Due to the high incidence and unsatisfactory prognosis, we sought strategies to prevent BC.[2, 3] Diet and nutrition are one of the directions, and numerous studies have examined the association between food intake and BC incidence[75–77]. Among them, regarding dairy products, poor consistency was observed among the results of many studies[39, 55, 78]. Besides that, dairy products contain complex nutritional components, and their consumption is widespread in daily life.[79] Many studies have indicated that dairy products may also impact health problems such as obesity[80], diabetes[81, 82], cancer[83], and coronary heart disease[84, 85], but whether dairy products play a protective or harmful role against BC occurrence in female population is still controversial. In light of these considerations, this study conducted a meta-analysis approach to analyze the retrieved observational studies.
According to data analysis, we found that total dairy products have a protective effect on female populations, which can reduce the incidence rate of BC, especially ER + and PR + BC. For ER+/PR + BC, there is a trend of protection without reaching a statistical significance. Fermented dairy products can reduce BC risk in postmenopausal population but have no protective effect on premenopausal population. Non-fermented dairy products have no significant effect on BC occurrence. There is a trend towards harm in the female population with high-fat dairy products. Although a statistically significant difference was not reached, it is suggested that excess high-fat dairy intake may elevate BC incidence. In contrast, low-fat dairy products can protect the premenopausal population and reduce the incidence rate of BC.
Because of their complex composition, no clear mechanism of action has been proposed to explain the contribution of dairy products to BC incidence.[17] Recent works have shown that dairy products have both pro- and anti-carcinogenic effects. Dairy products are rich in calcium, vitamin D, and conjugated linoleic acid (CLA), which affect cell proliferation and differentiation and can inhibit tumor development.[86–89] Conversely, large amounts of fats, saturated fats, and potentially carcinogenic contaminants (pesticides, estrogen metabolites, and growth factors including insulin-like growth factors-1 (IGF-1)) in dairy products increase BC risk.[86, 90, 91]
Dairy calcium can have antiproliferative properties by reducing the ability of circulating lipids to induce cell proliferation.[89, 92, 93] Vitamin D was shown to interrupt insulin and IGF-1 and reduce carcinogenesis.[94, 95] Insulin stimulates the elevation of free IGF-1, which can promote cell cycle progression and angiogenesis and has anti-apoptotic properties.[41, 67, 96] CLA is a mixture of positional and geometric isomers of linoleic acid, which are mainly from dairy products (60%) and beef products (32%).[97] It has anticancer effect in animal experiments and is an effective anticancer drug with cytotoxic effects on BC cells[70, 98], potentially by inhibiting the cyclooxygenase-2 or lipoxygenase pathways or by inducing the expression of apoptotic genes.[68, 99] Accordingly, we speculate that dairy products may have a protective effect on female populations through many of the above pathways to reduce the risk rate of BC. In addition to these, studies have shown that dairy partial ingredients and their bioactive components can exert inhibitory effects on BC by downregulating ER-αexpression and activity, inhibiting proliferation, metastasis, and angiogenesis of breast tumor cells, inducing apoptosis and cell cycle arrest and sensitizing breast tumor cells to radiotherapy and chemotherapy, which may explain the reason why dairy products have apparent protective effects on hormone receptor-positive BC.[33, 100–102]
Fermented dairy products, such as yogurt and cheese, have a superior nutritional value over non-fermented milk due to their high concentrations in beneficial bacteria, calcium, riboflavin, vitamin B6, and vitamin B12.[103] Studies have shown that the probiotic Lactobacillus acidophilus, present in yogurt, modulates immune responses against BC in mouse models.[104] By increasing age, the gut exhibits declining levels of bifidobacteria with antitumor effects, which allow toxin-induced tumor growth producing bacteria to thrive, such that postmenopausal populations have lower levels of intestinal probiotics than premenopausal populations.[105] Whereas fermented dairy products can provide probiotics to supplement and balance the gut microbial community, thereby reducing cancer risk.[106] In addition, IGF-I amount that may increase BC risk is significantly reduced in processed heat-treated or fermented dairy products.[107, 108] Combined with the above mechanisms, it may help explain that consumption of fermented dairy products can reduce BC incidence in postmenopausal population. Similarly, non-fermented dairy products that lack probiotics cannot balance the intestinal microbial community and may not affect BC occurrence.
Two potential reasons exist for the detrimental effect of high-fat dairy products on female BC. The first explanation is that high-fat dairy products directly induce BC.[109] Animal experiments have found that a high-fat diet increases mammary epithelial cell proliferation, particularly "hormonally driven" hyperproliferation during mammary gland outgrowth and development in young animals, enhancing the promotion of chemically induced mammary carcinogenesis and increasing BC risk in adulthood.[93, 110] The second explanation is that consuming high-fat dairy products leads to obesity and type 2 diabetes mellitus (T2D), thus increasing the risk of cancer. In particular, BC is more strongly associated with obesity than other cancers.[111, 112] Obese people also have metabolic syndrome, which increases their risk of T2D. Obesity and T2D result in hyperinsulinemia[113, 114], increased insulin-like growth factor[115, 116], hyperglycemia[117], dyslipidemia[118, 119], increased adipokines[120], increased inflammatory cytokines[121] and changes in intestinal microorganisms[122, 123], which may increase the risk of cancer through various mechanisms.
Although consuming fats increases BC risk[124, 125], the effect of different types of fatty acids on BC varies, with studies suggesting that saturated fatty acids are associated with increased BC risk. However, no clear association between total, monounsaturated, or polyunsaturated fatty acids and BC has been observed.[126, 127] Whereas low-fat dairy products differ from high-fat dairy products in the amount of fatty acids obtained by refining full-fat dairy products to extract most saturated fatty acids while preserving unsaturated fatty acids.[108] In addition to this, studies have demonstrated that the association between dairy intake and BC risk is stronger in premenopausal women than in postmenopausal women[41, 67, 128], possibly caused by more robust interactions between calcium, vitamin D, and IGF-1 than in postmenopause.[128–130] This may explain why low-fat dairy products appear to have a risk-lowering impact in the premenopausal population but have no apparent effect in the postmenopausal population.
Zang J et al.[108] also conducted a meta-analysis of the relationship between BC and dairy products, but their study did not conduct multifactorial subgroup analysis. In addition, their study considered cheese and butter as the same exposure factor and yogurt as another exposure factor. However, since cheese and yogurt are fermented dairy products and butter is not, they did not analyze fermented dairy products as a whole and misclassified cheese and butter, implying that the result could lack reliability. There was also inadequacy in the meta-analysis of Dong JY et al.[131] First, their studies were only retrieved from PubMed one database, probably ignoring the potentially included studies. Second, they only performed a subgroup analysis on the role of milk and could not provide a comprehensive conclusion.
Although this meta-analysis reached comprehensive and objective conclusions, there are still potential limitations that need to be considered. First, all risk estimates of included studies used multivariate models, but potential risk factors were not adjusted in the same way between studies. Therefore, we cannot exclude the possibility that inadequate control of potential risk confounders contributed to biased results. Second, most studies assessed diet using self-reported FFQs based on their plausibility and reproducibility, although a few used interviews, implying that dietary assessment inevitably produces assessment or measurement error, most likely biasing the findings. Third, consideration should be given to classification biases in dairy products. Because dietary assessments across studies were conducted based on different databases and different FFQs, there may be inconsistencies in the conclusions resulting from different methods of classifying dairy products. Fourth, there may be significant heterogeneity among the included studies due to reported population variations in terms of lifestyle preferences, living locations, and so on. Fifth, not all trials had accessible relevant subgroup data, such as subgroup data by BC type, subgroup data by menopausal status, and thus large-scale observational studies are still required to further validate the relevant conclusions.
Beyond limitations, this meta-analysis also has its own strengths. First, this study included many observational studies with more than one million participants in Asia, Europe, and the Americas. The larger observational population increases the reliability and authenticity of this study's conclusions. Secondly, this study selected HR data that adjusted the largest number of potential confounding factors for statistical analysis to improve accuracy. Third, the study grouped the abstracted data (by BC type, menopausal status, or type of dairy products) and did subgroup analyses to comprehensively screen the possibility of effects of different dairy products on different populations and different BC types. Overall, this meta-analysis leads to meaningful conclusions that may allow new recommendations on BC prevention in female populations.