editorial

nature publishing group

Issues and Misconceptions

About Obesity
J.P. Flatt*
Obesity (2011) 19, 676686. doi:10.1038/oby.2011.7

Energy is a concept of universal importance. In

relation to body weight regulation, one is so used
to consider energy intake, energy expenditure,
and energy balance that one forgets that energy is
an abstract notion without physiological equivalent. The bodys regulatory functions may strive to
maintain carbohydrate (CHO), protein and/or fat
balances, but not energy balance. The body ignores
that 1g of fat contains more than twice the energy
present in 1g of carbohydrate, an element of information needed to determine the energy balance. In
applying this knowledge to consider body weight
regulation, one introduces a biologically irrelevant
fact. One may learn to be in awe at the bodys ability to maintain approximate energy balances over
extended periods of an individuals life, as a consequence of the organisms ability to maintain substrate balances, but attempts to understand this as
the result of a regulatory process directed at the
energy balance itself are doomed to be frustrated.
Not surprisingly, failure to be aware of the inappropriate use of the energy and energy balance concepts
in considering body weight regulation and obesity
has spawned numerous misconceptions (1).
The intent of this article is to provide direct
access to the considerations that pertain to particular notions commonly cited in the context of
obesity. Their presentation under separate headings should be helpful, in particular to individuals
not necessarily immersed in obesity research who
want to understand the meaning of specific terms
and concepts.
1. Problems in applying the energy
balance concept

The energy balance equation is often invoked to

frame issues relating to body weight regulation
and obesity. The equation states that
Energy Balance = Energy Intake Energy Expenditure

which appears to support the statement that obesity is due to a positive energy balance, i.e., that

energy intake is exceeding energy expenditure.

Unfortunately, this seemingly obvious pronouncement sets up several misconceptions (1). First, this
view confuses past and present, obscuring the fact
that obese individuals, like lean subjects, tend to
reach a state of approximate energy balance, with
similar short-term fluctuations in the energy balance. Thus, one fails to recognize that the really
important difference between lean and obese
individuals is the degree of adiposity at which, on
average, their energy intake tends to adjust itself
to their energy expenditure. This varies widely
between individuals, in a manner influenced by
genetic as well as by environmental conditions.
Thus obesity is brought about by a failure of the
interactions between body composition and food
intake regulation to restrain eating when the
bodys energy stores are more than adequate for
health (Table1).
2. Problems with the metabolic
efficiency concept

Weight gains or losses generally do not exceed 1 or

2kg per year during long periods of an individuals life (2), reflecting an error of only 12% in the
adjustment of food intake to energy expenditure.
This corresponds to an average daily difference
between energy intake and expenditure in the
order of 2550kcal. Thus even minor differences
between energy intake and expenditure appear
to be able, in their cumulative long-term effect,
to cause or to prevent the progressive development of obesity. The second misconception promoted by notions based on the energy balance
equation is that obesity could therefore be due
to unusually low metabolic rates. This impression has been enhanced because rates of energy
expenditures per kg body weight or per kg fatfree mass (FFM) appear to be lower in obese
than in lean subjects, which is a simple artefact
due to a positive intercept in the correlations
between energy expenditure and body weight

Department of Biochemistry and Molecular Biology, University of Massachusetts Medical School, Worcester, Massachusetts, USA.
Correspondence:J.P. Flatt (jpaflatt@verizon.net)

676

VOLUME 19 NUMBER 4 | april 2011 | www.obesityjournal.org

editorial

Table 1 Issues and misconceptions about obesity

1. Problems in applying the energy balance concept
2. Problems with the metabolic efficiency concept
3. The misleading emphasis on the importance of low resting metabolic rates
4. Misleading expectations about the importance of adaptive thermogenesis
5. Problem in judging the importance of de novo lipogenesis and of its metabolic cost
6. The irrelevance of the nutrient partitioning concept
7. Failure to recognize the greater impact of energy intake than energy expenditure
8. Difficulties in understanding food intake regulation
9. Conditions for body weight stability: settling point vs. set point
10. Problems with the application of the RQ/FQ concept
11. The defense of body weight concept
12. The different roles played by CHO and fat in energy metabolism
13. Food intake regulation and carbohydrate balance
14. The difficulty in obtaining experimental evidence about the role of carbohydrate balance in food intake regulation
15. The need to distinguish between the role of carbohydrate balance in food intake regulation and the role of habitual
glycogen levels in body weight regulation
16. Understanding the recent increases in the preponderance of obesity
17. Why dont people eat even more?
18. Confusion about the leverage of exercise on body weight
19. Is dietary fat or is dietary CHO the major culprit in causing weight gain?
20. How can inherited traits influence body weight regulation?
21. The leverage of inherited vs. noninherited factors
22. BMI vs. % ideal body weight
CHO, carbohydrate; FQ, food quotient; RQ, respiratory quotient.

or FFM. Nevertheless it has led to speculations

about higher metabolic efficiency in obesity,
as a way to account for a higher than normal
proportion of energy retention. Such a notion,
sometimes tossed around without proper definition of metabolic efficiency (3), should long
ago have been dispelled by the realization that
total energy expenditure is raised in obesity, due
to the increase in lean body mass associated with
weight gain and the greater costs associated with
moving a heavier body. The utter confusion introduced by the energy efficiency concept becomes
apparent when one realizes that the large daily
variations in food consumption that occur when
food is freely available cause energy balance to
be positive on some days and negative on others.
Thus, during period of weight stability, metabolic
efficiency would oscillate between positive and
negative values over the short term, whereas it is
essentially equal to zero over extended periods.
3. The misleading emphasis on the importance
of low resting metabolic rates

Arguments about the potential long-term

impact on body weights of differences in resting energy expenditure, or of small diet or
obesity | VOLUME 19 NUMBER 4 | april 2011

physical activityinduced increases in energy

expenditure, hinge on the assumption that such
differences would not be offset by adjustments
in energy intake. This is a totally unwarranted
assumption, as there is no evidence indicating
that average energy intakes in subjects eating at
their own discretion are clamped at some particular level (4).
Several facts speak against the notion that
minor differences in metabolic efficiency or in
resting energy expenditure may be important
for body weight regulation. (i) Stature, which is
positively correlated with resting and with total
energy expenditures, has no impact on adiposity
(5). (ii) There is no correlation between % body
fat and deviations from predicted (or normal)
basal energy expenditure (Figure1) (6).
It is generally considered that causes cannot
safely be inferred from correlations, especially
from cross-sectional correlations. In the case considered here, however, the crucial fact is the lack
of correlation. This is much simpler to interpret,
as this evidence directly and simply demonstrates
that factors other than deviations from predicted
resting metabolic rate are overwhelming the possible effect of such deviations in determining the
677

editorial

70
Women

50
40
30
20

Men

50
40
30
20
10

10

0
60

70
60

% Body fat (m)

% Body fat (f)

60

40

20

20

40

60

% Deviation from predicted BEE

60

40

20

20

40

60

% Deviation from predicted BEE

Figure 1 Adiposity and basal energy expenditure (BEE). (a) % Body fat contents of 433 women whose ages,
anthropometric data, and BEE are listed in the dietary reference intake (DRI) database (45), plotted against the
deviation (expressed in terms of %) from their predicted (i.e., normal) daily BEE (as kcal/day) calculated using
the multiple regression equation in (45). % Body fat contents were predicted from their BMI, using the equation
correlating BMI and measurements of body fat by bio-impedance in the National Health and Nutrition Examination
Survey (NHANES) III data (45). (b) % Body fat contents of 335 men whose ages, anthropometric data, and BEE
are listed in the DRI database (45), plotted against the deviation (expressed in terms of %) from their predicted
(i.e., normal) daily BEE (as kcal/day) calculated using the multiple regression equation in (45). % Body fat
contents were predicted from using the equation correlating BMI and measurements of body fat by bio-impedance
in the NHANES III data (45).

degree of adiposity associated with the steady

state of weight maintenance.
Various arguments have been made about the
potential roles of futile cycles (3) and of brown
adipose tissue and uncoupling proteins (UCP)
in affecting adiposity by raising energy expenditure (7). However, their impact on overall energy
expenditure in man appears to be small at best.
Since differences in basal energy expenditure
have no statistically recognizable impact on body
fat contents, speculations about the potential
significance of UCP and futile cycles need to be
made cautiously. Given that long-term CHO balance will be maintained under all circumstances,
one should only expect an effect on body fat if
the resulting increases in energy expenditure
were largely covered by increments in fat oxidation, setting up an effect similar to that of physical activity (see #18 below).
At any rate, it is of interest to note that spontaneous food intake regulation is powerful enough
to obliterate the impact that meals and foods
served in uniform proportions could be expected
to exert on fatness among individuals whose
energy expenditures differ, for example, due to
differences in stature (5).
4. Misleading expectations about the
importance of adaptive thermogenesis

Adaptive thermogenesis describes changes in

resting energy expenditure which serve, or have
the effect of diminishing weight gains or weight
losses during periods of overconsumption or
starvation, relative to the weight changes which
678

would be expected if changes in resting energy

expenditure were solely due to changes in body
size and in the thermic effect of food (8). This
phenomenon has thus been considered to
reflect a kind of metabolic adaptation. While
adaptive thermogenesis has been found to be
substantial in some animal models, they are
modest in humans, to the point that it has been
difficult to establish them unambiguously. The
fact that differences in resting energy expenditure have no statistically recognizable impact on
adiposity (see Figure1) argues strongly against
the view that differences in adaptive thermogenesis, which occur only occasionally, play a
significant role in preventing or promoting the
preponderance of obesity.
5. Problem in judging the importance of
denovo lipogenesis and of its metabolic costs

Several reactions in the fatty acid synthesizing

pathway require adenosine triphosphate (ATP),
so conversion of glucose into fat requires a substantial investment of energy. If the costs for
prior conversion of glucose into glycogen as
well as for the transport of fatty acid synthesized
in the liver to adipose tissue are also included,
the cost for conversion of dietary carbohydrate
into fat may be assessed at some 25% (9). The
net ATP yield, i.e., the total number of ATP generated minus the number of ATP expended for
the transport, activation, and remodeling in the
metabolic process, was evaluated in refs. 3 and
10. This yield is higher during fatty acid oxidation than during glucose oxidation. The net
VOLUME 19 NUMBER 4 | april 2011 | www.obesityjournal.org

editorial

ATP yield from ingested glucose followed by

the oxidation of that amount of fat is thus not
much affected by initial conversion of glucose to
fat, or by the concomitant occurrence of lipogenesis and oxidation of a matching amount
of fat (10). Thus, contrary to a still commonly
held expectation, dissipation of dietary energy
by conversion of glucose into fat cannot explain
why high-carbohydrate diets are less conducive
to obesity than high-fat diets (11). In fact, the
issue is essentially moot, mainly because fatty
acid synthesis from glucose, estimated at some
10g/day (12) in subjects consuming a Western
diet, is of minor quantitative significance and
not sufficient to compensate for concomitant fat
oxidation even after massive CHO loads (13).
6. The irrelevance of the
nutrient-partitioning concept

The concept of nutrient partitioning has been

developed in the context of meat production,
where the proportions of nutrients consumed
retained as muscle or fat is important. It is relevant for periods of rapid growth, but it has
also been considered in discussing the problem
of obesity, in attempts to understand the differences in degrees of adiposity reached by different individuals (14). However, in lean as well as
in overweight adults whose weight is relatively
stable, essentially all the nutrients consumed
over a period of a few days are oxidized, regardless of diet, level of physical activity, and degree
of adiposity. Furthermore, the large variations in
physical activity, food consumption, and energy
balance occurring under free-living conditions
cause nutrient balances to vary considerably
from day to day. Thus the nutrient-partitioning
concept is rather meaningless when applied to
adult humans.
7. Failure to recognize the different impacts
of energy intake and energy expenditure
on energy balance

Food intake elicits an increase in energy expenditure, known as the thermic effect of food (TEF).
It is mainly due to the need to regenerate the ATP
used for the absorption, transport and storage of
the nutrients consumed (15). It is usually considered to amount to about 10% of the food energy
consumed when living on a mixed diet. In addition, modest increases or declines in resting energy
expenditure occur during periods of excessive
or deficient energy intake, which are commonly
referred to as adaptive thermogenesis (cf. #4). The
sum of the latter plus TEF is often referred to as
diet-induced thermogenesis. These changes can
obesity | VOLUME 19 NUMBER 4 | april 2011

only slightly attenuate, but not reverse, the impact

of changes in energy intake on the energy balance.
Unfortunately, the energy balance equation suggests that energy intake and energy expenditure
occupy equivalent roles in determining energy
balance, when in fact the factors governing energy
intakes influence the energy balance far more
powerfully than the factors determining resting
energy expenditure. This important fact does not
become evident by a simplistic consideration of
the energy balance equation.
8. Difficulties in understanding
food intake regulation

In spite of the multitude of physiological phenomena known to contribute to the regulation of food
consumption, variations in daily food intake are
very large. The coefficients of variation for intraindividual energy intake were found to average
23% (16). Furthermore, changes in food intake
are not closely synchronized with variations in
energy expenditure, which can also be substantial
(17). Evidently, metabolic functions can be readily sustained in spite of large daily deviations from
energy balance. That is why evolution was not
compelled to develop a precise control over daily
food consumption. The physiological mechanisms
involved in controlling food intake should therefore be expected to explain only a minor part of
the observed variability in daily food intake. This
has made it difficult to elucidate the mechanisms
involved. Furthermore, there appears to be considerable redundancy in this regulation. Thus,
inactivation or stimulation at one control site may
lead to temporary disruptions, but weight maintenance generally tends to become re-established
subsequently, albeit possibly at a different level.
Since most individuals maintain stable body
weights during long periods of their lives (2), and
since changes in resting energy expenditure cannot compensate for changes in intake (cf. #7), it
can be inferred that food intake tends to adjust
itself remarkably well to energy expenditure over
the long term, even though large daily deviations
from energy balance occur. Therefore, adjustment
of intake to expenditure in humans must obviously be happening over periods of several days.
This view is supported by the recent recognition
of corrective responses occurring with 34-day
delays (18).
9. Conditions for body weight stability:
settling point vs. set point

Although less obvious than the fact that energy

intake must be equal to energy expenditure, weight
stability also requires that the substrate mixture
679

editorial

oxidized be equivalent, on average, to the composition of the nutrient mix consumed. When substrate balance is not achieved, changes in body
composition occur, which in time are bound to
elicit adjustments in food intake (15,19).
The contributions made by carbohydrate and
by fat to the fuel mix oxidized is reflected in the
ratio of CO2 produced to O2 consumed. This ratio
is known as the respiratory quotient or RQ.
It varies between the values of 1.0, when CHO is
the predominant fuel, and 0.7, when oxidation of
fat provides most of the bodys energy. The ratio
of CO2 produced to O2 consumed during the biological oxidation of a representative sample of the
diet consumed is defined as the food quotient
or FQ (15). Stable body compositions will only
be sustained if the average RQ matches the average FQ of the diet.
The composition of the fuel mix oxidized and
hence the average RQ are influenced by the size
of the bodys substrate reserves. The steady state
of weight maintenance thus tends to become
established for a particular body composition in
a given individual living under a particular set
of circumstances. This corresponds to a settling
point (20). Such a view accommodates the fact
that circumstances cause weight stability to occur
for various degrees of adiposity. Thus it seems to
fit reality much better than the concept of a set
point or ponderostat (21) often invoked to
explain weight stability. In fact, such a concept
would seem to be utterly inconsistent with the
rise in the preponderance of obesity, since setpoints would have to be seen as preventing the
impact of changing circumstances. It has sometimes been considered that set-points are reset
for different conditions, but in effect this argument reduces the set-point phenomenon to a
settling point.
Rapid weight changes take place during growth,
as well as sometimes in adults. Such changes reflect
the fact that their body composition is still substantially different from that which corresponds
to the steady state of weight maintenance.

intakes are small and greatly influenced by daily

events. Except under tightly controlled conditions, the RQ/FQ ratio cannot be established with
sufficient precision to predict whether weight
gain or loss would take place. Reports finding
no correlation between RQ/FQ data and body
weights or body weight changes should therefore
not be taken to challenge the validity of the RQ/
FQ concept.

10. Problems with the application

of the RQ/FQ concept

Dietary carbohydrates and fats both provide substrates for the regeneration of the ATP expended
to sustain the bodys functions. Carbohydrates
generally contribute 4060% of dietary energy
and daily CHO intakes in adults consuming
mixed diets range from 200g in relatively small
and sedentary individuals to >500g in physically
active persons. The bodys glycogen reserves may
be estimated to vary between 1.0 and 1.5 times
the amount of carbohydrate consumed and oxidized during a typical day (22). The bodys fat

An RQ greater than the FQ indicates that CHO

makes a greater, and fat a smaller, contribution
to the substrate mix oxidized than their relative
proportions in the diet (3). On the other hand, a
RQ/FQ ratio below 1.0 reflects the oxidation of
a fuel mix containing a higher proportion of fat
than that provided by the diet. While this insight
is theoretically valid, the differences between RQ
and FQ under conditions of variable nutrient
680

11. The defense of body weight concept

The common tendency of individual body

weights to return to their original value after a
weight-changing intervention is often explained
as the manifestation of a mechanism tending to
defend a particular body composition. The
problem with this concept is that it appears
to imply that mechanisms exist to actively
drive the fat mass to a particular level, much
as one would expect if a set-point mechanism
existed (21). It fails to take into consideration
that before the intervention, body composition for a given individual had already evolved
until a steady state of weight maintenance had
become established. The body composition
that prevailed before such an intervention will
naturally tend to re-establish itself when the
disrupting intervention has ceased, not because
a given adiposity is defended or targeted, but
because weight maintenance tends to become
re-established at the settling point, i.e., at the
particular body composition which that individual had previously reached under this set of
circumstances (19).
Discussion of differing interpretation of such
concepts as settling point, set point, ponderostat,
and defense of body weight, which all appear to
provide tenable explanations for common observations, may seem to be futile, until one tries to
understand how environmental and dietary factors can influence adiposity and the preponderance of obesity.
12. The different roles played by carbohydrates
and fat in energy metabolism

VOLUME 19 NUMBER 4 | april 2011 | www.obesityjournal.org

editorial

content commonly amounts to 1015kg, being

much larger in obesity. The triglyceride reserve
typically ranges from 100,000 to 200,000kcal in
adults and it is thus some 100 times greater than
the CHO reserve of 1,0002,000kcal present in
the form of glycogen.
Given its small size in relation to its turnover
and the importance of maintaining adequate
blood glucose levels, maintenance of appropriate
glycogen stores present a considerable metabolic
challenge. Evolution was therefore compelled
to develop appropriate regulatory features to
adjust glucose oxidation to carbohydrate availability, through adjustments of the activity of
key enzymes and by hormonal signals, notably
those conveyed by insulin and glucagon. Thanks
to these, large variations in carbohydrate intake
can be accommodated without noticeable stress.
Other phenomena cause the oxidation of amino
acids to adjust themselves to protein intake, as
recognized long ago by monitoring nitrogen balances. On the other hand, the bodys large fat
stores are hardly affected by daily gains or losses
and adjustment of fat oxidation to fat intake has
received much lower priority in metabolic regulation. Indeed, consumption of fat has little or no
effect on fat oxidation, which declines even following fatty meals (23)! The rate of fat oxidation
varies greatly during the day, depending on nutrient intake and physical activity, but the amount
of fat oxidized in a 24-h period is in effect primarily determined by the difference between total
energy expenditure and the energy provided by
CHO plus protein intakes (19). While numerous
mechanisms operate to prevent large deviations
from CHO and protein balances, no such functions are at work to limit daily fat imbalances.
Not surprisingly, deviations from fat balance can
be much greater than deviations from CHO and
protein (or nitrogen) balances (24). It follows that
energy balance cannot be expected to be more
accurately maintained than fat balance. The roles
which CHO and fat play in the bodys fuel economy and in body weight regulation are therefore
markedly different, another fact that is neglected
when only the energy balance is considered (1).
Over the long term, changes in free fatty acid
levels and in insulin sensitivity brought about by
gains or losses in the adipose tissue mass influence
the average rate of fat oxidation (19,25). Thanks
to this long-term or chronic influence, body
composition will drift toward the degree of adiposity for which the average rate of fat oxidation
has become commensurate with fat intake (19).
Another important consideration needs to be
made here, namely that fat oxidation is restrained
obesity | VOLUME 19 NUMBER 4 | april 2011

by high glycogen levels (26), so a greater expansion of the fat mass is needed in individuals who
maintain relatively high glycogen reserves.
13. Food intake regulation
and carbohydrate balance

Spontaneous food consumption is increased after

periods of food deprivation, and inhibited after
deliberate weight gain, but only until the status
prevailing under the habitual ad libitum conditions is re-established. This makes it glaringly
apparent that food intake regulation does not
have maintenance of energy balance as its primary purpose! In addition, it becomes evident
that body composition is an important parameter
in the regulation of food intake (19).
Maintenance of appropriate glycogen levels
under conditions of ad libitum food intake would
be facilitated if, in addition to adjustment of glucose oxidation to carbohydrate intake, food intake
were to be regulated in a manner helping to sustain the balance between the use of glucose and
the influx of dietary carbohydrate. Making such
considerations at a time when the existence of
specific glucose receptors in the brain had just
been revealed, Mayer (27) developed the concept
of glucostatic regulation of food intake, in which
it is presumed that monitoring of blood glucose
levels by the brain generates signals for the regulation of food intake by the central nervous system. Indeed, it is now known that transient small
declines in blood glucose levels tend to elicit feeding (28). However, blood glucose levels vary greatly
during the day, and there is much overlap between
the values that prevail under different nutritional
and metabolic conditions. Subsequently, it was
therefore considered that changes in liver glycogen
levels (29) or in liver substrate oxidation rates (30)
would be more suitable or more likely to provide
appropriate feedback signals to the brain, as such
inputs can be based on some integrated parameter of substrate and carbohydrate utilization and
availability. Little is known about the mechanisms
and signals that may be involved, though there is
a possibility that they could be transmitted to the
central nervous system via the autonomous nervous system (31).
One could also entertain the notion that the
rate of change instead of, or in addition to, the
size of the bodys glycogen stores could provide
signals affecting hunger. This would account for
the fact that hunger is greatly attenuated after a
few days of total starvation and relatively well
tolerated on protein-sparing diets (32), once the
organism has become adapted to low but stable
glycogen and blood glucose levels.
681

editorial

14. The difficulty in obtaining experimental

evidence about the role of carbohydrate
balance in food intake regulation

In one of the few studies suitable to examine this

issue, Stubbs et al. (33) used continuous indirect
calorimetry in a respiratory chamber to established 24-h substrate balances in young men
eating ad libitum over 7 consecutive days. They
observed a negative correlation between 24-h
carbohydrate imbalances and subsequent food
intakes, which accounted for 510% of the variance in the next days energy balances and hence
energy intakes. Considering that most factors
influencing food consumption and variations in
the daily energy balance occur pretty much at
random once a particular lifestyle has become
established, even modest, but systematic, regulatory effects have the potential to exert significant long-term steering effects. These may be
difficult to establish with statistical certainty
during the short experimental periods feasible
in humans. The situation is similar to casino
games, where the small percentage gains programmed for the house would be difficult to
characterize in the short run, but lead in time to
significant profits.
15. The need to distinguish between the role
of carbohydrate balance in food intake
regulation and the role of habitual glycogen
levels in body weight regulation

Regulation of food intake in a manner helping to

maintain carbohydrate balance provides means
to achieve body weight stability, but only when
the situation has been reached for which the
composition of the fuel mix oxidized matches,
on average, the nutrient distribution in the diet.
However, in relation to body weights and health,
the mechanisms involved in stabilizing body
weights are less important than the factors determining the body fat content for which weight
maintenance tends to occur.
Glycogen levels are spontaneously maintained
at levels sufficient to prevent hypoglycemia and
to carry out habitual physical tasks. They also
spontaneously remain far below the level at
which rapid conversion of carbohydrate into
fat is induced in adults consuming mixed diets
(12,13,34). Glycogen levels in adults can be estimated to vary between a lower limit of some 150
200g and an upper limit of 500600g. Glycogen
levels are known to be higher on high-carbohydrate than on low-carbohydrate diets (22). They
are therefore not likely to vary over the entire
range when conditions are stable, but to oscillate
within a narrower range. Since neither the lower
682

nor the upper limits are specifically determined,

the range within which glycogen levels are habitually maintained can differ among individuals
and be affected by circumstances as well.
Although actual measurements of glycogen
levels in human populations are not yet available
from which one could gauge the influence of glycogen levels in reducing fat oxidation, this effect
can be recognized in experiments performed in
respiratory chambers, where changes in glycogen
content could be calculated (26,33). When the
range within which glycogen levels are habitually maintained rises, daily fat oxidation can be
expected to decrease, promoting fat accumulation. Thus, the role which increased habitual
glycogen levels will play in promoting obesity in
humans needs to be recognized! An attempt was
made to assess the scope of this effect in a computer model of human metabolism that reproduces
some of the behaviors known to regulate substrate
utilization in humans (35). Minor upward shifts
in habitual glycogen levels led to substantial body
weight gains, whereas decline in glycogen levels
entrained significant weight losses. The models
behavior thus supports the view that changes in
glycogen levels too small to be documented by
current means conceivably can explain recent
gains in adiposity.
16. Understanding the recent increases
in the preponderance of obesity

During the past few decades a continuing increase

in the prevalence of obesity in the United States
has occurred (36), even though neither the population gene pool nor the parameters considered
to have caused the progressive increase in obesity in affluent societies, such as higher dietary
fat content, greater buying power, and decreased
physical activity, appear to have changed substantially during this period. However, food diversity
and its appetizing qualities (37), changes in its
energy density and increased portion size, wider
choices, and advertising efforts have continued to
increase. In various ways, these factors encourage
food consumption so that higher glycogen levels
are reached after meals. Furthermore, depletion
of glycogen levels between meals is counteracted
by the ubiquitous availability and promotion of
appetizing snacks and caloric drinks. It will be
readily appreciated that together these environmental changes are bound to raise the range
within which glycogen levels are habitually maintained. By influencing the size of the fat mass necessary to cause fat oxidation to be commensurate
with fat intake, shifts in habitual glycogen levels
modify the steady-state conditions, although the
VOLUME 19 NUMBER 4 | april 2011 | www.obesityjournal.org

editorial

mechanisms regulating food intake remain the

same. The quantitative impact of raised habitual
glycogen levels has been examined in a computer model built to reflect the overall interactions
between glycogen levels and fat stores in humans.
It indicates that an increase by 25g in the range
within which glycogen levels are habitually maintained would require an increase in body fat content of 5% to restore fat oxidation to the previous
level, which would correspond to a weight gain
of some 7kg and to an increase of about 4 in the
BMI (BMI = Weight (kg)/(Height (m))2 (ref. 35).

Adjustments of food intake related to gains

or losses of glycogen can be expected to result
in a modulation of intake, effectively serving
to maintain glycogen levels sufficient to prevent
hypoglycemic events, while remaining far below
the levels at which appreciable de novo lipogenesis would be induced. Thus, the regulatory effects
serving to limit glycogen accumulation offer a
potential and plausible mechanism for restraining food intake when faced with an unlimited
supply of appetizing foods, but only once the
average rate of fat oxidation matches the average
fat intake.

17. Why dont people eat even more?

In humans, restraint on food intake can be due to

conscious efforts, to a degree susceptible to cultural
influences and to the prevailing level of adiposity
in the public. Unfortunately all such efforts fail too
frequently to prevent excess weight gain. It would
appear that in a large segment of the population,
control over food intake is largely dependent on
physiological mechanisms. These dont appear to
have been shaped by evolution to prevent excessive weight gain in conditions of an overabundant food supply, but can become effective once
the fat mass has enlarged. A number of signals
influencing food intake originate in the gastrointestinal tract and others are elicited by circulating
substrate and hormone levels. For the most part,
however, these decay overnight and do not carry
over much information from one day to the next
(38,39). Furthermore, these signals are commonly
thought to be more powerful in preventing energy
deficits than weight gain (40). Other mechanisms
must therefore be at work in modulating food
intake from one day to the next. In particular, it
is evident that even when highly desirable foods
are readily available, people eat substantially less
on most days than the amounts that they readily
consume on high-intake days(41). Even among
individuals who appear to combine dietary and
lifestyle habits believed to promote the development of obesity, weight gain over the years is much
slower than one would expect considering the
pleasure associated with eating, the loose regulation of daily food intake, and the great tolerance
for excessive intakes. We know little about the
phenomena limiting food intake, but their power
is made evident by the weight loss that spontaneously takes place after periods of deliberate overfeeding (4244). Further evidence can be seen in
the fact that there is no correlation between adiposity and stature, even though resting and total
energy expenditure are positively correlated with
stature (45) and many meals of standard size are
served to short and tall people (5).
obesity | VOLUME 19 NUMBER 4 | april 2011

18. Confusion about the effect of exercise

on the energy balance

Regular physical activity effectively reduces the

likelihood of developing obesity (46), and the
recommendation to increase exercise as a means
of controlling body weight is constantly emphasized to sedentary individuals (47). However, it
has been difficult to find a consistent explanation
to account for the fact that exercise causes a negative energy balance in some situations but not in
others. The issue is complicated by the fact that
physical exertion causes major perturbations in
metabolism, as it demands rapid and intensive fuel
mobilization to provide the substrates needed to
sustain ATP regeneration. Furthermore, exercise
affects metabolism beyond the period of physical activity itself. It causes the so-called excess
postexercise oxygen consumption (48), as well
as changes in substrate utilization.
What can be said is that in the absence of exercise, increases in the adipose tissue mass are often
required to raise fat oxidation to a rate commensurate with the diets fat content. The synergistic
effect of high dietary fat and lack of exercise in
promoting obesity is well illustrated in children.
Their spontaneously high physical activities used
to limit childhood obesity, until increasing inactivity caused a rise in its prevalence (49).
Since exercise leads to a steady state of weight
maintenance at lower levels of adiposity, it can be
inferred that exercise enhances fat oxidation more
than CHO oxidation (50). This is the case even
though glucose often makes a major contribution
to the fuel mix oxidized at the onset of exercise, as
this is evidently compensated later by an increase
in fat oxidation (51). Thus, introducing physical
activities into the daily routine enhances fat oxidation relative to CHO oxidation. Furthermore,
physical training increases the bodys ability
to use more fatty acids and less glucose (52).
Thus, the additional amount of food needed
to maintain CHO balance is then less than the
683

editorial

e xercise-induced increase in energy expenditure,

resulting in weight loss. In individuals who have
reached a steady state accommodating the impact
of habitual physical activity on fuel metabolism,
eating to maintain glycogen levels then again provides fat calories in amounts appropriate to maintain energy balance. With these considerations in
mind, one can readily understand why exercise
sometimes induces negative energy balance and
sometimes does not (50).
In the computer model cited above (35) the
impact of exercise is brought about by the greater
glycogen depletion induced by exercise between
meals. The relevance of this explanation can be
appreciated simply by considering meal-to-meal
intervals in terms of calories burned, rather than
in terms of time elapsed. In a corollary manner,
the model shows that consumption of snacks
between meals, by sustaining higher glycogen
levels between meals, leads to increased fat accumulation (35).
19. Is dietary fat or is dietary CHO the major
culprit in causing weight gain?

In view of the considerations made above, it is

not surprising that a high incidence of obesity is
typically encountered in sedentary populations
consuming diets providing substantial amounts
of fat. As shown recently for free-living and unrestricted eating conditions, energy intakes are
positively correlated with the fat content of the
foods consumed on a particular day (18). In most
animal models high-fat diets similarly promote
fat accumulation.
This underlying epidemiological evidence
appears to be contradicted by the often remarkably Good weight-losing results achieved on regimens severely restricting CHO but not fat intake,
such as the Atkins-type diets. It is important
here to distinguish between situations of acute
weight reduction, where intake is controlled by
a specific diet plan, and long-term efforts to control or reduce body weights. In regard to the latter, numerous testimonials report that reducing
dietary CHO leads to weight loss, while other
individuals experienced weight loss by limiting
fat intake. This leads to the conclusion in the first
case that CHO, and in the second case that fat,
is the dietary component most responsible for
excess fat accumulation. Such conflicting impressions, often promoted with much publicity, have
caused great confusion in the public. However,
these seemingly contradictory conclusions can
be accommodated simply by realizing that severe
restriction on the consumption of any major
macronutrient has the effect of substantially
684

reducing the selection of desirable foods from

which to partake. This will reduce the incentives
to overeat (37), even without having to engage in
formal calorie-counting!
Thus the answer to the question asked above
is that the major culprit is the unrestricted and
ubiquitous availability of a mixed diet, offering
numerous appetizing foods, often in large portions, in which sugar, and to an even greater
extent fat, contribute to raise the energy density.
20. How can inherited traits influence body
weight regulation?

One can readily appreciate that inherited personality traits, as well as learned and cultural behavioral traits, will influence food selection, ones
ability to control eating behaviors, the tendency
to engage in physical activities, and thereby
steady-state adiposity. There is no reason to think
that the way in which the organism is poised in
regulating the relative use of CHO and fat should
not be affected by inherited metabolic and hormonal differences (53). Indeed, in high CHOburners a greater expansion of the fat mass is
needed to achieve a given proportion of fat-toCHO oxidation. It has indeed been observed that
persons maintaining relatively high RQs under
standardized conditions are likely to be affected
by greater subsequent weight gain (54,55). This
effect is likely to be magnified when dietary CHO
is accompanied by substantial amounts of fat. An
inherited or permanently acquired effect affecting the mechanisms regulating the sparing effect
of glucose on fat oxidation, and vice versa, should
be expected to exert a potentially powerful leverage on body weight regulation. Indeed, this is
clearly demonstrated in the two-compartment
computer model (35). On the other hand, one
should not expect adiposity to be much influenced by genetic effects causing resting metabolic
rates to be higher or lower than the average (6).
21. The leverage of inherited vs. noninherited
factors

The relative impacts of inherited and noninherited

factors on body composition are commonly judged
by the contribution that they make to the variance
around the mean of BMI data. However, this does
not appropriately recognize the impact of factors
causing shifts in the means. To take such shifts
into account, it is useful to subdivide the noninherited factors into those determining the general
environment in which a population lives, i.e., the
environmental factors, and those affecting an individuals condition within this population, which
may be described as the circumstantial factors.
VOLUME 19 NUMBER 4 | april 2011 | www.obesityjournal.org

editorial

The importance of this distinction is that environmental factors can shift the means. For example,
the mean BMI (s.d.) of Nigerian men living in
Nigeria was reported to be 21.0 4.2 when living
in Nigeria, whereas it was 27.0 5.4 when living in
Chicago (56). The data summarized by Costa and
Steckel (57) show the average BMI of adult men
in the United States as rising from 22.5 in 1894
to 23.6 in 1944, and to 25.7 in 1991. While variability about the means reflects equal proportions
of values higher and lower than average, environmental forces capable of changing the mean have a
unidirectional impact. The latter is bound to have
a more powerful influence on the proportion of
individuals whose BMI will exceed certain limits
than the variations about the mean attributable to
circumstantial factors, such as socioeconomic and
lifestyle variables. Thus, environmental, circumstantial, and genetic factors contribute in different
ways to adiposity and to the obesity epidemic. In
most situations, only the circumstantial factors
should be thought to be controllable by individuals. Strategies for weight control should take this
into account by offering opportunities to strive for
weight control within families or in group settings.
This could provide a helpful measure of change
in an individuals environment. Encouraging the
development of physical activity habits, notably in
children and adolescents, also offers the potential
for improving the environment, if they could
become a routine in a population.
22. BMI vs. % ideal body weight

For many decades, body weights were assessed

relative to ideal body weights (IBW), i.e., the
weight associated with the longest life expectancy,
as judged from life insurance data. It requires the
use of tables giving IBW as a function of height.
This has now been supplanted by the use of the
BMI, which does not require such tables. In its
simplicity, BMI is well suited to establish epidemiological data. A BMI between 25 and 30 indicates
overweight, and BMI >30 obesity. However, the
BMI is a highly abstract notion and information in
terms of the BMI does not convey nearly as much
information to individuals and patients as learning
how their weight relates to IBW. Does their weight
fall into the 2040% above IBW range indicative
of overweight, or by how much does it exceed the
140% of IBW value corresponding to obesity?
Now that hand-held computers can easily be programmed to generate % IBW data (which could
be upgraded to take age into account), one could
again consider informing patients about their
weight status according to the more understandable, and potentially alarming, % IBW criterion.
obesity | VOLUME 19 NUMBER 4 | april 2011

Conclusions

The problem of body weight regulation and obesity ultimately boils down to two questions: (i) what
are the mechanisms which operate to correct the
imbalances created by the large short-term variations in energy intake (cf. #8, 1315,17), whether
they result in exact energy balances or in slow
rates of gain or loss of fat over time, and (ii) why
is the state of approximate weight maintenance
reached for such widely differing levels of adiposity (cf. #27, 912, 1516,1821)? This second
issue is one of practical significance, but it must be
addressed in a manner consistent with the mechanisms that bring about weight stability. A plausible explanation, compatible with this condition,
can be proposed to account for the recent increase
in the preponderance of obesity. It attributes the
influence of the food supply to its effect in raising
the range within which glycogen levels are habitually maintained. Furthermore, it seems important
to overcome several misconceptions that appear
to stand in the way of a better understanding of
weight maintenance. This may ultimately help in
learning how to enhance the effectiveness with
which the physiological regulation of food intake
and conscious efforts are capable of attenuating
environmental pressures to overeat.
2011 The Obesity Society

REFERENCES
1. Flatt JP. How NOT to approach the obesity problem. Obes
Res 1997;5:632633.
2. Nooyens AC, Visscher TL, Verschuren WM et al. Age,
period and cohort effects on body weight and body mass
index in adults: The Doetinchem Cohort Study. Public
Health Nutr 2009;12:862870.
3. Flatt JP, Tremblay A. Energy expenditure and substrate
oxidation. In: Bray GA, Bouchard C, James WPT (eds).
Handbook of obesity. Marcel Dekker: New York, 1997, pp
705731.
4. Schoeller DA. The energy balance equation: looking back
and looking forward are two very different views. Nutr Rev
2009;67:249254.
5. Flatt JP. Adaptive changes in thermogenesis are NOT
important in the aetiology of obesity. Progr Obes Res
2003;9:713717.
6. Flatt JP. Differences in basal energy expenditure and
obesity. Obesity (Silver Spring) 2007;15:25462548.
7. Nedergaard J, Golozoubova V, Matthias A, Asadi A,
Jacobsson A, Cannon B. UCP1: the only protein able
to mediate adaptive non-shivering thermogenesis
and metabolic inefficiency. Biochim Biophys Acta
2001;1504;82106.
8. Frayn KN. Metabolic Regulation: A Human Perspective.
Blackwell Science: Oxford, UK, 2003.
9. Acheson KJ, Schutz Y, Bessard T et al. Glycogen
storage capacity and de novo lipogenesis during massive
carbohydrate overfeeding in man. Am J Clin Nutr
1988;48:240247.
10. Flatt JP. Physiological and metabolic control of
macronutrient balance. In: Berthoud HR, Seeley RJ (eds).
Neural control of macronutrient selection. CRC Press:
Boca Raton, 2000, pp 143164.
685

editorial

11. Bray GA, Popkin BM. Dietary fat intake does affect obesity!
Am J Clin Nutr 1998;68:11571173.
12. Strawford A, Antelo F, Christiansen M, Hellerstein
MK. Adipose tissue triglyceride turnover, de novo
lipogenesis, and cell proliferation in human adipose tissue
measured with 2H2O. Am J Physiol Endocrinol Metab
2004;286:E577E588.
13. Acheson KJ, Flatt JP, Jquier E. Glycogen synthesis versus
lipogenesis after a 500 gram carbohydrate meal in man.
Metabolism 1982;31:12341240.
14. Shadid S, Jensen MD. Nutrient partitioning. In: Bray GA,
Bouchard C (eds). Handbook of Obesity, 2nd edn. Marcel
Dekker: New York, 2004, pp 753766.
15. Flatt JP. The biochemistry of energy expenditure. Rec Adv
Obes Res 1978;2:211228.
16. Bingham SA, Gill C, Welch A et al. Comparison of dietary
assessment methods in nutritional epidemiology: weighed
records v. 24h recalls, food-frequency questionnaires and
estimated-diet records. Br J Nutr 1994;72:619643.
17. Edholm OG, Adam JM, Healy MJ et al. Food intake
and energy expenditure of army recruits. Br J Nutr
1970;24:10911107.
18. Bray GA, Flatt JP, Volaufova J, Delany JP, Champagne
CM. Corrective responses in human food intake identified
from an analysis of 7-d food-intake records. Am J Clin Nutr
2008;88:15041510.
19. Flatt JP. McCollum Award Lecture, 1995: diet, lifestyle, and
weight maintenance. Am J Clin Nutr 1995;62:820836.
20. Wirschafter D, Davis JD. Set points, settling points and
control of body weight. Physiol Behav 1977;19:7578.
21. Keesey RE, Hirvonen MD. Body weight set-points:
determination and adjustment. J Nutr 1997;127:1875S1883S.
22. Hultman E, Nilsson LH. Factors influencing carbohydrate
metabolism in man. Nutr Metab 1975;18 Suppl 1:4564.
23. Flatt JP, Ravussin E, Acheson KJ, Jquier E. Effects
of dietary fat on postprandial substrate oxidation
and on carbohydrate and fat balances. J Clin Invest
1985;76:10191024.
24. Abbott WG, Howard BV, Christin L et al. Short-term energy
balance: relationship with protein, carbohydrate, and fat
balances. Am J Physiol 1988;255:E332E337.
25. Flatt JP. Role of the increased adipose tissue mass in the
apparent insulin insensitivity of obesity. Am J Clin Nutr
1972;25:11891192.
26. Schrauwen P, Lichtenbelt WD, Saris WH, Westerterp KR.
Fat balance in obese subjects: role of glycogen stores. Am
J Physiol 1998;274:E1027E1033.
27. Mayer J. Regulation of energy intake and the body weight:
the glucostatic theory and the lipostatic hypothesis.
Ann NY Acad Sci 1955;63:1543.
28. Campfield LA, Smith FJ. Transient declines in blood
glucose signal meal initiation. Int J Obes 1990;14 Suppl
3:1531; discussion 31.
29. Russek M. Current status of the hepatostatic theory of food
intake control. Appetite 1981;2:137143.
30. Friedman MI. Body fat and the metabolic control of food
intake. Int J Obes 1990;14 Suppl 3:5366; discussion 66.
31. Tian Q, Nagase H, York DA, Bray GA. Vagal-central
nervous system interactions modulate the feeding
response to peripheral enterostatin. Obes Res
1994;2:527534.
32. Wadden TA, Stunkard AJ, Day SC, Gould RA, Rubin CJ.
Less food, less hunger: Reports of appetite and symptoms
in a controlled study of protein-sparing modified fast. Int J
Obes 1987;11:239249.
33. Stubbs RJ, Harbron CG, Murgatroyd PR, Prentice AM.
Covert manipulation of dietary fat and energy density: effect
on substrate flux and food intake in men eating ad libitum.
Am J Clin Nutr 1995;62:316329.
34. Acheson KJ, Schutz Y, Bessard T et al. Nutritional
influences on lipogenesis and thermogenesis after a
carbohydrate meal. Am J Physiol 1984;246:E62E70.

686

35. Flatt JP. Carbohydrate-fat interactions and obesity

examined by a two-compartment computer model. Obes
Res 2004;12:20132022.
36. Flegal KM, Carroll MD, Kuczmarski RJ, Johnson CL.
Overweight and obesity in the United States: prevalence
and trends, 1960-1994. Int J Obes 1998;22:3947.
37. Rolls BJ, Rowe EA, Rolls ET et al. Variety in a meal enhances
food intake in man. Physiol Behav 1981;26:215221.
38. Blundell JE, Cotton JR, Delargy H et al. The fat
paradox: fat-induced satiety signals versus high fat
overconsumption. Int J Obes 1995;19:832835.
39. Johnstone AM, Stubbs RJ, Harbron CG. Effect of
overfeeding macronutrients on day-to-day food intake in
man. Eur J Clin Nutr 1996;50:418430.
40. Mattes RD, Pierce CB, Friedman MI. Daily caloric intake
of normal-weight adults: response to changes in dietary
energy density of a luncheon meal. Am J Clin Nutr
1988;48:214219.
41. Flatt JP. What do we most need to learn about food intake
regulation? Obes Res 1998;6:307310.
42. Roberts SB, Young VR, Fuss P et al. Energy expenditure
and subsequent nutrient intakes in overfed young men. Am
J Physiol 1990;259:R461R469.
43. Diaz EO, Prentice AM, Goldberg GR, Murgatroyd PR,
Coward WA. Metabolic response to experimental
overfeeding in lean and overweight healthy volunteers.
Am J Clin Nutr 1992;56:641655.
44. Pasquet P, Apfelbaum M. Recovery of initial body weight
and composition after long-term massive overfeeding in
men. Am J Clin Nutr 1994;60:861863.
45. Dietary Reference Intakes. National Academies Press:
Washington, DC, 2002.
46. Besson H, Ekelund U, Luan J et al. A cross-sectional
analysis of physical activity and obesity indicators in
European participants of the EPIC-PANACEA study. Int J
Obes (Lond) 2009;33:497506.
47. Wadden TA, Vogt RA, Andersen RE et al. Exercise in the
treatment of obesity: effects of four interventions on body
composition, resting energy expenditure, appetite, and
mood. J Consult Clin Psychol 1997;65:269277.
48. Bahr R, Ingnes I, Vaage O, Sejersted OM, Newsholme EA.
Effect of duration of exercise on excess postexercise O2
consumption. J Appl Physiol 1987;62:485490.
49. Dietz WH Jr, Gortmaker SL. Do we fatten our children at
the television set? Obesity and television viewing in children
and adolescents. Pediatrics 1985;75:807812.
50. Flatt JP. Integration of the overall effects of exercise. Int J
Obes 1995;19:S31S40.
51. Warren A, Howden EJ, Williams AD, Fell JW, Johnson
NA. Postexercise fat oxidation: effect of exercise duration,
intensity, and modality. Int J Sport Nutr Exerc Metab
2009;19:607623.
52. Smith SR, de Jonge L, Zachwieja JJ et al. Fat and
carbohydrate balances during adaptation to a high-fat diet.
Am J Clin Nutr 2000;71:450457.
53. Galgani JE, Moro C, Ravussin E. Metabolic flexibility
and insulin resistance. Am J Physiol Endocrinol Metab
2008;295:E1009E1017.
54. Zurlo F, Lillioja S, Esposito-Del Puente A et al. Low ratio of
fat to carbohydrate oxidation as predictor of weight gain:
study of 24-h RQ. Am J Physiol 1990;259:E650E657.
55. Eckel RH, Hernandez TL, Bell ML et al. Carbohydrate
balance predicts weight and fat gain in adults. Am J Clin
Nutr 2006;83:803808.
56. James WPT. The epidemiology of obesity. In: Chadwick DJ,
Cardew G (eds). The origins and consequences of obesity. Ciba
Foundation Symposium 201. Wiley: Chichester, 1996, pp 116.
57. Costa DL, Steckel RH. Long-term trends in health,
welfare, and economic growth in the United States.
In:Steckel RH, Floud R, (eds). Health and welfare during
industrialization. Univ Chicago Press: Chicago, 1997,
pp 4789.

VOLUME 19 NUMBER 4 | april 2011 | www.obesityjournal.org