|Home | About | Journals | Submit | Contact Us | Français|
We estimated trends in breast cancer incidence rates for specific Asian populations in California to determine if disparities exist by immigrant status and age.
To calculate rates by ethnicity and immigrant status, we obtained data for 1998 through 2004 cancer diagnoses from the California Cancer Registry and imputed immigrant status from Social Security Numbers for the 26% of cases with missing birthplace information. Population estimates were obtained from the 1990 and 2000 US Censuses.
Breast cancer rates were higher among US- than among foreign-born Chinese (incidence rate ratio [IRR] = 1.84; 95% confidence interval [CI] = 1.72, 1.96) and Filipina women (IRR = 1.32; 95% CI=1.20, 1.44), but similar between US- and foreign-born Japanese women. US-born Chinese and Filipina women who were younger than 55 years had higher rates than did White women of the same age. Rates increased over time in most groups, as high as 4% per year among foreign-born Korean and US-born Filipina women. From 2000–2004, the rate among US-born Filipina women exceeded that of White women.
These findings challenge the notion that breast cancer rates are uniformly low across Asians and therefore suggest a need for increased awareness, targeted cancer control, and research to better understand underlying factors.
Breast cancer is the most common cancer diagnosed among US women.1 Nevertheless, in the United States, incidence rates vary substantially by standard racial/ethnic categories, with the lowest rates reported for Asians as a single group.2 However, within this group, breast cancer incidence rates vary further by specific ethnicity,2,3 with a nearly 3-fold difference between populations with the highest rate (Japanese women: 126 per 100000) and the lowest rate (Laotian women: 44 per 100000).2 Moreover, within each of the Asian populations, there is substantial heterogeneity in lifestyles, health care practices, and risk factors4–14 that may correspond to important differences in breast cancer rates.
Higher breast cancer incidence rates among Asian women living in the United States than among those living in Asian countries,15 together with elevated breast cancer risks associated with immigration to and longer residence in the United States,3,16,17 have been centerpiece evidence for major roles for environmental, nongenetic factors in breast cancer causation. Earlier research has shown that immigrant status (US-born vs foreign-born) is associated with breast cancer risk through changes in reproductive factors (e.g., higher age at first live birth, lower breast feeding rates, earlier onset of menstruation)18,19 and lifestyle factors (e.g., diet)3 but could also indicate variations in other environmental exposures.20
Examination of breast cancer incidence rates by immigrant status could inform these and other issues in breast cancer etiology. However, calculation of the accurate incidence rate trends by immigration status has been hampered by (1) nonrandom missing data on immigrant status for approximately 30% of cancer registry cases21–23 and (2) the lack of annual US Census estimates of Asian population counts according to immigrant status. To overcome these limitations, we augmented data on cancer cases from the population-based California Cancer Registry (CCR) through statistical imputation of immigrant status for cancer cases with missing birthplace and then used robust demographic methods to compute corresponding population estimates. With the resulting enhanced data resource, we sought to estimate breast cancer incidence rates by immigrant status, age, and time for specific Asian populations in California, the US state with the largest and most diverse Asian population.24
We obtained information from the CCR regarding all females residing in California who were diagnosed with a primary invasive breast cancer (International Classification of Disease for Oncology, 3rd Edition, [ICD-O-3] site codes C50.0–50.9) from January 1, 1988, through December 31, 2004. The CCR comprises 3 of the National Cancer Institute’s Surveillance, Epidemiology, End Results (SEER) program registries and collects information on patient race/ethnicity and birthplace from hospital medical records, which is obtained primarily through self-report; by assumption of hospital personnel; on inference from other information including race/ethnicity of parents, maiden name, surname, and birthplace; and from death records.25
In this study, we included 21147 breast cancer cases among the 6 Asian ethnic populations that together comprised 91% of all Asian patients with breast cancer in the CCR during the study period. Of these, 5732 (27.1%) cases were among Chinese women, 3888 (18.4%) were among Japanese women, 7583 (35.9%) were among Filipina women, 1304 (6.2%) were among Korean women, 1130 (5.3%) were among South Asian women (including Asian Indians, Pakistanis, Sri Lankans, and Bangladeshis), and 1510 (7.1%) were among Vietnamese women. We excluded from further consideration cancer patients classified by the registry as “ Other Asian or Asian, not otherwise specified” (n = 1099).
Our prior research showed that patients in the cancer registry with unknown birthplace data were more likely to be US-born than those with available data.21–23 Thus, we expect that rates based on random imputation of immigrant status would be underestimated for the US born and overestimated for the foreign born. To more accurately classify patient immigrant status, we used 2 sources of information about birthplace: (1) registry-based data available for15387 patients, or 72.8% of cases (69.7% from hospital medical records and 3.1% from death certificates) and (2) statistical imputation of immigrant status using the first 5 digits of the patient’s Social Security Number (SSN), indicating the state and year of issuance for those who had unknown birthplace (n = 5409, or 25.6% of patients).26,27 With this approach, we imputed immigrant status as follows: individuals who received their SSN before age 25 years were considered US born, and those who received their SSN at or after age 25 years were considered foreign born. The age cut-point was set at 25 years based on self-reported nativity in previously interviewed cancer patients (n = 1836)22 and maximization of the area under the receiver-operating characteristic curve. The optimal positive predictive value of this single age cut-point was confirmed by using a logistic regression model with age at SSN issue as a continuous predictor of foreign-born status. This age cut-point resulted in immigrant status classifications associated with 84% sensitivity and 80% specificity for detecting foreign-born (i.e., 80% sensitivity for detecting US-born individuals), and was similar across Asian populations. The 351 individuals (1.7%) with missing or invalid SSNs were assigned an immigrant status based on the ethnicity–sex–age nativity distribution of the overall sample.
From the 1990 and 2000 Census Summary File 3 (SF-3),28,29 we obtained population counts by gender, race/ethnicity, immigrant status, and 5-year age group for the state of California. Data from the 20% Integrated Public-Use Microdata Sample of the Census also were used to estimate age- and immigrant status–specific population counts for the 6 Asian groups30–32 by smoothing the distribution with a spline-based function.33 For intercensal years, we estimated the percentage of foreign-born individuals by using cohort component interpolation and extrapolation methods, adjusting estimates to the populations by age and year provided by the California Department of Finance for years1988 to1989 and by the US Census for years1990 to 2004 because of data availability. Incidence rates for US-born Koreans, South Asians, and Vietnamese were not stable enough to report because of 2 consequences of the recent immigration of these groups. First, the US-born populations were significantly smaller than the foreign-born populations, and, second, US-born populations have considerably younger age distributions, reducing stability of age-adjusted rates for cancers, which predominantly occur at older ages.
We used SEER*Stat software34 to compute age-adjusted incidence rates (standardized to the US standard million population for 2000) and 95% confidence intervals (CIs) among women. We compared incidence rates between US- and foreign-born Asian populations with incidence rate ratios (IRRs). We examined time trends between 1988 and 2004 using Joinpoint Regression software35 to calculate the annual percentage changes (APC). We calculated APCs by fitting a series of least squares regression lines to the natural logarithm of the age-adjusted incidence rates, with calendar year as a regression variable.35
During the period 1988–2004, invasive breast cancer incidence rates were lower for all of the Asian ethnic groups than for non-Hispanic White women (Table 1). Table 1, which compares rates across ethnicity and immigrant status, shows that US-born Chinese and Filipina women had cancer rates that were approximately 80% and 30% higher, respectively, than were rates for their foreign-born counterparts. By contrast, rates were comparable for US-born and foreign-born Japanese women (106 and 103 per 100000, respectively). Incidence rates were significantly higher for US-born Chinese (122 per 100000; 95% CI = 115, 129) and Filipina women (130 per 100000; 95% CI = 119, 141) than for US-born Japanese women.
We observed an increase in incidence rates across time periods in both the US-born and foreign-born aggregated Asian populations, although the increase appeared to be greater in the US-born Asian populations (Table 2, Figure 1). US-born Filipina and foreign-born Korean women had the largest increases in breast cancer incidence (4% yearly) over time. The annual increase in incidence rates was statistically significant for all groups except foreign-born Vietnamese and Japanese women. Rates in foreign-born Japanese women increased steadily from1988 to 1999 before dropping substantially in 2001 to 2004. Across all time periods, the highest breast cancer incidence rate was observed for US-born Filipina women diagnosed from 2000 to 2004 (162.5 per 100000; 95% CI = 144.3, 182.2).
In age-specific data (Table 3), rates were higher among US-born than foreign-born groups across all age groups examined, except among Japanese women. Notably, incidence rates for US-born Chinese and Filipina women aged 44 years or older and those aged 45 to 54 years at diagnosis were higher than were corresponding rates for non-Hispanic White women, whereas incidence rates for women aged 55 years and older were considerably lower.
Using new, rigorously derived cancer and population data estimates for women by immigrant status, we identified subpopulations of US Asian women who appear to have higher burdens of breast cancer than were previously described. Specifically, we found that US-born Asian women have incidence rates of invasive breast cancer nearly 2-fold higher than do foreign-born Asian women in all groups examined except Japanese women. These findings suggested that breast cancer rates reported for specific US Asian populations are sensitive to the proportion of foreign-born women. Our results also suggest that rates for US-born Chinese, Japanese, and Filipina women approached and, among Filipina women exceeded, those for non-Hispanic White women; this pattern was consistent with a strong environmental influence on breast cancer risk. Our findings also suggested that US-born Asian women younger than 55 years in particular have higher risks of breast cancer than previously estimated. To our knowledge, our study is the first to suggest that rates among pre- and peri-menopausal US-born Chinese and Filipina women were higher than were those for non-Hispanic White women.
These relatively higher rates in Asian populations together with their younger age distributions lead to a considerably greater proportion of breast cancer cases diagnosed among young women than occurs in other racial/ethnic groups. For example, from 1988–2004, the percentage of breast cancers diagnosed among women younger than 55 years was 67.5% among US-born Filipina women but only 28.9% among non-Hispanic White women. Furthermore, age-specific incidence patterns among US- and foreign-born Asian women differed from patterns seen in non-Hispanic White women, with rates differing very little between women 55 years and older and those aged 45–54 years. By contrast, among non-Hispanic White women, incidence rates were nearly 2-fold higher in the 55 years and older age group. In this regard, the age-specific incidence patterns we estimated for US-born Chinese and Filipinas in particular were noteworthy in light of the age-specific incidence for African American women, which have been shown to be higher than that for non-Hispanic Whites among women diagnosed who were younger 40 years, but “crossing over” to be relatively lower for women diagnosed at older ages.36,37 Although our1988–2004 data show that the rates for women younger than 40 years were indeed higher for African American (15.1 per 100000) than they were for non-Hispanic White women (12.8 per 100000), the rates for US-born Chinese and Filipina women were even higher (20.3 and19.8 per 100000, respectively). Thus, the observed ethnic and immigrant differences in the age-specific patterns of breast cancer incidence relative to menopause may lend insights into breast cancer etiology among Asian women, particularly as these age-specific patterns relate to changes in acculturation that may be occurring during this critical time period.
Previous studies have reported comparable breast cancer incidence rates for US-born Japanese and non-Hispanic White women,38,39 despite the lower prevalence among Japanese women of particular breast cancer risk factors (e.g., lower weight or body mass index and less alcohol consumption).40 We identified increasing rates of breast cancer for US Japanese women, regardless of immigrant status, during the period 1988–2000, mirroring the rising rates of breast cancer in Japan,41 which42 nearly doubled between 1978 and 1998.41,43,44 Women in Japan have become increasingly similar to US White women with respect to the population prevalence of lifestyle-related breast cancer risk factors. After World War II, the number of women in the Japanese workforce has risen,45 which has produced a declining birthrate.46,47 This societal shift may have resulted in relevant changes in the population distribution of breast cancer risk factors, such as higher socioeconomic status,48 higher body mass index, fewer births, and later age at first birth.42,49 Consequently, breast cancer risk profiles may be more similar for foreign- and US-born Japanese women than for other Asian populations, causing a narrowing in the incidence rate gap between Japanese immigrants to the United States and those who are US-born. In addition, it is likely that the foreign-born Japanese immigrated earlier in life, and thus may be more acculturated, than other foreign-born Asians, based on the younger age of SSN issuance among Japanese observed in our data.
Some of our results differ from those reported in the only prior examination of breast cancer incidence rates by immigrant status.17 In an analysis based on 1973 to 1986 SEER data from Seattle, Washington, and San Francisco, Oakland, and Los Angeles, California, US-born Chinese and Japanese women had 50% higher incidence rates than did their foreign-born counterparts, whereas rates among US-born Filipinas were slightly lower than those among foreign-born Filipinas.17 The difference in main findings may be because of the choice in this prior study to randomly distribute immigrant status among the 13% to 22% of cancer cases with unknown registry birthplace data, which resulted in underestimated rates among US-born and overestimated rates among foreign-born women.22 Second, whereas the 1973–1986 rates were not directly comparable to ours given different populations used for age standardization, the rates among US-born Filipinas were substantially lower in the earlier analysis17 than they were in our study. The higher rates observed in our study may suggest a large increase in incidence rates in US-born Filipinas, as indicated by the 4% increase per year during the period 1988–2004, as well as immigrant cohort differences in risk factor profiles.
Our study had several possible limitations. Our imputation of immigrant status based on SSN, although an improvement over prior methods,50 had an 84% sensitivity and 80% specificity of classification. The remaining misclassification had a greater relative impact on the smaller immigrant group, usually US-born. The assignment of immigrant status in some groups is fairly sensitive to the SSN age cutoff for imputing immigrant status; for example, using an age cutoff of 20 years instead of 25 years would have yielded 21%, 8%, and 34% fewer US-born Chinese, Japanese, and Filipina breast cancer cases but would have been subject to higher misclassification rates than the cutoff we used. However, the primary function of immigrant status in our study was to indicate acculturation, with those characterized as US-born considered to be more acculturated than those characterized as foreign-born. Even if imputed immigrant status was not completely accurate, it was likely to be correlated with acculturation level.
Our population estimates for Asian populations by immigrant status may also have been subject to error, particularly for specific age groups, which may have biased overall or age-specific rates. Although the Census data we used were the most definitive for estimating populations, counts stratified by immigrant status were available only for a population sample, which reduced their robustness. To evaluate the accuracy of the assumptions underlying our methods for estimating annual populations, we compared our 2004 population estimates to those from the 2005 American Community Survey (ACS), a 2.3% stratified sample of the population, 51 for California. We found less than a 2% difference in the estimates for US- and foreign-born Filipinas, suggesting that our findings for Filipinas were unlikely to be biased by inaccuracies in the denominator data. For Chinese women, our population estimates were higher than the ACS estimates by about 8% for US-born Chinese and about 1.5% for foreign-born Chinese; however, if the ACS estimates were more accurate, the bias was conservative, as we would have underestimated the relative difference in rates between US- and foreign-born Chinese women. Our estimates for US- and foreign-born Japanese women were both 15% higher than the ACS estimates; however, even if the ACS data were more accurate, this difference would be unlikely to explain our finding of a lack of difference in breast cancer rates between US- and foreign-born Japanese women, given that the direction and degree of difference were similar.
Findings from our study support the notion that immigrant status is an important variable for assessing patterns and trends in breast cancer incidence in dynamic populations, like that of California, with significant immigration. It is likely that this measure of acculturation reflects population distributions of nongenetic breast cancer risk factors such as diet, age at menarche, age at first birth, breastfeeding, and other lifestyle and reproductive factors. Future studies should examine the prevalence of established risk factors by immigrant status and socioeconomic status among US Asian populations to understand further how these and other environmental factors explain observed differences in breast cancer incidence rates between US- and foreign-born Asian women.52 Of particular interest to prevention among US-born Asian women of postmenopausal age will be the prevalence of use of combined hormone therapy and 2 or more alcoholic beverages per day. Further exploration of recent age-specific changes in known and suspected breast cancer risk factors may provide valuable clues to as-yet undiscovered causes of breast cancer (e.g. adolescent diet).
Our findings challenge the current notion based on aggregated data that breast cancer rates are uniformly low among US Asian women by identifying apparent disparities in this diverse population group.53 Specifically, using novel and robust estimating methods to characterize immigration status, we found that breast cancer rates were high and increasing among US-born Chinese and Filipina women, to a point where contemporary rates may equal or exceed those of non-Hispanic Whites. We also estimated higher rates among younger US-born Chinese and Filipina women than among non-Hispanic White women of comparable age, and a sustained increase in breast cancer incidence rates over time, particularly among US-born Filipina and foreign-born Korean women. Together, our findings demonstrate the need to consider cancer control and policy efforts targeted specifically to these groups, especially among US-born Asian women, and to educate clinicians and women as to the possibility of these previously undocumented risks, especially because the number of US-born Asians is projected to rise in the future. Our data suggest that, within this generation, some groups of Asians are experiencing unprecedentedly high rates of breast cancer. This situation must be further evaluated but also addressed now so that ongoing and future efforts to reduce the breast cancer burden do not overlook Asian women, particularly those becoming more acculturated to the Western lifestyle. Work targeted to these groups should include concerted efforts, such as those with Asian community organizations, to educate young Asian women about lifestyle modifications that can help them reduce their risk of developing breast cancer in the future.
Health disparities and inequities experienced by Asian Americans largely have been underdescribed because of stereotypes about positive health profiles, together with the lack of detailed data about the heterogeneous subpopulations. Examining data enhanced to enable study of cancer incidence in US immigrant groups is an important step toward building an evidence base that can inform health policies for US Asian populations.
This study was supported by the Safeway Foundation and a Surveillance, Epidemiology and End Results Rapid Response Surveillance Study contract (N01-PC-35136 and N01-PC-35139). The collection of cancer incidence data used in this study was supported by the California Department of Health Services as part of the statewide cancer reporting program mandated by California Health and Safety Code Section 103885; the National Cancer Institute’s Surveillance, Epidemiology and End Results Program (contract N01-PC-35136 awarded to the Northern California Cancer Center, contract N01-PC-35139 awarded to the University of Southern California, and contract N02-PC-15105 awarded to the Public Health Institute); and the Centers for Disease Control and Prevention’s National Program of Cancer Registries grant (agreement U55/CCR921930-02) awarded to the Public Health Institute.
The authors thank Tim Miller, Rita Leung, Sarah Shema, June Kristine Winters, and Laura McClure for their help with this article.
ContributorsS. L. Gomez supervised the collection of data, originated the study, and supervised all aspects of its implementation. T. Quach conducted the data analysis and led the writing. J. T. Pham and M. Cockburn assisted with the study. All authors helped to conceptualize ideas, interpret findings, and review drafts of the article. All authors certify that they have contributed substantially to the article and approve the final version.
Note. The ideas and opinions expressed herein are those of the author and endorsement by the State of California, Department of Health Services, the National Cancer Institute, and the Centers for Disease Control and Prevention or their contractors and subcontractors is not intended nor should be inferred.
Human Participant Protection
The Northern California Cancer Center institutional review board approved this study.
Scarlett Lin Gomez, Northern California Cancer Center, Fremont, CA.
Thu Quach, Northern California Cancer Center, Fremont, CA.
Pamela L. Horn-Ross, Northern California Cancer Center, Fremont, CA.
Jane T. Pham, Northern California Cancer Center, Fremont, CA.
Myles Cockburn, Keck School of Medicine, University of Southern California, CA.
Ellen T. Chang, Northern California Cancer Center, Fremont, CA.
Theresa H.M. Keegan, Northern California Cancer Center, Fremont, CA.
Sally L. Glaser, Northern California Cancer Center, Fremont, CA.
Christina A. Clarke, Northern California Cancer Center, Fremont, CA.