Maturitas
Volume 66, Issue 1 , Pages 5-15, May 2010

Physical activity, diet, adiposity and female breast cancer prognosis: A review of the epidemiologic literature

Cancer Prevention and Control Program, Moores UCSD Cancer Center, University of California, 3855 Health Sciences Drive #0901, San Diego, La Jolla, CA 92093-0901, United States

Received 23 December 2009; received in revised form 6 January 2010; accepted 6 January 2010. published online 15 January 2010.

Article Outline

Abstract 

Given the increasing numbers of long-term survivors of breast cancer, research specific to prevention of recurrence, new breast cancer events, and mortality is of considerable public health importance. The objective of this report is to present a review of the published epidemiologic research on lifestyle and breast cancer outcomes among women with a history of breast cancer. This review focused on physical activity, diet, and adiposity; and the primary outcomes were additional breast cancer events and mortality. The most consistent finding from observational studies was that adiposity was associated with a 30% increased risk of mortality. Although the observational data were not as consistent (or abundant), physical activity appeared to be associated with a 30% decreased risk of mortality. These data do not indicate that alcoholic drinks are a risk factor. Based only on the observational studies, total dietary fat appeared to be a risk factor, fiber was protective, and information on micronutrients and specific foods was sparse. However, the null results of 2 dietary intervention trials in survivors suggests that lowering fat intake or increasing consumption of fruits, vegetables, and fiber will not lead to improved prognosis in breast cancer survivors. Given that a high proportion of breast cancer patients appear to be both sedentary and obese/overweight, clinical trials are needed to investigate whether the combination of increased physical activity and reduced adiposity can improve breast cancer prognosis.

Keywords: Breast cancer, Diet, Physical activity, Adiposity

 

Back to Article Outline

1. Introduction 

Advances in diagnosis and treatment have rendered breast cancer a chronic disease in countries with modern health care. For example, the American Cancer Society estimates that 89% of US women diagnosed with invasive breast cancer in 2009–2010 will be alive 5 years after diagnosis and 82% will be alive 10 years after diagnosis [1]. The National Cancer Institute estimates that approximately 2.5 million women with a history of breast cancer were alive in January 2006. However, little is known about lifestyle factors that modify breast cancer prognosis. Given the increasing numbers of long-term survivors, research specific to prevention of recurrence, new breast cancer events, and mortality is of considerable public health importance.

There are numerous reviews and meta-analyses addressing lifestyle and incident cancer. The most comprehensive and detailed synthesis of the scientific evidence regarding the extent to which lifestyle could modify the risk of cancer was published in 2007: Report on Food, Nutrition, Physical Activity and the Prevention of Cancer: a global perspective [2]. For this report, an expert panel was convened to assess and judge the evidence. The panel produced a set of public health goals and recommendations as follows:

1.Be as lean as possible within the normal range of body weight.

2.Be physically active as part of everyday life.

3.Limit consumption of energy-dense foods and avoid sugary drinks.

4.Eat mostly foods of plant origin.

5.Limit intake of red meat and avoid processed meat.

6.Limit alcoholic drinks.

7.Limit consumption of salt and avoid mouldy cereals (grains) or pulses (legumes).

8.Aim to meet nutritional needs through diet alone.

The expert panel also concluded that the evidence on survivorship was not clear enough to make firm judgments. Therefore, a special recommendation was made for cancer survivors indicating that they should follow the recommendations for cancer prevention.

In addition to the general recommendations given above for all cancers, this 2007 report reviewed the evidence regarding lifestyle and breast cancer. The panel considered the evidence for pre-menopausal cancer separately from post-menopausal cancer. For pre-menopausal breast cancer, the panel concluded the evidence was convincing that alcoholic drinks increased risk, probable that body fatness increased risk, and suggestive that low physical activity increased the risk of cancer. For post-menopausal cancer, the panel concluded that the evidence was convincing that alcoholic drinks and fatness increased risk; probable that low physical activity, abdominal fatness, and adult weight gain increased risk; and suggestive that total dietary fat increased risk of cancer. Overall, evidence regarding dietary factors (with the exception of total fat) was deemed limited.

There is considerable research on mechanisms for how lifestyle factors could influence breast cancer, although it is generally not specific to survivorship. Breast cancer is hormone related and the association between endogenous sex hormone concentrations and breast cancer risk in post-menopausal women is well-established. In particular, a recent meta-analysis reported that estradiol was associated with a 2- to 3-fold increased risk of breast cancer (p<0.01) [3]. Chronic hyperinsulinemia is associated with increased availability of IGF-1 and concomitant changes in the cellular environment that favor tumor formation. Epidemiologic studies have shown a consistent, positive association of fasting insulin concentrations or type 2 diabetes mellitus with incident [4], [5] and recurrent breast cancer [6], [7]. Finally, inflammation in the tumor microenvironment appears to aid in the proliferation and survival of malignant cells, promote angiogenesis and metastasis, subvert adaptive immune responses, and alter responses to hormones and chemotherapeutic agents [8]. Dozens of circulating markers of inflammation have been identified, including adiponectin, IL-6, and TNF-α. High-sensitivity C-reactive protein (CRP) is a general marker of chronic inflammation at the vascular level that has been investigated extensively in relation to cardiovascular disease [9], [10], [11], [12] and more recently associated with increased cancer risk [13]. Notably, one recent study found that elevated CRP was associated with reduced survival in breast cancer patients [14].

Although the data are conflicting, research indicates that lifestyle can influence the mechanisms summarized above. In particular, obesity has been consistently correlated with higher concentrations of estradiol, fasting insulin, and inflammatory markers [15]. Randomized trials have shown that increased physical activity and diets low in fat and high in vegetables and fruit reduce circulating levels of estradiol and estrone [16], [17]. Physical activity, hypocaloric diets, and possibly low-glycemic diets may reduce fasting insulin levels [18]. Physical activity and dietary patterns characterized as Mediterranean, anti-inflammatory, or having a low glycemic load have been associated with reductions in CRP [19], [20]. These associations of obesity, physical activity, and diet with breast cancer markers suggest that lifestyle has the potential to modify breast cancer risk.

The objective of this report is to present a review of the epidemiologic research on physical activity, diet, and adiposity with breast cancer clinical outcomes among women with a history of breast cancer.

Back to Article Outline

2. Methods 

We conducted a bibliographic database search for randomized controlled trials, cohort studies, or case-control studies of breast cancer outcomes that contained the words “breast cancer” plus “recurrence,” “mortality” or “survival” plus keywords related to the exposures of interest (e.g., “alcohol,” “diet,” “obesity”). We searched the bibliographies of publications for other relevant studies. We included all studies published in the prior 10 years (1999–2009) as well as a few large studies published more than 10 years ago. Papers were included only if there were at least 500 participants. We did not include studies of exposure biomarkers such as serum nutrient concentrations. Finally, we did not review studies of intermediate markers of breast cancer (e.g., mammographic density) or noninvasive lesions (e.g., ductal carcinoma in situ).

As indicated above, this review focuses on physical activity, diet, and adiposity. For brevity, we refer to these risk factors as lifestyle throughout this review. The primary outcomes are additional breast cancer events and mortality. An additional breast cancer event is defined as a recurrence from the original cancer or developing a new invasive breast cancer. Some studies only considered recurrence and we have included them with a footnote. Overall survival is the time to death from all causes (i.e., mortality). Some studies only considered breast cancer mortality, and we have included them with a footnote.

Details regarding the studies included in this review (e.g., sample size, follow-up) are presented in tabular format. We present the adjusted hazard ratio (HR) or relative risk (RR) and 95% confidence intervals (CI) for highest versus lowest level of exposure in forest plots. In general, we present one major finding per study in each plot to avoid apparent weighting of evidence due to multiple, related measures of an exposure. For example, many studies present hazard ratios for adiposity assessed as waist circumference, waist/hip ratio, and body mass index. If the authors did not identify the primary exposure measure, we selected the one deemed most relevant to this review. When breast cancer mortality and all-cause mortality were presented, preference was given to all-cause mortality. Only results adjusted for other covariates are presented. In addition, specific inclusion criteria were applied for each exposure as follows:

•Physical activity: Because the majority of studies measured only leisure-time (i.e., sports/recreational) activity, we focused on this measure to enhance comparability of the findings.

•Macronutrients and dietary patterns: When macronutrients were presented in absolute amounts (e.g., grams fat per day) and relative amounts (e.g., percent energy from fat), preference was given to the relative amount. We focused on total fat and did not include fat subtypes (e.g., saturated fat) in the plots.

•Micronutrients and fruits and vegetables: We limited this exposure to those vitamins and minerals that can reasonably be obtained by self-report: vitamins A, C, E, niacin, folate, calcium, iron, and zinc. When nutrients were presented with and without supplements, preference was given to the measure inclusive of supplements.

•Adiposity: Data regarding absolute weight and weight gain are presented separately given that they may represent different types of risk. When multiple measures of adiposity were presented, preference was given to body mass index (BMI). We present data on pre-diagnosis weight or BMI only if no other measure of adiposity was reported.

For the lifestyle exposures included in this review where 3 or more studies reported an association, we present a weighted hazard ratio (HR) by outcome in the text. Specifically, hazard ratios were averaged over exposures and weighted with the inverse of each study's within-study variance [21]. This method allowed for an average point estimate that considers each study's standard error. As with any meta-analysis, inference is limited due to factors such as varying sample characteristics and different follow-up times. Therefore, to avoid drawing any potentially erroneous conclusions, averaged hazard ratios are not presented with confidence intervals. It is our intention that these averaged estimates will serve as a general combined estimate of effect—they do not represent the results of a strict meta-analysis. Nonetheless, we believe that these weighted estimates provide further insight into exposures that are difficult to standardize across studies.

The experimental design of a randomized controlled trial (RCT) provides a high degree of assurance about the validity of the results, far superior to that possible with any observational study. Therefore, we present and discuss the results of RCTs separately from observational studies.

Back to Article Outline

3. Results 

3.1. Observational studies of physical activity, diet and obesity and breast cancer outcomes 

Table 1 gives details regarding the observational studies of lifestyle and breast cancer outcomes. Specifically, we present the reference, sample size, number and type of events, years of follow-up, menopausal status of the participants, and the specific exposures. Information from this table supplements the summary data (HR and CI) presented in the forest plots.

Table 1. Studies of physical activity, diet, and adiposity on additional breast cancer events and mortality among women with a history of breast cancer.
Author (year) [ref]NEventsYears (f/up)Menopausal statusExposure(s)
Physical activity
Friedenrich (2009) [22]1,231327 new BCa events
226 BC deaths		8BothLifetime physical activity
Holmes (2005) [25]2,987280 BC deaths
370 recurrences		8BothPost-diagnosis physical activity
Pierce (2007) [24]1,490252 additional events
149 all-cause deaths		6.7BothPost-diagnosis physical activity
Sternfeld (2009) [23]1,970107 BC deaths
225 recurrences
187 all-cause death		7.3BothPost-diagnosis physical activity
Borugian (2004) [26]603112 BC deaths10Both, stratifiedPhysical activity at diagnosis
Irwin (2008) [29]933115 BC deaths
56 recurrences
40 new primaries		6BothPre-diagnosis physical activity
Post-diagnosis physical activity
Holick (2008) [28]4,482109 BC deaths
412 all-cause deaths		5.5BothPost-diagnosis physical activity
Dal Maso (2008) [30]1,453398 BC deaths
503 all-cause deaths		12.6BothPre-diagnosis physical activity
Enger (2004) [27]717251 BC deaths10.4Pre-menopausalPre-diagnosis physical activity

Macronutrients, food groups, and dietary patterns
Flatt (2010) [31]3,088518 additional events
315 all-cause deaths		7.3BothAlcohol
Holmes (1999) [36]1,982326 BC deaths
378 all-cause deaths		18BothAlcohol, carbohydrates, fiber, protein
Borugian (2004) [26]603112 BC deaths10Both, stratifiedAlcohol, carbohydrates, fat, fiber, protein
Barnett (2008) [34]4,560564 all-cause deaths6.8BothAlcohol
Dal Maso (2008) [30]1,453398 BC deaths
503 all-cause deaths		12.6BothAlcohol, fat, protein
Reding (2008) [40]1,286BC deaths
364 all-cause deaths		10+Pre-menopausalAlcohol
Franceschi (2009) [35]1,453503 all-cause deaths12.6BothAlcohol
McEligot (2006) [39]51641 BC deaths
96 all-cause deaths		6.7Post-menopausalCarbohydrates, fat, fiber, protein
Jain (1994) [37]67876 BC deaths
83 all-cause deaths		7.7BothFat
Zhang (1995) [41]69856 all-cause deaths2.9Post-menopausalFat
Kroenke (2005) [38]2,619242 BC deaths
414 all-cause deaths		9BothPrudent diet
Western diet

Micronutrients, fruits, and vegetables
Jain (1994) [37]67876 BC deaths
83 all-cause deaths		7.7BothCalcium, vitamin A, vitamin C, vitamin E
Holmes (1999) [36]1,982326 BC deaths
378 all-cause deaths		18BothCalcium, folate, iron, niacin, vitamin A, vitamin C, vitamin E, zinc, fruits, vegetables
McEligot (2006) [39]51641 BC deaths
96 all-cause deaths		6.7Post-menopausalCalcium, folate, vitamin A, vitamin C, vitamin E, zinc, fruits, vegetables
Dal Maso (2008) [30]1,453398 BC deaths
503 all-cause deaths		12.6BothFruits and vegetables (combined)

Adiposity
Loi (2005) [46]1,360264 distant recurrences5.0BothPre-diagnosis BMIb
Berclaz (2004) [43]6,792Relapses+any death (n: N/A)
All-cause deaths (n: N/A)		14Both, stratifiedPost-diagnosis BMI
Caan (2008) [44]1,692207 BC recurrences7.0BothPre-diagnosis BMI
99 BC deaths Post-diagnosis BMI
162 all-cause deaths Pre- and post-diagnosis weight gain
Majed (2008) [45]14,7093780 metastases
555 2nd cancers
4876 DFS events
3693 all-cause deaths		8.0BothBMI at diagnosis
Kroenke (2005) [42]5,204681 recurrence (not BC specific)9Both, stratifiedPre-diagnosis weight
533 BC deaths Pre- and post-diagnosis weight gain
860 all-cause deaths
Zhang (1995) [41]69856 all-cause deaths2.9Post-menopausalPre-diagnosis BMI
Daling (2001) [47]1,177283 BC deaths
317 All-cause deaths		8+Pre-menopausalPre-diagnosis BMI
Enger (2004) [27]717251 BC deaths10.4Pre-menopausalPre-diagnosis BMI
Dal Maso (2008) [30]1,453398 BC deaths12.6BothPost-diagnosis BMI
503 all-cause deaths BMI Δ from age 30 to diagnosis
Dignam (2003) [51]4,0771326 additional events (772 BC recurrences 439 2nd primaries 115 prior deaths)VariesBothPost-diagnosis BMI
Abrahamson (2006) [49]1,254290 all-cause deaths9.8BothPre-diagnosis BMI, waist-to-hip ratio
Post-diagnosis BMI, waist-to-hip ratio
Nichols (2009) [53]3,993121 BC deaths6.3BothPre-diagnosis BMI
419 all-cause deaths Post-diagnosis BMI
Post-diagnosis weight gain
Olsson (2009) [54]2,974142 BC deaths10.0BothPost-diagnosis BMI
Whiteman (2005) [55]3,9241347 BC deaths14.6Both, stratifiedAge 18 BMI
1671 all-cause deaths Adult BMI
Cleveland (2007) [50]1,508127 BC deaths
196 all-cause deaths		5.6BothPre-diagnosis weight gain

Diabetes
Lipscombe (2008) [52]6,1071426 all-cause deaths4.9Post-menopausalDiabetes
Patterson (2010) [48]2,542406 additional events
232 all-cause deaths		7.3BothDiabetes

aBreast cancer.

bBody mass index.

3.2. Observational studies of physical activity 

We identified four studies of physical activity and additional breast cancer events [22], [23], [24], [25] (Fig. 1). None of these studies found a statistically significant association, although there was a trend toward a protective effect: the weighted average was 0.86.

  • View full-size image.
  • Fig. 1. 

    Physical activity exposure on breast cancer outcome, by study. Outcome defined as additional events (recurrence or new primary) and mortality (all cause or death due to breast cancer). Studies sorted by exposure and year of study. 1Includes recurrence and metastasis (does not include new primary events). 2Breast cancer mortality.

We identified 9 studies of physical activity and breast cancer mortality [22], [23], [24], [25], [26], [27], [28], [29], [30], of which 3 showed a statistically significant protective effect [25], [28], [29] (Fig. 1). The weighted average for the association of lifetime, at diagnosis, or post-diagnosis physical activity with mortality was 0.71.

3.3. Observational studies of macronutrients and dietary patterns and breast cancer outcomes 

We identified only 2 studies with additional events as an outcome, one on alcohol [31] and one comparing prudent versus western dietary pattern [32] (Fig. 2). For these exposures, the HRs were approximately 1.0. Although not included in the figures, a recent abstract [33] reported that alcohol consumption was a risk factor for recurrence (HR 1.42, 95% CI, 1.05 to 1.93).

  • View full-size image.
  • Fig. 2. 

    Macronutrient exposure on breast cancer outcome, by study. Outcome defined as additional events (recurrence or new primary) and mortality (all cause or death due to breast cancer). Studies sorted by exposure and year of study. 1Relative risk. 2Breast cancer mortality.

Overall, there were 12 studies of macronutrients and dietary patterns with mortality as an outcome [26], [30], [31], [32], [34], [35], [36], [37], [38], [39], [40], [41] (Fig. 2). For the 7 studies on alcohol [26], [30], [31], [34], [35], [36], [40], findings were mixed although 3 studies reported a statistically significant protective effect [31], [34], [40]: the weighted average was 0.92. While not shown on the figure, a recent abstract reported that alcohol was a nonsignificant risk factor for mortality (HR 1.25, 95% CI, 0.89 to 1.75) [33]. For the 3 studies on carbohydrates [26], [36], [39], findings were mixed with 1 report of a statistically significant protective effect [39]: the weighted average was 0.74. For the 5 studies on dietary fat [26], [30], [37], [39], [41], 4 HRs showed a trend toward increased risk although only one had a statistically significant finding [39]: the weighted HR was 1.53. The 3 studies of dietary fiber showed a trend toward being protective [26], [36], [39] and 2 were statistically significant [36], [39]: the weighted average was 0.63. Of the 4 studies that reported on protein [26], [30], [36], [39], 2 found a statistically significantly protective association [26], [36]: the weighted average was 0.72. A prudent dietary pattern (characterized by high intakes of fruits, vegetables, whole grains, legumes, poultry and fish) appeared protective [32], [38], with one statistically significant finding [32]. Conversely, the Western dietary pattern (characterized by high intakes of refined grains, processed and red meats, desserts, high-fat dairy products, and French fries) was associated with a nonsignificant increase in risk [32], [38].

3.4. Observational studies of micronutrients, fruits, and vegetables with breast cancer outcomes 

We did not identify any observational studies that reported on micronutrient, fruit or vegetable intake with additional breast cancer events.

There were only 4 studies of micronutrients and/or fruits and vegetables with mortality [30], [36], [37], [39], although these publications typically included many exposures (Fig. 3). For example, Holmes et al. reported on 93 dietary exposures in relation to all-cause mortality (of which 17 were statistically significant) [36]. Therefore, readers interested in all possible dietary exposures should refer to the original papers. Overall, the HRs in Fig. 3 showed a trend toward being protective. Among the vitamins, minerals, fruits and vegetables in this review; statistically significant protective associations were seen for higher intakes of calcium [36], vitamin C [37], and vegetable consumption [39]. Given the limited number of total studies and the multiplicity of dietary exposures reported in each study, we did not consider it appropriate to calculate a weighted average for the exposures presented in this review.

  • View full-size image.
  • Fig. 3. 

    Micronutrient exposure on breast cancer outcome, by study. Outcome defined as additional events (recurrence or new primary) and mortality (all cause or death due to breast cancer). Studies sorted by exposure and year of study. 1Breast cancer mortality. 2Relative risk.

3.5. Observational studies of adiposity and breast cancer outcomes 

We identified 5 studies on adiposity and additional breast cancer events [42], [43], [44], [45], [46], [47] (Fig. 4). One paper reported a statistically significant association of pre-diagnosis BMI with additional events [46]. Among the 3 studies of post-diagnosis BMI, 2 showed a modest (∼10%) increased risk that was statistically significant [43], [45] and one was null [44]; the weighted HR was 1.09. Otherwise, findings were mixed for 2 papers on pre- to post-diagnosis weight gain [42], [44]. We included one study of type 2 diabetes mellitus because of the strong association of obesity and diabetes. This study found a statistically significant, 2-fold increase in the risk of additional events with type 2 diabetes mellitus [48].

  • View full-size image.
  • Fig. 4. 

    Adiposity on breast cancer outcome, by study. Outcome defined as additional events (recurrence or new primary) and mortality (all cause or death due to breast cancer). Studies sorted by exposure and year of study. BMI: body mass index (kg/m2). 1Includes recurrence and metastasis (does not include new primary events). 2Never smokers. 3Breast cancer mortality. 4Included due to strong association of obesity and development of type 2 diabetes mellitus.

There were 17 studies on adiposity and mortality [27], [30], [41], [42], [43], [44], [45], [46], [47], [48], [49], [50], [51], [52], [53], [54], [55] (Fig. 2). Among the 7 studies that reported on post-diagnosis BMI [43], [44], [45], [49], [51], [53], [54], all showed increased risk and 5 had a statistically significant finding [43], [45], [49], [51], [54]: the weighted HR was 1.27. Four [30], [41], [46], [47] of 5 studies [27], [30], [41], [46], [47] found increased risk with pre-diagnosis BMI and one finding was statistically significant [46]: the weighted HR was 1.25. Findings on weight gain were mixed [42], [50], [53], [56], with a study showing a statistically significant increase in risk with pre- to post-diagnosis weight gain [42] and with post-diagnosis weight gain [53]. There was one study of “usual adult BMI” that found a statistically significant risk with mortality [55]. Both studies of type 2 diabetes mellitus reported a statistically significant increase in the risk of mortality [48], [52]. While not included on the forest plot, we identified one study that reported that waist/hip ratio was associated with breast cancer deaths (HR 3.3, 95% CI, 1.1 to 10.4 for post-menopausal; HR 1.2, 95% CI, 0.4 to 3.4 for pre-menopausal cancer) [57].

3.6. Randomized trials of role of dietary change in breast cancer outcomes 

The early suggestive data from the observational studies led to two large randomized trials in the United States in the mid-1990s. Both addressed prognosis in early stage breast cancer survivors who had completed treatment. The Women's Intervention Nutrition Study (WINS) focused on the efficacy of a major reduction in fat intake whereas the Women's Healthy Eating and Living (WHEL) study focused on the efficacy of a major change in dietary pattern (including a reduction in fat).

Between 1994 and 2001, WINS1 enrolled 2437 post-menopausal women who had completed primary treatment for early stage breast cancer (55% AJCC IV stage I) and had been diagnosed within the previous year from 39 centers across the USA. One third (n=975) were randomized to an intensive intervention that targeted a reduction in energy from fat from 29.6% at baseline to 15% [58]. Between 1995 and 2000, 7 centers (4 in California) in the WHEL trial enrolled 3088 women (2448 were post-menopausal) who had completed treatment for early stage breast cancer (40% AJCC IV stage I) and had been diagnosed within the previous 4 years (average 24 months). One half was randomized to an intensive intervention that targeted a dietary pattern with very high vegetables, fruit and fiber and low fat [59].

The differences in the study populations were: (a) the cancer characteristics profile indicated a better prognosis for WINS patients, (b) the WHEL study allowed a 4-year enrollment window and hence cannot address early recurrence and death, and (c) WINS could not address pre-menopausal breast cancer. However, when comparing just the post-menopausal sample of WHEL with WINS, there was a remarkable similarity between the two study populations on age, race, education, BMI and dietary energy from fat [60].

Both study interventions were associated with major dietary changes at 12 months as assessed by multiple 24-h dietary recalls. At 12 months, the WHEL intervention reduced energy from fat to 22.7% while increasing vegetables/fruits to 12 daily servings and increasing and fiber to 29g/day [61]. In a completers analysis (the only analysis presented), the WINS intervention reduced energy from fat to 20.3% with fiber relatively unchanged at 19.5g/day [58].

Its important to note that over time, WINS had much lower completion rates than the WHEL study. For example, the WINS completion rate at 5 years was 39% for the intervention and 44% for the CONTROL group; while the WHEL study completion rate at 4 years was 83% for the intervention and 89% for the control group. Using an intent to treat analysis, WHEL found that the between group difference was approximately −5 to −4 percentage points for energy from fat at 3 and 4 years, respectively [61]. WINS reported a between group differential of −8 percentage points for energy from fat at 3 and 5 years. However, assuming that the approximately 60% of WINS intervention participants lost to follow-up were not adhering to the dietary intervention, this effect size would be closer to −4 percentage points in an intent-to-treat analysis [60]. This comparison suggests that the level of fat reduction achieved in these 2 trials was similar. However, the WHEL trial also targeted other dietary behaviors and reported between group differences at 4 years as follows: +65% for vegetables; +25% for fruits; +30% for fiber; which was confirmed through plasma carotenoid analysis [62].

Both studies solicited health status on a regular basis and confirmed self-reports with medical record review. Restricting the WHEL sample to post-menopausal women allows a comparison in event rates between the two studies [60]. Given the difference in tumor characteristics at diagnosis, the distal recurrence rate was, as expected, 45% lower in WINS compared to WHEL. However, the between group (control–intervention) differences in either study were similar for both regional and distal recurrences (Table 2). There were differences between the studies in the local and new primaries data. In particular, the WINS comparison group had a higher rate of local and new primaries than the WHEL comparison group (5.2% vs. 3.7%). In WINS, the intervention group had a 25% lower rate of local and new primaries, whereas in WHEL, it was 13% higher (+1.3% vs. −0.5%). Thus, there were similar rates for the intervention groups in each study (3.9% vs. 4.2%).

Table 2. Comparison of WINSa and WHELb event rates for post-menopausal women with a history of breast cancer.
WINSWHEL
ComparisonC-Ic differenceComparisonC-I difference
Local or new primary5.2%1.3%3.7%−0.5%
Regional recurrence0.8%0.2%0.7%0.1%
Distal recurrence6.4%1.1%11.6%0.9%
Death without breast cancer1.3%−0.2%1.7%−0.3%

aWomen's Intervention Nutrition Study.

bWomen's Healthy Eating and Living.

cControl minus intervention.

Neither study reported an intervention effect on overall survival. For WINS, the HR was 0.89 (95% CI, 0.65 to 1.21; p=0.56). For WHEL the HR was 0.97 (95% CI, 0.78 to 1.22; p=0.82). Further, the primary analysis for each study did not achieve statistical significance for disease-free survival (WINS, HR=0.76, p=.077; WHEL HR=0.99, p=0.87). However, both studies reported a significant effect in a subgroup that was not hypothesized at the time of the protocol development. In WINS, a Cox model analysis demonstrated that in the ER-/PR-subgroup (15% of sample), the intervention had a HR of 0.44 (95% CI, 0.25 to 0.77). However, in this ER-/PR-subgroup in the WHEL study (10% of sample), the intervention group had a HR of 1.14 (95% CI, 0.80 to 1.61). The lack of concordance in this subgroup data suggests that the effect may be an artifact. The WHEL study reported that there was a significant intervention effect in the group of women who did not report hot flashes (suggestive of higher circulating estrogen concentrations) with a HR of 0.69 (95% CI, 0.51 to 0.93; p=0.02) [63]. Hot flash status data were not reported for the WINS study. However, a subsequent analysis of WHEL data revealed that in this subgroup, participants with the best quality diet at baseline (highest in fruit, vegetable, fiber and lowest in fat) experienced the greatest intervention effect [64]. This unexpected finding suggests that the intervention effect seen in the hot flash subgroup might also be an artifact.

Back to Article Outline

4. Discussion 

The most consistent finding from the observational studies was the increased risk of adiposity or body fatness in relation to mortality. We found a nearly 30% increased risk for both post- and pre-diagnosis adiposity. Although the observational data were not as consistent (or abundant), physical activity appears to be associated with a 30% decreased risk of mortality. Data from 7 cohort studies did not suggest that alcoholic drinks are a risk factor.

Based on observational studies, total dietary fat appeared to be a risk factor, fiber appeared protective, and information on micronutrients, foods, and dietary patterns was sparse. A major limitation of observational studies concerns the problem of systematic and random error in self-reported exposures. In particular, it is notable that the strength and consistency of the associations seen here correlate with the degree of measurement error. That is, the strongest associations are seen for BMI, which has low measurement error; and the weakest associations are seen for diet, which has considerable measurement error [65], [66]. The dietary data from observational studies need to be weighed against the results of the WHEL and WINS randomized trials that had null findings for a low-fat dietary intervention (WINS, WHEL) and a high fruit, vegetable, and fiber intervention (WHEL). It is possible that the degree of dietary change achieved in these interventions was not sufficient to change cancer outcomes, the wrong dietary components were targeted, or only certain subgroups of women or cancer types benefit from a dietary intervention. Nonetheless, data from these trials indicate that in a general population of breast cancer survivors, dietary interventions without weight loss or physical activity are not sufficient to improve breast cancer prognosis.

In Table 3, we present the conclusions from the Report on Food, Nutrition, Physical Activity and the Prevention of Cancer on incident cancer [2] compared to the conclusions drawn from this review of epidemiologic studies of breast cancer prognosis. Both reviews concluded that the evidence is convincing that body fatness (i.e., adiposity) increases the risk of breast cancer. Between studies of incident cancer and breast cancer outcomes, there is a discrepancy in the findings regarding alcohol consumption that warrants additional research. Given the overall null effects of 2 randomized trials conducted in survivors, we conclude that the current data do not support the hypothesis that total dietary fat (or other dietary constituents) can modify the incidence of additional breast cancer events or mortality.

Table 3. Lifestyle factors associated with increased risk of post-menopausala breast cancer: conclusions from the Report on Food, Nutrition, Physical Activity and the Prevention of Cancer on incident cancer (WCRF/AICR 2007) compared to research on breast cancer outcomes.
Strength of evidenceIncident cancerBreast cancer prognosis
ConvincingBody fatnessBody fatness
Alcoholic drinks

ProbablyLow physical activityLow physical activity
Abdominal fatness
Adult weight gain

Limited-suggestiveTotal fat

Limited-no conclusionDiet (excepting total fat)Alcoholic drinks
Adult weight gain
Abdominal fatness

No effect Total fat
Low fiber
Fruits and vegetables

aThere are insufficient data regarding lifestyle and pre-menopausal breast cancer outcomes to draw any conclusions.

There are important factors to consider with regards to the data presented here. Overall, there are much less published data regarding additional events or recurrence in comparison to mortality. The association of physical activity, diet and adiposity with additional events appears attenuated compared to mortality, although there are not enough data to confirm this observation. Additionally, only three studies in this review [27], [40], [47] reported on pre-menopausal breast cancer specifically, while the majority of studies focused on post-menopausal or both cancers. Finally, there is no way to assess the degree of publication bias that exists because authors chose not to publish null findings.

The data from observational studies indicate that obesity is an important, modifiable risk factor for breast cancer and breast cancer prognosis. Although it may be a positive note to conclude that cancer survivors can take personal steps to reduce their risk of poor outcomes, this observation comes in the face of sobering increases in worldwide obesity prevalence [67] and research indicating that most people who achieve weight loss through lifestyle modification regain most of the weight lost over time [68]. Although we identified only 2 studies showing that type 2 diabetes (a disease largely caused be obesity) increased the risk of breast cancer mortality, there is an emerging body of literature on diabetes and cancer risk that supports the necessity for more research in this area [69], [70], [71]. An important question concerns the degree of weight loss that can reduce the risk of poor breast cancer outcomes, or even whether a fit-but-fat phenotype could also reduce risk [24]. Given that a high proportion of breast cancer patients appear to be both sedentary and obese/overweight, clinical trials are needed to investigate whether the combination of increased physical activity and reduced adiposity can improve breast cancer prognosis.

Back to Article Outline

Conflict of interest 

All authors (Patterson, Cadmus, Emond, Pierce) declare that they have no conflicts of interest.

Back to Article Outline

Funding 

Funding for this study was supported by a gift from The Safeway Foundation. The funding source was not involved in the study design, collection, analysis, and interpretation of data, writing the report, or the decision to submit the paper for publication.

Back to Article Outline

Contributors 

Drs. Patterson, Cadmus and Pierce were responsible for the manuscript design and analysis plan, interpretation of the analysis, and drafting and reviewing the article for important intellectual content. Ms. Emond performed the statistical analysis and reviewed the article for intellectual content.

Back to Article Outline

Provenance and peer review 

Commissioned and externally peer reviewed.

Back to Article Outline

References 

  1. ACS . Cancer facts and figures. Atlanta, GA: American Cancer Society; 2009–2010;
  2. WCRF . Food, nutrition and the prevention of cancer: a global perspective. 2nd ed.. Washington, DC: World Cancer Research Fund, American Institute for Cancer Research; 2007;
  3. Cummings SR, Tice JA, Bauer S, et al. Prevention of breast cancer in postmenopausal women: approaches to estimating and reducing risk. J Natl Cancer Inst. 2009;101(March (6)):384–398
  4. Lipscombe LL, Goodwin PJ, Zinman B, McLaughlin JR, Hux JE. Increased prevalence of prior breast cancer in women with newly diagnosed diabetes. Breast Cancer Res Treat. 2006;98(August (3)):303–309
  5. Kabat GC, Kim M, Caan BJ, et al. Repeated measures of serum glucose and insulin in relation to postmenopausal breast cancer. Int J Cancer. 2009;125(December (11)):2704–2710
  6. Goodwin PJ, Ennis M, Pritchard KI, et al. Fasting insulin and outcome in early-stage breast cancer: results of a prospective cohort study. J Clin Oncol. 2002;20(January (1)):42–51
  7. Pollak M, Chapman J, Shephard L, et al. Insulin resistance, estimated by serum C-peptide level, is associated with reduced event-free survival for postmenopausal women in NCIC CTG MA.14 adjuvant breast cancer trial. 2006 ASCO annual meeting proceedings (post-meeting edition). J Clin Oncol. 2006;24(18S, June Suppl.):524
  8. Mantovani A, Allavena P, Sica A, Balkwill F. Cancer-related inflammation. Nature. 2008;454(July (7203)):436–444
  9. Libby P. Inflammation in atherosclerosis. Nature. 2002;420(December (6917)):868–874
  10. Devaraj S, Rosenson RS, Jialal I. Metabolic syndrome: an appraisal of the pro-inflammatory and procoagulant status. Endocrinol Metab Clin North Am. 2004;33(June (2)):431–453[table of contents]
  11. Danesh J, Wheeler JG, Hirschfield GM, et al. C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease. N Engl J Med. 2004;350(April (14)):1387–1397
  12. Malik I, Danesh J, Whincup P, et al. Soluble adhesion molecules and prediction of coronary heart disease: a prospective study and meta-analysis. Lancet. 2001;358(September (9286)):971–976
  13. Heikkila K, Harris R, Lowe G, et al. Associations of circulating C-reactive protein and interleukin-6 with cancer risk: findings from two prospective cohorts and a meta-analysis. Cancer Causes Control. 2009;20(February (1)):15–26
  14. Pierce BL, Ballard-Barbash R, Bernstein L, et al. Elevated biomarkers of inflammation are associated with reduced survival among breast cancer patients. J Clin Oncol. 2009;27(July (21)):3437–3444
  15. Renehan AG, Roberts DL, Dive C. Obesity and cancer: pathophysiological and biological mechanisms. Arch Physiol Biochem. 2008;114(February (1)):71–83
  16. McTiernan A, Tworoger SS, Ulrich CM, et al. Effect of exercise on serum estrogens in postmenopausal women: a 12-month randomized clinical trial. Cancer Res. 2004;64(April (8)):2923–2928
  17. Prentice RL, Caan B, Chlebowski RT, et al. Low-fat dietary pattern and risk of invasive breast cancer: the Women's Health Initiative Randomized Controlled Dietary Modification Trial. JAMA. 2006;295(February (6)):629–642
  18. Pittas AG, Roberts SB. Dietary composition and weight loss: can we individualize dietary prescriptions according to insulin sensitivity or secretion status?. Nutr Rev. 2006;64(October (10 Pt 1)):435–448
  19. Cavicchia PP, Steck SE, Hurley TG, et al. A new dietary inflammatory index predicts interval changes in serum high-sensitivity C-reactive protein. J Nutr. 2009;139(December (12)):2365–2372
  20. Salas-Salvado J, Garcia-Arellano A, Estruch R, et al. Components of the Mediterranean-type food pattern and serum inflammatory markers among patients at high risk for cardiovascular disease. Eur J Clin Nutr. 2008;62(May (5)):651–659
  21. Evertitt BS, Hothorn T. In: A handbook of statistical analyses using R. Boco Raton, FL: Chapman & Hall/CFC; 2006;
  22. Friedenreich CM, Gregory J, Kopciuk KA, Mackey JR, Courneya KS. Prospective cohort study of lifetime physical activity and breast cancer survival. Int J Cancer. 2009;124(April (8)):1954–1962
  23. Sternfeld B, Weltzien E, Quesenberry CP, et al. Physical activity and risk of recurrence and mortality in breast cancer survivors: findings from the LACE study. Cancer Epidemiol Biomarkers Prev. 2009;18(January (1)):87–95
  24. Pierce JP, Stefanick ML, Flatt SW, et al. Greater survival after breast cancer in physically active women with high vegetable–fruit intake regardless of obesity. J Clin Oncol. 2007;25(June (17)):2345–2351[PMCID: 2274898]
  25. Holmes MD, Chen WY, Feskanich D, Kroenke CH, Colditz GA. Physical activity and survival after breast cancer diagnosis. JAMA. 2005;293(May (20)):2479–2486
  26. Borugian MJ, Sheps SB, Kim-Sing C, et al. Insulin, macronutrient intake, and physical activity: are potential indicators of insulin resistance associated with mortality from breast cancer?. Cancer Epidemiol Biomarkers Prev. 2004;13(July (7)):1163–1172
  27. Enger SM, Bernstein L. Exercise activity, body size and premenopausal breast cancer survival. Br J Cancer. 2004;90(June (11)):2138–2141
  28. Holick CN, Newcomb PA, Trentham-Dietz A, et al. Physical activity and survival after diagnosis of invasive breast cancer. Cancer Epidemiol Biomarkers Prev. 2008;17(February (2)):379–386
  29. Irwin ML, Smith AW, McTiernan A, et al. Influence of pre- and postdiagnosis physical activity on mortality in breast cancer survivors: the health, eating, activity, and lifestyle study. J Clin Oncol. 2008;26(August (24)):3958–3964
  30. Dal Maso L, Zucchetto A, Talamini R, et al. Effect of obesity and other lifestyle factors on mortality in women with breast cancer. Int J Cancer. 2008;123(November (9)):2188–2194
  31. Flatt SW, Thomson CA, Gold EB, Pierce JP. Low to moderate alcohol intake is not associated with increased mortality after breast cancer. Cancer Epidemiol Biomarkers Prev 2010; March (19):xxx-xxxx.
  32. Kwan ML, Weltzien E, Kushi LH, Castillo A, Slattery ML, Caan BJ. Dietary patterns and breast cancer recurrence and survival among women with early-stage breast cancer. J Clin Oncol. 2009;27(February (6)):919–926
  33. Kwan M, Kushi L, Weltzien E, Castillo A, Caan B. Alcohol and breast cancer survival in a prospective cohort study. In: Presented at the 32nd CTRC-AACR San Antonio breast cancer symposium. December 9–13, 2009, San Antonio, TX. 2009;[abstract 17]
  34. Barnett GC, Shah M, Redman K, Easton DF, Ponder BA, Pharoah PD. Risk factors for the incidence of breast cancer: do they affect survival from the disease?. J Clin Oncol. 2008;26(July (20)):3310–3316
  35. Franceschi S, Dal Maso L, Zucchetto A, Talamini R. Alcohol consumption and survival after breast cancer. Cancer Epidemiol Biomarkers Prev. 2009;18(March (3)):1011–1012[author reply 1012–1013]
  36. Holmes MD, Stampfer MJ, Colditz GA, Rosner B, Hunter DJ, Willett WC. Dietary factors and the survival of women with breast carcinoma. Cancer. 1999;86(September (5)):826–835
  37. Jain M, Miller AB, To T. Premorbid diet and the prognosis of women with breast cancer. J Natl Cancer Inst. 1994;86(September (18)):1390–1397
  38. Kroenke CH, Fung TT, Hu FB, Holmes MD. Dietary patterns and survival after breast cancer diagnosis. J Clin Oncol. 2005;23(December (36)):9295–9303
  39. McEligot AJ, Largent J, Ziogas A, Peel D, Anton-Culver H. Dietary fat, fiber, vegetable, and micronutrients are associated with overall survival in postmenopausal women diagnosed with breast cancer. Nutr Cancer. 2006;55(2):132–140
  40. Reding KW, Daling JR, Doody DR, O’Brien CA, Porter PL, Malone KE. Effect of prediagnostic alcohol consumption on survival after breast cancer in young women. Cancer Epidemiol Biomarkers Prev. 2008;17(August (8)):1988–1996
  41. Zhang S, Folsom AR, Sellers TA, Kushi LH, Potter JD. Better breast cancer survival for postmenopausal women who are less overweight and eat less fat. The Iowa Women's Health Study. Cancer. 1995;76(July (2)):275–283
  42. Kroenke CH, Chen WY, Rosner B, Holmes MD. Weight, weight gain, and survival after breast cancer diagnosis. J Clin Oncol. 2005;23(March (7)):1370–1378
  43. Berclaz G, Li S, Price KN, et al. Body mass index as a prognostic feature in operable breast cancer: the International Breast Cancer Study Group experience. Ann Oncol. 2004;15(June (6)):875–884
  44. Caan BJ, Kwan ML, Hartzell G, et al. Pre-diagnosis body mass index, post-diagnosis weight change, and prognosis among women with early stage breast cancer. Cancer Causes Control. 2008;19(December (10)):1319–1328
  45. Majed B, Moreau T, Senouci K, Salmon RJ, Fourquet A, Asselain B. Is obesity an independent prognosis factor in woman breast cancer?. Breast Cancer Res Treat. 2008;111(September (2)):329–342
  46. Loi S, Milne RL, Friedlander ML, et al. Obesity and outcomes in premenopausal and postmenopausal breast cancer. Cancer Epidemiol Biomarkers Prev. 2005;14(July (7)):1686–1691
  47. Daling JR, Malone KE, Doody DR, Johnson LG, Gralow JR, Porter PL. Relation of body mass index to tumor markers and survival among young women with invasive ductal breast carcinoma. Cancer. 2001;92(August (4)):720–729
  48. Patterson RE, Flatt SW, Saquib N, et al. Medical comorbidities predict mortality in women with a history of early stage breast cancer. Breast Cancer Res Treat, doi:10.1007/S10549-010-0732-3.
  49. Abrahamson PE, Gammon MD, Lund MJ, et al. General and abdominal obesity and survival among young women with breast cancer. Cancer Epidemiol Biomarkers Prev. 2006;15(October (10)):1871–1877
  50. Cleveland RJ, Eng SM, Abrahamson PE, et al. Weight gain prior to diagnosis and survival from breast cancer. Cancer Epidemiol Biomarkers Prev. 2007;16(September (9)):1803–1811
  51. Dignam JJ, Wieand K, Johnson KA, Fisher B, Xu L, Mamounas EP. Obesity, tamoxifen use, and outcomes in women with estrogen receptor-positive early-stage breast cancer. J Natl Cancer Inst. 2003;95(October (19)):1467–1476
  52. Lipscombe LL, Goodwin PJ, Zinman B, McLaughlin JR, Hux JE. The impact of diabetes on survival following breast cancer. Breast Cancer Res Treat. 2008;109(May (2)):389–395
  53. Nichols HB, Trentham-Dietz A, Egan KM, et al. Body mass index before and after breast cancer diagnosis: associations with all-cause, breast cancer, and cardiovascular disease mortality. Cancer Epidemiol Biomarkers Prev. 2009;18(May (5)):1403–1409
  54. Olsson A, Garne JP, Tengrup I, Zackrisson S, Manjer J. Body mass index and breast cancer survival in relation to the introduction of mammographic screening. Eur J Surg Oncol. 2009;35(December (12)):1261–1267
  55. Whiteman MK, Hillis SD, Curtis KM, McDonald JA, Wingo PA, Marchbanks PA. Body mass and mortality after breast cancer diagnosis. Cancer Epidemiol Biomarkers Prev. 2005;14(August (8)):2009–2014
  56. Caan BJ, Emond JA, Natarajan L, et al. Post-diagnosis weight gain and breast cancer recurrence in women with early stage breast cancer. Breast Cancer Res Treat. 2006;99((Sep) 1):47–57
  57. Borugian MJ, Sheps SB, Kim-Sing C, et al. Waist-to-hip ratio and breast cancer mortality. Am J Epidemiol. 2003;158(November (10)):963–968
  58. Chlebowski RT, Blackburn GL, Thomson CA, et al. Dietary fat reduction and breast cancer outcome: interim efficacy results from the Women's Intervention Nutrition Study. J Natl Cancer Inst. 2006;98(December (24)):1767–1776
  59. Pierce JP, Faerber S, Wright FA, et al. A randomized trial of the effect of a plant-based dietary pattern on additional breast cancer events and survival: the Women's Healthy Eating and Living (WHEL) Study. Control Clin Trials. 2002;23(December (6)):728–756
  60. Pierce JP. Diet and breast cancer prognosis: making sense of the Women's Healthy Eating and Living and Women's Intervention Nutrition Study trials. Curr Opin Obstet Gynecol. 2009;21(February (1)):86–91
  61. Pierce JP, Natarajan L, Caan BJ, et al. Influence of a diet very high in vegetables, fruit, and fiber and low in fat on prognosis following treatment for breast cancer: the Women's Healthy Eating and Living (WHEL) randomized trial. JAMA. 2007;298(July (3)):289–298[PMCID: 2083253]
  62. Pierce JP, Newman VA, Natarajan L, et al. Telephone counseling helps maintain long-term adherence to a high-vegetable dietary pattern. J Nutr. 2007;137(October (10)):2291–2296[NIHMSID: 30821]
  63. Gold EB, Pierce JP, Natarajan L, et al. Dietary pattern influences breast cancer prognosis in women without hot flashes: The Women's Healthy Eating and Living Trial. J Clin Oncol. 2009;27(January (3)):352–359
  64. Pierce JP, Natarajan L, Caan BJ, et al. Dietary change and reduced breast cancer events among women without hot flashes after treatment of early-stage breast cancer: subgroup analysis of the Women's Healthy Eating and Living Study. Am J Clin Nutr. 2009;89(May (5)):1565S–1571S
  65. Natarajan L, Flatt SW, Sun X, et al. Validity and systematic error in measuring carotenoid consumption with dietary self-report instruments. Am J Epidemiol. 2006;163(April (8)):770–778
  66. Kristal AR, Peters U, Potter JD. Is it time to abandon the food frequency questionnaire?. Cancer Epidemiol Biomarkers Prev. 2005;14(December (12)):2826–2828
  67. WHO. Obesity and overweight (factsheet). Geneva, Switzerland: World Health Organization 2006. http://www.who.int/dietphysicalactivity/media/en/gsfs_obesity.pdf [accessed 12.11.2009]
  68. Hill JO. Can a small-changes approach help address the obesity epidemic? A report of the Joint Task Force of the American Society for Nutrition. Institute of Food Technologists, and International Food Information Council. Am J Clin Nutr. 2009;89(February (2)):477–484
  69. Vigneri P, Frasca F, Sciacca L, Pandini G, Vigneri R. Diabetes and cancer. Endocr Relat Cancer. 2009;16(December (4)):1103–1123
  70. Hjartaker A, Langseth H, Weiderpass E. Obesity and diabetes epidemics: cancer repercussions. Adv Exp Med Biol. 2008;630:72–93
  71. Barone BB, Yeh HC, Snyder CF, et al. Long-term all-cause mortality in cancer patients with preexisting diabetes mellitus: a systematic review and meta-analysis. JAMA. 2008;300(December (23)):2754–2764

PII: S0378-5122(10)00005-8

doi:10.1016/j.maturitas.2010.01.004

Maturitas
Volume 66, Issue 1 , Pages 5-15, May 2010

Article Tools

* Image Usage & Resolution