International Journal of Obesity (2004) 28, 1039–1047
& 2004 Nature Publishing Group All rights reserved 0307-0565/04 $30.00
www.nature.com/ijo
PAPER
Insulin resistance and weight gain in postmenopausal
women of diverse ethnic groups
BV Howard1*, L Adams-Campbell2, C Allen3,w, H Black4, M Passaro1, RJ Rodabough5, BL Rodriguez6,
M Safford7, VJ Stevens8 and LE Wagenknecht9
1
MedStar Research Institute, Washington, DC, USA; 2Howard University, Washington, DC, USA; 3University of Wisconsin,
Madison, WI, USA; 4Rush-Presbyterian St. Luke’s Medical Center, Chicago, IL, USA; 5Fred Hutchinson Cancer Research
Center, Seattle, WA, USA; 6Women’s Health Hawaii, Honolulu, HI, USA; 7Women’s Health Initiative, University of Medicine
and Dentistry of New Jersey, Newark, NJ, USA; 8Women’s Health Initiative, Kaiser Center for Health Research, Portland, OR,
USA; and 9Women’s Health Initiative, Southeast Region, Winston-Salem, NC, USA
OBJECTIVE: This study was conducted to examine the influence of insulin resistance on weight change in postmenopausal
women of various ethnic groups.
SUBJECTS: Data were obtained from 3389 women (60% White, 20% Black, 12% Hispanic, and 8% Asian/Pacific Islander), ages
50–79, enrolled in either the Women’s Health Initiative Clinical trial or Observational Study, whose blood samples were selected
randomly from the full cohort of 161 809 women for analyses.
MEASUREMENTS: Glucose, insulin, and lipids were measured on fasting serum samples drawn at baseline and after 3 y of
follow-up. Weight, height, waist circumference, and blood pressure were measured. Physical activity and energy intake were
assessed via questionnaire. Insulin resistance was estimated using the HOMA (homeostasis model) calculation.
RESULTS: Average age was 62 y, average BMI (body mass index) was 27.4 kg/m2, and average weight change was a gain of
0.4 kg in 3 y. In a multivariate analysis, insulin resistance and insulin concentrations were independent predictors of increases in
weight in White women (P ¼ 0.002 and 0.004, respectively) and in the combined group (P ¼ 0.027 and 0.039). For the whole
group, after adjustment for other covariates, those in the highest quartile of insulin resistance gained 0.4 kg in 3 y, whereas those
in the lowest quartile lost 0.06 kg. Similar trends were found for insulin resistance and weight gain in Hispanic and Asian/Pacific
Islander women, but they did not reach statistical significance. In Black women, no relation was seen between either insulin or
insulin resistance and weight change. A significant interaction between obesity and insulin resistance was observed (P ¼ 0.002
for White women and 0.032 for the whole group), so that there is weight gain with increasing insulin resistance in the leaner
women, but weight loss with increasing insulin resistance in the most obese.
CONCLUSION: Insulin resistance appears to be a predictor of weight gain in postmenopausal women, except for the most
obese women. The effect is more pronounced in women who have a lower BMI, and the effect was not seen in the Black women
who as a group had a higher BMI.
International Journal of Obesity (2004) 28, 1039–1047. doi:10.1038/sj.ijo.0802645
Keywords: insulin resistance; weight gain; postmenopausal women; diverse ethnic groups; women’s health initiative
Introduction
Obesity is a major health problem in the United States.
Weight gain occurs with age in both genders, and rates of
obesity are highest in postmenopausal women.1–3 Although
energy intake in excess of energy expenditure is the under-
*Correspondence: Dr BV Howard, MedStar Research Institute, 6495 New
Hampshire Ave., Ste. 201, Hyattsville, MD 20783, USA
E-mail: [email protected]
w
Deceased
Received 17 September 2003; revised 18 February 2004; accepted 22
February 2004
lying cause of weight gain, genetic and environmental
factors influence this energy imbalance.1–4 Thus, it is
important to better define the risk factors leading to weight
gain in various age groups.
Considerable research has been focused on the relationship between insulin resistance and weight gain. This focus
began with the Neel hypothesis, in which it was proposed
that insulin resistance might provide a survival advantage to
populations subject to limited availability of food.4,5 According to this hypothesis, insulin-resistant individuals living a
Westernized lifestyle, with a continual abundance of food,
are predisposed to weight gain. The current analysis is
Insulin resistance and weight gain in postmenopausal women
BV Howard et al
1040
intended to explore the hypothesis that insulin resistance
predicts weight gain in postmenopausal women of differing
ethnic groups.
Studies of children6 and young adults7,8 have found an
association between insulin resistance, as measured by
fasting insulin, and weight gain. In the Normative Aging
Study, researchers found that in White adults a higher
insulin concentration predicted increases in weight,9 and in
Chinese men, researchers found that higher insulin levels
predicted weight gain.10 Other researchers examining relations between insulin resistance and weight gain in adults,
however, have suggested that insulin sensitivity or low
insulin levels could predispose people to gain weight. This
predisposition was first described in Pima Indian adults11
and has been observed in Mexican Americans and nonHispanic Whites in the San Antonio12 and San Luis Valley
Heart studies,13 in middle-aged Whites and Blacks in the
ARIC cohort,14 and in older Whites in the Rancho Bernardo
Study.15 Some of these investigators postulated that ethnic
differences might exist in the relationship between insulin
sensitivity and weight gain, consistent with the known
ethnic differences in prevalence rates of insulin resistance.
The Women’s Health Initiative (WHI) affords an opportunity to examine the influence of insulin resistance on weight
gain in a large cohort of postmenopausal women of diverse
ethnic groups.16 In the current analysis, the homeostasis
(HOMA) model is used to examine the influence of insulin
resistance on weight gain in White, Black, Hispanic, and
Asian/Pacific Islander postmenopausal women. The ethnic
and geographic diversity and careful characterization of
women in the WHI afford an opportunity to assess the
relationship between insulin resistance and weight change
in a multiethnic group of postmenopausal women.
Materials and methods
Participants
The current analysis was conducted with data from participants enrolled in the WHI Observational Study (OS) and
clinical trials of hormone replacement therapy and diet
modification; all participants gave informed consent. Institutional review committees from all 40 clinical centers
approved the study, and the study conformed to the ethical
guidelines of the 1975 declaration of Helsinki.
The WHI is a continuing study of major health issues in
postmenopausal women. Descriptions of the overall study
and experimental designs have been published previously.17
A total of 161 809 women, ranging in age from 50 to 79 y old,
were enrolled in the observational study or one or both of
the clinical trials. They were recruited at 40 clinical centers
throughout the United States. Exclusion criteria for the
observational study included medical or personal conditions
that would preclude a 3-y follow-up visit. Additional
exclusion criteria for the diet trial included previous
diagnosis of breast or colon cancer, current use of oral
International Journal of Obesity
corticosteroids, gastrointestinal conditions that required a
high fiber diet, Type I diabetes, and current consumption of
a diet containing less than 32% of energy from fat.
Additional exclusion criteria for the hormone trial included
an acute cardiovascular event during the previous 6 months,
current use of hormone replacement therapy or oral
corticosteroids, history of breast cancer, and history of
invasive cancer during the past 10 y. For all components,
mental illness or other factors that would preclude informed
consent also were exclusion criteria.
Sample selection
Six percent of the women enrolled in the hormone
replacement therapy (HRT) trial or the dietary trial and 600
women enrolled in the observational study were selected
randomly to have their blood samples analyzed for glucose,
insulin, lipids, and homeostatic factors. The random sampling procedure was stratified by age, clinical center, and
hysterectomy status. A larger proportion of samples from
minority women were selected to increase the sample size of
Black, Hispanic, and Asian-Pacific women. For the current
analysis, women with diabetes, defined as either a prior
diagnosis of the disease or fasting blood glucose level greater
than or equal to 7.0 mmol/l (126 mg/dl) or those taking oral
hypoglycemic agents or insulin, were excluded. Also excluded were women who did not attend the 3-y follow-up
visit, those who quit smoking between the baseline and the
follow-up visit, those who were diagnosed with cancer
between the baseline and the follow-up visit, and participants taking corticosteroids or weight loss medication, such
as amphetamines, sibutramine, or orlistat.
Clinical examinations
The baseline exams for the WHI included physical measurements, fasting blood samples, and interviewer and selfadministered questionnaires. Clinical measurements included weight, height, and blood pressure. Body mass index
(BMI) was calculated as weight in kilograms divided by
height in meters squared (kg/m2). Waist circumference was
determined with the participants standing, with measurement at the natural waist or the narrowest part of the torso.
Measurement was made with the participants clothed, but
belts or heavy objects were removed from the waist area.
Participants were asked to refrain from taking nonsteroidal,
anti-inflammatory drugs or aspirin for 48 h prior to their
assessment visit. A blood sample was collected after a 12-h
fast. Participants also were asked to refrain from vigorous
physical activity for at least 12 h before the blood draw and
to refrain from smoking for at least 1 h. Aliquots of plasma
and serum were isolated from the blood samples and frozen
at 701C within 2 h after being drawn. Self-reported
questionnaires included assessment of demographic factors,
medical and reproductive history, and health habits. Information on the dose of all current medications, including
Insulin resistance and weight gain in postmenopausal women
BV Howard et al
1041
hormones, was ascertained by the interviewers who administered the questionnaires. Use of vitamins and supplements also was assessed. Physical activity was assessed via a
standardized questionnaire, and dietary intake was measured
using the WHI Food Frequency Questionnaire.18
Blood specimens were analyzed by medical research
laboratory staff (Highland Heights-Kentucky). Glucose was
measured using the glucose oxidase method19,20 on the
Hitachi 747 analyzer; coefficient of variation was less than
2%. Insulin was measured using a step-wise ELISA procedure;
coefficient of variation ranged from 3.2 to 9.5%.21 The
HOMA (homeostasis model) was used to estimate insulin
sensitivity from fasting insulin and glucose values.22 This
estimate has been found to correlate well with measures of
insulin resistance obtained by the euglycemic clamp technique,23 and is now commonly used in large population
studies where metabolic measures are not tenable. Using this
model, insulin resistance was calculated as [FI (mIU/ml) * FG
(mmol/1)/(0.0555*18)]/22.5.
Statistical analysis
All analyses were restricted to participants in whom baseline
blood draws had been analyzed. This restriction resulted in
2031 White, 681 Black, 405 Hispanic, 272 Asian/Pacific
Islander, and 61 American Indian women. Analyses were not
performed on American Indian women because of the small
number available. Distributions for baseline variables by
ethnicity are displayed in the form of medians and ranges for
continuous variables and frequencies for categorical variables. Associations of median values for each characteristic
with ethnicity were tested using Brown–Mood multisample
median tests, and associations of categorical variables with
ethnicity were tested using w2 tests. These tests were
performed to determine whether the distribution of each
characteristic differed across levels of ethnicity. Two multivariate models were developed to examine the separate
relationships of fasting insulin (mIU/ml) and insulin resistance (HOMA-IR) with 3-y weight change within each ethnic
group and for the group as a whole. Each model was
performed on complete case data and adjusted for age at
screening, BMI (kg/m2), waist circumference (cm), categories
of physical activity, total daily energy intake (kcal), and
fasting glucose (mmol/l), as well as for the interaction
between BMI and either fasting insulin or HOMA-IR. BMI,
waist circumference, energy intake, and fasting insulin or
HOMA were not normally distributed and were log-transformed for all multivariate analyses. Least-squares means
were computed for weight change from the resultant multivariate models. Distributions of variables prior to restricting
the data set were all similar to the complete case.
Results
Metabolic variables for each ethnic group are shown in
Table 1. P-values indicate whether the distribution of a given
characteristic differs by ethnicity. In each ethnic group, there
is a wide range of HOMA insulin resistance, but median
values indicate high prevalence of insulin resistance in each
ethnic group. Median insulin concentrations and insulin
resistance differ significantly across ethnic groups
(Po0.0001), with apparently higher values in Black and
Hispanic women. Women in all four ethnic groups were
overweight, with the median BMI for Blacks being in the
obese range and for White and Hispanic women in the
overweight range. Only Asian/Pacific Islander women had a
median BMI less than 25 kg/m2.
Age, BMI, glucose and insulin concentrations, and waist
circumference all showed significant univariate associations
with insulin resistance (data not shown). Final regression
models examining factors that predict weight change were
run, including either HOMA-IR (see Table 2) or insulin
concentration (see Table 3) for each ethnic group and for the
group as a whole. In all models, age was a significant
negative predictor of weight gain in each ethnic group. In all
models, physical activity and total energy intake showed no
meaningful relationships with weight change. Insulin resistance was a significant predictor of weight gain in White
women and in the combined group (Table 2). A similar effect
was found for weight gain in White women and in the total
group with increasing insulin concentrations (Table 3). The
effect of body size on weight gain was dependent on insulin
resistance or insulin level, as evidenced by the significant
interaction terms for BMI IR or BMI Insulin observed in
White women and in the combined group (Tables 2 and 3).
The main effect of BMI was not statistically significant in the
overall models when including the interaction with insulin
resistance or insulin, but it was significant without the
interaction term (data not shown). Thus, we included both
the main effect and interaction term in the final models to
more fully control for the effect of BMI on weight change. In
Hispanic and Asian/Pacific Islander women, regression
coefficients for insulin or insulin resistance and weight
change and the interaction terms with BMI were similar in
magnitude to those in White women, but did not reach
statistical significance, presumably because of the smaller
sample size (Tables 2 and 3). On the other hand, in Black
women, the coefficients for the relations between IR or
insulin and weight change tended to be negative.
To further illustrate the interaction between BMI and
insulin resistance, an age-adjusted multivariate model was
developed, including the main effect of BMI, and the
resulting relation was plotted for various levels of BMI
(Figure 1) using data from the combined group. Although
the main effect of BMI was not significant when including
the interaction of BMI with insulin resistance in the overall
multivariate model, we again included it in this model
to more fully control for the effect of BMI. The results
show increasing weight gain with increasing HOMA-IR in
the leaner BMI ranges, and weight loss with increasing
HOMA-IR in the higher BMI ranges (P for BMI IR
interaction ¼ 0.006).
International Journal of Obesity
1042
International Journal of Obesity
Baseline characteristics by ethnicity
White
N
Median
Min
Black
Max
N
Median
Min
Hispanic
Max
N
Median
Min
Asian/Pacific Islander
Max
N
Median
Min
Max
Total
N
Median
Min
Max
Age at screening (y)**
2031
63.0
50
79
681
60.0
50
79
405
60.0
50
79
272
63.0
50
78
3389
62.0
50
79
BMI (kg/m2)**
2023
26.9
15.9
69.4 676
30.7
18.4
57.7 402
27.9
16.7
46.4 272
24.5
17.4
42.3 3373
27.4
15.9
69.4
Waist (cm)**
2024
85.0
54.0
174.5 680
91.0
37.5
144.2 402
85.5
64.5
121.3 270
78.4
58.5
114.5 3376
85.5
37.5
174.5
Weight (kg)**
2031
71.0
40.6
148.0 681
80.6
45.0
147.5 405
69.6
41.2
116.4 272
59.1
37.0
111.0 3389
71.5
37.0
148.0
Weight change (AV3baseline)* 2031
0.2 37.9
41.5 681
0.5 48.2
29.4 405
1.0 37.6
30.9 272
0.3 35.0
9.6 3389
0.4 48.2
41.5
Physical activity (METs)**
1808
7.5
0.0
90.2 668
4.5
0.0
119.0 397
6.3
0.0
76.2 271
8.9
0.0
90.2 3144
7.0
0.0
119.0
Dietary energy (kcal)**
2005 1599.9 603.2 4987.4 651 1493.8 608.1 4798.2 393 1513.0 603.1 4971.5 261 1371.0 600.5 4956.1 3310 1559.3 600.5 4987.4
Fasting glucose (mmol/l (mg/dl))* 2031 5.1 (92.0) 3.7 (66) 6.9 (124) 681 5.2 (93.0) 3.2 (58) 6.9 (124) 405 5.2 (94.0) 3.6 (64) 6.9 (124) 272 5.3 (95.0) 4.0 (72) 6.8 (123) 3389 5.2 (93.0) 3.2 (58) 6.9 (124)
F Insulin (mIU/ml)**
1949
8.7
3.0
53.4 670
11.2
3.3
171.1 396
10.2
3.5
61.5 265
8.8
3.5
31.7 3280
9.4
3.0
171.1
HOMA insulin resistance**
1949
2.0
0.7
15.9 670
2.6
0.6
43.1 396
2.3
0.7
17.3 265
2.0
0.7
8.1 3280
2.2
0.6
43.1
N
%
N
%
N
%
N
%
N
%
Education**
0–8 y
Some high school
High school diploma/GED
School after high school
College degree or higher
14
53
442
750
762
0.7
2.6
21.9
37.1
37.7
19
52
85
274
247
2.8
7.7
12.6
40.5
36.5
55
41
70
128
105
13.8
10.3
17.5
32.1
26.3
3
8
37
95
129
1.1
2.9
13.6
34.9
47.4
91
154
634
1247
1243
2.7
4.6
18.8
37.0
36.9
HRT usage status**
Never used
Past user
Current user
999
360
669
49.3
17.8
33.0
393
135
152
57.8
19.9
22.4
220
59
126
54.3
14.6
31.1
126
53
93
46.3
19.5
34.2
1738
607
1040
51.3
17.9
30.7
P-values for association of median values with ethnicity were computed from Brown–Mood tests. P-values for categorical variables were computed from w2 tests. *P-valueo0.01. **P-valueo0.0001.
Insulin resistance and weight gain in postmenopausal women
BV Howard et al
Table 1
Insulin resistance and weight gain in postmenopausal women
BV Howard et al
1043
Table 2
Regression model for insulin resistance and other variables predicting weight change, by ethnicity
White
Coefficient
P
Physical activity (METs) (vs none)
40–o6
6–o15
15+
0.18
0.18
0.40
0.14
3.2 (0.02)
Log total calories (kcal/day)
Fasting glucose (mmol/l (mg/dl))
Insulin resistance
Log HOMA insulin resistance
11.48
Log BMI Log HOMA insulin resistance 3.34
Model R2
Hispanic
Coefficient
P
o0.0001 0.09
0.156 2.72
0.275 0.58
0.09
2.08
1.61
Age at screening (y)
Log BMI (kg/m2)
Log waist (cm)
Black
Asian/Pacific Islander
Coefficient
P
0.011 0.14
0.386 3.29
0.854 0.37
0.0001 0.07
0.330 4.38
0.906
5.29
0.385
0.097
0.217
0.391
0.882
0.102
0.635
0.629
0.291
0.55
1.16
0.88
0.660
0.233
0.64
0.228 1.06
0.055
0.38 (0.02) 0.429 0.09 (0.01) 0.859
0.002
0.002
0.048
6.87
1.82
Coefficient
0.58
0.10
1.17
0.373 9.70
0.422 2.90
0.025
0.79
0.68
0.95
0.265
0.260
0.060
P
Total
Coefficient
o0.0001
0.173
0.500
0.091 0.10
0.227 1.59
0.221 0.81
0.357
0.423
0.257
P
0.27
0.32
0.51
0.331
0.259
0.077
0.56
0.404 0.01
0.79 (0.04) 0.171 0.27 (0.01)
0.978
0.182
13.68
4.31
0.027
0.032
0.034
0.141 6.29
0.136 1.83
0.102
Multivariate regression models included all tested variables. They also included an indicator variable for the women who were randomized to the intervention group
of the WHI diet trial and were adjusted for hormone use.
Table 3
Regression model for insulin concentrations and other variables predicting weight change, by ethnicity
White
Coefficient
Age at screening (y)
Log BMI (kg/m2)
Log waist (cm)
Physical activity (METs) (vs none)
40–o6
6–o15
15+
0.09
2.84
1.60
0.18
0.18
0.40
Black
P
Coefficient
o0.0001 0.09
0.342 5.56
0.278 0.56
Hispanic
P
Coefficient
Asian/Pacific Islander
P
0.011 0.14
0.393
8.97
0.860 0.38
0.0001 0.07
0.202
1.97
0.905
5.20
0.387
0.098
0.217
0.394
0.877
0.103
0.630
0.624
0.289
0.55
1.16
0.88
Log total calories (kcal/day)
0.15
Fasting glucose (mmol/l (mg/dl)) 0.25 (0.01)
0.654
0.312
0.64
0.228 1.06
0.055
0.26 (0.01) 0.547 0.11 (0.01) 0.820
Insulin
Log insulin
Log BMI log insulin
Model R2
0.004
0.006
0.047
11.46
3.34
7.11
1.89
Coefficient
0.57
0.11
1.17
0.404 11.74
0.451 3.50
0.025
0.213
0.210
0.061
0.78
0.67
0.93
P
Total
Coefficient
o0.0001
0.603
0.506
0.094 0.10
0.792
1.22
0.229
0.80
0.361
0.429
0.267
P
0.27
0.32
0.51
0.329
0.257
0.076
0.55
0.409 0.01
0.83 (0.05) 0.116 0.24 (0.01)
0.975
0.199
13.75
4.33
0.039
0.044
0.034
0.178
6.45
0.173 1.88
0.101
Multivariate regression models included all tested variables. They also included an indicator variable for the women who were randomized to the intervention group
of the WHI diet trial and were adjusted for hormone use.
To evaluate other possible confounders in this analysis, the
analyses were repeated, eliminating those women who were
taking antidepressant or thyroid replacement medications,
women who were undergoing the low-fat diet intervention,
or women who developed CVD or diabetes within the 3-y
period; the results were similar. Also, prior estrogen use was
not related to weight change (data not shown).
To quantify the magnitude of the impact of insulin
resistance on weight change, change in weight was computed by quartiles of HOMA-IR or insulin concentrations
(see Table 4); this was done only for the whole group and
White women, because of the smaller numbers in the other
three ethnic groups. After adjustment for covariates, those in
the highest quartile of insulin resistance showed approximately 0.4 kg of weight gain over the average 3-y period,
while those in the most insulin-sensitive quartile showed
weight loss (P for trends ¼ 0.120 and 0.070, respectively).
Discussion
In this cohort of postmenopausal women, those with insulin
resistance were more likely to have subsequent weight gain,
but there was an interaction between obesity and insulin
resistance. Weight gain was more likely in leaner women
with increased insulin resistance than in heavier women
International Journal of Obesity
Insulin resistance and weight gain in postmenopausal women
BV Howard et al
1044
3 -Year Weight Change
2.5
BMI = 23
BMI = 28
BMI = 33
BMI = 38
BMI = 43
1.5
0.5
-0.5
-1.5
-2.5
0
1
2
3
4
5
6
7
8
HOMA IR
Weight change = 9.33 - 0.09*age - 1.02*ln(BMI) + 7.47*ln(IR) - 2.23*ln(BMI)*ln(IR)
Figure 1 Effect of BMI (kg/m2) and HOMA insulin resistance on 3-y weight change (multiple regression lines for a 63-y-old participant).
Table 4 Adjusted mean 3-year weight changes in White women and the
whole group, by quartiles of fasting insulin and insulin resistance
White
Total
Adjusted mean
weight change (kg)
Adjusted mean
weight change (kg)
Insulin (mIU/ml)
o6.8
6.8–o9.4
9.4–o13.3
13.3+
P-value for trend
0.12
0.18
0.14
0.53
0.067
0.09
0.09
0.15
0.42
0.147
Insulin resistance (HOMA)
o1.50
1.50–o2.17
2.17–o3.21
3.21+
P-value for trend
0.24
0.13
0.01
0.48
0.070
0.06
0.03
0.26
0.40
0.120
with increased insulin resistance. This trend was not seen in
the Black women, whose mean BMI as a group was higher.
Some researchers have shown a positive association
between insulin resistance or fasting insulin concentrations
and weight gain in adults. This includes the analysis of the
longitudinal data from the Normative Aging Study;9 in this
cohort ages and BMI were similar to those of the WHI. Hodge
et al.10 showed that in Chinese in Mauritius, insulin
resistance predicted weight gain. Further evidence that
insulin resistance predicts adiposity was presented by Gould
et al,24 who showed that insulin resistance predicts central
adiposity, and by Boyko et al,25 who found that increases in
fasting insulin predicted increases in visceral fat in Japanese
men. Studies of children6,7,26 and of young adults in the
International Journal of Obesity
CARDIA Study14 also showed that insulin resistance predicted weight gain.
Other researchers, however, in some studies of adults, have
reported a significant association between insulin concentrations or insulin resistance and weight loss. Some possible
distinguishing features separate the current cohort from
those of other studies; most of the others examined younger
adults and in some the participants were quite obese. The
Black and White adults in the ARIC Study8 averaged age 54
with a BMI of 27 kg/m2. In the study of Pima Indian adults,11
the average age was 25 y and average BMI was 34. Similarly,
in a study of Mexican Americans and non-Hispanic Whites,
the individuals ranged from 25 to 64 y, with a median BMI of
approximately 26.12 In the analysis of White adults in the
Rancho Bernardo Study, the mean age was similar to women
in the WHI and the average baseline BMI was 24.7. The data
from our study provide insight concerning the inconsistencies among studies. We found that, although there was no
significant interaction of insulin resistance with age, there
was one with BMI, so that the effect of insulin resistance on
weight gain was more pronounced in leaner women. Also, in
most of the previous studies, data on energy intake, physical
activity, and education were not available.
Ethnic diversity has been invoked in several earlier studies
to explain the inconsistencies in findings.27 The current data
indicate that insulin resistance appears to be greater in
Hispanic and African American women compared with
White women.27 In Hispanic and Asian/Pacific Islander
women, trends were found between insulin resistance and
weight change, which did not reach statistical significance.
On the other hand, there appeared to be no relation between
insulin resistance and weight change in the Black women in
this study. This possible ethnic difference may be related to
the higher BMI of Black women (average of 31). However,
Insulin resistance and weight gain in postmenopausal women
BV Howard et al
1045
the relatively small number and nonrandom selection of
subjects make the observation tenuous; more studies of
insulin resistance and weight are needed in diverse population groups of all ages.
Insulin resistance often is related to increases in visceral
adiposity, which is commonly observed with aging.28 In this
study, however, waist circumference was not shown to be a
correlate of weight gain or a significant predictor in the
multivariate analysis. Adjusting for baseline waist circumference had no effect on the relationship between insulin
resistance and weight gain.
Although it is established that obesity and insulin
resistance are associated, the causative relationship between
insulin resistance and obesity has been debated. Explanations of the Neel hypothesis4 focus on the concept that
resistance to insulin-mediated glucose disposal in peripheral
tissues, such as muscle, would lead to energy deposition in
adipose tissue. More recently, it has been proposed29 that
insulin acts in the brain to provide feedback inhibition of
food intake, and that insulin resistance in the central
nervous system might promote increased energy intake
and storage.30 If this latter construct is operative, then in
individuals who are not already overweight, peripheral
insulin resistance could be expected to promote weight gain.
On the other hand, in obese individuals, the effect of insulin
resistance may be blunted in some way.
Another hypothesis that must be considered in evaluating
the current observations of the association between insulin
resistance and weight gain concerns the role of inflammation. A recent report31 shows that fibrinogen, white blood
cell count, factor VIII, and Von Willebrand factor all are
positive predictors of weight gain in middle age adults. All of
these inflammatory markers are known to be elevated in
insulin resistance. Thus, chronic inflammation, which
increases with age, may lead to insulin resistance and weight
gain in postmenopausal women. The observation that the
effect of insulin resistance is more pronounced in lean
individuals may suggest that inflammatory mediators produced by adipocytes may in part mediate the relationship.
A number of potential biases must be considered in this
analysis. There were a number of medications that could
confound the analysis of changes in weight. Women who
were on corticosteroids were eliminated from the analysis.
The analysis was repeated after eliminating women on
thyroid medication or antidepressants, but those changes
did not influence the conclusions. Women who developed
diabetes during the 3-y follow-up also were eliminated, with
no effect on the results. A few women reported smoking
cessation between baseline and 3-y follow-up, and these
women also were dropped from the analysis. Some controversy exists about the role of postmenopausal estrogen
use and weight change in women; however, estrogen use had
no effect in our analysis.
Finally, no significant relationships were observed between
either energy intake or physical activity and weight change.
Self-reported data on both of these parameters are imprecise.
Given that the average weight change during 3 y was small in
most cases, it is possible that neither of these measures has
the precision to detect the small amount of energy or
physical activity imbalance that would be responsible for the
average weight changes.
In conclusion, overweight and obesity are major health
problems in postmenopausal women. In this cohort of
women averaging 62 y of age, slight increases were found
in weight on average during a 3-y period. Over a wide range
of insulin resistance, insulin resistance appeared to be a
predictor of weight gain in White women, with a similar
trend in Hispanic and Asian/Pacific Islander women. This
effect was more pronounced in leaner women, and was not
observed in Black women. All women with high levels of
insulin resistance, because they are at risk for the metabolic
syndrome and diabetes, should receive increased encouragement to implement lifestyle changes to maintain or reduce
insulin resistance.
Acknowledgements
The Women’s Health Initiative program is funded by the
National Heart, Lung and Blood Institute, U.S. Department
of Health and Human Services.
References
1 Centers for Disease Control and Prevention. Prevalence of overweight and obesity among adults: United States, 1999, Centers for
Disease Control and Prevention U.S. Dept. of Health and Human
Services: Hyattsville, MD; 1999.
2 Clinical Guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: executive summary.
Expert Panel on the identification, evaluation, and treatment of
overweight in adults. Am J Clin Nutr 1998; 68: 899–917.
3 Mokdad AH, Serdula MK, Dietz WH, Bowman BA, Marks JS,
Koplan JP. The spread of the obesity epidemic in the United
States, 1991–1998. JAMA 1999; 282: 1519–1522.
4 Neel JV. Diabetes mellitus: a ‘thrifty’ genotype rendered detrimental by ‘progress’. Am J Hum Genet 1962; 14: 353–362.
5 Mayer-Davis EJ, Kirkner GJ, Karter AJ, Zaccaro DJ. Metabolic
predictors of five-year change in weight and waist circumference
in a tri-ethnic population: the Insulin Resistance Atherosclerosis
Study. Am J Epidemiol 2003; 157: 592–601.
6 Odeleye OE, de Courten M, Pettit DJ, Ravussin E. Fasting
hyperinsulinemia is a predictor of increased body weight
gain and obesity in Pima Indian children. Diabetes 1997; 46:
1341–1345.
7 Srinivasan SR, Myers L, Berenson GS. Temporal association
between obesity and hyperinsulinemia in children, adolescents,
and young adults: the Bogalusa Heart Study. Metabolism 1999; 48:
928–934.
8 The ARIC Investigators. The atherosclerosis risk in communities
(ARIC) study: design and objectives. Am J Epidemiol 1989; 129:
687–702.
9 Lazarus R, Sparrow D, Weiss S. Temporal relations between
obesity and insulin: longitudinal data from the normative aging
study. Am J Epidemiol 1998; 147: 173–179.
10 Hodge AM, Dowse GK, Alberti KG, Gareeboo H, Tuomilehto J,
Zimmet PZ. Relationship of insulin resistance to weight gain in
nondiabetic Asian Indian, Creol, and Chinese Mauritians.
Metabolism 1996; 45: 627–633.
International Journal of Obesity
Insulin resistance and weight gain in postmenopausal women
BV Howard et al
1046
11 Swinburn BA, Nyombe BL, Saad MF, Zurlo F, Raz I, Knowler WC,
Lillioja S, Bogardus C, Ravussin E. Insulin resistance associated
with lower rates of weight gain in Pima Indians. J. Clin. Invest
1991; 88: 168–173.
12 Valdez R, Mitchell BD, Haffner SM, Hazuda HP, Morales PA,
Monterrosa A, Stern MP. Predictors of weight change in a biethnic population: the San Antonio heart study. Int J Obes 1994;
18: 85–91.
13 Hoag S, Marshall JA, Jones RE, Hamman RF. High fasting insulin
levels associated with lower rates of weight gain in persons with
normal glucose tolerance: the San Luis Valley diabetes study. Int J
Obes Relat Metab Disord 1995; 19: 175–180.
14 Folsom AR, Vitelli LL, Lewis CE et al. Is fasting insulin
concentration inversely associated with rate of weight gain?
Contrasting findings from the CARDIA and ARIC study cohorts.
Int J Obes Relat Metab Disord 1998; 22: 48–54.
15 Wedick NM, Mayer-Davis EJ, Wingard DL et al. Insulin resistance
precedes weight loss in adults without diabetes: the Rancho
Bernardo study. Am J Epidemiol 2001; 153: 1199–1205.
16 Howard BV, The Women’s Health Initiative Study Group. Design
of the Women’s Health Initiative clinical trial and observational
study. Controlled Clin Trials 1998. 61–109.
17 Anonymous. Design of the Women’s Health Initiative clinical
trial and observational study. The Women’s Health Initiative
Study Group. Controlled Clin Trials 1998; 19: 61–109.
18 Howard BV. Diabetes in postmenopausal women. Women’s
Health Initiative Grant Application. MedStar Research Institute:
Washington, DC; 2000.
19 Schmidt FH In: Bergmeyer Hu (ed). Methods of enzymatic
analysis. 2nd edn. Academic Press: New York, NY; 1974.
p 1196.
20 Peterson JI, Young DS. Evaluation of the hexokinase-glucose-6phosphate dehydrogenase method of determination of glucose in
urine. Ann Biochem 1968; 23: 301–316.
21 Tietz NW. Fundamentals of clinical chemistry., 3rd edn.
Philadelphia, PA: WB Saunders Co.; 1987.
22 Ivandic A, Prpic-Krizevac I, Bozic D, Barbir A, Peljhan V, Balog Z,
Glasnovic M. Insulin resistance and androgens in healthy women
with different body fat distributions. Wien Klin Wochenschr 2002;
114: 321–326.
23 Resnick HE, Bergman RN, Henderson RJ, Nez-Henderson P,
Howard BV. Utility of a surrogate measure of insulin resistance
in American Indians: the Strong Heart Study. Ethnicity and
Disease, in press.
24 Gould AJ, Williams DEM, Byrne CD, Hales CN, Wareham NJ.
Prospective cohort study of the relationship of markers of
insulin resistance and secretion with weight gain and changes
in regional adiposity. Int J Obes Relat Metab Disord 1999; 23:
1256–1261.
25 Boyko EJ, Leonett DL, Bergstrom RW, Newell-Morris L, Fujimoto
WY. Low insulin secretion and high fasting insulin and C-peptide
levels predict increased visceral adiposity: 5-year follow-up
among initially nondiabetic Japanese-American men. Diabetes
1996; 45: 1010–1015.
26 Travers SH, Jeffers BW, Eckel RH. Insulin resistance during
puberty and future fat accumulation. J Clin Endocrinol Metab
2002; 87: 3814–3818.
27 Haffner SM, Howard G, Mayer E, Bergman RN, Savage PJ,
Rewers M et al. Insulin sensitivity and acute insulin response in
African-Americans, non-Hispanic Whites, and Hispanics with
NIDDM. The Insulin Resistance and Atherosclerosis Study.
Diabetes 1997; 46: 63–69.
28 Cefalu WT, Wang ZQ, Werbel S, Bell-Farrow A, Crouse III JR,
Hinson WH, Terry JG, Anderson R. Contribution of visceral
fat mass to the insulin resistance of aging. Metabolism 1995; 44:
954–959.
29 Porte D, Seeley RJ, Woods SC, Baskin DG, Figlewicz DP,
Schwartz MW. Obesity, diabetes and the central nervous system.
Diabetologia 1998; 41: 863–881.
International Journal of Obesity
30 Schwartz MW, Boyko EJ, Kahn SE, Ravussin E, Bogardus C.
Reduced insulin secretion: an independent predictor of body
weight gain. J Clin Endocrinol Metab 1995; 80: 1571–1576.
31 Duncan BB, Schmidt MI, Chambless LE, Folsom AR, Carpenter M,
Heiss G. Fibrinogen, other putative markers of inflammation, and
weight gain in middle-aged adultsFthe ARIC study. Obes Res
2000; 4: 279–286.
Appendix A: Long list of WHI investigators
Program Office: (National Heart, Lung, and Blood Institute,
Bethesda, MD)
Barbara Alving, Jacques Rossouw, Linda Pottern, Shari
Ludlam, Joan McGowan.
Clinical Coordinating Center: (Fred Hutchinson Cancer
Research Center, Seattle, WA) Ross Prentice, Garnet Anderson, Andrea LaCroix, Ruth Patterson, Anne McTiernan,
Barbara Cochrane, Julie Hunt, Lesley Tinker, Charles Kooperberg, Martin McIntosh, C. Y. Wang, Chu Chen, Deborah
Bowen, Alan Kristal, Janet Stanford, Nicole Urban, Noel
Weiss, Emily White; (Wake Forest University School of
Medicine, Winston-Salem, NC) Sally Shumaker, Pentti
Rautaharju, Ronald Prineas, Michelle Naughton; (Medical
Research Labs, Highland Heights, KY) Evan Stein, Peter
Laskarzewski (University of California at San Francisco, San
Francisco, CA) Steven Cummings, Michael Nevitt, Maurice
Dockrell (University of Minnesota, Minneapolis, MN) Lisa
Harnack; (McKesson BioServices, Rockville, MD) Frank
Cammarata, Steve Lindenfelser (University of Washington,
Seattle, WA) Bruce Psaty, Susan Heckbert.
Clinical Centers: (Albert Einstein College of Medicine,
Bronx, NY) Sylvia Wassertheil-Smoller, William Frishman,
Judith Wylie-Rosett, David Barad, Ruth Freeman; (Baylor
College of Medicine, Houston, TX) Jennifer Hays, Ronald
Young, Jill Anderson, Sandy Lithgow, Paul Bray; (Brigham
and Women’s Hospital, Harvard Medical School, Boston,
MA) JoAnn Manson, Julie Buring, J. Michael Gaziano,
Kathryn Rexrode, Claudia Chae; (Brown University, Providence, RI) Annlouise R. Assaf, Carol Wheeler, Charles Eaton,
Michelle Cyr; (Emory University, Atlanta, GA) Lawrence
Phillips, Margaret Pedersen, Ora Strickland, Margaret Huber,
Vivian Porter; (Fred Hutchinson Cancer Research Center,
Seattle, WA) Shirley A.A. Beresford, Vicky M. Taylor, Nancy F.
Woods, Maureen Henderson, Mark Kestin; (George Washington University, Washington, DC) Judith Hsia, Nancy Gaba,
Joao Ascensao; (Harbor-UCLA Research and Education
Institute, Torrance, CA) Rowan Chlebowski, Robert Detrano,
Anita Nelson, James Heiner, John Marshall; (Kaiser Permanente Center for Health Research, Portland, OR) Cheryl
Ritenbaugh, Barbara Valanis, Patricia Elmer, Victor Stevens,
Njeri Karanja; (Kaiser Permanente Division of Research,
Oakland, CA) Bette Caan, Stephen Sidney, Geri Bailey Jane
Hirata; (Medical College of Wisconsin, Milwaukee, WI) Jane
Morley Kotchen, Vanessa Barnabei, Theodore A. Kotchen,
Mary Ann C. Gilligan, Joan Neuner; (MedStar Research
Institute/Howard University, Washington, DC) Barbara V.
Howard, Lucile Adams-Campbell, Maureen Passaro,
Insulin resistance and weight gain in postmenopausal women
BV Howard et al
1047
Monique Rainford, Tanya Agurs-Collins; (Northwestern
University, Chicago/Evanston, IL) Linda Van Horn, Philip
Greenland, Janardan Khandekar, Kiang Liu, Carol Rosenberg;
(Rush-Presbyterian St. Luke’s Medical Center, Chicago, IL)
Henry Black, Lynda Powell, Ellen Mason; (Stanford Center
for Research in Disease Prevention, Stanford University,
Stanford, CA) Marcia L. Stefanick, Mark A. Hlatky, Bertha
Chen, Randall S. Stafford, Linda C. Giudice; (State University
of New York at Stony Brook, Stony Brook, NY) Dorothy Lane,
Iris Granek, William Lawson, Gabriel San Roman, Catherine
Messina; (The Ohio State University, Columbus, OH)
Rebecca Jackson, Randall Harris, Electra Paskett, W. Jerry
Mysiw, Michael Blumenfeld; (University of Alabama at
Birmingham, Birmingham, AL) Cora E. Lewis, Albert Oberman, James M. Shikany, Monika Safford, Brian K. Britt;
(University of Arizona, Tucson/Phoenix, AZ) Tamsen Bassford, John Mattox, Marcia Ko, Timothy Lohman; (University
at Buffalo, Buffalo, NY) Jean Wactawski-Wende, Maurizio
Trevisan, Ellen Smit, Susan Graham, June Chang; (University
of California at Davis, Sacramento, CA) John Robbins, S.
Yasmeen, Karen Lindfors, Judith Stern; (University of
California at Irvine, Orange, CA) Allan Hubbell, Gail Frank,
Nathan Wong, Nancy Greep, Bradley Monk; (University of
California at Los Angeles, Los Angeles, CA) Howard Judd,
David Heber, Robert Elashoff; (University of California at
San Diego, LaJolla/Chula Vista, CA) Robert D. Langer,
Michael H. Criqui, Gregory T. Talavera, Cedric F. Garland,
R. Elaine Hanson; (University of Cincinnati, Cincinnati, OH)
Margery Gass, Suzanne Wernke, Nelson Watts; (University of
Florida, Gainesville/Jacksonville, FL) Marian Limacher, Michael Perri, Andrew Kaunitz, R. Stan Williams, Yvonne
Brinson; (University of Hawaii, Honolulu, HI) David Curb,
Helen Petrovitch, Beatriz Rodriguez, Kamal Masaki, Santosh
Sharma; (University of Iowa, Iowa City/Davenport, IA)
Robert Wallace, James Torner, Susan Johnson, Linda Snetselaar, Bradley VanVoorhis; (University of Massachusetts/
Fallon Clinic, Worcester, MA) Judith Ockene, Milagros Rosal,
Ira Ockene, Robert Yood, Patricia Aronson; (University of
Medicine and Dentistry of New Jersey, Newark, NJ) Norman
Lasser, Bali Singh, Vera Lasser, John Kostis; (University of
Miami, Miami, FL) Mary Jo O’Sullivan, Linda Parker, R.
Estape, Diann Fernandez; (University of Minnesota, Minneapolis, MN) Karen L. Margolis, Richard H. Grimm, Donald B.
Hunninghake, June LaValleur, Sarah Kempainen; (University
of Nevada, Reno, NV) Robert Brunner, William Graettinger,
Vicki Oujevolk; (University of North Carolina, Chapel Hill,
NC) Gerardo Heiss, Pamela Haines, David Ontjes, Carla
Sueta, Ellen Wells; (University of Pittsburgh, Pittsburgh, PA)
Lewis Kuller, Jane Cauley, N. Carole Milas; (University of
Tennessee, Memphis, TN) Karen C. Johnson, Suzanne
Satterfield, Raymond W. Ke, Jere Vile, Fran Tylavsky;
(University of Texas Health Science Center, San Antonio,
TX) Robert Brzyski, Robert Schenken, Jose Trabal, Mercedes
Rodriguez-Sifuentes, Charles Mouton; (University of
Wisconsin, Madison, WI) Gloria Sarto, Douglas Laube,
Patrick McBride, Julie Mares-Perlman, Barbara Loevinger;
(Wake Forest University School of Medicine, Winston-Salem,
NC) Denise Bonds, Greg Burke, Robin Crouse, Lynne
Parsons, Mara Vitolins; (Wayne State University School of
Medicine/Hutzel Hospital, Detroit, MI) Susan Hendrix,
Michael Simon, Gene McNeeley, Pamela Gordon, Paul
Makela.
Former WHI Investigators: Catherine Allen (University of
Wisconsin, Madison, WI) *, Sandy Dougherty* (University of
Nevada, Reno, NV), Richard Carleton*(Brown University,
Providence, RI).
International Journal of Obesity