Abstract

The authors conducted a population-based case-control study of 832 endometrial cancer cases and 846 frequency-matched controls in Shanghai, China (1997–2001), to examine the association of overall adiposity and body fat distribution with disease risk. Overall adiposity was estimated using weight and body mass index (BMI); upper body fat distribution was evaluated using waist circumference and waist:hip ratio. Overall and upper-body obesity were both associated with an elevated risk of endometrial cancer. Adjusted odds ratios and 95% confidence intervals for highest-versus-lowest quartile comparisons were 2.6 (95% confidence interval (CI): 2.0, 3.5) for weight, 2.9 (95% CI: 2.2, 3.9) for BMI, 4.7 (95% CI: 3.4, 6.4) for waist circumference, and 3.5 (95% CI: 2.6, 4.8) for waist:hip ratio. The positive associations with weight and BMI vanished after results were controlled for waist circumference, while associations with waist circumference and waist:hip ratio persisted after adjustment for BMI. The positive association with upper-body obesity was more pronounced among younger women, women who had never used oral contraceptives, and women with a history of diabetes mellitus (p for multiplicative interaction < 0.05). Upper-body obesity was related to increased risk among women with low BMI. These results suggest that obesity, particularly upper-body fat deposition, is associated with an increased risk of endometrial cancer.

Obesity has been associated with elevated endogenous estrogen levels, decreased levels of sex hormone-binding globulin, and reduced progesterone production, all of which are believed to create an environment favorable to the development of endometrial cancer (16). A variety of epidemiologic studies have examined the association between obesity and endometrial cancer (720). The evidence for the adverse effects of overall adiposity is quite convincing. The association of body fat distribution with endometrial cancer has been less well characterized. While some studies have shown a positive association with upper-body (1214) or central (2, 9, 10) obesity, others have found no association with waist:hip ratio (9) or have found an association that disappeared after adjustment for body mass index (BMI) (1, 2, 19). However, the majority of studies on fat distribution have been conducted in Western populations, in which the prevalence of obesity is high and accurate measurements of fat distribution are difficult to obtain, particularly among extremely obese women.

The incidence of endometrial cancer and average body weight have been rising in parallel over the past two decades among women in the urban area of Shanghai, China (20, 21). However, the prevalence of obesity in this population is still low compared with that of women in Western countries, and extreme obesity is rare among women in Shanghai; this offered us a unique opportunity to evaluate the effect of fat distribution on endometrial cancer independently of overall obesity. We used data from a population-based case-control study in urban Shanghai to examine the association of endometrial cancer with measures of overall adiposity and body fat distribution.

MATERIALS AND METHODS

Eligible cases, identified from the population-based Shanghai Cancer Registry, included female permanent residents of urban Shanghai aged 30–69 years who were newly diagnosed with endometrial cancer between January 1997 and December 2001. The study was approved by the relevant institutional review boards for human research in China and the United States. In-person interviews were performed for 832 of 982 (84.7 percent) eligible cases by 22 trained nurses and physicians. Among case nonparticipants, 73 declined to be interviewed (7.4 percent), 36 were deceased (3.7 percent), 22 could not be located (2.2 percent), 10 did not live in Shanghai during the study period (1.0 percent), and nine were excluded for other miscellaneous reasons (0.9 percent).

Controls were selected from female permanent residents of the Shanghai urban area through the Shanghai Resident Registry and were frequency matched to cases. The number of controls in each 5-year age group was determined in advance according to the ages of the incident endometrial cancer cases in 1996. Women who had had a hysterectomy (identified during the survey) were not eligible (n = 36). A total of 1,165 eligible controls were identified for possible participation, and 846 (72.6 percent) were enrolled in the study. Among control nonparticipants, 277 (23.8 percent) declined to be interviewed, and 42 (3.6 percent) did not live in Shanghai during the study period. After written consent was obtained, a structured in-person interview was conducted to elicit information on demographic factors, menstrual and reproductive history, hormone use, prior disease history, tobacco and alcohol use, family history of cancer, and height and weight history during adolescence and adulthood. Usual dietary habits over the previous 5 years were assessed through in-person interviews with a validated, quantitative food frequency questionnaire. Physician-diagnosed diabetes mellitus was ascertained by asking, “Have you ever been told by a doctor that you have diabetes?” Age at diagnosis was also requested. Menopause was defined as cessation of the menstrual cycle for 12 months or longer, excluding those periods caused by pregnancy or breastfeeding.

Each participant was also measured for her current weight, sitting and standing height, and waist and hip circumference. Height was measured to the nearest 0.1 cm using a stadiometer. Girth measurements, recorded to the nearest 0.1 cm, were taken with a cloth tape. Waist circumference was measured at a level 2.5 cm above the umbilicus. Hip circumference was defined as the maximum girth reading between the waist and the thigh. Body weight was measured with electronic scales to the nearest 0.1 kg. Two measurements were taken, with tolerances of <1 cm for height and circumference and 1 kg for weight. A third measurement was taken if the difference between the first two exceeded the tolerance. The mean of the replicates was used in the analyses. The median interval between diagnosis and interview for cases was 5 months (interquartile range: 3, 8).

Adiposity was estimated using both body weight and BMI, which was computed as weight in kilograms divided by height in meters squared (kg/m2). Body fat distribution was estimated using both waist circumference and waist:hip ratio. For analysis, the anthropometric indices were grouped into quartiles based on distributions among the controls.

We estimated the risk of endometrial cancer associated with each anthropometric factor by calculating odds ratios and 95 percent confidence intervals derived from unconditional logistic regression (22), with adjustment for age and confounding factors. Potential confounders evaluated in the study included reported risk factors and factors that were related to endometrial cancer risk in the study population. Only factors that altered obesity and endometrial cancer risk estimates by 10 percent were considered as confounders and adjusted for in the final model. Multiplicative interaction between two study variables was examined by including the main effect and a cross-product term of two variables in the regression model. Collinearity between variables was checked before variables were included in the regression model. Tests for trend were performed by entering the categorical variables as continuous parameters in the models.

RESULTS

The descriptive characteristics of cases and controls are shown in table 1. There were no significant (χ2 test: p > 0.10) case-control differences with regard to age, family income, regular exercise, total energy intake, or use of hormone replacement therapy. However, compared with controls, cases were more likely to have higher education, a history of diabetes mellitus, an earlier age at menarche, a later age at menopause, more cumulative years of menstruation, fewer pregnancies, and a positive family history of any cancer and endometrial cancer. In addition, cases were less likely to drink alcohol or use oral contraceptives. Of these factors, only years of menstruation and number of pregnancies were found to confound the association of obesity with endometrial cancer, and we adjusted for these factors in subsequent analyses. Because it was the primary match variable, we adjusted for age in the logistic model, while we included education in the model to control for potential selective participation bias.

TABLE 1.

Distribution of cases and controls according to demographic factors and selected endometrial cancer risk factors, Shanghai, China, January 1997–December 2001*


 

Cases (n = 832)
 
 
Controls (n = 846)
 
 
p value from χ2 test
 

 
No. or %
 
SD or IQR
 
No. or %
 
SD or IQR
 
 
Mean age (years) 55.3 8.6 55.7 8.6 0.30 
Education (%)      
    No formal education or elementary school only 24.5  27.7   
    Middle or high school 60.5  60.6   
    College or above 15.0  11.7  0.08 
Marital status (%)      
    Unmarried 1.7  1.2   
    Married or cohabiting 87.0  87.7   
    Separated/divorced/widowed 11.3  11.1  0.68 
Per capita income in previous year (%)      
    ≤$508.1 27.6  28.8   
    $508.2–$762.2 29.3  28.7   
    $762.3–$1,016.3 6.9  5.9   
    >$1,016.3 36.2  36.5  0.83 
No. of pregnancies (%)      
    0 7.5  4.1   
    1 16.5  12.9   
    2 23.9  24.6   
    3 23.3  24.5   
    4 17.0  18.6   
    ≥5 11.9  15.4  <0.01 
Diabetes mellitus (%)      
    Yes 14.6  6.9   
    No 85.0  92.9   
    Unknown 0.5  0.2  <0.01 
Polycystic ovary syndrome (%)      
    Yes 0.9  0.2   
    No 98.5  99.4   
    Unknown 0.6  0.4  0.18 
Family history of endometrial cancer (%) 1.4  0.2  <0.01 
Cancer among first-degree relatives (%) 35.1  27.1  <0.01 
Postmenopause (%) 60.7  64.5  0.10 
Regular alcohol consumption (%) 2.4  5.0  <0.01 
Hormone replace therapy (%) 4.2  4.0  0.85 
Oral contraceptive use (%) 17.7  24.5  <0.01 
Regular exercise (%) 30.4  33.9  0.12 
Median age (years) at menarche 14 13, 16 15 13, 16 <0.01 
Median age (years) at menopause (postmenopausal women) 50.1 48.6, 52.5 49.4 47.1, 51.1 <0.01 
Median years of menstruation 33.2 30.0, 36.1 31.5 27.8, 34.5 <0.01 
Median usual energy intake (kcal/day)
 
2,171.2
 
1,871.4, 2,497.0
 
2,141.5
 
1,840.8, 2,485.2
 
0.71
 

 

Cases (n = 832)
 
 
Controls (n = 846)
 
 
p value from χ2 test
 

 
No. or %
 
SD or IQR
 
No. or %
 
SD or IQR
 
 
Mean age (years) 55.3 8.6 55.7 8.6 0.30 
Education (%)      
    No formal education or elementary school only 24.5  27.7   
    Middle or high school 60.5  60.6   
    College or above 15.0  11.7  0.08 
Marital status (%)      
    Unmarried 1.7  1.2   
    Married or cohabiting 87.0  87.7   
    Separated/divorced/widowed 11.3  11.1  0.68 
Per capita income in previous year (%)      
    ≤$508.1 27.6  28.8   
    $508.2–$762.2 29.3  28.7   
    $762.3–$1,016.3 6.9  5.9   
    >$1,016.3 36.2  36.5  0.83 
No. of pregnancies (%)      
    0 7.5  4.1   
    1 16.5  12.9   
    2 23.9  24.6   
    3 23.3  24.5   
    4 17.0  18.6   
    ≥5 11.9  15.4  <0.01 
Diabetes mellitus (%)      
    Yes 14.6  6.9   
    No 85.0  92.9   
    Unknown 0.5  0.2  <0.01 
Polycystic ovary syndrome (%)      
    Yes 0.9  0.2   
    No 98.5  99.4   
    Unknown 0.6  0.4  0.18 
Family history of endometrial cancer (%) 1.4  0.2  <0.01 
Cancer among first-degree relatives (%) 35.1  27.1  <0.01 
Postmenopause (%) 60.7  64.5  0.10 
Regular alcohol consumption (%) 2.4  5.0  <0.01 
Hormone replace therapy (%) 4.2  4.0  0.85 
Oral contraceptive use (%) 17.7  24.5  <0.01 
Regular exercise (%) 30.4  33.9  0.12 
Median age (years) at menarche 14 13, 16 15 13, 16 <0.01 
Median age (years) at menopause (postmenopausal women) 50.1 48.6, 52.5 49.4 47.1, 51.1 <0.01 
Median years of menstruation 33.2 30.0, 36.1 31.5 27.8, 34.5 <0.01 
Median usual energy intake (kcal/day)
 
2,171.2
 
1,871.4, 2,497.0
 
2,141.5
 
1,840.8, 2,485.2
 
0.71
 
*

Subjects with missing values were excluded from the analysis.

SD, standard deviation; IQR, interquartile range (25th percentile, 75th percentile).

Age at diagnosis for cases and age at interview for controls.

Table 2 presents the odds ratios for the anthropometric measurements. Increased body weight, BMI, waist circumference, hip circumference, and waist:hip ratio were all positively associated with endometrial cancer (p for trend < 0.01). The adjusted odds ratios for the highest quartile versus the lowest were 2.6 (95 percent confidence interval (CI): 2.0, 3.5) for body weight, 2.9 (95 percent CI: 2.2, 3.9) for BMI, 4.7 (95 percent CI: 3.4, 6.4) for waist circumference, 3.1 (95 percent CI: 2.3, 4.2) for hip circumference, and 3.5 (95 percent CI: 2.6, 4.8) for waist:hip ratio.

TABLE 2.

Associations between anthropometric measurements and endometrial cancer risk, Shanghai, China, January 1997–December 2001



 

No. of cases
 

No. of controls
 

OR1*
 

95% CI*
 

OR2
 

95% CI
 
Body height (cm)§       
    ≤153 169 217 1.0  1.0  
    154–157 221 209 1.3 1.0, 1.8 1.1 0.8, 1.5 
    158–161 236 226 1.3 1.0, 1.7 0.9 0.7, 1.3 
    >161 203 192 1.3 1.0, 1.8 0.9 0.6, 1.2 
        p for trend   0.15  0.20  
Body weight (kg)       
    ≤53 135 235 1.0  1.0  
    54–59 165 219 1.3 1.0, 1.7 0.9 0.7, 1.2 
    60–65 230 195 2.0 1.5, 2.7 1.1 0.7, 1.5 
    >65 297 196 2.6 2.0, 3.5 0.9 0.6, 1.3 
        p for trend   <0.01  0.70  
Body mass index#       
    ≤21.4 120 211 1.0  1.0  
    21.5–23.7 170 211 1.4 1.0, 1.9 1.0 0.7, 1.4 
    23.8–26.2 214 212 1.7 1.3, 2.3 0.9 0.6, 1.3 
    >26.2 322 210 2.9 2.2, 3.9 1.0 0.6, 1.5 
        p for trend   <0.01  0.79  
Waist circumference (cm)**       
    ≤73 102 229 1.0  1.0  
    74–79 157 198 1.9 1.4, 2.7 1.8 1.3, 2.6 
    80–86 215 207 2.6 1.9, 3.6 2.4 1.7, 3.3 
    >86 357 211 4.7 3.4, 6.4 3.9 2.5, 5.9 
        p for trend   <0.01  <0.01  
Hip circumference (cm)       
    ≤91 117 213 1.0  1.0  
    92–97 196 230 1.6 1.2, 2.2 1.1 0.8, 1.5 
    98–103 211 201 2.0 1.5, 2.7 1.0 0.7, 1.4 
    >103 307 201 3.1 2.3, 4.2 0.9 0.6, 1.4 
        p for trend   <0.01  0.48  
Waist:hip ratio**       
    ≤0.782 106 212 1.0  1.0  
    0.783–0.814 158 211 1.6 1.1, 2.2 1.4 1.0, 1.9 
    0.815–0.855 248 212 2.5 1.8, 3.4 2.0 1.5, 2.7 
    >0.855 319 210 3.5 2.6, 4.8 2.6 1.9, 3.6 
        p for trend
 

 

 
<0.01
 

 
<0.01
 

 


 

No. of cases
 

No. of controls
 

OR1*
 

95% CI*
 

OR2
 

95% CI
 
Body height (cm)§       
    ≤153 169 217 1.0  1.0  
    154–157 221 209 1.3 1.0, 1.8 1.1 0.8, 1.5 
    158–161 236 226 1.3 1.0, 1.7 0.9 0.7, 1.3 
    >161 203 192 1.3 1.0, 1.8 0.9 0.6, 1.2 
        p for trend   0.15  0.20  
Body weight (kg)       
    ≤53 135 235 1.0  1.0  
    54–59 165 219 1.3 1.0, 1.7 0.9 0.7, 1.2 
    60–65 230 195 2.0 1.5, 2.7 1.1 0.7, 1.5 
    >65 297 196 2.6 2.0, 3.5 0.9 0.6, 1.3 
        p for trend   <0.01  0.70  
Body mass index#       
    ≤21.4 120 211 1.0  1.0  
    21.5–23.7 170 211 1.4 1.0, 1.9 1.0 0.7, 1.4 
    23.8–26.2 214 212 1.7 1.3, 2.3 0.9 0.6, 1.3 
    >26.2 322 210 2.9 2.2, 3.9 1.0 0.6, 1.5 
        p for trend   <0.01  0.79  
Waist circumference (cm)**       
    ≤73 102 229 1.0  1.0  
    74–79 157 198 1.9 1.4, 2.7 1.8 1.3, 2.6 
    80–86 215 207 2.6 1.9, 3.6 2.4 1.7, 3.3 
    >86 357 211 4.7 3.4, 6.4 3.9 2.5, 5.9 
        p for trend   <0.01  <0.01  
Hip circumference (cm)       
    ≤91 117 213 1.0  1.0  
    92–97 196 230 1.6 1.2, 2.2 1.1 0.8, 1.5 
    98–103 211 201 2.0 1.5, 2.7 1.0 0.7, 1.4 
    >103 307 201 3.1 2.3, 4.2 0.9 0.6, 1.4 
        p for trend   <0.01  0.48  
Waist:hip ratio**       
    ≤0.782 106 212 1.0  1.0  
    0.783–0.814 158 211 1.6 1.1, 2.2 1.4 1.0, 1.9 
    0.815–0.855 248 212 2.5 1.8, 3.4 2.0 1.5, 2.7 
    >0.855 319 210 3.5 2.6, 4.8 2.6 1.9, 3.6 
        p for trend
 

 

 
<0.01
 

 
<0.01
 

 
*

OR, odds ratio; CI, confidence interval.

Adjusted for age, educational level, years of menstruation, and number of pregnancies.

Adjusted for age, educational level, years of menstruation, and number of pregnancies, with additional adjustment for weight (§), waist circumference (¶), or body mass index (**), as indicated.

§

Additionally adjusted for weight.

Additionally adjusted for waist circumference.

#

Weight (kg)/height (m)2.

**

Additionally adjusted for body mass index.

In models examining the effect of waist circumference, additional adjustment for BMI resulted in a weaker association for women in the upper quartile of waist circumference (the odds ratio decreased from 4.7 to 3.9, or by approximately 17 percent (table 2)). Controlling for waist circumference in the BMI and body weight models completely attenuated their positive associations with endometrial cancer, with the odds ratios decreasing from 2.9 and 2.6 to 1.0 and 0.9, respectively (table 2). Controlling for waist:hip ratio in the BMI models only slightly attenuated the association with BMI (the odds ratio for the upper quartile decreased from 2.9 to 2.1; data not shown in table). Analyses using the conventional definition of obesity showed similar results (data not shown in table). Compared with women with BMIs <25, odds ratios for women with BMIs of 25–<30 and ≥30 were 1.6 (95 percent CI: 1.3, 2.0) and 3.3 (95 percent CI: 2.3, 4.7), respectively, without adjustment for waist:hip ratio or waist circumference. The corresponding odds ratios were 1.3 (95 percent CI: 1.1, 1.7) and 2.6 (95 percent CI: 1.8, 3.7) after adjustment for waist:hip ratio and 0.9 (95 percent CI: 0.7, 1.1) and 1.1 (95 percent CI: 0.7, 1.7) after adjustment for waist circumference.

Table 3 examines the independent and combined effects of fat distribution by BMI level. Higher BMIs and waist:hip ratios were positively associated with the risk of endometrial cancer, and no interaction between BMI and waist:hip ratio was observed (interaction test: p = 0.90). Stratified analysis indicated that central adiposity was related to increased risk regardless of BMI level (table 3). After adjustment for waist circumference, BMI was no longer a significant predictor of risk. No interaction between BMI and waist circumference was observed (p = 0.68).

TABLE 3.

Joint effects of body mass index and waist:hip ratio and body mass index and waist circumference on the risk of endometrial cancer, Shanghai, China, January 1997–December 2001*


Quartile of waist:hip ratio or waist circumference
 

Quartile of body mass index
 
               
p for interaction
 
 ≤21.4 (Q1§)
 
   21.5–23.7 (Q2)
 
   23.8–26.2 (Q3)
 
   >26.2 (Q4)
 
    
 No. of cases
 
No. of controls
 
OR§
 
95% CI§
 
No. of cases
 
No. of controls
 
OR
 
95% CI
 
No. of cases
 
No. of controls
 
OR
 
95% CI
 
No. of cases
 
No. of controls
 
OR
 
95% CI
 
 
Waist:hip ratio                 0.90 
    ≤0.782 (Q1) 45 106 1.0  27 54 1.1 0.6, 2.0 29 35 2.1 1.1, 3.9 11 19 1.6 0.7, 3.6  
    0.783–0.814 (Q2) 37 52 1.9 1.1, 3.4 27 65 1.1 0.6, 1.9 51 52 2.2 1.3, 3.7 42 42 2.7 1.5, 4.8  
    0.815–0.855 (Q3) 28 35 2.0 1.1, 3.7 67 54 3.0 1.8, 5.0 60 68 2.2 1.3, 3.6 90 55 4.6 2.8, 7.7  
    >0.855 (Q4) 16
 
20
 
2.4
 
1.1, 5.2
 
49
 
38
 
3.8
 
2.1, 6.6
 
74
 
57
 
3.5
 
2.1, 5.9
 
179
 
94
 
5.5
 
3.5, 8.7
 
 
 Q1 and Q2 combined
 
       Q3 and Q4 combined
 
        
Waist circumference (cm)                 0.68 
    ≤73 (Q1) 98 217 1.0      14 1.5 0.6, 3.6      
    74–79 (Q2) 106 126 2.1 1.5, 3.0     50 72 1.6 1.0, 2.5      
    80–86 (Q3) 71 70 2.6 1.7, 4.0     142 136 2.6 1.8, 3.6      
    >86 (Q4)
 
21
 
11
 
5.6
 
2.5, 12.3
 

 

 

 

 
335
 
200
 
4.6
 
3.3, 6.3
 

 

 

 

 

 

Quartile of waist:hip ratio or waist circumference
 

Quartile of body mass index
 
               
p for interaction
 
 ≤21.4 (Q1§)
 
   21.5–23.7 (Q2)
 
   23.8–26.2 (Q3)
 
   >26.2 (Q4)
 
    
 No. of cases
 
No. of controls
 
OR§
 
95% CI§
 
No. of cases
 
No. of controls
 
OR
 
95% CI
 
No. of cases
 
No. of controls
 
OR
 
95% CI
 
No. of cases
 
No. of controls
 
OR
 
95% CI
 
 
Waist:hip ratio                 0.90 
    ≤0.782 (Q1) 45 106 1.0  27 54 1.1 0.6, 2.0 29 35 2.1 1.1, 3.9 11 19 1.6 0.7, 3.6  
    0.783–0.814 (Q2) 37 52 1.9 1.1, 3.4 27 65 1.1 0.6, 1.9 51 52 2.2 1.3, 3.7 42 42 2.7 1.5, 4.8  
    0.815–0.855 (Q3) 28 35 2.0 1.1, 3.7 67 54 3.0 1.8, 5.0 60 68 2.2 1.3, 3.6 90 55 4.6 2.8, 7.7  
    >0.855 (Q4) 16
 
20
 
2.4
 
1.1, 5.2
 
49
 
38
 
3.8
 
2.1, 6.6
 
74
 
57
 
3.5
 
2.1, 5.9
 
179
 
94
 
5.5
 
3.5, 8.7
 
 
 Q1 and Q2 combined
 
       Q3 and Q4 combined
 
        
Waist circumference (cm)                 0.68 
    ≤73 (Q1) 98 217 1.0      14 1.5 0.6, 3.6      
    74–79 (Q2) 106 126 2.1 1.5, 3.0     50 72 1.6 1.0, 2.5      
    80–86 (Q3) 71 70 2.6 1.7, 4.0     142 136 2.6 1.8, 3.6      
    >86 (Q4)
 
21
 
11
 
5.6
 
2.5, 12.3
 

 

 

 

 
335
 
200
 
4.6
 
3.3, 6.3
 

 

 

 

 

 
*

Odds ratios were adjusted for age, educational level, years of menstruation, and number of pregnancies.

Weight (kg)/height (m)2.

Classified by median value in joint-effect analyses with waist circumference.

§

Q, quartile; OR, odds ratio; CI, confidence interval.

Summarized in table 4 are results for the combined effects of waist circumference and BMI with age (<50 years/≥50 years), menopausal status (premenopause/postmenopause), oral contraceptive use (ever/never), and ever having been diagnosed with diabetes mellitus (never/ever). The magnitude of the association with waist circumference was significantly greater among younger or premenopausal women than in older or postmenopausal women; p values for interaction were <0.01 and 0.07, respectively. Further analyses showed that age rather than menopausal status was the main effect modifier. Excluding women (seven cases and two controls) with polycystic ovary syndrome did not change the results (data not shown). The association between waist circumference and endometrial cancer risk was modified by oral contraceptive use (mainly combined oral contraceptives or oral contraceptives containing only progesterone) (p for interaction = 0.02) and diabetes (p for interaction < 0.01), with the positive association being more pronounced among women who never used oral contraceptives and being seen among only nondiabetic women. Among 178 women with a history of diabetes, only nine (5 percent) were diagnosed at an age of 40 years or younger. Excluding those probable type I diabetes mellitus cases from the analysis did not change the study results (data not shown). A similar pattern of effect modification was observed for BMI, but only the interaction with diabetes reached statistical significance.

TABLE 4.

Associations of endometrial cancer with waist circumference and body mass index, by age, menstrual status, oral contraceptive use, and diabetes mellitus, Shanghai, China, January 1997–December 2001*


Stratum
 

Quartile 1
 
   
Quartile 2
 
   
Quartile 3
 
   
Quartile 4
 
   
p for interaction
 
 No. of cases
 
No. of controls
 
OR
 
95% CI
 
No. of cases
 
No. of controls
 
OR
 
95% CI
 
No. of cases
 
No. of controls
 
OR
 
95% CI
 
No. of cases
 
No. of controls
 
OR
 
95% CI
 
 
Waist circumference                  
Age (years)                  
    <50 38 104 1.0  56 56 2.7 1.6, 4.5 75 41 4.7 2.7, 8.0 77 22 9.7 5.3, 18.0  
    ≥50 65 126 1.0 0.6, 1.8 101 142 1.6 0.9, 2.7 140 166 1.9 1.1, 3.2 280 189 3.4 2.0, 5.8 <0.01 
Menopausal status                  
    Premenopausal 50 124 1.0  76 76 2.4 1.5, 3.8 99 63 3.6 2.3, 5.7 103 38 6.4 3.9, 10.6  
    Postmenopausal 53 106 1.5 0.9, 2.6 81 122 2.3 1.4, 3.9 116 144 2.9 1.8, 4.7 254 173 5.4 3.4, 8.8 0.07 
Age (years) and menopausal status                  
    Age <50 38 104 1.0  56 56 2.7 1.6, 4.5 75 41 4.7 2.7, 8.0 77 22 9.7 5.3, 18.0  
    Age ≥50, premenopausal 16 27 1.0 0.5, 2.1 24 24 1.6 0.8, 3.2 32 25 2.0 1.0, 3.9 36 18 2.9 1.4, 5.9 <0.01 
    Age ≥50, postmenopausal 49 99 1.1 0.6, 2.0 77 118 1.6 0.9, 2.9 108 141 1.9 1.1, 3.4 244 171 3.6 2.1, 6.4  
Oral contraceptive use                  
    Ever use 17 46 1.0  34 42 2.5 1.2, 5.2 31 52 1.8 0.9, 3.8 65 67 3.0 1.5, 5.9  
    Never use 86 184 1.1 0.6, 2.0 123 156 2.0 1.1, 3.6 184 155 3.1 1.7, 5.7 292 144 5.9 3.2, 10.9 0.02 
Diabetes mellitus                  
    No 109 234 1.0  131 187 1.6 1.1, 2.2 185 188 2.4 1.7, 3.2 287 179 4.3 3.1, 5.9  
    Yes 18.1 2.0, 159.2 25 10 7.4 3.4, 16.4 24 17 3.5 1.8, 7.0 65 30 6.0 3.6, 10.1 <0.01 
Body mass index                  
Age (years)                  
    <50 50 80 1.0  59 66 1.4 0.9, 2.3 66 51 2.0 1.2, 3.3 76 27 4.5 2.5, 8.0  
    ≥50 76 133 0.7 0.4, 1.1 111 145 0.9 0.5, 1.4 148 161 1.0 0.6, 1.7 246 183 1.6 1.0, 2.6 0.06 
Menopausal status                  
    Premenopausal 57 103 1.0  82 80 1.8 1.1, 2.8 90 68 2.2 1.4, 3.4 102 51 3.3 2.1, 5.4  
    Postmenopausal 69 110 1.4 0.9, 2.4 88 131 1.5 0.9, 2.4 124 144 1.9 1.2, 3.1 220 159 3.4 2.1, 5.4 0.36 
Age (years) and menopausal status                  
    Age <50 50 80 1.0  59 66 1.4 0.8, 2.3 66 51 2.0 1.2, 3.3 76 27 4.5 2.5, 8.0  
    Age ≥50, premenopausal 12 30 0.4 0.2, 0.8 29 18 1.4 0.7, 2.8 32 21 1.2 0.6, 2.5 33 25 1.0 0.5, 2.1  
    Age ≥50, postmenopausal 64 103 0.8 0.4, 1.4 82 127 0.8 0.4, 1.4 116 140 1.0 0.6, 1.7 213 158 1.8 1.0, 3.1 0.13 
Oral contraceptive use                  
    Ever use 19 42 1.0  42 52 1.9 1.0, 3.8 31 59 1.1 0.5, 2.2 59 55 2.6 1.3, 5.0  
    Never use 107 171 1.2 0.7, 2.3 128 159 1.5 0.8, 2.8 183 153 2.3 1.3, 4.2 263 155 3.6 2.0, 6.6 0.47 
Diabetes mellitus                  
    No 125 214 1.0  143 201 1.2 0.9, 1.6 187 193 1.6 1.2, 2.2 258 180 2.7 2.0, 3.7  
    Yes
 
13
 
4
 
7.5
 
2.3, 24.3
 
26
 
9
 
5.7
 
2.5, 12.7
 
21
 
18
 
2.0
 
1.0, 3.9
 
59
 
27
 
4.3
 
2.5, 7.3
 
<0.01
 

Stratum
 

Quartile 1
 
   
Quartile 2
 
   
Quartile 3
 
   
Quartile 4
 
   
p for interaction
 
 No. of cases
 
No. of controls
 
OR
 
95% CI
 
No. of cases
 
No. of controls
 
OR
 
95% CI
 
No. of cases
 
No. of controls
 
OR
 
95% CI
 
No. of cases
 
No. of controls
 
OR
 
95% CI
 
 
Waist circumference                  
Age (years)                  
    <50 38 104 1.0  56 56 2.7 1.6, 4.5 75 41 4.7 2.7, 8.0 77 22 9.7 5.3, 18.0  
    ≥50 65 126 1.0 0.6, 1.8 101 142 1.6 0.9, 2.7 140 166 1.9 1.1, 3.2 280 189 3.4 2.0, 5.8 <0.01 
Menopausal status                  
    Premenopausal 50 124 1.0  76 76 2.4 1.5, 3.8 99 63 3.6 2.3, 5.7 103 38 6.4 3.9, 10.6  
    Postmenopausal 53 106 1.5 0.9, 2.6 81 122 2.3 1.4, 3.9 116 144 2.9 1.8, 4.7 254 173 5.4 3.4, 8.8 0.07 
Age (years) and menopausal status                  
    Age <50 38 104 1.0  56 56 2.7 1.6, 4.5 75 41 4.7 2.7, 8.0 77 22 9.7 5.3, 18.0  
    Age ≥50, premenopausal 16 27 1.0 0.5, 2.1 24 24 1.6 0.8, 3.2 32 25 2.0 1.0, 3.9 36 18 2.9 1.4, 5.9 <0.01 
    Age ≥50, postmenopausal 49 99 1.1 0.6, 2.0 77 118 1.6 0.9, 2.9 108 141 1.9 1.1, 3.4 244 171 3.6 2.1, 6.4  
Oral contraceptive use                  
    Ever use 17 46 1.0  34 42 2.5 1.2, 5.2 31 52 1.8 0.9, 3.8 65 67 3.0 1.5, 5.9  
    Never use 86 184 1.1 0.6, 2.0 123 156 2.0 1.1, 3.6 184 155 3.1 1.7, 5.7 292 144 5.9 3.2, 10.9 0.02 
Diabetes mellitus                  
    No 109 234 1.0  131 187 1.6 1.1, 2.2 185 188 2.4 1.7, 3.2 287 179 4.3 3.1, 5.9  
    Yes 18.1 2.0, 159.2 25 10 7.4 3.4, 16.4 24 17 3.5 1.8, 7.0 65 30 6.0 3.6, 10.1 <0.01 
Body mass index                  
Age (years)                  
    <50 50 80 1.0  59 66 1.4 0.9, 2.3 66 51 2.0 1.2, 3.3 76 27 4.5 2.5, 8.0  
    ≥50 76 133 0.7 0.4, 1.1 111 145 0.9 0.5, 1.4 148 161 1.0 0.6, 1.7 246 183 1.6 1.0, 2.6 0.06 
Menopausal status                  
    Premenopausal 57 103 1.0  82 80 1.8 1.1, 2.8 90 68 2.2 1.4, 3.4 102 51 3.3 2.1, 5.4  
    Postmenopausal 69 110 1.4 0.9, 2.4 88 131 1.5 0.9, 2.4 124 144 1.9 1.2, 3.1 220 159 3.4 2.1, 5.4 0.36 
Age (years) and menopausal status                  
    Age <50 50 80 1.0  59 66 1.4 0.8, 2.3 66 51 2.0 1.2, 3.3 76 27 4.5 2.5, 8.0  
    Age ≥50, premenopausal 12 30 0.4 0.2, 0.8 29 18 1.4 0.7, 2.8 32 21 1.2 0.6, 2.5 33 25 1.0 0.5, 2.1  
    Age ≥50, postmenopausal 64 103 0.8 0.4, 1.4 82 127 0.8 0.4, 1.4 116 140 1.0 0.6, 1.7 213 158 1.8 1.0, 3.1 0.13 
Oral contraceptive use                  
    Ever use 19 42 1.0  42 52 1.9 1.0, 3.8 31 59 1.1 0.5, 2.2 59 55 2.6 1.3, 5.0  
    Never use 107 171 1.2 0.7, 2.3 128 159 1.5 0.8, 2.8 183 153 2.3 1.3, 4.2 263 155 3.6 2.0, 6.6 0.47 
Diabetes mellitus                  
    No 125 214 1.0  143 201 1.2 0.9, 1.6 187 193 1.6 1.2, 2.2 258 180 2.7 2.0, 3.7  
    Yes
 
13
 
4
 
7.5
 
2.3, 24.3
 
26
 
9
 
5.7
 
2.5, 12.7
 
21
 
18
 
2.0
 
1.0, 3.9
 
59
 
27
 
4.3
 
2.5, 7.3
 
<0.01
 
*

Odds ratios were adjusted for age, educational level, years of menstruation, and number of pregnancies.

OR, odds ratio; CI, confidence interval.

Weight (kg)/height (m)2.

We carefully examined potential confounding from other nonanthropometric factors, such as regular exercise, alcohol consumption, breastfeeding, and any cancer history among first-degree relatives. There was no evidence of confounding from any of these factors. In addition, excluding subjects who had ever used hormone replacement therapy or women with induced menopause did not appreciably change the risk estimates. Finally, we assessed our results for effect modification by energy intake and regular exercise and found no evidence of effect modification by these factors.

DISCUSSION

The effect of obesity on endometrial cancer risk has been of interest for decades. Since Mack et al. (23) reported a positive association between adiposity and risk of endometrial cancer in the early 1970s, a number of investigators have explored the role of obesity in the development of endometrial cancer, primarily using measurements of body weight and indices of relative weight as an indicator of overall adiposity. Most of these investigators have reported that overall adiposity is a strong risk factor for endometrial cancer, with risk estimates ranging from 2 to 10 (1, 2, 719, 24). However, studies examining the relation between endometrial cancer and body fat distribution have produced mixed results. While several studies have found that body fat distribution confers additional risk for endometrial cancer (2, 9, 10, 12, 13), the positive association between waist:hip ratio and the disease disappeared after adjustment for BMI in two other studies (1, 19), and one Japanese study observed an increased risk for fat distribution but not for overall adiposity (14).

Results from this population-based case-control study suggest that overall adiposity, measured as body weight and BMI, increases the risk of endometrial cancer, which is consistent with most previous studies. Our finding of a more pronounced association of endometrial cancer with body fat distribution, as measured by waist circumference and waist:hip ratio, is consistent with some (10, 1214) but not all (1, 19) previous studies. BMI was highly correlated with measurements of body fat distribution, especially with waist circumference, in our study. The high correlations between these indices may have lowered the statistical power for detecting an interaction between BMI and waist:hip ratio or waist circumference, making it difficult to disentangle the effect of overall adiposity from central fat deposition in this study population. Nevertheless, analyses with mutual adjustment of BMI and waist:hip ratio or BMI and waist circumference suggest that central adiposity may be a more important predictor of endometrial cancer risk.

Exposure of the endometrium to estrogen unopposed by progesterone plays a pivotal role in the etiology of endometrial cancer. It is generally believed that adiposity influences endometrial cancer risk through modification of levels of endogenous sex hormones (e.g., estrogen, progesterone). In premenopausal women, the main source of circulating estrogen is ovarian production. Adiposity has little effect on serum concentrations of estradiol, but it causes an increase in the frequency of anovular and irregular menstrual cycles, resulting in a reduction of luteal-phase progesterone levels and ultimately increased exposure to “unopposed estrogen” (25, 26). Among postmenopausal women, obese women have greater estrogen exposure due to increased aromatization of androgen to estrogen (2729). Moreover, obese women, regardless of their menopausal status, have lower sex hormone-binding globulin levels, which can result in higher levels of biologically available estrogens (30). A slightly higher odds ratio for the effect of BMI on endometrial cancer risk among postmenopausal women in our study is in line with this mechanism. Interestingly, we found that the association of obesity indices, particularly waist circumference, with endometrial cancer was much stronger among younger women than among older women—an effect that appeared to be independent of menopausal status. Studies focused on the timing of the onset of obesity and on genetic susceptibility would help in understanding the nature of this interaction.

The role of fat deposition in the development of cancer has received much attention recently. It has been suggested that body fat distribution may be a better marker than overall obesity for estimating breast cancer risk associated with sex hormone levels (31). Two major types of fat distribution—upper-body versus lower-body and central versus peripheral—have been evaluated for their association with health risk, including endometrial cancer. The former is usually measured by waist:hip ratio (1, 2, 914, 19) and the latter by subscapular:triceps skinfold ratio (9, 10) or truck:limb circumference ratio (2). However, biologic differences between these two types of fat distribution are not well understood, and the two terms have sometimes been used interchangeably in the literature. Upper-body/central obesity has been associated with an increased level of estrogen and a reduced level of sex hormone-binding globulin (1, 2, 30, 32). The more pronounced association of waist circumference, but not BMI, with endometrial cancer among women who had never used oral contraceptives in the current study appears to suggest an estrogenic effect of upper-body obesity on endometrial cancer. Among oral contraceptive users, the effects of high estrogen levels caused by central adiposity might be counteracted by a continuous supply of progesterone (33, 34). However, other mechanisms may also be implicated. Hyperinsulinemia and insulin resistance are strongly related to increases in intraabdominal body fat, particularly among premenopausal women (3538). Central obesity has been related to increased levels of insulin and insulin-like growth factor 1 (IGF-1) (3941). IGF-1 is a peptidic growth factor involved in the proliferation of a wide variety of cell types, especially endometrial epithelial cells (42, 43). Insulin may directly enhance tumor formation through endometrial insulin receptors or stimulate endometrial tumor development by reducing levels of insulin-like growth factor binding protein 1, resulting in increased IGF-1 activity (44). Evidence directly linking circulating IGF-1 concentrations to endometrial cancer risk is available (45). The positive association between diabetes and endometrial cancer risk observed in the current study and in many previous studies favors the insulin pathway hypothesis (46, 47).

Waist circumference has been reported to be a better indicator of intraabdominal fat content than waist:hip ratio (48, 49). In this study, a high waist circumference was significantly related to risk of endometrial cancer independently of BMI, and the positive association was also seen for women whose BMI was less than 23.7. Our results suggest that waist circumference may be a better predictor of endometrial cancer risk, and they support the recommendation to include measurement of waist circumference as a practical tool for assessing abdominal fat, in conjunction with measures of overall adiposity as indicated by BMI, in both clinical and research settings.

This study is unique in that the degree of obesity in China is much lower than in Western countries. The lack of extreme obesity allows for easier evaluation of the independent effect of central adiposity. In particular, it is much easier to measure and interpret waist circumference in this population than in Western populations, where a waistline for the extremely obese is difficult to define. Strengths of this study also include the population-based design, the high response rates, and the low frequency of hysterectomy. In addition, we adjusted for or excluded confounding from major risk factors for endometrial cancer, such as oral contraceptive use, hormone replacement therapy, family history, energy intake, and physical activity. Importantly, our primary exposures in this research were anthropometric measurements taken by trained interviewers who were retired nurses and physicians.

The major limitation of this study is that the anthropometric measurements were taken after disease diagnosis. The presence of tumors may have led to differential misclassification bias, either toward or away from the null. In our study, anthropometric measurements were taken within 6 months of diagnosis for 531 cases and beyond 6 months for 298. Analyses stratified by the timing of measurement showed few differences in risk estimates (odds ratios for the highest quartiles vs. the lowest for waist circumference and BMI were 4.8 and 2.8 for women interviewed within 6 months of diagnosis and 5.0 and 3.2 for women interviewed more than 6 months after diagnosis). Several earlier studies observed a more pronounced association of endometrial cancer with adiposity in premenopausal women (50, 51); the effect was attributed to polycystic ovary syndrome, a disease affecting 5–10 percent of women of reproductive age in Western countries and characterized by chronically elevated fasting and nonfasting serum insulin levels, and often central obesity (52). To evaluate the possible influence of polycystic ovary syndrome on our results, we excluded women who reported having polycystic ovary syndrome, and the results were unchanged. However, we appreciate that polycystic ovary syndrome may be a condition that is underdiagnosed.

In summary, we found that both overall adiposity and body fat distribution were strongly associated with endometrial cancer risk among Chinese women, with the latter being a stronger disease predictor. The association between central obesity and endometrial cancer was more pronounced among younger women and was seen for both normal-weight and obese women.

This work was supported by grant R01CA92585 from the US National Cancer Institute.

The authors thank Dr. Fan Jin for her contributions in implementing the study in Shanghai and Bethanie Hull for technical assistance in manuscript preparation. This study would not have been possible without the support of the research staff of the Shanghai Endometrial Cancer Study.

References

1.
Folsom AR, Kaye SA, Potter JD, et al. Association of incident carcinoma of the endometrium with body weight and fat distribution in older women: early findings of the Iowa Women's Health Study.
Cancer Res
 
1989
;
49
:
6828
–31.
2.
Austin H, Austin JM Jr, Partridge EE, et al. Endometrial cancer, adiposity, and body fat distribution.
Cancer Res
 
1991
;
51
:
568
–72.
3.
MacDonald PC, Siiteri PK. Relationship between extraglandular production of estrone and the occurrence of endometrial neoplasia.
Gynecol Oncol
 
1974
;
2
:
2159
–63.
4.
Judd HL, Lucas WE, Yen SS. Serum 17 beta-estradiol and estrone levels in postmenopausal women with and without endometrial cancer.
J Clin Endocrinol Metab
 
1976
;
43
:
272
–8.
5.
Key TJ, Pike MC. The dose-effect relationship between ‘unopposed’ oestrogens and endometrial mitotic rate: its central role in explaining and predicting endometrial cancer risk.
Br J Cancer
 
1998
;
57
:
205
–12.
6.
Dividson BJ, Gambone JC, Lagasse LD, et al. Free estradiol in postmenopausal women with and without endometrial cancer.
J Clin Endocrinol Metab
 
1981
;
52
:
404
–8.
7.
Olson SH, Trevisan M, Marshal JR, et al. Body mass index, weight gain, and risk of endometrial cancer.
Nutr Cancer
 
1995
;
23
:
141
–9.
8.
Ballard-Barbash R, Swanson CA. Body weight: estimation of risk for breast and endometrial cancers.
Am J Clin Nutr
 
1996
;
63
(suppl):
437s
–41s.
9.
Shu XO, Brinton LA, Zheng W, et al. Relation of adiposity and body fat distribution to endometrial cancer in Shanghai, China.
Cancer Res
 
1992
;
52
:
3865
–70.
10.
Swanson CA, Potischman N, Wilbanks GD, et al. Relation of endometrial cancer risk to past and contemporary body size and body fat distribution.
Cancer Epidemiol Biomarkers Prev
 
1993
;
2
:
321
–7.
11.
Petridou E, Belechri M, Dessypris N, et al. Leptin and body mass index in relation to endometrial cancer risk.
Ann Nutr Metab
 
2002
;
46
:
147
–51.
12.
Elliott EA, Matanoski GM, Rosenshein NB, et al. Body fat patterning in women with endometrial cancer.
Gynecol Oncol
 
1990
;
39
:
253
–8.
13.
Schapira DV, Kumar NB, Lyman GH, et al. Upper-body fat distribution and endometrial cancer risk.
JAMA
 
1991
;
266
:
1808
–11.
14.
Iemura A, Douchi T, Yamamoto S, et al. Body fat distribution as a risk factor of endometrial cancer.
J Obstet Gynaecol Res
 
2000
;
26
:
421
–5.
15.
La Vecchia C, Parazzini F, Negri E, et al. Anthropometric indicators of endometrial cancer risk.
Eur J Cancer
 
1991
;
27
:
487
–90.
16.
Henderson BE, Casagrande JT, Pike MC, et al. The epidemiology of endometrial cancer in young women.
Br J Cancer
 
1983
;
47
:
749
–56.
17.
Weiderpass E, Persson I, Adami HO, et al. Body size in different periods of life, diabetes mellitus, hypertension, and risk of postmenopausal endometrial cancer.
Cancer Causes Control
 
2000
;
11
:
185
–92.
18.
Tretli S, Magnus K. Height and weight in relation to uterine corpus cancer morbidity and mortality: a follow-up study of 570,000 women in Norway.
Int J Cancer
 
1990
;
46
:
165
–72.
19.
Lapidus L, Helgesson O, Merck C, et al. Adipose tissue distribution and female carcinomas: a 12-year follow-up of participants in the population study of women in Gothenburg, Sweden.
Int J Obes
 
1988
;
12
:
361
–8.
20.
Jin F, Devesa S, Chow W, et al. Cancer incidence trends in urban Shanghai, 1972–1994: an update.
Int J Cancer
 
1999
;
83
:
435
–40.
21.
Popkin BM. The nutrition transition and obesity in the developing world.
J Nutr
 
2001
;
131
(suppl):
871s
–3s.
22.
Breslow NE, Day NE. Statistical methods in cancer research. Vol 1. The analysis of case-control studies. (IARC scientific publication no. 32). Lyon, France: International Agency for Research on Cancer,
1980
.
23.
Mack TM, Pike MC, Henderson BE, et al. Estrogens and endometrial cancer in a retirement community.
N Engl J Med
 
1976
;
294
:
1262
–7.
24.
International Agency for Research on Cancer. Weight control and physical activity. (IARC handbooks of cancer prevention, vol 6). Lyon, France: International Agency for Research on Cancer,
2002
:112–15.
25.
Rogers J, Mitchell GW. The relation of adiposity to menstrual disturbances.
N Engl J Med
 
1952
;
247
:
53
–5.
26.
Key TJ, Pike MC. The dose-effect relationship between ‘unopposed’ oestrogens and endometrial mitotic rate: its central role in explaining and predicting endometrial cancer risk.
Br J Cancer
 
1988
;
57
:
205
–12.
27.
Kirschner MA, Samojlik E, Drejka M, et al. Androgen-estrogen metabolism in women with upper body versus lower body adiposity.
J Clin Endocrinol Metab
 
1990
;
70
:
473
–9.
28.
Evans DJ, Hoffmann RG, Kalkhoff RK, et al. Relationship of androgenic activity to body fat topography, fat cell morphology, and metabolic aberrations in premenopausal women.
J Clin Endocrinol Metab
 
1983
;
57
:
304
–9.
29.
Haffner SM, Stern MP, Hazuda HP, et al. Do upper-body and centralized adiposity measure different aspects of regional body-fat distribution?
Diabetes
 
1987
;
36
:
43
–51.
30.
Anderson DC. Sex hormone binding globulin.
Clin Endocrinol
 
1974
;
6
:
69
–72.
31.
Barnett JB, Woods MN, Rosner B, et al. Sex hormone levels in premenopausal African-American women with upper and lower body fat phenotypes.
Nutr Cancer
 
2001
;
41
:
47
–56.
32.
Boyapati SM, Shu XO, Gao YT, et al. Correlation of blood sex steroid hormones with body size, body fat distribution, and other known risk factors for breast cancer in post-menopausal Chinese women.
Cancer Causes Control
 
2004
;
15
:
305
–11.
33.
International Agency for Research on Cancer. Hormonal contraception and postmenopausal hormonal therapy. (IARC monographs on the evaluation of carcinogenic risks to humans, vol 72). Lyon, France: International Agency for Research on Cancer,
1998
.
34.
Weiderpass E, Adami HO, Baron JA, et al. Use of oral contraceptives and endometrial cancer risk (Sweden).
Cancer Causes Control
 
1999
;
10
:
277
–84.
35.
Peiris AN, Sothmann MS, Hennes MI, et al. Relative contribution of obesity and body fat distribution to alterations in glucose insulin homeostasis: predictive values of selected indices in premenopausal women.
Am J Clin Nutr
 
1989
;
49
:
758
–64.
36.
Kissebah AH, Vydelingum N, Murray R, et al. Relation of body fat distribution to metabolic complications of obesity.
J Clin Endocrinol Metab
 
1982
;
54
:
254
–60.
37.
Evans DJ, Hoffmann RG, Kalkhoff RK, et al. Relationship of body fat topography to insulin sensitivity and metabolic profiles in premenopausal women.
Metabolism
 
1984
;
33
:
68
–75.
38.
Krotkiewski M, Seidell JC, Bjorntorp P. Glucose tolerance and hyperinsulinaemia in obese women: role of adipose tissue distribution, muscle fiber characteristics and androgens.
J Intern Med
 
1990
;
228
:
385
–92.
39.
Sandhu MS, Gibson JM, Heald AH, et al. Association between insulin-like growth factor-I:insulin-like growth factor-binding protein-1 ratio and metabolic and anthropometric factors in men and women.
Cancer Epidemiol Biomarkers Prev
 
2004
;
13
:
166
–70.
40.
Gomez JM, Maravall FJ, Gomez N, et al. Interactions between serum leptin, the insulin-like growth factor-I system, and sex, age, anthropometric and body composition variables in a healthy population randomly selected.
Clin Endocrinol
 
2003
;
58
:
213
–19.
41.
Bjormtorp P. Body fat distribution, insulin resistance, and metabolic diseases.
Nutrition
 
1997
;
13
:
795
–803.
42.
Yakar S, Wu Y, Setser J, et al. The role of circulating IGF-I: lessons from human and animal models.
Endocr J
 
2002
;
19
:
239
–48.
43.
Pierro E, Minici F, Alesiani O, et al. Stromal-epithelial interactions modulate estrogen responsiveness in normal human endometrium.
Biol Reprod
 
2001
;
64
:
831
–8.
44.
Lee PD, Conover CA, Powell DR. Regulation and function of insulin-like growth factor-binding protein-1.
Proc Soc Exp Biol Med
 
1993
;
204
:
4
–29.
45.
Ayabe T, Tsutsumi O, Sakai H, et al. Increased circulating levels of insulin-like growth factor-I and decreased circulating levels of insulin-like growth factor binding protein-1 in postmenopausal women with endometrial cancer.
Endocr J
 
1997
;
44
:
419
–24.
46.
Parazzini F, Vecchia CL, Negri E, et al. Diabetes and endometrial cancer: an Italian case-control study.
Int J Cancer
 
1999
;
81
:
539
–42.
47.
Anderson KE, Anderson E, Mink PJ, et al. Diabetes and endometrial cancer in the Iowa Women's Health Study.
Cancer Epidemiol Biomarkers Prev
 
2001
;
10
:
611
–16.
48.
Van der Kooy K, Leenen R, Seidell JC, et al. Waist-hip ratio is a poor predictor of changes in visceral fat.
Am J Clin Nutr
 
1993
;
38
:
327
–33.
49.
Pouliot MC, Despres JP, Lemieux S, et al. Waist circumference and abdominal sagittal diameter: best simple anthropometric indexes of abdominal visceral adipose tissue accumulation and related cardiovascular risk in men and women.
Am J Cardiol
 
1994
;
73
:
460
–8.
50.
Gallup DG, Stock RJ. Adenocarcinoma of the endometrium in women 40 years of age and younger.
Obstet Gynecol
 
1984
;
64
:
417
–20.
51.
Evans-Metcalf ER, Brooks SE, Reale FR, et al. Profile of women 45 years of age and younger with endometrial cancer.
Obstet Gynecol
 
1998
;
91
:
349
–54.
52.
Chang RJ. Polycystic ovary syndrome: diagnostic criteria. In: Chang RJ, Heindel JJ, Dunaif A, eds. Polycystic ovary syndrome. New York, NY: Marcel Dekker, Inc,
2002
:361–5.