Background: The loss of self-control or inability to resist tempting/rewarding foods, and the development of less healthful eating habits may be explained by three key neural systems: (1) a hyper-functioning striatum system driven by external rewarding cues; (2) a hypo-functioning decision-making and impulse control system; and (3) an altered insula system involved in the translation of homeostatic and interoceptive signals into self-awareness and what may be subjectively experienced as a feeling.
Methods: The present study examined the activity within two of these neural systems when subjects were exposed to images of high-calorie versus low-calorie foods using functional magnetic resonance imaging (fMRI), and related this activity to dietary intake, assessed by 24-hour recall. Thirty youth (mean BMI = 23.1 kg/m2, range = 19.1 - 33.7; age =19.7 years, range = 14 - 22) were scanned using fMRI while performing food-specific go/nogo tasks.
Results: Behaviorally, participants more readily pressed a response button when go trials consisted of high-calorie food cues (HGo task) and less readily pressed the response button when go trials consisted of low-calorie food cues (LGo task). This habitual response to high-calorie food cues was greater for individuals with higher BMI and individuals who reportedly consume more high-calorie foods. Response inhibition to the high-calorie food cues was most difficult for individuals with a higher BMI and individuals who reportedly consume more high-calorie foods. fMRI results confirmed our hypotheses that (1) the "habitual" system (right striatum) was more activated in response to high-calorie food cues during the go trials than low-calorie food go trials, and its activity correlated with participants’ BMI, as well as their consumption of high-calorie foods; (2) the prefrontal system was more active in nogo trials than go trials, and this activity was inversely correlated with BMI and high-calorie food consumption.
Conclusions: Using a cross-sectional design, our findings help increase understanding of the neural basis of one’s loss of ability to self-control when faced with tempting food cues. Though the design does not permit inferences regarding whether the inhibitory control deficits and hyper-responsivity of reward regions are individual vulnerability factors for overeating, or the results of habitual overeating.
Alex’s Notes: A handful of studies have sought to determine the underlying mechanisms for why obese individuals have such an intimate relationship with food. There is evidence suggesting that the difficulty to resist highly palatable, calorie-rich foods represents a special case of addictive behavior with similarities to other addictions. Using past research as a base, the current study utilized fMRI to investigate brain activity during high- and low-calorie food-specific tasks and hypothesized that,
- There is a hyper-functioning striatum driven by external highly rewarding foods (impulsive system);
- There is a hypo-functioning decision-making and impulse control center (reflective system);
- The strength of the two previous neural systems is modulated by urges and cravings.
The subjects were 30 healthy 14-22 year old (avg. 19.7-years) men and women with an average BMI of 23.1, no neuropsychiatric disorders, medications, or health issues that could impact the results. This age range of subjects was purposely chosen as the prefrontal cortex has not yet fully developed, increasing the potential for making poor food choices.
The subjects came to the lab twice; first to complete the Structured Clinical Interview for DSM-IV (SCID) to exclude individuals who meet the criteria for current psychoses, anxiety, or bipolar disorders, as well as two behavioral tasks (an IQ test and a working memory and executive functioning test), and second, to undergo the fMRI scan. Importantly, the subjects were asked to eat normally before they arrived for the fMRI scan, so as not to be in a “deprived state,” and this was confirmed via a hunger rating scale. A single, in-person 24-hour dietary recall was also conducted, with total daily servings of low-calorie foods equating total intake of fruit and vegetables, while high-calorie foods were the sum of fatty foods and sugar-sweetened foods.
When actually within the fMRI scanner,
“Participants performed two food-specific go/nogo tasks in the scanner as follows: 1) a low-calorie food go and high-calorie food nogo task (LGo task), and 2) high-calorie food go and low-calorie food nogo task (HGo task)… Participants were asked to press a button as soon as possible to the go trials (low-calorie food pictures in LGo task and high-calorie food pictures in HGo task), and to withhold responses to the nogo trials (high-calorie food pictures in LGo/HNogo task and low-calorie food pictures in HGo/LNogo task). Examples of low-calorie food images included cucumbers, celery, broccoli, and carrots. Examples of high-calorie food images included chocolate bars, cookies, ice cream, and potato chips. All images of the foods observed are commonly available in everyday life.”
The baseline data revealed only that there was a significant gender difference in the consumption of low-calorie foods, with females consuming about 3 servings per 1000 kcal compared to the males 1.6 servings per 1000 kcal. Overall results from the task itself demonstrated that there was a higher false alarm rate in the LGo task, as well as a smaller decision bias, together suggesting that the subjects made more errors and had a harder time inhibiting responses to high-calorie food cues in the LGo task. When controlling for age and gender,
“Reaction time for the go trials in the HGo task was negatively correlated with both BMI (r = -.60, p < .01) and high-calorie food consumption (r = -.50, p < .05), suggesting the habitual response to the high-calorie foods was greater for individuals with higher BMI, and individuals who reportedly consumed more high-calorie foods. The decision bias C for the LGo task negatively correlated with both BMI (r = -.47, p < .05) and high-calorie food consumption (r = -.49, p < .05), suggesting the inhibiting response to the high-calorie foods was more difficult for individuals with higher BMI and individuals who reported consuming more high-calorie food.”
So what was going on upstairs?
In line with the authors hypothesizes, it was found that the reflective system showed more activation during the nogo trials than the go trials for high-calorie foods, and this was significantly negatively correlated with BMI and reported high-calorie food consumption. Females also showed a trend to have more reflective system activation that approached statistical significance (p=0.06). During the go trials, only the high-calorie food cues were associated with higher activity in the right striatum, which was positively associated with both BMI and high-calorie food consumption.
So what does this mean?
First, it is clear that the response to high-calorie food stimuli was significantly greater for those with higher BMI and who more regularly consumed high-calorie foods, and it was more difficult for these same individuals to inhibit their response to high-calorie food stimuli. Additionally, the fMRI scanning supported the hypothesizes that the impulsive system was more activated and the reflective system less activated in individuals with greater BMI and who consume more high-calorie foods. However, there was no significant correlation between BMI and high-calorie food consumption itself, which could simply be a measurement error from the dietary recall, but puts a bump in the findings. Accordingly, those with greater BMI and those who consume more high-calorie foods can be considered separate groups, thus raising two possible scenarios:
- They have a greater BMI and/or eat more high-calorie food because they have a hyper-functioning impulsive system and a hypo-functioning reflective system, or
- The neural changes are a result of their eating habits and/ or BMI.
The second scenario seems silly regarding BMI, but is definitely realistic for eating habits. When we consider that there may have been reporting errors, the scenario could become that the neural changes and greater BMI are the result of habitual high-calorie food intake rather than vice-versa.
Overall, the correlational nature of the study prevents determining a causal relationship, but it does provide food for thought.