task, 32 were included in the visual food cues fMRI task, and 16 were unique to the IP paradigm (see Supplemental Figure 2). As such, participants only viewed each food image a maximum of 3 times during the study day, to minimize potential habituation effects. In the control group, the primes were scrambled, unrecognizable images that matched the complexity and brightness of the active IP images. In both IP conditions, participants saw a brief flash before each food image, but were not able to discern image content. Participant inability to discern the active/control conditions, as well as high-speed photographic confirmation of timing accuracy of the 20-ms prime presentation, were ascertained during task development and in our previous behavioral study (11). In addition, participants completed a debriefing questionnaire after the completion of the MRI session, to determine if primes were consciously detected. No participants reported seeing any of the negative primes. Twelve participants in the active IP group reported seeing ≥1 of the positive primes, although none of them reported seeing >4 positive priming images in total (5 reported seeing Minnie or Mickey Mouse, 10 reported seeing a baby, 4 reported seeing “people,” and 1 reported seeing an “80s woman”). To ensure that participants were focusing on and actively thinking about each presented food image, they were instructed to indicate by button press, as quickly as possible, if they would categorize the food image as “healthy” or “not healthy.” These categories were chosen in an effort to support generalization of effects; if participants focus on how each food image fits within the constructs of healthy/not healthy, rather than simply focusing on specific image features, it may promote generalization to similar foods within those constructs [e.g., (22)]. Participants were informed that the images would progress regardless of button press timing (i.e., timing of image presentation was consistent, regardless of response time). The intervention consisted of a single run (∼11 min), with 20 blocks of high-calorie and 20 blocks of low-calorie food images (presented in a fixed random order). Each block included 4 prime/food image pairings, each lasting 4 s (including a 500-ms fixation cross and 20-ms prime), for a total of 16 s/block. Sessions also included 2 blocks of a baseline fixation condition (fixation cross), for a total of 16 s. Four additional volumes were automatically acquired and discarded before all fMRI runs to minimize saturation effects.

Data analyses

fMRI data were preprocessed and analyzed using SPM12 (Wellcome Centre for Human Neuroimaging; www.fil.ion.ucl.ac.uk/spm/). Functional data were realigned to the first echo-planar image, coregistered to the individual's T1 anatomical image, normalized to Montreal Neurological Institute (MNI) standard space, and smoothed with an 8-mm full width at half maximum Gaussian kernel. After preprocessing, voxel size was 3 × 3 × 3 mm. Movement parameters derived from the realignment procedures were included in the model to reduce the effects of residual motion-related noise. The hemodynamic response was modeled with a double γ function, without temporal derivatives, using the general linear model in SPM12. A 128-s high-pass filter was applied to remove low-frequency fluctuation in the BOLD signal. To account for within-group and within-subject variance, a random-effects analysis was implemented. The primary analysis focused on the visual food cues task, for which parameter estimates for each individual's first-level analysis (SPM contrast images) contrasting high-calorie food cues to nonfood objects were entered into second-level repeated-measures ANOVA. A secondary analysis contrasted low-calorie food cues to nonfood objects, testing the hypothesis that active IP would be associated with increased responsivity to low-calorie food images. Group comparisons (active compared with control) were evaluated using directional contrasts (SPM t-contrasts) in a full factorial model. Given that the intention of the IP intervention pairings with high-calorie foods is to target disgust, we anticipated individual propensity to disgust would affect intervention responsivity. To control for this, we included individual ratings of disgust propensity (DS-R score) as a regressor in analyses. Results were considered significant at a whole-brain level if they exceeded a voxel-level family-wise error (FWE) threshold of P < 0.05 (P_FWE).

In addition to a whole-brain analysis, we tested intervention (group-by-time) effects using a region of interest (ROI) analysis in the insula, striatum, and amygdala, brain regions of particular relevance to this study owing to their established roles in both food-related reward processing (23, 24) and disgust (25–27). As a secondary analysis, we investigated the response to high-calorie food cues (relative to an implicit baseline) during the IP paradigm itself, focusing on the same ROIs (insula, striatum, and amygdala). Bilateral ROIs were defined anatomically, using a probabilistic atlas (28). The MarsBaR toolbox (http://marsbar.sourceforge.net/) in SPM was used to extract the mean value across all voxels for each ROI. ROI results were corrected for multiple comparisons using Bonferroni correction for an α of 0.05 for 3 ROIs, resulting in a statistical significance threshold of P = 0.017.

Analyses of demographic, behavioral, and body composition measures were performed with SPSS 27 (IBM Corp.). Independent-samples t tests were used to assess group differences (active compared with control) in age, body composition measures (BMI, lean body mass, fat mass, body fat), and behavioral measures (TFEQ, PFS, FCQ-S, VAS ratings, DS-R), with an α of 0.05. Group differences in sex were assessed with a chi-square test. Effects of intervention (active compared with control) on food image ratings (pre- compared with post-treatment) were assessed using repeated-measures ANOVA, with an α of 0.05. Effects of intervention (active compared with control) and food category (high- compared with low-calorie) on behavioral response time to categorize foods as “healthy” or “unhealthy” during the priming paradigm were also assessed using repeated-measures ANOVA (α = 0.05). As with the fMRI analyses, individual ratings of disgust propensity (DS-R scores) were included as a covariate in repeated-measures ANOVA analyses. Pearson's correlations were used to determine if changes in food image ratings were related to subsequent food intake.

Results

Table 1 reports participant characteristics and behavioral measures. By chance, the active group had greater percent body fat and fat mass than the control group, but these characteristics were not significantly correlated with baseline behavioral food ratings or change in food ratings (P > 0.05) (see the Supplemental Results for details). Also by chance, women comprised a greater percentage of the active group than the control group. This was likely a contributing factor to the increased percent body fat and fat mass observed in the active group, because significantly greater percent body fat [t (39) = 7.19, P < 0.001] and fat mass [t (39) = 3.94, P < 0.001] were observed in female than in male participants in the study. There were no sex differences observed, however, in baseline behavioral food ratings [high-calorie: t (34) = 1.64, P = 0.110; low-calorie: t (34) = −0.48, P = 0.634] or change in food ratings [high-calorie: t (34) = 0.23, P = 0.816; low-calorie: t (34) = 0.11, P = 0.910].

TABLE 1.

Participant characteristics and behavioral measures¹

	Group		Group differences
Characteristic	Active (n = 22)	Control (n = 19)	P
Age,² y	41.5 ± 3.1	37.3 ± 3.1	0.347
BMI,² kg/m²	30.8 ± 1.2	28.8 ± 0.7	0.172
Sex,³ % female	81.8	52.6	0.045
Lean body mass,² kg	114.4 ± 5.5	125.6 ± 5.5	0.158
Fat mass,² kg	80.6 ± 5.5	63.8 ± 5.4	0.036
Body fat,² %	41.0 ± 1.6	33.5 ± 2.5	0.013
TFEQ: Restraint²	8.1 ± 0.9	9.8 ± 1.2	0.255
TFEQ: Disinhibition²	7.6 ± 0.7	7.0 ± 0.6	0.581
TFEQ: Hunger²	5.3 ± 0.6	4.9 ± 0.8	0.675
PFS²	58.7 ± 2.9	53.4 ± 3.3	0.282
FCQ-S²	32.3 ± 2.1	32.5 ± 1.7	0.941
DS-R²	45.4 ± 3.4	38.1 ± 2.8	0.113

Open in a new tab

Values are mean ± SEM unless otherwise indicated. DS-R, Disgust Scale-Revised; FCQ-S, Food Cravings Questionnaire-State; PFS, Power of Food Scale; TFEQ, Three Factor Eating Questionnaire.

Group differences examined using independent-samples t test.

Group differences examined using chi-square test.

Behavioral food ratings, food intake, and appetite

Participant ratings of “desire to eat” high-calorie food images (active group, n = 19; control group, n = 17), which were administered both before and after the priming intervention, demonstrated a significant time-by-group interaction [F (1,33) = 4.70, P = 0.038], such that the active group reduced “desire to eat” high-calorie foods compared with the control group after the intervention (Figure 3). When focusing on only novel high-calorie food images (i.e., those not included in the priming paradigm), there was a similar time-by-group interaction [F (1,33) = 7.64, P = 0.009], with a decrease in image ratings in the active group compared with the control group. A significant time-by-group interaction was also observed for low-calorie food ratings [F (1,33) = 9.02, P = 0.005], with an increase in “desire to eat” low-calorie foods in the control group, compared with the active group, after the intervention. For novel low-calorie food images, a similar time-by-group interaction was observed [F (1,33) = 10.83, P = 0.002], with an increase in ratings in the control group compared with the active group. Effects do not appear to have lasted through the follow-up period (active group, n = 16; control group, n = 16), because we did not observe significant time-by-group interactions when focusing on change from pre-intervention ratings to follow-up ratings for either high-calorie foods [F (1,29) = 0.07, P = 0.799] or low-calorie foods [F (1,29) = 0.86, P = 0.361]. No group differences were observed at any of the measurement times for hunger, satiety, or prospective food consumption (P > 0.05 for all) (see Supplemental Table 1). In the active IP group, greater reduction in high-calorie food ratings was associated with reduced ad libitum intake during lunch (r = 0.47, P = 0.040), an effect not observed in the control group (r = −0.03, P = 0.921) (active group, n = 19; control group, n = 17). Change in low-calorie food ratings was not associated with caloric intake during lunch for either group (active: r = 0.26, P = 0.285; control: r = 0.15, P = 0.564).

Change in mean ± SEM high-calorie and low-calorie food image “desire to eat” ratings from pre- to post-intervention for active (n = 19) and control (n = 17) groups. Ratings of “desire to eat” high-calorie food images demonstrated a significant time-by-group interaction [F (1,33) = 4.70, P = 0.038], such that the active group reduced high-calorie food ratings compared with the control group after the IP intervention. A significant time-by-group interaction was also observed for low-calorie food ratings [F (1,33) = 9.02, P = 0.005], with increased low-calorie food ratings in the control group compared with the active group after IP. IP, implicit priming.

Baseline neuronal response to visual food cues

At baseline, across all participants, viewing high-calorie food images (compared with objects) engaged similar regions to those we have observed in previous studies with a similar visual food cue paradigm (29, 30), including the precuneus, posterior cingulate cortex, lateral parietal cortex, thalamus, and insula (Table 2) (whole-brain analysis, FWE voxel-level corrected P < 0.05).

TABLE 2.

Baseline response to visual food cues (high-calorie food images > objects), across all participants¹

Brain region	MNI coordinates x, y, z²			k ³	t value⁴	P _FWE ⁵ (voxel-level)	P _FWE ⁶ (cluster-level)
Precuneus (L)	−9	−67	41	603	7.60	<0.001	<0.001
Posterior cingulate cortex (L)	−3	−25	32	—	6.23	0.003	—
Lateral parietal cortex (L)	−48	−37	38	389	5.92	0.008	<0.001
Lateral parietal cortex (R)	54	−52	47	403	5.51	0.023	<0.001
Lateral parietal cortex (R)	39	−39	41	—	5.34	0.036	—
Insula (L)	−39	−4	8	53	5.52	0.022	0.163
Thalamus (R)	3	−25	5	151	5.33	0.027	0.008

Open in a new tab

n = 40. One-sample t tests. FWE, family-wise error; k, cluster size; L, left hemisphere; MNI, Montreal Neurological Institute; R, right hemisphere.

Stereotactic coordinates in MNI space.

k at P < 0.001, uncorrected.

⁴

t values reported for local maxima within clusters.

⁵

FWE voxel-level corrected P values.

⁶

FWE cluster-level corrected P values.

IP effects on neuronal response to visual food cues

In a whole-brain analysis, we observed a significant group- (active compared with control) by-time (pre- compared with post-IP) interaction, with reduced dorsolateral prefrontal cortex (dlPFC) response to high-calorie food images (compared with nonfood objects) in the active compared with control group (active group, n = 21; control group, n = 19), after the IP intervention (Figure 4) (MNI peak voxel coordinates x, y, z (mm): −39, 29, 23; t = 4.93, P_FWE = 0.033, d = 1.60; MNI peak voxel coordinates x, y, z (mm): −36, 44, −1; t = 4.89, P_FWE = 0.038, d = 1.59; both local maxima included in 1 cluster of 311 voxels, cluster-level P_FWE < 0.001). In the ROI analysis, a similar significant group-by-time interaction was observed, with reduced response to high-calorie food images (compared with nonfood objects) in the striatum (t = 2.40, P = 0.009, d = 0.71) and insula (t = 2.38, P = 0.010, d = 0.72) after the IP intervention in the active compared with control group. A similar trend was observed in the amygdala ROI (t = 1.89, P = 0.032, d = 0.60). A secondary analysis tested if pairing positively valenced images with low-calorie foods would increase responsivity to food cues; no significant effects were observed for this in either the whole-brain or ROI analyses (all P > 0.05) (see the Supplemental Results). Given the observed group differences in sex distribution and percent body fat, analyses of IP effects on the neuronal response to visual food cues were repeated controlling for these variables. Controlling for sex or percent body fat did not affect the study results (see Supplemental Table 2).

Greater reduction in response to high-calorie food cues after active IP (n = 21) than after control IP (n = 19). The difference in neuronal response to high-calorie food cues (compared with nonfood objects) after active IP compared with control IP (compared with pre-intervention), is shown (assessed via repeated-measures ANOVA). Greater post-intervention reduction in response was observed in the active compared with the control group in the dorsolateral prefrontal cortex, insula, and striatum. Statistical maps use a voxel-wise threshold of P < 0.001, uncorrected, for visualization. Data are shown in the neurological convention (i.e., right hemisphere on the right). IP, implicit priming.

Neuronal response during IP paradigm

For a secondary analysis, we investigated response to high-calorie food cues (which were paired with negatively valenced images in active IP, but with neutral, scrambled images in control IP) during the IP paradigm itself (active group, n = 21; control group, n = 19), in the same ROIs as for the visual food cues task (insula, striatum, amygdala). Response to high-calorie food cues was significantly greater during active IP than during control IP in the insula (t = 2.37, P = 0.011). Trends in the same direction were observed for both striatum (t = 1.42, P = 0.082) and amygdala (t = 2.10, P = 0.021) ROIs. A significant main effect of food type was observed in response times to categorize foods as “healthy” or “unhealthy” via button press during the priming task, with faster response times for low-calorie (mean ± SEM: 906.17 ± 24.41 ms) than for high-calorie food cues (mean ± SEM: 1056.36 ± 31.16 ms) [F (1,38) = 5.83, P = 0.021]. No significant effects on response time were observed for group [active compared with control; F (1,38) = 1.77, P = 0.192] or for group-by-food-type interaction [F (1,38) = 0.13, P = 0.719].

Discussion

In adults with overweight/obesity, a reduced response to high-calorie food cues was observed after active IP compared with control IP in dlPFC, insula, and striatum. Because these brain regions have been associated with food-based motivation, reward processing, and stimulus salience (23, 24), this supports reduced valuation of and/or attention toward high-calorie food cues after IP.

The observed reduction in insula and striatum response to high-calorie food cues after active compared with control IP suggests reductions in cue salience. The insula, which contains the primary taste cortex, is involved in food-based motivation and reward processing, including integration of sensory inputs (e.g., food cues) with homeostatic information and interoception (23, 24, 26, 31). The insula is an integral component of the salience network, which plays an important role in attentional control, including automatic (i.e., passive, bottom-up) filtering processes involved in directing attention to salient stimuli (32). The insula is also considered part of the appetitive network, as is the striatum (23, 24, 31). Both regions play a key role in conditioned food responses (23, 31, 33), which is particularly relevant to the IP paradigm because it aims to alter these responses. The striatum is also involved in food-related reward processing, playing a key role in encoding stimulus salience (24, 34, 35). Increased responsivity to food cues in the insula and striatum has been associated with obesity (24, 36–38) and subsequent weight gain (23, 24, 39, 40). As such, the observed reduction in insula and striatum responses to high-calorie food cues after active compared with control IP is consistent with IP diminishing the salience of those cues.

Previous studies have observed associations between increased dlPFC responsivity to food cues and greater weight loss during dietary and/or exercise interventions, perhaps reflecting increased inhibitory control while viewing food images (24, 38, 41–43), because the dlPFC plays a key role in cognitive control and inhibition (44). Similarly, cognitive reappraisal techniques, such as thinking of the long-term costs of unhealthy eating or suppressing cravings, have been associated with increased dlPFC response to food cues (14, 45, 46). Although these results may initially seem at odds with the current finding of reduced dlPFC response to high-calorie foods, a consideration of the theoretical approach of the IP strategy reveals that this is not the case. The goal of the IP intervention is to affect automatic responses to food cues, thus representing a “bottom-up” approach (i.e., passive, automatic) rather than a cognitively driven “top-down” approach. As such, in this context, the observed reduction in dlPFC response to high-calorie food cues after active compared with control IP may reflect the success of this approach, such that diminished salience of food cues leads to fewer attentional resources directed to those cues and perhaps reflecting a reduced engagement of cognitive control strategies (e.g., dietary restraint) when viewing foods not compatible with health goals.

During the IP intervention itself, comparing the active group, which viewed high-calorie foods primed with images chosen to evoke disgust, with the control group, which viewed high-calorie foods primed with neutral, scrambled images, revealed greater neuronal response in the insula. Similar trends during active IP were observed in striatum and amygdala. Given the role of these brain regions, particularly the insula, in disgust processing and conditioned disgust responses (26, 47), engagement of these regions suggests that the IP intervention appears to be targeting the desired biological construct.

Behaviorally, results were similar to those in our preliminary study with this intervention (11). “Desire to eat” high-calorie foods was reduced after active compared with control IP, in both the current study and our previous work. This effect was also observed when only “novel” food images (i.e., those not in the priming intervention) were included in the analysis, suggesting generalization beyond foods specifically primed. Interestingly, low-calorie ratings increased after control compared with active IP. This may also fit with our previous study, in which a small low-calorie rating increase was observed after control compared with active IP, although to a lesser extent than observed here. In active IP, low-calorie foods were paired with positive images. As noted in our preliminary work (11), these pairings may be less effective than high-calorie/negative pairings owing to the greater salience of disgust images (8). That we did not observe increased responsivity to low-calorie foods after active compared with control IP during fMRI is consistent with this being less effective than high-calorie/disgust pairings. Although it is intriguing that low-calorie ratings increased after control IP, this should be interpreted with caution, because baseline low-calorie ratings were lower in the control than in the active group.

Our findings are also consistent with previous work supporting the potential efficacy of IP approaches in altering affective associations relating to food or consumption. For example, implicitly priming a brand of beverage increased the likelihood of choosing that brand over another (48) and implicitly priming eating-related verbs with positively valenced words was associated with greater saliva production than priming with neutrally valenced words (49). In women with anorexia nervosa, implicitly primed fear cues elicited negative appraisal of food images (50). Findings are also consistent with those unrelated to eating, such as implicitly pairing target pictures with pictures or words of positive/negative valence (e.g., pairing a flower picture with the word “stench”), which increased liking/disliking of the target picture (5).

At the follow-up visit in the current study, high-calorie ratings were still reduced from baseline in the active group, but they were also reduced compared with baseline in the control group. How this relates to the intervention is not clear. Low-calorie rating reductions were also observed between baseline and follow-up in both groups, so this could reflect habituation or fatigue with the task. We observed similar high- and low-calorie rating reductions 3–5 d after active IP in our previous study (11). In that study, however, only the active group completed the follow-up, so it is unknown if similar reductions would have been observed in the control group. It is possible that effects of a single IP administration do not extend to the time at which we performed follow-up assessments. A future study could assess intervention dosing, to determine if multiple administrations may confer longer-lasting effects. Lastly, high-calorie rating reductions after active IP were associated with less caloric intake during an ad libitum lunch meal. This was not observed in the control group, or for low-calorie ratings in either group. Although further assessment is needed, this could suggest that high-calorie rating reductions associated with active IP may translate to alterations in subsequent food intake. The current study only included a single ad libitum meal, however, so we were unable to assess intervention-associated change in intake, an effect that should be examined in future studies.

A potential limitation was that behavioral ratings were performed at different times of day, with the first assessment in the morning (pre-breakfast) and the second in the afternoon (pre-lunch), which could have influenced ratings. This concern is somewhat mitigated, however, by results being concordant with our previous study (11), in which ratings were performed immediately before and after IP (and, as such, at the same time of day). Because neuronal response to food cues was the primary outcome, the current study design focused on assessing immediate IP effects on neuronal rather than behavioral responses. Both behavioral rating periods followed a period of fasting, although fasting length was longer for morning measures (overnight fast) than for afternoon measures (∼3.5-h fast). Future studies should further assess behavioral effects of IP, ideally at the same time of day and satiety state.

A potential concern with this intervention could be the possibility of diminishing eating enjoyment overall. Although reducing desire to consume unhealthy foods is important in supporting weight loss/maintenance and healthy eating habits, diminished eating enjoyment could potentially affect quality of life, a possibility that future studies should assess. In addition, the current study focused on individuals with overweight/obesity because this group may have more to gain from the intervention than healthy-weight individuals. Because healthy eating goals apply to individuals across all weights, however, it will also be important to assess effects in healthy-weight participants. Another potential limitation of the study is that it did not control for handedness. Because there is not a clear effect of handedness on structural/functional brain organization outside that relating to language or sensorimotor control, we would not have anticipated this to affect study results (51). Nonetheless, this should be considered in future studies. Finally, because the sample size in the current study was modest, a larger-scale investigation of IP will be a useful future direction to further assess reliability and replicability of effects, and to examine effects across a broader population.

In conclusion, the current study supports the potential of this novel IP paradigm for altering automatic associations to food cues. Reduced responsivity to high-calorie food cues after active compared with control IP in brain regions associated with salience, motivation, and reward processing suggests reduced valuation of and/or attention toward those cues after the IP intervention. This effect could support weight-loss or maintenance efforts by curbing the degree of cognitive restraint that may be required to maintain healthy eating habits; i.e., reducing the salience of unhealthy foods may facilitate better choices from a bottom-up perspective, because it could be easier to avoid unhealthy foods even when facing challenges to self-control (e.g., fatigue, convenience, habit). Supporting this, the observed reduced response to high-calorie food cues in the dlPFC after active IP is consistent with inhibitory control engagement being reduced when encountering cues depicting unhealthy food options. As such, the strategy examined in the current study may represent a promising approach for supporting healthier eating choices and weight goals.

Supplementary Material

nqac009_Supplemental_File

Click here for additional data file.^{(303.1KB, pdf)}

Acknowledgments

The authors’ responsibilities were as follows—KTL, M-AC, and JRT: designed the research; KTL, CE, and BPL: conducted the research study; KTL, M-AC, BPL, EK, and JRT: provided essential research materials; KTL, JJB, EK, and JRT: analyzed the data; KTL, CE, and JRT: wrote the manuscript; M-AC, BPL, and JJB: critically reviewed the manuscript; KTL: had primary responsibility for the final content; and all authors: read and approved the final manuscript. The authors report no conflicts of interest.

Notes

Supported by Colorado Nutrition Obesity Research Center grant P30 DK048520 (to KTL); National Institute of Mental Health grant R01 MH102224 (to JRT); and National Institute of Diabetes and Digestive and Kidney Diseases grants R21 DK102052 (to JRT), R01 DK089095 (to M-AC and JRT), K01 DK100445 (to KTL), and R01 DK119236 (to KTL). Support was also provided by Department of Veterans Affairs Clinical Science Research and Development Merit Review Awards I01CX001414 (to JRT) and I01CX001949 (to KTL) and Research Career Scientist Award IK6CX002178 (to JRT).

Supplemental Figures 1 and 2, Supplemental Results, and Supplemental Tables 1 and 2 are available from the “Supplementary data” link in the online posting of the article and from the same link in the online table of contents at https://academic.oup.com/ajcn/.

Abbreviations used: AHWC, Anschutz Health and Wellness Center; CS, conditioned stimulus; dlPFC, dorsolateral prefrontal cortex; DS-R, Disgust Scale-Revised; EC, evaluative conditioning; FCQ-S, Food Cravings Questionnaire-State; FOV, field of view; FWE, family-wise error; IAPS, International Affective Picture System; IP, implicit priming; MNI, Montreal Neurological Institute; PFS, Power of Food Scale; ROI, region of interest; TE, echo time; TFEQ, Three Factor Eating Questionnaire; TR, repetition time; US, unconditioned stimulus; VAS, visual analog scale.

Contributor Information

Kristina T Legget, Department of Psychiatry, University of Colorado School of Medicine, Anschutz Medical Campus, Aurora, CO, USA; Research Service, Rocky Mountain Regional VA Medical Center, Aurora, CO, USA.

Marc-Andre Cornier, Research Service, Rocky Mountain Regional VA Medical Center, Aurora, CO, USA; Division of Endocrinology, Metabolism and Diabetes, Department of Medicine, University of Colorado School of Medicine, Anschutz Medical Campus, Aurora, CO, USA; Anschutz Health and Wellness Center, University of Colorado Anschutz Medical Campus, Aurora, CO, USA; Division of Geriatric Medicine, Department of Medicine, University of Colorado School of Medicine, Anschutz Medical Campus, Aurora, CO, USA.

Christina Erpelding, Department of Psychiatry, University of Colorado School of Medicine, Anschutz Medical Campus, Aurora, CO, USA.

Benjamin P Lawful, Department of Psychiatry, University of Colorado School of Medicine, Anschutz Medical Campus, Aurora, CO, USA.

Joshua J Bear, Department of Pediatrics, Section of Neurology, Children's Hospital Colorado, Aurora, CO, USA; Department of Pediatrics, Section of Neurology, University of Colorado School of Medicine, Anschutz Medical Campus, Aurora, CO, USA.

Eugene Kronberg, Department of Psychiatry, University of Colorado School of Medicine, Anschutz Medical Campus, Aurora, CO, USA; Department of Neurology, University of Colorado School of Medicine, Anschutz Medical Campus, Aurora, CO, USA.

Jason R Tregellas, Department of Psychiatry, University of Colorado School of Medicine, Anschutz Medical Campus, Aurora, CO, USA; Research Service, Rocky Mountain Regional VA Medical Center, Aurora, CO, USA.

Data Availability

Data described in the article, code book, and analytic code will be made available upon request pending application and approval. Group-level statistical maps are available at http://neurovault.org/collections/12215/.

References

1. Fryar CD, Carroll MD, Afful J. Prevalence of overweight, obesity, and severe obesity among adults aged 20 and over: United States, 1960–1962 through 2017–2018 [Internet]. NCHS Health E-Stats. Hyattsville (MD): National Center for Health Statistics; 2020. [Cited 2021 Aug 24]. Available from: https://www.cdc.gov/nchs/data/hestat/obesity-adult-17-18/obesity-adult.htm. [Google Scholar]
2. Cornell CE, Rodin J, Weingarten H. Stimulus-induced eating when satiated. Physiol Behav. 1989;45(4):695–704. [DOI] [PubMed] [Google Scholar]
3. Fedoroff I, Polivy J, Herman CP. The specificity of restrained versus unrestrained eaters’ responses to food cues: general desire to eat, or craving for the cued food?. Appetite. 2003;41(1):7–13. [DOI] [PubMed] [Google Scholar]
4. Boswell RG, Kober H. Food cue reactivity and craving predict eating and weight gain: a meta-analytic review. Obes Rev. 2016;17(2):159–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
5. Fulcher EP, Hammerl M. When all is revealed: a dissociation between evaluative learning and contingency awareness. Conscious Cogn. 2001;10(4):524–49. [DOI] [PubMed] [Google Scholar]
6. Jones CR, Fazio RH, Olson MA. Implicit misattribution as a mechanism underlying evaluative conditioning. J Pers Soc Psychol. 2009;96(5):933–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
7. March DS, Olson MA, Fazio RH. The implicit misattribution model of evaluative conditioning. Soc Psychol Bull. 2018;13(3):1–25. [Google Scholar]
8. Rozin P. One-trial acquired likes and dislikes in humans: disgust as a US, food predominance, and negative learning predominance. Learn Motiv. 1986;17(2):180–9. [Google Scholar]
9. Batsell WR, Brown AS. Human flavor-aversion learning: a comparison of traditional aversions and cognitive aversions. Learn Motiv. 1998;29(4):383–96. [Google Scholar]
10. Eertmans A, Baeyens F, Van den Bergh O. Food likes and their relative importance in human eating behavior: review and preliminary suggestions for health promotion. Health Educ Res. 2001;16(4):443–56. [DOI] [PubMed] [Google Scholar]
11. Legget KT, Cornier M-A, Rojas DC, Lawful B, Tregellas JR. Harnessing the power of disgust: a randomized trial to reduce high-calorie food appeal through implicit priming. Am J Clin Nutr. 2015;102(2):249–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
12. Garner DM, Olmsted MP, Bohr Y, Garfinkel PE. The Eating Attitudes Test: psychometric features and clinical correlates. Psychol Med. 1982;12(4):871–8. [DOI] [PubMed] [Google Scholar]
13. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007;39(2):175–91. [DOI] [PubMed] [Google Scholar]
14. Siep N, Roefs A, Roebroeck A, Havermans R, Bonte M, Jansen A. Fighting food temptations: the modulating effects of short-term cognitive reappraisal, suppression and up-regulation on mesocorticolimbic activity related to appetitive motivation. Neuroimage. 2012;60(1):213–20. [DOI] [PubMed] [Google Scholar]
15. Mifflin MD, St Jeor ST, Hill LA, Scott BJ, Daugherty SA, Koh YO. A new predictive equation for resting energy expenditure in healthy individuals. Am J Clin Nutr. 1990;51(2):241–7. [DOI] [PubMed] [Google Scholar]
16. Stunkard AJ, Messick S. The three-factor eating questionnaire to measure dietary restraint, disinhibition and hunger. J Psychosom Res. 1985;29(1):71–83. [DOI] [PubMed] [Google Scholar]
17. Lowe MR, Butryn ML, Didie ER, Annunziato RA, Thomas JG, Crerand CE, Ochner CN, Coletta MC, Bellace D, Wallaert Met al. The Power of Food Scale. A new measure of the psychological influence of the food environment. Appetite. 2009;53(1):114–8. [DOI] [PubMed] [Google Scholar]
18. Burger KS, Cornier MA, Ingebrigtsen J, Johnson SL. Assessing food appeal and desire to eat: the effects of portion size & energy density. Int J Behav Nutr Phys Act. 2011;8:101. [DOI] [PMC free article] [PubMed] [Google Scholar]
19. Lang P, Bradley MM, Cuthbert BN. International Affective Picture System (IAPS): instruction manual and affective ratings. Gainesville (FL): The Center for Research in Psychophysiology, University of Florida; 2001. [Google Scholar]
20. Cepeda-Benito A, Gleaves DH, Fernandez MC, Vila J, Williams TL, Reynoso J. The development and validation of Spanish versions of the State and Trait Food Cravings Questionnaires. Behav Res Ther. 2000;38(11):1125–38. [DOI] [PubMed] [Google Scholar]
21. Olatunji BO, Williams NL, Tolin DF, Abramowitz JS, Sawchuk CN, Lohr JM, Elwood LS. The Disgust Scale: item analysis, factor structure, and suggestions for refinement. Psychol Assess. 2007;19(3):281–97. [DOI] [PubMed] [Google Scholar]
22. Chen Z, Holland RW, Quandt J, Dijksterhuis A, Veling H. When mere action versus inaction leads to robust preference change. J Pers Soc Psychol. 2019;117(4):721–40. [DOI] [PubMed] [Google Scholar]
23. Tang DW, Fellows LK, Small DM, Dagher A. Food and drug cues activate similar brain regions: a meta-analysis of functional MRI studies. Physiol Behav. 2012;106(3):317–24. [DOI] [PubMed] [Google Scholar]
24. Neseliler S, Han JE, Dagher A. The use of functional magnetic resonance imaging in the study of appetite and obesity. In: Harris RBS (editor). Appetite and food intake: central control. 2nd ed. Boca Raton (FL): CRC Press/Taylor & Francis; 2017. 117–34. [Google Scholar]
25. Stark R, Schienle A, Walter B, Kirsch P, Sammer G, Ott U, Blecker C, Vaitl D. Hemodynamic responses to fear and disgust-inducing pictures: an fMRI study. Int J Psychophysiol. 2003;50(3):225–34. [DOI] [PubMed] [Google Scholar]
26. Chapman HA, Anderson AK. Understanding disgust. Ann N Y Acad Sci. 2012;1251(1):62–76. [DOI] [PubMed] [Google Scholar]
27. Bhikram T, Abi-Jaoude E, Sandor P. OCD: obsessive-compulsive … disgust? The role of disgust in obsessive-compulsive disorder. J Psychiatry Neurosci. 2017;42(5):300–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Hammers A, Allom R, Koepp MJ, Free SL, Myers R, Lemieux L, Mitchell TN, Brooks DJ, Duncan JS. Three-dimensional maximum probability atlas of the human brain, with particular reference to the temporal lobe. Hum Brain Mapp. 2003;19(4):224–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Cornier MA, Salzberg AK, Endly DC, Bessesen DH, Tregellas JR. Sex-based differences in the behavioral and neuronal responses to food. Physiol Behav. 2010;99(4):538–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Cornier M-A, McFadden KL, Thomas EA, Bechtell JL, Eichman LS, Bessesen DH, Tregellas JR. Differences in the neuronal response to food in obesity-resistant as compared to obesity-prone individuals. Physiol Behav. 2013;110–111:122–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Dagher A. The neurobiology of appetite: hunger as addiction. In: Dubé L, Bechara A, Dagher A, Drewnowski A, Lebel J, James P, Yada RY (editors). Obesity prevention: the role of brain and society on individual behavior. San Diego (CA): Academic Press; 2010. p. 15–22. [Google Scholar]
32. Menon V, Uddin LQ. Saliency, switching, attention and control: a network model of insula function. Brain Struct Funct. 2010;214(5–6):655–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Kusumoto-Yoshida I, Liu H, Chen BT, Fontanini A, Bonci A. Central role for the insular cortex in mediating conditioned responses to anticipatory cues. Proc Natl Acad Sci U S A. 2015;112(4):1190–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
34. O'Doherty JP, Buchanan TW, Seymour B, Dolan RJ. Predictive neural coding of reward preference involves dissociable responses in human ventral midbrain and ventral striatum. Neuron. 2006;49(1):157–66. [DOI] [PubMed] [Google Scholar]
35. Martin-Soelch C, Linthicum J, Ernst M. Appetitive conditioning: neural bases and implications for psychopathology. Neurosci Biobehav Rev. 2007;31(3):426–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
36. Rothemund Y, Preuschhof C, Bohner G, Bauknecht HC, Klingebiel R, Flor H, Klapp BF. Differential activation of the dorsal striatum by high-calorie visual food stimuli in obese individuals. Neuroimage. 2007;37(2):410–21. [DOI] [PubMed] [Google Scholar]
37. Stoeckel LE, Weller RE, Cook EW 3rd, Twieg DB, Knowlton RC, Cox JE. Widespread reward-system activation in obese women in response to pictures of high-calorie foods. Neuroimage. 2008;41(2):636–47. [DOI] [PubMed] [Google Scholar]
38. Pursey KM, Stanwell P, Callister RJ, Brain K, Collins CE, Burrows TL. Neural responses to visual food cues according to weight status: a systematic review of functional magnetic resonance imaging studies. Front Nutr. 2014;1:7. [DOI] [PMC free article] [PubMed] [Google Scholar]
39. Demos KE, Heatherton TF, Kelley WM. Individual differences in nucleus accumbens activity to food and sexual images predict weight gain and sexual behavior. J Neurosci. 2012;32(16):5549–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
40. Murdaugh DL, Cox JE, Cook EW 3rd, Weller RE. fMRI reactivity to high-calorie food pictures predicts short- and long-term outcome in a weight-loss program. Neuroimage. 2012;59(3):2709–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
41. Kahathuduwa CN, Davis T, O'Boyle M, Boyd LA, Chin SH, Paniukov D, Binks M. Effects of 3-week total meal replacement vs. typical food-based diet on human brain functional magnetic resonance imaging food-cue reactivity and functional connectivity in people with obesity. Appetite. 2018;120:431–41. [DOI] [PubMed] [Google Scholar]
42. Neseliler S, Hu W, Larcher K, Zacchia M, Dadar M, Scala SG, Lamarche M, Zeighami Y, Stotland SC, Larocque Met al. Neurocognitive and hormonal correlates of voluntary weight loss in humans. Cell Metab. 2019;29(1):39–49.e4. [DOI] [PubMed] [Google Scholar]
43. Szabo-Reed AN, Martin LE, Hu J, Yeh H-W, Powell J, Lepping RJ, Patrician TM, Breslin FJ, Donnelly JE, Savage CR. Modeling interactions between brain function, diet adherence behaviors, and weight loss success. Obes Sci Pract. 2020;6(3):282–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
44. Miller EK, Cohen JD. An integrative theory of prefrontal cortex function. Annu Rev Neurosci. 2001;24:167–202. [DOI] [PubMed] [Google Scholar]
45. Kober H, Mende-Siedlecki P, Kross EF, Weber J, Mischel W, Hart CL, Ochsner KN. Prefrontal–striatal pathway underlies cognitive regulation of craving. Proc Natl Acad Sci U S A. 2010;107(33):14811–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
46. Hollmann M, Hellrung L, Pleger B, Schlogl H, Kabisch S, Stumvoll M, Villringer A, Horstmann A. Neural correlates of the volitional regulation of the desire for food. Int J Obes. 2012;36(5):648–55. [DOI] [PubMed] [Google Scholar]
47. Klucken T, Schweckendiek J, Koppe G, Merz CJ, Kagerer S, Walter B, Sammer G, Vaitl D, Stark R. Neural correlates of disgust- and fear-conditioned responses. Neuroscience. 2012;201:209–18. [DOI] [PubMed] [Google Scholar]
48. Karremans JC, Stroebe W, Claus J. Beyond Vicary's fantasies: the impact of subliminal priming and brand choice. J Exp Soc Psychol. 2006;42(6):792–8. [Google Scholar]
49. Amd M, Baillet S. Neurophysiological effects associated with subliminal conditioning of appetite motivations. Front Psychol. 2019;10:457. [DOI] [PMC free article] [PubMed] [Google Scholar]
50. Soussignan R, Jiang T, Rigaud D, Royet JP, Schaal B. Subliminal fear priming potentiates negative facial reactions to food pictures in women with anorexia nervosa. Psychol Med. 2010;40(3):503–14. [DOI] [PubMed] [Google Scholar]
51. Bailey LM, McMillan LE, Newman AJ. A sinister subject: quantifying handedness-based recruitment biases in current neuroimaging research. Eur J Neurosci. 2020;51(7):1642–56. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

nqac009_Supplemental_File

Click here for additional data file.^{(303.1KB, pdf)}

Data Availability Statement

[bib1] 1. Fryar CD, Carroll MD, Afful J. Prevalence of overweight, obesity, and severe obesity among adults aged 20 and over: United States, 1960–1962 through 2017–2018 [Internet]. NCHS Health E-Stats. Hyattsville (MD): National Center for Health Statistics; 2020. [Cited 2021 Aug 24]. Available from: https://www.cdc.gov/nchs/data/hestat/obesity-adult-17-18/obesity-adult.htm. [Google Scholar]

[bib2] 2. Cornell CE, Rodin J, Weingarten H. Stimulus-induced eating when satiated. Physiol Behav. 1989;45(4):695–704. [DOI] [PubMed] [Google Scholar]

[bib3] 3. Fedoroff I, Polivy J, Herman CP. The specificity of restrained versus unrestrained eaters’ responses to food cues: general desire to eat, or craving for the cued food?. Appetite. 2003;41(1):7–13. [DOI] [PubMed] [Google Scholar]

[bib4] 4. Boswell RG, Kober H. Food cue reactivity and craving predict eating and weight gain: a meta-analytic review. Obes Rev. 2016;17(2):159–77. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib5] 5. Fulcher EP, Hammerl M. When all is revealed: a dissociation between evaluative learning and contingency awareness. Conscious Cogn. 2001;10(4):524–49. [DOI] [PubMed] [Google Scholar]

[bib6] 6. Jones CR, Fazio RH, Olson MA. Implicit misattribution as a mechanism underlying evaluative conditioning. J Pers Soc Psychol. 2009;96(5):933–48. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib7] 7. March DS, Olson MA, Fazio RH. The implicit misattribution model of evaluative conditioning. Soc Psychol Bull. 2018;13(3):1–25. [Google Scholar]

[bib8] 8. Rozin P. One-trial acquired likes and dislikes in humans: disgust as a US, food predominance, and negative learning predominance. Learn Motiv. 1986;17(2):180–9. [Google Scholar]

[bib9] 9. Batsell WR, Brown AS. Human flavor-aversion learning: a comparison of traditional aversions and cognitive aversions. Learn Motiv. 1998;29(4):383–96. [Google Scholar]

[bib10] 10. Eertmans A, Baeyens F, Van den Bergh O. Food likes and their relative importance in human eating behavior: review and preliminary suggestions for health promotion. Health Educ Res. 2001;16(4):443–56. [DOI] [PubMed] [Google Scholar]

[bib11] 11. Legget KT, Cornier M-A, Rojas DC, Lawful B, Tregellas JR. Harnessing the power of disgust: a randomized trial to reduce high-calorie food appeal through implicit priming. Am J Clin Nutr. 2015;102(2):249–55. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] 12. Garner DM, Olmsted MP, Bohr Y, Garfinkel PE. The Eating Attitudes Test: psychometric features and clinical correlates. Psychol Med. 1982;12(4):871–8. [DOI] [PubMed] [Google Scholar]

[bib13] 13. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007;39(2):175–91. [DOI] [PubMed] [Google Scholar]

[bib14] 14. Siep N, Roefs A, Roebroeck A, Havermans R, Bonte M, Jansen A. Fighting food temptations: the modulating effects of short-term cognitive reappraisal, suppression and up-regulation on mesocorticolimbic activity related to appetitive motivation. Neuroimage. 2012;60(1):213–20. [DOI] [PubMed] [Google Scholar]

[bib15] 15. Mifflin MD, St Jeor ST, Hill LA, Scott BJ, Daugherty SA, Koh YO. A new predictive equation for resting energy expenditure in healthy individuals. Am J Clin Nutr. 1990;51(2):241–7. [DOI] [PubMed] [Google Scholar]

[bib16] 16. Stunkard AJ, Messick S. The three-factor eating questionnaire to measure dietary restraint, disinhibition and hunger. J Psychosom Res. 1985;29(1):71–83. [DOI] [PubMed] [Google Scholar]

[bib17] 17. Lowe MR, Butryn ML, Didie ER, Annunziato RA, Thomas JG, Crerand CE, Ochner CN, Coletta MC, Bellace D, Wallaert Met al. The Power of Food Scale. A new measure of the psychological influence of the food environment. Appetite. 2009;53(1):114–8. [DOI] [PubMed] [Google Scholar]

[bib18] 18. Burger KS, Cornier MA, Ingebrigtsen J, Johnson SL. Assessing food appeal and desire to eat: the effects of portion size & energy density. Int J Behav Nutr Phys Act. 2011;8:101. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib19] 19. Lang P, Bradley MM, Cuthbert BN. International Affective Picture System (IAPS): instruction manual and affective ratings. Gainesville (FL): The Center for Research in Psychophysiology, University of Florida; 2001. [Google Scholar]

[bib20] 20. Cepeda-Benito A, Gleaves DH, Fernandez MC, Vila J, Williams TL, Reynoso J. The development and validation of Spanish versions of the State and Trait Food Cravings Questionnaires. Behav Res Ther. 2000;38(11):1125–38. [DOI] [PubMed] [Google Scholar]

[bib21] 21. Olatunji BO, Williams NL, Tolin DF, Abramowitz JS, Sawchuk CN, Lohr JM, Elwood LS. The Disgust Scale: item analysis, factor structure, and suggestions for refinement. Psychol Assess. 2007;19(3):281–97. [DOI] [PubMed] [Google Scholar]

[bib22] 22. Chen Z, Holland RW, Quandt J, Dijksterhuis A, Veling H. When mere action versus inaction leads to robust preference change. J Pers Soc Psychol. 2019;117(4):721–40. [DOI] [PubMed] [Google Scholar]

[bib23] 23. Tang DW, Fellows LK, Small DM, Dagher A. Food and drug cues activate similar brain regions: a meta-analysis of functional MRI studies. Physiol Behav. 2012;106(3):317–24. [DOI] [PubMed] [Google Scholar]

[bib24] 24. Neseliler S, Han JE, Dagher A. The use of functional magnetic resonance imaging in the study of appetite and obesity. In: Harris RBS (editor). Appetite and food intake: central control. 2nd ed. Boca Raton (FL): CRC Press/Taylor & Francis; 2017. 117–34. [Google Scholar]

[bib25] 25. Stark R, Schienle A, Walter B, Kirsch P, Sammer G, Ott U, Blecker C, Vaitl D. Hemodynamic responses to fear and disgust-inducing pictures: an fMRI study. Int J Psychophysiol. 2003;50(3):225–34. [DOI] [PubMed] [Google Scholar]

[bib26] 26. Chapman HA, Anderson AK. Understanding disgust. Ann N Y Acad Sci. 2012;1251(1):62–76. [DOI] [PubMed] [Google Scholar]

[bib27] 27. Bhikram T, Abi-Jaoude E, Sandor P. OCD: obsessive-compulsive … disgust? The role of disgust in obsessive-compulsive disorder. J Psychiatry Neurosci. 2017;42(5):300–6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib28] 28. Hammers A, Allom R, Koepp MJ, Free SL, Myers R, Lemieux L, Mitchell TN, Brooks DJ, Duncan JS. Three-dimensional maximum probability atlas of the human brain, with particular reference to the temporal lobe. Hum Brain Mapp. 2003;19(4):224–47. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib29] 29. Cornier MA, Salzberg AK, Endly DC, Bessesen DH, Tregellas JR. Sex-based differences in the behavioral and neuronal responses to food. Physiol Behav. 2010;99(4):538–43. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib30] 30. Cornier M-A, McFadden KL, Thomas EA, Bechtell JL, Eichman LS, Bessesen DH, Tregellas JR. Differences in the neuronal response to food in obesity-resistant as compared to obesity-prone individuals. Physiol Behav. 2013;110–111:122–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib31] 31. Dagher A. The neurobiology of appetite: hunger as addiction. In: Dubé L, Bechara A, Dagher A, Drewnowski A, Lebel J, James P, Yada RY (editors). Obesity prevention: the role of brain and society on individual behavior. San Diego (CA): Academic Press; 2010. p. 15–22. [Google Scholar]

[bib32] 32. Menon V, Uddin LQ. Saliency, switching, attention and control: a network model of insula function. Brain Struct Funct. 2010;214(5–6):655–67. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib33] 33. Kusumoto-Yoshida I, Liu H, Chen BT, Fontanini A, Bonci A. Central role for the insular cortex in mediating conditioned responses to anticipatory cues. Proc Natl Acad Sci U S A. 2015;112(4):1190–5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib34] 34. O'Doherty JP, Buchanan TW, Seymour B, Dolan RJ. Predictive neural coding of reward preference involves dissociable responses in human ventral midbrain and ventral striatum. Neuron. 2006;49(1):157–66. [DOI] [PubMed] [Google Scholar]

[bib35] 35. Martin-Soelch C, Linthicum J, Ernst M. Appetitive conditioning: neural bases and implications for psychopathology. Neurosci Biobehav Rev. 2007;31(3):426–40. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib36] 36. Rothemund Y, Preuschhof C, Bohner G, Bauknecht HC, Klingebiel R, Flor H, Klapp BF. Differential activation of the dorsal striatum by high-calorie visual food stimuli in obese individuals. Neuroimage. 2007;37(2):410–21. [DOI] [PubMed] [Google Scholar]

[bib37] 37. Stoeckel LE, Weller RE, Cook EW 3rd, Twieg DB, Knowlton RC, Cox JE. Widespread reward-system activation in obese women in response to pictures of high-calorie foods. Neuroimage. 2008;41(2):636–47. [DOI] [PubMed] [Google Scholar]

[bib38] 38. Pursey KM, Stanwell P, Callister RJ, Brain K, Collins CE, Burrows TL. Neural responses to visual food cues according to weight status: a systematic review of functional magnetic resonance imaging studies. Front Nutr. 2014;1:7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib39] 39. Demos KE, Heatherton TF, Kelley WM. Individual differences in nucleus accumbens activity to food and sexual images predict weight gain and sexual behavior. J Neurosci. 2012;32(16):5549–52. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib40] 40. Murdaugh DL, Cox JE, Cook EW 3rd, Weller RE. fMRI reactivity to high-calorie food pictures predicts short- and long-term outcome in a weight-loss program. Neuroimage. 2012;59(3):2709–21. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib41] 41. Kahathuduwa CN, Davis T, O'Boyle M, Boyd LA, Chin SH, Paniukov D, Binks M. Effects of 3-week total meal replacement vs. typical food-based diet on human brain functional magnetic resonance imaging food-cue reactivity and functional connectivity in people with obesity. Appetite. 2018;120:431–41. [DOI] [PubMed] [Google Scholar]

[bib42] 42. Neseliler S, Hu W, Larcher K, Zacchia M, Dadar M, Scala SG, Lamarche M, Zeighami Y, Stotland SC, Larocque Met al. Neurocognitive and hormonal correlates of voluntary weight loss in humans. Cell Metab. 2019;29(1):39–49.e4. [DOI] [PubMed] [Google Scholar]

[bib43] 43. Szabo-Reed AN, Martin LE, Hu J, Yeh H-W, Powell J, Lepping RJ, Patrician TM, Breslin FJ, Donnelly JE, Savage CR. Modeling interactions between brain function, diet adherence behaviors, and weight loss success. Obes Sci Pract. 2020;6(3):282–92. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib44] 44. Miller EK, Cohen JD. An integrative theory of prefrontal cortex function. Annu Rev Neurosci. 2001;24:167–202. [DOI] [PubMed] [Google Scholar]

[bib45] 45. Kober H, Mende-Siedlecki P, Kross EF, Weber J, Mischel W, Hart CL, Ochsner KN. Prefrontal–striatal pathway underlies cognitive regulation of craving. Proc Natl Acad Sci U S A. 2010;107(33):14811–6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib46] 46. Hollmann M, Hellrung L, Pleger B, Schlogl H, Kabisch S, Stumvoll M, Villringer A, Horstmann A. Neural correlates of the volitional regulation of the desire for food. Int J Obes. 2012;36(5):648–55. [DOI] [PubMed] [Google Scholar]

[bib47] 47. Klucken T, Schweckendiek J, Koppe G, Merz CJ, Kagerer S, Walter B, Sammer G, Vaitl D, Stark R. Neural correlates of disgust- and fear-conditioned responses. Neuroscience. 2012;201:209–18. [DOI] [PubMed] [Google Scholar]

[bib48] 48. Karremans JC, Stroebe W, Claus J. Beyond Vicary's fantasies: the impact of subliminal priming and brand choice. J Exp Soc Psychol. 2006;42(6):792–8. [Google Scholar]

[bib49] 49. Amd M, Baillet S. Neurophysiological effects associated with subliminal conditioning of appetite motivations. Front Psychol. 2019;10:457. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib50] 50. Soussignan R, Jiang T, Rigaud D, Royet JP, Schaal B. Subliminal fear priming potentiates negative facial reactions to food pictures in women with anorexia nervosa. Psychol Med. 2010;40(3):503–14. [DOI] [PubMed] [Google Scholar]

[bib51] 51. Bailey LM, McMillan LE, Newman AJ. A sinister subject: quantifying handedness-based recruitment biases in current neuroimaging research. Eur J Neurosci. 2020;51(7):1642–56. [DOI] [PubMed] [Google Scholar]

PERMALINK

An implicit priming intervention alters brain and behavioral responses to high-calorie foods: a randomized controlled study

Kristina T Legget

Marc-Andre Cornier

Christina Erpelding

Benjamin P Lawful

Joshua J Bear

Eugene Kronberg

Jason R Tregellas

ABSTRACT

Background

Objectives

Methods

Results

Conclusions

Introduction

Methods

Participants

Study design and randomization

FIGURE 1.

Eucaloric run-in diet day

Study day

Follow-up visit

fMRI data acquisition

Visual food cues task

IP intervention

FIGURE 2.

Data analyses

Results

TABLE 1.

Behavioral food ratings, food intake, and appetite

FIGURE 3.

Baseline neuronal response to visual food cues

TABLE 2.

IP effects on neuronal response to visual food cues

FIGURE 4.

Neuronal response during IP paradigm

Discussion

Supplementary Material

Acknowledgments

Notes

Contributor Information

Data Availability

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases