A possible major contributor to the rising incidence of breast cancer in the developed world is the increasing prevalence of obesity in these countries (Flegal et al., 1998;Popkin and Doak, 1998). Positive energy balance has been implicated in mammary carcinogenesis since the 1940s when Tannenbaum demonstrated that the incidence of breast cancer tumours increased when mice were given diets high in total calories and dietary fat (Tannenbaum, 1942). Following these early experimental studies, de Waard and colleagues postulated in 1960 that overnutrition and obesity were key determinants of postmenopausal breast cancer (de Waard et al., 1960). Since that time, numerous epidemiological investigations have examined the association between anthropometric factors and breast cancer. The evidence for a role of these factors in breast cancer aetiology is becoming clearer and stronger, although uncertainties remain regarding these associations. Given that few modifiable risk factors have been identified for breast cancer and because of the established health benefits for weight control and maintenance for numerous other chronic diseases, the public health need and possible benefit for clarification on the nature of the associations between obesity and breast cancer is very high.
Given this background, the aim of this review is to provide a comprehensive overview of the state of scientific evidence for the association between anthropometric risk factors and breast cancer. The specific objectives of this literature review are: (1) to identify and systematically review the epidemiological literature for the association between height, weight, body mass index, fat deposition patterns, weight change, and breast size and breast cancer risk; (2) to highlight gaps in scientific knowledge for the association of anthropometric factors and breast cancer risk; (3) to identify the possible underlying biologic mechanisms; and (4) to recommend areas for future research.
The scientific literature on the association between anthropometric factors and breast cancer was reviewed. A search was conducted on Medline and Pubmed for all publications on weight, height, body mass index, anthropometric factors and breast cancer in human populations and was supplemented by hand searching of all major relevant journals. No restrictions on language or year of publication were made. The literature search included all publications up to October 2000. To be included in this review, the studies needed to have focused on some aspect of anthropometric risk factors in relation to breast cancer risk. Recently written major review papers were identified (Hunter and Willett, 1993;Ballard-Barbash, 1994;Ursin et al., 1995;Ballard-Barbash and Swanson, 1996;Ziegler, 1997;Pujol et al., 1997;Schindler, 1997;Carroll, 1998;Cold et al., 1998;McTiernan, 2000), in addition to a few other papers that examined specific aspects of the relation between anthropometric factors and breast cancer (Stoll, 1992, 1995, 1998, 1999). This report updates the literature review since the most comprehensive recent review included studies published to 1996 (Cold et al., 1998) and expands on previous reviews by identifying gaps in current knowledge and suggesting avenues for future investigation.
An individual's body build is not a simple construct that can be described in one variable. Rather, build is represented by a complex interaction of various measures and related metabolic activities that are each associated with specific aspects of breast cancer risk (Cold et al., 1998). The variables most commonly used in epidemiological studies to capture the anthropometric profile of the study subjects are: height, weight, adiposity as assessed by body mass index (BMI) (weight/height 2 ), fat deposition patterns as defined by waist–hip ratio (WHR), weight change, including weight gain and weight loss over lifetime and weight cycling, and breast size. Each of these anthropometric characteristics has been assessed as an independent risk factor for breast cancer; however, several are closely linked. Hence, when examining the aetiologic role of these factors, consideration must be given to how these factors are related to each other.
Two main methodological issues arise in epidemiological studies of anthropometric factors: (1) how the measurements were taken; and (2) when in life these measurements were taken. Generally, studies have relied on self-reported measures of height, weight, and hip and waist circumferences and have been restricted to self-reports for recalled weights over lifetime. More recent investigations have used interviewers to measure these parameters directly from the study participants, thereby reducing the possibility of random and systematic measurement error. For retrospective studies, the current height and weight data obtained during data collection are relatively valid, provided that the breast cancer cases are interviewed soon after their diagnosis and that they have not experienced any postdiagnosis, treatment-induced weight change. However, overweight study subjects have been found to underestimate and the short and underweight tend to overestimate. Studies investigating associations between disease with height and weight using self-reported measures will underestimate the effects (Gunnell et al., 2000).
For prospective studies, another major problem can arise when anthropometric data taken at baseline are used to predict risk of disease occurring several years after enrolment. In those studies that do not measure these factors again during the follow-up, severe bias can occur since changes in body weight and adiposity occur over lifetime (Williamson et al., 1990). In women, the rate of increase of weight or central adiposity changes at defined life periods including menarche, pregnancy and menopause.
Some of the inconsistencies across studies and some of the lack of associations found between these anthropometric measures and breast cancer risk can be attributed to errors in measurement and to changes in these measures that occur over the respondents’ lifetimes that are not appropriately assessed. Another source of inconsistency across these studies is the definition used for the cut-off points for measures of obesity and central adiposity. No direct comparisons can be made across studies on these factors since definitions particular to each study population have been used to categorize the study respondents into the quantiles of the anthropometric characteristics.
There are two methods for decreasing inconsistency between studies. The first is to design separate studies prospectively and to pool the data at the analysis stage. If such planning was not undertaken, then the second method is to pool the original data from epidemiological studies and re-analyse them using common definitions for all variables. The Pooling Project in Diet and Cancer (Hunter et al., 1996) has used this latter method since it combined and re-analysed the original, individual-level data from seven prospective cohort studies conducted worldwide. The data set includes 337 819 women and 4385 incident invasive breast cancer cases. The data on height, weight and body mass index were analysed in the Pooling Project and are included in this review as some of the strongest evidence to date for those risk factors.
The epidemiological evidence for an association between height, weight, body mass index, fat deposition patterns (including waist–hip ratio, waist circumference), weight change (including weight gain, weight loss, weight cycling), and breast size and breast cancer risk are presented in this section.
Height. The most recent and comprehensive assessment of the association between height and breast cancer risk was the pooled analysis of seven prospective cohort studies conducted by the Pooling Project of Diet and Cancer (van den Brandt et al., 2000). In multivariate analyses controlling for reproductive, dietary and other risk factors, the pooled relative risk of breast cancer per height increment of 5 cm was 1.02 (95% confidence interval (CI) 0.96–1.10) in premenopausal women and 1.07 (95% CI 1.03–1.12) in postmenopausal women. The Pooling Project did not include all prospective cohort studies on anthropometric factors and breast cancer risk but did include all studies with data on dietary intake that could potentially confound the associations with anthropometric factors.
Most previous studies have also found a positive association between breast cancer and height (Cold et al., 1998). The evidence is somewhat stronger from the cohort than the case–control studies, possibly because most cohort studies used direct measures of height while the majority of case–control studies used self-reported height. Among those studies that could examine the risks by menopausal status, weaker associations were found for premenopausal women. Overall, the range of risks estimated from these studies for taller, as compared to shorter, women is 0.8–2.0 for premenopausal women and 1.3–1.9 for postmenopausal women (Ballard-Barbash, 1994). Women with a family history of breast cancer experience a greater risk with increased height (RR∼2.0) than do women without such a family history (RR∼1.2) (Ballard-Barbash, 1994).
An increased risk of breast cancer has been observed in populations worldwide. However, stronger associations have been generally found in populations in which inadequate energy intake and nutrients in childhood and adolescence limited growth, such as developing countries and European nations that experienced severe food deprivation during and after World War II (Vatten and Kvinnsland, 1990). Consequently, attained height has been proposed as an indicator of childhood energy intake and it has been suggested that early exposures that possibly affect mammary mass (Albanes and Winick, 1988) may also be critical in breast carcinogenesis (Swanson et al., 1988;Hunter and Willett, 1993). Two studies have found that age when maximum height was attained was related to risk rather than attained height (Li et al., 1997, 2000). These investigators suggested that the physiologic basis for this observation may be a delayed effect of growth hormone and insulin-like growth factor I activity on the breast. They hypothesized that if women reach their maximum height later, their breasts mature later and, consequently, they have less time between their pubertal breast development and the protective breast proliferation that occurs at the time of the first live birth (Li et al., 1997).
Height may also reflect the number of ductal stem cells that develop in the breast in utero, which implicates prenatal exposures in breast cancer aetiology (Trichopoulos and Lipman, 1992). Attained height is also probably influenced by inherited patterns in endogenous hormones and growth factors that influence risk at puberty when breast tissue is rapidly developing in addition to promoting effects later in life. Dietary exposures other than energy deprivation may influence height, including an overabundance of energy and fat and variation in macronutrient intake in the years before puberty (Ziegler et al., 1996). Stoll (1992) has suggested that better nutrition accelerates growth hormone release, which then increases insulin-like growth factor (IGF) levels. The adolescent growth spurt involves stimulation by growth hormone, insulin, IGF and sex steroids, and Stoll (1998) hypothesizes that the combination of IGF and sex steroids results in mitogenic effects on developing mammary tissue in adolescence and, therefore, an increased risk of epithelial atypia and carcinogenesis. (See Biologic mechanisms section below for further discussion of the aetiologic role of IGF in breast cancer.)
Weight. The Pooling Project on Diet and Cancer (van den Brandt et al., 2000) also examined breast cancer risk by menopausal status associated with weight measured at baseline. An inverse association was found for premenopausal women, with a risk of 0.58 (95% CI 0.40–0.83) among women weighing 80 kg or more. For postmenopausal women, the opposite association was found as an increased risk of breast cancer was observed in all categories above the referent group of less than 60 kg. The highest weight category had a risk of 1.25 (95% CI 1.02–1.52). For both pre- and postmenopausal women the linear trends in risk were statistically significant.
The results from the Pooling Project for premenopausal women are in agreement with most earlier investigations (Cold et al., 1998). An inverse association between weight and breast cancer has been found in most earlier studies of premenopausal women conducted in western, high-risk countries (Adami et al., 1977;Paffenbarger et al., 1980;Helmrich et al., 1983;Willett et al., 1985;Kampert et al., 1988;London et al., 1989;Bouchardy et al., 1990;Brinton and Swanson, 1992;Bruning et al., 1992a;Pathak and Whittemore, 1992;Vatten and Kvinnsland, 1992;Törnberg and Cartensen, 1994;Franceschi et al., 1996;Swanson et al., 1996;Trentham-Dietz et al., 1997). A decreased risk among heavy, young Asian–American women has also been found in a recent case–control study (Ziegler et al., 1996).
For postmenopausal women, significant positive associations have been observed in many case–control studies while the results from cohort studies have been more inconsistent (Cold et al., 1998). Although some cohort studies have reported significant positive associations (de Waard et al., 1974;Törnberg et al., 1988;London et al., 1989;Folsom et al., 1990;de Stavola et al., 1993;den Tonkelaar et al., 1994;Goodman et al., 1997;van den Brandt et al., 1997), some have shown weak positive or null associations among postmenopausal women (Swanson et al., 1989;Tretli, 1989;Yong et al., 1996). The inconsistencies found in the cohort studies may be attributable to some methodological limitations inherent in those study designs. Since weight was sometimes measured at baseline and not re-measured during follow-up assessments, and because women gain weight as they age, the ability to detect an association between weight measured several years before diagnosis may be quite diminished. Some studies have found stronger relations between weight and breast cancer among older postmenopausal women (Franceschi et al., 1996;Yong et al., 1996), suggesting that the effect may be stronger, or more easily detected, at older ages. In the Pooling Project, no significant effect modification by age at diagnosis was observed, but there was a suggestion that women over 65 years at diagnosis had increased risks compared with younger women (van den Brandt et al., 2000).
Overall, the case–control studies have found risks ranging from 0.6 to 1.1 for premenopausal women and 1.0–1.9 for postmenopausal when comparing heavier with leaner women as measured as current weight or body mass index (Ballard-Barbash, 1994). For cohort studies, premenopausal women were also found to have risks around 0.6. Postmenopausal women experience risks ranging from 1.0 to 2.2, with the higher risks observed among those with a family history of breast cancer.
Body mass index. Body mass index (BMI) showed significant inverse and positive associations with breast cancer in pre- and postmenopausal women, respectively, in the Pooling Project (van den Brandt et al., 2000). Although the trends were statistically significant for both groups, the associations were non-linear, particularly for postmenopausal women. Compared with premenopausal women with a BMI of less than 21 kg/m 2 , women with a BMI greater than 31 kg/m 2 experienced a risk of 0.54 (95% CI 0.34–0.95). Among postmenopausal women, the risks did not continue to increase after a BMI of 28 kg/m 2 was attained, at which level the risk was 1.26 (95% CI 1.09–1.46). In the Pooling Project, there was no statistically significant heterogeneity among the studies for these associations. Furthermore, the associations were not confounded or modified by other risk factors, with the exception of menopausal status. These results are in agreement with previous investigations that have generally found inverse associations between BMI and risk in premenopausal women and positive associations for postmenopausal women (Cold et al., 1998). The results are relatively consistent in the case–control studies and, as found for weight and breast cancer, less consistent for the cohort studies (Cold et al., 1998). No clear cut-off point for BMI has emerged from these studies – above which there is a decreased risk for premenopausal women and an increased risk among postmenopausal women – because no consistent definitions of cut-off points for BMI have been used.
Ziegler conducted a case–control study (Ziegler et al., 1996) among Asian–American women and found associations with adiposity that were stronger and occurred at younger ages than those observed in other American studies of breast cancer, where the risks ranged from 1.0 to 1.5 in postmenopausal women (Hunter and Willett, 1993;Ballard-Barbash, 1994). Peacock et al. (1999) were the first investigators to find that age was a strong effect modifier of the relation between body mass and breast cancer risk. They found a strong inverse relation between BMI and risk of breast cancer among women aged 21–35 years but did not find this association for women aged 36–45.
Pathak and Whittemore (1992) performed a meta-analysis of case–control data from seven countries worldwide at high, moderate and low incidence of breast cancer and found that incidence rates consistently increased with adiposity in both pre- and postmenopausal women, except for premenopausal women in high-risk countries where an inverse association was found. A subsequent meta-analysis of BMI and risk of premenopausal breast cancer conducted by Ursin et al. (1995) that included 23 case–control and cohort studies reported between 1966 and 1992 also found an inverse association. For a BMI difference of 8 kg/m 2 (i.e. the difference between a thin and obese person), the risk from the four cohort studies was 0.7 (95% CI 0.54–0.91) and from the 19 case–control studies was 0.88 (95% CI 0.76–1.02). The authors found significant heterogeneity across the studies and evaluated potential sources of this heterogeneity (Ursin et al., 1995). They found stronger inverse associations among the case–control studies that had been better designed and properly conducted as compared to the weaker studies. Most studies used body weight measured at the time of the interview, when the cases may have gained weight because of chemotherapy.
The inverse association between adiposity and breast cancer in younger women was initially attributed to earlier detection of breast tumours in leaner women; however, further analyses that have accounted for delay in detection discounted this hypothesis (Brinton and Swanson, 1992;London et al., 1989). Other hypotheses, involving hormonal mechanisms, for this discrepancy are currently being evaluated (see Biological mechanisms section).
Fat deposition patterns. Several investigators have reported that women with increased abdominal fat deposition, or central adiposity, have an increased risk of postmenopausal breast cancer that is independent of adult adiposity (Ballard-Barbash et al., 1990a, bFolsom et al., 1990;Schapira et al., 1990;Bruning et al., 1992a;den Tonkelaar et al., 1994;Männistö et al., 1996;Kaaks et al., 1998;Huang et al., 1999). Greater upper or central body fat distribution is associated with multiple hormonal and metabolic changes including insulin resistance, hyperinsulinaemia, decreases in sex hormone-binding globulin levels, increases in androgen levels, and the conversion of androgen to oestrogen in adipose tissue (Ballard-Barbash, 1994). Thus, women with central adiposity may have a higher risk of breast cancer than women whose fat is mainly distributed subcutaneously over the hips, buttocks and lower extremities. The pattern of abdominal fat deposition is believed to be genetically determined and is linked to increased risk of diabetes mellitus, hypertension and cardiovascular disease (Folsom et al., 2000).
In the Nurses’ Health Study cohort I of women aged 30–55 years at baseline, waist circumference and waist–hip ratio were moderately associated with an increased risk of breast cancer, particularly among postmenopausal women (Huang et al., 1999). The risk for the highest versus lowest quintile of waist circumference for postmenopausal women was 1.34 (95% CI 1.05–1.72). When the analysis was restricted to women who had never used hormone replacement therapy, the risk was increased to 1.88 (95% CI 1.25–2.85). After controlling for body mass index, the positive association was slightly attenuated to 1.83 (95% CI 1.12–2.99). The risk for waist–hip ratio among postmenopausal women, when fully adjusted including body mass index, was 1.22 (95% CI 0.96–1.55) and when restricted to never users of hormone replacement therapy, was 1.85 (95% CI 1.25–2.74).
Three other prospective studies have examined regional adiposity and breast cancer risk (Ballard-Barbash et al., 1990b;Folsom et al., 1990;Sellers et al., 1992;den Tonkelaar et al., 1994;Kaaks et al., 1998). Each study used different measures of adiposity, including a central adiposity ratio derived from skin fold measurements (Ballard-Barbash et al., 1990b), skin fold measurements (den Tonkelaar et al., 1994) and waist–hip ratio (Folsom et al., 1990;Kaaks et al., 1998). A positive association with postmenopausal breast cancer risk was found within each study for these measures of central adiposity, with the exception of one analysis from the Dutch DOM cohort (den Tonkelaar et al., 1994). This lack of association may be attributable to the measure used, which was restricted to subscapular and triceps skin fold thicknesses and did not include gluteofemoral fatness as used in other studies. A weaker association was found for premenopausal women in these studies.
Waist–hip ratio, and particularly waist circumference, has also been associated with postmenopausal breast cancer in most (Ballard-Barbash et al., 1990b;Schapira et al., 1990;Bruning et al., 1992a;Männistö et al., 1996) but not all (den Tonkelaar et al., 1992;Petrek et al., 1993) of the case–control studies that have examined central adiposity and its association with breast cancer. An increased risk for premenopausal breast cancer has been observed for waist–hip ratio (Männistö et al., 1996), but most case–control studies have shown only weak or null associations between these factors and premenopausal breast cancer (Bruning et al., 1992a;Petrek et al., 1993;den Tonkelaar et al., 1995;Swanson et al., 1996).
Waist circumference has been found to be a stronger predictor of breast cancer risk than waist–hip ratio (Huang et al., 1999), probably because it is a direct measure of abdominal adiposity. Waist–hip ratio, the other most commonly used measure of central adiposity, is a ratio of two variables, both of which contribute to cancer risk. Furthermore, waist–hip ratio includes more measurement error because it includes errors in measure from both the waist and hip circumferences (Rimm et al., 1990). Waist–hip ratio and waist circumference, in turn, have been stronger predictors of breast cancer risk than body mass index in some studies (Kaaks et al., 1998;Huang et al., 1999), perhaps because they are more sensitive indicators of obesity. It is also possible that waist–hip ratio is more specifically related to alterations in hormone metabolism that are related to breast cancer risk.
Overall, having higher central adiposity increased breast cancer risk among postmenopausal women from 1.4 to 5.2 times that of women with lower central body fat (Ballard-Barbash, 1994).
Weight change. Weight gain through adult life has consistently been shown in retrospective and prospective studies to increase postmenopausal breast cancer risk and this effect is evident even in cohort studies that found no association between baseline relative weight and subsequent risk (Paffenbarger et al., 1980;Le Marchand et al., 1988;London et al., 1989;Ingram et al., 1989;Ballard-Barbash et al., 1990a;Folsom et al., 1990;Chu et al., 1991;Brinton and Swanson, 1992;Harris et al., 1992;Radimer et al., 1993;Barnes-Josiah et al., 1995;Kumar et al., 1995;Franceschi et al., 1996;Männistö et al., 1996;Ziegler et al., 1996;Huang et al., 1997Trentham-Dietz et al., 1997, 2000van den Brandt et al., 1997;Magnusson et al., 1998). These studies have shown increased risks between 1.2 and 2.3 for the highest categories of weight gain versus the lowest categories for weight gained between age 18 or 20 and the reference age. These results are consistent despite the heterogeneity among studies in the definitions of weight gain, use of weight gain versus increases in body mass index, and relevant ages (Trentham-Dietz et al., 2000). The consistent associations observed for adult weight gain may be attributable to the fact that weight gain reflects mainly an increase in body fat and therefore is a more accurate measure of adiposity than body weight, which includes both lean and fat mass (Ballard-Barbash et al., 1990a;Kumar et al., 1995).
Weight gain during particularly susceptible time periods in breast carcinogenesis (e.g. pregnancy, menopause) may be critical, or weight gain during periods of hormonal transition may indicate high risk metabolic patterns (Ballard-Barbash, 1994). One recent study by Coates et al. (1999) found that young women who were heavier and who were lighter than average in early adulthood (18–25 years) were at reduced risk of breast cancer. Weight gain after age 20 resulted in reduced risk; however, the effect was restricted to early stage and, more specifically, lower grade breast cancer. Thus, the association between body weight at different ages and breast cancer risk in young women is complex and effect modification by tumour stage and type may exist (Coates et al., 1999). Another recent investigation has found that earlier menarche, extremely lean body mass at age 10 years, and taller adult height predicted increased breast cancer risk (Berkey et al., 1999). These same three factors were also found to predict higher peak growth velocities during adolescence, which supports the hypothesis that more rapid adolescent growth may increase the risk of breast cancer development (Berkey et al., 1999).
Huang et al. (1997), using data from the Nurses’ Health Study I cohort, demonstrated the dual effects of hormone replacement therapy and weight gain on breast cancer risk. In that cohort, adult weight gain in postmenopausal women increased the risk of both breast cancer incidence and mortality. Much stronger risks were found in postmenopausal women who never used hormone replacement therapy as compared to current or past hormone users. A statistically significant interaction between weight change and postmenopausal hormone use was found. This study also noted that magnitude of excess risk was greater for adult weight gain than for higher recent body weight, a finding supported by other studies as well (Kumar et al., 1995;Le Marchand et al., 1988). Huang et al. (1997) estimated that as much as one-third of new cases of postmenopausal breast cancer may be attributable to adult weight gain, replacement hormone therapy use, or both. This study demonstrated the importance of considering subgroups of the population separately when a potential risk factor such as obesity is believed to operate through the same aetiologic pathway (i.e. increased oestrogen exposure) as one or more of the subgroups at elevated risk (i.e. users of hormone replacement therapy) (Kelsey and Baron, 1997). In the Nurses’ Health Study, about half of the women had used hormone replacement therapy; hence the lack of effect among non-users could substantially dilute any association among the non-users. The results from the Nurses’ Health Study are supported in other studies that also found an attenuation of the excess risk associated with weight gain among menopausal hormone users (Harris et al., 1992;Franceschi et al., 1996;Magnusson et al., 1998;Trentham-Dietz et al., 2000). These results may be attributable to the higher oestrogen levels obtained from hormone therapy than those derived from adipose tissue (Jurgens et al., 1992;Hankinson et al., 1998a).
The timing of weight gain may also be important in evaluating breast cancer risk (Stoll, 1995). Excessive weight gain during periods of hormonal change such as puberty (Stoll, 1998), pregnancy, lactation and menopause (Stoll, 1999) could have different physiologic effects because of the differences in the location of fat deposition during those periods. Weight gained during puberty is primarily in the hips and buttocks (gynoid), while weight gain during pregnancy and menopause is characterized by an increase in central body fat distribution (android) (Kumar et al., 1995). Several studies have shown that women with a predominant upper body-fat distribution are at increased risk for breast cancer (Ballard-Barbash et al., 1990a, bFolsom et al., 1990;Schapira et al., 1990;Sellers et al., 1992;Kumar et al., 1995;Männistö et al., 1996;Ng et al., 1997;Magnusson et al., 1998;Sonnenschein et al., 1999;Hall et al., 2000). For one of these recent studies (Sonnenschein et al., 1999), effect modification between waist–hip ratio and body mass index was found. In that study, premenopausal women with a high waist–hip ratio who were overweight had an increased risk while those with a low waist–hip ratio who were overweight experienced no increased risk (Sonnenschein et al., 1999). Adult weight gain that occurs primarily during childbearing and menopause, and in which the fat deposition is mainly in the upper body, may theoretically pose the highest risk for breast cancer (Kumar et al., 1995). Excessive weight gain during these times of hormonal change may result in metabolic dysfunction leading to hyperinsulinaemia (Stoll, 1995).
Weight loss, particularly later in life, may decrease breast cancer risk (Kyogoku et al., 1990;Männistö et al., 1996;Ziegler et al., 1996Trentham-Dietz et al., 1997, 2000van den Brandt et al., 1997;Magnusson et al., 1998). This decrease in risk may be mediated through changes in hormonal status and body fat distribution (Kumar et al., 1995). Studies have shown an increase in sex hormone-binding globulin with weight loss (Enriori et al., 1986;Schapira et al., 1994). Women with a predominant central body fat distribution have been shown to decrease fat cell size with caloric reduction, thereby reducing abdominal obesity. This reduction could change the pattern of obesity and alter breast cancer risk (Schapira et al., 1994). Schapira and colleagues, using a risk model for breast cancer and body fat localization, estimated that breast cancer risk could be reduced by 45% for an entire group of women who lost more than 4.5 kg of body weight (Schapira et al., 1991).
Weight cycling, defined as successive weight loss and gain, does not appear to be related to breast cancer risk; however, few studies have examined this risk factor (Trentham-Dietz, 2000). It appears that sustained weight gain, rather than fluctuations in weight, increases postmenopausal breast cancer risk. There have been suggestions that weight cycling can be detrimental to health through effects on metabolic rate, body composition and fat distribution, or dietary preference for foods high in dietary fat; however, no strong or consistent evidence has supported these hypotheses (Prentice et al., 1992;Wing, 1992;Muls et al., 1995;Jeffrey, 1996).
Overall, weight gain throughout adult life was associated with a risk ranging from 0.5 to 1.2 in premenopausal women and 1.4–2.5 in postmenopausal women (Ballard-Barbash, 1994). Weight loss, on the other hand, was associated with a risk of 0.7–0.9 in premenopausal women and 0.8–1.5 in postmenopausal women (Ballard-Barbash, 1994).
Breast size. Breast size has also been investigated as a possible predictor of breast cancer risk since it is considered a plausible proxy for the volume of ductal epithelium at risk. Fat depots in the breast also contribute to local oestrogen levels and may act as a repository for lipid-soluble carcinogens (Kohlmeier and Kohlmeier, 1995). Only a few studies (Dupont and Page, 1987;Hsieh and Trichopoulos, 1991;Kato et al., 1995) have supported an association between breast size and cancer risk. Egan and colleagues (1999) suggested that the lack of association found in most studies (Wynder et al., 1960;Valaoras et al., 1969;Katariya et al. 1974;Hirohata et al., 1977;Soini, 1977;Adami and Rimsten, 1978;Kolonel et al., 1986;Senie et al., 1993;Tavani et al., 1996;Thurfjell et al., 1996;Scutt et al., 1997) could be attributable to the small sample sizes, the lack of control for overall adiposity, and misclassification of the exposure since breast size after, rather than before childbearing was used. Egan et al. (1999) examined the risk of preparous bra size dimensions in the Collaborative Breast Cancer Study, a large population-based case–control study that included women aged 50–79 years at diagnosis. They found an increased risk of breast cancer only among women who were large-breasted and lean as young adults and who had a small back circumference (i.e. chest size under 34 inches). Egan et al. (1999) suggested that the lack of association in previous studies may be attributable to a lack of consideration of effect modification by overall adiposity. In their investigation, no association for breast size was apparent in the crude analysis that combined all categories of back circumferences. Back circumference, they argue, may be a more important variable for assessing cup size associations than body mass index since the latter measure does not take into consideration body fat distribution. These investigators felt that breast cancer aetiology could be determined early in life when processes controlling growth and development determine body size and also the number of breast cells at risk for malignancy.
Summary of epidemiological evidence for anthropometric factors and breast cancer. In summary, increasing height is associated with an increasing risk of breast cancer in both pre- and postmenopausal women. Attained height may be influenced by childhood and adult nutrition, genetic predisposition, prenatal exposures and IGF levels. Increased weight, as assessed as either weight or body mass index, decreases breast cancer risk before menopause but increases risk after menopause. Central adiposity, as assessed by waist circumference or waist–hip ratio, is positively associated with postmenopausal breast cancer but has no impact on premenopausal breast cancer. Weight gain through adult life increases postmenopausal breast cancer risk and this effect may be modified by hormone replacement therapy, since non-users had much greater risks than current or past hormone users. Adult weight gain that occurs primarily in the upper body appears to pose the greatest risk for breast cancer. Weight loss, particularly occurring later in life, decreases risk and weight cycling is not associated with risk while breast size may be a risk factor for breast cancer, the evidence is currently inconclusive. Besides breast size, the evidence for the associations between the anthropometric factors considered here and breast cancer risk is fairly strong and consistent across studies. Biologic plausibility for these associations also exists; the main mechanisms are discussed in the next section.
Several hypothesized biologic mechanisms have been postulated and investigated to explain how anthropometric factors might influence breast cancer risk. These biologic mechanisms include: (1) endogenous sex hormones, (2) insulin, (3) growth factors, (4) genetic factors, and (5) fat tissue storage.
Endogenous sex hormones. Oestrogens, and possibly androgens, are important in the aetiology of breast cancer. Early research showed that oestrogens induce and promote mammary tumours in rodents (MacKenzie, 1955) and that the risk of developing breast cancer is increased among women with early menarche or late menopause (Kelsey et al., 1993). Incidence rates of breast cancer rise less steeply in all populations after menopause, when ovarian oestrogen production ceases (Key and Pike, 1988). Breast cancer risk also decreases after bilateral oophorectomy (Lilienfeld, 1956) or through the use of anti-oestrogenic drugs (Bush and Helzlsouer, 1993). Prospective cohort studies reported in the 1990s have found that increased levels of oestrone, oestradiol and bioavailable oestradiol may be associated with strong increased risks of breast cancer (Toniolo, 1997). Serum total oestradiol levels have been shown to decrease with increasing BMI in premenopausal women but to increase in postmenopausal women (Potischman et al., 1996). The reversal in this endogenous oestrogen level could explain the differing relation between obesity and breast cancer before and after menopause (Carroll, 1998). More recent evidence has shown that women with the thickest bone density, which may be a surrogate for cumulated exposure to endogenous sex hormones, experience greatly increased risk of breast cancer as compared to women with thin bones (Cauley et al., 1996;O’Brien and Caballero, 1997;Zhang et al., 1997;Nguyen et al., 2000). Circulating hormone levels may increase as a result of an overall increase in ovarian and adrenal secretion occurring or persisting after menopause.
Studies on both pre- and postmenopausal women have shown that obesity is associated with increased plasma concentrations of testosterone (Evans et al., 1983;Krotkiewski et al., 1990;Kaye et al., 1991;Maggino et al., 1993) and decreased concentrations of sex hormone-binding globulin (SHBG) (Soler et al., 1989;Kaye et al., 1991;Campagnoli et al., 1992;Haffner et al., 1992;Maggino et al., 1993). SHBG is the main protein carrier of oestradiol and the percentage of free oestradiol is related inversely to the level of SHBG. As testosterone levels increase and SHBG levels decrease in obese women, the levels of free oestrogen increase (Kaye et al., 1991). The effect of obesity on the bioavailable level of oestrogen is further influenced by menopausal status. Before menopause, more oestrogens are produced in the ovaries than in peripheral adipose tissue. Obese premenopausal women have reduced progesterone levels, because of an increased frequency of an ovulation and a decreased production of progesterone in the luteal phase (Henderson et al., 1985). Moreover, leptin levels, which increase with increasing fat stores, inhibit ovarian oestrogen production and may, thereby, contribute to reduced risk of breast cancer in heavier, younger women (Spicer and Francisco, 1997). After menopause, there is an increased level of production of oestrogens from androgens (Meldrum et al., 1981;Moore et al., 1982;Reed et al., 1983;Begg et al., 1987;Ingram et al., 1990;Shifren and Schiff, 2000) and decreased oestrogen–protein binding because of decreases in SHBG levels (Kaye et al., 1991;Bruning et al., 1992a;Maggino et al., 1993). Androgens may also increase breast cancer risk directly through increasing breast cell proliferation after binding to androgen receptors (Bryan et al., 1984).
Insulin, glucose, triglycerides. Obesity, especially central adiposity, may also increase breast cancer risk through increasing circulating concentrations of insulin, glucose or triglycerides. Insulin levels in premenopausal women may be positively associated with breast cancer (Del Guidice et al., 1998) and cross-sectionally with adiposity and subsequent weight gain (Folsom et al., 1998). In postmenopausal women, increased adiposity is related to increased plasma glucose and hyperinsulinaemia (Okosun et al., 2000). Kaaks (1996) has argued that nutritionally induced hyperinsulinaemia and insulin resistance are the fundamental metabolic changes that result in breast cancer development. This hypothesis offers a model that combines overnutrition, obesity, low physical activity, chronic changes in the endocrine secretion of steroid hormones (particularly ovarian androgens) and reduced production of SHBG by the liver. Toniolo (1997) has suggested that in high-risk populations, nutritionally induced endocrine disregulation would begin early in life and that the key to understanding breast cancer aetiology would be to understand metabolic and hormonal alterations that occur during childhood and adolescence and that continue throughout a woman's lifetime.
Insulin-like growth factor. Insulin has been shown to be a growth factor for breast cancer cells (Pollak, 2000) and C-peptide, a marker of hyperinsulinaemia and insulin resistance, predicts breast cancer risk (Bruning et al., 1992b). Women with breast cancer have been reported to have higher serum levels of C-peptide than comparable controls (Bruning et al., 1992a). C-peptide levels were also correlated with waist–hip ratio (Bruning et al., 1992a). Waist–hip ratio may be a more specific marker of the metabolic consequences of obesity, including hyperinsulinaemia, hypertriglyceridaemia and hyperandrogenicity (Björntorp, 1988). Although hyperinsulinaemia may promote breast cancer, non-insulin-dependent diabetes mellitus, a late, non-sensitive marker for hyperinsulinaemia, has not been found to be a consistent risk factor for breast cancer (Weiss et al., 1999). Serum insulin concentrations are negatively associated with serum SHBG levels (Kaye et al., 1991), which could either reflect or explain the positive association between obesity and serum oestrogens.
Receptors for insulin-like growth factors, which are synthesized by fat tissues, are expressed by breast tumour stromal cells (Yee et al., 1989) and may play an aetiologic role in the promotion of breast cancer by oestrogens (Stewart et al., 1990). Recent research has shown that insulin-like growth factor I (IGF-I) increases premenopausal breast cancer risk (Hankinson et al., 1998b). Obesity in postmenopausal women also is related to increased insulin-like growth factor I levels (Campagnoli et al., 1992). There is further indirect evidence for an association between IGF-I and breast cancer risk involving anthropometric factors. Height, a risk factor for breast cancer, is positively related with IGF-I concentrations (Juul et al., 1994). Low birthweight is associated with lower IGF-I concentrations (Osorio et al., 1996) and higher birthweight increases breast cancer risk (Michels et al., 1996). Reducing energy intake and thereby reducing body weight decreases IGF-I levels and decreases breast cancer incidence in rodents (Ruggeri et al., 1998).
Genetic factors. Genetic factors are also involved in the association between anthropometric factors and breast cancer risk. It is known that genetic factors are implicated in the predisposition to being overweight (Ravussin et al., 1992;Bouchard et al., 1993) and in determining body fat distribution and insulin resistance (Wing et al., 1992;Bouchard et al., 1993). However, developing central obesity and experiencing insulin resistance still require overnutrition to occur and genetic predisposition alone is not sufficient to cause obesity.
Fat tissue storage. Fat tissue has the capacity to store toxins, medications and certain vitamins, and stored toxins can serve as a continuous source of carcinogens (Kohlmeier and Kohlmeier, 1995). Hence, women with more fat tissue could be exposed to a higher level of carcinogens than women who have less body fat (McTiernan, 2000). Alternatively, fat-soluble antioxidant vitamins A, D and E stored in fat could also decrease cancer risk. Storage of carcinogens could be particularly relevant for some racial and ethnic groups, such as Hispanic migrant agricultural workers, that might be exposed occupationally to high levels of carcinogens (McTiernan, 2000).
Recommendations for future research
The field of anthropometric measurements and their association with breast cancer risk has been actively researched and a clearer understanding of the aetiologic roles of weight, height, adiposity, fat distribution patterns, and weight change throughout lifetime is emerging. A few gaps in the epidemiological evidence exist and much more research is needed to clarify the biologic mechanisms that are operative. Finally, intervention studies that test the impact of weight control programs on breast cancer risk are the ultimate level of research that will provide a more definitive understanding of the underlying aetiology and biology for these associations. This next section highlights these gaps in current understanding and provides recommendations for future research.
Gaps in current knowledge
Although it is clear that obesity and weight gain in life are risk factors for postmenopausal breast cancer, the precise aetiologic role for these risk factors remains unclear. Specifically, it is currently unknown if weight gain itself is sufficient to cause physiological changes that influence breast cancer risk or if other factors, such as lack of physical activity, are also necessary for risk increases. The independent effects of weight at different time points in life versus weight gain over lifetime have not been adequately studied. Although some evidence exists to suggest that weight gain in specific times in life is particularly aetiologically relevant, the question of timing of weight gain requires more clarification. Likewise, some initial research has suggested that weight loss can reduce breast cancer risk; however, it remains uncertain the extent to which weight loss can reverse the effects of weight gain. The interaction of anthropometric factors with other breast cancer risk factors has received some preliminary assessment in recent investigations but requires further clarification to identify particular subgroups of the population that may be at particularly high risk (e.g. women with family history of breast cancer). Furthermore, preliminary evidence suggests that the stage and type of breast tumour may also be important effect modifiers for the association between body size and shape and breast cancer risk. Since these investigations have used differing definitions for the highest risk category, more research is needed to define the ranges of body size and shape over which the effects on breast cancer risk will be observed.
Data are still relatively sparse on the influence of anthropometric factors on breast cancer among women of different ethnic and racial backgrounds, since the majority of research studies have been conducted on Caucasian women (McTiernan, 2000). Several ongoing cohort studies in the US have included large numbers of racial and ethnic minority women notably the Women's Health Initiative in the US that has enrolled more than 28 000 African–American, Hispanic, American Indian and Asian–American women into long-term clinical trials and a cohort study (Women’s Health Initiative Study Group, 1998).
The hormonal and metabolic mechanisms that underlie the associations of anthropometric factors and breast cancer need further clarification since several of the postulated mechanisms have not been appropriately tested in epidemiological studies.
Methodological improvements for future epidemiological studies
This review has highlighted the epidemiological evidence for the main variables used thus far to measure anthropometric characteristics and has discussed how some of these variables are related. More detailed, standardized, reliable and validated assessments of body size and shape are needed. Recent advances in some technologies such as bioelectric impedance for the assessment of body composition may result in affordable and portable methods of assessing body composition that could be used in epidemiological investigations (Ballard-Barbash, 1999). For fat distribution, better classification according to fat localization is needed so that differences in fat deposition can be appropriately measured (Hall et al., 2000). As described, self-reported measures were commonly used in earlier epidemiological studies and only a limited number of measures have been used. Several other measures of anthropometric characteristics already exist and others may still be developed. Existing ones that could be used are measures of sitting height, distribution of adipose tissue, and the relative percentage of the absolute amount of adipose tissue in the body. These variables have seldom been measured but could have biologic importance. In fact, Cold and colleagues (1998) suggested that these anthropometric variables should be linked to biologic measures such as hormone levels, composition of fatty tissue, hormone receptor status and information on menopausal status, diet and physical activity. Such studies would clarify how each risk factor is implicated in breast cancer aetiology.
Observational epidemiological studies. Future epidemiological studies of anthropometric factors and breast cancer should be designed to address the gaps identified above. Such studies would use more standardized, validated definitions of these anthropometric characteristics and would expand to incorporate new measures that have not been studied or have been incompletely studied to date. They would attempt to measure weight throughout lifetime so that the influence of weight and weight gain at critical life periods and breast cancer risk could be fully evaluated. Where possible, data from medical and school records should supplement recalled information on body size. The effects of weight gain during pregnancy and lactation and weight cycling caused by dieting need to be examined. These studies could also investigate the resulting changes in hormonal status and body fat distribution to define further the role of timing of weight gain during the critical periods in a woman's life (Kumar et al., 1995).
Prospective cohort studies are needed that have repeat assessments of anthropometric measures and associated metabolic factors over several time periods as well as follow-up assessments of menopausal status. Such studies would permit an analysis of the temporal relations between breast cancer risk and anthropometric measures and would reduce misclassification bias since menopausal status at the time of diagnosis would be available.
Future studies should examine confounding and effect modification by other breast cancer risk factors that have not yet been consistently examined. To date, menopausal status, age, and a family history of breast cancer has been identified as important effect modifiers of the association between weight and body fat distribution and breast cancer risk (Sellers et al., 1992;Peacock et al., 1999). Future research could identify more precisely the population subgroups that experience particularly increased or decreased risks. Analyses in future studies should also consider the possibility of complex, possibly non-linear relations since the association between body size and shape differs with population subgroups (Coates et al., 1999). Li and colleagues (2000) demonstrated how the associations between anthropometric factors and breast cancer risk can be missed if women at the different extremes of these variables are combined into categories (e.g. quartiles) with women whose weight is much closer to the normal range than theirs. If the women in the extreme categories are analysed separately, clearer patterns can be detected. These investigators found women in the very lowest category, who would be considered anorexic, had a decreased risk of breast cancer while women in the most obese category had an elevated risk (Li et al., 2000). Hence, they suggest that division of the data into smaller and more clinically relevant strata is needed to detect these relations. Careful consideration of other related factors and biologic interactions is also important to reveal these complex underlying associations. Future observational epidemiological studies should incorporate biologic characteristics of the breast cancers into the investigations (Coates et al., 1999).
McTiernan (2000) suggested that case–control studies designed specifically to determine the effects of anthropometric factors in racial and ethnic minorities would be helpful since these groups have been insufficiently studied to date. She also suggested that the data from currently on-going studies that include these minorities should be pooled since the sample size of a particular group could be too small to permit inferences from one study alone. One recent case–control study of Black and White women (Hall et al., 2000) found similar associations between anthropometric factors and breast cancer risk, although the body size profiles of the two groups were dissimilar.
Intervention studies. This review has provided considerable evidence to support the suggestion by Ballard-Barbash and Swanson (1996) that the prevention of postmenopausal weight gain and accumulation of central body fat during adult life may reduce the risk of breast cancer. Since these factors are truly modifiable they are particularly well suited for breast cancer prevention studies. Ziegler et al. (1996) concluded that the twofold to threefold changes in breast cancer risk observed for weight, weight change and height in their case–control study were at least as strong as those associated with established breast cancer risk factors and that they are definitely stronger than those found for specific dietary factors or endogenous hormones. They also argued that weight loss has the potential to have an impact, within a decade, on breast cancer risk. Given that further clarification of some aspects of the aetiologic association are all that remains to be elucidated at this time, there is currently a strong rationale to proceed with intervention studies among postmenopausal overweight women to determine how breast cancer risk can be reduced.
Intervention studies are needed to examine whether or not weight loss and/or maintenance can reduce breast cancer incidence and if such weight loss and/or maintenance must be accompanied by specific dietary patterns or levels of physical activity (Ziegler et al., 1996). McTiernan and colleagues (1999a) have provided detailed arguments and a clear rationale for intervention trials for cancer prevention and are providing leadership in this field with their currently on-going randomized controlled trial entitled the Physical Activity for Total Health (PATH) study (McTiernan et al., 1999b). In the PATH study, McTiernan and colleagues are studying the influence of exercise on changing biomarkers of breast carcinoma risk in sedentary obese postmenopausal women. McTiernan et al. (1999a), in their detailed description of the various steps needed to establish sufficient evidence before embarking on large-scale clinical trials of breast cancer prevention, have noted that small-scale clinical trials should be used to determine the effect of weight loss on endogenous sex and metabolic hormones. These experimental studies can provide more clarity on the exact role of these anthropometric characteristics on breast cancer risk. They can also provide direct data on how biologic factors are influenced when weight, body fat content and distribution are changed under a controlled setting.
These trials can also be conducted in different racial and ethnic groups to determine how weight loss influences risk across different population subgroups (McTiernan, 2000). Trials in different populations could be used to determine the cultural acceptance of weight loss in these populations, data that would be very relevant when designing health promotion strategies for breast cancer prevention that are targeted at these groups. After conducting these smaller scale studies and studies within different populations, large-scale intervention trials could be conducted that would combine studies on the influence of changing these anthropometric characteristics on breast cancer risk with investigations of the underlying biologic mechanisms that are operative.
Some preliminary studies of weight reduction and markers for breast cancer have been conducted. Schapira et al. (1991) examined weight reduction in healthy women and found favourable changes in anthropometric factors. O’Dea et al. (1979) found that, as compared with healthy non-obese controls, levels of sex hormone-binding globulin increased to a higher level among obese postmenopausal women who underwent a weight-reduction programme. Much more research potential exists in this field since relatively little research has been conducted on the effectiveness and efficacy of weight reduction on breast cancer risk.
Studies of anthropometric factors and other breast cancer lifestyle risk factors
To answer questions regarding the specific roles of weight versus physical activity and dietary intake in breast cancer aetiology, specific research studies are needed that combine these factors. In a review of the research issues in the field of weight gain and fat distribution changes with menopause, Astrup (1999) argued that there is a need to evaluate the effect of increased physical activity and fitness as a tool for the prevention of changes in body composition associated with menopause and ageing in normal women. He also identified a need to clarify the independent effect on breast cancer risk of abdominal fat gain as distinguished from the risk attributable to total fat gain associated with menopause. Furthermore, he suggested that the efficiency of different programmes of exercise training as treatment options in subjects with existing obesity should be investigated alone and in combination with diet interventions. These questions could be most adequately examined in a well-designed randomized controlled trial that should be sufficiently large to allow subgroup analyses of factors such as hormone replacement therapy, ethnic origin, race and socio-economic factors as possible effect modifiers. These trials should also use valid and reliable methods to measure exercise compliance, physical fitness, body composition and intra-abdominal adipose tissue. Hence, intervention trials that examine the effects of physical activity and dietary changes on weight and body fat loss would be an ultimate objective in a research programme. Such trials would clarify the combined and individual impacts of these factors on breast cancer risk.
Research on biologic mechanisms
More research is needed to clarify how the hypothesized biologic mechanisms are operative. Toniolo (1997) has suggested that new prospective cohort studies should be conducted to examine the role of endogenous sex hormones, as measured in blood and urine samples obtained early in the natural history of breast cancer, along with an assessment of bone density and other important risk factors including body weight, mammographic density, physical activity. Markers of individual susceptibility should also be assessed, since they may indicate increased risk through an effect on the metabolism of endogenous hormones or through specific metabolic and physiologic responses to over nutrition. These studies would focus primarily on hormonal and metabolic imbalances associated with breast cancer in adult life; however, their results could create new opportunities to study associations between metabolic biomarkers and lifestyle determinants earlier in life. This research would provide the necessary evidence to design effective strategies for breast cancer prevention.
Weight control interventions for breast cancer prevention
Population attributable risk estimates indicate that 10–16% of postmenopausal breast cancer may be attributable to obesity and weight gain (Brownson et al., 1993;Mezzetti et al., 1998). However, women generally do not believe that obesity is a potential cause of breast cancer (Brinton et al., 1994). Postmenopausal obesity is likely the only established risk factor for breast cancer that could be influenced by behaviour modification (Kelsey and Bernstein, 1996). Obesity is already clearly associated with numerous other chronic diseases including diabetes mellitus, hypertension, dyslipidaemia, coronary artery disease, cerebrovascular disease, osteoarthritis and other cancers (Solomon and Manson, 1997;Must et al., 1999). The prevalence of obesity is increasing worldwide and has reached epidemic levels in most western countries (VanItallie, 1996;Mokhad et al., 1999). The direct costs of inactivity and obesity were estimated recently by Colditz (1999) to account for 9.4% of the national health care expenditures in the United States. Although much research has examined the treatment of obesity and weight loss maintenance, far less research has examined primary prevention of obesity (Glenny et al., 1997). Community-based education strategies, especially combined with financial incentives, may help prevent obesity in adults (Glenny et al., 1997). Interventions, strategies and programmes for weight maintenance as well as weight reduction are, therefore, clearly a high public health priority. Given the prevalence of obesity and the potential reduction in breast cancer incidence and mortality that could occur with decreases in postmenopausal obesity among women, more intervention research on the treatment and prevention of obesity is another essential component of future research in primary prevention of breast cancer. The two main environmental strategies to prevent obesity are, simply, to increase physical activity and to decrease food intake. Both strategies are difficult to implement on a population-wide basis (Hill and Melanson, 1999). Since physical activity and total caloric intake are also related to breast cancer risk, a combined programme that integrates these risk factors may have the greatest potential for breast cancer prevention.
Anthropometric factors are clearly implicated in breast cancer aetiology. Several aspects of these associations and the underlying biologic mechanisms, however, require further clarification. Priorities for research include: better assessment of anthropometric characteristics; more aetiologic studies of these factors and their influence on risk throughout life, with particular emphasis on population subgroups and ethnic minorities that have not been sufficiently investigated; intervention trials of weight loss and maintenance; intervention trials of diet, physical activity and weight control to determine the relative contribution of each to breast cancer risk reduction; and more research on the biologic mechanisms that may be operative. Despite the gaps in current understanding of these aetiologic associations, there is sufficient evidence to support avoiding weight gain throughout adult life as a means of reducing postmenopausal breast cancer risk. Furthermore, given the established health consequences of obesity, interventions and public health recommendations to maintain ideal weight throughout life are clearly warranted and should be a very high priority for health policy.
de Waard and Baanders-van Halewijn, 1974;Ravussin and Swinburn, 1992
C M Friedenreich is supported by a National Health Research Scholar Award from Health Canada. This work was originally commissioned by the Canadian Breast Cancer Initiative of Health Canada.
1. Adami HO , Rimsten Å ( 1978 ). Adipose tissue and aetiology
of breast cancer
. Lancet 2 : 677 – 678 .
2. Adami HO , Rimsten Å , Stenkvist B , Vegelius J ( 1977 ). Influence of height
and obesity on risk of breast cancer
in an unselected Swedish population . Br J Cancer 36 : 787 – 792 .
3. Albanes D , Winick M ( 1988 ). Are cell number and cell proliferation risk factors for cancer? . J Natl Cancer Inst 80 : 772 – 775 .
4. Astrup A ( 1999 ). Physical activity and weight
gain and fat distribution changes with menopause: current evidence and research issues . Med Sci Sports Exerc 31 : S564 – S567 .
5. Ballard-Barbash R ( 1994 ). Anthropometry and breast cancer
. Body size – a moving target . Cancer 74 : 1090 – 1100 .
6. Ballard-Barbash R ( 1999 ). Energy balance, anthropometry, and cancer . In Nutritional Oncology (eds Heber D , Blackburn GL , Go VLW ); Academic Press , New York , pp. 137 – 151 .
7. Ballard-Barbash R , Swanson CA ( 1996 ). Body weight
: estimation of risk for breast and endometrial cancers . Am J Clin Nutr 63 : 437S – 441S .
8. Ballard-Barbash R , Schatzkin A , Taylor PR , Kahle LL ( 1990 ). Association of change in body mass with breast cancer
. Cancer Res 50 : 2152 – 2155 .
9. Ballard-Barbash R , Schatzkin A , Carter CL et al. ( 1990 ). Body fat distribution and breast cancer
in the Framingham Study . J Natl Cancer Inst 82 : 286 – 290 .
10. Barnes-Josiah D , Potter JD , Sellers TA , Himes JH ( 1995 ). Early body size and subsequent weight
gain as predictors of breast cancer
incidence (Iowa, United States) . Cancer Causes Control 6 : 112 – 118 .
11. Begg L , Kuller LH , Gutai JP , Caggiula AG , Wolmark N , Watson CG ( 1987 ). Endogenous sex hormone levels and breast cancer
risk . Genet Epidemiol 4 : 233 – 247 .
12. Berkey CS , Frazier LA , Gardner JD , Colditz GA ( 1999 ). Adolescence and breast carcinoma risk . Cancer 85 : 2400 – 2409 .
13. Björntorp P ( 1988 ). The associations between obesity, adipose tissue distribution and disease . Acta Med Scand 723 : 121 – 134 .
14. Bouchardy C , Le MG , Hill C ( 1990 ). Risk factors for breast cancer
according to age at diagnosis in a French case–control study . J Clin Epidemiol 43 : 267 – 275 .
15. Bouchard C , Desprès JP , Mauriège P ( 1993 ). Genetic and non-genetic determinants of regional fat distribution in humans . Endocr Rev 14 : 72 – 93 .
16. Brinton LA , Swanson CA ( 1992 ). Height
at various ages and risk of breast cancer
. Ann Epidemiol 2 : 597 – 609 .
17. Brinton LA , Malone KE , Stanford JL , Liff J , Schoenberg JB ( 1994 ). Re: Should we consider a subject's knowledge of the etiologic hypotheses in the analysis of case–control studies? . Am J Epidemiol 140 : 1054 – 1055 .
18. Brownson RC , Reif JS , Alavanja MCR , Bal DG ( 1993 ). Cancer . In Chronic Disease Epidemiology and Control (eds Brownson RC , Remington PL , Davis JR ); American Public Health Association , Washington, DC , pp. 150 – 153 .
19. Bruning PR , Bonfrer JMG , Hart AAM et al. ( 1992 ). Body measurements, estrogen availability and the risk of human breast cancer
– a case–control study . Int J Cancer 51 : 14 – 19 .
20. Bruning PR , Bonfrer JMG , van Noord PAH et al. ( 1992 ). Insulin resistance and breast cancer
risk . Int J Cancer 52 : 511 – 516 .
21. Bryan RM , Mercer RJ , Bennett RC et al. ( 1984 ). Androgen receptors in breast cancer
. Cancer 54 : 2436 – 2440 .
22. Bush TL , Helzlsouer KJ ( 1993 ). Tamoxifen for the primary prevention of breast cancer
: a review
and critique of the concept and trial . Epidemiol Rev 15 : 233 – 244 .
23. Campagnoli C , Biglia N , Belforte P et al. ( 1992 ). Post-menopausal breast cancer
risk: oral estrogen treatment and abdominal obesity induce opposite changes in possibly important biological variables . Eur J Gynaecol Oncol 13 : 139 – 154 .
24. Carroll KK ( 1998 ). Obesity as a risk factor for certain types of cancer . Lipids 33 : 1055 – 1059 .
25. Cauley JA , Lucas FL , Kuller LH et al. ( 1996 ). Bone mineral density and risk of breast cancer
in older women . JAMA 276 : 1404 – 1408 .
26. Chu SY , Lee NC , Wingo PA et al. ( 1991 ). The relationship between body mass and breast cancer
among women enrolled in the Cancer and Steroid Hormone Study . J Clin Epidemiol 44 : 1197 – 1206 .
27. Coates RJ , Uhler RJ , Hall HI et al. ( 1999 ). Risk of breast cancer
in young women in relation to body size and weight
gain in adolescence and early adulthood . Br J Cancer 81 : 167 – 184 .
28. Cold S , Hansen S , Overvad K , Rose C ( 1998 ). A woman's build and the risk of breast cancer
. Eur J Cancer 34 : 1163 – 1174 .
29. Colditz GA ( 1999 ). Economic costs of obesity and inactivity . Med Sci Sports Exerc 31 : S663 – S667 .
30. de Stavola BL , Wang DY , Allen DS et al. ( 1993 ). The association of height
, menstrual and reproductive events with breast cancer
: results from two prospective studies on the island of Guernsey (United Kingdom) . Cancer Causes Control 4 : 331 – 340 .
31. de Waard F , Baanders-van Halewijn EA ( 1974 ). A prospective study in general practice on breast cancer
risk in postmenopausal women . Int J Cancer 14 : 153 – 160 .
32. de Waard F , Laive JWJ , Baanders-van Halewijn EA ( 1960 ). On the bimodal age distribution of mammary carcinoma . Br J Cancer 14 : 437 – 451 .
33. Del Guidice ME , Fantus IG , Ezzat S et al. ( 1998 ). Insulin and related factors in premenopausal breast cancer
risk . Breast Cancer
Res Treat 47 : 111 – 120 .
34. den Tonkelaar I , Seidell JC , Collette HJ , de Waard F ( 1992 ). Obesity and subcutaneous fat patterning in relation to breast cancer
in postmenopausal women participating in the Diagnostic Investigation of Mammary Cancer Project . Cancer 69 : 2663 – 2667 .
35. den Tonkelaar I , Seidell JC , Collette HJ , de Waard F ( 1994 ). A prospective study on obesity and subcutaneous fat patterning in relation to breast cancer
in post-menopausal women participating in the DOM project . Br J Cancer 69 : 352 – 357 .
36. den Tonkelaar I , Seidell JC , Collette HJ ( 1995 ). Body fat distribution in relation to breast cancer
in women participating in the DOM-project . Breast Cancer
Res Treatment 34 : 55 – 61 .
37. Dupont WD , Page DL ( 1987 ). Breast cancer
risk associated with proliferative disease, age at first birth, and a family history of breast cancer
. Am J Epidemiol 125 : 769 – 779 .
38. Egan KM , Newcomb PA , Titus-Ernstoff L et al. ( 1999 ). The relation of breast size to breast size risk in postmenopausal women (United States) . Cancer Causes Control 10 : 115 – 118 .
39. Enriori CL , Osini W , Cremona MD et al. ( 1986 ). Decrease of circulating levels of SHBG in post-menopausal obese women as a risk factor in breast cancer
: reversible effect of weight
loss . Gynecol Oncol 23 : 77 – 86 .
40. Evans D , Hoffman RG , Kalkhoff RK , Kissebah AH ( 1983 ). Relationship of androgenic activity to body fat topography, fat cell morphology, and metabolite aberrations in premenopausal women . J Clin Endocrinol Metab 57 : 304 – 310 .
41. Flegal K , Carroll M , Kuczmarski R , Johnson C ( 1998 ). Overweight and obesity in the United States: prevalence and trends, 1960–1994 . Int J Obes 22 : 39 – 47 .
42. Folsom AR , Vitelli LL , Lewis CE et al. ( 1998 ). Is fasting insulin concentration inversely associated with rate of weight
gain? Contrasting findings from the CARDIA and ARIC study cohorts . Int J Obes 22 : 48 – 54 .
43. Folsom AR , Kaye SA , Prineas RJ et al. ( 1990 ). Increased incidence of carcinoma of the breast associated with abdominal adiposity in postmenopausal women . Am J Epidemiol 131 : 794 – 803 .
44. Folsom AR , Kushi LH , Anderson KE et al. ( 2000 ). Associations of general and abdominal obesity with multiple health outcomes in older women: the Iowa Women's Health Study . Arch Intern Med 160 : 2117 – 2128 .
45. Franceschi S , Favero A , La Vecchia C et al. ( 1996 ). Body size indices and breast cancer
risk before and after menopause . Int J Cancer 67 : 181 – 186 .
46. Glenny A-M , O'Meara S , Melville A , Sheldon TA , Wilson C ( 1997 ). The treatment and prevention of obesity: a systematic review
of the literature . Int J Obes 21 : 715 – 737 .
47. Goodman MT , Cologne JB , Moriwaki H et al. ( 1997 ). Risk factors for primary breast cancer
in Japan: 8-year follow-up of atomic bomb survivors . Prev Med 26 : 144 – 153 .
48. Gunnell D , Berney L , Holland P et al. ( 2000 ). How accurately are height
and leg length reported by the elderly, and how closely are they related to measurements recorded in childhood? . Int J Epidemiol 29 : 456 – 464 .
49. Haffner SM , Dunn JF , Katz MS ( 1992 ). Relationship of sex-hormone binding globulin to lipid, lipoprotein, glucose, and insulin concentrations in postmenopausal women . Metabolism 41 : 278 – 284 .
50. Hall IJ , Newman B , Millikan RC , Moorman PG ( 2000 ). Body size and breast cancer
risk in black women and white women. The Carolina Breast Cancer
Study . Am J Epidemiol 151 : 754 – 764 .
51. Hankinson SE , Willett WC , Colditz GA et al. ( 1998 ). Circulating concentrations of insulin-like growth factor-I and risk of breast cancer
. Lancet 351 : 1393 – 1396 .
52. Hankinson SE , Willett WC , Manson JE et al. ( 1998 ). Plasma sex steroid hormone levels and risk of breast cancer
in postmenopausal women . J Natl Cancer Inst 90 : 1292 – 1299 .
53. Harris RE , Namboodiri KK , Wynder EL ( 1992 ). Breast cancer
risk: effects of estrogen replacement therapy and body mass . J Natl Cancer Inst 84 : 1575 – 1582 .
54. Helmrich SP , Shapiro S , Rosenberg L et al. ( 1983 ). Risk factors for breast cancer
. Am J Epidemiol 117 : 35 – 45 .
55. Henderson BE , Ross RK , Judd HL , Krailo MD , Pike MC ( 1985 ). Do regular ovulatory cycles increase breast cancer
risk? . Cancer 56 : 1206 – 1208 .
56. Hill JO , Melanson EL ( 1999 ). Overview of the determinants of overweight and obesity: current evidence and research issues . Med Sci Sports Exer 31 : S515 – S521 .
57. Hirohata T , Nomura AM , Kolonel LN ( 1977 ). Breast size and cancer . BMJ 2 : 641 .
58. Hsieh C , Trichopoulos D ( 1991 ). Breast size, handedness and breast cancer
risk . Eur J Cancer 27 : 131 – 135 .
59. Huang Z , Hankinson SE , Colditz GA et al. ( 1997 ). Dual effects of weight
gain on breast cancer
risk . JAMA 278 : 1407 – 1411 .
60. Huang Z , Willett WC , Colditz GA et al. ( 1999 ). Waist circumuference, waist:hip ratio, and risk of breast cancer
in the Nurses’ Health Study . Am J Epidemiol 150 : 1316 – 1324 .
61. Hunter DJ , Willett WC ( 1993 ). Diet, body size, and breast cancer
. Epidemiol Rev 15 : 110 – 132 .
62. Hunter DJ , Spiegelman D , Adami H-O et al. ( 1996 ). Cohort studies of fat intake and the risk of breast cancer
– a pooled analysis . N Engl J Med 334 : 356 – 361 .
63. Ingram D , Nottage E , Ng S et al. ( 1989 ). Obesity and breast disease: the role of the female sex hormones . Cancer 64 : 1049 – 1053 .
64. Ingram DM , Nottage EM , Willcox DL , Roberts A ( 1990 ). Oestrogen binding and risk factors for breast cancer
. Br J Cancer 61 : 303 – 307 .
65. Jeffrey RW ( 1996 ). Does weight
cycling present a health risk? . Am J Clin Nutr 63 : 452S – 455S .
66. Jurgens RW , Downey LJ , Abernethy WD , Cutler NR , Conrad J ( 1992 ). A comparison of circulating hormone levels in postmenopausal women receiving hormone replacement therapy . Am J Obstet Gynecol 167 : 459 – 460 .
67. Juul A , Bang P , Hertel NT et al. ( 1994 ). Serum insulin-like growth factor I in 1030 healthy children, adolescents, and adults: relation to age, sex, stage of puberty, testicular size, and body mass index
. J Clin Endocrinol Metab 78 : 744 – 752 .
68. Kaaks R ( 1996 ). Nutrition, hormones, and breast cancer
: Is insulin the missing link? . Cancer Causes Control 7 : 605 – 625 .
69. Kaaks R , van Noord PAH , Den Tonkelaar I et al. ( 1998 ). Breast-cancer incidence in relation to height
and body-fat distribution in the Dutch ‘Dom’ cohort . Int J Cancer 76 : 647 – 651 .
70. Kampert JB , Whittemore AS , Paffenbarger RS Jr ( 1988 ). Combined effects of childbearing, menstrual events, and body size on age-specific breast cancer
risk . Am J Epidemiol 128 : 962 – 979 .
71. Katariya RN , Forrest PM , Gravelle JH ( 1974 ). Breast volumes in cancer of the breast . Br J Cancer 29 : 270 – 273 .
72. Kato I , Clifford B , Bleich A et al. ( 1995 ). A nested case–control study of mammographic patterns, breast volume, and breast cancer
(New York City, NY, United States) . Cancer Causes Control 6 : 431 – 438 .
73. Kaye SA , Folsom AR , Soler JT , Prineas RJ , Potter JD ( 1991 ). Associations of body mass and fat distribution with sex hormone concentrations in postmenopausal women . Int J Epidemiol 20 : 151 – 156 .
74. Kelsey JL , Baron J ( 1997 ). Weight
and risk for breast cancer
. JAMA 278 : 1448 – 1449 .
75. Kelsey JL , Bernstein L ( 1996 ). Epidemiology and prevention of breast and endometrial cancers . Am J Clin Nutr 63 : 437S – 441S .
76. Kelsey JL , Gammon MD , John EM ( 1993 ). Reproductive factors and breast cancer
. Epidemiol Rev 15 : 36 – 47 .
77. Key TJA , Pike MC ( 1988 ). The role of oestrogens and progestogens in the epidemiology and prevention of breast cancer
. Eur J Cancer Clin Oncol 24 : 29 – 43 .
78. Kohlmeier L , Kohlmeier M ( 1995 ). Adipose tissue as a medium for epidemiologic exposure assessment . Environ Health Perspect 103 : 99 – 106 .
79. Kolonel LN , Nomura AM , Lee J , Hirohata T ( 1986 ). Anthropometric indicators of breast cancer
risk in postmenopausal women in Hawaii . Nutr Cancer 8 : 247 – 256 .
80. Krotkiewski M , Seidell JC , Björntorp P ( 1990 ). Glucose tolerance and hyperinsulinemia in obese women: role of adipose tissue distribution, muscle fiber characteristics and androgens . J Intern Med 228 : 385 – 392 .
81. Kumar NB , Lyman GH , Allen K , Cox CE , Schapira DV ( 1995 ). Timing of weight
gain and breast cancer
risk . Cancer 76 : 243 – 249 .
82. Kyogoku S , Hirohata T , Takeshita S , Hirota Y , Shigematsu T ( 1990 ). Anthropometric indicators of breast cancer
risk in Japanese women in Fukuoka . Jpn J Cancer Res 81 : 731 – 737 .
83. Le Marchand L , Kolonel LN , Earl ME , Mi M-P ( 1988 ). Body size at different periods of life and breast cancer
risk . Am J Epidemiol 128 : 137 – 152 .
84. Li CI , Malone KE , White E , Daling JR ( 1997 ). Age when maximum height
is reached as a risk factor for breast cancer
among young U.S. women . Epidemiology 8 : 559 – 565 .
85. Li CI , Stanford JL , Daling JR ( 2000 ). Anthropometric variables in relation to risk of breast cancer
in middle-aged women . Int J Epidemiol 29 : 208 – 213 .
86. Lilienfeld AM ( 1956 ). The relationship of cancer of the female breast to artificial menopause and marital status . Cancer 9 : 927 – 934 .
87. London SJ , Colditz GA , Stampfer JJ et al. ( 1989 ). Prospective study of relative weight
, and risk of breast cancer
. JAMA 262 : 2853 – 2858 .
88. MacKenzie I ( 1955 ). The population of mammary cancer in rats using oestrogens . Br J Cancer 9 : 284 – 299 .
89. Maggino T , Pirrone F , Velluti F , Bucciante G ( 1993 ). The role of endocrine factors and obesity in hormone-dependent gynecological neoplasias . Eur J Gynaecol Oncol 14 : 119 – 126 .
90. Magnusson C , Baron J , Persson I et al. ( 1998 ). Body size in different periods of life and breast cancer
risk in postmenopausal women . Int J Cancer 76 : 29 – 34 .
91. Männistö S , Pietinen P , Pyy M et al. ( 1996 ). Body size indicators and risk of breast cancer
according to menopause and estrogen-receptor status . Int J Cancer 68 : 8 – 13 .
92. McTiernan A ( 2000 ). Associations between energy balance and body mass index
and risk of breast carcinoma in women from diverse racial and ethnic backgrounds in the U.S . Cancer 88 : 1248 – 1255 .
93. McTiernan A , Schwartz RS , Potter J , Bowen D ( 1999 ). Exercise clinical trials in cancer prevention
research: a call to action . Cancer Epidemiol Biomarkers Prev 8 : 201 – 207 .
94. McTiernan A , Ulrich CM , Yancey D et al. ( 1999 ). The Physical Activity for Total Health (PATH) Study: rationale and design . Med Sci Sports Exerc 31 : 1307 – 1312 .
95. Meldrum D , Davidson B , Tataryn L , Judd HL ( 1981 ). Changes in circulating steroids with aging in postmenopausal women . Obstet Gynecol 57 : 624 – 628 .
96. Mezzetti M , La Vecchia C , Decarli A et al. ( 1998 ). Population attributable risk for breast cancer
: diet, nutrition, and physical exercise . J Natl Cancer Inst 90 : 389 – 394 .
97. Michels KB , Trichopoulos D , Robins JM et al. ( 1996 ). Birthweight as a risk factor for breast cancer
. Lancet 348 : 1542 – 1546 .
98. Mokhad AH , Serdula MK , Dietz WH et al. ( 1999 ). The spread of the obesity epidemic in the United States, 1991–1998 . JAMA 282 : 1519 – 1522 .
99. Moore J , Clark G , Bulbrook R et al. ( 1982 ). Serum concentrations of total and non-protein bound oestradiol in patients with breast cancer
and in normal controls . Int J Cancer 29 : 17 – 21 .
100. Muls E , Kempen K , Vansant G , Saris W ( 1995 ). Is weight
cycling detrimental to health? A review
of the literature in humans . Int J Obes 19 : S46 – S50 .
101. Must A , Spadano J , Coakley EH et al. ( 1999 ). The disease burden associated with overweight and obesity . JAMA 282 : 1523 – 1529 .
102. Ng E-H , Gao F , Ji C-H , Ho G-H , Soo K-C ( 1997 ). Risk factors for breast carcinoma in Singaporean Chinese women . Cancer 80 : 725 – 731 .
103. Nguyen TV , Center JR , Eisman JA ( 2000 ). Association between breast cancer
and bone mineral density: the Dubbo Osteoporosis Epidemiology study . Maturitas 36 : 27 – 34 .
104. O'Brien K , Caballero B ( 1997 ). High bone mass as a marker for breast cancer
risk . Nutr Rev 55 : 284 – 286 .
105. O'Dea JP , Wieland RG , Hallberg MC , Llerena LA , Zorn EM , Genuth SM ( 1979 ). Effect of dietary weight
loss on sex steroid binding sex steroids, and gonadotropins in obese postmenopausal women . J Lab Clin Med 93 : 1004 – 1048 .
106. Okosun IS , Liao Y , Rotimi CN , Prewitt TE , Cooper RS ( 2000 ). Abdominal adiposity and clustering of multiple metabolic syndrome in white, black and Hispanic Americans . Ann Epidemiol 10 : 263 – 270 .
107. Osorio M , Torres J , Moya F et al. ( 1996 ). Insulin-like growth factor (IGFs) and IGF binding proteins-1, -2 and -3 in newborn serum: relationships to fetoplacental growth at term . Early Hum Dev 46 : 15 – 26 .
108. Paffenbarger RS , Kampert JB , Chang H ( 1980 ). Characteristics that predict risk of breast cancer
before and after menopause . Am J Epidemiol 112 : 258 – 268 .
109. Pathak DR , Whittemore AS ( 1992 ). Combined effects of body size, parity, and menstrual events on breast cancer
incidence in seven countries . Am J Epidemiol 135 : 153 – 168 .
110. Peacock SL , White E , Daling JR , Voigt LF , Malone KE ( 1999 ). Relation between obesity and breast cancer
in young women . Am J Epidemiol 149 : 339 – 346 .
111. Petrek JA , Peters M , Cirrincione C , Rhodes D , Bajorunas D ( 1993 ). Is body fat topography a risk factor for breast cancer
? . Ann Intern Med 118 : 356 – 362 .
112. Pollak M ( 2000 ). Insulin-like growth factor physiology and cancer risk . Eur J Cancer 36 : 1224 – 1228 .
113. Popkin BM , Doak CM ( 1998 ). The obesity epidemic is a worldwide phenomenon . Nutr Rev 56 : 106 – 114 .
114. Potischman N , Swanson CA , Siiteri P et al. ( 1996 ). Reversal of relation between body mass and endogenous estrogen concentrations with menopausal status . J Natl Cancer Inst 88 : 756 – 758 .
115. Prentice AM , Jebb SA , Goldberg GR et al. ( 1992 ). Effects of weight
cycling on body composition . Am J Clin Nutr 56 : 209S – 216S .
116. Pujol P , Galtier-Dereure F , Bringer J ( 1997 ). Obesity and breast cancer
risk . Human Reprod 12 : 116 – 125 .
117. Radimer K , Siskind V , Bain C , Schofield F ( 1993 ). Relation between anthropometric indicators and risk of breast cancer
among Australian women . Am J Epidemiol 138 : 77 – 89 .
118. Ravussin E , Swinburn BA ( 1992 ). Pathophysiology of obesity . Lancet 340 : 404 – 408 .
119. Reed M , Cheng R , Noel C , Dudley HA , James VH ( 1983 ). Plasma levels of estrone, estrone sulfate, and estradiol and the percentage of unbound estradiol in postmenopausal women with and without breast disease . Cancer Res 43 : 3940 – 3943 .
120. Rimm EB , Stampfer MJ , Colditz GA et al. ( 1990 ). Validity of self-reported waist and hip circumferences in men and women . Epidemiology 1 : 466 – 473 .
121. Ruggeri B , Klurfeld D , Kritchevsky D , Furlanetto RW ( 1998 ). Calorie restriction and 7,12–dimethylbenz(a)anthracene-induced mammary tumour growth in rats: alterations in circulation insulin, insulin-like growth factors I and II, and epidermal growth factor . Cancer Res 49 : 4135 – 4141 .
122. Schapira DV , Kumar NB , Lyman GH , Cox CE ( 1990 ). Abdominal obesity and breast-cancer risk . Ann Intern Med 112 : 182 – 186 .
123. Schapira DV , Kumar NB , Lyman GH ( 1991 ). Estimate of breast cancer
risk reduction with weight
loss . Cancer 67 : 2622 – 2625 .
124. Schapira DV , Wolff PA , Kumar NB et al. ( 1994 ). The effect of weight
loss on estimated breast cancer
risk and sex hormone levels . Oncol Reports 1 : 613 – 617 .
125. Shifren JL , Schiff I ( 2000 ). The aging ovary . J Womens Health Gend Based Med 9 : S3 – S7 .
126. Schindler AE ( 1997 ). Obesity and cancer risk in women . Arch Gynecol Obstet 261 : 21 – 24 .
127. Scutt D , Manning JT , Whitehouse GH , Leinster SJ , Massey CP ( 1997 ). The relationship between breast asymmetry, breast size and the occurrence of breast cancer
. Br J Radiol 70 : 1017 – 1021 .
128. Sellers TA , Kushi LH , Potter JD et al. ( 1992 ). Effect of family history, body-fat distribution, and reproductive factors on the risk of post-menopausal breast cancer
. N Engl J Med 326 : 1323 – 1329 .
129. Senie RT , Saftla AF , Brinton LA , Hoover RN ( 1993 ). Is breast size a predictor of breast cancer
risk or the laterality of the tumor? . Cancer Causes Control 4 : 203 – 208 .
130. Soini I ( 1977 ). Risk factors of breast cancer
in Finland . Int J Epidemiol 6 : 365 – 373 .
131. Soler JT , Folsom AR , Kaye SA , Prineas RJ ( 1989 ). Associations of abdominal adiposity, fasting insulin, sex hormone binding globulin, and estrone with lipids and lipoproteins in post-menopausal women . Atherosclerosis 79 : 21 – 27 .
132. Solomon CG , Manson JE ( 1997 ). Obesity and mortality: a review
of the epidemiologic data . Am J Clin Nutr 66 : 1044S – 1050S .
133. Sonnenschein E , Toniolo P , Terry MB et al. ( 1999 ). Body fat distribution and obesity in pre- and post-menopausal breast cancer
. Int J Epidemiol 28 : 1026 – 1031 .
134. Spicer LJ , Francisco CC ( 1997 ). The adipose gene product, leptin: evidence of a direct inhibitory role in ovarian function . Endocrinology 135 : 3374 – 3379 .
135. Stewart AJ , Johnson MD , May FEB , Westley BR ( 1990 ). Role of insulin-like growth factors and the type I insulin-like growth factor receptor in the estrogen-stimulated proliferation of human breast cancer
cells . J Biol Chem 254 : 2117 – 2218 .
136. Stoll BA ( 1992 ). Does extra height
justify a higher risk of breast cancer
? . Ann Oncol 3 : 29 – 30 .
137. Stoll BA ( 1995 ). Timing of weight
gain in relation to breast cancer
risk . Ann Oncol 6 : 245 – 248 .
138. Stoll BA ( 1998 ). Teenage obesity in relation to breast cancer
risk . Int J Obes 22 : 1035 – 1040 .
139. Stoll BA ( 1999 ). Perimenopausal weight
gain and progression of breast cancer
precursors . Cancer Detect Prev 23 : 31 – 36 .
140. Swanson CA , Brinton LA , Taylor PR et al. ( 1989 ). Body size and breast cancer
risk assessed in women participating in the Breast Cancer
Detection Demonstration Project . Am J Epidemiol 130 : 1133 – 1141 .
141. Swanson CA , Jones DY , Schatzkin A , Brinton LA , Ziegler RG ( 1988 ). Breast cancer
risk assessed by anthropometry in the NHANES I Epidemiological Follow-up Study . Cancer Res 48 : 5363 – 5367 .
142. Swanson CA , Coates RJ , Schoenberg JB et al. ( 1996 ). Body size and breast cancer
risk among women under age 45 years . Am J Epidemiol 143 : 698 – 706 .
143. Tannenbaum A ( 1942 ). Genesis and growth of tumors. II. Effects of caloric restriction per se
. Cancer Res 2 : 460 – 467 .
144. Tavani A , Pregnolato A , La Vecchia C et al. ( 1996 ). Breast size and breast cancer
risk . Eur J Cancer Prev 5 : 337 – 342 .
145. Thurfjell E , Hsieh C-C , Lipworth L et al. ( 1996 ). Breast size and mammographic pattern in relation to breast cancer
risk . Eur J Cancer Prev 5 : 37 – 41 .
146. Toniolo PG ( 1997 ). Endogenous estrogens and breast cancer
risk: the case for prospective cohort studies . Environ Health Perspect 105 : 587 – 592 .
147. Törnberg SA , Cartensen JM ( 1994 ). Relationship between Quetelet's index and cancer of breast and female genital tract in 47,000 women followed for 25 years . Br J Cancer 69 : 358 – 361 .
148. Törnberg SA , Holm LE , Cartensen JM ( 1988 ). Breast cancer
risk in relation to serum cholesterol, serum beta-lipoprotein, height
, and blood pressure . Acta Oncol 27 : 31 – 37 .
149. Trentham-Dietz A , Newcomb PA , Storer BE et al. ( 1997 ). Body size and risk of breast cancer
. Am J Epidemiol 145 : 1011 – 1019 .
150. Trentham-Dietz A , Newcomb PA , Egan KM et al. ( 2000 ). Weight
change and risk of postmenopausal breast cancer
(United States) . Cancer Causes Control 11 : 533 – 542 .
151. Tretli S ( 1989 ). Height
in relation to breast cancer
morbidity and mortality. A prospective study of 570,000 women in Norway . Int J Cancer 44 : 23 – 30 .
152. Trichopoulos D , Lipman RD ( 1992 ). Mammary gland mass and breast cancer
risk . Epidemiology 3 : 523 – 526 .
153. Ursin G , Longnecker MP , Haile RW , Greenland S ( 1995 ). A meta-analysis of body mass index
and risk of premenopausal breast cancer
. Epidemiology 6 : 137 – 141 .
154. Valaoras VG , McMahon B , Trichopoulos D , Polychronopoulou A ( 1969 ). Lactation and reproductive histories of breast cancer
patients in greater Athens, 1965–67 . Int J Cancer 4 : 350 – 363 .
155. van den Brandt PA , Dirx MJM , Ronckers CM , van den Hoogen P , Goldbohm RA ( 1997 ). Height
change, and postmenopausal breast cancer
risk: the Netherlands Cohort Study . Cancer Causes Control 8 : 39 – 47 .
156. van den Brandt PA , Spiegelman D , Yaun S-S et al. ( 2000 ). Pooled analysis of prospective cohort studies on height
, and breast cancer
risk . Am J Epidemiol 152 : 514 – 527 .
157. VanItallie TB ( 1996 ). Prevalence of obesity . Endocrinol Metab Clin North Am 25 : 887 – 905 .
158. Vatten LJ , Kvinnsland S ( 1990 ). Body mass index
and risk of breast cancer
: a prospective study of 23,826 Norwegian women . Int J Cancer 45 : 440 – 444 .
159. Vatten LJ , Kvinnsland S ( 1992 ). Prospective study of height
, body mass index
and risk of breast cancer
. Acta Oncol 31 : 195 – 200 .
160. Weiss HA , Brinton LA , Potischman NA et al. ( 1999 ). Breast cancer
risk in young women and history of selected medical conditions . Int J Epidemiol 5 : 816 – 823 .
161. Willett WC , Browne ML , Bain C et al. ( 1985 ). Relative weight
and risk of breast cancer
among premenopausal women . Am J Epidemiol 122 : 731 – 740 .
162. Williamson DF , Kahn HS , Remington PL , Anda RF ( 1990 ). The 10-year incidence of overweight and major weight
gain in US adults . Arch Intern Med 150 : 665 – 672 .
163. Wing RR ( 1992 ). Weight
cycling in humans: a review
of the literature . Ann Behav Med 14 : 113 – 119 .
164. Wing RR , Matthews KA , Kuller LH et al. ( 1992 ). Environmental and familial contributions to insulin levels and change in insulin levels in middle-aged women . JAMA 268 : 1890 – 1895 .
165. Women's Health Initiative Study Group. ( 1998 ). Design of the Women's Health Initiative Clinical Trial and Observational Study . Controlled Clin Trials 19 : 61 – 109 .
166. Wynder EL , Bross IJ , Hirayama T ( 1960 ). A study of the epidemiology of cancer of the breast . Cancer 13 : 559 – 601 .
167. Yee D , Paik S , Levovic GS et al. ( 1989 ). Analysis of insulin-like growth factor I gene expression in malignancy: evidence for a paracrine role in human breast cancer
. Mol Endocrinol 3 : 509 – 517 .
168. Yong LC , Brown CC , Schatzkin A , Schairer C ( 1996 ). Prospective study of relative weight
and risk of breast cancer
: the Breast Cancer
Detection Demonstration Project follow-up study, 1979 to 1987–1989 . Am J Epidemiol 143 : 985 – 995 .
169. Zhang Y , Kiel DP , Kreger BE et al. ( 1997 ). Bone mass and the risk of breast cancer
among postmenopausal women . N Engl J Med 336 : 611 – 617 .
170. Ziegler RG ( 1997 ). Anthropometry and breast cancer
. J Nutr 127 : 924S – 928S .
171. Ziegler RG , Hoover RN , Nomura AMY et al. ( 1996 ). Relative weight
, and breast cancer
risk in Asian–American women . J Natl Cancer Inst 88 : 650 – 660 .