SEARCH

SEARCH BY CITATION

Keywords:

  • breast cancer;
  • body weight at age 20;
  • weight change;
  • risk;
  • estrogen receptor;
  • progesterone receptor

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Few prospective studies have investigated the association between BMI at age 20 years (BMI20y) and breast cancer risk with consideration to estrogen/progesterone receptor status (ER/PR). We evaluated the association between BMI20y and ER/PR-defined breast cancer risk among 41,594 women in the population-based Japan Public Health Center-based Prospective Study. Anthropometric factors were assessed using self-reported questionnaires. Relative risks (RRs) were estimated by Cox proportional hazards regression models. Through to the end of 2006, 452 breast cancer cases were identified. We observed a statistically significant inverse association between BMI20y and breast cancer incidence [multivariable-adjusted RR for each 5-unit increment 0.75 (95%CI = 0.61–0.92)], which was not modified by menopausal or recent BMI status. In contrast, recent BMI and subsequent BMI gain were not associated with increased risk among premenopausal women, but were substantially associated with increased risk among postmenopausal women [corresponding RRrecent BMI = 1.31 (95%CI = 1.07–1.59); RRsubsequent BMI gain = 1.32 (95%CI = 1.09–1.60)]. In subanalyses by receptor status (∼50% of cases), the observed inverse association of BMI20y with risk was consistent with the result for ER–PR– [0.49 (95%CI = 0.27–0.88)], while the observed positive associations of BMI gain with postmenopausal breast cancer risk appeared to be confined to ER+PR+ tumors [corresponding RRfor subsequent BMI gain =2.24 (95%CI = 1.50–3.34)]. Low BMI at age 20 years was substantially associated with an increased risk of breast cancer. In contrast, high recent BMI and subsequent BMI gain from age 20 were associated with increased risk of postmenopausal ER+PR+ tumors.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Despite the lower prevalence of obesity in Japan than Western countries,1 the incidence rate of breast cancer in this country has increased rapidly for a quarter of a century, and this cancer is now the most prevalent malignancy among women.2 A national survey has identified a high overall prevalence of leanness rather than obesity, particularly among younger generations, and more than 20% of young Japanese female adults in their 20s and 30s are underweight.3, 4 Further, a recent nationwide cross-sectional survey showed that young female adults became thinner at an early life-stage.5 In contrast, the prevalence of overweight among women tends to increase as age exceeds 50 years.3

A number of epidemiological studies have reported that both early adult body weight6–16 and a subsequent change in body weight6, 7, 9, 10, 12, 15–19 are associated with breast cancer risk. Several of these have reported an inverse association between body weight in early adulthood and the incidence of breast cancer.6, 10, 12, 18, 19 Almost all these previous studies were conducted in Western populations, however, in which the prevalence of obesity is high. This largely explains why the proposed biological mechanism for this inverse association involves a decrease in levels of estradiol20 due to premenopausal obesity, including anovulatory disorder. However, the Nurses' Health Study (NHS) II reported that the observed inverse association of BMI in early adulthood with risk was not eliminated after adjustment for ovulatory disorders,14 suggesting the presence of other biological mechanisms apart from anovulation.

As an alternative, we hypothesized that a certain level of body fat in the mammary gland (i.e., mammary gland fat pad) might be essential to healthy differentiation in breast tissue,21 particularly in early adulthood. Lean BMI might be an epidemiological indicator of a low level of fat tissue in the mammary gland, associated with an increased risk of breast cancer in later life resulting from the interruption of healthy differentiation in maturation in the breast in young adult women.

In this study, we prospectively investigated the impact of relative body weight at age 20 years (BMI20y) on the development of breast cancer among 41,594 Japanese women, with a relatively low prevalence of obesity, in the Japan Public Health Center-based Prospective Study (JPHC Study). We also evaluated the association of recent BMI and a subsequent change in BMI from age 20 years with breast cancer risk.

Material and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Study population

The JPHC Study has been described in detail elsewhere.22 The cohort was started in 1990 to evaluate the association between lifestyle factors and cancer and cardiovascular disease in the Japanese population. The study population consisted of all Japanese aged 40–59 years in Cohort I (the Iwate-Ninohe, Akita-Yokote, Nagano-Saku, Okinawa-Chubu, Tokyo-Katsushika public health center (PHC) areas) and 40–69 years in Cohort II (the Ibaraki-Mito, Niigata-Nagaoka, Kochi-Chuohigashi, Nagasaki-Kamigoto, Okinawa-Miyako and Osaka-Suita PHC areas) who were enrolled in the residential registries. Initially, 140,420 subjects were invited to the JPHC cohort, of whom 71,698 were female. For this study, subjects from one PHC area (Tokyo-Katsushika; n = 4,178) were excluded due to a lack of complete information on cancer incidence. A total of 55,907 women completed the baseline questionnaire (response rate 82.8%). All eligible cohort members received two further follow-up questionnaires for the 5-year (1995–1998; response rate 79.4%) and 10-year follow-up surveys (2000–2003; response rate 77.4%).

We excluded ineligible subjects (n = 21), women who moved before the start of follow-up or who could not be followed (n = 48), and those with a self-reported history of cancer before the start of follow-up (n = 1,509). In this study, we excluded women with missing or unreliable information on current BMI or BMI at age 20 (<14 or ≥40) (n = 10,146), alcohol drinking status, smoking and leisure-time physical activity (n = 1,954); women with a family history of breast cancer at baseline (n = 215); and those who reported unreasonable estimates of total energy intake (±3SD) (n = 420). In this study, we defined menopausal status based on information from self-reported questionnaires, which asked subjects to describe menstrual bleeding in the three classifications of (i) yes, natural; (ii) no, natural menopause; and (iii) no, surgical menopause. Postmenopausal women were asked about age at menopause; if this information was not available (0.053% of the cohort), we considered those aged over 56 years at administration of the questionnaire as postmenopausal, since ∼99% of subjects had stopped menstruating before this age. The final study cohort consisted of 41,594 women.

Exposure measurement

Information on weight and height was assessed through self-reported questionnaires in the baseline and 5- and 10-year follow-up surveys, while that on weight at age 20 years was collected in the baseline and 10-year follow-up surveys. In the baseline questionnaire, however, the question on weight at age 20 years was not included for Cohort I, so that we were unable to obtain any information on BMI20y among 22,273 women, or 53.5% of the study cohort. In the 10-year follow-up survey, in contrast, all questionnaires included an inquiry about weight at age 20, with responses received from 36,880 women (88.7%). Accordingly, we mainly used information from the 10-year follow-up survey, supplemented by that obtained at baseline.

Relative body weight was evaluated by body mass index (BMI), calculated as the weight in kilograms divided by the square of height in meters (kg/m2). We previously reported a high correlation between self-reported and measured BMI in a subgroup of the JPHC study (Spearman rank correlation coefficient r = 0.9).23 BMI20y was also calculated as weight at age 20 years in kilograms divided by the square of height in meters (kg/m2). Reproducibility of self-reported BMI20y was assessed by comparison of baseline and 10-year follow-up survey information for those who answered both questionnaires in the JPHC cohort, giving a Spearman correlation coefficient of 0.81.

The change in BMI from age 20 to recent age was calculated as the difference between BMI at recent age and that at age 20, updated with the respective questionnaire cycle. Relative risks (RRs) according to ER/PR-defined tumor status were estimated by including exposure information in the model as a continuous variable, and presented per 5 kg/m2 increment.

Information on other lifestyle-related factors, such as reproductive information (i.e., parity, age at first birth, age at menarche, age at menopause), alcohol drinking status, and smoking status, was also collected using a self-reported questionnaire at the baseline survey and updated by the respective follow-up surveys, if available.

In the JPHC study, dietary information was accessed using a validated FFQ at baseline,24 and in the 5- and 10-year follow-up surveys. In the present analyses, however, we used dietary information from the baseline survey only, because the number of food items in the 5-year and 10-year follow-up FFQs differed from that in the FFQ in the baseline survey.

Ascertainment of breast cancer cases and follow-up of the cohort

Breast cancer incident cases were identified by active patient notification from major local hospitals in the study area and data linkage with population-based cancer registries, with permission from the local governments responsible for the registries. Breast cancer cases were defined as codes C500–509 in accordance with the Third Edition of the International Classification of Diseases for Oncology.25 Eight cases (1.8% of cases) were identified through information on death certificates (i.e., Death Certificate Notification), of which 5 (1.1% of cases) had no information on diagnosis (i.e., Death Certificate Only). Diagnosis was microscopically verified for 97% of all cases. ER and PR status were evaluated by either immunohistochemical assay or enzyme-linked immunoassay. The cut-off point for positivity for ER and PR in breast tumors was decided by clinical estimation at the hospital treating the case or as specified by the assay method at the clinical laboratory performing the assay.

We started follow-up on the date of administration of the baseline questionnaire. Participants contributed person-time from baseline to the date of diagnosis of breast cancer, date of death, date of moving away from the study area, or end of follow-up (Dec 31st, 2006), whichever occurred first. Date of death was verified through linkage with death registries at the PHCs, which are required by the Ministry of Health, Labour and Welfare. Date of moving was verified through linkage with the residential registries at the regional PHCs.

Statistical analysis

To estimate relative risks (RRs) and 95% confidence intervals (CIs), we used a time-dependent multivariate Cox proportional hazards regression model with age as the time scale.26 The proportional hazards assumptions were verified using Kaplan-Meier curves.27 In primary analyses, women were subdivided into five categories (BMI20y and recent BMI: <18.5, 18.5–19.9, 20–23.9, 24–28.9, ≥29 kg/m2; with the cut-off point of 18.5 based on the WHO classification; 20 as a recommended cut-off point for international comparison;28 and 24 as overweight and 29 as obesity for Japanese populations, in accordance with the WHO expert consultation28). For BMI20y, however, because the prevalence of obesity was too low to analyze (1% at age 20 years), we divided women into four categories (<18.5, 18.5–19.9, 20–23.9, ≥24) in the final analyses. According to the change in BMI from age 20 years to recent age, women were also subdivided into four groups, as follows: loss (<–2.5 BMI units), maintain (≥–2.5 to <2.5 BMI units), gain (≥2.5 to <5 BMI units) and major gain (≥5 BMI units). In the main analysis, we adjusted for age (time-scale), area, age at menarche (≤13, 14, 15, ≥16 years, missing), age at first birth (nulliparous, <26, ≥26 years, missing), parity (nulliparous, 1–2, 3, ≥4 children, missing), menopausal status (premenopausal, age at menopause ≤48, 48–53, ≥54 years), use of exogenous female hormones (EFH) (never, ever, missing), smoking status (never, ever), leisure-time physical activity (no or 1–3 days/month, >1 days/week, 3–4 days/week, every day), alcohol intake (past-drinker, never-drinker, occasional drinker and regular drinker ≤150 or regular drinker >150 g of ethanol/week), total energy-adjusted intake of green-yellow vegetables (quintiles), total energy-adjusted intake of meat and meat products (quintiles) and total energy-adjusted intake of isoflavones (quintiles) as potential confounders on the basis that these covariates were likely associated with risk,29–31 and correlated with the exposures of interest. Trend tests were performed using a continuous value of exposure in the model.

We assessed the association of BMI20y, recent BMI, and change in BMI from age 20 years to recent age with breast cancer incidence with stratification by menopausal status at baseline survey, by BMI at age 20 (<20 or ≥20) or recent BMI (<25 or ≥25), and by use of EFH (never- or ever-use).

Cross-product terms of these factors and BMI at 20 years, recent BMI or change in BMI were introduced into the Cox proportional hazards regression model. The P-value for interaction was calculated by a likelihood ratio test which compared models with and without the interaction terms. All analyses were performed using the PROC PHREG procedure of the SAS statistical package version 9.1 (SAS Institute, Cary, NC). All statistical tests were two-sided, and statistical significance was defined as p < .05.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

After an approximate average of 14 years' follow-up, corresponding to 581,934 person-years, 452 invasive breast cancer cases were identified among 41,594 women.

Baseline characteristics of the study population are shown in Table 1. Compared to those with a high BMI20y, women with a low BMI20y were more likely to be younger, have a lower BMI, have fewer children, have a high intake of meat products, a low intake of isoflavones, and a higher prevalence of smoking and alcohol drinking. Women who gained BMI (≥5 units) tended to have a lower BMI20y, higher BMI, be younger at first birth, have more children, have a high intake of meat products and green-yellow vegetables, a low intake of isoflavones, and a higher EFH than women who lost BMI (<−2.5 units BMI).

Table 1. Baseline characteristics according to category of BMI at age 20 years and change in BMI from age 20 to recent age among 41,594 women in the Japan Public Health Center-based Prospective Study, Cohort I (1990-) and Cohort II (1993-)
inline image

Evaluation of the association between BMI20y and incidence of breast cancer revealed an inverse association [multivariable-adjusted RR for each 5-unit increment for BMI20y = 0.75 (95%CI = 0.61–0.92); Table 2].

Table 2. Multivariable relative risks (RRs)1 and 95% confidence intervals (CIs) for the association between relative body weight at age 20 years and breast cancer risk with stratification by menopausal status and BMI at the time of questionnaires (recent BMI) over 581,934 person-years in 41,594 women in the Japan Public Health Center-based Prospective Study, 1990–2006
inline image

In analyses stratified by menopausal status, the observed inverse association was similar across menopausal status (Pinteraction = 0.48; Table 2).

In stratification by level of recent BMI, RRs for the association between BMI20y and breast cancer incidence between the nonoverweight (recent BMI <24) and overweight groups (recent BMI ≥24) were not statistically heterogeneous (Pinteraction = 0.64; Table 2).

In this study, women with major weight gain (over 5 units BMI) were more likely to have a low body weight at age 20 years. To evaluate whether the observed inverse association of low BMI20y with risk was attributable to the impact of subsequent BMI gain, we performed subgroup analysis among 31,705 women who did not gain more than 5 units BMI. The results also support our observed substantial inverse association [multivariable-adjusted RR for each 5-unit increment for BMI20y= 0.73 (95%CI =0.57–0.93); text only].

Recent BMI was not associated with breast cancer risk among premenopausal women [multivariable-adjusted RR for each 5-unit increment = 1.02 (95%CI =0.81–1.27); Table 3], but was positively associated with increased risk among postmenopausal women [1.31 (95%CI =1.07–1.59); Table 3]. We also observed a statistically significant positive trend among women with BMI20y ≥20 (Ptrend 0.016). However, there was no evidence for effect modification by these factors (Pinteraction for menopausal status= 0.61; Pfor BMI20y=0.82).

Table 3. Multivariable relative risks (RRs) and 95% confidence intervals (CI) for the association of BMI at the time of the questionnaires (recent BMI) and change in BMI from age 20 years to recent age in relation to breast cancer risk with stratification by menopausal status as well as level of BMI at age 20 years over 581,934 person-years in the Japan Public Health Center-based Prospective Study, 1990–2006
inline image

Similarly, change in BMI from age 20 to recent age was not associated with breast cancer risk among premenopausal women [multivariable-adjusted RR for increase in each 5-unit increment = 1.04 (95%CI =0.84–1.30)], but was statistically significantly associated with increased risk among postmenopausal women [corresponding multivariable-adjusted RR= 1.32 (95%CI =1.09–1.60); Pinteraction = 0.042; Table 3]. This observed positive association among postmenopausal women was not modified by BMI20y level (<20 vs. ≥20) (Pinteraction = 0.31; Table 3).

In analyses stratified by EFH use among postmenopausal women, the observed inverse association between BMI20y and breast cancer risk was not modified by EFH use (Pinteraction = 0.69; Table 4). Substantial positive associations of recent BMI and subsequent BMI gain from age 20 years with the development of postmenopausal breast cancer were confined to never-users of EFH. However, there was no statistical evidence for effect modification by EFH use (Pinteractions for recent BMI = 0.28; Pfor change in BMI = 0.77; Table 4).

Table 4. Multivariable relative risks (RRs) and 95% confidence intervals (CIs) for the association of BMI at age 20 years, recent BMI, and change in BMI from age 20 to recent age with the incidence of breast cancer stratified by use of exogenous female hormones among 23,708 postmenopausal women with information on the use of exogenous female hormones in the Japan Public Health Center-based Prospective Study, 1990–2006
inline image

With regard to ER/PR status, information about joint ER/PR status was available for 211 cases. Among these, 94 (45% of known cases) were ER+PR+, 45 (21%) were ER+PR−, and 60 (28%) were ER−PR−. The number of ER−PR+ tumor cases (n = 12) was too small to allow separate analyses.

We performed subanalyses by receptor status in ∼50% of cases. Our finding of an overall inverse association of BMI20y with the incidence of breast cancer was not consistent for ER+PR+ tumors [RR for each 5-unit increment = 1.10 (95%CI = 0.71–1.70)], but was consistent for ER—PR— tumors [RR for ER—PR— = 0.49 (95%CI = 0.27–0.88); Table 5].

Table 5. Multivariable relative risks (RRs) and 95% confidence intervals (CIs) for the association of BMI at age 20 years and change in BMI (per increment of 5 kg/m2) with the risk of breast cancer defined by estrogen and progesterone receptor status in the Japan Public Health Center-based Prospective Study, 1990–2006
inline image

In contrast, the positive association of BMI gain from age 20 years to recent age with the development of postmenopausal breast cancer was consistent with the results for ER+PR+ [RR for each 5-unit increment =2.24(95%CI = 1.50–3.34)], but not for other tumor subtypes. These results for the association of recent BMI with the risk of ER/PR-defined breast cancer were consistent with those for the change in BMI from age 20 years (text only).

Compared to the stable BMI group (i.e., range of BMI change from –2.5 to <2.5), an ∼2.4 times' higher increase in risk for ER+PR+ tumors was observed among postmenopausal women who gained BMI ≥5 (RR = 2.44: 95%CI = 1.10–5.40; Ptrend = 0.0002; text only).

Because this study used information on BMI20y mainly from the 10-year follow-up survey, we performed sensitivity analyses using information mainly from the baseline survey, supplemented by that from the 10-year follow-up survey. These analyses gave similar results. Further, risk estimates for further sensitivity analyses based on a statistical model with height were also similar to those in Table 2 [multivariable-adjusted RR for increase in each 5-unit increment = 0.75 (95%CI =0.61–0.93) text only].

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

To our knowledge, this is the first large population-based prospective cohort study in Japan to evaluate the association between BMI20y and the incidence of ER/PR-defined breast cancer. Our observed inverse association was consistent with three prospective cohort14, 16, 32 and four case-control studies,7, 8, 11, 15 but not with others.33 Several studies6, 12, 15 have suggested that this inverse association is more pronounced among younger/premenopausal than older/postmenopausal women, but this was not fully consistent with the present and previous results.7, 11 In our cohort, age at baseline was ≥40 years, and thus follow-up did not completely cover the premenopausal period.

With regard to ER/PR status, NHS II14 reported that the association with BMI at age 18 years was strongest for ER+ [hazard ratio≥25 vs. 20–22.4 0.76] but their corresponding result for ER– was similar. Further, the most recent study (including NHS I and II) suggested that the inverse association between adolescent body fatness and breast cancer risk was stronger for ER– than ER+ tumors.34 The main contribution to our inverse association appeared to derive from ER–PR– tumors, but ER/PR status was verified in fewer than half of the cases, and this result should therefore be interpreted with caution.

Regarding change in BMI from age 20, our null association among premenopausal women was consistent with several studies.6, 18 It has been reported that weight gain from age 18 years was inversely associated with premenopausal breast cancer risk, but that this association was attenuated by adjustment for BMI at enrollment.16 Among postmenopausal women, our finding of a substantial positive association agrees well with most6, 7, 12, 17, 19 but not all previous studies.9, 10

Since women who gained BMI (≥5 units) from age 20 tended to have a lower BMI20y, our inverse association between BMI20y and risk might have been partly enhanced by the longitudinal amplitude of weight gain among lean women in early adulthood. When the analysis was restricted to women who maintained BMI (amplitude –2.5 to +2.5 units), the inverse association appeared attenuated, although this could be explained by lower power due to stratification. The lack of effect modification by BMI20y is consistent with a previous report.19 Further, we observed two contrasting results, the inverse association of BMI20y with ER–PR– tumor incidence and positive association of BMI gain from age 20 years with postmenopausal ER+PR+ tumors. These associations therefore appear independent, albeit that receptor information was limited.

A meta-analysis with ER/PR status35 agreed with our finding of a substantial positive association between recent BMI and ER+PR+ postmenopausal breast cancer risk. Further, our finding of a substantial positive association among EFH never-users is consistent with previous studies.13, 36 These results might indirectly support the validity of our information on BMI, EFH use and ER/PR status. Meanwhile, the observed inverse association between BMI20y and breast cancer risk was not modified by EFH use and was not consistent with one previous report.34

Plausible explanations for the biological mechanism underlying the inverse association between BMI20y and breast cancer risk include irregular menstruation and anovulation due to premenopausal obesity. These conditions might decrease exposure to ovarian hormones.37 The inverse trend in our results for BMI20y ≥20, and in previous epidemiological studies among Western populations might be explained by this premenopausal overweight/obesity-related (decreased risk) mechanism.14

However, our finding among Japanese women, who have a low prevalence of overweight (overweight (9%) or obesity (0.65%) in our cohort), may suggest a nonobesity-related mechanism, because the inverse association was found not only for those over 20 (i.e., BMI20y ≥20) but also those below 20 (i.e., BMI20y <20). The inverse trend might thus be explained in two dimensions, namely obesity-related (i.e., decreased risk) and lean-related (i.e., increased risk) biological mechanisms.

Plausible lean-related mechanisms include various vital roles of the mammary fat pad in normal mammary gland morphogenesis,21, 38 possibly in close conjunction with other hormones, such as estrogens and progesterone.39 Low BMI in early adulthood might indirectly indicate an insufficient mammary fat pad or progesterone deficiency, since progesterone may stimulate body fat deposition.40 Incomplete differentiation in early adulthood due to either or both factors might predispose to breast cancer in later life.21, 41 Progression stage of mammary epithelial cells from undifferentiated ER-negative mammary stem cells to differentiated cells may be linked to tumor subtypes.42

In contrast, our finding for a positive association between recent BMI, BMI gain from age 20 years and postmenopausal ER+PR+ breast cancer risk could be explained by classic estrogen-dependent mechanism.43 After menopause, the major source of endogenous estrogens shifts from the ovary to body fat44 due to increased endogenous estrogen production by aromatization of androgens in peripheral fat tissue.45 The obscure impact of BMI on postmenopausal breast cancer risk among EFH ever-users in our results might be explained by a stronger impact of EFH use on the risk than postmenopausal endogenous estrogen of body fat-origin.46, 47

Several limitations warrant consideration. Some measurement error was inevitable, because exposure information was evaluated by self-reported weight values, which tend to be underreported.23, 48 In particular, information of body weight at age 20 years was obtained retrospectively. Nevertheless, BMI20y at baseline and at 10-year follow-up survey was highly correlated, supporting tolerable reproducibility.49 Receptor status misclassification due to different assay methods or interlaboratory variation is also possible, although good agreement between immunohistochemical assay and enzyme-linked immunoassay50 has been reported. Possible selection bias due to the high percentage of unknown cases should be also considered. However, results for unknown ER/PR tumors were similar to overall results, suggesting the unlikelihood of any marked selection bias.

Major strengths of our study are its prospective population-based cohort design and large sample size. Three repeated exposure assessments of change in BMI from age 20 may have reduced misclassification due to the long follow-up. The prospective cohort study design meant that recall bias was rarely encountered, because exposure information was collected before diagnosis. If present, any misclassification of exposure was likely nondifferential, and would likely have moved the results toward the null (i.e., move RR closer to 1). Further, the biological plausibility of a positive association between BMI gain and postmenopausal ER+PR+ tumors indirectly supports the validity of the data.

In summary, BMI in early adult life was inversely associated with breast cancer incidence in a Japanese population. This inverse association was partly attributable to increased risk due to leanness at age 20 years. In contrast, a subsequent BMI gain from age 20 was substantially positively associated with postmenopausal ER+PR+ tumors. Optimum weight for breast cancer prevention might change with women's life-stage. Further epidemiological study of the generalizability of our results to other populations is required.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The authors thank all the staff members in each study area and in the central offices for their cooperation and technical assistance. They also thank the Iwate, Aomori, Ibaraki, Niigata, Osaka, Kochi, Nagasaki and Okinawa Cancer Registries for their provision of incidence data.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Murakami Y, Miura K, Ueshima H. [Comparison of the trends and current status of obesity between Japan and other developed countries]. Nippon Rinsho 2009; 67: 24552.
  • 2
    Matsuda T, Marugame T, Kamo KI, Katanoda K, Ajiki W, Sobue T. Cancer incidence and incidence rates in Japan in 2003: based on data from 13 population-based Cancer Registries in the monitoring of cancer incidence in Japan (MCIJ) project. Jpn J Clin Oncol 2009; 39: 8508.
  • 3
    National Health and Nutrition Examination Survey. Current status of the nations Health and Nutrition in Japan, 2004 edn. Tokyo, Japan: Daiich Publisher, 2006.
  • 4
    Takimoto H, Yoshiike N, Kaneda F, Yoshita K. Thinness among young Japanese women. Am J Public Health 2004; 94: 15925.
  • 5
    Funatogawa I, Funatogawa T, Yano E. Do overweight children necessarily make overweight adults? Repeated cross sectional annual nationwide survey of Japanese girls and women over nearly six decades. BMJ 2008; 337: a802.
  • 6
    Paffenbarger RS, Jr, Kampert JB, Chang HG. Characteristics that predict risk of breast cancer before and after the menopause. Am J Epidemiol 1980; 112: 25868.
  • 7
    Brinton LA, Swanson CA. Height and weight at various ages and risk of breast cancer. Ann Epidemiol 1992; 2: 597609.
  • 8
    Willett WC, Browne ML, Bain C, Lipnick RJ, Stampfer MJ, Rosner B, Colditz GA, Hennekens CH, Speizer FE. Relative weight and risk of breast cancer among premenopausal women. Am J Epidemiol 1985; 122: 73140.
  • 9
    Coates RJ, Uhler RJ, Hall HI, Potischman N, Brinton LA, Ballard-Barbash R, Gammon MD, Brogan DR, Daling JR, Malone KE, Schoenberg JB, Swanson CA. Risk of breast cancer in young women in relation to body size and weight gain in adolescence and early adulthood. Br J Cancer 1999; 81: 16774.
  • 10
    de Vasconcelos AB, Azevedo e Silva Mendonca G, Sichieri R. Height, weight, weight change and risk of breast cancer in Rio de Janeiro, Brazil. Sao Paulo Med J 2001; 119: 626.
  • 11
    Hirose K, Tajima K, Hamajima N, Takezaki T, Inoue M, Kuroishi T, Miura S, Tokudome S. Effect of body size on breast-cancer risk among Japanese women. Int J Cancer 1999; 80: 34955.
  • 12
    Ziegler RG, Hoover RN, Nomura AM, West DW, Wu AH, Pike MC, Lake AJ, Horn-Ross PL, Kolonel LN, Siiteri PK, Fraumeni JF, Jr. Relative weight, weight change, height, and breast cancer risk in Asian-American women. J Natl Cancer Inst 1996; 88: 65060.
  • 13
    Huang Z, Hankinson SE, Colditz GA, Stampfer MJ, Hunter DJ, Manson JE, Hennekens CH, Rosner B, Speizer FE, Willett WC. Dual effects of weight and weight gain on breast cancer risk. JAMA 1997; 278: 140711.
  • 14
    Michels KB, Terry KL, Willett WC. Longitudinal study on the role of body size in premenopausal breast cancer. Arch Inter Med 2006; 166: 2395402.
  • 15
    Trentham-Dietz A, Newcomb PA, Storer BE, Longnecker MP, Baron J, Greenberg ER, Willett WC. Body size and risk of breast cancer. Am J Epidemiol 1997; 145: 10119.
  • 16
    Weiderpass E, Braaten T, Magnusson C, Kumle M, Vainio H, Lund E, Adami HO. A prospective study of body size in different periods of life and risk of premenopausal breast cancer. Cancer Epidemiol Biomarkers Prev 2004; 13: 11217.
  • 17
    Eliassen AH, Colditz GA, Rosner B, Willett WC, Hankinson SE. Adult weight change and risk of postmenopausal breast cancer. JAMA 2006; 296: 193201.
  • 18
    Lahmann PH, Schulz M, Hoffmann K, Boeing H, Tjonneland A, Olsen A, Overvad K, Key TJ, Allen NE, Khaw KT, Bingham S, Berglund G, et al. Long-term weight change and breast cancer risk: the European prospective investigation into cancer and nutrition (EPIC). Brit J Cancer 2005; 93: 5829.
  • 19
    Barnes-Josiah D, Potter JD, Sellers TA, Himes JH. Early body size and subsequent weight gain as predictors of breast cancer incidence (Iowa, United States). Cancer Causes Control 1995; 6: 1128.
  • 20
    Potischman N, Swanson CA, Siiteri P, Hoover RN. Reversal of relation between body mass and endogenous estrogen concentrations with menopausal status. J Natl Cancer Inst 1996; 88: 7568.
  • 21
    Sakakura T, Nishizuka Y, Dawe CJ. Capacity of mammary fat pads of adult C3H/HeMs mice to interact morphogenetically with fetal mammary epithelium. J Natl Cancer Inst 1979; 63: 7336.
  • 22
    Tsugane S, Sobue T. Baseline survey of JPHC study—design and participation rate. Japan public health center-based prospective study on cancer and cardiovascular diseases. J Epidemiol 2001; 11: S249.
  • 23
    Tsugane S, Sasaki S, Tsubono Y. Under- and overweight impact on mortality among middle-aged Japanese men and women: a 10-y follow-up of JPHC study cohort I. Int J Obes Relat Metab Disord 2002; 26: 52937.
  • 24
    Tsubono Y, Takamori S, Kobayashi M, Takahashi T, Iwase Y, Iitoi Y, Akabane M, Yamaguchi M, Tsugane S. A data-based approach for designing a semiquantitative food frequency questionnaire for a population-based prospective study in Japan. J Epidemiol 1996; 6: 4553.
  • 25
    World Health Organization. International classification of diseases for oncology, 3rd edn. Geneva, Switzerland: World Health Organization, 2000.
  • 26
    Korn EL, Graubard BI, Midthune D. Time-to-event analysis of longitudinal follow-up of a survey: choice of the time-scale. Am J Epidemiol 1997; 145: 7280.
  • 27
    Collett D. Modelling survival data in medical research, texts in statistical scienceed. New York: Chapman & Hall/CRC, 1993.
  • 28
    WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 2004; 363: 15763.
  • 29
    Iwasaki M, Inoue M, Otani T, Sasazuki S, Kurahashi N, Miura T, Yamamoto S, Tsugane S. Plasma isoflavone level and subsequent risk of breast cancer among Japanese women: a nested case-control study from the Japan public health center-based prospective study group. J Clin Oncol 2008; 26: 167783.
  • 30
    Iwasaki M, Otani T, Inoue M, Sasazuki S, Tsugane S. Role and impact of menstrual and reproductive factors on breast cancer risk in Japan. Eur J Cancer Prev 2007; 16: 11623.
  • 31
    Hanaoka T, Yamamoto S, Sobue T, Sasaki S, Tsugane S. Active and passive smoking and breast cancer risk in middle-aged Japanese women. Int J Cancer 2005; 114: 31722.
  • 32
    Li HL, Gao YT, Li Q, Liu DK. [Anthropometry and female breast cancer: a prospective cohort study in urban Shanghai]. Zhonghua Liu Xing Bing Xue Za Zhi 2006; 27: 48893.
  • 33
    Kolonel LN, Nomura AM, Lee J, Hirohata T. Anthropometric indicators of breast cancer risk in postmenopausal women in Hawaii. Nutr Cancer 1986; 8: 24756.
  • 34
    Baer HJ, Tworoger SS, Hankinson SE, Willett WC. Body fatness at young ages and risk of breast cancer throughout life. Am J Epidemiol 2010; 171: 118394.
  • 35
    Suzuki R, Orsini N, Saji S, Key TJ, Wolk A. Body weight and incidence of breast cancer defined by estrogen and progesterone receptor status-A meta-analysis. Int J Cancer 2009; 124: 698712.
  • 36
    Feigelson HS, Jonas CR, Teras LR, Thun MJ, Calle EE. Weight gain, body mass index, hormone replacement therapy, and postmenopausal breast cancer in a large prospective study. Cancer Epidemiol Biomarkers Prev 2004; 13: 2204.
  • 37
    Rich-Edwards JW, Goldman MB, Willett WC, Hunter DJ, Stampfer MJ, Colditz GA, Manson JE. Adolescent body mass index and infertility caused by ovulatory disorder. Am J Obstet Gynecol 1994; 171: 1717.
  • 38
    Sakakura T, Nishizuka Y, Dawe CJ. Mesenchyme-dependent morphogenesis and epithelium-specific cytodifferentiation in mouse mammary gland. Science 1976; 194: 143941.
  • 39
    Shyamala G. Progesterone signaling and mammary gland morphogenesis. J Mammary Gland Biol Neoplasia 1999; 4: 89104.
  • 40
    Kalkhoff RK. Metabolic effects of progesterone. Am J Obstet Gynecol 1982; 142: 7358.
  • 41
    Russo J, Hu YF, Silva ID, Russo IH. Cancer risk related to mammary gland structure and development. Microsc Res Tech 2001; 52: 20423.
  • 42
    Prat A, Perou CM. Mammary development meets cancer genomics. Nat Med 2009; 15: 8424.
  • 43
    Key TJ, Appleby PN, Reeves GK, Roddam A, Dorgan JF, Longcope C, Stanczyk FZ, Stephenson HE, Jr, Falk RT, Miller R, Schatzkin A, Allen DS, et al. Body mass index, serum sex hormones, and breast cancer risk in postmenopausal women. J Natl Cancer Inst 2003; 95: 121826.
  • 44
    Siiteri PK. Adipose tissue as a source of hormones. Am J Clin Nutr 1987; 45: 27782.
  • 45
    Cleland WH, Mendelson CR, Simpson ER. Effects of aging and obesity on aromatase activity of human adipose cells. J Clin Endocrinol Metab 1985; 60: 1747.
  • 46
    Tworoger SS, Missmer SA, Barbieri RL, Willett WC, Colditz GA, Hankinson SE. Plasma sex hormone concentrations and subsequent risk of breast cancer among women using postmenopausal hormones. J Natl Cancer Inst 2005; 97: 5952.
  • 47
    Hankinson SE, Willett WC, Manson JE, Colditz GA, Hunter DJ, Spiegelman D, Barbieri RL, Speizer FE. Plasma sex steroid hormone levels and risk of breast cancer in postmenopausal women. J Natl Cancer Inst 1998; 90: 12929.
  • 48
    Elgar FJ, Stewart JM. Validity of self-report screening for overweight and obesity. Evidence from the Canadian community health survey. Can J Public Health 2008; 99: 4237.
  • 49
    Must A, Willett WC, Dietz WH. Remote recall of childhood height, weight, and body build by elderly subjects. Am J Epidemiol 1993; 138: 5664.
  • 50
    Elledge RM, Green S, Pugh R, Allred DC, Clark GM, Hill J, Ravdin P, Martino S, Osborne CK. Estrogen receptor (ER) and progesterone receptor (PgR), by ligand-binding assay compared with ER, PgR and pS2, by immuno-histochemistry in predicting response to tamoxifen in metastatic breast cancer: a Southwest Oncology Group Study. Int J Cancer 2000; 89: 1117.