0021-972X/05/$15.00/0
Printed in U.S.A.
The Journal of Clinical Endocrinology & Metabolism 90(2):849 – 854
Copyright © 2005 by The Endocrine Society
doi: 10.1210/jc.2004-0455
Cardiovascular Fitness and Exercise as Determinants of
Insulin Resistance in Postpubertal Adolescent Females
J. Z. Kasa-Vubu, C. C. Lee, A. Rosenthal, K. Singer, and J. B. Halter
Departments of Pediatrics (J.Z.K.-V., A.R., K.S.) and Internal Medicine (C.C.L., J.B.H.), University of Michigan, Ann Arbor,
Michigan 48019
estimated by homeostasis model assessment index (HOMAIR)
derived from fasting glucose and insulin concentrations.
BMI was not related to HOMAIR (P ⴝ 0.20), RDE showed a
marginal relationship (P ⴝ 0.049), whereas percent body fat
and VO2max were significantly related to HOMAIR (P ⴝ 0.01
and 0.0008, respectively). In a multiple regression model,
VO2max was a more critical determinant of insulin resistance
than percent body fat (P ⴝ 0.03 vs. P ⴝ 0.67) or RDE (P ⴝ 0.01
vs. 0.51).
For prevention strategies in youth, physical inactivity may
represent a greater metabolic risk than obesity alone. (J Clin
Endocrinol Metab 90: 849 – 854, 2005)
In obese adolescents, body mass index (BMI) is a poor predictor of insulin resistance, and the potential role of diminished
physical activity has not been quantitated. We measured possible determinants of sensitivity to insulin in 53 adolescent
females with a BMI between the 10th and the 95th percentile.
We hypothesized that across weight and fitness spectra, relative fat mass, rather than BMI, and cardiovascular fitness
would be predictors of insulin resistance.
We measured body composition by total-body dual x-ray
absorptiometry. Self-reported weekly frequency of aerobic
exercise for 1 h (RDE) was recorded, and maximal oxygen
consumption (VO2max) was measured. Insulin sensitivity was
T
centiles is unknown. Identifying risk factors for decreased
sensitivity to insulin in youth would facilitate the development of preventative strategies against the onset of type 2
diabetes at younger ages (13).
Several methods can be used for the determination of
insulin resistance. The homeostasis model assessment index
for insulin resistance (HOMAIR) has been used as a surrogate
measure for insulin resistance and has been validated in
adolescents (11, 12). We recruited healthy young women in
the immediate postpubertal period. We tested the hypothesis
that BMI, relative fat mass, and cardiovascular fitness would
be predictors of insulin resistance in adolescent women
across the weight and fitness spectra.
HE OBESITY EPIDEMIC (1) and recent reports about the
increased prevalence of type 2 diabetes in youth (2)
bring to the forefront the need for prevention. The most
frequently identified risk factors for the onset of type 2 diabetes in youth are puberty, obesity, and evidence of insulin
resistance (3). A sedentary lifestyle as well as the heavy
consumption of high-energy foods and beverages (4) are
thought to largely contribute to the obesity epidemic now
reaching children of all ages (3). The relative contributions of
diet and exercise as risk factors for obesity and type 2 diabetes are still open for debate (3, 5), while the time dedicated
to physical activity in school physical education programs is
decreasing (6). Trends of relative inactivity increase markedly in girls as they mature through puberty (7). The substantial decrease in physical activity observed in multiethnic
cohorts of adolescent girls is associated with a steady increase of body mass index (BMI) (7).
Although insulin resistance is an important risk factor for
diabetes, little information is available about insulin resistance in adolescent females (8). In adults, obesity is a strong
determinant of insulin resistance (9, 10), and risk assessment
is often based on body weight (3). In contrast to adults, recent
studies in very obese children and adolescents (11, 12) suggest that an elevated BMI by itself does not appear to predict
the degree of insulin resistance. However, the relationship of
insulin resistance to BMI in adolescents across the BMI per-
Subjects and Methods
Subjects
The protocol was approved by the Institutional Review Board of
the University of Michigan Health System. Fifty-three healthy adolescent and young women ages 16 –21 yr with a BMI between the 10th
and the 95th percentile for age (CDC Growth Charts from the National
Center for Health Statistics; http://www.cdc.gov/nchs/data/nhanes/
growthcharts/set3/chart%2016.pdf) were studied a minimum of 2 yr
after the age at menarche. Participants were recruited from local university campuses and high schools and were enrolled after signing an
Institutional Review Board-approved consent form. An additional signature was required from one parent or legal guardian for all subjects
less than 18 yr of age. All participants were healthy, had completed
pubertal maturation, and were taking no medications. Adolescents who
were using a hormonal method of birth control or had a history of
hirsutism were excluded. TSH was normal in all subjects. Pregnancy was
excluded by serum measurement of human chorionic gonadotropin. All
participants were instructed to contact the Principal Investigator at the
start of the menstrual cycle and were admitted to the General Clinical
Research Center (GCRC) in the follicular phase of the cycle, which was
confirmed by the determination of serum progesterone.
There was a self-reported assessment of physical activity as well as
an interview-based measure of fitness. To quantify individual levels of
fitness, maximal oxygen consumption (VO2max) was determined from
First Published Online November 30, 2004
Abbreviations: BMI, Body mass index; DEXA, dual x-ray absorptiometry; FAI, free androgen index; GCRC, General Clinical Research
Center; RDE, reported days of exercise; VO2max, maximal oxygen
consumption.
JCEM is published monthly by The Endocrine Society (http://www.
endo-society.org), the foremost professional society serving the endocrine community.
849
850
J Clin Endocrinol Metab, February 2005, 90(2):849 – 854
a standard exercise tolerance test. All patients were exercised on a
treadmill using the Bruce protocol (14). Measurements during the test
included continuous 12-lead electrocardiogram monitoring with blood
pressure measurements made during the last minute of each stage of
exercise. An electrocardiogram was obtained at 2 and 3 min of each stage
of exercise.
Respiratory measurements made during testing included minute
ventilation, oxygen consumption (VO2), carbon dioxide production
(VCO2), and respiratory exchange. These were measured continuously
during each test by a Sensor Medic Vmax cardiopulmonary exercise
system.
Patients were encouraged to exercise as long as they physically could.
Testing was stopped if there were any significant changes or symptoms;
otherwise, the patients exercised until they were too fatigued to continue. To assure a maximal test was performed and a true VO2max was
obtained, at least two of the following criteria needed to be met: 1)
respiratory exchange rate at maximal exercise at least 1.10, 2) achievement of more than 90% of age-predicted maximal heart rate (220 –age),
and 3) a plateau in VO2 (increase in VO2 ⬍ 2 ml/kg䡠min) despite increased exercise workload.
We estimated exercise level by asking each participant the number of
days per week during which she exercised for at least 1 h of continuous
aerobic exercise recorded as reported days of exercise (RDE) per week;
this activity self-report provided a qualitative assessment of the participant’s perception of her own level of exercise. The duration of 1 h was
chosen for its practicality, given that most aerobic classes or exercise
videos are limited to 1 h. The types of physical activity reported included
jogging, aerobics, field hockey, volleyball, soccer, in-line skating, skiing,
diving, daily biking, resistance training, and exercise videotapes.
Body composition and relative body fat were obtained by dual x-ray
absorptiometry (DEXA) using a total-body scanner (model DPX-L, Lunar Radiation Corp., Madison, WI) (15). No blood was drawn during the
day or for the 36 h after either the treadmill or DEXA tests. Special effort
was made to complete all studies within the span of one menstrual cycle.
Outpatient visits for both treadmill and DEXA tests were within10.4 ⫾
0.1 (se) and 10.7 ⫾ 0.1 d, respectively, from the determination of insulin
resistance by HOMAIR.
To obtain an estimation of daily caloric intake and dietary composition, a prospective 3-d food diary was obtained and followed by an
interview with the GCRC dietary staff. Nutrient calculations were performed using the Nutrition Data System for Research software, version
4.03, developed by the Nutrition Coordinating Center, University of
Minnesota, Minneapolis, MN, Food and Nutrient Database 31, released
November 2000 (16). The participants were not asked to change their
dietary habits, and while in the GCRC for 36 h, they were fed a selfselected diet from the hospital menu with standard meal times. After an
overnight fast from 2100 h the night before, three blood samples were
obtained at 10-min intervals between 0500 and 0800 h for measurements
of glucose and insulin. Because IGF-I and SHBG have a potential modulating role on insulin sensitivity (17, 18), their levels were determined
on these samples. The testosterone level was determined at the time of
admission to the GCRC.
Assays
Serum samples were stored at – 80 C until assayed. SHBG and IGF-I
concentrations were measured using a solid-phase, two-site chemiluminescent enzyme immunometric assay for use with the Immulite automated analyzer (Immulite Diagnostic Product Corp., Los Angeles,
CA). For SHBG, assay sensitivity was 0.2 nmol/liter; intra- and interassay coefficients of variation were 6.5 and 8.7%, respectively. For IGF-I,
assay sensitivity was 20 ng/ml; intra- and interassay coefficients of
variation were 2.5 and 10.7%, respectively.
Progesterone was measured to confirm the phase of the menstrual
cycle, and testosterone was measured for the determination of the free
androgen index (FAI). Both steroids’ concentrations were measured
using the VITROS progesterone reagent pack and VITROS immunodiagnostic products progesterone calibrators on the VITROS ECi immunodiagnostic system. The assay uses a competitive immunoassay technique. For progesterone, assay sensitivity was 0.079 ng/ml (0.25 nmol/
liter); intra- and interassay coefficients of variation were 5.2 and 9.5%,
respectively. For testosterone, assay sensitivity was 0.865 ng/dl (0.03
Kasa-Vubu et al. • Insulin Resistance in Girls
nmol/liter); intra- and interassay coefficients of variation were 2.7 and
5.6%, respectively.
The glucose assay was performed on the Cobas Mira Chemistry
Analyzer from Roche Diagnostics Corp (Indianapolis, IN). The reagent
used for the analysis was purchased from Diagnostic Chemicals Limited
(Oxford, CT). Glucose was measured by the hexokinase enzymatic
method. Intra- and interassay coefficients of variation were, respectively, 0.6 and 3.2%. Insulin was measured by double-antibody RIA.
Sensitivity of the insulin assay is 2.1 ␮U/ml (15.1 pmol/liter). Intra- and
interassay coefficients of variation were 3.8 and 4.8%, respectively. The
cross-reactivity with proinsulin was 71.6%.
Calculations
SHBG levels are inversely related to the level of androgenicity. FAI
was calculated with the following formula: 3.467 ⫻ total testosterone
(ng/dl)/SHBG (nmol/liter) (19). The insulin resistance index was estimated using the HOMAIR: mean fasting insulin (␮U/ml) ⫻ fasting
glucose (mmol/liter)/22.5 (20, 21). As an estimate of fitness, VO2max
was calculated by dividing the highest recorded VO2 during exercise
with the participant’s total body mass. Percent body fat was calculated
from body composition analysis by DEXA.
Data and statistical analysis
The results are expressed as mean and sd. The relationships between
the dependent variable insulin resistance and the independent variables
BMI, percent body fat, diet, androgenicity, physical activity, and
VO2max were analyzed using univariate linear regression. Correlations
between all these variables were examined using Pearson correlations.
Tests for collinearity were performed to assess for significant multicollinearity. Stepwise multiple regression models were constructed using
Mallow’s C(p) criterion for the dependent variable of insulin resistance
and the independent variables that were significant by univariate linear
regression. Statistical analysis was performed using SAS version 6.12
(SAS Institute, Inc., Cary, NC). A value of P ⱕ 0.05 was selected to
indicate statistical significance.
Results
Subjects’ characteristics
Table 1 describes the participants’ nutritional and hormonal characteristics. Among the 53 patients, 36 were nonHispanic-Whites, 12 were African-Americans, four were
Asian-Americans, and one was Hispanic-White. IGF-I, testosterone, and FAI were comparable to age-specific norms
(age-dependent reference ranges from Esoterix, Austin, TX:
TABLE 1. Subjects’ hormonal and nutritional characteristics
Age (yr)
HOMAIR
Glucose (mg/dl)
Insulin (␮U/ml)
BMI (kg/m2)
RDE (h/wk)a
VO2 max (ml/kg 䡠 min)a
% Body fat
Fat mass (g)
Testosterone (ng/dl)b
FAIb
Dietary intake (kcal/d)c
% Diet carbohydratec
% Diet fatc
% Diet proteinc
IGF-I (ng/ml)
n ⫽ 52.
n ⫽ 46.
c
n ⫽ 50.
a
b
Mean (SD)
Range
18.7 (1.3)
2.9 (1.1)
93.1 (7.6)
12.7 (4.7)
23.3 (3.1)
3.8 (2.5)
40.5 (6.8)
31.1 (8.3)
19918 (8552)
32.3 (14.9)
4.7 (3.4)
1799.0 (462.9)
67.0 (8.2)
15.2 (5.3)
17.8 (6.1)
307.0 (84.5)
16–21
0.8–5.2
75.3–112.7
3.6–25.7
18.2–30.4
0.0–7.0
25.7–52.2
16.9– 49.7
(7672– 48112)
10.0–90.0
0.9–22.1
989.0–3027.0
40.3– 82.9
6.4–30.4
8.3– 43.9
110.0–573.0
Kasa-Vubu et al. • Insulin Resistance in Girls
J Clin Endocrinol Metab, February 2005, 90(2):849 – 854
http://www.esoterix.com/files/expected_values.pdf; and
Ref. 22). The average BMI was at the 75th percentile for this
age group, and the range was 18 –30 kg/m2, spanning from
underweight to overweight and moderately obese. The
range for HOMAIR was 0.8 –5.2 with 60% of values above
2.5, a cutoff previously used for insulin resistance in healthy
adult males (20).
Cardiovascular fitness relative to total body mass was in
the normal range for this age group with an average VO2max
of 40.5 (6.8) ml/kg䡠min (14). The average caloric intake was
1799.0 (462.9) calories or 30% greater than the estimated
resting energy expenditure of 1350 cal/d for this age group
(23) and compatible with mild to moderate physical activity
(23). RDE correlated with VO2max (P ⫽ 0.0003; r ⫽ 0.49).
As expected, BMI was significantly correlated with percent
body fat (P ⫽ 0.0001; r ⫽ 0.85) and total fat mass (P ⫽ 0.0001;
r ⫽ 0.74).
851
FIG. 1. Univariate association between HOMAIR and percent body
fat. Insulin resistance increases with greater percent body fat (r ⫽
0.34; P ⫽ 0.01).
Univariate relationships
The results of the univariate regression for HOMAIR for all
the variables analyzed are summarized in Table 2. BMI was
a poor predictor of HOMAIR (P ⫽ 0.20), but percent body fat
was significantly linked to HOMAIR (r ⫽ 0.34; P ⫽ 0.01) (Fig.
1) as well as RDE (P ⫽ 0.049). As shown in Fig. 2, VO2max
was also highly associated with HOMAIR (r ⫽ 0.45; P ⫽
0.0008), and for every unit increase in VO2max, HOMAIR
decreased by 0.07. The estimated daily carbohydrate intake
over 3 d, fat intake, and protein intake all had poor predictive
TABLE 2. Univariate regression table for HOMAIR as a
dependent variable (n ⫽ 53)
Independent variable
␤ (SE)
t
statistic
P
value
r
value
BMI
BMI percentile
% Body fat
Total fat mass
VO2max
RDE
IGF-I (ng/ml)
Testosteronea
FAIa
Dietary carbohydrate (%)b
Dietary protein (%)b
Dietary fat (%)b
0.06 (0.05)
0.01 (0.01)
0.04 (0.02)
0.0 (0.0)
⫺0.07 (0.02)
⫺0.12 (0.06)
0.002 (0.002)
0.003 (0.011)
0.033 (0.050)
⫺0.002 (0.019)
⫺0.030 (0.026)
0.046 (0.029)
1.29
4.97
2.54
0.80
⫺3.58
⫺2.07
1.36
0.29
0.66
⫺0.13
⫺1.176
1.55
0.20
0.11
0.01c
0.42
0.0008c
0.049a
0.18
0.77
0.51
0.90
0.25
0.13
0.18
0.22
0.34
0.11
⫺0.45
⫺0.27
0.19
0.04
0.10
⫺0.02
⫺0.16
0.21
Nutrient intake was from 3-d diary.
n ⫽ 46.
b
n ⫽ 50.
c
Statistically significant.
a
TABLE 3. Multivariate regression models for HOMAIR as a
dependent variable
Independent
variable
Model A: n ⫽ 53;
R2 ⫽ 0.21
Model B: n ⫽ 52;
R2 ⫽ 0.21
a
% Body fat
␤ (SE)
t statistic
P value
0.01 (0.02)
0.4
VO2max
⫺0.64 (0.03)
⫺2.2
0.03a
RDE
⫺0.45 (0.06)
⫺0.7
0.51
VO2max
⫺0.07 (0.02)
⫺2.7
0.01a
Statistically significant.
0.67
FIG. 2. Univariate association between HOMAIR and VO2max. Insulin resistance decreases with increased cardiovascular fitness (r ⫽
– 0.45; P ⫽ 0.0008).
value for HOMAIR. Neither IGF-I nor FAI was significantly
associated with HOMAIR.
Multiple regression models
We used multiple regression analysis to further explore
the relative contributions of body composition and fitness to
insulin resistance as summarized in Table 3. The dependent
variable in the model was HOMAIR. The independent variables tested in the model were VO2max and percent body fat
and accounted for 21% of the variance. VO2max remained a
significant predictor even when accounting for percent body
fat, whereas percent body fat did not retain significance.
Thus, insulin sensitivity was predicted by the VO2max
univariate model, accounting for 20% of the variance in
HOMAIR (P ⫽ 0.0008).
When both measures of physical activity in these young
women were simultaneously included in a regression model,
RDE was no longer statistically significant, indicating
VO2max was a better measure of the exercise-HOMAIR relationship (Table 3).
The stepwise regression models were also analyzed with
the C(p) statistic (Table 4). Among the three predictors that
852
J Clin Endocrinol Metab, February 2005, 90(2):849 – 854
TABLE 4. Stepwise regression model using C(p) statistic for
HOMAIR as a dependent variable
a
Independent variable
R2
C(p) statistic
VO2max
% Body fat
RDE
VO2max, % body fat
VO2max, RDE
0.20
0.12
0.08
0.21
0.21
0.56a
5.04
7.33
2.37
2.12
Smallest C(P) statistic.
were significant by univariate analyses and included into the
model using the C(p) statistic, VO2max remained the best
predictor of HOMAIR with the smallest C(p) statistic. Tests
for collinearity for VO2max, percent body fat, and RDE
yielded variation inflation factors that were all less than 10,
ranging from 1.33–2.15. No condition indices were more than
30. Thus, the degree of collinearity did not confound the
results from the multiple regression models.
Discussion
In adult women, a healthy weight considerably reduces
the risk of type 2 diabetes (9). In contrast to older populations,
the relationship between obesity and the risk for insulin
resistance and type 2 diabetes is not known in younger
women. In this multiethnic cohort of postpubertal adolescent
females, we found that BMI had poor predictive value for
insulin resistance, whereas cardiovascular fitness was a predictor of HOMAIR even when accounting for percent body
fat.
In youth, the impact of physical activity has been inferred
by indirect methods such as activity reports (4, 24), the documentation of sedentary lifestyle (25, 26), physical activity
recall (7), or the participation in organized sports (27). The
debate about how much exercise is desirable for health maintenance is still open given that there is significant subjectto-subject variation (28). The recommendations for daily
physical activity are estimated at a minimum of 30 min a day
but more recent reports have proposed at least 1 h a day for
the best metabolic benefits (29, 30). In addition, practical
considerations in our study design took into account that
most exercise classes at the local gymnasiums or exercise
videotapes favored by some participants were available in
1-h modules. Our goal was to measure fitness and compare
that assessment to the participants’ own assessment of their
physical activity based on self-report, similar to the type of
information that would be exchanged in a health maintenance visit.
The fitness range in our participants was similar to those
reported in previous studies. The adolescent girls studied by
Andersen et al. (31) showed a decrease of VO2max during late
puberty as body weight increased. In the more recent study
of Ahmad et al. (14), VO2max of adolescent girls showed a
modest increase between 12 and 14 yr of age followed by a
drop for the 16- to 18-yr group with an average of 40.1 ⫾ 6.1
ml/kg䡠min. This stood in sharp contrast to boys in both
studies, who showed an increase in cardiovascular fitness in
late puberty. Taken together with the dramatic decline in
physical activity reported by Kimm et al. (7), these reports
highlight the vulnerability of adolescent girls whose fitness
Kasa-Vubu et al. • Insulin Resistance in Girls
tends to stagnate in face of increasing weight. The contribution of our study is to show the link between insulin resistance and cardiovascular fitness within the range of VO2max
measured in this age group, because those with higher than
average fitness had a lower HOMAIR index. Longitudinal
studies will help determine whether deliberate strategies to
optimize VO2max in adolescent girls can minimize the risk
for insulin resistance.
We sought to measure physical activity in the participants
by reducing the subjectivity inherent to survey methods (32).
By measuring VO2max we were able to quantify physical
fitness rather than relying solely on self-reported physical
activity (26). Although self-reported exercise had a significant positive relationship with HOMAIR, VO2max was a
stronger explanatory variable than RDE, which was not significant when included within the same model as VO2max.
Our findings identify relative physical inactivity as an important predictor of insulin resistance in young women. In
the same adolescents studied, relative fat mass was a predictor of HOMAIR, but it lost its predictive value if cardiovascular fitness was introduced in the statistical model. Thus,
in this population, adiposity does not appear to matter as
much as fitness as a predictor of insulin resistance.
The lack of association between BMI and insulin resistance
found in our study is consistent with two recent studies with
a focus on the relationship between obesity and type 2 diabetes in children and adolescents. In a multiethnic cohort of
167 obese children, Sinha et al. (12) found a high prevalence
of impaired glucose tolerance, yet the BMI had a poor predictive value for risk factors associated with type 2 diabetes,
possibly because the population studied was already severely and homogeneously obese. Similar findings were recently reported in a larger cohort of obese European adolescents, in which the degree of obesity was not linked to
insulin resistance per se (11). However, we cannot exclude an
important relationship between BMI and HOMAIR in a larger
group of participants and a broader range of BMI including
leaner as well as heavier subjects than those studied here.
The norm for insulin resistance as determined by HOMAIR
in postpubertal young women has not been defined before
(11, 33), and clearly more age- and gender-specific studies are
needed. Over 60% of our participants had HOMAIR greater
than the normal cutoff of 2.5 used by others (11) to identify
impaired insulin sensitivity. Our findings parallel those reported by Invitti et al. (11) in their large cohort of children of
both genders across pubertal stages where 64% of obese
children had HOMAIR greater than 2.5. However, whether
cardiovascular fitness was a predictor of relative sensitivity
to insulin in that group was not determined (11).
Despite its ease of determination, HOMAIR does have
some limitations in the assessment of responsiveness to insulin. First, it may not account for the effects of oscillatory
insulin release, although this pitfall can be partially overcome by repeated sampling as was done in our study (20).
Second, HOMAIR does not assess insulin sensitivity with the
breadth and accuracy of the steady-state euglycemic clamp
(33), which remains the gold standard for this measurement.
HOMAIR derives from basal insulin level rather than a quantitative measured response to an insulin challenge as in the
euglycemic clamp or response to dynamic changes of en-
Kasa-Vubu et al. • Insulin Resistance in Girls
dogenous or exogenous insulin as in the frequently sampled
iv glucose tolerance test (34). Basal insulin depends on a
sensitive insulin assay and may be confounded by the measurement of proinsulin, which could vary among subjects
(35). More studies are needed to establish whether the
HOMAIR has a predictive value for insulin-stimulated states
in adolescent females.
Because overt evidence of androgen excess such as hirsutism was an exclusion criteria, polycystic ovary syndrome
was an unlikely contributor to insulin resistance (36) in our
population. This was confirmed by the lack of association
between FAI and HOMAIR. In an effort to standardize for the
enhancing effect of puberty on insulin resistance (33, 37), all
participants were recruited after the completion of puberty
and a minimum of 2 yr after menarche, a time when participants were expected to be near final height. To further
control for insulin resistance associated with puberty, IGF-I
was measured and showed no significant relationship with
HOMAIR.
A role for the relative contribution of dietary protein, carbohydrate, and fat to insulin sensitivity has been proposed
(38, 39). We were not able to detect a relationship between
dietary composition and HOMAIR, although the data collection was limited to only 3 d. Long-term dietary studies with
a more detailed analysis of dietary fats and high-fiber carbohydrates, which was not done in this study, might be more
informative on the potential for relationship of dietary factors to insulin resistance in young women.
Adolescent females show a dramatic decrease of physical
activity as they progress through puberty (7), which places
them at risk for obesity. There have been mounting reports
about the lack of physical activity in children across the
United States (6). However, much focus has been placed on
the deleterious effects resulting from the frequent consumption of high-energy foods (4, 40). If VO2max is a strong
predictor of insulin resistance in adolescent girls, then an
effective strategy to reduce metabolic risk may be to reverse
the serious trend of decreasing physical activity both in the
schools and in the more general extracurricular environment.
Our subjects were in the latter part of adolescence and
would for the most part qualify as young adults. More studies are needed in younger pediatric populations. Albeit well
characterized, our study sample was small, thus limiting the
ability to generalize about the relationships observed. However, our study echoes the recent report that total fat mass
reduction does not by itself improve metabolic outcomes
(41). As the fight against obesity gains momentum (42, 43)
and more radical treatments are being considered for youth
(44, 45), it will be important to determine whether they provide long-term protection against type 2 diabetes without a
concomitant long-term exercise intervention that is sustained
enough to improve cardiovascular fitness.
Acknowledgments
We thank Sheila Gahagan for her assistance in reviewing this manuscript and Terrin Meckmongkol for technical assistance. We are most
grateful to Andrzej Galecki for his statistical expertise.
Received March 11, 2004. Accepted November 15, 2004.
J Clin Endocrinol Metab, February 2005, 90(2):849 – 854
853
Address all correspondence and requests for reprints to: Josephine
Z. Kasa-Vubu, M.D, M.S., Department of Pediatrics, University of
Michigan Medical Center, Ann Arbor, Michigan 48019-0718. E-mail:
[email protected].
This study was funded by Grant 1K12HD01438-01 BIRCWH from the
Office of Women’s Health Research and the National Institute of Child
Health (J.Z.K.-V.), a Department of Veterans Affairs’ Career Development Award (to C.C.L.), and GCRC Grant M01-RR-0042.
References
1. Ogden CL, Flegal KM, Carroll MD, Johnson CL 2002 Prevalence and trends
in overweight among US children and adolescents, 1999 –2000. JAMA 288:
1728 –1732
2. Pinhas-Hamiel O, Dolan LM, Daniels SR, Standiford D, Khoury PR, Zeitler
P 1996 Increased incidence of non-insulin-dependent diabetes mellitus among
adolescents. J Pediatr 128:608 – 615
3. Goran MI, Ball GD, Cruz ML 2003 Obesity and risk of type 2 diabetes and
cardiovascular disease in children and adolescents. J Clin Endocrinol Metab
88:1417–1427
4. Schwartz RP 2003 Soft drinks taste good, but the calories count. J Pediatr
142:599 – 601
5. Schmitz MK, Jeffery RW 2000 Public health interventions for the prevention
and treatment of obesity. Med Clin North Am 84:491–512
6. Nader PR, National Institute of Child Health and Human Development
Study of Early Child Care and Youth Development Network 2003 Frequency
and intensity of activity of third-grade children in physical education. Arch
Pediatr Adolesc Med 157:185–190
7. Kimm SY, Glynn NW, Kriska AM, Barton BA, Kronsberg SS, Daniels SR,
Crawford PB, Sabry ZI, Liu K 2002 Decline in physical activity in black girls
and white girls during adolescence. N Engl J Med 347:709 –715
8. Stern MP 1995 Diabetes and cardiovascular disease: the “common soil” hypothesis. Diabetes 44:369 –374
9. Lawlor DA, Davey Smith G, Ebrahim S 2003 Life course influences on insulin
resistance: findings from the British Women’s Heart and Health Study. Diabetes Care 26:97–103
10. Liese AD, Mayer-Davis EJ, Haffner SM 1998 Development of the multiple
metabolic syndrome: an epidemiologic perspective. Epidemiol Rev 20:157–172
11. Invitti C, Guzzaloni G, Gilardini L, Morabito F, Viberti G 2003 Prevalence
and concomitants of glucose intolerance in European obese children and adolescents. Diabetes Care 26:118 –124
12. Sinha R, Fisch G, Teague B, Tamborlane WV, Banyas B, Allen K, Savoye M,
Rieger V, Taksali S, Barbetta G, Sherwin RS, Caprio S 2002 Prevalence of
impaired glucose tolerance among children and adolescents with marked
obesity.[Erratum (2002) 346:1756] N Engl J Med 346:802– 810
13. American Academy of Pediatrics Committee on Nutrition 2003 Policy statement: prevention of pediatric overweight and obesity. Pediatrics 112:424 – 430
14. Ahmad F, Kavey R, Kveselis D, Gaum W, Smith F 2001 Responses of nonobese white children to treadmill exercise. J Pediatr 139:284 –290
15. Symons JP, Sowers M-FR, Harlow SD 1997 Relationship of body composition
measures and menstrual cycle length. Ann Hum Biol 24:107–116
16. Schakel SF, Sievert YA, Buzzard IM 1988 Sources of data for developing and
maintaining a nutrient database. J Am Diet Assoc 88:1268 –1271
17. Hoffman RP, Vicini P, Sivitz WI, Cobelli C 2000 Pubertal adolescent malefemale differences in insulin sensitivity and glucose effectiveness determined
by the one compartment minimal model. Pediatr Res 48:384 –388
18. Lindstedt G, Lundberg PA, Lapidus L, Lundgren H, Bengtsson C, Bjorntorp
P 1991 Low sex-hormone-binding globulin concentration as independent risk
factor for development of NIDDM: 12-yr follow-up of population study of
women in Gothenburg, Sweden. Diabetes 40:123–128
19. Wilke TJ, Utley DJ 1987 Total testosterone, free-androgen index, calculated
free testosterone, and free testosterone by analog RIA compared in hirsute
women and in otherwise-normal women with altered sex-hormone-bindingglobulin. Clin Chem 33:1372–1375
20. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC
1985 Homeostasis model assessment: insulin resistance and ␤-cell function
from fasting plasma glucose and insulin concentrations in man. Diabetologia
28:412– 419
21. Bonora E, Kiechl S, Willeit J, Oberhollenzer F, Egger G, Targher G, Alberiche
M, Bonadonna R C, Muggeo M 1998 Prevalence of insulin resistance in
metabolic disorders: the Bruneck Study. Diabetes 47:1643–1649
22. Ibanez L, Potau N, Georgopoulos N, Prat N, Gussinye M, Carrascosa A 1995
Growth hormone, insulin-like growth factor-I axis, and insulin secretion in
hyperandrogenic adolescents. Fertil Steril 64:1113–1119
23. Food and Nutrition Board, Commission on Life Sciences, National Research
Council 1989 Recommended dietary allowances. 10th ed. Washington DC:
National Academy Press
24. Serdula MK, Collins ME, Williamson DF, Anda RF, Pamuk E, Byers TE 1993
Weight control practices of United states adolescents and adults. Ann Intern
Med 119:667– 671
854
J Clin Endocrinol Metab, February 2005, 90(2):849 – 854
25. Sothern MS 2001 Exercise as a modality in the treatment of childhood obesity.
Pediatr Clin North Am 48:995–1015
26. Goran MI, Treuth MS 2001 Energy expenditure, physical activity, and obesity
in children. Pediatr Clin North Am 48:931–953
27. Kawabe H, Murata K, Shibata H, Hirose H, Tsujioka M, Saito I, Saruta T 2000
Participation in school sports clubs and related effects on cardiovascular risk
factors in young males. Hypertens Res Clin Exp 23:227–232
28. Bouchard C, Tremblay A, Despres JP, Theriault G, Nadeau A, Lupien PJ,
Moorjani S, Prudhomme D, Fournier G 1994 The response to exercise with
constant energy intake in identical twins. Obes Res 2:400 – 410
29. Jakicic JM, Clark K, Coleman E, Donnelly JE, Foreyt J, Melanson E, Volek
J, Volpe SL 2001 American College of Sports Medicine position stand: appropriate intervention strategies for weight loss and prevention of weight
regain for adults. Med Sci Sports Exerc 33:2145–2156
30. Schoeller DA, Shay K, Kushner RF 1997 How much physical activity is
needed to minimize weight gain in previously obese women? Am J Clin Nutr
66:551–556
31. Andersen LB, Henckel P, Saltin B 1989 Risk factors for cardiovascular disease
in 16 –19-year-old teenagers. J Intern Med 225:157–163
32. Sirard JR, Pate RR 2001 Physical activity assessment in children and adolescents. Sports Med 31:439 – 454
33. Wallace TM, Matthews DR 2002 The assessment of insulin resistance in man.
Diabet Med 19:527–534
34. Quon MJ 2001 Limitations of the fasting glucose to insulin ratio as an index
of insulin sensitivity. J Clin Endocrinol Metab 86:4615– 4617
35. Fritsche A, Madaus A, Stefan N, Tschritter O, Maerker E, Teigeler A, Haring
Kasa-Vubu et al. • Insulin Resistance in Girls
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
H, Stumvoll M 2002 Relationships among age, proinsulin conversion, and
␤-cell function in nondiabetic humans. Diabetes 51:S234 –S239
Palmert M, Gordon C, Kartashov A, Legro R, Emans S, Dunaif A 2002
Screening for abnormal glucose tolerance in adolescents with polycystic ovary
syndrome. J Clin Endocrinal Metab 87:1017–1023
Travers SH, Jeffers BW, Bloch CA, Hill JO, Eckel RH 1995 Gender and Tanner
stage differences in body composition and insulin sensitivity in early pubertal
children. J Clin Endocrinol Metab 80:172–178
Hung T, Sievenpiper JL, Marchie A, Kendall CW, Jenkins DJ 2003 Fat versus
carbohydrate in insulin resistance, obesity, diabetes and cardiovascular disease. Curr Opin Clin Nutr Metab Care 6:165–176
Ludwig DS 2002 The glycemic index: physiological mechanisms relating to
obesity, diabetes, and cardiovascular disease. JAMA 287:2414 –2423
Anonymous 2002 You are what you eat. Lancet Oncology 3:517
Klein S, Fontana L, Young VL, Coggan A R, Kilo C, Patterson B W, Mohammed BS 2004 Absence of an effect of liposuction on insulin action and risk
factors for coronary heart disease. N Engl J Med 350:2549 –2557
Kuczmarski RJ, Flegal KM, Campbell SM, Johnson CL 1994 Increasing prevalence of overweight among US adults: the National Health and Nutrition
Examination Surveys, 1960 to 1991. JAMA 272:205–211
Wang G, Dietz WH 2002 Economic burden of obesity in youths aged 6 to 17
years: 1979 –1999. Pediatrics [Erratum (2002) 109:1195] 109:E81–1
Garcia VF, Langford L, Inge TH 2003 Application of laparoscopy for bariatric
surgery in adolescents. Curr Opin Pediatr 15:248 –255
Yanovski JA 2001 Intensive therapies for pediatric obesity. Pediatr Clin North
Am 48:1041–1053
JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the
endocrine community.