Representation of food preference in the brain in relation to visceral obesity and stress

Obesity is rapidly developing into a great threat of epidemic proportions.
Especially visceral obesity is a key component in the development of the
metabolic syndrome. Obesity develops due to an exceeding food intake compared
to the energy requirements and especially visceral obesity is often correlated
to stress. Stress is indicated by an increased *hypothalamic-
pituitary-adrenal-axis* (HPA-axis) activity, represented by an increased
concentration of its end product, cortisol. Visceral obese subjects have higher
serum cortisol levels in comparison to normal weight people. In stressed
conditions they choose foods high in fat and carbohydrates because these foods
have a high rewarding value. However, foods high in carbohydrates may increase
the stress (cortisol)-response even more in visceral obese subjects thereby
inducing a vicious cycle. To break the vicious cycle that stress causes in
visceral obesity, the relationship between food choice and stress response
(cortisol levels) needed to be clarified and quantified. In order to do this,
first the rewarding value of foods needed to be quantified (MEC 07-3-053).
The rewarding value is represented as brain activity in certain brain areas
that have been described previously in the literature. To identify the exact
brain areas that are involved and to quantify the representation of the
rewarding value of foods in the brain, a computer game was played while
*functional magnetic resonance imaging* (fMRI) was performed in different
conditions (stress vs. rest) within different subjects (visceral overweight vs.
normal weight). The computer game gave us the ability to determine the
rewarding value of foods and explain the food choice. The involved brain areas
were identified with the fMRI.
First results show that visceral overweight subjects in stress and in the
absence of hunger prefer and consume foods that are crispy, full of taste, high
in fat and carbohydrate, in contrast to normal weight subjects. Moreover,
reward related brain areas such as the putamen show reduced activation. A study
completed by Martens et al. (MEC 08-3-076) showed that carbohydrates increased
cortisol, while protein and fat did not when compared with control (water). A
study by Vicennati et al. (2002) showed that, in contrast to a high protein/fat
meal, a high carbohydrate meal significantly increased the cortisol levels in
visceral obese subjects. Moreover, Lacroix et al. (2004) showed that in rats
high protein/fat foods reduced cortisol levels. Further research aims to
determine the effects of consumption of highly rewarding (crispy, full of
taste) high protein/fat foods compared with high carbohydrate foods on the
stress response and on the representation of rewarding/satiating effects of
those foods in the brain, on food preference, food choice and food intake,
acutely and in the long-term.

This study focuses on visceral overweight and stress, related to intake of
highly rewarding foods and food choice. The main objective of the study is
determination of the effects of consumption of highly rewarding high
protein/fat foods compared with high carbohydrate foods on the physiological
stress response (cortisol) and on the representation of rewarding/satiating
effects of those foods in the brain, on rewarding value of other foods and food
intake, acutely and in the long-term.

The study has a 4-arm cross-over design, with randomized conditions. Conditions
are rest or stress with the consumption of high protein/fat foods and rest or
stress with the consumption of high carbohydrate foods. The subjects are
brought in the rest or stress condition and will subsequently play a computer
game in the fMRI apparatus. The computer game is used to determine the
rewarding value of food while fMRI measures the representation of this
rewarding value in the brain. Per condition, the fMRI session consists of two
parts (30 min each), one before and one after the meal. Stress and rest
conditions are induced with a mathematical test. The rewarding value of food is
compared between normal weight and overweight subjects in all the different
conditions. Hereby the relationship between visceral overweight and food choice
in stress and rest conditions will be quantified. Each test-session 5 blood
samples of 5 ml each are taken for cortisol concentration measurements that
represent the HPA-axis activity.
Subjects will be tested in stress/rest, in high protein/fat vs. carbohydrate
condition, before and after a dietary phase (*4 weeks) of daily protein/fat vs.
carbohydrate-consumption as snacks.
In total the study consists of 3 times 4 test-sessions per subject: before and
after the dietary phase of daily protein/fat-snack consumption (or
carbohydrate) and after the dietary phase of daily carbohydrate-snack
consumption (or protein/fat). A complete test-session takes 3.5 h.

1. Stressed condition
The stressed condition is a situation where the subjects are brought in a state
of actual mental stress. The control condition is a condition of rest applied
to the same subjects. The two conditions will be compared

2. High protein/fat vs. carbohydrate consumption
During the test-sessions (meal) as well as on a daily basis (snacks) subjects
will consume high protein/fat vs. high carbohydrate foods. The food items will
be highly rewarding, i.e. full of taste, creamy and crispy.
The amount of foods consumed will be calculated according to the subjects daily
energy requirements, 25 % per test-meal during the test-days and also 25% per
day during the snack-dietary phase of daily consumption.

As this is a cross-over study, all subjects will undergo the experiment in the
stress and rest condition and in the high protein/fat and carbohydrate
condition.
The responses of the visceral overweight subjects will be compared to those of
the normal weight subjects per condition. Within the subject groups the
different conditions will be compared.

This research is neither beneficial nor harmful to the subjects. The study
consists of 3 times 4 test-sessions per subject of 3.5 h each. All sessions
include two blocks of 30 min that will be performed in the fMRI apparatus. fMRI
is a non-invasive standard brain-imaging method without any significant risks
(See chapter K4A for standardized and approved methods for conducting fMRI
experiments involving human subjects). It is a technique that utilizes magnetic
fields and low energy radio frequencies to visualize brain structures and brain
function. During screening subjects with metallic fragments in their body will
be excluded from the study since the fMRI magnet exerts a force on
ferromagnetic objects. The blood sampling in this study (5 blood samples of 5
ml each test-session) does not include any other risks for the subjects, apart
from its usual risk of minor bruising. The computer game is a standard memory
game with pictures of food items and contains questionnaires regarding the food
preference.