Breast cancer is the second most common type of cancer among Korean women [1] and the most common among women globally [2]. The incidence of breast cancer is lower among women in Asia than among women in Western countries [3], with different trends in age at diagnosis [4]. These epidemiological differences may be due to biological variation in estrogen receptor (ER) status among racial/ethnic groups [5] and the prevalence of risk factors such as mammographic breast density [6].

Breast density has been shown to be associated with reproductive and hormonal factors [7], and among Western women, breast density is considered one of the strongest known risk factors for breast cancer development [8]. Breast density also varies among racial groups and may explain some of the racial disparity in breast cancer incidence [9].

Mammographic density and the age-specific incidence of breast cancer differ markedly between women in Asian and Western countries [10-13]. Therefore, there may be differences in the relationship between breast density and breast cancer risk in Asian women [9], including Korean women [14,15], compared to women in Western countries.

Evaluating the distribution of breast density by age group is necessary to determine whether there is an effect of breast density on the risk of breast cancer [16]. Two studies have examined breast density among healthy Korean women [15,17], although these data were collected in 2002 [15] and 1998 [17]. Because the incidence rate of breast cancer has increased in Asian countries, including South Korea [1,11], the aim of the present study was to describe the distribution of mammographic breast density by age group in a Korean population.

The distribution of PDB by age group is shown in Table 1. In total, 231,058 participants were enrolled in the study, and the overall PDB was 78.15%. The highest PDB was observed in the youngest age group (30-34 years, 94.87%). PDB decreased as age increased. The greatest difference in PDB was observed between the 55-59 years and 60-64 years age group (-20.04%).

Figure 1 and Table 2 illustrate the difference in PDB distribution between the KMI database and the dataset of Kim et al. [17], stratified by age group. There was an increase in PDB after ten years in all age groups, except women 35-39 years old. The greatest increases were observed among the age groups of women older than 50 years. The youngest age group with a PDB less than 50% in Kim et al. [17] was 50-54 years, which shifted to the 60-64 age group after a 10-year period.

Our results show that the distribution of dense breasts by age group has changed over a ten-year period, with an increase in PDB observed for all age groups, except those aged 35-39 years. These results suggest that the different patterns in age at diagnosis of breast cancer between Asian and Western women may be related to the difference in PDB. It is also possible that changes in the PDB could be associated with the rising incidence of breast cancer.

Our study had three main limitations. The first issue is the possibility of self-selection bias, a type of selection bias. Because the study participants voluntarily participated in the screenings, they are not representative of all Korean women. However, the aim of this study was not to evaluate any benefit from the screening intervention, but rather to describe the distribution of dense breasts. Since the participants did not know whether they had dense breasts as the time of the screening, they are likely to be representative of women who would volunteer for a screening mammography study. The second issue is the reliability of mammographic density determination. PDB was determined using the BI-RADS four-category scoring system for mammographic density [19], not the BI-RADS five-category scoring system for recommendataion. By the way, Kerlikowske et al. [20] reported that the kappa value in breast density was highest (0.81) among BI-RADS parameters such as lesion finding, assessment category and recommendation. The last limitation is the comparability of our results with those in Kim et al. [17]. The source population in both studies was healthy women who underwent screening mammography at a medical institute, and both studies used the BI-RADS reporting system [19]. However, while comparable trends in PDB between the two studies can be discussed, a direct comparison of PDB is not possible due to the differences in the medical setting and the mammographic machine used.

The phenomenon of rapidly increasing incidence rates of breast cancer among Asian women could be explained by the combination of a diet rich in saturated fat and a sedentary lifestyle, as well as early menarche, decreased parity, and delayed childbearing, commonly referred to as the Westernization of lifestyle [21]. Shin et al. [11] and Sehn et al. [21] have argued that the changing trends in breast cancer incidence among Asian women are due to a birth cohort effect that corresponds to the adoption of a Western lifestyle. However, data published by Bae et al. showing incidence trends in Korean women over a ten-year period [4] cast doubt on the birth cohort effect hypothesis. In addition, Westernization of lifestyle has been associated with increased breast density [15]. Thus, an age-period-cohort analysis is needed to evaluate an association between changes of distribution of dense breasts and trends in incidence rates among Asian women.

Because PDB is higher among women from Asian countries than women from Western countries [17], the accuracy of screening mammography could be reduced, leading to higher rates of interval cancers and worsening prognosis [22,23]. If dense breasts are shown to be an independent risk factor for breast cancer in Asian women [24,25], women with dense breasts could be considered to be at high risk for breast cancer development and included in high-risk screening programs [19,26]. Further cohort studies incorporating follow-up screening mammography are therefore required to evaluate the relative risk of breast cancer among women whose breasts remain dense and women whose breasts develop adipose tissue over time [27].

In conclusion, this study suggests that over a ten-year period, PDB changes with age in Korean women. This suggests an association between the age distribution of women with dense breasts and trends in the incidence rate of breast cancer. Additional studies are needed to estimate the change in incidence rate by age group and the accumulation of breast fatty tissue in Korean women.