Abstract CN06-02: Soy foods and cancer risk: How does the relation vary in individuals with different ethnicity, diet, and lifestyle?

Abstract

Abstract Soy foods have been part of traditional Asian diets for many centuries, while soy products were introduced in Western countries only a few decades ago. Migrants from Asia maintained their food habits to a certain degree when they moved to the U.S., but they also adopted Western foods (1). Inspired by studies among migrants and by the low incidence rates for several cancers among Asians, research into the protective effects of isoflavones in soy foods was initiated. Due to their estrogen-like structure, much of the research has focused on hormone-dependent cancers, i.e., breast, endometrial, ovarian, and prostate cancer, but some work on colorectal cancer has also been performed. During years of experimental and epidemiologic research, it has become evident that not all populations may derive the same benefit from consuming soy foods. One of the more consistent findings in soy research is the fact that epidemiologic studies report a stronger protective effect of soy foods against breast cancer among women who grew up in Asian countries and in those who consumed soy foods throughout childhood and adolescence (2). This presentation will focus on epidemiologic studies that have investigated the association between soy foods and cancer risk among populations with different ethnic backgrounds. In addition, some underlying mechanisms that may be responsible for ethnic differences, such as intestinal equol production, genetic polymorphisms, and timing of exposure will be discussed. Epidemiologic studies among women of Asian, primarily Japanese and Chinese, ancestry show a more consistent protective effect against breast cancer than investigations among Western populations; the odds ratio (OR) was 0.71 (95% CI: 0.60-0.85) among Asians when the highest and lowest intake group were compared, while soy food was unrelated to breast cancer risk in Western studies (OR = 1.04; 95%CI: 0.97-1.11) (2). Similarly, a meta-analysis of prostate cancer studies found a lower risk associated with soy consumption only among Asian (OR = 0.52; 95%CI: 0.33-0.81) but not Western populations (OR = 0.99; 95% CI: 0.85-1.16) (3). On the other hand, meta-analyses for endometrial, ovarian, and colorectal cancer described only small differences in associations by ethnicity (4, 5). Recent case-control analyses nested within the Multiethnic Cohort in Hawaii and Los Angeles examined urinary isoflavone excretion as a biomarker of soy intake and reported similar results for breast but not prostate cancer. While the risk associated with breast cancer was 0.69 (95% CI: 0.51-0.92) among Japanese Americans and 0.98 (95% CI: 0.61-1.55) among Caucasians (6), the risk estimates did not differ by ethnicity for prostate cancer (7). Evidence from observational and intervention studies also suggests possible differential effects of soy intake on biomarkers of breast cancer risk (8-10), e.g., mammographic density and concentrations of serum estrogens and testosterone, urinary estrogen metabolite patterns, IGF-I, and PSA. Whereas the higher intake and possibly the type of soy foods in Asian populations may explain the observations for breast and prostate cancer risk (11), bioavailability of bioactive substances may also be a contributing factor (12). For example, the capacity of the intestinal bacteria to metabolize the isoflavone dadzein into the metabolite equol varies across populations and may confer a greater protection against disease than the other isoflavonoids (13, 14). Another reason that not all individuals may benefit from soy food exposure to the same degree is genetic variation. Genetic polymorphisms may alter enzyme activity or modulate the expression of genes involved in pathways influenced by isoflavones and relevant to cancer risk. Menopausal status, obesity, and other dietary components are other causes for isoflavones to exert differential actions across populations. Alternatively, the overall effect of soy foods on carcinogenesis may depend on the time of life when isoflavones were administered because isoflavones have estrogenic and anti-estrogenic effects. A number of case-control studies that assessed soy intake during childhood or adolescence found a stronger protection for early life than adult soy intake. Since experimental studies indicate that estrogen exposure in young animals induces protection against cancer development, the weak estrogenic effects of isoflavones in soy beans, if consumed early in life, may achieve or accelerate differentiation of breast tissue structures similar to an early pregnancy and, thereby, make the tissue less susceptible to carcinogens and prevent tumor development later in life. Since the reports about beneficial effects of soy against cancer come primarily from Asian populations who have consumed soy foods since childhood and whose diet and lifestyle differs in many other ways from Western countries, the expectations of individuals in Western countries who start eating soy products later in life may not be fulfilled. Therefore, future research in different populations should focus on how soy affects cancer risk throughout life.