Neural reactivity to visual food stimuli is reduced in some areas of the brain during evening hours compared to morning hours: an fMRI study in women
- First Online:
- Cite this article as:
- Masterson, T.D., Kirwan, C.B., Davidson, L.E. et al. Brain Imaging and Behavior (2016) 10: 68. doi:10.1007/s11682-015-9366-8
- 1.5k Downloads
The extent that neural responsiveness to visual food stimuli is influenced by time of day is not well examined. Using a crossover design, 15 healthy women were scanned using fMRI while presented with low- and high-energy pictures of food, once in the morning (6:30–8:30 am) and once in the evening (5:00–7:00 pm). Diets were identical on both days of the fMRI scans and were verified using weighed food records. Visual analog scales were used to record subjective perception of hunger and preoccupation with food prior to each fMRI scan. Six areas of the brain showed lower activation in the evening to both high- and low-energy foods, including structures in reward pathways (P < 0.05). Nine brain regions showed significantly higher activation for high-energy foods compared to low-energy foods (P < 0.05). High-energy food stimuli tended to produce greater fMRI responses than low-energy food stimuli in specific areas of the brain, regardless of time of day. However, evening scans showed a lower response to both low- and high-energy food pictures in some areas of the brain. Subjectively, participants reported no difference in hunger by time of day (F = 1.84, P = 0.19), but reported they could eat more (F = 4.83, P = 0.04) and were more preoccupied with thoughts of food (F = 5.51, P = 0.03) in the evening compared to the morning. These data underscore the role that time of day may have on neural responses to food stimuli. These results may also have clinical implications for fMRI measurement in order to prevent a time of day bias.
KeywordsVisual stimuli Food Neural reactivity fMRI Time Morning Evening
Food-seeking and eating behaviors are complex but may be influenced by time of day (Allison et al. 2014; de Castro 2004; LeCheminant et al. 2013). For example, there is evidence to suggest that the evening meal is the largest of the day and is important in determining overall energy intake (de Castro 2004). In addition, data are emerging that suggest that delays in the time of eating (i.e., night-time eating) can influence energy balance (Allison et al. 2014). However, the precise mechanisms for why or how the evening hours reportedly influence eating behavior is poorly understood and not well investigated. This path of scientific inquiry is important and has practical implications with regard to weight management and obesity (Ogden et al. 2014).
Admittedly, food seeking and eating behaviors can be influenced by a variety of environmental cues, as well as social, behavioral, and individual factors (I. C. Fedoroff et al. 1997; I. Fedoroff et al. 2003; Lumeng and Burke 2006; Prinsen et al. 2013). However, responses to visual cues may have a particularly potent effect on dietary choices (Brignell et al. 2009; Ng et al. 2011; Yokum et al. 2011). Accordingly, a growing body of research has begun to objectively examine the influence of visual food cues by specifically measuring responses inside the brain when presented with food stimuli (Bruce et al. 2013; Ng et al. 2011; Rothemund et al. 2007; Yokum et al. 2011).
Previous research has shown higher activation in areas of the brain related to reward and motivation when participants are presented with food stimuli compared to neutral stimuli (Killgore et al. 2003; Sweet et al. 2012; Yokum et al. 2011). This is important as the reward response may index how motivated a person is to seek food (Holsen et al. 2012). In addition, neural responses appear greater to high-energy food stimuli compared to low-energy food stimuli (Killgore et al. 2003; Sweet et al. 2012; Yokum et al. 2011).
We are aware of only one recent study that has investigated the effect of time-of-day on neural reward circuitry in humans (Hasler et al. 2014). In this paper, Hasler et al. showed increased activation in reward pathways (using monetary stimuli) in the “afternoon” compared to the morning, which they suggested followed the natural circadian cycles (Hasler et al. 2014). Similarly, of the functional magnetic resonance imaging (fMRI) experiments cited in the present paper (Brignell et al. 2009; Bruce et al. 2013; Killgore et al. 2003; Murdaugh et al. 2012; Ng et al. 2011; Rothemund et al. 2007; Sweet et al. 2012; Yokum et al. 2011) only two reported the timing of their scans. Murdaugh et al. reported scanning subjects between the hours of 4 and 7 pm (Murdaugh et al. 2012) while Yokum et al. reported similar hours between 4 and 6 pm (Yokum et al. 2011); Yokum also reported a subset of individuals being scanned between the hours of 11 am and 1 pm (Yokum et al. 2011).
In this context, it would be valuable to know to what extent the evening hours influence neural responses to visual food stimuli compared to other times of the day.
Functional MRI has become a widely-used and accepted measurement tool in the human neuroscience and psychology fields (McGonigle 2012). Functional MRI measures the signal change produced by the blood oxygenation level dependent (BOLD) response, i.e., the hemodynamic response in which the relative proportion of oxygenated hemoglobin increases relative to deoxygenated hemoglobin in response to increases in neural activity. Functional MRI is recognized as one of the most effective ways to examine metabolic changes in the brain as it processes various types of stimuli (McGonigle 2012).
Therefore, this study attempts to add to the scientific literature by utilizing fMRI to compare the neural response to visual food pictures in the morning vs. evening; and second, to determine differences in response to low- and high-energy visual food stimuli. Given studies showing a higher energy intake in the evening, we hypothesized that the neural response to visual food stimuli would be greater in the evening compared to the morning. Furthermore, there would be a greater neural response to high-energy foods as opposed to low-energy foods regardless of time of day. We hypothesized that we would see significant difference within a whole brain analysis in areas such as the ventral tegmental area, the nucleus accumbens, the prefrontal cortex, amygdala, and the hippocampus.
The Institutional Review Board at Brigham Young University approved this study and all participants gave written informed consent prior to beginning. Fifteen premenopausal women were recruited to complete two experimental conditions (morning condition and evening condition). This design included 30 total fMRI scans: 15 in the morning, 15 in the evening. The conditions were counterbalanced and the order was determined at random using a stratified randomization protocol to ensure that an equivalent proportion of participants would start in each condition. Both the participant and the research assistant were blinded to randomization prior to when the first randomized condition was revealed. During the morning condition, the neural response to pictures of food was assessed in each participant between the hours of 6:30–8:30 am. During the evening condition, the neural response to pictures of food was assessed in each participant between the hours of 5:30–7:30 pm. During each condition, fMRI scans were collected as participants viewed pictures of food. The food pictures used contained a mix of high energy-dense/highly palatable foods, low energy-dense foods, blurred versions of the food pictures (used as a control condition to factor out visual processing), and visually complex pictures of vegetation and minerals (used as a second control condition to factor out neural activation to visually complex stimuli).
Except for the time of day of the fMRI scans, the morning and evening conditions were nearly identical. Participants were scanned on the same day of the week with a separation of 1 week between conditions. Participants were instructed to follow the same daily schedule during both conditions. In addition, assessments were not conducted during holidays, vacations, or any break not typical of the participant’s normal schedule. For both conditions, participants were required to have slept at least 7 h the night before, discontinued eating the previous night by 8 pm, and refrained from vigorous-intensity exercise, caffeine, or alcohol consumption for at least 24 h prior. During both conditions, each participant consumed her same self-selected diet (based on habitual food consumption). Additionally, each meal was eaten at approximately the same times on the days of both conditions.
Participants were weight stable (±5 lbs), had normal sleep patterns, and were regular consumers of breakfast, lunch, and dinner during the six months previous to the study. Participants were excluded if they were claustrophobic or if they had night eating syndrome (NES). The NES criteria used are those proposed by Stunkard et al. (Stunkard et al. 2009). Additionally, participants were excluded if they were highly active (>4 times and 20 min per bout of vigorous activity per week), competitive or collegiate athletes (including marathon runners and triathletes), actively dieting or participating in any sort of commercial weight management program, had a metabolic disease, neurological or psychiatric disorder, eating disorder, used tobacco products, abused alcohol (no more than three standard drinks in a day or more than seven standard drinks in a week), were pregnant or lactating, or were otherwise incompatible with MRI (e.g., have pacemakers or recent tattoos). Participants were compensated for their time and participation.
For qualifying participants, assessments began with a visit to the Body Composition Laboratory and two visits to the MRI Research Facility; both facilities are located at Brigham Young University. On the initial visit, each participant received an overview of the study requirements, was screened for safety, informed of the protocol of the fMRI facility, and then gave informed consent to participate in the study. Subsequently, each participant was assessed for body composition, height and weight, and was instructed on how to complete a computerized, multiple-pass, 24-h dietary food recall by completing one for the previous day with the aid of study personnel. Between the first and second visits, participants were required to complete three additional online 24-h dietary recalls (days were determined randomly and included one weekend day). Prior to the second visit, each participant was instructed that she should eat an identical diet on the day of both fMRI scans and was taught how to select foods that best represented her typical dietary intake. Each participant’s typical dietary intake was based on her 24-h dietary recall data.
On the second visit, participants presented at the MRI research facility and received fMRI testing while in the first condition (morning or evening). For the duration of the day of each condition, participants recorded their food consumption by completing a weighed food record for all food and beverages consumed. Weighed food records were used on the specific days of the scans rather than 24-h recalls to ensure that the exact same foods and amounts were consumed on both days, since variance in content or energy profile could affect study outcomes. Between the second and third visits, participants completed three separate 24-h dietary recalls (including one weekend day); also determined randomly. On the third visit (on the same day of the week as the second visit), participants were scanned under their second condition (the condition not received previously). Participants consumed an identical breakfast and lunch at the same time of the day on the second scan as the first scan. This was confirmed using the weighed food records.
For the morning condition, participants were scanned in the MRI machine between 6:30–8:30 am in a fasted state (participants had been fasting for approximately 10 h). Participants subsequently consumed their typical breakfast before 9:30 am and their typical lunch between 11:30 am and 1:00 pm. These meal timeframes were designed to mimic typical meal hours. Snacking or food consumption after lunch was not allowed. For the evening condition, participants consumed the same breakfast and lunch at the same time of day as the morning condition. The lunchtime of 11:30 am-1:00 pm allowed for a significant delay between the lunch and evening meal (~6–8 h). In addition, during the evening condition, participants completed their fMRI scan between 5:30–7:30 pm and prior to their evening meal. As with the morning condition, no snacking was allowed on the day of the evening condition.
During the initial meeting with the participant, body composition was assessed via dual-energy x-ray absorptiometry (DXA) using a GE Lunar iDXA (GE Lunar, Madison, WI). The DXA has been shown to be a valid and widely-accepted method of body composition measurement (Toombs et al. 2012). Height was measured using a digital stadiometer and weight was measured on a standard scale.
Habitual dietary intake
Habitual dietary intake, using 24-h dietary recalls, was ascertained primarily to ensure participants ate normally on the days of each fMRI scan and also to provide a comprehensive picture of their habitual diet for reporting purposes. Habitual dietary intake of participants was assessed using an online multiple-pass 24-h recall (Automated Self-Administered 24-h Dietary Recall, 2014, Washington D.C.). A total of 7 dietary recalls were obtained, including: one during their first visit, three during the week prior to the first condition, and three during the week prior to the second condition. The first four dietary recalls were used to inform each participant’s diet on the day of each fMRI scan. The days of the recalls were determined randomly and included two weekdays and one weekend day. Studies have shown the validity of online 24-h recall tests. Studies have also shown computerized recalls to have relatively high accuracy (Conway et al. 2003, 2004) as well as having the advantage of reducing the burden on participants and researchers compared to other dietary assessment methods (Probst and Tapsell 2005).
Dietary intake on the day of fMRI scans
On the day of the first condition, participants were encouraged to consume a breakfast and lunch based on the information from their dietary recalls (as noted above), including: 1) similar macronutrient content; 2) similar energy content; and 3) similar amounts of food from specific food groups. Each participant was provided a portable weigh scale (Ohaus, Parsippany, NJ) and a blank food record in order to increase precision of recording. On the subsequent condition, each participant consumed foods identical to their first condition (including amounts) which was verified with weighed food records. In this way, dietary intake was representative of habitual eating and nearly identical on the days of each fMRI scan. Finally, the food record results were then entered into ESHA Food Processor software (ESHA, Salem, OR) for processing and analysis.
Visual analog scales
Prior to fMRI scanning on each condition, participants were given a visual analog scale (VAS), which asked several questions including: 1) How hungry do you feel, 2) How full do you feel, 3) How strong is your desire to eat, 4) How much to you think you could eat now, 5) What is your urge to eat, and 6) What is your preoccupation with thoughts of food? These questions have been used previously in other studies (Beechy et al. 2012). Furthermore, VAS have been shown to be a quick and accurate way to obtain pertinent data (Beechy et al. 2012; Ernst and Vingiano 1989). Individuals answered these questions by making a mark on a 100 mm straight line, with the two extreme answers to each question placed on opposite ends of the line. Distance measurements from the extreme answers provided participative measures of how strongly a participant felt about the specific question towards one extreme or the other.
Functional MRI task stimuli
Pictures used for the fMRI behavioral task were comprised of four groups: 1) low-energy foods (n = 120), including sub groups of vegetables (n = 34), fruits (n = 74), fish (n = 6), and whole grains (n = 6); 2) high-energy foods (n = 120), including sub-groups of candy (n = 10), baked goods (n = 32), ice cream (n = 18), and high-fat restaurant foods (n = 60); 3) complex visually appealing distractor pictures (n = 120), including sub groups of vegetation (n = 16), flowers (n = 80) and minerals (n = 24) in order to control for visual richness and general interest; and 4) blurred pictures (n = 360) corresponding to each picture used from all of the above categories. The blurred pictures were also shown to control for visual stimulation due to the visual complexity and colors presented by each picture. Forty of the pictures from each group were selected from a previous study that has shown high test-retest reliability (Killgore et al. 2003). The other 80 pictures selected for each group were similar foods and presentations from the reference study (Killgore et al. 2003). One group of 60 pictures from each subset (low-energy, high-energy, and distractor) was randomly selected for the first fMRI scan on the first visit. The remaining group of 60 pictures from each group was used in the subsequent fMRI scan on the second visit. Equal numbers of pictures from each subgroup were present in either group of 60 pictures. This was done to ensure that each picture used in the scans maintained a novel effect for the participant while at the same time it assured that differences in scans were not due to the participant being presented with personally appealing or unappealing foods during one condition, as all pictures had an equivalent content picture in either group. Furthermore, this assured that the high- and low-energy groups were balanced for food types in both conditions.
Images were displayed by means of an MR-compatible LCD monitor (Cambridge Research Systems, Rochester, UK) and mirror mounted on the head coil in the MRI scanner.
Participants were shown blocks of photos of the same type of stimuli with 10 photos in each block (order randomized); blocks were alternated with blocks of blurred images in random order. During each block, the participants were asked to respond to a question by pressing a button with their right hand. For food items, participants were asked to decide whether the picture was a breakfast item or a dinner item. For distractor stimuli, participants were asked whether the picture was predominantly warm or cool in color. For blurred control stimulus pictures, participants were asked to press a single button every time the photo changed. These discriminations were used to help eliminate mind wandering by keeping the participant on task and engaged. Each picture was displayed for 2.5 s with an inter-stimulus interval of 0.5 s. Participants were presented with a total of 36 blocks consisting of 10 photos each over the two conditions.
MRI data acquisition
Functional MRI data were acquired using a Siemens TIM-Trio 3.0 T MRI scanner (Siemens Trio, Erlangen, Germany). A 7-min high resolution T1-weighted MPRAGE structural scan with the following parameters was collected from each participant at the beginning of the scan session: TE = 2.26 ms; TR = 1900 ms; flip angle = 9°; matrix size = 256 × 215 mm; field of view = 218 × 250 mm; 176 slices; slice thickness = 1 mm; voxel size = 0.977 × 0.977 × 1 mm.
Participants then completed two functional runs with each run lasting approximately 10 min with a small break in between. Total time in the scanner was approximately 30 min. We collected T2*-weighted images using the following parameters: TE = 28 ms; TR = 2000 ms; flip angle = 90°; matrix size = 64 × 64; field of view = 220 × 220 mm; 40 slices; slice thickness = 3 mm; voxel size = 3.4 × 3.4 × 3 mm; 270 total acquisitions.
The fMRI data were preprocessed and analyzed using the Analysis of Functional NeuroImages (AFNI) suite of software (Cox 1996). All functional runs were time shifted, corrected for participant motion, and spatially filtered using a 5 mm FWHM Gaussian kernel. Structural scans were co-registered with the functional scans. Initial spatial normalization was accomplished by aligning structural scans to the atlas of Talairach and Tournoux (1988). Further normalization was accomplished by aligning all structural scans to a custom template using Advanced Normalization Tools software (ANTs; Version 1.9; http://sourceforge.net/projects/advants/) (Avants et al. 2008; Lacy et al. 2011; Motley and Kirwan 2012; Yassa et al. 2011). A regression analysis was conducted on the functional data set using AFNI program 3dDeconvolve. Six regressors coding for motion (three translations and three rotations) were included as conditions of no interest. Three additional regressors were created coding for high-energy blocks, low-energy blocks, and distractor stimuli blocks. Blocks were modeled by convolving the standard hemodynamic response function with a 30-s boxcar function. The blurred control image blocks were used as an implicit baseline in the model. Data were then spatially normalized and a group analysis was performed. The conditions of interest were: morning low-energy, morning high-energy, evening low-energy, evening high-energy. The distractor picture data were used to assure that the reactions were not similar to nonfood items. A quality check for comparison was also completed by performing an ANOVA test using time of day and stimulus type (high-energy, control).
For the group-level analysis, we conducted a 3-way ANOVA whole brain analysis with scan time (morning, night) and stimulus type (high-, low-calorie, and distractor) as fixed factors and participants as a random factor. In order to correct for multiple comparisons the results of these tests were thresholded with a voxel-wise p-value of P < 0.001 and 40 contiguous voxels for the spatial extent threshold. These parameters were determined by conducting Monte Carlo simulations to yield an overall p-value of P < 0.05. Finally, thresholded statistical maps were subjected to visual inspection to find meaningful effects. Average beta coefficients were extracted from within the areas of activation and were subjected to further analysis (ANOVA and T-tests) in SPSS.
Regression analyses, condition differences, and Condition × Food Type interactions were conducted in SPSS. Habitual food intake, participant demographics, and VAS scale scores were analyzed with fMRI results in areas of activation for main effect of time and main effect of stimulus looking for correlations using the statistical software PC-SAS (version 9.3, SAS Institute, Inc., Cary, NC). Statistical significance for all results were p < 0.05. Paired t-tests were used to determine differences in VAS scores and energy intake between conditions.
Participant characteristics (n = 15)
23.33 ± 1.02
167.91 ± 6.09
64.38 ± 7.04
22.93 ± 2.69
Body fat (%)
32.46 ± 5.99
Habitual dietary intake and dietary intake on the day of each fMRI scan
Based on the 24-h dietary recalls, habitual dietary intake was 1953 ± 466 kcal per day. Dietary intake on the day of the fMRI, assessed using the weighed food records, was 2009 ± 544 kcal on the day of the morning condition and 1975 ± 504 kcal on the day of the evening condition (less than a 2 % deviation on their scan days). The difference between energy intake on the scan days was not significant between conditions (F = 1.16, P = 0.78).
Visual analog scales
Visual analog scales
Morning condition n = 15
Evening condition n = 15
How hungry do you feel?
5.29 ± 2.61
6.55 ± 2.65
How full do you feel?
1.54 ± 1.45
2.04 ± 2.02
How strong is your desire to eat?
4.94 ± 2.56
6.74 ± 3.00
How much to you think you could eat now?
5.05 ± 1.82
6.48 ± 2.08
What is your urge to eat?
4.45 ± 2.52
6.27 ± 2.86
What is your preoccupation with thoughts of food?
3.00 ± 1.90
5.04 ± 3.13
The group-level analysis revealed several significant clusters that demonstrated main effects for time of day and for stimulus type as described below. However, no significant brain activations were found for a direct interaction between time of day and food stimulus type (P > 0.05). The quality check performed using stimulus type (high-energy, distractor) and time of day showed two small clusters of activation in the right occipital lobe where there was a significant time of day by stimulus (high-energy, distractor) interaction. However these clusters did not overlap with any of our other functionally-defined ROIs.
Main effect of time of day (morning vs. evening)
Regions of interest (ROI) for main effect of time
Stimulus × time interaction
R. Lingual gyrus
R. Middle temporal gyrus
R. Middle occipital
R. Ventral striatum/amygdala
Main effect of food stimulus type (high- vs. low-energy)
Regions of interest (ROI) for main effect of high energy vs. low energy stimulus
Stimulus × time interaction
R. Middle occipital gyrus
L. Lingual gyrus
R. Fusiform gyrus/parahippocampal cortex
R. Superior parietal lobule
R. Medial frontal gyrus
L. Fusiform gyrus/parahippocampal cortex
L. Medial frontal gyrus
L. Superior temporal gyrus
Anatomical region of interest analysis
As our experimental design was similar to that of other studies, we ran similar masked ROI analyses for archival purposes to assure that our results were similar to past studies. We compiled a list of the specific regions of interest (ROIs) from previous studies examining the effect of food type on fMRI activation (Bruce et al. 2013; Killgore et al. 2003; Murdaugh et al. 2012; Sweet et al. 2012; Yokum et al. 2011). We examined 41 specific ROIs listed in Supplemental Table 1 using segmentations obtained for each individual participant using the FreeSurfer software package. For nearly all areas, high-energy stimuli produced higher responses than low-energy stimuli. Areas and statistical results are reported in a supplemental table.
The results of this study suggest that brain activation is consistently higher in response to high-energy food pictures compared to low-energy food pictures for several areas of the brain, regardless of time of day. This was consistent with our original hypothesis and agrees with previous research (Killgore et al. 2003; Sweet et al. 2012; Yokum et al. 2011).
In addition, when examining the main effect of time of day (morning vs. evening) on neural activation, significant results were seen in six areas of the brain: the putamen, lingual gyrus, middle temporal gyrus, middle occipital, parahippocampus/ hippocampus, and ventral striatum/amygdala. In each of these areas, there was a reduction in activation during the evening time compared to morning. Although high-energy food stimuli still produced greater activation than low-energy food stimuli, these results were contrary to our initial hypothesis that activations would be greater in the evening to both high- and low-energy food. Several of these areas (e.g., lingual gyrus, middle temporal gyrus, and middle occipital gyrus) are part of pathways involved in the processing of visual stimuli. One of the reported areas of significant activation, the amygdala, is thought to play an important role in the visual evaluation of food stimuli and is thought to be responsible for evaluating novel stimuli to determine if it may be used as a possible food source (Rolls 1999).
The two other significant areas reported were the hippocampus and ventral striatum which have been implicated, along with the amygdala, as part of dopamine pathways related to motivation and reward (Bromberg-Martin et al. 2010). The final reported area of interest is the putamen, which has been shown to be involved specifically in reward history-based action selection (Muranishi et al. 2011).
The simplest interpretation of this study is that high-energy foods produce a greater neural response than low-energy foods in some areas of the brain. Thus, if visual responses to high-energy foods indeed predict greater motivation for food and increased energy intake, then avoiding pictures, advertisements, other high-energy food stimuli, or actual high-energy food is warranted for those actively attempting to reduce energy intake. Furthermore, susceptibility to visual stimuli seems to be greater in the morning; thus, on the surface, these data suggest that greater care should be taken to avoid visual food cues in the morning.
However, considered in the context that food consumption may be greatest during the evening (not the morning), significant questions arise from these data. Could decreased sensitivity to food stimuli in the areas of the brain associated with reward (as shown in the present study) actually evoke a greater motivation for food in the evening (in order to attain the same level of reward as earlier in the day)? Interestingly, we showed that even though hunger and fullness levels from the VAS’s were not different by condition, thoughts and preoccupation with food were higher during the evening. Is this due to physiology (Allison et al. 2014) or habitual or environmental factors? Our study cannot answer these speculative questions but underscore the need for more research to understand the relationship among visual food stimuli, neural reward responses, and time of day of assessment, and especially, whether or not these are related to actual eating behavior.
Another potential implication from the results of this experiment may relate to scan protocol. As the results of this study suggest that time of day can influence the neural responses to visual food stimuli, careful consideration should be given to when study participants are scanned or tested. If groups of participants are scanned too many hours apart, systematic biases in neural responses may occur. Thus, consistency in time of day of fMRI assessments among participants may be needed to answer certain research questions.
Strengths and limitations
While there are significant strengths in this study, there are also limitations. The population studied was fairly homogenous. All participants were young females and almost all were Caucasian, with the exception of two participants. Furthermore, they were mainly in the normal weight BMI category. Therefore, generalization to a population with a larger BMI is not possible. Another limitation was the small sample size used, which may have compromised power for several of the VAS scale responses that approached significance. In addition, the study only looked at acute effects as no intervention was used.
This study confirms the scientific literature that there is a greater neural response to high-energy foods compared to low-energy foods, regardless of time of day. Furthermore, our data adds to the literature by indicating that there is a dampening effect in several important brain areas related to reward and visual processing in the evening compared to the morning. These findings highlight the importance that time of day of assessment of visual food stimuli can play both practically, and perhaps clinically. Further study of the role of time of day on visual food processing and how this relates to eating behavior is needed.
TDM, CBK, LED, and JDL designed research; TDM and CBK conducted research; TDM, CBK, and JDL analyzed data; TDM, CBK, LED and JDL wrote the paper; TDM had primary responsibility for final content. All authors read and approved the final manuscript.
All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki declaration of 1975, and the applicable revisions at the time of the investigation. Informed consent was obtained from all patients for being included in the study.
Conflict of interest
Travis Masterson, James D. LeCheminant, C. Brock Kirwan, and Lance E. Davidson state that they have no conflict of interested associated with this project.