Abstract
Obesity is an increasing worldwide epidemic. With an increased prevalence of obesity there is an increase in obesity-related cancers, such as breast cancer. Although reproductive and lifestyle choices are among the best-recognized risk factors for breast cancer, few of these can be modified readily by the individual. Obesity is unlike these other risk factors since it can be modified and controlled. Breast cancer prognosis is worse in patients who are obese, and epidemiological data suggests that obesity is a significant risk factor for postmenopausal breast cancer. Addressing the obesity epidemic, at both an individual and public health level, is expected to have a significant impact on breast cancer incidence and mortality.
Keywords
Obesity is a major health problem. Despite of considerable effort by the scientific and healthcare professions to understand and successfully treat obesity, its incidence continues to rise. Obesity-related costs to society are staggering [1,2]. The health consequences associated with obesity include an increased risk of noninsulin-dependent diabetes mellitus (DM), cancer, cardiovascular disease and the metabolic syndrome [3]. Calle and colleagues concluded that excess body weight is the cause of approximately 20% of cancer deaths in the USA in women aged 50 years or older [4]. Breast cancer accounts for 31% of all cancers and 15% of cancer deaths among women in the USA, with over 200,000 women reportedly newly diagnosed with breast cancer each year [5]. With the increasing prevalence of both obesity and breast cancer, it is critical to understand how obesity may influence susceptibility to breast cancer. This article will examine evidence that the incidence of breast cancer and its mortality is linked to body weight and body fat distribution. The proposed mechanisms by which obesity and breast cancer are associated will be briefly discussed.
It is now generally accepted that the maintenance of body weight, and especially of body-fat stores, depends on a remarkably efficient regulatory system [3]. Body fat varies among different environments as well as among individuals, the variation due in part to genetics, early nutritional experience and average exercise level, among other variables. Nonetheless, in any given environment, individuals with adequate food availability rigorously maintain and defend a particular quantity of fat [3]. Obesity has been identified as a risk factor for breast cancer, especially in postmenopausal women [6–8]. Improved therapy and earlier diagnosis have decreased the mortality rate from breast cancer in the USA, but with aging demographics and the increase in postmenopausal obesity, breast cancer incidence is expected to increase. In fact, women in the USA have among the highest breast cancer rates in the world. Importantly, this cannot be due to increased genetic susceptibility, as less than 10% of breast cancers are attributed to gene mutations, such as BRCA1, and immigrants experience higher rates of breast cancer in the USA than in their home countries. The rising rate of obesity and interactions between low-penetrance, cancer-susceptibility genes and environmental factors including nutrition and lifestyle, are all thought to be major contributors to the increased risk for breast cancer observed in American women. Importantly, the frequency of breast cancer in western countries parallels the prevalence of insulin resistance and the metabolic syndrome, both consequences of obesity. This is most significant for postmenopausal women, where the overweight have up to a twofold greater risk of breast cancer. However, the association between excess body weight/obesity and cancer susceptibility in premenopausal women is less consistent [9].
Etiology of breast cancer
Risk factors for breast cancer include being female, early age of menarche, late age of menopause, nulliparity or late first full-term pregnancy, increased height and a family history of breast cancer [10]. Most of these risk factors are linked to breast cancer as they relate to hormonal stimulation of the breast. Epidemiological and animal studies clearly demonstrate that breast cancer is promoted by estrogen and progesterone stimulation. Lifestyle and genetic factors that tend to increase exposure to these hormones increase the risk. However, the relationship between sex hormones and breast cancer is complex and likely interacts with other factors. In addition to hormonal status, breast cancer is also stimulated by a number of growth hormones and cytokines, which may increase risk when present in excessive amounts. Importantly, the factors and cytokines are increased in obese women and this may be one mechanism by which increased body weight is directly related to increased incidence of breast cancer. For example, breast cancer may be promoted by insulin and insulin-like growth factor (IGF)-1, both of which are elevated in insulin-resistant states including diabetes and obesity. In support of this, noninsulin-dependent DM has been associated with an increased relative risk of breast cancer of 1.25 [11]. In a similar vein, the mechanisms by which obesity and Type II diabetes are associated with breast cancer could be attributed to body-fat distribution, particularly when fat is distributed in the visceral depot, also known as central obesity. Centrally stored adipose tissue, as is found in men and postmenopausal women, is strongly associated with cardiovascular disease and DM [12]. This occurs because the adipose tissues in the trunk, particularly the visceral stores, have endocrine functions that are uniquely regulated relative to the peripheral adipose depots [13]. Central adiposity increases the risk of the metabolic syndrome, which is characterized by insulin resistance, dyslipidemias and increased circulating IGF-1 levels [13].
Elevated estrogen levels have also been determined to promote the development of postmenopausal breast cancer [14]. Circulating estrogens are synthesized in the stromal cells of the adipose tissue by enzymatic aromatization of C19 steroid androstenedione to yield estrone, and this occurs even after menopause, when ovarian production of estrogen has ceased [15]. In fact, adipose tissue can be considered an ‘extra-ovarian tissue’ where there is a conversion of estrogen to the more potent estradiol (reviewed by Rose [15,16]), and this conversion is thought to be more prevalent in obese individuals. Obesity-related increases in circulating estrogens have been associated with an elevated breast cancer risk and enhanced progression of estrogen receptor (ER)-positive cancers [7]. One large epidemiological study in Norway found that women who were obese in the years before their breast cancer diagnosis of ER-positive tumors were significantly more likely to die of their disease than lean women with ER-positive breast cancer [17].
Obesity & postmenopausal breast cancer
With few exceptions in the postmenopausal years, obesity is strongly correlated with breast cancer risk. The American Cancer Society's large, prospective Cancer Prevention Study II found that increasing body mass index (BMI) substantially increased breast cancer mortality, and they estimated that being overweight contributed to 30–50% of breast cancer deaths in the USA [18]. In addition, both the Framingham Study and the Nurses' Health Study demonstrated a direct correlation between current BMI and postmenopausal breast cancer risk, and this was associated with ‘early’ weight gain, that began when the women were in their early- or mid-twenties [19,20]. Finally, an estimated 16% of postmenopausal breast cancer could be attributed to early weight gain.
One important question is, what is the influence of adiposity on breast cancer risk in postmenopausal women? Is obesity, which is rampant in western culture, the primary causal factor or is it a confounder of other factors in a typical western lifestyle? Only limited data are available, but they suggest that obesity is an independent risk factor across populations. One study in Asian–American women found that increased BMI correlated to an increased risk of breast cancer in women in their forties and fifties, and that weight gain was particularly deleterious for women in their fifties. As with the other breast cancer risk factors, obesity seems to exert its effects in the postmenopausal years via an increase in estrogen synthesis and availability (Figure 1). As previously mentioned, during the period of diminishing ovarian estrogen synthesis, adrenal androgens, such as androstenedione, are readily converted to estrogens, such as estrone, via aromatase (cytochrome P19) present in the adipose tissue stroma [21]. Obesity-related increases in circulating estrogens have been associated with an elevated breast cancer risk and enhanced progression of ER-positive cancers [7]. An increased BMI is also correlated with a decrease in the levels of sex hormone-binding globulin (SHBG). Decreased SHBG levels allow increased levels of free estrogen, so under conditions of constant estrogen but diminished SHBG production, there is more available estrogen for receptor activation. Increased production and bioavailability of estrogen is undoubtedly one mechanism for increased risk of breast cancer in the obese during the postmenopausal years. One might then consider adipose tissue as a form of endogenous hormone replacement therapy.

Stimulation of breast tumors by adipose tissue.
Another important unanswered question is whether the risk is associated with overall increased adiposity or with a specific distribution of body fat. One recent meta-analysis of epidemiological studies suggests that postmenopausal breast cancer is related primarily to overall adiposity, as measured by the BMI, rather than any particular distribution of adipose tissue [8]. However, there are large discrepancies in the methods of fat measurement and estimates of distribution vary dramatically in their level of accuracy. For example, a number of studies have compared traditional measures, such as BMI, waist circumference and waist:hip ratio with radiological measurements of central adiposity. One study found that waist girth was a better predictor of visceral adiposity in men and women, independent of BMI, but that this was age dependent [22]. Similarly, a second study demonstrated that waist circumference was the best predictor of abdominal visceral fat – in their analyses, waist:hip ratio was the worst predictor [23]. Using bioimpedance as a measure of adiposity, another study found that waist circumference correlated better with bioimpedance than waist:hip ratio [24]. These types of data further suggest that measures may necessarily vary with study population (i.e., lean vs obese, young vs aged) and that a number of parameters will be necessary to discriminate how the biology and anthropometrics inter-relate. Given that not all studies included in recent meta-analyses used similar methodologies to determine body composition, strong conclusions remain elusive. One additional caveat is that as women age, they naturally accrue more adipose tissue in the central or visceral depot. Therefore, central obesity is common in postmenopausal women [25] and separating obesity itself from central obesity is extremely difficult in this population. Nonetheless, what can be concluded is that insulin resistance, which is associated with central obesity, is a risk factor for breast cancer. Therefore, fat accrual in the central adipose depot would be expected to be associated with susceptibility for breast cancer. Standardization with respect to body-fat measurements and distribution measurements will confirm this association.
Obesity & premenopausal breast cancer
Data from epidemiological studies that include women with premenopausal breast cancer overall support the concept that increased adiposity is protective [9]. Data suggest that free estradiol may be higher in premenopausal women with central adiposity [26], but estrogen:androgen ratios are decreased [27]. The question is whether premenopausal breast cancer is in some way related to specific adipose depots, some of which may be protective and some of which may promote cancer formation.
A large number of prospective and case-control studies from highly disparate populations indicate an inverse correlation between body weight and breast cancer risk (reviewed in [9,19,28–30]). In the New York University Women's Health Study and European Prospective Investigation into Cancer and nutrition (EPIC) study, premenopausal breast cancer risk was inversely correlated with BMI; however, when adjusted for BMI, waist and hip circumference were associated with increased breast cancer risk [31,32]. Other studies also suggest that central adiposity and adult weight gain are contributors to an increased risk of breast cancer in premenopausal women [33–36]. Yet a large, multicenter US case-control study in women aged under 45 years found that while increased BMI was protective, there was no effect of central adiposity [37]. However, some controversy exists, as indicated in the meta-analysis by Pathak and colleagues [38]. This analysis suggested that the obesity-dependent reduction in breast cancer risk in premenopausal women is unique to high-risk populations. Consistent with this hypothesis, BMI in low- (Taiwan and Japan) and moderate-risk (Brazil, Greece and Yugoslavia) countries had a positive correlation with premenopausal breast cancer [38]. However, there are other studies that found no effect of BMI at diagnosis on premenopausal risk [39–41]. Finally, an interesting population case-control study in Alberta (Canada) found that in nulliparous women, increased BMI increased risk, but decreased the risk in parous women [42].
An admittedly tenuous interpretation of these data may be as follows: in populations of low-risk and low-BMI women, obesity increases the incidence of premenopausal breast cancer. In high-risk, obese populations, there may be a promotional effect of central adiposity. An important exception may be women who have BRCA1 mutations, in whom weight loss between the ages of 18 and 30 years was strongly protective against breast cancer between the ages of 30 and 40 years [43]. Mechanistic explanations as to how obesity may be protective in high-risk populations include decreased estrogen levels in obese women who have anovulatory cycles [44], but the epidemiological evidence to support this is inconsistent [37,45].
Obesity & breast cancer prognosis
As previously discussed, there is solid evidence that for women who have breast cancer, obesity is associated with an increased mortality risk. In a recent comprehensive review [46], as well as in a more recent study [47], women with increased body weight and BMI were significantly more prone to disease recurrence and death at 5 or 10 years after diagnosis. Importantly, women who gained weight after diagnosis were also significantly more likely to die of their disease [46]. Unlike conflicting data on obesity and menopausal status, obesity and breast cancer prognosis appears to be important in both pre- and postmenopausal women, as indicated in a study of premenopausal cancer patients, which indicated that the higher the BMI, the more likely women are to die of their disease [48]. As with almost any other medical condition, obesity is a comorbid factor that may alter treatment strategies and consequences, which for breast cancer patients may include wound healing, lymphedema, congestive heart failure, endometrial cancer, sentinel node biopsy failure and insufficient chemotherapy [46,49,50].
Adipose tissue & the role of adipocyte hormones in breast cancer
In addition to adipose tissue estrogen synthesis, there are a number of other mechanisms by which obesity may increase the risk of breast cancer. Obesity, particularly obesity associated with an increase in central or visceral fat, is associated with hyperinsulinemia and the insulin-resistance syndromes [13]. There is both experimental and epidemiological support for insulin resistance increasing breast cancer risk [7]. Hyperinsulinemia has been correlated with both BMI and the risk of recurrence and mortality in breast cancer [51]. There are several mechanisms by which insulin may increase breast cancer risk (Figure 2). Both insulin and a similar peptide, IGF-1, exert a mitogenic effect on normal and neoplastic breast epithelial cells [7]. Insulin reduces levels of circulating binding proteins (BPs), IGFBP-1 and −3, which increase the bioavailable levels of IGF-1 [52]. Furthermore, both insulin and IGF-1 increase ovarian androgen synthesis and inhibit the production of SHBG. Increased androgen synthesis provides the substrate for aromatization into estrogen in the adipose tissue (Figure 1) [52]. Decreased SHBG allows more free, bioavailable estrogen to circulate. Thus, three stimulators of breast cell proliferation are available: insulin, IGF-1 and estrogen.

Visceral adiposity influences insulin resistance.
In addition to its effects on insulin, IGF-1 and sex steroids, adipose tissue secretes a wide range of enzymes, hormones and growth factors, including a number of adipocyte-derived biologically active polypeptides, which were grouped together by Matsuzawa and colleagues and termed adipocytokines [53–55]. The adipocytokines constitute a group of polypeptide growth factors and cytokines that are produced exclusively, or substantially, by white adipose tissue preadipocytes and mature adipocytes [13]. There are seven adipocytokines: leptin, adiponectin, tumor necrosis factor (TNF)-α, interleukin (IL)-6, heparin-binding epidermal growth factor (HB-EGF), vascular endothelial growth factor (VEGF) and hepatocyte growth factor (HGF), which differ with respect to their relevance to obesity and breast cancer (Figure 3).

Influence of dipocytokines on breast cancer.
Leptin is produced in adipose tissue and is therefore directly proportional to the degree of adiposity [56]. Plasma leptin concentrations in healthy women are positively correlated with both BMI and total body fat [57–59]. Leptin stimulates estrogen biosynthesis by induction of aromatase activity [60]. Leptin interferes with insulin signaling and, in Type 2 diabetes, plasma leptin levels were found to be correlated with the degree of insulin resistance, a relationship that was independent of BMI and body fat mass [61,62]. Leptin has a direct mitogenic effect on human breast cancer cells [63], and leptin's involvement in the insulin-resistance syndrome suggests that they act synergistically to stimulate breast cancer growth and metastasis. Thus, insulin resistance is accompanied by hyperleptinemia [64], which allows for excessive endocrine activity of these proteins at target sites, which potentially include breast epithelial tissue and vascular endothelial cells.
Adipose tissue TNF-α expression is also positively correlated with plasma insulin concentrations [65]. TNF-α is a powerful local regulator within adipose tissue, with TNF-α acting in both an autocrine and a paracrine manner, regulating the production of several cytokines and other adipokines [66]. Increased TNF-α secretion from adipocytes was shown to accompany diminished insulin sensitivity in obese individuals [67]. In abdominal obesity, high circulating TNF-α levels were associated with hyperinsulinemia and insulin resistance [68]. Insulin resistance has also been associated with human adipose tissue-derived IL-6 ([67] and references therein). Like leptin and TNF-α, IL-6 can stimulate estrogen biosynthesis by the induction of aromatase activity [69]. Human adipocytes secrete sizeable amounts of IL-6, and plasma IL-6 positively correlates with BMI [67].
Breast cancer is dependent upon angiogenesis, or new blood vessel formation, for its development and metastasis [70]. Of the adipocytokines, leptin, TNF-α, VEGF, HGF and HB-EGF have all been reported to be angiogenic in vivo [70]. While not directly angiogenic, IL-6 stimulates VEGF synthesis [71]. VEGF is a well-studied target for breast cancer therapy, as other adipocytokines may be in the near future [72]. Unlike the other adipocytokines, adiponectin is inversely associated with total adiposity and inversely associated with postmenopausal breast cancer risk, independent of BMI [73,74]. A potential mechanism for this inverse association may be adiponectin's antiangiogenic activity found in some [75], but not all, circumstances [76].
Is weight loss a viable mechanism for reducing breast cancer risk & increasing survival?
The foods we consume are thought to be tightly linked to an increased risk of disease. For breast cancer, the food items that are most tightly linked to breast carcinogenesis are alcohol and charred or grilled meats [77]. There are several other dietary components that may be linked to breast cancer susceptibility, but there is currently no strong evidence of a direct association. The first of these components is caloric excess. High energy intakes and higher levels of body adiposity hasten the onset of ménarche, and there is some indication that early ménarche may be associated with increased risk [78]. Dietary fatty acids, or diets high in fats, have not been demonstrated to be a major risk factor for breast cancer. However, one impact of high-fat diets is obesity. Furthermore, epidemiological as well as experimental data indicate that when a diet with a high fat content is consumed on a regular basis, the defended level of body adiposity actually increases, likely due in part to leptin resistance. There is a significant positive correlation between average dietary fat intake and the incidence of obesity, and when the average amount of fat in the diet increases over time, the incidence of obesity also increases.
One prediction based on the findings that increased body weight may be promotional with respect to breast cancer might be that weight loss would decrease the relative risk of the disease. Although the biochemical and epidemiological data would suggest that weight loss in adulthood may decrease the risk of breast cancer, particularly in postmenopausal women, there are very few prospective studies to verify this hypothesis. However, there are data from the Iowa Women's Health Study that suggest that weight loss may actually be protective [79]. Postmenopausal women who had an intentional weight loss of 20 lbs or more were found to have a significantly decreased risk of breast cancer. Furthermore, women who had lost weight and were not currently overweight had risks similar to nonoverweight women who had never lost weight [79]. This verifies a previous study that showed weight loss of greater than 20 lbs significantly decreased cancer-related mortalities [80]. Long-term intervention trials to decrease incidence or alter prognosis of breast cancer are not yet available. However, small-scale trials indicate that weight reduction in the breast cancer patient population is achievable [81–83]. Therefore, based on these studies, one could conclude that modest weight loss is also associated with an improvement in metabolic profiles.
In addition, dietary/energy restriction (DER) is arguably the most potent physiological approach to the prevention of experimentally induced breast cancer that has been identified to date [84]. Despite the fact that DER is a form of energy restriction, weight loss is not a critical component of DER. Rather, the key feature of DER is to provide limited access to nutrients [84]. It has been suggested that DER exerts its effect by altering one or more aspects of cell cycle regulation, due to the findings that energy restriction inhibits cell proliferation and increases cell death due to apoptosis [84]. Note that in a large meta-analysis of spontaneous mammary tumors in mice, energy restriction uniformly reduced tumor incidence. However, the beneficial effect was reduced with prolonged periods of DER treatment. This suggests that dietary restriction may delay but not prevent tumor formation. Importantly, the only study included in that meta-analysis that did not find any protective effect used the lowest level of energy restriction (23%) and the longest duration (2.5 years) [85].
Conclusion
Breast cancer incidence continues to increase around the world and is correlated with westernization. Similarly, obesity is becoming an international epidemic. Hormonal and lifestyle factors that modify breast cancer risk are well known but not readily modifiable on an individual basis. Obesity thus becomes one of the more urgent breast cancer risk factors to examine, since it is modifiable on an individual and population basis. Although prospective weight reduction trials are not yet available, current data suggest that normalizing peri- and postmenopausal weight would reduce breast cancer by 30%. How much weight reduction is needed to significantly decrease risk in our increasingly obese population is unknown. At all ages, weight normalization is expected to have beneficial effects on the prognosis of breast cancer patients and should be a long-term goal for women in conjunction with required surgery, radiation and chemotherapy. Within some populations, obesity appears to be somewhat protective against breast cancer in the premenopausal years. However, given the number and severity of other health consequences of obesity, including cardiovascular, metabolic and orthopedic problems, one is hesitant to encourage adiposity to prevent breast cancer. Furthermore, much more research is needed to examine premenopausal subpopulations relative to inheritance, family history and reproductive history to determine in whom and how obesity might have a protective effect. Most important is that obesity reversal/prevention be viewed at a public health level, even more than as an area of personal concern. Willpower and support groups, to date, are not up to the challenge. Employers, government, educators and public health providers must all participate in changing the ‘culture of adiposity’.
Future perspective
Obesity is a multifactorial epidemic. Current approaches for weight loss emphasize both caloric resitriction and increased exercise, as well as public health adjustments related to food sources and living spaces [86]. To curb the obesity epidemic, research will elucidate specifics in gene–environment interactions that will allow individually tailored dietary/exercise programs with pharmacological support. A future goal may be a screen for plasma chemistries and genetic polymorphisms that will permit individually designed therapies. Fat distribution will be key in this; if an individual has more visceral fat, the design would focus on exercise and increased metabolic rate to mobilize the visceral fat. A successful paradigm over the previous decades has been the reduction in cardiovascular disease, a disease linked to obesity, by reducing cholesterol via a multipronged approach of dietary modification, exercise and pharmaceuticals targeted towards lipid absorption and synthesis. Current knowledge on mechanisms of satiety, food craving, nutrient absorption pathways, and adipocyte distribution and metabolism will provide targets for pharmaceutical products for weight loss. However, individualization will be important with therapies based upon age, gender, family history, BMI, body fat distribution, insulin sensitivity and genetics. For individuals with an increased risk of breast cancer, there is no time to wait for pharmaceutical support. Despite the lack of prospective studies, current data and common sense suggest that reducing weight to within the normal range should be a goal due to the compelling data suggesting that obesity is a risk factor for a multitude of cancers, cardiovascular disease and stroke.
Executive summary
Overweight and obesity is a preventable cause of postmenopausal breast cancer incidence and mortality.
Overweight and obesity in high-risk populations is protective for breast cancer incidence in the premenopausal years. However, obesity may worsen prognosis once breast cancer is acquired.
Central adiposity and insulin resistance, termed the metabolic syndrome could, in theory, contribute to breast cancer risk in all age groups; more data are needed to answer this question.
The role of adipocytokines, their release and function related to breast cancer are currently being investigated.
Preliminary studies in breast cancer populations indicate that biochemical parameters associated with risk are diminished with weight loss. Prospective trials to examine risk reduction and to discover the most effective programs to facilitate long-term weight reductions are needed.
Obesity itself carries its own risk factors for breast cancer regardless of estrogen. One estrogen-independent mechanism for which there is both experimental and epidemiological support involves hyperinsulinemia and the insulin-resistance syndrome, which are both associated with obesity and the metabolic syndrome. Hyperinsulinemia has been correlated with both body mass index and the risk of recurrence and mortality in breast cancer; therefore, reductions in hyperinsulinemia, which is usually associated with weight loss, may decrease the risk and mortality from breast cancer.
