A Simple Model Predicting Individual Weight Change in Humans

April 19, 2010
6:0
Journal of Biological Dynamics
NHANES

Vol. 00, No. 00, Month 2009, 122
GUIDE
A Simple Model Predicting Individual Weight Change in
Humans
Diana M. Thomas1 and Corby K. Martin2 and Steven Heymsfield 3 and Leanne M.
Redman 2 and Dale A. Schoeller 4 and James A. Levine5
1
Department of Mathematical Sciences, Montclair State University, Montclair, NJ;

Pennington Biomedical Research Center, Baton Rouge, LA ; 3 Merck & Company,
Rahway, NJ ; 4 Department of Nutritional Sciences, University of Wisconsin-Madison
5
Department of Medicine, Endocrine Research Unit, Mayo Clinic and Mayo Foundation,
Rochester, MN
2
(v2.2 released November 2008)
Keywords: energy balance equation, metabolic adaptation, non-exercise activity

thermogenesis, dietary adherence
AMS Subject Classification: 58F15, 58F17, 53C35
Excessive weight in adults is a national concern with over 2/3 of the US population deemed
overweight. Because being overweight has been correlated to numerous diseases such as heart
disease and type 2 diabetes, there is a need to understand mechanisms and predict outcomes
of weight change and weight maintenance. A simple mathematical model that accurately
predicts individual weight change offers opportunities to understand how individuals lose and
gain weight and can be used to foster patient adherence to diets in clinical settings. For
this purpose, we developed a one dimensional differential equation model of weight change
based on the energy balance equation is paired to an algebraic relationship between fat free
mass and fat mass derived from a large nationally representative sample of recently released
data collected by the Centers for Disease Control. We validate the models ability to predict
individual participants weight change by comparing model estimates of final weight data
from two recent underfeeding studies and one overfeeding study. Mean absolute error and
standard deviation between model predictions and observed measurements of final weights
are less than 1.8 1.3 kg for the underfeeding studies and 2.5 1.6 kg for the overfeeding
study. Comparison of the model predictions to other one dimensional models of weight change
shows improvement in mean absolute error, standard deviation of mean absolute error, and
group mean predictions. The maximum absolute individual error decreased by approximately
60% substantiating reliability in individual weight change predictions. The model provides a
viable method for estimating individual weight change as a result of changes in intake and
determining individual dietary adherence during weight change studies.
1.
Introduction
The Centers for Disease Control currently estimates that approximately 67% of
the US adult population is overweight, with body mass index (BMI) between 25
and 29.9 and 34% is obese (BMI > 30). BMI levels above 25 have been linked to
Corresponding
author. Email: thomasdia@mail.montclair.edu
ISSN: 1751-3758 print/ISSN 1751-3766 online

c 2009 Taylor & Francis

DOI: 10.1080/1751375YYxxxxxxxx
http://www.informaworld.com
April 19, 2010
6:0

2
NHANES
Taylor & Francis and I.T. Consultant
diseases and health related consequences such as coronary heart disease, type 2 diabetes, cancers (endometrial, breast, and colon), hypertension, dyslipidemia, stroke,
liver and gallbladder disease, sleep apnea and respiratory problems,osteoarthritis,
and gynecological problems [66]. In fact, it is estimated obesity related diseases
currently account for approximately 9.1% of total national health care costs [66]
With such a large proportion of the population considered overweight, mathematical models can provide intuition, insight, and solutions for effective weight
loss and weight loss maintenance [2, 46, 18, 19, 31, 32, 53, 55]. These differential
equation models apply state variables that track the quantity of macronutrients
that provide energy, namely, carbohydrates, protein, and fat. The common goal
of the models is to predict weight change as a result of changes in dietary intake
and/or energy expenditure.
One particular application of weight change models is to foster individual adherence to diets during caloric restriction. Using weight predicted curves, an individuals actual weight loss can be compared to predicted weight loss to test for
compliance. Although there are a variety of existing weight prediction models,
they either require numerous individual parameter estimates, posing challenges for
clinical implementation, or the maximum individual subject error obtained during
validation is too high to reliably use for adherence purposes. To fill this knowledge
gap, we sought to develop a model that predicts weight change that simultaneously
satisfies the following conditions:
(1) The model requires minimal input of easily measurable baseline information
(age, height, weight and gender).
(2) The model predicts final weight during weight loss with a high degree of
accuracy (low mean absolute error in predicted weight versus actual weight
when tested on individual weight loss data sets).
(3) The model provides reliable individual estimates (low standard deviation
in the mean absolute error of predicted weight versus actual weight and low
maximum absolute error).
The model presented here is the first model to simultaneously satisfy these goals
and provides a significant advance in application of weight loss models to clinical
settings.
In the next section, we provide a brief history of existing weight change models
from the literature followed by the development of our model. Section 3 provides
existence and non-negativity of solutions criteria for our developed model. In Section 4 we validate the models capacity to reliably predict individual weight change
from two recently conducted weight loss experiments and one over-feeding study.
We also compare the model results with a recently published model that satisfies
our first criterion and show that the developed model provides more reliable and
accurate estimates of final individual weight.
2.
2.1.
Model
History of Dynamic Weight Change Models
One approach to modeling human weight change is to incorporate all reasonable

physiological factors in order to reflect reality. This type of model provides insight
into how perturbations to any component involved during changes in energy intake
ultimately effect weight change. Applying this perspective has led to the carefully developed system of compartmental equations with state variables tracking
changes in energy derived from protein, fat and carbohydrates [18, 19]. The Hall
April 19, 2010
6:0
NHANES
model identifies detailed movement of energy from intake to expenditure providing

important information into the mechanisms behind weight change.
The Hall model requires estimation of the amount of grams of protein, carbohydrates and fat a subject consumes per day along with numerous other parameters
[18, 19]. In order to reduce the amount of initial data and parameter estimations,
several one and two-dimensional models have been developed by applying simplifying assumptions [2, 46, 31, 32, 55]. These simplifications are derived by assuming
glycogen stores can be modeled by a time averaged constant, resulting in elimination of the carbohydrate mass equation. Elimination of the protein equation is
obtained by algebraically correlating fat mass to fat free mass (F F M ).
Some models [2, 46, 31, 32, 55] either assume that F F M is a linear function
of fat mass or assume F F M can be modeled as a time averaged constant. Chow
and Hall explored a more sophisticated formulation by coupling the F F M model
proposed by the Forbes equation through a two dimensional dynamic model [4].
Forbes conjectured that F F M and fat mass were companions; increases/decreases
in fat mass will be followed by increases/decreases in F F M [12]. Based on this
hypothesis, Forbes fit mean data for 167 women of similar stature to a log linear
regression equation to arrive at what is now known as the Forbes curve:
F F M = 10.4 ln(F/D)
where F represents fat mass in kg and D 2.55.
We develop here a one-dimensional model that incorporates a newly developed
F F M F relationship [54] along with several other terms based on recent experimental observations. We refer the reader to [55] for complete biological details on
the full model development and present here a description of the essential components of the model, highlighting the modifications beginning with the energy
balance principle.
2.2.
The Energy Balance Principle
The energy balance principle discussed in physiology and nutrition literature is

based on the application of the first law of thermodynamics to an open system [11].
The human body is considered an open system because energy can be added to
the system by input of mass flow in the form of food. The human energy balance
equation takes the form,
R = I E,
(1)
where R is the rate of energy stored/lost, I is the rate of energy intake, and E is
the rate of energy expenditure. The quantities R, I, and E are typically measured
in kcal/d [27, 41].
2.3.
The rate of energy stored/lost: R
Energy that is not used by the body is stored in the form of glycogen, lipids and
protein [24]. Glycogen is a storage molecule of the bodys short-term energy resource, glucose [14]. Because of the regulatory mechanism that works to maintain
the glucose/glycogen level in a narrow range, we can model glucose/glycogen energy measured in kcal by a time-averaged constant, G. We remark here that this
assumption has been compared successfully to observed data [4].
April 19, 2010
6:0

4
NHANES
Lipids (mainly in the form of fatty acids) are the nutrients containing the most
energy per unit mass[24]. Unlike glucose, fat can be stored in large quantities for
extended lengths of time in the form of triacylglycerols within the adipose tissue
spread throughout the human body. Therefore, fat mass is the main long-term
energy storage mechanism of the human body. We denote the total kg of fat mass
of the body at time t by the function F (t).
Proteins building blocks, amino acids, can be broken down and transformed to
glucose to be used for energy through amino acid metabolism [24]. Although there
are small amounts of protein stored in the liver, the major portion is contained in
the bodys muscle tissue [14]. Protein is a component of F F M and so we track
changes in protein energy through changes in F F M as discussed previously [4].
As mentioned in the introduction, the two-dimensional model [4] relies on the
Forbes F F M function of fat mass (F ). Specifically, Forbes observed an algebraic
relationship between fat mass and F F M [12] by fitting a log linear curve through
mean data for 167 women of similar stature [12],
F F M (t) = 10.4 ln(F (t)/D)
An analogous male Forbes model was developed using body composition data from
approximately 200 men of average stature [54]:
F F M (t) = 13.8 ln(F (t)/S)
where S 0.29.
The Forbes model has been established to predict the change in F F M in numerous weight change studies with a high degree of accuracy for group mean
data[17, 54]. It may be tempting to think of the Forbes curve as a trajectory
that body composition travels during weight loss, however, the model does not
determine change in F F M by estimating the baseline and final time body composition [54]. The Forbes curve estimates for the change in F F M rest on the fact
that individual data follow a parallel curve of identical slope and in this manner
accurately determines change without identifying beginning and endpoints.
As a stand alone equation, the parameter D can be computed using baseline
body composition data to determine the correct translate. It was shown [54] that
D is a reflection of age, height,and race. In addition, the Forbes model has a range
of positive fat mass that yields negative F F M values. Moreover, the intercept of
the Forbes model is undefined.
The two dimensional model from [4] successfully incorporates the Forbes model
by considering F and F F M as state variables and applying the p-ratio (proportion
of body energy mobilized as protein during weight change) derived from the Forbes
model [4]. Reduction of the two-dimensional model to a one-dimensional model
(Equation 25 [4]) reduces the dimension of the model, but will require calculation
of the parameter D.
In order to develop a universal F F M function that is representative of the current national population, has a biologically realistic intercept when F = 0, and can
be immediately applied for each individual subject in a one dimensional energy
balance model, a class of fourth order polynomials of F F M as a function of F
that considers age, height, gender and race was formulated based on the recently
released body composition data collected by the Centers for Disease Control under
the The National Health and Nutrition Examination Survey (NHANES). NHANES
is a program designed to assess the health and nutritional status of adults and children in the United States. NHANES performs a continuous, nationally representa-
April 19, 2010
6:0
NHANES
tive health survey of the civilian, non-institutionalized United States population,

collecting data on about 5000 persons each year from interviews, physical examinations, and medical tests including bone densitometry. In 1999 NHANES began
performing dual energy X-ray absorptiometry (DXA) whole body measurements
on survey subjects 8 years old and older in three mobile examination centers. DXA
measurements are widely accepted as the gold standard for human body composition measurements. The DXA whole body data obtained from the mobile exam
centers was compiled by the NHANES study group and released on the Center
for Disease Control (CDC) website [29]. The F F M models developed earlier [54]
provided comparable estimates for F F M to the Forbes model after entering age,
height, and gender as initial inputs.
The first major modification to our prior model [55] is the replacement of the
linear F F M function of by the F F M functions based on NHANES data. These
F F M model terms were statistically determined through a regression analysis
[54],
Females:
F F M (t) = 72.1 + 2.5F (t) 0.04(A0 + t/365) + 0.7H

0.002(A0 + t/365) 0.01F H 0.04F (t)2
+ 0.00003F (t)2 (A0 + t/365)
+ 0.0000004F (t)4 + 0.0002F (t)3 + 0.0003F (t)2 H
0.000002F (t)3 H
Males:
F F M (t) = 71.7 + 3.6F (t) 0.04(A0 + t/365) + 0.7H

0.002F (t)(A0 + t/365) 0.01F (t)H + 0.00003F (t)2 (A0 + t/365)
0.07F (t)2 + 0.0006F (t)3 0.000002F (t)4 + 0.0003F (t)2 H
0.000002F (t)3 H
where F F M (t) represents the kg of F F M on day t, F (t) is the kg of fat mass on
day t, H denotes height in cm, and A0 is baseline age.
Separating R from Equation 1 in the three different components, kg of glucose/glycogen, F F M (t), and F (t) we have

R = cl
dF F M
dt

+ cf
dF
dt
(2)
where cl is the energy density of F F M , cf is the energy density of fat mass, and
the derivative of F F M is taken implicitly. The values of cl and cf were obtained
from [26, 51] and appear in Table 2.
2.4.
The rate of energy intake, I
The rate of energy intake, I has been historically difficult to measure [19]. Past
dynamic models have assumed that I during weight change is constant and de-
April 19, 2010
6:0

6
NHANES
termined by target energy intake requirements [2, 5, 6, 18, 19, 32, 55? ]. We now
describe the limitations behind this assumption and provide a novel method based
on modern technology and measurements to assess I as a time varying function
dependent on an individual subject.
First, baseline intake is determined by applying steady state (zero energy balance) at the conception of the study:
I E =0
The gold standard for determining E is through doubly labeled water (DLW). The
technique involves enriching the body water of a subject with an isotope of hydrogen and an isotope of oxygen, and then determining the washout kinetics of
both isotopes as their concentrations decline exponentially toward natural abundance levels. It is not feasible to obtain DLW data for every study or individual so
we applied individual DLW data from National Academy of Sciences/Institute of
Medicine (NAS/IOM) database of over 200 subjects to fit a quadratic regression
formula that defines E as a function of body mass, W . Specifically,
EF = 0.0278W 2 + 9.2893W + 1528.9
(3)
EM = 0.0971W 2 + 40.853W + 323.59

where the subscript F represents females and M represents males.
As stated earlier, when a study prescribes caloric restriction, it is tempting to
model I as the prescription demands. For example, if we prescribe a change in
that is 25% below baseline intake, we would model I = I I
baseline intake (I)
where
I = 0.25I
and I represents baseline energy intake.
However, compliance to such dietary restrictions are rare. In an analysis of 181
participants placed on popular diets, only one subject in the study was 100% adherent by the end of one year [1]. Furthermore, it has been consistently observed
that compliance is a decreasing function of time [1, 7, 25]. To determine I, we applied data from studies which calculated energy intake using DXA measurements
of changes in energy stores and DLW measurments of energy expenditures during
weight loss. Specific details on these calculations appear in the model validation
simulations in Section 4.
2.5.
The rate of energy expenditure, E
The rate of energy expenditure consists of four different quantities; the kcal/day
used for dietary induced thermogenesis (DIT), volitional physical activity (PA),
basal metabolic rate (RMR), and spontaneous physical activity (SPA) [15]:
E = DIT + P A + RM R + SP A.
(4)
DIT is the energy involved in processing food. This consists of digestion, absorption, metallization, storage and transport of ingested food and is estimated to
account for 4-15% of total energy expenditure [15]. Although DIT was observed to
be sensitive to amount of protein ingested, we do not examine this aspect of DIT
[58]. In the case of an aggregated composition of food intake, we assume that DIT
April 19, 2010
6:0
NHANES
is a direct proportion of the rate of energy intake, I, with proportionality constant,

. In [20], the adjustment to calorie reduction on DIT was examined by multiplying by an adjustment factor. In fact, such an adjustment has also been observed
in overfeeding studies by a factor of up to 19% [20, 37]. We model the adjustment
by a multiplier that is greater than 1 for the overfeeding case and less than one in
the calorie restriction case. As in [20] we absorb the multiplier into one constant
with ranges provided in Table 1.
Therefore, the energy expended for dietary induced thermogenesis is
DIT = I.
By physical activity, PA, we specifically mean volitional exercise such as sports
and fitness related activity. PA can accounts for 20-40% of total energy expenditure
[15]. Strenuous exercise or starvation cause glycogen stores to deplete [14]. As a
result the liver begins to produce ketone bodies as an alternative energy source.
Because we are not modeling extreme cases such as starvation, we assume the
physical activity is light to moderate and does not deplete glycogen stores. Some
PA is what we refer to as weight bearing activity and some of PA is non-weight
bearing. Weight bearing activities involve activities that require us to carry our
body weight. For example, walking and running are examples of weight bearing
activities. As in the two dimensional model [4], we assume that energy used for
non-weight bearing activity is negligible.
Thus, the rate of weight bearing exercise is estimated as a direct proportion of
body mass and so we model PA by,
P A = mW
where m is the proportionality constant in kcal/kg/day and W , total body mass,
is the sum of F F M and F .
An individuals RMR is the rate of energy required to sustain life. RMR is measured in controlled conditions. The subject must be in a relaxed (preferably just
having awakened), postabsorptive state (12 hours or more of fasting) [15]. Thus,
the direct determination of RMR is not simple. There exists several simple statistical formulas that estimate RMR depending on sex, total body mass, height, and
age. These formulas are based on extensive experimental data which was analyzed
using predictive regression equations. Although these estimates work well in most
situations, they tend not to be as reliable in extreme cases such as obesity [49].
Popular statistical formulas assume that RMR is a linear function of body mass,
age, height, and gender [21, 42]. There is evidence that RMR is a nonlinear function
of body mass of the form W p where 0 < p < 1 [30, 60]. The Livingston-Kohlstadt
RMR equations developed using the NAS/IOM database take the form of body
mass to a power and provides accurate estimates of RMR when validated against
other databases [38]. The Livingston-Kohlstadt formula is a function of a power of
body mass and baseline age:
RM RLK = (ai W pi yi (A0 + t/365)) .
(5)
where aM = 293, aF = 248, pM = 0.4330, pF = 0.4356, yM = 5.92, yF = 5.09, W

represents body mass, and A0 represents baseline age.
We point out that the units of the coefficients are non-standard as the exponent
of body mass is statistically determined. While analysis of data sets indicate that
the power of body mass is fractional, the exact fraction is still debated and the
April 19, 2010
6:0

8
NHANES
physiological reasoning behind the fractional power is not fully understood at this
time. There are several mathematical explanations, for example, the fractal based
derivation [57], however, the science behind human RMR models is still in its
infancy and the Livingston-Kohlstadt model provides the best known data driven
fractional power formula to date.
It has been observed that calorie restriction results in RMR that is lower than
expected [25, 39]. The expected basal metabolic rate based on body mass in our
model is given by Equation (5). If a is the percent of metabolic adaptation where
0 a 1, then we model the contribution of RMR with adaptation to energy
expenditure by adjusting RMR:
RM R = (1 a) (ai W pi yi (A0 + t/365)) .
Spontaneous physical activity (SPA) is defined as the total energy expended in
activities of daily living, change of posture and fidgeting [61]. During weight gain
due to overfeeding, it was observed that the ratio of the change in SPA to change
in total energy expenditures remained relatively constant [37]:
SP A
= s.
E
where represents final time quantity minus baseline quantity.
SPA is difficult to measure even in highly controlled settings, however, by expanding E using (4) and applying the linearity of the operation we arrive at:
SP A = s(DIT + P A + RM R + SP A)
s
SP A =
(DIT + P A + RM R)
1s
Integration over time allows us to solve for SP A in terms of the remaining energy
expenditure:
SP A =
s
(DIT + P A + RM R) + C
1s
where C is a constant of integration. The constant C is determined by solving

SP A(0) =
s
(DIT (0) + P A(0) + RM R(0)) + C
1s
By examining data measured in [34, 35], it was found that baseline SPA is approximately 32.6% of baseline energy expenditure determined by DLW; SP A(0) =
0.326E(0). As a result, we obtain a closed form expression for SP A in terms of the
other components of energy expenditure. The closed form expression is an increasing function of weight supporting observations that SPA is found to change during
both over-feeding experiments [33, 37] and caloric restriction in obese individuals
[10, 46].
There are only a few overfeeding studies that provide individual DLW, RMR,
and DIT data which presents challenges in locking down the precise value of s.
To estimate s, we analyzed several well known overfeeding studies with published
individual data. Within these published studies, we carefully detailed which data
sets were applicable. Individual subject data was either published or provided by
the author in the overfeeding studies of Bandini [3], Diaz [9], Levine [37], Pasquet
[43], and Siervo [52]. The Pasquet study examined the effects of overfeeding during
April 19, 2010
6:0
NHANES
Guru-Walla, a traditional Cameroon fattening season. The reported baseline energy expenditures of the Pasquet study were higher than expected when compared
to the NAS/IOM regression formulas for E. This was due to increased physical
activity of the subjects during the collection of baseline information. Subjects in
the study were performing physically demanding agricultural work which raised
their baseline energy expenditures and therefore moved them out of energy balance. Thus, calculations of the change in E over time (E) were negative or close
to zero for most of the subjects and could not be applied to determine s. The
Siervo study was unusual as they overfed participants for a period of a few weeks
and then allowed a relaxation of the requirements for a few weeks [52]. This process was repeated for three cycles with the amount overfed and the length of the
periods increasing. Although baseline and final time energy expenditures were reported, computation of E was confounded by the non-monotonic feeding pattern.
Therefore, we removed the Siervo study from consideration.
Computing the value of s using individual data from Bandini, Levine, and Diaz
studies yielded a value of s = 0.4, 0.67, and s = 0.6 respectively. The Bandini
study overfed 13 adolescents for a period of 2 weeks, the Levine study overfed 16
subjects by 1000 kcal/d over baseline requirements for a period of 8 weeks, and the
Diaz study overfed 9 male participants 50% over baseline for a period of 42 days.
Subjects in the Diaz study conducted both cycling and stepping activity during
baseline and overfeeding. Similar to the Pasquet study, baseline energy expenditures for most subjects was higher than expected when compared to the NAS/IOM
predictions. The Diaz study, however, continued the stepping and cycling physical
activity during the overfeeding study and thus, resulting in positive E for most
subjects. Explanations for the negative E measurements were due to external
circumstances and were not considered in studys group mean results [9]. Similarly, we restrict our calculation of s to the positive E measurements. Based on
the available current data, we estimate s as the average of the 0.4, 0.6, and 0.67,
yielding s = 0.56.
Computations of s for the case of caloric restriction indicates that s = 0.67
[22, 44, 48, 50].
We now summarize the formulation for E:
E = RM R + P A + SP A + DIT
(1) RMR is modeled by the Livingston-Kohlstadt equation for RM R. Resting metabolic rate travels along the trajectory for weight determined by
the differential equation. Age is continuously increased from initial age, A0 .
RM R = ai (W (t))p yi (A0 + t/365), where aF , aM , yF , yM are the Livingston Kohlstadt parameters subscripted for gender.
(2) DIT is modeled as a direct proportion of time varying intake (I). If intake
is being increased then there is an increase in the proportion of DIT which
is reported from 14-19%. So we have DIT = I where is in the range of
0.075 0.086 (determined based on overfeeding or caloric restriction).
Baseline DIT is estimated as 0.075I.
(3) PA at baseline is determined by applying baseline SP A observed to be
around 32.6% of baseline E [34, 35, 37]. Baseline P A is estimated then
from T EE(0) DIT (0) SP A(0) RM R(0). If this expression is less
than zero, then there is no contribution of physical activity so P A(0) = 0.
Otherwise P A(0) = T EE(0) DIT (0) SP A(0) RM R(0). We model
the dynamic portion of P A by P A = mW where m = P A(0)/W (0) and
W is time varying weight.
April 19, 2010
6:0

10
NHANES

Notation
Energy Balance Terms
R
I
E, EF , EM
DIT
PA
RMR
SPA
State Variable
Formulas dependent
on state variable
F (t)
F F M (t)
W
G
Parameters
s
cl
cf
m
aF , aM
pF , pM
yF , yM
A0
A
H
a
D, S
F
Acronyms
DLW
DXA
NAS/IOM
NHANES
Table 1.
Description
Estimation
Units
Rate of energy stored/lost

Rate of energy intake
obtained from DXA and DLW
measurements
Rate of energy expended
subscripts represent female and male
Dietary induced thermogenesis
Volitional activity
Resting energy requirements
Activity expended in
daily living
Fat mass on day t
N/A
time dependent step function
kcal/day
kcals/day
Measured by
DLW
N/A
N/A
Livingston-Kohlstadt
N/A
kcals/day
kcals/day
kcals/day
kcals/day
kcals/day
kcals/day
N/A
kg
Fat free mass on day t

Total body mass
Time averaged
kcals of glucose
Ratio of SP A to
E
Energy density of
fat free mass
Energy density of
fat mass
DIT = I
Proportion of body mass
related to PA
Proportionality constant
Mass Exponent
in Livingston-Kohlstadt
Age constant
from Livingston-Kohlstadt
Initial Age
Age
Height
Percent of metabolic
adaptation
Translation parameter in the
Forbes model
Value of fat mass
at F F M = 0
Doubly labeled
water
Dual X-ray
absorptiometry
National Academy of Science
Institute of Medicine
National Health and Nutrition
Examination Survey
F F M = 1 F + b
W=FFM(t)+F(t)
N/A
kg
kg
kcal
N/A
kcals/kg
0.075 0.086
kcals/kg
N/A
N/A
aF = 248
aM = 293
pF = 0.4356
pM = 0.4330
yF = 5.09
yM = 5.92
A0 > 0
A = A0 + t/365
kcals/kg/d
N/A
0a1
N/A
N/A
N/A
years
days
cm
N/A
N/A
kg
N/A
N/A
N/A
Legend of Notation
(4) SPA is modeled by integrating

SP A = sE
where s = 0.56 and solving for the constant of integration by applying
baseline data.
Combining all expressions obtained for R, I and E and substituting into Equation
(1) we obtain the final one dimensional model,
cf
dF F M
dF
+ cl
= I (RM R + DIT + SP A + P A)
dt
dt
(6)
with initial condition

F (0) = F0 > 0
We remark that the state variable of Equation 6 is F (t) since F F M (t) is slaved to
F (t) through the NHANES regression formulas. The derivative of F F M (t) on the
left hand side of Equation 6 is taken implicitly.
April 19, 2010
6:0
NHANES
3.
11
Mathematical analysis
In this section we address the existence, uniqueness and non-negativity of solutions

to (6). The model (6) when written in the form
dF
= G(F (t))
dt
has a locally Lipschitz right hand side and therefore is guaranteed a unique solution
on an interval of existence, [0, ) where > 0 [16].
In our analysis, we consider I constant. Individual changes in I during weight
loss or weight gain are behaviorial and would move to a different trajectory based
on the change to I and hence we consider existence and uniqueness of a single
trajectory by assuming a constant I
The model defined by Equation 6 is non-autonomous due to the time varying
age parameter A = A0 + t/365. Our goal is to first prove existence, uniqueness and
non-negative for a related autonomous equation and use differential inequalities to
extrapolate the result to our non-autonomous equation.
If we replace the non-autonomous expression A0 +t/365 by A0 in our expenditure
term and A = A0 + /365 in the R term, the modified equation is autonomous.
Applying differential inequalities, we can quickly see that the autonomous equation
(AE) is a lower bound for (6). Thus, our strategy is to prove non-negativity for
the (AE), which guarantees non-negativity for (6) on [0, ). Finally, to guarantee
existence for all time, we prove using contradiction that solutions to (6) do not
have finite time blow up.
To prove non-negativity of solutions for the autonomous equation we need to
prove that dF/dt 0 when F = 0. This condition is called the quasi-positivity
criteria [40]. If an autonomous differential equation satisfies the quasi-positivity
criteria, then the solutions to the equation are non-negative by a result in [40].
In order, to determine what restriction on the parameters will guarantee quasipositivity, we substitute F = 0 into (AE) and calculate the left hand side of AE:

= cl
=
d
3.6 F (t) 0.002
dt

d
d
d
F (t) A 0.01
F (t) H + cf F (t)
dt
dt
dt
dF
H, cl , cf ))
((A,
dt
Similarly, setting F = 0 in the right hand side of AE yields a function of parameters

(I, A0 , H, , ai , yi , C, m)
By solving the inequality dF/dt 0, we have that quasi-positivity is guaranteed
as long as
(I, A0 , H, , ai , yi , C, m)
0
H, cl , cf )
(A,
(7)
The biological meaning of this condition restricts energy intake to be larger than
energy expenditures at zero fat. If we define critical intake to be the intake equal to
expenditures at zero fat, the restriction determined by Inequality 7 relates critical
intake to minimal energy expenditures from F as a function of age and height
[54]. Direct computation of this condition is not elegant but simple using computer
April 19, 2010
6:0

12
NHANES
algebra system software. Numerical experimentation with the intake parameter

shows that if intake is too low (on the order of below 1000 kcal/d), the restriction
on the parameters is violated. This is consistent with the observation from the
Minnesota Starvation Experiment that continued reduced intake of 1800 kcal/d for
a 24 week resulted in very low BMI surmised to be the near fatal in the Minnesota
Starvation Experiment [13]. A separate model describing the dynamics of starvation
was previously developed [53] and can be applied to understand lower and zero
levels of intake.
To guarantee existence for all time, we only need to show that finite time blow
up cannot occur. This is because we have already proved that on the interval
of existence the solution is non-negative. If there is a loss of existence, then the
solution must blow up. We will assume on the contrary that finite time blow up
occurs and arrive at a contradiction.
Let us assume that the restriction on parameters (7) holds. Assume on the contrary that there is a b0 such that the limtb0 F (t) = . Then for some T > 0,
there is a subsequence tk with tk b0 and F (tk ) increasing to infinity.
If F increases to infinity, then W also increases to infinity. Moreover, since E is
an increasing function of W , E also goes to infinity. But then substitution into the
right hand side of (6) yields
I E(W (tk ))
which implies F 0 (tk ) and creates a contradiction. Therefore under condition (7) a nonnegative solution to the initial value problem exists for all time
t 0.
We summarize these observations in the theorem below:
Theorem 3.1 : Suppose that the parameters in (6) satisfy the restriction (7).
Then a nonnegative unique solution for the equation (6) exists for all t 0.
The theorem restricts the rate of energy intake to be greater than energy expenditures at zero fat mass, providing lower bounds for energy intake that avoid
starvation.
3.1.
Long-term dynamics
In the case where age is constant, and I is constant, solutions to (AE) converge
monotonically to the steady state where intake equals expenditure, I = E [55].
Increases in age act as incremental decreases in energy expenditure which do not
allow solutions to completely rest, but force a small pulling away from the steady
state of (AE). To see this, we simulated the model with age held constant at A0
alongside the non-autonomous model. One can see the phenomena described in
Figure 1.
4.
Validation and Comparison of the Model
In this section we validate the model using individual data from three studies; two
caloric restriction studies and one overfeeding study. Our choice of weight change
study to validate the model were defined by three criteria:
(1) Baseline data was carefully measured to simulate zero energy balance as
closely as possible. No lifestyle changes in physical activity or diet were
prescribed during the baseline measurements.
April 19, 2010
6:0
NHANES
13
Figure 1. A simulation of individual weight loss for a period of five years with age held constant (dashed
curve) and increasing age (solid curve). The y-axis represents weight in kg and the x-axis time in days.
Baseline weight, height, and age are 77 kg, 172 cm, and 44 years. The target caloric intake is 2200 kcal/d.
(2) Changes in stored energy were measured during weight change through
body composition measurements. These measurements are used to estimate
the value of R.
(3) Energy expenditures were determined during weight change through DLW
measurements. These measurements are used to estimate E.
Because it is well known that subject adherence to energy intake are not guaranteed, we can use the estimate of R and E to approximate energy intake: I = R +E.
Comparisons of actual I and estimated I through DXA/DLW has been successfully
demonstrated [8, 28]
We also compare the model results to the one dimensional model (Equation 25
[4]):
cl
dF F M
dF
+ cf
=I E
dt
dt
applying the energy expenditure regression model from [4]:

E = 238.85(0.14F F M + 0.05F + 1.55),
F F M is modeled by the Forbes model with D and S set as the average value
from [12] and [54], and the derivative of F F M is determined implicitly on the left
hand side of the equation. There are several developed one dimensional models
that describe weight change [2, 46, 31, 32]. A comparison of a differential equation that uses the Forbes model for F F M to other one dimensional models was
conducted previously [20] and shown to outperform in predictive power to other
one dimensinoal models in the steady state case. In our own comparisons, the one
dimensional equation (Equation 25 [4]) provided more accurate predictions in the
dynamic state than the model in [20], therefore we focus our comparison solely to
Equation 25 from [4].
April 19, 2010
6:0

14
NHANES
Parameters and baseline values that are dependent on individual data are outlined in a table under each simulation. Table 2 provides the values for all universal
parameters used in every simulation.
Table 2. Constants that are universal to all numerical simulations of the differential equation model.
parameter a is zero for overfeeding simulations.
Constant
s
cl , cf
aM , aF , pM , pF , yM , yF
4.1.
Value
underfeeding: s = 0.67
overfeeding: s = 0.56
cl = 1100, cf = 9500,
aM = 293, aF = 248, aM = 0.4330
pF = 0.4356, yM = 5.92, yF = 5.09
= 0.075
= 1.19: overfeeding, = 1: caloric restriction
a = 0.02
The
Reference
[22, 44, 48, 50]
[3, 9, 37]
[26, 51]
[38]
[58]
[20, 37]
[39]
Caloric Restriction - CALERIE PHASE I study
The Phase I of the Comprehensive Assessment of Long-term Effects of Reducing Intake of Energy (CALERIE) trial tested the effects of calorie restriction on
biomarkers of longevity [23]. Twelve of the CALERIE subjects were placed on a
very low calorie diet of 890 kcal/d (LCD), twelve were placed on a low calorie diet
of 25% below baseline energy requirements (CR), and twelve were prescribed a
combination of caloric restriction (12.5% below baseline energy requirements) and
exercise (physical activity increased to 12.5% above baseline total energy expenditure). Thirty of the subjects were overweight and six of the subjects had a baseline
BMI classifying them as obese. The 24 subjects that were reducing their weight
solely through caloric restriction (CR and LCD) were used to test the validity of
our model.
The metabolic adaptation parameter a was determined by comparing the result
of regression formulas for the resting metabolic rate to actual resting metabolic
rate in the CALERIE Phase I study [39]. The value of a was not calibrated using the dynamical systems model presented here, nor was it computed using the
Livingston-Kohlstadt model and is treated as a universal parameter.
Subject energy intake was determined using the energy balance equation. The
value of R was estimated by,
R cl
F
F F M
+ cf
t
t
where body composition changes were measured by DXA. E was measured using
DLW and then I was estimated by computing R+E. DXA and DLW measurements
were collected for all subjects at 4, 12, and 24 weeks. We restrict our attention to
the 24 CR and LCD subjects and remark that the 12 combined caloric restriction
and exercise subjects would require an extra measurement determining adherence
to the exercise prescription.
In order to estimate actual intake as close as possible to known data, we set
intake as a piecewise defined step function
: 0 t 28
(1 c4 )I0
I = (1 c12 )I0 : 28 < t 84
(1 c24 )I0 : 84 < t 168
April 19, 2010
6:0
NHANES
15
Table 3. Mean data and mean absolute error for observed final weights versus model predictions of final weights
measured in kg.
Study
CALERIE [23]
Racette [44]
Final
Body Mass (kg)
72.6 10.6
85.0 9.6
Overfeeding [34, 35]
79.6 20.4
Heymsfield Heymsfield Model 1-D Chow-Hall1-D Chow-Hall Model

Model (kg) Mean Abs. Error (kg) Model (kg) Mean Abs. Error (kg)
73.3 10.9
1.8 1.3
72.6 13.3
4.5 3.3
85.1 10.4
1.7 1.2
85.6 9.3
3.0 3.0
80.7 20.5
2.5 1.6
92.1 19.2
5.5 2.1
where I0 is initial intake and c4 , c12 , c24 are the individual subjects measured percent decrease from baseline intake at 4, 12, and 24 weeks respectively. The piecewise
definition of I will result in a continuous solution but not one that is differentiable
everywhere. This is because we are guaranteed a unique nonnegative solution for
each branch of I by our existence theorem. The continuation along the same trajectory (feeding the end values as initial data for the next leg) guarantees continuity,
however, differentiability does not hold since all 19 subjects presented here varied
their intake, thereby changing the slope.
For the sake of comparison, we refer to the model presented here as the Heymsfield model. At 24 weeks of calorie restriction, the Heymsfield model predicts the
mean 24 week weight as 73.9 kg versus the actual mean observed weight of 72.6
kg (Table 1). Simulations of the one dimensional model in [4] yield a mean weight
of 72.6 kg. Error between each subjects individual final predicted and observed
weight at 24 weeks was computed by
i = |Wactual,i Westimate,i |
where the subscript i denotes subject number, and Wactual,i represents the measured weight, and Westimatei represents the model predicted weight for subject i.
We refer to the mean absolute error as the mean of the set of {i }19
i=1 .
The mean absolute error in predicting individual final weight for the Heymsfield
model was 1.8 1.3 kg whereas the one dimensional model (Equation 25 [4]) was
4.5 3.3 kg. The Heymsfield model predicted final body mass consistently with
a maximum absolute error of 4.3 kg. Model reliability is also substantiated by
the low standard deviation in the mean absolute error for the Heymsfield model.
In terms of individual predictions, the one dimensional model (Equation 25 [4])
yielded a maximum absolute error of 12 kg and predictions for 9 subjects produced
absolute error over the maximum absolute error held by the Heymsfield model.
Individual subject final measured 24 week weight and model predictions for both
the Heymsfield model and the one dimensional model (Equation 25 [4]) appear in
Figure 2.
Subjects were provided with all food using an infeeding paradigm during the
first 12 weeks of the study which appeared to affect the level of adherence to the
prescribed restrictions in intake. Direct inspection of individual weight predictions
against measured data at 4 weeks, 6 weeks, 12 weeks, and 24 weeks, revealed
although the weight loss trend was modeled well by the adherence measurements,
the initial drop of weight appeared higher than predicted by the model. Moreover,
subjects appeared to adhere at the 24 week measured adherence percentages for a
few weeks after infeeding was discontinued.
We surmised that subjects may have been highly compliant during the inception
of the study with compliance decreasing over time. To test this assumption, we
modeled intake by a piecewise defined step function applying the assumption that
subjects were 100% compliant for the first 28 days (represented by the percentage
c0 ), followed by a decrease in intake to the observed adherence numbers c4 for
28 t 60. We then assumed that participants attempted to match their infeeding
behavior at home for a few weeks after infeeding was discontinued. This is reflected
by a decrease in intake to by the measured percentage c12 for 60 < t 100. Finally,
April 19, 2010
6:0

16
NHANES
Figure 2. The 24-week body mass in kg (y-axis) is plotted for each individual calorie restricted subject
(x-axis) for actual CALERIE PHASE I measurements (square), Heymsfield model predictions (solid circle)
and one dimensional Chow-Hall model predictions (solid triangle). [23].
Table 4.
Baseline information used to simulate model compared to data in [23].
Baseline/Parameter
RMR
A0
H
W (0)
F (0)
F (0)
F F M (0)
c2 , c3
Method Obtained
Provided from study data
Heymsfield Model: Estimated using NHANES F F M formula [54]
1-D Chow-Hall model: Estimated using Forbes model
Obtained from W(0)-F(0)
Obtained from study data
after 100 days, intake decreased to the adherence percentages measured at 168 days:
(1 c0 )I0
: 0 t 28
(1 c4 )I0
: 28 < t 60
I=
(1 c12 )I0 : 60 < t 100
(1 c24 )I0 : 100 < t 168

Applying the intake step function, we discover a remarkable fit to the individual
observed measurements. The mean absolute error in prediction for the Heymsfield
model reduces to 1.5 1.6 kg. The one dimensional Chow-Hall model predictions
remain relatively similar to the previous case with a mean absolute error of 4.62.7
kg.
A sample simulation for a single subject describing the impact of each definition
of intake (full compliance, adherence data applied, full compliance is temporarily
attained followed by the measured adherence values) is depicted in Figure 3.
4.2.
The Racette study
To compare the effects of exercise and composition of reduced energy intake, twenty
three obese women were assigned to four groups; control, exercise, low carbohy-
April 19, 2010
6:0
NHANES
17
Figure 3. Each graph is a plot of body mass in kg during weight change (y-axis) versus time in days
(x-axis). The solid curves are simulations of the Heymsfield model for an individual subject in the CALERIE Phase I study assuming energy intake was 100% compliant (a),estimating energy intake from
DLW/DXA measurements (b), and assuming 100% compliance for 4 weeks followed by intake estimated
from DLW/DXA measurements (c). Circles represent actual measurements of body mass.
drate and low fat. A 12 week weight reduction phase followed a 5-week baseline
measurement period. A third phase, where subjects maintained their weight reductions was also conducted, however, for the purposes of model validation we
are solely interested in the 12-week weight reduction phase. Total daily energy expenditure was measured at baseline and twelve weeks using doubly labeled water
and body composition was determined at baseline and at the end of the weight
reduction period using under water weighing. From these two measurements, we
estimate actual intake as R + E where R is determined by
R = cl F F M/84 + cf F/84
and E is given by DLW measurement at 12 weeks. We restricted our analysis to
the 13 subjects who were assigned to the high fat or low carbohydrate diets.
Parameters for model simulations are provided in Table 5 and individual subject
results for the observed, Heymsfield model, and one dimensional Chow-Hall model
(Equation 25 [4]) appear for each of the 13 subjects in Figure 4. Similar to the CALERIE study, the one dimensional Chow-Hall model predicted mean group results
accurately with a mean final weight of 85.3 9.3 kg, however the maximum mean
absolute error was 11.1 kg as opposed to the Heymsfield model which yielded maximum absolute error of 4.3 kg. Mean absolute error in prediction for the Heymsfield
model was 1.7 1.2 kg and mean absolute error for the one dimensional Chow-Hall
model was 2.9 3.0. We again note the low standard deviation in mean absolute
error for the Heymsfield model indicating low variance in the error for individual
subject predictions.
4.3.
Overfeeding
For model validation purposes in the case of overfeeding, we apply overfeeding data
from 22 subjects who increased their caloric intake to 1000 kcal/d over baseline requirements for a period of 8 weeks[34, 35]. Similar to the caloric restriction studies,
subject compliance was determined using the energy balance equation; namely R
was measured by DXA, E was measured through DLW and then I was estimated
April 19, 2010
6:0

18
NHANES
Figure 4. The 12-week body mass in kg (y-axis) is plotted for each individual calorie restricted subject
(x-axis) for actual Racette study measurements (square), Heymsfield model predictions (solid circle) and
one dimensional Chow-Hall model predictions (solid triangle) [44].
Table 5.
Baseline information used to simulate model compared to data in [44].
Baseline/Parameter
RMR
A0
H
W (0)
F (0)
F (0)
F F M (0)
I
Table 6.
Method Obtained
!-D Chow-Hall model: Estimated using Forbes model
Baseline information used to simulate model compared to data in [34, 35].
Baseline/Parameter
RMR
A0
H
W (0)
F (0)
F (0)
F F M (0)
I
Method Obtained
1-D Chow-Hall Model: Estimated using Forbes model
Used IOM/NAS regression formula (4)
by computing R + E. From these measurements, it was determined that subjects

were on average 98% compliant to the 1000 kcal/d overfeeding prescription. Input
parameters and baseline information appear in Table 6. Results of the simulations
show that the Heymsfield model predicts with a mean absolute error of 2.5 1.6 kg
the actual weight at 8 weeks and the one dimensional Chow-Hall model predicts
with a mean absolute error of 5.5 2.1 kg.
In summary, the Heymsfield model predictions of weight during weight change
had very good agreement with actual measured weights Figure 6. Figure 6 depicts
the plot of individual actual final weight from all three studies considered here
April 19, 2010
6:0
NHANES
19
Figure 5. The 8-week body mass in kg (y-axis) is plotted for each individual overfed subject (x-axis) for
actual measurements (square), Heymsfield model predictions (solid circle) and one dimensional Chow-Hall
model predictions (solid triangle) [34, 35]
Figure 6. Final measured body mass in kg (y-axis) plotted against the Heymsfield model predicted body
mass in kg (x-axis) for subjects in all three studies.
(y-axis) against model predictions.
April 19, 2010
6:0

20
5.
NHANES
Conclusion
Our previously developed model [55] was modified by coupling the differential
equation to the F F M functions of fat mass developed in [54]. The revision results
in improved predictions of final individual weight for three weight change studies.
Comparison of the model to other one dimensional models demonstrates that the
model provides improved predictions of individual final weight for both caloric
restriction and overfeeding studies. The model also yields an increase in reliability
measured through very low standard deviation of the mean absolute error and a
reduction by over 60% in the maximum absolute error.
As part of our efforts to validate the model using data from weight change studies
that also collected measurements of subject compliance, it is confirmed that subject
dietary adherence to study prescriptions are not guaranteed. Additional challenges
presented by our analysis of overfeeding studies indicates a need for more tightly
controlled overfeeding studies where baseline data is obtained as close to energy
balance as possible, overfeeding is a consistent and constant, and compliance to
the prescription is determined.
The model presented here can be applied in a clinical setting by simulating the
expected rate of weight change (and final weight) of participants engaging in dieting
or overfeeding and comparing to observed weight change. Differences in model
predictions and observed weight change identify the need for clinical intervention
to foster adherence.
To this date, validation of models of weight change focused on model accuracy for
predicting group mean data. In order to apply models in a clinical setting, reliable
and valid individual-level predictions are required. Importantly, the model reported
herein provides accurate estimates for both group-level and individual-level data,
demonstrating the ability to use the model to accurately predict individual patients
weight loss and objectively measure adherence to calorie prescriptions.
Acknowledgements
This work was supported by the following grants:The CALERIE study was supported by NIH U01 AG20478 (PI: Eric Ravussin, Ph.D.). Corby K. Martin is supported by NIH K23 DK068052 (PI: Corby Martin, Ph.D.). Leanne M. Redman,
Ph.D. is supported by a Training Fellowship awarded by the NHMRC of Australia
(ID 349553). Dale Schoeller is supported by NIH RO1 DK30031. James A. Levine
is supported by NIH grants DK56650, DK63226, DK 662760 and M01 RR00585,
and the Mayo Foundation.
References
[1] S. Alhassan, S. Kim, A. Bersamin,A.C. King, C.D. Gardner, Dietary adherence and weight loss success
among overweight women: results from the A TO Z weight loss study Int. J. Obes. (Lond). 32 (2008),
pp. 985-91.
[2] S.S. Alpert, The cross-sectional and longitudinal dependence of the resting metabolic rate on the
fat-free mass Metabolism. 56 (2007), pp. 363-72.
[3] L.G. Bandini, D.A. Schoeller, J. Edwards, V.R, Young, S.H. Oh, W.H. Dietz Energy expenditure
during carbohydrate overfeeding in obese and nonobese adoles-cents Am. J. Physiol. 256 (1989), pp.
E357-E367.
[4] C. C. Chow, K.D. Hall The dynamics of human weight change PLoS Comp. Biolg. 4 (2008), pp. 1-11.
[5] E. Christiansen, L. Garby Prediction of body weight changes caused by changes in energybalance.
Eur. J. Clin. Invest. 32 (2002), pp. 826-30.
[6] E. Christiansen, L.Garby, T. I.A. Srensen, Quantitative analysis of the energy requirements for development of obesity J. Theor. Biol. 234 (2005), pp. 99106.
[7] M.L. Dansinger, J.A. Gleason, J.L. Griffith, H.P. Selker, E.J. Schaefer, Comparison of the Atkins, Or-
April 19, 2010
6:0
NHANES
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
21
nish, Weight Watchers, and Zone diets for weight loss and heart disease risk reduction: a randomized
trial. JAMA. 293 (2005), pp. 43-53.
P. Del Corral, P. C. Chandler-Laney, K. Casazza, B. A. Gower, and G. R. Hunter Effect of Dietary
Adherence with or without Exercise on Weight Loss: A Mechanistic Approach to a Global Problem
J. Clin. Endocrinol. Metab. 94(2009),pp. 1602 - 1607.
E.O. Diaz, A.M. Prentice, G.R. Goldberg, P.R. Murgatroyd, W.A. Coward, Metabolic response to
experimental overfeeding in lean and overweight healthy volunteers Am. J. Clin. Nutr., 56 (1993), pp.
641-655.
E.Doucet, S. St-Pierre,N. Almeras, J. Despres, C. Bouchard,A. Tremblay, Evidence for the existence
of adaptive thermogenesis during weight loss Br. J. Nutr. 85 (2001), pp. 715-723.
H. Folger, Elements of Chemical Reaction Engineering, 4th Edition, Prentice Hall, (1986).
G.B. Forbes, Lean body mass-body fat interrelationships in humans Nutr. Rev. 45 (1987), pp. 225-231.
K.E. Friedl, R.J. Moore, L.E. Martinez-Lopez, J.A. Vogel, E.W. Askew, L.J. Marchitelli, R.W. Hoyt,
C.C. Gordon, Lower limit of body fat in healthy active men. J. Appl. Physiol. 77 (1994), pp. 933-40.
G.F. Gilbert, Basic Concepts in Biochemistry, McGraw-Hill Health Professions Division, (2000).
S. Gropper,J. Smith, J. Groff, Advanced Nutrition and Human Metabolism, Thomson Wadsworth
Pub., (2005).
J. Hale, Ordinary Differential Equations, 2nd Ed., Krieger, (1980).
K.D. Hall Body fat and fat-free mass interrelationships: Forbes Theory revisited Br. J. Nutr. 97
(2007), pp. 1059-1063.
K.D. Hall Computational model of in vivo human energy metabolism during semi starvation and
refeeding Am. J. Physiol. Endroc. Metabl. 291 (2006), pp. E23-27.
K.D. Hall, Predicting metabolic adaptation, body weight change, and energy intake in humans Am.
J. Physiol. Endocrinol. Metab. (2010) E449 - E466.
K. Hall, P. Jordan, Modeling Weight Loss Maintenance To Help Prevent Body Weight Regain Amer.
J. Clin. 88 (2008), pp. 1495-503.
J. Harris, F. Benedict, A Biometric Study of Human Basal Metabolism Proc. Sci. US. 4 (1918), pp.
370373.
M. Heyman, V.R. Young, P. Fuss,R. Tsay, L. Joseph, S.B.Roberts, Underfeeding and body weight
regulation in normal weight young men, Am. J. Physiol. Regul. Integr. Comp. Physiol. 263: (1992),pp.
R250 R257.
L. K. Heilbronn, L. Jonge, M. I. Frisard, J. P. DeLany, D. E. Larson-Meyer, J.Rood, T. Nguyen, C. K.
Martin, J.Volaufova, M. M. Most, F.L. Greenway, S. R. Smith, W. A. Deutsch, D. A. Williamson, E.
Ravussin, Effect of 6-Month Calorie Restriction on Biomarkers of Longevity, Metabolic Adaptation,
and Oxidative Stress in Overweight Individuals A Randomized Controlled Trial JAMA 295 (2006),
pp. 1539-1548.
M. Hein,L.R. Best,S. Pattison, S. Arena, Introduction to General, Organic, and Biochemistry Sixth
Edition, Brooks/Cole, (1997).
S.B. Heymsfield, J.B. Harp, M.L. Reitman, Why do obese patients not lose more weight when treated
with low calorie diets? A mechanistic perspective, Am. J. Clin. Nutr. 85 (2007), pp. 346-54.
S.B. Heymsfield, M. Waki, J. Kehayias, S. Lichtman, F.A. Dilmanian, Y. Kamen, J.Wang, R.N.
Pierson RN Jr., Chemical and elemental analysis of humans in vivo using improved body composition
models. Am. J Physiol., 261(1991), pp. 191-198.
E. Jequier, L. Tappy, L., Regulation of Body Weight in Humans, Physiol. Rev. 79 (1999), pp. 451480.
L. de Jonge, J.P. DeLany, T. Nguyen,J. Howard,E. Hadley,L. Redman,E. Ravussin, Validation study
of energy expenditure and intake during calorie restriction using doubly labeled water and changes
in body composition. Am. J. Clin. Nutr. 85(2007), pp. 7379.
T.L. Kelly, K.E. Wilson, S.B. Heymsfield SB, Dual Energy X-Ray Absorptiometry Body Composition
Reference Values from NHANES. PLoS ONE 4 (2009), pp e7038.
M. Kleiber, The Fire of Life New York: John Wiley & Sons Inc., (1961).
F. P. Kozusko, Body weight, setpoint, metaboic adaptation and human starvation, Bull. Math. Biol.
63 (2001),pp. 393-403.
F. Kozusko, The effects of body composition on set point based weight loss Math and Comp. Modelling
29 (2002), pp. 1-7.
R. L. Leibel, M. Rosenbaum, M.,J. Hirsch, Changes in energy expenditure resulting from altered body
weight in man NEJM 332 (1995), pp. 621-627.
Levine JA, McCrady SK, Lanningham-Foster LM, Kane PH, Foster RC, Manohar CU The role of
free-living daily walking in human weight gain and obesity Diabetes 57 (2008), pp. 548-554.
J.A. Levine, L.M. Lanningham-Foster, S.K. McCrady, A.C. Krizan, L.R. Olson, P.H. Kane, M.D.
Jensen, M.M. Clark, Inter-individual variation in Posture Allocation: Possible role in human obesity.
Science 28 (2005), pp. 584-6.
Nonexercise activity thermogenesis (NEAT): environment and biology Am. J. Physiol. Endocrinol.
Metab. 286 (2004), pp. E675E685.
J.A. Levine, N.L. Eberhardt, M.D. Jensen, Role of nonexercise activity thermogenesis in resistance
to fat gain in humans. Science, 283 (1999),pp. R212214.
E. Livingston, I. Kohlstadt Simplified resting metabolic rate-predicting formulas for normal-sized and
obese individuals Obes.Res. 13 (2005), pp. 1255-1262.
C.K. Martin, L.K. Heilbronn, L. de Jonge, J.P. DelLany, J. Volaufova, S.D. Anton, L.M. Redman,
S. R. Smith, E. Ravussin, Effect of Calorie Restriction on Resting Metabolic Rate and Spontaneous
Physical Activity Obesity 15 (2007),pp. 2965-2973.
R.H. Martin, Nonlinear operators and differential equations in Banach spaces., Robert E. Krieger.,
Melbourne, FL., (1987).
W.D. McArdle, F.I. Katch, V.L. Katch, Exercise Physiology: Energy, Nutrition, and Human Performance, 5th Edition, Lippincott, Williams and Wilkins, (2001).
M.D. Mifflin, S.T. St. Jeor, L.A. Hill, B.J. Scott, S.A. Daugherty, Y.O. Koh, A new predictive equation
April 19, 2010
6:0

22
NHANES
for resting energy expenditure in healthy individuals. Amer. J. Clin. Nutr. 51 (1990), pp. 241-247.
[43] P. Pasquet,L. Brigant,A. Froment,G.A. Koppert,D. Bard, T.de Garine, M. Apfelbaum, Massive overfeeding and energy balance in men: the Guru Walla model, Am. J. Clin. Nutr., 56 (1992), pp. 483490.
34. Prentice AM. Manipulation of dietary fat and energy density
[44] S.B. Racette, D.A. Schoeller, R.F. Kushner, K.M. Neil, and K. Herling-Iaffaldano, Effects of aerobic
exercise and dietary carbohydrate on energy expenditure and body composition during weight reduction
in obese women Am. J. Clin. Nutr. 61 (1995), pp. 486-494.
[45] E. Ravussin, S. Lillioja, T. E. Anderson, L. Christin and C. Bogardus, Determinants of 24-hour energy
expenditure in man. Methods and results using a respiratory chamber. J. Clin. Invest. 78 (1986), pp.
15681578.
[46] L. M. Redman, L.K. Heilbronn, C. K. Martin, A.Alfonso, S. R. Smith, E. Ravussin, Effect of Calorie
Restriction with or without Exercise on Body Composition and Fat Distribution J. Clin. End. Metab.
92 (2007),pp. 865872.
[47] L.M. Redman, L.K. Heilbronn, C.K. Martin, L. de Jonge, D.A. Williamson, J.P. Delany, E. Ravussin,
Pennington CALERIE Team, Metabolic and behavioral compensations in response to caloric restriction: implications for the maintenance of weight loss PLoS One. 4 (2009), e4377 Epub 2009.
[48] S.B. Roberts, I, Rosenberg, Nutrition and aging: changes in the regulation of energy metabolism with
aging Physiol. Rev. 86 (2005), pp. 651667.
[49] T. Roozalia, R. Rising., D. Brown, F. Lifshitz, Comparison of several equations and derivation of a
new equation for calculating basal metabolic rate in obese children, J. Amer. Col. Nutr. 4 (1998), pp.
333-336.
[50] E.Saltzman, S.B. Roberts, The role of energy expenditure in energy regulation: findings from a decade
of research Nutr. Rev. 53 (1995), pp. 209-220.
[51] Y. Schutz, Glossary of energy terms and factors used for calculations of energy metabolism in human studies. Van Es AJH. Human Energy Metabolism: Physical Activity and Energy Expenditure
Measurements in Epidemiological Research Based Upon Direct and Indirect Calorimetry (1984).
[52] M Siervo,G.Fruhbeck,A. Dixon, G.R. Goldberg, W.A. Coward, P.R Murgatroyd, A. M. Prentice,S.A.
Jebb, Efficiency of autoregulatory homeostatic responses to imposed caloric excess in lean men Am.
J. Physiol. Endocrinol. Metab. 294 (2008), pp. E416 - E424.
[53] B. Song, D. M. Thomas, Dynamics of Starvation in Humans, J. of Math. Biol.,54 (2007),pp. 27-43.
[54] D. Thomas, S. Das, J.A. Levine, C.K. Martin, L Mayer, A. McDougall, B.J. Strauss, S.B. Heymsfield
New Fat Free Mass - Fat Mass Model for use in Physiological Energy Balance Equations, Nutr. and
Metab. (to appear) 2010.
[55] D.M. Thomas, A. Ciesla,J.A. Levine,J.G. Stevens, C.K. Martin A mathematical model of weight
change with adaptation 6 (2009), pp. 873 - 887.
[56] S. Toubro, N. J. Christensen and A. Astrup, Reproducibility of 24-h energy expenditure, substrate
utilization and spontaneous physical activity in obesity measured in a respiration chamber. Int. J.
Obes. Relat. Metab. Disord. 19 (1995), pp. 544-549.
[57] G.B. West, J.H. Brown, B.J. Enquist, A general model for the origin of allometric scaling laws in
biology Science 276 (1997), pp. 122-126.
[58] K.R. Westerterp, Diet induced thermogenesis, Nutr. Metab. 1 (2004), pp. 1-5.
[59] C. R. White, R.S. Seymour, Mammalian basal metabolic rate is proportional to body mass 2/3 PNAS
100 (2003), pp. 40464049.
[60] I. Zakeri,M.R. Puyau,A.L. Adolph,F.A. Vohra,N.F. Butte Normalization of Energy Expenditure Data
for Differences in Body Mass or Composition in Children and Adolescents Amer.Soc.Nutr. J. Nutr.
136 (2006), pp. 1371-1376.
[61] Draft
Energy
Requirements
report
scientific
consultation
November
2009
http://www.sacn.gov.uk/reports position statements/index.html
[62] Obesity and Overweight Facts, (2003) WHO report.
[63] Statistics Related to Overweight and Obesity, NIH Publication No. 03-4158.
[64] Body Measurements (Data are for the U.S.,1999-2002), National Center for Health Statistics.
[65] 2004 Vital Health Statitics, Trends in Intake of Energy and Macronutrients United States, 1971
2000.
[66] CDC website on Overweight and Obesity http://www.cdc.gov/obesity/index.html

A Simple Model Predicting Individual Weight Change in Humans

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Simple Model Predicting Individual Weight Change in Humans

Uploaded by

Copyright:

Available Formats

April 19, 2010

Journal of Biological Dynamics

Journal of Biological Dynamics

Department of Mathematical Sciences, Montclair State University, Montclair, NJ;

(v2.2 released November 2008)

Keywords: energy balance equation, metabolic adaptation, non-exercise activity

author. Email: thomasdia@mail.montclair.edu

ISSN: 1751-3758 print/ISSN 1751-3766 online

April 19, 2010

Journal of Biological Dynamics

Taylor & Francis and I.T. Consultant

One approach to modeling human weight change is to incorporate all reasonable

April 19, 2010

Journal of Biological Dynamics

model identifies detailed movement of energy from intake to expenditure providing

The Energy Balance Principle

The energy balance principle discussed in physiology and nutrition literature is

The rate of energy stored/lost: R

April 19, 2010

Journal of Biological Dynamics

Taylor & Francis and I.T. Consultant

April 19, 2010

Journal of Biological Dynamics

tive health survey of the civilian, non-institutionalized United States population,

F F M (t) = 72.1 + 2.5F (t) 0.04(A0 + t/365) + 0.7H

F F M (t) = 71.7 + 3.6F (t) 0.04(A0 + t/365) + 0.7H

The rate of energy intake, I

April 19, 2010

Journal of Biological Dynamics

Taylor & Francis and I.T. Consultant

EM = 0.0971W 2 + 40.853W + 323.59

The rate of energy expenditure, E

April 19, 2010

Journal of Biological Dynamics

is a direct proportion of the rate of energy intake, I, with proportionality constant,

where aM = 293, aF = 248, pM = 0.4330, pF = 0.4356, yM = 5.92, yF = 5.09, W

April 19, 2010

Journal of Biological Dynamics

Taylor & Francis and I.T. Consultant

where C is a constant of integration. The constant C is determined by solving

April 19, 2010

Journal of Biological Dynamics

April 19, 2010

Journal of Biological Dynamics

Taylor & Francis and I.T. Consultant

Energy Balance Terms

Rate of energy stored/lost

Fat free mass on day t

(4) SPA is modeled by integrating

with initial condition

April 19, 2010

Journal of Biological Dynamics

In this section we address the existence, uniqueness and non-negativity of solutions

Similarly, setting F = 0 in the right hand side of AE yields a function of parameters

April 19, 2010

Journal of Biological Dynamics

Taylor & Francis and I.T. Consultant

algebra system software. Numerical experimentation with the intake parameter

Validation and Comparison of the Model

April 19, 2010

Journal of Biological Dynamics

applying the energy expenditure regression model from [4]:

April 19, 2010

Journal of Biological Dynamics