Introduction

Breast cancer (BC) remains one of the commonest cancers – one in eight women will be diagnosed with BC in her lifetime [1] – and a leading cause of death from cancer. However, it remains a significant scientific and medical challenge. One of the crucial gaps identified is how to implement a sustainable preventive lifestyle strategy [2]. Both risk factors and protective factors must be considered. Some risk factors, such as genetic predisposition, cannot be modified, whereas others (unhealthy diet, sedentary lifestyle) can be avoided. Increasing protective factors may be crucial for women at high risk (as evaluated by the National Cancer Institute’s (NCI’s) BC Risk Assessment Tool [3]) and in preventing recurrence and improving survival after BC diagnosis. Decreasing the length of time a woman’s breast tissue is exposed to estrogens may help prevent BC [13], although the main ways of achieving this (first pregnancy before the age of 20 years, breast feeding, late menstruation and early menopause) are difficult to control.

There are strong links between environmental/lifestyle factors and BC, suggesting that modifying these factors may result in decreasing BC risk, although there has been no randomized trial that clearly demonstrated this. As changing these factors has been shown also to definitely reduce the risk of fatal diseases, notably cardiovascular diseases in randomized trials, it is reasonable to propose these changes to high-risk women and to women who wish to reduce their BC risk.

For instance, dietary fats have been extensively studied in the prevention of BC [4]. Neither animal fat nor a low-fat diet has been linked to BC risk, whereas marine omega-3 fatty acids (n-3) may be protective [4]. In a meta-analysis of 21 independent prospective cohort studies, Zheng et al. found a significant reduction of BC risk with marine n-3 [5]. However, this meta-analysis highlights the difficulties in assessing the effects of specific dietary fats on BC risk. In subgroup analyses, Zheng et al. found that the inverse associations between marine n-3 and BC risk were significant only in post-menopausal women, were stronger in East Asian populations compared with western populations and were more evident without adjustment for body mass index (BMI). This suggests that marine n-3 may influence BC risk partly through an effect on BMI or related factors (insulin or adipokines), whereas the importance of BMI on BC risk during the pre-menopausal and post-menopausal periods is still a matter of controversy [6]. The East Asian population issue suggests that other foods/nutrients and related factors may be involved. For instance, omega-6 fatty acids (n-6) may play a role in BC risk [7]. It has long been suspected that n-6 increase the risk of cancers, and this has been confirmed in controlled trials in which n-6 intakes were modified. In the Los Angeles Trial, there were more cancers in the experimental group with high n-6 intake, whereas in the Lyon Diet Heart Study, there were fewer cancers in the group with low n-6 [8, 9]. Thus, when analyzing the associations between n-3 and BC risk, it is crucial that n-6 is included in the analyses, as Yang et al. did in their recent study published in BMC Cancer[10].

Breast cancer risk and n-3/n-6 ratios

Yang et al. used the ratio of n-3/n-6 in a meta-analysis comprising 274,135 women with a total of 8,331 BC events from 11 independent prospective studies [10]. Women with a higher n-3/n-6 ratio had a significantly lower risk of BC: pooled RR 0.90, 95% CI 0.82 to 0.99. When the authors analyzed only dietary intake, they found a 6% reduction in BC risk per one-tenth increment of the n-3/n-6 ratio. The association between the blood phospholipid n-3/n-6 ratio and BC risk (in four studies only) did not reach statistical significance. The lack of a significant relationship between BC risk and blood n-3/n-6 ratio is not unexpected, because n-3 and n-6 measured in phospholipids do not accurately reflect dietary intake. Several factors (discussed in the next paragraph) interfere with the levels of both n-3 and n-6 in each class of phospholipids. In addition, the fatty acid composition of each phospholipid is not identical: it differs in serum, red cell membranes, and mitochondria, for example, and these various phospholipids have a different physiology and a different effect (if any) on cancers [7]. There are nonetheless correlations between dietary n-6 and n-3 and the corresponding fatty acids in blood and cells, and it is not illogical to pool data from diet and blood when analyzing relationships with BC risk. It is, however, crucial to bear these differences in mind when interpreting the results, and potential confounders should be included in the analyses if possible; that is, if they have been recorded in a timely and accurate fashion. In the next section, we highlight some lifestyle, environmental, and pharmacological factors that influence n-3/n-6 ratios, and examine their relationships with BC risk. In most studies included in the meta-analysis by Yang et al., these factors were not measured, which is likely to have weakened the associations between n-3/n-6 ratio and BC risk.

Factors influencing n-3/n-6 ratios and BC risk

Among the main risk factors of BC, estrogen exposure, lack of physical exercise and being overweight are well known [13].

Studies support an association between endogenous sex hormone levels and BC risk for post-menopausal women, whereas the association is less clear for pre-menopausal women [11]. One possible explanation is that the high estrogen levels present before menopause increase blood marine n-3 [12], which in turn may partly counteract the effect of estrogens, as marine n-3 are protective [5, 7, 10].

Besides increased intake of n-3 and decreased intake of n-6 through consumption of foods rich in n-3 and poor in n-6 [79], other substances are known to affect the n-3/n-6 ratio. The endogenous synthesis of marine n-3 from their plant substrate alpha-linolenic acid (ALA) is stimulated by the plant pigment polyphenols found in purple fruits such as grapes (and wines), plums, and blueberries [1315]. Polyphenol flavonoids increase marine n-3 by 30% without altering n-6 levels, resulting in a significant increase in the n-3/n-6 ratio. As expected from the above data, flavonoids are associated with a decreased BC risk [16]. These data are encouraging, as it is difficult to accurately determine the intake of each flavonoid, and this probably weakens the association between specific flavonoids and BC risk. In addition, the bioavailability and biological effect of flavonoids are dependent on many factors, including the gut microbiota [17], which again is likely to weaken the associations. A typical example of such complexity is provided by the soy flavonoids.

Soy isoflavones, a major class of phytoestrogens, may reduce BC risk, but epidemiologic studies have yielded inconsistent results [18]. A meta-analysis suggests that protection is only observed in studies conducted in East Asian and not in western populations [18], suggesting that protection may require that women consume the high levels of soy typical of East Asian diets. In addition, the food source of isoflavones, timing of exposure to isoflavones (starting at or not before adolescence), the woman’s menopausal status, and the ability of the gut microbiota to transform isoflavones into equol probably modify the association between soy isoflavones and BC risk. Equol is more biologically active (that is, more anti-estrogenic) than dietary isoflavones, but only about 30% of women in the USA and Australia are equol producers [19]. Finally, the small studies examining whether equol exposure is associated with BC risk have produced inconsistent results [20]. Other phytoestrogens more characteristic of the western diet are the polyphenol lignans present in seeds, grains, fruits, and vegetables. High dietary lignans or high exposure to enterolignans, the metabolites resulting from transformation by the gut microbiota, are associated with lower BC risk and better survival in post-menopausal BC [21, 22]. This is not unexpected, as we have shown in rats that lignans increase blood n-3 without affecting n-6, resulting in a higher n-3/n-6 ratio [7, 13].

Another crucial point is the fact that organic plant foods contain more polyphenols than similar conventional foods [2326]. Organic animal fat also - for instance, milk and milk products - do have a higher n-3/n-6 ratio than conventional products [2729]. Organic-fed animals mainly consume non-contaminated fresh grass with a high polyphenol content, rather than polyphenol-poor concentrates potentially contaminated with pesticides, and this may partly explain the favorable n-3/n-6 ratio [30, 31].

Regarding food contaminants, recent studies have shown a strong association between estrogenic PCBs (polychlorobiphenyls) congeners and BC risk [32]. In this context, it is crucial to identify ’hidden’ groups of women whose occupational exposure to carcinogens is underrepresented in epidemiological studies. Studies have recently shown significant associations between BC risk and endocrine disruptors in women with specific professional exposure [33]. Endocrine disruptors (particularly phthalates) increase insulin resistance, and the risk of diabetes and obesity [34], all of which increase BC risk [13].

Other commonly encountered substances that potentially influence both n-3/n-6 ratio and BC risk are the cholesterol-lowering statins. Statins decrease n-3/n-6 ratio [35], increase insulin resistance and diabetes risk [36], lower cholesterol, and are toxic to mitochondria [37]. Whether or not statins, by modifications of the above processes [3538], are involved in increased BC risk, remains an area of controversy. Therefore, when weighing up the benefits and potential risks of statins, caution should be applied prior to prescription. Following the same line of reasoning, high-risk women should do everything they can to reduce their risk of insulin resistance, metabolic syndrome and diabetes [13]. Beside optimal physical activity, which is a well-known strategy for decreasing both diabetes and BC [13] risks, high fiber intake, high flavonoids, and high n-3 are all inversely associated with diabetes risk. Accordingly, fiber intake, flavonoids [16], and n-3 [5, 7, 10] are inversely associated with BC risk. Finally, consumption of foods with a lower glycemic impact – that is, a low glycemic index (GI) – is associated with a lower incidence of diabetes and a lower BC risk [13].

These data are highly consistent, and the combination of high fiber, high n-3/n-6 ratio, high polyphenols, and low-GI foods represents what many experts call a healthy dietary pattern – for instance, the Mediterranean diet – which has been associated with a lower BC risk and better survival for women with early-stage BC.

Reducing BC risk

In summary (see Table 1), the accumulated evidence shows that in addition to genetic predisposition and estrogen exposure, a number of lifestyle, environmental, and pharmacological factors play an important role in BC risk and BC survival. Optimal physical activity decreases insulin resistance, diabetes risk, and the risk and progression of BC [39]. Adhering to a healthy dietary pattern – a modernized Mediterranean diet [40], or a similar healthy diet adapted to specific populations, such as the Okinawan diet in East Asians [41] – has been shown to be effective. Emphasis should be given to increasing plant and animal (including marine) n-3, and decreasing plant and animal n-6. High polyphenol consumption, in particular of flavonoids, which increase the synthesis of marine n-3 and result in a higher n-3/n-6 ratio, is associated with lower BC risk. To reduce insulin resistance and diabetes, which are associated with an increased BC risk, women should increase their consumption of n-3 and fiber and favor low-GI foods. In addition, organic foods have been shown to contain more polyphenols and have a higher n-3/n-6 ratio than non-organic foods. As a higher n-3/n-6 ratio has been associated with decreased BC risk, consumption of organic foods may be beneficial. Care must be taken to reduce (professional and non-professional) exposure to environmental chemical contaminants. Drugs that decrease the n-3/n-6 ratio, and/or increase insulin resistance and diabetic risk (in particular statins) should be taken with caution.

Table 1 How to reduce breast cancer risk and improve survival after diagnosis

Conclusions

National and international cancer associations regularly release anti-cancer guidelines. For instance, every five years the American Cancer Society publishes its Nutrition and Physical Activity Guidelines, which reflect current scientific evidence and focus on recommendations for individual choices regarding diet and physical activity patterns [42]. These Guidelines state: ‘for people who do not use tobacco, the most important modifiable determinants of cancer risk are weight control, dietary choices, and levels of physical activity’ and ‘although genetic susceptibility influences the risk of cancer, most of the variation in cancer risk across populations and among individuals is due to factors that are not inherited’. Regarding BC specifically, the Guidelines state, ‘the best advice to reduce the risk of breast cancer is to engage in regular, intentional physical activity; to minimize lifetime weight gain through the combination of caloric restriction (in part by consuming a diet rich in vegetables and fruits) and regular physical activity; and to avoid or limit intake of alcoholic beverages’[42].

We agree wholeheartedly with this advice. However, we feel it is time to go further and to be more specific. A specific healthy dietary pattern such as the modernized Mediterranean diet, and not simply “consuming a diet rich in vegetables and fruits”, should be adopted to decrease BC risk. This is also an effective way of maintaining a healthy weight, and preventing diabetes and cardiovascular disease. The focus must be on the n-3/n-6 ratio and polyphenols. Organic foods are shown to contain higher levels of these than non-organic foods, and so may be beneficial. Great care should be taken when using drugs that potentially increase BC risk. This also applies to BC survivors to prevent recurrence and improve survival [43].