Body size in relation to cancer of the uterine corpus in 1 million Norwegian women

Authors

  • Tone Bjørge,

    Corresponding author
    1. Department of Public Health and Primary Health Care, Section for Epidemiology and Medical Statistics, University of Bergen, N-5018 Bergen, Norway
    2. Division of Epidemiology, Norwegian Institute of Public Health, N-5018 Bergen/N-0403 Oslo, Norway
    • Department of Public Health and Primary Health Care, Section for Epidemiology and Medical Statistics, University of Bergen, N-5018 Bergen, Norway
    Search for more papers by this author
    • Fax: +47-55-58-61-30.

  • Anders Engeland,

    1. Department of Public Health and Primary Health Care, Section for Epidemiology and Medical Statistics, University of Bergen, N-5018 Bergen, Norway
    2. Division of Epidemiology, Norwegian Institute of Public Health, N-5018 Bergen/N-0403 Oslo, Norway
    Search for more papers by this author
  • Steinar Tretli,

    1. The Cancer Registry of Norway, Institute of Population-Based Cancer Research, Montebello, N-0310 Oslo, Norway
    Search for more papers by this author
  • Elisabete Weiderpass

    1. The Cancer Registry of Norway, Institute of Population-Based Cancer Research, Montebello, N-0310 Oslo, Norway
    2. Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, SE-171 77 Stockholm, Sweden
    Search for more papers by this author

Abstract

A positive association between overweight/obesity and endometrial cancer has been observed. It has been hypothesized that obesity is mostly associated with a subtype described as estrogen-dependent (Type I tumors), constituting about 80% of the endometrial tumors. Few epidemiologic studies have, however, analyzed different histological subtypes separately. The present study aimed at exploring the relations between body size and histological subtypes of cancer of the uterine corpus. Height and weight were measured in over 1 million Norwegian women aged 20–74 during 1963–2001. During follow-up, 9,227 cancers of the uterine corpus were diagnosed. The tumors were classified as Type I tumors (mostly endometrial adenocarcinomas with subgroups), Type II tumors (papillary, serous, and clear cell adenocarcinomas and some poorly differentiated carcinomas), sarcomas, and mixed tumors. Relative risks (RRs) of cancer of the uterine corpus were estimated using Cox proportional hazards regression. Compared with women with normal BMI, overweight and obese women had an overall RR of cancer of the uterine corpus of 1.36 (95% CI: 1.29–1.42) and 2.51 (95% CI: 2.83–2.66). The increase in risk was most pronounced for Type I tumors, but was also seen for Type II tumors, sarcomas and mixed tumors. The overall RR of corpus uteri cancer associated with a 10-cm increase in height was 1.09 (95% CI: 1.05–1.13), and was mostly observed for Type I tumors. © 2006 Wiley-Liss, Inc.

Cancer of the corpus uteri is the 7th most common cancer and the 13th cause of death from cancer in women worldwide (3.9% of cases and 1.7% of cancer deaths).1 The highest incidence rates are in North America and Europe. Rates are low in southern and eastern Asia and most of Africa. The incidence increases steadily with increasing age, and increasing incidence trends among postmenopausal women have been observed in many European countries.2 Five-year survival, however, is rather good, about 80% in European cancer registries.1 In Norway, the age-adjusted (world standard population) incidence rate per 100,000 person-years was 14.8 in 1999–2003.3 The lifetime risk for women to develop cancer of the uterine corpus was 2.2% (based on incidence rates in 1997–2001).

Several cancer forms, including pre- and postmenopausal endometrial cancer, have been associated with overweight/obesity.4 Overweight and obese women appear to be at a 2- to 3-fold increased risk. Adult weight gain seems to be a better predictor of risk than middle-age obesity, and has been reported to be associated with risk in a linear dose-dependent fashion.

Previous studies on the relationship between height and endometrial cancer risk have given conflicting results. Most studies have, however, found an increased risk with increasing height.5, 6, 7

Two major types of endometrial carcinomas have been described—Type I and Type II tumors.8, 9 Type I tumors, mostly endometrioid adenocarcinomas, have been described as estrogen-dependent, and are generally associated with endometrial hyperplasia, hyperlipidemia, obesity and with an indolent clinical course.10 Type II tumors, mainly serous and clear cell adenocarcinomas, have been described as less estrogen-dependent, arising from the atrophic endometrium in elderly women and having a more aggressive clinical behavior.

Type I tumors constitute the vast majority (about 80%) of the cancers of the uterine corpus,11 and therefore a large number of study subjects are necessary to compare differences in effect for the different histological groups. Most epidemiologic studies have analyzed all types of endometrial cancer as a single entity, rather than considered different histological types. To our knowledge, the differences between Type I and Type II tumors in relation to body size have not been studied in large prospective studies, although it has been hypothesized that obesity is mostly associated with Type I tumors. We present here results form a large prospective cohort study carried out in Norway, where we explore the associations between body size and cancers of the uterine corpus, with special emphasis on Type I and Type II tumors.

Material and methods

The study population and the methods used in the present study have been described previously.12

Study population

In several Norwegian health surveys carried out during 1963–2001, height and weight were measured in a standardized way in 1,038,018 women aged 20–74 years.

A major part of the health surveys in the period 1963–75 was included in a nationwide screening-program aimed at detecting tuberculosis.13 This screening was compulsory for all Norwegians aged 15 years and above, and the attendance was about 85%.

In 1963–64 and in 1972–2001, height and weight were also measured in other health surveys in different parts of Norway.14, 15 The main objective of these studies was to assess risk factors for coronary heart diseases. The attendance in the mid 1970s was 85–90%, but decreased to about 75% in the mid 1990s.15

Deaths, emigrations and cancers of the uterine corpus (International Classification of Diseases, seventh revision (ICD-7): 172) in this cohort were identified by linkage to the Cancer Registry of Norway3 and the Death Registry at Statistics Norway.16 These registries are population-based and cover the entire Norwegian population. A unique 11-digit identification number assigned to all individuals living in Norway after 1960, simplified the linkages.

In the present study, only histologically verified cancers were included. One thousand one hundred and nine women with a cancer diagnosis prior to the height and weight measurements were excluded. Follow-up started at the date of body size measurements and ended at the date of cancer diagnosis, emigration, age 100 years, death or December 31, 2003, whichever occurred first. Altogether, 1,036,909 women were eligible for inclusion in the study, and 32 women were lost during follow-up. By the end of follow-up, 64% of the women were alive without a diagnosis of cancer of the uterine corpus, 35% were dead and 1% had cancer of the uterine corpus.

The cancers of the uterine corpus were subclassified into the following histological groups: Type I tumors (endometrioid adenocarcinomas with subgroups and mucinous adenocarcinomas), Type II tumors (papillary, serous and clear cell adenocarcinomas, and some poorly differentiated carcinomas), sarcomas, mixed endothelial and mesenchymal tumors, and other tumors.9, 10, 11, 17

From the tuberculosis screening program during 1963–75, only data on height and weight measurements were available in addition to time of measurement, age and sex. Data on oral contraceptive (OC) use, age at menarche, use/type/duration of hormone replacement therapy (HRT), history of diabetes or hypertension and physical activity, i.e., known confounders of endometrial cancer risk, were not available in this database.

Most health surveys carried out during 1972–2001, however, included questions about smoking habits (self-administered questionnaire) as well. The study subjects were divided into never, former and current smokers. Information on smoking habits was available for 324,704 women after exclusions. The same exclusion criteria and follow-up procedure as in the main analysis were used for this subsample.

Statistical methods

Cox proportional hazards regression models,18 with time since measurement as the time variable, were fitted to obtain relative risk (RR) estimates of cancer of the uterine corpus. In the multivariate analyses, categorized variables for age at measurement, year of birth, BMI ((weight in kilograms)/(height in meters)2) and height were included. BMI was categorized using the WHO-categorization19: BMI < 18.5 (underweight), 18.5–24.9 (normal), 25.0–29.9 (overweight) and ≥30.0 (obese).

Analyses were performed, including units of BMI and height in 10-cm categories, respectively, as continuous variables, to test for trend in the risk of uterine corpus cancer.

Separate analyses were performed for the different histological groups. As we have no information on age at menopause for the study subjects, we used age 50 years as a proxy for age at menopause, and stratified some analyses according to ages 20–49 years and 50–74 years at body size measurement. Because we had no information on OC use or HRT, we performed separate analyses for women born before 1930 and in 1930 or later. Women born before 1930 were older at measurements (55 vs. 34 years), had higher BMI (26.5 vs. 23.6) and were lower (161 cm vs. 165 cm). The analyses were also stratified according to attained age.

The statistical program package SPSS20 was used for estimating RRs of uterine corpus cancer with 95% confidence intervals (CIs).

The hazard function of malignancy by BMI in the Cox model was estimated using penalized spline functions in S-plus,21 with 4 degrees of freedom. BMI was not categorized in these analyses, but cutpoints were defined in the estimation of splines by the limits between the categories used in the main analyses. The RR of point a compared with point b on the x-axis can be calculated by using the figure to find the corresponding y-values, ya and yb, and then calculate the RR by emath image.

Results

The 1,036,877 women included in this study were followed for an average of 25 years (range 0–41 years), constituting 25 million person-years (Table I). The mean age at body size measurement was 44 years, and 13% of the women were obese. During follow-up, 9,227 cancers of the uterine corpus were diagnosed among the study subjects. The mean age at diagnosis was 64.5 years for all cancers of the uterine corpus, and 64.6 years for Type I tumors, 65.2 years for Type II tumors, 58.5 years for sarcomas and 68.8 years for mixed tumors. The cases had their body size measurements taken on average 18 years prior to the cancer diagnosis.

Table I. Number of Observed Cases of Cancer of the Uterine Corpus (N), Person-Years (Person-yrs), and Overall Incidence Rates (INC. Rate)
VariableType IType IIAll
nInc. rate1nInc. rate1nPerson-yrsInc. rate1
  • 1

    Number of uterine cancer cases per 100,000 person-years.

Time since measurement
 0–45881224159585,097,73819
 5–98581926261,2474,620,47527
 10–141,1072714031,3764,028,94434
 15–191,119326021,3193,538,86537
 20–241,237407931,5103,116,76148
 25–291,1914410641,4872,696,75755
 ≥301,0645010451,3302,140,00862
Age at measurement
 20–29761144019385,593,46517
 30–391,5032910621,8345,265,19335
 40–492,4053427643,0387,008,26243
 50–591,6553733482,2144,437,40350
 60–697123018881,0102,409,53142
 70–7412824489193525,69337
Year of birth
 <190018126558258683,12438
 1900–19097272920181,0352,541,21041
 1910–19191,6863733272,2554,523,83850
 1920–19292,2104026652,7945,542,41050
 1930–19391,419299221,7014,820,97535
 1940–1949796164119955,107,83419
 ≥19501457501892,020,1569
BMI (kg/m2)
 <18.5821541101548,63018
 18.5–24.92,9602136633,83414,395,73327
 25.0–29.92,3613236953,0747,399,65542
 ≥30.01,7616125392,2182,895,53077
Height (cm)
 <15011130195139365,46238
 150–1592,0382935052,6617,018,81038
 160–1694,1182853645,31814,551,10637
 ≥170897278731,1093,304,16934
Total7,1642899249,22725,239,54837

The risk of cancer of the uterine corpus increased with increasing BMI (RR per unit of increase in BMI = 1.08; 95% CI: 1.08–1.09). Compared with women with normal BMI, overweight and obese women had an overall RR of cancer of the uterine corpus of 1.36 (95% CI: 1.29–1.42) and 2.51 (95% CI: 2.83–2.66) (Table II). Splitting the upper BMI-category into 3 revealed a consistent and strong increase in risk by increasing BMI. The RR in women with BMI above 40 compared to normal-weighted was 4.96 (95% CI: 4.26–5.78). Analyses stratified on age at measurement showed a marked increase in risk for obese compared to normal-weighted women at all ages (data not shown). The increase in risk was most pronounced for Type I tumors, but was also seen for Type II tumors, sarcomas and mixed tumors. The RR of Type I tumors per unit increase in BMI was 1.09 (95% CI: 1.09–1.10), while the similar RRs for Type II tumors, sarcomas and mixed tumors were 1.06 (95% CI: 1.05–1.08), 1.06 (95% CI: 1.04–1.08) and 1.06 (95% CI: 1.03–1.09).

Table II. Relative Risk (RR) of Cancer of the Uterine Corpus with 95% Confidence Intervals (CIs) Obtained in Cox Regression Analyses
VariableType IType IISarcomasMixed tumorsAll
RR95% CIRR95% CIRR95% CIRR95% CIRR95% CI
  • Age at measurement and birth cohort were included in the model in addition to either body mass index (BMI) or height.

  • 1

    BMI and height, respectively, were included as continuous variables.

BMI (kg/m2)
 <18.50.900.72–1.120.420.16–1.131.090.58–2.050.820.26–2.570.860.71–1.05
 18.5–24.91.00Referent1.00Referent1.00Referent1.00Referent1.00Referent
 25.0–29.91.391.32–1.471.261.09–1.461.220.99–1.501.481.14–1.921.361.29–1.42
 ≥30.02.722.56–2.901.941.64–2.301.881.46–2.411.971.44–2.712.512.38–2.66
 Test for trend1p < 0.001p < 0.001p < 0.001p < 0.001p < 0.001
Height (cm)
 <1500.990.82–1.190.830.52–1.310.370.12–1.170.720.27–1.950.910.77–1.08
 150–1590.930.88–0.990.960.84–1.100.750.61–0.930.910.71–1.180.920.88–0.97
 160–1691.00Referent1.00Referent1.00Referent1.00Referent1.00Referent
 ≥1701.141.06–1.231.100.88–1.390.950.73–1.241.190.83–1.701.111.04–1.19
 Test for trend1p < 0.001p = 0.1p = 0.21p = 0.30p < 0.001
Total7,164 cases992 cases521 cases316 cases9,227 cases

There was a modest increase in risk of cancer of the uterine corpus with increasing height (RR per 10 cm increase in height = 1.09; 95% CI: 1.05–1.13). The association with height was most evident for Type I tumors (Table II).

The association between cancer of the uterine corpus and overweight (for Type I tumors) and obesity (for both Type I and II tumors) was similar both in women aged 20–49 or 50–74 years at body size measurement. When body measurement was taken after age 50, overweight was not associated with risk of Type II tumors (Table III). The increase in risk with increasing height (Type I tumors) was only seen for women measured at the age of 20–49 years.

Table III. Relative Risk (RR) of Cancer of the Uterine Corpus with 95% Confidence Intervals (CIs) Obtained in Cox Regression Analyses
VariableType IType IIAll
RR95% CIRR95% CIRR95% CI
  1. Age at measurement and birth cohort were included in the model in addition to either body mass index (BMI) or height. Analysis stratified on age at measurement.

20–49 years
 BMI (kg/m2)
  <18.50.980.78–1.230.710.26–1.920.950.78–1.17
  18.5–24.91.00Referent1.00Referent1.00Referent
  25.0–29.91.401.31–1.501.451.17–1.801.371.29–1.46
  ≥30.02.702.49–2.932.061.56–2.732.542.36–2.73
 Height (cm)
  <1501.070.78–1.461.030.38–2.771.010.76–1.34
  150–1590.960.90–1.030.980.78–1.230.940.88–1.00
  160–1691.00Referent1.00Referent1.00Referent
  ≥1701.171.08–1.261.070.81–1.431.131.05–1.22
  Total4,669 cases422 cases5,810 cases
50–74 years
 BMI (kg/m2)
  <18.50.320.12–0.850.000.00–∞0.320.15–0.72
  18.5–24.91.00Referent1.00Referent1.00Referent
  25.0–29.91.381.25–1.531.110.91–1.361.321.21–1.44
  ≥30.02.692.42–2.981.771.44–2.192.422.21–2.64
 Height (cm)
  <1500.920.72–1.160.780.46–1.310.840.68–1.04
  150–1590.890.82–0.970.950.80–1.130.890.83–0.96
  160–1691.00Referent1.00Referent1.00Referent
  ≥1701.050.88–1.271.180.81–1.741.050.89–1.23
Total2,495 cases570 cases3,417 cases

The RR of cancer of the uterine corpus in obese compared with normal-weighted women increased with attained age (Table IV). The increasing RR was only observed for Type I tumors (data not shown).

Table IV. Relative Risk (RR) of Cancer of the Uterine Corpus with 95% Confidence Intervals (CIs) by Attained Age1 from Cox Regression Analysis, Adjusted for Birth Cohort and Age at Measurement
VariableAttained age (years)
40–4950–5960–69≥70
RR95% CIRR95% CIRR95% CIRR95% CI
  • 1

    Only 66 cases were diagnosed at the age of 30–39 years, and none at the age of 20–29 years.

BMI (kg/m2)
 <18.51.160.77–1.760.990.75–1.310.540.32–0.910.610.33–1.14
 18.5–24.91.00Referent1.00Referent1.00Referent1.00Referent
 25.0–29.91.140.96–1.351.291.17–1.411.421.30–1.551.401.29–1.54
 ≥30.01.751.39–2.212.111.88–2.372.562.32–2.842.772.52–3.04
Height (cm)
 <1502.321.33–4.021.040.70–1.550.850.61–1.190.770.59–0.98
 150–1590.820.68–1.000.960.88–1.060.930.86–1.020.890.83–0.96
 160–1691.00Referent1.00Referent1.00Referent1.00Referent
 ≥1700.890.73–1.071.191.06–1.321.171.03–1.321.070.93–1.24
Total803 cases2,580 cases2,715 cases3,063 cases

In women born in 1930 or later, a higher RR for Type II tumors was found for women with BMI above 30. A slightly weaker positive association between height and Type I tumors was found.

The hazard functions of the risk of cancer of the uterine corpus by BMI and height were estimated using spline functions (Fig. 1). The association between the RR and BMI was nearly loglinear, while the association between the RR and height was nearly loglinear above the height of 150 cm. The slope was clearly steepest for the association with BMI.

Figure 1.

The logarithm of the hazard function of cancer of the uterine corpus, with 95% CIs (dashed lines), by BMI from spline functions with 4 degrees of freedom. Adjusted for birth cohort and age at measurement.

To exclude the possibility that weight was influenced by the presence of an undiagnosed cancer of the uterine corpus at the time of the BMI measurement, we repeated all analysis excluding all person-years and cancer cases diagnosed during the first 5 years of follow-up. Similar results as presented earlier were found.

In the subsample with known smoking habits, analyses were performed for women born before 1930 (Table V). The current smokers had a reduced risk of cancer of the uterine corpus compared with never smokers (RR = 0.67, 95% CI = 0.47–0.97). No major differences were found for the associations between BMI and height and cancer of the uterine corpus in the subanalyses compared with the main analyses. In the subsample, there were too few Type II tumors to perform meaningful analyses.

Table V. Relative Risk (RR) of Cancer of the Uterine Corpus with 95% Confidence Intervals (CIs) Obtained in Cox Regression Analyses
 Type ITotal
RR95% CIRR95% CI
  • Age at measurement and birth cohort were included in the model in addition to smoking habits, body mass index (BMI) and height. Only persons born before 1930 with known smoking habits.

  • 1

    BMI and height, respectively, were included as continuous variables.

Smoking
 Never1.00Referent1.00Referent
 Former1.140.77–1.681.050.73–1.51
 Current0.540.35–0.830.670.47–0.97
BMI (kg/m2)
 <18.50.000.00–∞0.000.00–∞
 18.5–24.91.00Referent1.00Referent
 25.0–29.91.811.28–2.551.831.34–2.49
 ≥30.02.261.52–3.342.411.65–3.33
 Test for trend1p < 0.001p < 0.001
Height (cm)
 <1500.980.40–2.820.750.31–1.84
 150–1590.940.69–1.270.900.68–1.18
 160–1691.00Referent1.00Referent
 ≥1701.220.71––2.101.080.65–1.80
 Test for trend1p = 0.2p = 0.2
Total193 cases242 cases

Discussion

In the present study, a Norwegian cohort of more than 1 million women was followed prospectively with regard to the risk of cancer of the uterine corpus for an average of 25 years. More than 9,200 cancer cases were observed. The risk of cancer of the uterine corpus increased with BMI, most pronounced for Type I tumors, but also observed for Type II tumors, sarcomas and mixed tumors. There was a moderately increasing risk with increasing height for Type I tumors mostly.

A weakness of the present study is the lack of control for possible confounders as nulliparity, diabetes, intake of OCs and HRT.22 Because we had no information on OC use or HRT, we performed separate analyses for women born before 1930. Similar results as in the main analysis were found. The use of OC and HRT was probably low in these women. OCs were not introduced in Norway until 1967, and from another Norwegian cohort study of women born between 1927 and 1965, it is known that about 90% of the women aged 60–70 at the start of follow up (1991–97) had never used OCs.23 Previously, Norwegian women were rather restrictive in their use of HRT.24 During 1985–88, less than 6% of postmenopausal women aged 40–54 years, in some Norwegian counties, was current users of HRT.25 However, these patterns changed somewhat in the 90s. In a Norwegian study, including more than 30,000 postmenopausal women aged 45–64 years, 35% was current users of HRT (1996–98).26 Among the users of estrogen–progestagen preparations, there was no increased risk of endometrial cancer.

Tobacco is carcinogenic to humans, and is an important cause of different cancers as lung cancer and cancer of the urinary bladder.27 Nevertheless, smoking reduces the risk of endometrial cancer through its effect on estrogen production and metabolism.22 In the present study, information on smoking habits was available for a subcohort only. Current smokers had a reduced risk of cancer of the uterine corpus when compared with never smokers. No major differences were found for the associations between BMI and height and cancer of the uterine corpus in the sub analyses when compared with the main analyses for women born before 1930.

In Europe, about 50% of the adult population is either overweight or obese, and the proportion of endometrial cancers attributed to obesity in European women has been estimated to be about 40%.28 In Norway, the BMI in the female population has increased since the late 1970s, approaching an obesity prevalence of 12.5% in the late 1990s.29 Excess weight is thought to influence the risk of endometrial cancer through changes in endogenous hormone metabolism, including sex steroids, insulin and insulin-like growth factors.11, 30 It has also been speculated about the possible role of adipocytokines (leptin and adiponectin) in the regulation of carcinogenesis.31, 32, 33 Most case-control studies on endometrial cancer risk and BMI have found a 200–400% linear increase in risk in persons with BMI above 25.30 Cohort studies have also found a similar increased risk, although of a smaller magnitude. Studies examining the association between obesity and endometrial cancer risk, separately for pre- and postmenopausal women, have found some higher risk for older women.4 Further, adult weight gain appears to be a stronger predictor of risk. In the present study, the association between BMI, height and cancer of the uterine corpus was most pronounced in women measured at the age of 20–49. The cancer risk increased steadily with increasing BMI, although the RR was highest in older age groups (attained age). Besides, measurements were taken at one point in time only, so that changes in BMI over time could not be assessed.

In the last 2 decades, 2 major types of endometrial carcinomas have been described, based mostly on clinicopathologic and molecular genetic studies. Type I tumors are mostly endometrioid, are estrogen dependent and develop through the hyperplasia-carcinoma sequence.8, 9, 10, 11, 17, 34 Type II tumors are mostly of serous or clear cell histology, are less estrogen dependent and generally develop from atrophic endometrial tissue in elderly women. Factors associated with unopposed estrogenic stimulation, such as obesity, have been hypothesized as related to the development of Type I tumors mostly.11 The only known risk factor for Type II tumors described so far is age.10 In the present study, the risk of both Type I and Type II tumors as well as sarcomas and mixed tumors increased steadily with increasing BMI, although the risk increase was strongest for Type I tumors. Further, the mean ages at diagnosis for the Type I and Type II tumors were virtually the same (65 years).

Previously, Tretli and Magnus35 have shown that both BMI and height were risk factors for cancer of the uterine corpus, using parts of the present data set. The present study is, however, an extension both with regard to the number of persons and length of follow-up (extended from 13.5 to 25 million person-years), and included 4 times as many cancer cases. Further, we focused now on the possible differences between Type I and Type II tumors in relation to body size. As we expected, we found a stronger positive association between obesity and the risk of Type I tumors compared to other histological types. Some authors10, 17 have suggested that Type II tumors may not be estrogen-dependent, but our study clearly contradicts this theory.

A large number of studies have reported a positive weak association between height and cancer risk. In a systematic review on height and cancer risk from 2001 by Gunnell et al., positive associations with endometrial cancer were reported in 3 out of 6 cohort studies, and in only 3 of 17 case-control studies.5 In the present study, only a modest increase in risk of cancer of the uterine corpus was seen with increasing height for Type I tumors mostly.

In the present study, we found an association both to height and BMI, telling us that it may be a combination of these 2 measures that is optimal in predicting uterine corpus cancer. In the literature, the body surface area has been proposed as such a factor. We estimated the body-surface area using a method suggested by Du Bois and Du Bois as weight0.425 × height0.725 × 71.84.36 Body surface area was portioned into quintiles. The risk of cancer of the uterine corpus (all histology groups) increased with increasing body-surface area.

In summary, we observed an increasing risk of cancer of the uterine corpus with increasing BMI, most pronounced for Type I tumors, but also for Type II tumors, sarcomas and mixed tumors. A modest increase in risk was seen with increasing height for Type I tumors.

Acknowledgements

We are grateful to those who during almost 40 years collected the data used in the present study. These are persons connected to the former National Health Screening Service, The Nord-Trøndelag Health Survey (HUNT), The Hordaland Health Survey (HUSK) and The Tromsø Study. We are also grateful to Dr. Liane Deligedish (Mount Sinai Hospital, New York), who kindly assisted us in the classification of the cancers of the uterine corpus.

Ancillary