Early suggestions of an association between dietary fat and breast carcinogenesis, with evidence strongest in animal models (Rose 1997), led to the design of several small and two large-scale randomized controlled trials focused on the effect of a diet change intervention on intermediate or prognosis-related outcomes in breast cancer patients (Demark-Wahnefried et al. 2012; Chlebowski et al. 2006; Pierce et al. 2007). The diet change interventions have included reduced fat (typically a goal of less than 20 % of daily calories from fat), increased vegetables and fruits, increased fiber, or various combinations of these components. Durations have ranged from a few months to several years, but most have been of 6 or 12 months’ duration. Reporting of effect size varied among studies, with some showing absolute change in weight, others reporting relative change, and a smaller number providing data on other body composition variables such as waist circumference or image-derived body fat. Biological endpoints have included insulin and insulin resistance markers, inflammation-related biomarkers, sex hormones (estrogens, androgens, sex hormone binding proteins), and various adipokines (Scott et al. 2013). Few trials have examined the effect of weight loss on important quality-of-life endpoints in breast cancer survivors.

Two full-scale randomized clinical trials evaluated dietary change in the adjuvant breast cancer setting (Chlebowski et al. 2006; Pierce et al. 2007). The WINS and WHEL study enrolled different populations and studied different dietary patterns, but both aimed to reduce dietary fat intake. Neither targeted weight loss nor physical activity. In the WINS trial, while weight loss was not a specific intervention target, there was a statistically significant (P = 0.005), 6-pound lower mean body weight in the intervention group at 5 years. There were more recurrence events in the control (181 of 1,462, 12.4 %) compared to the intervention group (96 of 975, 9.8 %, hazard ratio (HR) 0.76, 95 % CI 0.60–0.98, p = 0.034). The WINS results suggest that a lifestyle intervention reducing dietary fat intake and associated with modest weight loss may improve outcome of breast cancer patients receiving conventional cancer management.

Following the observations that overweight and obesity adversely affect prognosis, a number of randomized controlled trials have tested the effect of weight loss on various health factors in women with breast cancer. None, however, have been specifically designed or sufficiently powered to test the effect of weight loss on survival. Several diet or diet plus exercise intervention trials have tested weight loss interventions on health factors other than survival. Earlier studies used individual in-person counseling to deliver guidance on caloric-restriction, while more recently, group-based or telephone support have been used (Goodwin et al. 2014).

The Lifestyle Intervention Study in Adjuvant Treatment of Early Breast Cancer (LISA) Weight Loss randomized controlled trial enrolled 338 women with early stage estrogen receptor positive breast cancer to either a telephone-based weight loss intervention or educational control group (Goodwin et al. 2014). The initial aim was to assess weight loss effect on disease-free survival but the trial was stopped due to lack of funding. Eligibility included diagnosis of Stage I-III breast cancer, BMI ≥ 24 kg/m², and treatment with letrozole. The weight loss intervention, based on the Diabetes Prevention Program lifestyle change intervention (DPP 2002a, b), focused on weight reduction through calorie restriction and increased physical activity. The weight loss intervention arm lost significantly more weight than the control arm, with mean reductions of 5.3 vs. 0.7 % at 6 months (p < 0.001) and 3.6 vs. 0.4 % at 24 months (p < 0.001).

A12-month trial with 48 obese stage I-II breast cancer patients, produced weight losses of <1 % in controls, 8.4 % with individualized counseling, and 9.8 % with individualized counseling paired with Weight Watchers^® group classes (Djuric et al. 2002). Two other group-based randomized clinical trials in breast cancer survivors, i.e., the Healthy Weight Management Study (n = 85) (Mefferd et al. 2007), and the Survivors Health And Physical Exercise (SHAPE) trial (n = 258) tested the effect of a cognitive-behavioral weight loss program plus telephone counseling vs. wait-list controls (Taylor et al. 2010). The Healthy Weight Management intervention produced an 8 % weight loss at 12 months, while the SHAPE intervention yielded a 4.5 % weight loss at 18 months. The weight loss interventions were also associated with favorable changes in self-esteem, depression and serum concentrations of sex hormone binding globulin, estradiol, bioavailable estradiol, insulin, leptin and total and LDL cholesterol.

The ongoing Exercise and Nutrition to Enhance Recovery and Good Health for You (ENERGY) Trial, is a multi-site randomized controlled trial designed to promote and sustain a 7 % weight loss over a 2-year period in 693 overweight or obese women who have been diagnosed with early stage breast cancer (Rock et al. 2013a, b). Secondary aims are to evaluate weight loss at 24 months according to time since diagnosis and type of tumor and therapy; to assess the impact of the intervention on quality of life; and to prospectively collect biological samples for future biomarker studies to help explain the mechanism and probable differential response across subgroups. The group-class weight loss intervention addresses breast cancer specific issues and promotes an energy-restricted diet, plus increased physical activity, behavioral strategies, cognitive restructuring, skills to facilitate and maintain good choices, social support, self-nurturing, and body image and self-acceptance.

As in persons without cancer (Butryn et al. 2011; DPP 2002a, b), randomized trials in breast cancer survivors indicate that optimal weight loss effects result from multicomponent behavior change interventions that target dietary calorie reduction to reach a deficit of 500–1,000 kcal/day, moderate or greater intensity physical activity for at least 150 min/week, and behavior change principles including goal setting, self-monitoring, and stimulus control. Interventions that include group behavior change sessions have produced results equal to or greater than one-on-one counseling, although optimal results provide for some individual contact with a counselor/case-worker (Befort et al. 2014).

The considerable costs of delivering in-person individual or group interventions, and the difficulties accruing participants who live at some distance from research centers, has led to several trials testing home interventions with remote contacts with case-workers. For cancer survivors, these remote contacts have primarily been via telephone. Befort et al. delivered a group behavioral weight loss intervention by conference call to obese breast cancer survivors living in remote rural locations after first recruiting them in person with the assistance of their local caregivers (Befort et al. 2012). The intervention included a reduced calorie diet incorporating prepackaged entrees and low calorie high protein shakes, advice on physical activity which was gradually increased to 225 min/week of moderate intensity exercise, and weekly group phone sessions which included education about breast cancer as well as advice on how manage life on a diet. Adherence was excellent with a loss of 13.9 % of baseline weight and significant reductions in leptin and insulin. A follow-up randomized study of usual care vs. a structured weight loss and maintenance intervention patterned on the above has completed accrual.

Irwin et al. 2015a, b recently completed a 6-month diet- and exercise-induced randomized weight loss trial in overweight and obese breast cancer survivors who had completed adjuvant treatment, entitled the Lifestyle, Exercise and Nutrition (LEAN) Study (Irwin et al. 2014). The LEAN Study randomized 100 women to one of three arms: an 11-session weight loss counseling program occurring over 6 months delivered in-person (Arm 1) vs. 11-session weight loss counseling over 6 months delivered via telephone (Arm 2) vs. usual care group where women received AICR pamphlets on healthy eating and exercise (Arm 3). The weight loss counseling was adapted from the 2010 U.S. Dietary Guidelines, the Diabetes Prevention Program, and American Cancer Society and AICR publications. They found statistically significant decreases in body weight among women randomized to in-person (−6.2 % weight loss) and telephone (−5.8 % weight loss) counseling compared to usual care, as well as significant decreases in several biomarkers related to breast cancer including C-reactive protein, insulin, and leptin levels. In addition to being of potential beneficial for breast cancer survivors, the changes seen in these biomarkers could predict reduced risk of diabetes and heart disease for those in the weight loss groups.

Most of the previously reported randomized clinical trials of weight loss used diet change interventions for weight loss, without addition of an exercise program. While diet change to reduce calories and fat has been shown to be highly efficacious in inducing relatively long-term weight loss (Foster-Schubert et al. 2012), it does so at the expense of muscle loss (Mason et al. 2013a, b). This is a significant issue for cancer survivors, who have a high prevalence of sarcopenia, among both obese and non-obese survivors, and sarcopenia has been associated with poorer prognosis (Villasenor et al. 2012). In non-cancer populations, exercise aids with weight loss maintenance (Miller et al. 2013), and somewhat with weight loss efficacy (Foster-Schubert et al. 2012). Yet, there are no controlled clinical trial data comparing effects of diet alone, exercise alone, and combined diet plus exercise interventions on health outcomes in cancer survivors. Given the findings in non-cancer populations, more recent weight loss trials in cancer survivors should include an exercise component.

Other adverse effects of weight loss through caloric reduction have been observed in populations without cancer, including reduced white blood cell and neutrophil counts (Imayama et al. 2012a, b), but this have been largely unexplored in weight loss trials in breast cancer survivors. Weight loss programs that include exercise interventions could carry risk for musculoskeletal injuries or cardiovascular events (Campbell et al. 2012a, b; Dahabreh and Paulus 2011). These, too, have not been part of outcomes reporting for most weight loss trials in breast cancer survivors.

Ideally, resources would be available to determine effects of various weight loss, diet, and physical activity interventions on breast cancer prognosis. An alternative would be to launch a full-scale randomized controlled trial testing the effect of a weight loss intervention on overall and breast-cancer-specific survival with an intervention known to maximize weight loss while having high acceptability to survivors and a favorable risk profile. The same trial could also test mediating biomarkers that could then be used as endpoints in future trials of other weight loss interventions. Effects on health and quality of life factors relevant to breast cancer survivors should be assessed, including lymphedema, bone density, diabetes, cardiovascular disease, arthralgias, cognitive function, fatigue, anxiety, depression, and adverse effects should be enumerated. Such a trial has been proposed, although resources have not been available (Ballard-Barbash et al. 2009). While these definitive trials of weight loss and exercise on disease-free survival are critical for moving the field forward, dissemination and implementation of evidence-based lifestyle interventions needs to occur. Research is necessary on how to disseminate lifestyle interventions into the clinic and community that lead to clinically meaningful weight losses of at least 5 % that are maintained over time. Whether these interventions are more effective when implemented in cancer hospital survivorship clinics/centers or when implemented via referrals to community-based programs needs to be examined. In summary, a growing number of observational studies have consistently shown obesity to be associated with a higher risk of breast cancer and all-cause mortality, yet no randomized controlled trial of weight loss on disease-free survival has been conducted (see Table 13.1 for future research needs). A growing number of randomized weight loss trials on biomarkers or quality of life have been conducted, yet it is unknown if these findings will lead to implementation and reimbursement of weight management programs in the clinic or community. If so, then we could expect to see decreases in obesity in breast cancer survivors, as well as prevention of weight gain in women newly diagnosed with breast cancer.

One of the most plausible mechanisms of how exercise may reduce breast cancer risk, recurrence and mortality is by lowering estrogen concentrations through reduction in body fat and decreased estrogen production from aromatization of androgens. Two randomized controlled exercise trials, conducted in healthy women have shown an increase in sex hormone binding globulin and an ~10 % decrease in bioavailable estrogen and testosterone primarily in those women who lost body fat (McTiernan et al. 2004; Friedenreich et al. 2010). In a 4-arm randomized controlled trial, a far greater reduction in serum estradiol was observed with weight reduction through caloric restriction, with or without exercise, compared with controls or with an exercise-only intervention (Campbell et al. 2012a, b). The greater effect of dietary weight loss on serum estrogens compared with exercise alone is not surprising, since caloric reduction of about 500–1,000 kcal/day typically produces 10 % weight loss over 6–12 months, while exercise alone produces 1–2 % loss (DHHS 1996). The effect of weight loss on blood estrogens in women with breast cancer has been little studied, likely because of the potential for confounding effects of some treatments such as aromatase inhibitors and tamoxifen. One study in 220 survivors enrolled in a weight loss intervention found that postmenopausal women who lost ≥5 % of body weight at 6 months had lower estrone (P = 0.02), estradiol (P = 0.002), and bioavailable estradiol (P = 0.001) concentrations than women who did not lose at least 5 % of body weight (Rock et al. 2013a, b).

Elevated insulin levels have been linked to an increased risk of breast cancer, and several reports have demonstrated that women with higher levels of insulin at the time of breast cancer diagnosis are at increased risk of cancer recurrence and death (Duggan et al. 2011; Irwin et al. 2011; Goodwin et al. 2002). These findings showed that a lowering of insulin levels by 25 % may be associated with a 5 % absolute improvement in breast cancer mortality, and this strong association between fasting insulin levels and breast cancer mortality has led a number of oncologists and scientists to consider the targeting of insulin as a therapeutic modality in breast cancer.

A number of exercise and weight loss interventions have been shown to impact insulin in healthy women. One recent trial, conducted by Dr. McTiernan, randomized 439 overweight/obese, sedentary postmenopausal women to one of three energy balance interventions (dietary weight loss alone, exercise alone or dietary weight loss plus exercise) or to control and demonstrated that the weight loss groups experienced the most significant changes in insulin (−22.3 % in the dietary weight loss alone and −24 % in the combined diet and exercise group vs. −7.8 % in the exercise alone group and −1.9 % in the control group) (Mason et al. 2011).

There are fewer data regarding the impact of energy balance interventions upon insulin in breast cancer survivors. One study looked at the impact of three different dietary weight loss interventions (Weight Watchers, an individualized weight loss program or a combination of the two) vs. control on fasting insulin in 48 breast cancer survivors and demonstrated an average 12 % reduction in insulin levels in the three dietary intervention groups. Another study looked at the impact of a diet and exercise weight loss program on insulin levels in 35 rural breast cancer survivors and demonstrated a 17 % reduction in fasting insulin levels. Finally a few small studies have looked at the impact of exercise-only interventions upon insulin levels in breast cancer survivors. One exercise study demonstrated a 28 % reduction in fasting insulin levels in 101 inactive, overweight breast cancer survivors participating in a mixed strength and aerobic exercise intervention (p = 0.07) (Ligibel et al. 2008). The other exercise study looked at the impact of a moderate-intensity aerobic exercise intervention in 68 sedentary, overweight breast cancer survivors, and demonstrated an 8 % decrease in insulin levels in exercisers and a 20 % between group difference (p = 0.089) (Irwin et al. 2009a, b). Thus there is preliminary evidence in healthy populations that weight loss may be the most important factor in reducing insulin, but data are limited in breast cancer survivors. Metformin reduces insulin levels by 22 % in non-diabetic breast cancer survivors (Palmirotta et al. 2009), and a randomized trial of metformin vs. placebo is being tested in the adjuvant setting (clinicaltrials.gov NCT01101438), as well as trials of metformin alone or with exercise or with weight loss are being tested in the NCI-funded Transdisciplinary Research on Energetics and Cancer studies (clinicaltrials.gov NCT01340300 and NCT01302379). These findings will move the field forward in regards to the role of lifestyle factors compared to medication upon lowering insulin levels in breast cancer survivors.

Exercise training seems to lower both resting and post exercise inflammatory cytokine levels through reduction of circulating monocyte as well as tissue macrophage production and release (Kasapis and Thompson 2005). Preclinical studies suggest that exercise can have a profound effect on macrophage infiltration into adipose and muscle tissue with reduction in M1 macrophage concentration associated with cytokine production and chronic inflammation particularly in diet induced obesity (Kawanishi et al. 2010). Most moderate volume and intensity exercise intervention studies in the general population have found no significant change in inflammatory biomarkers (Marcell et al. 2005; Hammett et al. 2006; Arsenault et al. 2009). Those studies in which inflammatory markers particularly TNF-α, IL6, and/or CRP were favorably modulated with exercise tended to be those in which: (a) individuals were obese at baseline and thus had higher baseline levels of inflammatory cytokines (Kasapis and Thompson 2005; Christiansen et al. 2009; Arikawa et al. 2011; Phillips et al. 2012); (b) exercise volume and intensity were high enough to result in loss of weight and/or body fat (Christiansen et al. 2009); and/or (c) where cytokine production (TNF-α or IL6) was stimulated with lipopolysaccharide exposure (Phillips et al. 2012). Loss of 5–10 % of baseline weight through caloric reduction with or without an exercise program has been shown to reduce inflammation-related biomarkers such as CRP and IL-6 by 20–40 % (Imayama et al. 2012a, b). These effects far exceed those seen with exercise interventions in the absence of significant weight loss. A systematic review concluded that across lifestyle and surgical weight loss interventions, for each 1 kg of weight loss, the mean change in CRP level was −0.13 mg/L (with a weighted Pearson correlation of r = 0.85) (Selvin et al. 2007). Although future research in this area is definitely warranted, investigating more sensitive circulating as well as breast and adipose tissue based immune parameters is warranted.