Life history strategy, energetics & breast cancer: the role of breast density

Previous life history (LH) approaches to breast cancer have focused on hormone receptor subtypes. Oestrogen receptor positive (ER+) tumours are related to slow-LH traits, whereas ER- types may be typical of the fast LH strategy. The link between ER+ incidence with slow LH  is inconsistent with the theory that a trade-off between reproduction and cancer defence underlies higher incidence rates in species or groups who prioritise reproduction.

Breast density and life history strategy

A useful trait for understanding breast cancer aetiology from a life history perspective may be breast density, one of the strongest risk factors for breast cancer. Breast density increases with glandular and connective tissue volume – the biological structures involved in lactation – compared to fat. It seems intuitive to link a higher mass of this functional tissue in the breast to increased biological investment in lactation.

In the slow LH strategy, more biological investment is made in the growth and development of offspring, who are usually fewer in number. Just as higher investment in reproduction increases cancer risk in the fast LH strategy, this higher investment in each offspring would be expected to increase cancer risk in the biological structures involved.

The slow LH strategy involves heightened cancer defence, but this must be massively outweighed by the inevitable risks of investing in breast function. The likely mechanism of carcinogenesis of breast density is a larger number of cell divisions – a theory also applied to the risks of earlier reproduction.

From a LH perspective, breast density would be considered a slow trait, if it is indeed associated with higher somatic investment in the quality and/or quantity of breast milk. Several typically slow LH traits are predictive of denser breasts: higher socioeconomic status (SES), greater birth weight, longer birth length, taller adult height, lower BMI, higher bone density (which may reflect greater somatic investment), lower parity, and later age at first birth. However, higher educational attainment has been associated with both less and more dense breasts.

Breastfeeding duration in rich and poor countries

Breastfeeding duration might be expected to correlate with breast density, if both are measures of investment in offspring. Breastfeeding duration has been put forward as a slow LH trait, with longer durations at higher SES in high income countries (HICs). Consistent with this, breast cancer risk increases up the income scale.

However, if anything longer breastfeeding duration has been found to be protective against breast cancer. This seems to go against the findings in HICs, but in low and middle income countries (LMICs) the social gradient in breastfeeding is reversed. Low SES women breastfeed for longer, and in fact most women breastfeed for much longer than any group of women in rich countries. However, many factors affect breastfeeding duration, not least the availability and affordability of formula milk, and these may significantly skew any LH link with breastfeeding duration.

The protective effect of breastfeeding may come from this distinct pattern of extended lactation, which is closer to the pattern in our evolutionary history. Consistent with this is that the protective effect is often only found after a certain duration of breastfeeding, and then becomes stronger with increasing duration. If HR- cancers are fast LH typical, breastfeeding may appear protective as it is a slow trait, at least in HICs. Taking this effect out may leave little effect of breastfeeding on HR+ types, especially at shorter durations. Some studies have only found a protective effect of breastfeeding in pre-menopausal women, and indeed HR- types mainly affect women in their reproductive years.

Evidence on breast density in developing countries is scarce, but it might be predicted to be lower. If so, it could be a better candidate for explaining the lower cancer risk, with breastfeeding duration just a confound. There would be no positive relationship between density and breastfeeding, and indeed one study found breastfeeding for more than 8 months has an odds ratio for dense breasts of 0.72.

Some studies even find an elevated risk of developing breast cancer in women who breastfeed for an intermediate length of time, compared to women who don’t breastfeed at all after giving birth. Again, this won’t be because of a harmful effect of breastfeeding, but rather that those women who breastfeed for intermediate lengths of time tend to be high SES women in high income countries, who are at the highest risk of breast cancer for other reasons. Caution should be taken when comparing with women who don’t breastfeed however, as there are many reasons why women don’t or can’t breastfeed. They can’t therefore be taken as a homogenous group who might be characterised as fast LH.

It appears that breastfeeding follows fundamentally different patterns in developed and developing countries. To conflate the two in analyses of breastfeeding’s effect on breast cancer might therefore be inadvisable, especially given the widely varying rates of the disease between countries. A protective effect of breastfeeding in rich countries seems incompatible with the higher rates of the disease higher up the social gradient, in women who tend to breastfeed for longer.

Prolonged breastfeeding in rich countries

Prolonged breastfeeding isn’t confined to women in low-income countries though. Immigrants initially retain the breastfeeding pattern of their country of origin, only taking on the traits of their new country after a couple of generations. Even this is contingent on the level of cultural assimilation, and less assimilated groups may retain their traditional nursing practices. Women in rural areas may also be more likely to practice extended lactation.

There is the question of what drives different weaning timings – in low income countries, it may have to do with calorie and nutrient availability. These are generally lower, so infants may require a longer duration of breastfeeding, with what may be smaller amounts of less nutritious breast milk.

The presence of prolonged breastfeeding in high income countries means that it could influence the apparent effect of breastfeeding on breast cancer risk in these countries. The duration of breastfeeding itself may not be the causal factor, rather the women who breastfeed for prolonged durations would be hypothesised to have lower breast densities.

It is an open question to what extent this is mediated by the breastfeeding, and to what extent these women have less dense breasts already. A LH approach might suggest that the fast LH trait of low breast density is already present, lowering breast cancer risk. Breastfeeding duration here isn’t considered a slow LH trait, as there seem to be other considerations.

Energetics and age at menarche

The mixed, and perhaps counter-intuitive evidence on factors like breastfeeding duration and age at menarche from a fast/slow life history perspective, may be better understood when energy availability constraints are considered. Age at menarche tends to be later in populations where low energy availability from food limits growth rates. This prevents any tendency for a fast LH strategy to engender early menarche.

Prolonged breastfeeding, which also tends to be found in LMICs, may be analogous. Low energy availability limits the quantity and quality of breast milk, as well as nutrition in utero, and so it may be necessary to breastfeed for longer to ensure the viability of offspring. Alternative infant feeding methods may be less widely available too.

Older age at menarche and longer breastfeeding duration are both thought of as slow LH traits, certainly in WEIRD populations, but they are found in LMICs due to energy constraints. Indeed menarcheal age can be much older, and breastfeeding duration much longer, than even slow LH women in HICs. The fast traits of a younger age at menarche and shorter breastfeeding are not possible or adaptive in an energy-poor context. Overall, the “protective” effect of the energy constraints masks the risks of the slow traits. Women who reach menarche later and breastfeed for longer tend to have a lower risk of breast cancer, not because these traits are protective themselves, but because they tend to go along with other traits which are protective.

The secular trend in age at menarche

However, some studies have found later age at menarche to be protective against breast cancer in HICs, where energy constraints wouldn’t be expected to exist to the same extent as in LMICs. The quality as well as quantity of nutrition may be important, and increasing nutrition is likely to lower menarcheal age in a dose-response manner. This may underlie the secular trend of decreasing age at menarche in HICs and some middle income countries in previous decades.

A 1970s study linked a higher proportion of girls reaching menarche over the age of 16 in Poland and Tokyo to their lower incidence of breast cancer, compared to women in Athens and Boston. In China, SES is negatively related to age at menarche. The studies which the risk of early age at menarche is based on have been published over the last several decades, on women of all ages. They therefore study women who were born over the course of more than a century. Young women affected by food shortages around WWII, or widespread famine in China, faced significant energetic constraint.

In many HICs the secular decline has continued, suggesting that there continues to be gains to be made from improving nutrition. Young women today may be less constrained by energy availability, but this is a recent phenomenon, and they are only just starting to appear in studies. Life history speed would have been very much secondary to energetics for the majority of women in studies before now.

A cohort effect may be detectable in the two studies which distinguish between pre- and postmenopausal women, both in Canada. In both older cohorts, there is a significant protective effect of older age at menarche, whereas in the younger cohorts, there is no effect or even a risk of older menarche. This is consistent with the significant but declining importance of energetics on menarcheal age in HICs. By contrast, for parity, the effect is if anything the other way around – the premenopausal women show the biggest protective effect of having more children, with postmenopausal women showing less of an effect.

The decreasing incidence of infectious disease may be a factor too, freeing up energy spent on immune response to be invested in growth instead. Illness and injury are likely to show a social gradient, with lower SES girls having to invest more energy on immune response and healing. Girls are also less active than in previous generations, meaning less energy is expended on exercise.

The alternative, LH explanation would be that LHs are accelerating in HICs – it is not clear why this would be the case, and it goes against other evidence. As the secular trend continues, inevitably there will be inequalities within countries in energy balance, nutrition and age at menarche, and this may be what the studies on menarche and breast cancer risk are picking up on. Again, energetics and LH speed have opposing effects on a social gradient in age at menarche.

Other factors play a role in nutrition. Energy balance is important – physical exercise uses up energy that could otherwise go into development, as evidenced by delayed menarche in young female athletes. The very idea of everyone having enough food to eat in rich countries is given the lie to by evidence of widespread food poverty and soaring food bank use. There is a potential link between animal protein, an expensive food source, and increased breast cancer risk. Both may be related to earlier menarche.

Age at menarche is strongly inversely correlated with life expectancy across countries. Even within HICs there is unlikely to be a positive correlation, diminishing the idea of early menarche as an exclusively fast LH trait, or rather making the case that LH speed is not the only determinant. An international study of breast cancer risk found that late menarche was only protective in women who weren’t overweight or obese. This could be interpreted as supporting the theory that only late menarche caused by energy constraint is protective against breast cancer.

The subset of women who reach menarche later and breastfeed for longer because they are following the slow LH strategy may actually have a higher risk. This may not be apparent as they are closer to the average for these traits – they only mature later and breastfeed for longer compared to women who aren’t energetically constrained either, but are following the fast strategy. For instance, in HICs high SES women tend to breastfeed for longer, and may reach menarche later, than low SES women. However, these high SES women are at higher risk of breast cancer, which suggests that late menarche and longer breastfeeding aren’t in themselves protective. Again, these traits aren’t themselves risky, but they tend to go along with other traits that are, in slow-strategy women.

Breast density – investment in lactation

Breast density may be one such trait. The important distinction is that the slow trait – high density – is also the one which requires the more energy availability. Because of this, the risk profiles according to energy availability and LH speed go the same way, rather than being countervailing forces. Energetics accounts for more variance in age at menarche and breastfeeding duration, especially at older ages and longer durations. Because of this, energetics comes out as the statistically significant predictor of these outcomes.

As a result, breast density is an extremely strong risk factor for breast cancer, with dense breasts having 4-6 times the risk of carcinogenesis. Density would be expected to be higher on average both in HICs compared to LICs, and at high SES compared to low SES within countries. Other traits that are both adaptive in energy-rich environments and slow LH-typical are late age at first birth, and low parity. These are also correlated with breast density and breast cancer risk.

One exception is that breast density predicts both HR+ and HR- types of breast cancer, with HR- cancer possibly being fast LH typical, and breast density here being proposed as slow LH. Once again, energy availability may be the explanation – as it increases, both investing in breast density and following a fast LH become more possible and adaptive. They appear to be divergent paths, but they are both disproportionately present in a small group, mostly in HICs. There is greater variance in cancer risk between women with high and low energy availability, and this may be what the correlation between breast density and HR- types is picking up on.

It might also be predicted that breast milk quantity is another trait which fits into the cluster: high energy availability, slow LH, dense breasts and higher breast cancer risk. The composition of breast milk may also be different. All of these factors are linked to the function of the breast, and it may be fruitful to look at breast milk and associated physiological factors around the time of lactation – components of breast milk have previously been put forward as potential biomarkers of cancer risk. For example, most mammograms are carried out long after this stage, and information about disease risk may have been lost by then. The theory is that increased potential (if not actual) breast function, of which breast density is an indicator, carries with it an inherently higher risk of tumour development.

Summary

Age at menarche and breastfeeding duration follow different social gradients in low and high income countries because energetic constraints hold sway in LMICs, with LH strategy having more influence in HICs, with the effects being countervailing. With traits like breast density, parity and age at first birth, energetics and LH lead to similar social gradients, giving stronger risk or protective factors which are more universally applicable. Breast density may be a proximate mechanism for the slow LH and high energy availability strategy of investing more biological resources in each offspring. The increased cell division involved may be why breast density is such a strong risk factor for breast cancer.

Advertisements