Portal de Periódicos da CAPES

Effect of body size on breast-cancer risk among Japanese women

1999; Wiley; Volume: 80; Issue: 3 Linguagem: Inglês

10.1002/(sici)1097-0215(19990129)80

1097-0215

Kaoru Hirose, Kazuo Tajima, Nobuyuki Hamajima, Toshiro Takezaki, Manami Inoue, Tetsuo Kuroishi, Shigeto Miura, Shinkan Tokudome,

Metabolism, Diabetes, and Cancer

International Journal of CancerVolume 80, Issue 3 p. 349-355 Human CancerFree Access Effect of body size on breast-cancer risk among Japanese women Kaoru Hirose, Corresponding Author Kaoru Hirose khirose@aichigw.aichi-cc.pref.aichi.jp Division of Epidemiology, Aichi Cancer Center Research Institute, Nagoya, JapanDivision of Epidemiology, Aichi Cancer Center Research Institute, 1-1 Kanokoden Chikusa-ku, Nagoya 464-8681, Japan. Fax: (81) 52-763-5233.Search for more papers by this authorKazuo Tajima, Kazuo Tajima Division of Epidemiology, Aichi Cancer Center Research Institute, Nagoya, JapanSearch for more papers by this authorNobuyuki Hamajima, Nobuyuki Hamajima Division of Epidemiology, Aichi Cancer Center Research Institute, Nagoya, JapanSearch for more papers by this authorToshiro Takezaki, Toshiro Takezaki Division of Epidemiology, Aichi Cancer Center Research Institute, Nagoya, JapanSearch for more papers by this authorManami Inoue, Manami Inoue Division of Epidemiology, Aichi Cancer Center Research Institute, Nagoya, JapanSearch for more papers by this authorTetsuo Kuroishi, Tetsuo Kuroishi Division of Epidemiology, Aichi Cancer Center Research Institute, Nagoya, JapanSearch for more papers by this authorShigeto Miura, Shigeto Miura Department of Breast Surgery, Aichi Cancer Center Hospital, Nagoya, JapanSearch for more papers by this authorShinkan Tokudome, Shinkan Tokudome Department of Public Health, Nagoya City University Medical School, Nagoya, JapanSearch for more papers by this author Kaoru Hirose, Corresponding Author Kaoru Hirose khirose@aichigw.aichi-cc.pref.aichi.jp Division of Epidemiology, Aichi Cancer Center Research Institute, Nagoya, JapanDivision of Epidemiology, Aichi Cancer Center Research Institute, 1-1 Kanokoden Chikusa-ku, Nagoya 464-8681, Japan. Fax: (81) 52-763-5233.Search for more papers by this authorKazuo Tajima, Kazuo Tajima Division of Epidemiology, Aichi Cancer Center Research Institute, Nagoya, JapanSearch for more papers by this authorNobuyuki Hamajima, Nobuyuki Hamajima Division of Epidemiology, Aichi Cancer Center Research Institute, Nagoya, JapanSearch for more papers by this authorToshiro Takezaki, Toshiro Takezaki Division of Epidemiology, Aichi Cancer Center Research Institute, Nagoya, JapanSearch for more papers by this authorManami Inoue, Manami Inoue Division of Epidemiology, Aichi Cancer Center Research Institute, Nagoya, JapanSearch for more papers by this authorTetsuo Kuroishi, Tetsuo Kuroishi Division of Epidemiology, Aichi Cancer Center Research Institute, Nagoya, JapanSearch for more papers by this authorShigeto Miura, Shigeto Miura Department of Breast Surgery, Aichi Cancer Center Hospital, Nagoya, JapanSearch for more papers by this authorShinkan Tokudome, Shinkan Tokudome Department of Public Health, Nagoya City University Medical School, Nagoya, JapanSearch for more papers by this author First published: 08 November 1999 https://doi.org/10.1002/(SICI)1097-0215(19990129)80:3 3.0.CO;2-8Citations: 41AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract With the use of data from the hospital-based epidemiologic research program at Aichi Cancer Center (HERPACC), the effect of body size on the risk of breast cancer was evaluated among Japanese women, who are generally leaner than white women. In total, 1,359 breast-cancer cases were included, and 24,207 women, confirmed as free of cancer, were recruited as a reference group. Odds ratios (OR) and 95% confidence intervals (95% CI) were determined by multiple-logistic regression analysis. Separate analyses were performed for pre- and post-menopausal women. Furthermore, stratification by decade of age was done to evaluate the effect of body size on the development of breast cancer. The results obtained from the present study were as follows. (1) Current body-mass index (BMI) was positively associated with post-menopausal breast cancer (OR 2.08, 95% CI 1.49–2.92 for highest quintile vs. lowest), although higher BMI did not affect the risk in pre-menopausal women. (2) Estimates of risk were below unity for BMI at around age 20 in post-menopausal women. (3) After stratifying BMI at around age 20, gaining BMI in later life was positively associated with increased risk, regardless of BMI in early life. These findings suggest that avoidance of marked weight gain during adult life, especially after natural menopause and/or after age 60, may reduce the risk of breast cancer. Int. J. Cancer 80:349–355, 1999. © 1999 Wiley-Liss, Inc. Although the incidence rates of female breast cancer in most Asian countries are much lower than those in Western countries (Parkin et al., 1992), there has been a marked increase in recent years. The age distribution of breast cancer in Japanese women is entirely different from that in Western countries (Moolgavkar et al., 1979), i.e., the age trend falls after menopause, whereas the age-dependent elevation of risk in pre-menopausal women is similar. In Japan, the westernization of lifestyle has progressed rapidly since the end of World War II. This remarkable change may be associated with the anthropometric characteristics and chronological variation of cancer mortality, but the extent of the influence of altered lifestyle may vary among age groups to a considerable degree. Information on the effects of body size on breast cancer in different decades of age may well have preventive implications. The present study, therefore, was designed to elucidate the influence of anthropometric indices on breast cancer, with particular emphasis on the differences and similarities between pre- and post-menopausal women. In order to explore the importance of adiposity at different periods of adult life, further analysis was conducted separately in various strata by decade of age. Many studies have examined the relationship between anthropometric indices and breast cancer. Breast cancer and body-mass index (BMI) have been extensively studied, and negative correlations among pre-menopausal women and positive correlations among post-menopausal women have been the predominant findings (Willett et al., 1985; Le Marchand et al., 1988; Vatten and Kvinnsland, 1990). It has also been argued that change in weight is an important anthropometric determinant of breast cancer (Ballard-Barbash, 1994). It has been suggested that adult weight gain is important, particularly in women with low BMI in early adulthood (Barnes-Josiah et al., 1995). Few studies have addressed the effect of BMI and weight change on breast-cancer risk among Asian women, who are generally shorter and leaner than women in the USA and other Western countries, where the incidence rates of female breast cancer are much higher (Swanson et al., 1989). To confirm the effects of body size on Japanese women, we undertook a case-referent study of breast cancer with the use of data from the hospital-based epidemiologic research program at Aichi Cancer Center (HERPACC), Japan. In this HERPACC study, we examined the influence of BMI on breast-cancer risk according to menopausal status and decade of age. We also investigated any association of early adult body size and subsequent weight gain with the development of breast cancer, by stratification of BMI at around 20 years old. MATERIAL AND METHODS Data collection Beginning in 1988, we conducted the HERPACC study, which comprised a self-administered questionnaire completed by first-visit outpatients to the Aichi Cancer Center Hospital (ACCH). All questionnaires were collected after they had been checked by a trained interviewer for incomplete responses. The data were processed using the computer system of the Aichi Cancer Center Research Institute. Details of the questionnaire and data-collection procedures have been described elsewhere (Yoo et al., 1992; Hirose et al., 1995; Inoue et al., 1997). Of all the first-visit outpatients, totaling 57,974 between January 1988 and June 1995, 4,564 patients were excluded due to interviewer absence, age (younger than 18 years), or visit for consultation only. The questionnaire was finally distributed to 53,410 patients. Among these, 52,591 (98.5%) returned adequate responses. The questionnaire included items on occupation, medical history, marital status, family history (parents and siblings), smoking and drinking habits, dietary habits, sleeping habits, physical exercise, bowel habits and reproductive history. Items on height, current weight and weight at around 20 years old were added since 1989. Questions on socio-economic status and educational level were not included, since Japanese are, in general, reluctant to answer such questions. Inquiries about these items were made prior to the investigation of symptoms, and all the information was collected before diagnoses were made. Cases and referents The data collected were linked with hospital-based cancer registry files. Our analysis was restricted to women aged 30 and over who visited hospitals between January 1989 and June 1995. Among 3,712 female cancer patients, 1,359 women who were first diagnosed, within 6 months after the first-visit, as having breast cancer by histological examination were recruited as the case group. The referents were 24,207 female first-visit outpatients who had never been diagnosed as having cancer. Table I shows the 24,207 referents and 1,359 cases by age group. Table I. AGE-ADJUSTED ODDS RATIOS (OR) AND 95% CONFIDENCE INTERVALS (CI) FOR BREAST CANCER BY SELECTED CHARACTERISTICS (HERPACC, 1989–1995) Characteristic Cases (n = 1,359) Referents (n = 24,207) Age-adjusted (95% CI) Number (%) Number (%) OR Age 30–39 162 (11.9) 5,020 (20.7) 40–49 535 (39.4) 9,322 (38.5) 50–59 349 (25.7) 5,833 (24.1) 60+ 313 (23.0) 4,032 (16.7) Mean (yr) 51.2 48.4 Mean body size Height (cm) 154.0 154.3 Weight (kg) 53.511 Significantly higher. 52.5 Weight at around age 20 (kg) 48.7 49.0 Change of weight (kg) 4.8011 Significantly higher. 3.47 Current BMI 22.511 Significantly higher. 22.0 BMI at around age 20 20.5 20.6 Change of BMI 2.0111 Significantly higher. 1.45 Marriage Yes 1,261 (93.9) 22,924 (95.5) 1.00 No 82 (6.1) 1,094 (4.5) 1.51 (1.20–1.91)**** p < 0.01. Menarche ≤12 285 (21.1) 5,324 (22.1) 1.00 13–14 624 (46.2) 11,729 (48.7) 0.87 (0.75–1.01) ≧15 442 (32.7) 7,042 (29.2) 0.83 (0.70–0.99)** p < 0.05; Mean (yr) 13.9 13.8 Pregnancy No 145 (10.7) 2,086 (8.6) 1.00 Yes 1,214 (89.3) 22,116 (91.4) 0.77 (0.64–0.92)**** p < 0.01. Parity 0 190 (14.0) 2,802 (11.6) 1.00 1 188 (13.9) 3,126 (12.9) 0.88 (0.71–1.08) 2 640 (47.2) 12,128 (50.2) 0.78 (0.66–0.92)**** p < 0.01. 3+ 339 (25.0) 6,115 (25.3) 0.73 (0.61–0.88)**** p < 0.01. Mean 1.89 1.96 Age at first full-term pregnancy ≤23 299 (25.7) 6,044 (28.3) 1.00 24–26 470 (40.3) 9,146 (42.9) 1.11 (0.95–1.28) ≧27 396 (34.0) 6,155 (28.8) 1.39 (1.19–1.62)**** p < 0.01. Mean (yr) 25.8 25.3 Age at natural menopause ≤47 86 (16.9) 1,343 (18.1) 1.00 48–52 300 (58.8) 4,543 (61.1) 1.02 (0.80–1.31) ≧53 124 (24.3) 1,544 (20.8) 1.21 (0.91–1.61) Mean (yr) 50.3 49.9 Family history (mother or sister) No 1,234 (92.6) 22,866 (96.4) 1.00 Yes 98 (7.4) 846 (3.6) 2.15 (1.73–2.67)**** p < 0.01. 1 Significantly higher. * p < 0.05; ** p < 0.01. Statistical analysis Logistic regression was used to obtain odds ratios (OR) and 95% confidence intervals (95% CI) as estimates of relative risk. Quintiles were defined according to the distribution of body size among reference subjects. To control potential confounding by established risk factors for breast cancer, we carried out multivariate analyses using a model including age, age at menarche, parity, age at first full-term pregnancy, and family history of breast cancer. Calculations were performed using the SAS (SAS Institute) LOGISTIC procedure. BMI was calculated as weight (in kg)/height (in m2), according to Quetelet's formula. Since menopausal status was identified as a modifier of body size, data were stratified by menopausal status to evaluate the impact of anthropometric indices. To compare the differences and similarities of the effects of body size in the younger and older generation, OR and 95% CI stratified by 10-year intervals were calculated. RESULTS Table I shows the distribution of breast-cancer cases and referents according to age and selected risk factors. Cases had a higher proportion of single status and lower parity, and were older at first full-term pregnancy. Compared with referents, women with breast cancer were more likely to report a family history of breast cancer. Breast-cancer cases weighed more than referents and had higher BMI. One way in which the effects of weight and height were assessed was by comparing OR among the 9 groups of weight/height combinations. OR in tall (≧159 cm) and heavy (≧57 kg) women were higher (OR 1.51, 95% CI 1.14–1.98) than that in short (≤150 cm) and lighter (≤47 kg) women (Fig. 1). Similarly, OR were calculated for 9 groups which weight at around age 20 years old/height combinations were compared. In women >45 kg at around age 20, the taller and the heavier ones had increased OR of breast cancer (Fig. 2). Figure 1Open in figure viewerPowerPoint Odds ratios of breast cancer in 9 groups according to height and weight. Figure 2Open in figure viewerPowerPoint Odds ratios of breast cancer in 9 groups according to height and weight at around age 20. BMI is a better measure of adiposity than weight. Table II showed OR of current BMI (C-BMI) in separate strata by menopausal status and decade of age. In the overall data set, OR was 1.42 for the highest vs. the lowest quintile of C-BMI. Among post-menopausal women, the OR was very high (OR 2.08, 95% CI 1.49–2.92 for highest vs. lowest quintile), while no statistically significant or persuasive trends were found in pre-menopausal women. OR for breast cancer were highly associated with C-BMI in women aged 60 and over. We observed slightly higher OR among women aged 30 to 39 years, whereas there was no evidence for an association between C-BMI and breast-cancer risk among middle-aged groups (40–49 and 50–59 years old). To examine the excessive risk for women older than 50 years, we sub-divided the top category of C-BMI at BMI ≧ 26.50, defined as obesity in Japan. Post-menopausal women with obesity had OR of 2.31 (95% CI 1.56–3.42) compared with women in the lowest quintile of C-BMI (<19.73). OR for obese women aged 50 to 59 years and aged 60 and over were 1.62 (95% CI 1.01–2.62) and 2.13 (95% CI 1.30–3.48) respectively. Table II. OR AND 95% CI FOR BREAST CANCER, BY QUINTILE OF BODY-MASS INDEX (BMI), ACCORDING TO MENOPAUSAL STATUS AND GENERATION Cases Referents OR11 Adjusted for age, age at menarche, number of births, age at first full-term pregnancy and family history. (95% CI) Number Mean Mean Number Mean Mean Quintile of height weight height weight current BMI (cm) (kg) (cm) (kg) All women 1 (<19.73) 195 155.0 44.3 4,428 155.4 44.9 1.00 2 (19.73–21.07) 232 153.6 48.3 4,411 155.1 49.1 1.19 (0.98–1.45) 3 (21.08–22.37) 237 154.6 52.0 4,457 154.6 51.9 1.17 (0.96–1.42) 4 (22.38–24.13) 257 154.3 55.4 4,384 154.0 55.0 1.23 (1.01–1.49)** p < 0.05; 5 (>24.13) 307 153.7 62.7 4,315 153.3 61.9 1.42 (1.18–1.72)**** p < 0.01. p for trend <0.001 Pre-menopausal women 1 ( 24.13) 114 155.5 63.8 2,167 154.6 62.8 1.11 (0.85–1.45) p for trend 0.53 Post-menopausal women 1 (<19.73) 48 152.9 42.4 1,002 152.6 42.9 1.00 2 (19.73–21.07) 66 151.7 47.1 1,013 152.4 47.5 1.45 (0.99–2.14) 3 (21.08–22.37) 66 152.5 50.6 1,198 152.3 50.5 1.21 (0.83–1.78) 4 (22.38–24.13) 101 153.5 54.9 1,388 152.2 53.8 1.58 (1.11–2.26)** p < 0.05; 5 (>24.13) 158 152.4 61.8 1,661 151.8 60.8 2.08 (1.49–2.92)**** p < 0.01. p for trend <0.001 Women aged 30–39 1 ( 24.13) 25 157.0 64.8 491 155.9 64.4 1.81 (1.06–3.08)** p < 0.05; p for trend 0.05 Women aged 40–49 1 ( 24.13) 88 155.5 64.1 1,583 154.4 62.6 0.96 (0.70–1.30) p for trend 0.81 Women aged 50–59 1 ( 24.13) 87 153.3 62.7 1,367 152.8 61.3 1.27 (0.84–1.90) p for trend 0.33 Women aged 60 and over 1 ( 24.13) 107 151.7 61.1 874 150.5 59.9 2.19 (1.45–3.31)**** p < 0.01. p for trend <0.001 1 Adjusted for age, age at menarche, number of births, age at first full-term pregnancy and family history. * p < 0.05; ** p < 0.01. OR of BMI in young adults, around 20 years old (20-BMI), are shown in Table III. Women with higher 20-BMI generally displayed lower OR. Estimates of risk were below unity for 20-BMI among post-menopausal women. No relationship was found between 20-BMI and breast cancer among women aged 30 to 39 years, although OR in the third quintile was significantly higher (OR 1.70, 95% CI 1.07–2.72). In contrast, higher 20-BMI lowered OR among women aged 40 to 49 and those aged 50 to 59 (OR 0.78, 95% CI 0.58–1.05 and OR 0.66, 95% CI 0.46–0.96, respectively, for the highest vs. the lowest quintile). When current mean weight was compared with the mean weight at age 20, weight gain (kg) in the lowest quintile groups was greatest, while subjects in the highest quintile group changed weight only slightly (Table III). Table III. OR AND 95% CI FOR BREAST CANCER, BY QUINTILE OF BODY-MASS INDEX (BMI) AT AROUND AGE 20, ACCORDING TO MENOPAUSAL STATUS AND GENERATION Cases Referents OR11 Adjusted for age, age at menarche, number of births, age at first full-term pregnancy and family history. (95% CI) Number Mean [A] − [B] Number [A] − [B] Mean weight Mean Mean weight Mean weight at age 20 Mean weight at age 20 Quintile of BMI height (kg) (kg) height (kg) (kg) at around age 20 (cm) [A] [B] (cm) [A] [B] All women 1 ( 22.26) 220 153.3 56.9 56.2 0.7 4,240 153.2 55.9 56.0 −0.1 0.76 (0.63–0.91)**** p < 0.01. p for trend 0.02 Pre-menopausal women 1 ( 22.26) 91 155.1 56.8 56.5 0.3 2,113 155.1 57.3 56.9 0.4 0.81 (0.62–1.06) p for trend 0.49 Post-menopausal women 1 ( 22.26) 101 151.9 57.2 56.0 1.2 1,688 151.1 54.3 55.1 −0.8 0.72 (0.53–0.98)** p < 0.05; p for trend 0.04 Women aged 30–39 1 (<18.75) 31 155.5 47.7 43.3 4.4 1,230 157.4 47.8 44.2 3.6 1.00 2 (18.75–19.89) 33 156.7 50.6 47.6 3.0 1,150 157.3 50.5 47.9 2.6 1.09 (0.66–1.80) 3 (19.90–20.87) 44 155.2 52.0 49.1 2.9 979 156.7 51.6 50.1 1.5 1.70 (1.07–2.72)** p < 0.05; 4 (20.88–22.26) 28 156.1 56.9 52.4 4.5 848 156.2 53.6 52.5 1.1 1.19 (0.70–2.00) 5 (>22.26) 20 155.7 56.7 56.9 −0.2 667 156.2 57.6 57.8 −0.2 1.06 (0.60–1.89) p for trend 0.56 Women aged 40–49 1 (<18.75) 128 155.6 50.0 43.3 6.7 1,852 155.6 49.7 43.1 6.6 1.00 2 (18.75–19.89) 94 156.1 52.0 47.2 4.8 1,840 155.7 52.0 47.0 5.0 0.73 (0.56–0.96)** p < 0.05; 3 (19.90–20.87) 106 155.3 54.6 49.2 5.4 1,878 154.9 53.1 49.0 4.1 0.80 (0.62–1.05) 4 (20.88–22.26) 101 155.2 54.7 51.9 2.8 1,799 154.6 54.1 51.5 2.6 0.81 (0.62–1.06) 5 (>22.26) 79 154.9 56.9 56.2 0.7 1,413 154.7 57.3 56.6 0.7 0.78 (0.58–1.05) p for trend 0.18 Women aged 50–59 1 ( 22.26) 53 153.4 58.4 56.3 2.1 1,126 152.7 55.9 55.5 0.4 0.66 (0.46–0.96)** p < 0.05; p for trend 0.05 Women aged 60 and over 1 ( 22.26) 68 150.5 55.8 55.8 0.0 1,034 149.7 52.9 54.5 −1.6 0.75 (0.50–1.13) p for trend 0.12 1 Adjusted for age, age at menarche, number of births, age at first full-term pregnancy and family history. * p < 0.05; ** p < 0.01. Adjusted OR for breast cancer by quintile of change in BMI after 20 years of age are presented in Table IV. When the second quintile group, that is, the no-change group, was referenced, the OR of women whose BMI had changed at the highest level of BMI (>3.55) was 1.34 (95% CI 1.11–1.61) in the overall data set. These higher OR were more pronounced among post-menopausal women than among pre-menopausal women, and a marked increasing trend of OR was observed among women aged 60 and over (OR 2.11, 95% CI 1.34–3.30 for highest vs. second quintile; p value for linear trend <0.001). In these analyses, we were more interested in the independent effect of weight gain: accordingly, data were dichotomized at the mean of 20-BMI (20-BMI 20.6), since the effect of adult weight gain may vary with early weight. Examination of the greatest change in BMI categories showed a general increase in the OR of breast cancer, but moderate change did not increase risk, irrespective of the 20-BMI level (Table V). Among women aged 60 and over with lower 20-BMI, the OR was 2.35 (95% CI 1.12–4.92) for the highest vs. the second quintile (p value for linear trend = 0.01). 20-BMI did not appear to be a modulator of the association between change in BMI and breast-cancer risk. Table IV. ADJUSTED OR11 Adjusted for age, age at menarche, number of births, age at first full-term pregnancy and family history. FOR BREAST CANCER BY QUINTILE (Q) OF CHANGE IN BODY-MASS INDEX (BMI) AFTER AGE 20 Change of BMI after age 20 p for trend Q1 Q2 Q3 Q4 Q5 <−0.83 −0.84–+0.77 +0.78–+1.91 +1.92–+3.54 >+3.55 Number OR Number OR Number OR Number OR Number OR of cases of cases of cases of cases of cases All women 211 0.86 199 1.0022 Reference group. 247 1.11 243 1.07 317 1.3433 95% CI does not include 1. p < 0.001 Pre-menopausal 132 1.06 103 1.00 150 1.13 136 1.08 130 1.15 0.39 women Post-menopausal 50 0.83 78 1.00 71 1.21 86 1.27 148 1.7533 95% CI does not include 1. p < 0.001 women Age group 30–39 40 0.92 24 1.00 46 1.45 27 1.21 19 1.27 0.15 40–49 100 0.94 74 1.00 108 1.00 113 0.99 113 1.11 0.37 50–59 44 0.69 41 1.00 60 0.97 58 0.79 95 1.05 0.12 60+ 27 0.96 60 1.00 33 1.12 45 1.39 90 2.1133 95% CI does not include 1. p < 0.001 1 Adjusted for age, age at menarche, number of births, age at first full-term pregnancy and family history. 2 Reference group. 3 95% CI does not include 1. Table V. ADJUSTED OR11 Adjusted for age, age at menarche, number of births, age at first full-term pregnancy and family history. FOR BREAST CANCER BY QUINTILE (Q) OF CHANGE IN BMI AFTER AGE 20, ACCORDING TO BMI AT AROUND AGE 20 Change of BMI after 20 years old Q1 Q2 Q3 Q4 Q5 p for trend <−0.83 −0.84–+0.77 +0.78–+1.91 +1.92–+3.54 >+3.55 BMI <20.6 at around age 20 All women 0.97 1.0022 Reference group. 1.12 1.07 1.3333 95% CI does not include 1. 0.02 Pre-menopausal women 1.11 1.00 1.15 1.15 1.17 0.48 Post-menopausal women 0.99 1.00 1.44 1.14 1.9833 95% CI does not include 1. 0.001 Age group 30–39 1.37 1.00 1.55 1.11 1.30 0.82 40–49 0.73 1.00 0.88 1.05 1.06 0.23 50–59 0.88 1.00 1.18 0.80 1.07 0.88 60+ 1.40 1.00 1.06 1.03 2.3533 95% CI does not include 1. 0.01 BMI ≧20.6 at around age 20 All women 0.83 1.00 1.08 1.06 1.27 0.002 Pre-menopausal women 1.03 1.00 1.08 0.90 1.09 0.99 Post-menopausal women 0.76 1.00 1.03 1.48 1.53 p < 0.001 Age group 30–39 0.64 1.00 1.24 1.47 1.18 0.08 40–49 1.08 1.00 1.19 0.77 1.09 0.62 50–59 0.63 1.00 0.73 0.77 1.02 0.17 60+ 0.88 1.00 1.20 1.76 1.9033 95% CI does not include 1. p < 0.001 1 Adjusted for age, age at menarche, number of births, age at first full-term pregnancy and family history. 2 Reference group. 3 95% CI does not include 1. DISCUS