cron-web.org

[A1CR]

The previously presented paper (1) appeared to find no
significant
differences in the energy expended in the rhesus monkey CR
experiments, as
indicated and shown in http://tinyurl.com/yltrgz . The
paper below (2) describes background
http://en.wikipedia.org/wiki/Anthropometric features of the
model system
rhesus monkeys, to define normal anthropometry properties
used in (1) in
terms of. Table 3 in (2) appeared to be informative with
regard to the characteristics of body anthropometry
properties in the rhesus monkeys used in experiments for CR,
as related to
features such as the levels of glucose, insulin and
cholesterol in the blood
in well-controlled experiments in animals closely
representing humans and
fed identical diets.

1. Raman A, Ramsey JJ, Kemnitz JW, Baum ST, Newton W,
Colman RJ, Weindruch
R, Beasley MT, Schoeller DA.
Influences of calorie restriction and age on energy
expenditure in the
rhesus monkeys.
Am J Physiol Endocrinol Metab. 2006 Aug 8; [Epub ahead of print]
PMID: 16896169 http://tinyurl.com/nsc5p

2. Raman A, Colman RJ, Cheng Y, Kemnitz JW, Baum ST,
Weindruch R, Schoeller
DA.
Reference body composition in adult rhesus monkeys:
glucoregulatory and
anthropometric indices.
J Gerontol A Biol Sci Med Sci. 2005 Dec;60(12):1518-24.
PMID: 16424283

Rhesus monkeys have been used as models to study obesity and
disease. The
aim of this study was to define body mass indices for
underweight and
obesity in rhesus monkeys. Longitudinal data collected over
8-14 years from
40 male and 26 female rhesus monkeys were analyzed. Body
weight, insulin
sensitivity index, and disposition index were regressed
against percent body
fat (%BF). A minimal %BF beyond which further loss of body
weight resulted
in loss of lean mass was determined to be 11.5% in older
males, 8% in adult
females, and 9% in younger adult males. Insulin sensitivity
index and
disposition index reached minimum values at 23% fat in older
males, 18% in
adult females, and 21% in younger adult males, indicating
obesity. The
estimated reference range for %BF was 9%-23% in male and
8%-18% in female
monkeys, corresponding to body mass indices of 32-44 kg/m(2)
for male and
27-35 kg/m(2) for female monkeys.

... Monkeys of both sexes with excess body weight (BW) due
to increased fat
mass have been shown to have fasting hyperinsulinemia (1),
elevated insulin
response to intravenous glucose or marginally impaired
glucose tolerance
(2), and elevated fasting serum triglycerides (3). These
glucoregulatory and
liporegulatory abnormalities are similar to those of obese
humans;
nevertheless, there are no uniform definitions for
overweight and obesity in
rhesus monkeys. Similarly, large losses of lean body mass
can have
deleterious consequences such as damage to organs and
disturbances in
cardiac function due to attrition in the myocardial mass
(4); however, there
are also no uniform definitions of underweight in rhesus
monkeys. This
creates an ambiguity in the interpretation of results based
on the
non-uniform definition of underweight, overweight, or obese
animals when
used as models for human disease.

In humans, obesity is generally defined as a body mass index
(BMI) > 30
kg/m2 based on the morbidity risks of cardiovascular
diseases, hypertension,
diabetes, and associated symptoms (5-7), and underweight is
defined as BMI <
19 kg/m2 (8). In contrast, obesity and underweight in rhesus
monkeys has
been characterized using morphometric parameters such as BW,
BMI, and
abdominal circumference (AC), which are reliable predictors
of body fat
(3,9). For example, obesity in rhesus monkeys has been
characterized using a
BW greater than 2 standard deviations (SD) above the mean
for their sex (3)
and break-points for percentage body fat (%BF), such as 25%
BF (10) and 30%
BF (11).

The most important variable that addresses the majority of
the gluco- and
liporegulatory abnormalities in an individual is body fat
mass (3).
Hyperinsulinemia, hypertriglyceridemia, decreased glucose
clearance rate,
and glucose disposal can be seen with elevated %BF (12). Too
low a %BF,
however, may also be detrimental. Higher all-cause mortality
rates have been
observed in individuals with low BMI (13,14). The increase
in mortality rate
due to lower BMI is not fully understood, but factors such as
osteoporosis-induced fractures (15), decreased vitamin A
status [leading to
decreased survival rates for acute illnesses (16)], and
deficient levels of
body fat (16) have been suggested as possible mechanisms. In
addition, a
systematic analysis of the composition of weight loss has
shown that
mortality decreases when the weight loss is due to loss of
fat, but
increases when it is due to loss of fat-free mass (FFM)
(17). Unfortunately,
most of these definitions were not based on systematic
analysis of any
metabolic parameters or variables in rhesus monkeys in a
manner similar to
that which has been used to define underweight, overweight,
and obesity in
humans making it difficult to compare outcomes between
rhesus monkeys and
humans.

... %BF point can indicate the minimum %BF an animal should
have and hence
define underweight. Because not all primate research
institutions may have
ready access to body fat measuring equipment, reference
ranges of BMI will
be ascertained using the highly correlated relationship of
%BF and BMI. AC
is highly correlated with visceral adiposity in humans as
well as nonhuman
primates (9,18). Studies done on humans have shown that
increased abdominal
obesity is associated with increased risk of type 2
diabetes, cardiovascular
diseases, hypertension, and hypercholesterolemia (19,20).
Abdominal
adiposity is also associated with hyperinsulinemia, higher
plasma glucose
and insulin levels, and eventually glucose intolerance which
will be
reflected in the insulin sensitivity. Similar to those of
BMI, reference
ranges of AC will be ascertained using the highly correlated
relationship of
%BF and AC.

... Longitudinal data from 40 male and 26 female rhesus
monkeys which are
part of Wisconsin National Primate Research Center (WNPRC)
were used in this
analysis. These monkeys are part of an ongoing dietary
restriction and aging
study ... Data consisted of 639 values from 66 animals (34
calorie
restricted and 32 control) over a span of 14 years in older
animals and 8
years in adult animals. Monkeys were caged individually in
standard
stainless steel cages ... The cages had inside dimensions of
89 cm width, 86
cm depth, and 86 cm height. Room temperature was maintained
at 21°C, and the
animals were maintained on a 12-h light/dark cycle with
lights on between 6
AM and 6 PM. Animals were fed a semipurified diet (Teklad,
Madison, WI)
containing 15% lactalbumin, 10% corn oil, and 65%
carbohydrate. ...

... dual energy x-ray absorptiometry (DXA ... crown-rump
length (CRL) ...
Fasting triglycerides (TGb)

... RESULTS

The data used for this analysis are from an ongoing study,
and each animal
has multiple representations in this data set with the OM
measured for 14
years and the AF and AM measured for 8 years. The
characteristics of the
animals are summarized in Table 1. Within males, the OM and
AM differed in
KG, Gb, DI, plasma cholesterol, and TGb. Female monkeys were
significantly
different from males (OM and AM; p <.0001) in their BW, BF
(kg), BMI, AC,
and SI. Basal glucose concentrations (Gb) were significantly
higher in the
AM compared to the OM and AF (p <.0007), but basal insulin
levels were not
different among the three groups. TGb was lower in AF than
in AM but was not
different from the OM monkeys, whereas cholesterol levels
were higher in AM
than in OM. These findings prompted us to stratify the
analysis according to
age range and gender for the analyses that follow.

Table 1. Group Characteristics (Mean±SD).
=====================================================
Variable Units OM AF AM
=====================================================
Weight kg 11.9±3a 8.3±2b 11.9±2a
Body fat kg 2.9±2a 1.9±1b 2.4±2a
Body fat % 21.5±10 19.9±11 17.9±9
Body mass index kg/m2 42.0±9a 34.6±7b 41±7a
Abdominal circumference cm 51.0±11a 45.6±9b 50.1±10a
Basal glucose level mmol/L 3.4±0.4a 3.4±0.4a 3.6±1.2b
Basal insulin level pmol/L 285±289 254±274 205±195
Glucose disappearance rate % 6.4±3a 10.2±5 7.3±4c
Insulin sensitivityindex 10-5/min-1/ (pmol/L) 4.6±4a 7.1±6b
5.7±5a
Disposition index/min 377±282a 760±491b 526±360c
Fasting triglycerides mmol/L 1.5±1.4a 1.1±0.8a 2.0±5b
Plasma cholesterol mmol/L 4.7±0.9a 4.9±0.9ab 5.1±2.4b
=====================================================
Notes: a,b,c Groups with different letters are
significantly different (p
<.05).
SD = standard deviation; OM = older males; AF = adult
females; AM = adult
males.

Break-points in the relationships between %BF and SI and DI
were at a point
of change in slope between the dependent and independent
variables. When %BF
was regressed with SI, an exponential relationship was
observed in male (OM,
%BF = 28.216 * e(-0.091 * SI), p <.001; AM, %BF = 23.694 *
e(-0.08 * SI), p
<.0001) and female monkeys (%BF = 26.12 * e(-0.067 * SI), p
<.001) (Figure
1). Using the R software, the break-points for maximum
attainable %BF before
its SI becomes minimal were 23.2% in OM, 20.8% in AM, and
17.5% in AF
monkeys. DI also showed an exponential relationship with %BF
in male (AM,
%BF = 26.67 * e(-0.001 * DI), p <.001; OM, %BF = 26.025 *
e(-0.001 * DI), p
<.001) and female (%BF = 22.819 * e(-0.0004 * DI), p <.001)
monkeys.
Accordingly, the %BF break-points for DI were 23.2% in OM,
22% in AM, and
16.4% in AF monkeys. In a regression plot of %BF against SI,
AM and OM
showed a similar increase in SI with decreasing %BF compared
to AF. The mean
difference in %BF among all three groups was significantly
different at any
given SI (OM and AM = 3.5%, AF and AM = 2.1%, and OM and AF
= 1.5%; p <.05).
However, for a given %BF, females had higher absolute SI
values than males.
At a mean value of 5.7 SI units, the average %BF was 20% in
OM and 18% in AM
versus 21.3% in the AF monkeys. Using the interaction
between SI and group,
this difference proved to be significant (p <.0001).
However, there was no
significant effect of age on the relationship of SI and %BF
(interaction of
SI x Age) when animals in individual groups were analyzed.
Also, when the
males were grouped together there was no significant
interaction between SI
and age on %BF indicating that age in this group of animals
does not affect
the relationship between SI and %BF.

Besides SI and DI, TGb and cholesterol levels and additional
indices of
glucoregulation, KG, AIR, Ib, and Gb levels were examined
for any
associations with %BF (Table 2). Though Gb, Ib, TGb, and
cholesterol levels
showed similar relationships with %BF, a break-point
analysis using these
variables did not reach significance due to a high
variability in the data.
Hence these variables did not contribute to the
determination of the maximal
body fat level.

Table 2. Correlation of Systemic Metabolic Indices.
================================================
% Fat KG Gb, mmol/L Ib, pmol/L AIR, pmol/L
Cholesterol, mmol/L
================================================
OM
KG -0.49 - - - - -
Gb, mmol/L 0.36 -0.19 - - - -
Ib, pmol/L 0.37 -0.21 0.25 - - -
AIR, pmol/L 0.35 -0.09 0.05 0.51 - -
Cholesterol, mmol/L 0.04 -0.05 0.06 0.01 -0.21 -
TGb, mmol/L 0.39 -0.31 0.11 0.59 0.46 0.18
AM
KG -0.61
Gb, mmol/L 0.24 -0.24
Ib, pmol/L 0.49 -0.35 0.14
AIR, pmol/L 0.27 -0.01 -0.26 0.43
Cholesterol, mmol/L 0.03 -0.15 0.38 0.09 -0.19
TGb, mmol/L 0.18 -0.23 0.48 0.24 -0.14 0.87
AF
KG -0.40
Gb, mmol/L 0.24 -0.21
Ib, pmol/L 0.28 -0.17 0.36
AIR, pmol/L 0.37 -0.02 -0.18 0.41
Cholesterol, mmol/L 0.07 -0.11 0.06 -0.16 -0.23
TGb, mmol/L 0.44 -0.27 0.07 0.27 0.36 -0.00
================================================
Note: SD = standard deviation; KG = glucose
disappearance rate; Gb =
basal glucose level; Ib = basal insulin level; AIR = acute
insulin response;
TGb = fasting triglycerides.

The lower end of the range for %BF was ascertained using the
relationship
between BW and %BF. The %BF of male monkeys showed an
exponential
relationship with their BW (OM, %BF = 2.49 * e(0.168 * BW),
p <.001; AM, %BF
= 1.03 * e(0.225 * BW), p <.001) and female (%BF = 1.02 *
e(-0.335 * BW), p
<.001). Percent BF was regressed with BW sequentially to
identify the
break-point where the relationship indicates most of the
weight loss as FFM
(Figure 2). The minimum %BF an animal should have before
increasing loss of
lean body mass occurs was ascertained to be 8.5% in OM, 6%
in AM, and 5% in
AF monkeys.

The above break-point analysis does not provide a measure of
statistical
range around the break-point values. The values, therefore,
have
limitations. There may be individual variation among
animals, and the
measurement of %BF may not be exact. In either case, animals
at the lower
end of the reference body fat spectrum could be at a greater
risk for
negative health-related outcomes even with slight decreases
in body fat. It
is therefore prudent to add a safety factor to the low-end
break-point.
Based on a comparison of %BF measurement between DXA and
total body water,
we calculated a mean difference of 3% for the determination
of %BF and used
this as the safety level needed on the lower end of
reference %BF. In so
doing, the minimal %BF below which animals can be classified
as underweight
were 11.5% in OM, 9% in AM, and 8% in AF monkeys.

The %BF values can also be translated to BMICRL. Percent BF
showed
significant correlations with BMI in all three groups of
monkeys (%BF
= -19.3 + 0.97 * BMI; r2 = 0.7, p <.0001 in OM; %BF = -29.2
+ 1.4 * BMI; r2
= 0.8, p <.0001 in AF, and %BF = -30.7 + 1.2 * %BF; r2 =
0.8, p <.0001 in
AM; Figure 3) with the mean %BF (mean±SD) at 21.5±10% in OM,
19.9±11% in AM,
and 17.9±9% in AF monkeys. Hence a reference BMI of 32-44
kg/m2 in the AM,
34-44 kg/m2 in OM, and 26-38 kg/m2 in AF monkeys was deduced.

Similarly, these break-points can be translated to AC
values. Because %BF
and AC have a linear relationship (%BF = -11.9 + 0.66 * AC;
r2 = 0.6, p
<.0001 in OM; %BF = -29.5 + 1.08 * AC; r2 = 0.8, p <.0001 in
AF, and %BF
= -24.9 + 0.9 * AC; r2 = 0.9, p <.0001 in AM), we estimated
a reference AC
of 40-54 cm in AM, 35-53 cm in OM, and 35-44 cm in AF
monkeys using the
reference range of body fat (Figure 4).

DISCUSSION

Using glucoregulatory indices and changes in body
composition, we developed
a reference range of %BF a monkey can have before being
classified as
underweight or overweight and obese. Insulin-sensitivity
measures and
changes in FFM during weight change have been used to
identify the reference
range for %BF; hence, these data promise to be a good index
to define health
using a group of metabolic predictors in young and old
rhesus monkeys. This
is an effort in classifying rhesus monkeys into underweight,
reference, and
obese based on their %BF and will make it easier to compare
health outcomes
with humans.

The linear relationship between BMI and body fat can be used
effectively to
ascertain the %BF of an individual based on their BMI.
However, this
relationship needs to be analyzed with caution, because a
higher BW can be
due to a higher lean body mass, in which instance using BMI
may lead to
misclassification of the individual as overweight or obese.
Nonetheless, BMI
has been used effectively to assess obesity and underweight
in numerous
human studies (26-28).

Overweight and Obesity
The literature is replete with evidence of body fat being
strongly
associated with most of the glucoregulatory processes in the
body. A lower
%BF has been associated with better glucose regulation and
better insulin
sensitivity (29). Significant correlations between basal and
stimulated
insulin levels with various indices of obesity have been
noted in monkeys
(2,9) and humans (30,31). Hyperinsulinemia has been shown to
occur as one of
the initial consequences of increased BW or body fat (32).
In fact, this
relationship has been best reported in monkeys with body fat
greater than
30% of their BW (33,34). Conversely, a reduction in BW or
body fat has been
shown to decrease insulin dosage or eliminate the need for
supplemental
insulin in type 2 diabetics (35,36). Hence, we used insulin
sensitivity and
disposition indices to identify the upper end of reference %BF.

Hypertriglyceridemia and hypercholesterolemia have been
observed in obese
humans (37) and nonhuman primates with higher %BF (3,9), but
we were unable
to find a break-point associated with these variables.
Perhaps this is
because there is a strong genetic component to the elevated
plasma
triglycerides (32,38). This genetic component may have
obscured the
break-points, wherein the fasting triglyceride and total
cholesterol levels
were higher in animals with higher body fat but were highly
variable (CV was
0.3 for plasma cholesterol and 1.9 for TGb). Nonetheless,
the mean TGb
levels in the male and female monkeys with a reference %BF
was 1.1±0.9
mmol/L and 0.7±0.4 mmol/L, respectively, compared to 2.7±5
mmol/L and
1.4±0.9 mmol/L in obese animals (p <.001). The levels seen
in reference %BF
animals were within the ranges defined for humans (39).

Overweight Versus Obese
The relationship between %BF and glucoregulatory indices was
exponential and
could not be used to differentiate between overweight and
obese monkeys.
Nonetheless, obesity is associated with hyperinsulinemia,
insulin
resistance, and glucose intolerance (11). It has been shown
that animals
with increasing body fat may gradually become diabetic after
going through a
sequence of events of normoglycemia-normoinsulinemia,
normoglycemia-hyperinsulinemia, and
hyperglycemia-hyperinsulinemia. With
regard to this sequence, there are two diabetic monkeys in
the larger study
cohort that were not included in this analysis. The %BF of
these two animals
was 34.2% and 37.6% at the time of diagnosis. In addition,
Gresl and
colleagues (1) reported one additional animal in the larger
study that has
since died and was not included in the current data analysis
(this animal
became diabetic and had a %BF of 36% at the time of
diagnosis). In our
analysis, we concluded that the maximum body fat a male
animal could have
before SI became minimal was 22% of BW. Hence, we conjecture
that male
animals with %BF between 22% and 36% can be categorized as
overweight, above
which we see more animals with frank diabetes.

The findings of Hotta and colleagues (40) can be compared
with ours. Hotta
and colleagues grouped a cohort of rhesus monkeys according
to their fasting
plasma insulin and glucose levels using a priori values to
characterize the
animals as lean (normal) hyperinsulinemic or diabetic
(obese). They reported
that the "normal weight" monkeys in their study had a mean
%BF of 18.1±3.3%
and normal fasting plasma glucose and insulin
concentrations. The male
monkeys in our data set with a "reference" %BF (<22%; mean =
16±4.5%) were
also normoglycemic (3.4±0.4 mmol/L) and normoinsulinemic
(200±143 pmol/L);
these values are comparable to those of the animals of Hotta
and colleagues
(40) (Table 3). The obese group of Hotta and colleagues had
a mean %BF of
32.6±2.7% and were hyperinsulinemic but normoglycemic, thus
paralleling our
data in the animals above the upper %BF break-point. Among
the obese group
reported by Hotta and colleagues (40), noninsulin-dependent
diabetes
mellitus was observed in a group of monkeys with mean %BF
35.1±4%, similar
to the three diabetic animals in our larger study. These
data support our
conjecture on classifying overweight rhesus monkeys as
22-34%BF and obese
rhesus monkeys as 35%BF, although further evidence is needed
because the
number of animals on which this is based is still small.

Table 3. Characteristics of Monkeys When Assigned to
Underweight, Normal,
and Obese Categories (Mean 6 SD)
===================================================
Variable Weight, kg Body fat, % Gb, mmol/L Ib, pmol/L TGb,
mmol/L
Cholesterol, mmol/L
===================================================
Males
Underweight 8.8±1a 5±1a 3.1±0.3a 90±48a 0.6±0.3a 4.6±1a
Normal 10.7±2b 16±5b 3.4±0.4b 200±144b 1.1±1a 4.7±1a
Obese 14±2c 29±5c 3.6±1c 372±337c 2.8±5b 5.2±2b
Females
Underweight 5.9±1a 4.2±0.5a 3.1±0.3a 133±118a 0.6±0.2a 4.5±1a
Normal 7.0±1b 10±3.9b 3.4±0.4b 164±107a 0.7±0.4a 5.1±1b
Obese 9.3±1c 27±6.2c 3.5±0.5b 319±324b 1.4±1b 4.9±1b
===================================================
Notes: a,b,c Different letters indicate significant
differences between
underweight, normal, and obese groups within males and
females ( p <0.05).
SD=standard deviation; Gb=basal glucose level; Ib=basal
insulin level;
TGb=fasting triglycerides.

Underweight with Minimal Body Fat
On the other end of the spectrum of body fat, we identified
a minimal %BF to
define reference or normal weight of 10% in male and 8% in
female rhesus
monkeys. There are, however, few studies in literature that
provide data
against which we can compare these values. One study by
Altmann and
colleagues (41), however, reported that young adult baboons
foraging in the
wild had %BF as low as 2% (in adult females) and 1% (in
adult males).
Anthropometric data indicated that, despite their lower %BF,
growth among
the female baboons continued and that the animals maintained
reproductive
function. This might indicate that our minimal %BF values
were too
conservative; however, caution should be maintained when
comparing our data
to animals under free-living conditions, due to the high
fiber content of
the diet and the high activity level of the free-living
animals compared to
our caged animals. Both of these factors could
differentially influence the
relationship between SI and %BF.

Despite the data on %BF in wild animals, we were concerned
about using
minimal %BF obtained by break-point analysis to define
underweight. This
concern stemmed from the rapid weight loss and loss of FFM
that was observed
when some animals neared but were still above these critical
values and the
potential for serious negative health outcomes that could
accompany the loss
in FFM (13,14). This concern was amplified by the knowledge
that the
measurement of %BF is accompanied by a measurement error,
and thus when the
break-point values are applied to individual animals, %BF
might be
overestimated and the animal could be at risk of being
underweight despite
an apparently normal %BF. Because of these two factors, we
added a safety
margin of 3% to the %BF in an effort to reduce the risk in
individual
animals.

Finally, it should be noted that our data were derived from
animals that
were part of a long-term dietary restriction study. Thus,
there was a
possibility that the relationships between %BF and the other
variables in
the diet-restricted animals were a little different from
general colony
animals. No interactions between dietary treatment and the
parameters used
for the above analyses were observed, however; thus we
conclude that these
values to define underweight, overweight, and obesity are
reasonable. By
using these cut-offs to define nutritional status in rhesus
monkeys,
comparison of studies between those conducted in rhesus
monkeys with those
conducted in humans should be easier.