Racial/Ethnic Disparities in Prostate Cancer Incidence, Distant Stage Diagnosis, and Mortality by U.S. Census Region and Age Group, 2012–2015

Prostate cancer remains the most commonly diagnosed non-skin cancer and the second leading cause of cancer mortality among men in the United States (1). Prostate cancer incidence has been highest in the Northeast United States and among non-Hispanic black (NHB) men (2–7). An analysis of prostate cancer incidence by region and race from 1999 to 2008 showed that incidence rates for white men were highest in the Northeast census divisions of New England and Middle Atlantic while incidence rates for black men were highest in the Middle Atlantic and South Atlantic census divisions (2). NHB men also have the highest prostate cancer mortality rate nationally (7–9). Evidence from 2002 to 2011 indicates that prostate cancer mortality rates were higher in West and North Central states among non-Hispanic white (NHW) men and higher in the South among NHB men (10). Although regional differences in prostate cancer incidence and mortality have been examined for NHB and NHW men, they have not been as well characterized for other races/ethnicities.

Prostate cancer incidence peaked in 1992, then declined with an acceleration in the rate of decline from 2010 to 2014, as highlighted in the 2018 Annual Report to the Nation on the Status of Cancer (6). This accelerated rate of decline in recent years has been associated with changes to the United States Preventive Services Task Force (USPSTF) prostate-specific antigen (PSA) screening guidelines in 2012, which recommended against prostate cancer screening in all men (6, 11). Although overall prostate cancer incidence has been declining, an increase in distant stage prostate cancer has been reported from 2010 to 2014, with some suggesting it may also be a result of the changes to the PSA screening guidelines (6). Less is known about regional variations in prostate cancer incidence by race/ethnicity since the 2012 change in USPSTF guidelines. Examining rates by region using more recent data would help determine whether these patterns are occurring similarly in all regions and for all groups. Characterizing regional differences in the incidence of distant stage prostate cancer by race/ethnicity may help identify populations experiencing these increases.

In addition to the increased burden of prostate cancer incidence and mortality noted above, NHB men have a higher rate of distant stage prostate cancer diagnoses and a younger age of prostate cancer diagnosis than NHW men (6, 12, 13). Although the most recent guidelines suggest discussing risks, harms, and benefits of prostate cancer screening with NHB men, the USPSTF does not make separate screening recommendations for this higher risk population. Current USPSTF screening guidelines recommend that men between the ages of 55 and 69 years make an individual decision in conjunction with their health care provider regarding whether or not to be screened using the PSA test (14). An examination of the rates of distant stage prostate cancer diagnosis by age group may better inform screening decisions based on age.

This article examines regional variation in racial/ethnic differences in prostate cancer incidence and mortality from 2012 to 2015. Given recent findings of increasing distant stage prostate cancer diagnosis, we also characterize racial/ethnic differences in incidence of distant stage prostate cancer by region as well as by age group. Finally, because rates for racial/ethnic groups other than NHB and NHW men have been reported less frequently in the literature, this article characterizes these indicators for Hispanic, non-Hispanic Asian-American/Pacific Islander (NHAPI), and non-Hispanic American Indian/Alaska Native (NHAIAN) race/ethnicity as well.

We used incidence data from the United States Cancer Statistics (USCS) dataset, which is comprised of data from the Centers for Disease Control and Prevention's (CDC) National Program of Cancer Registries (NPCR) and National Cancer Institute's (NCI) Surveillance, Epidemiology, and End Results (SEER) Program (15). USCS includes cancer incidence data from 51 population-based central cancer registries in the United States. This report includes new cases of primary invasive prostate cancer diagnosed from 2012 to 2015. All statewide registries met USCS publication criteria for these years and were included in the analysis (15, 16). Thus for these years, cancer incidence data cover 100% of the United States population (15, 16). Prostate cancer mortality data were accessed through the SEER program's SEER*Stat software with underlying data provided by the CDC's National Center for Health Statistics (17).

Prostate cancer cases were defined using the International Classification of Diseases for Oncology, Third Edition topography code C61.9 (18). Cases identified only by autopsy or death certificate were excluded. We report prostate cancer incidence rates, rates of distant stage prostate cancer diagnoses, and prostate cancer mortality rates from 2012 to 2015. Rates (per 100,000 men) were reported by age group (all ages, <55 years, 55–69 years, and ≥70 years), race/ethnicity (NHW, NHB, Hispanic, NHAPI, and NHAIAN), year, and U.S. Census region (Northeast, Midwest, South, and West; ref. 16). In addition, we report prostate cancer incidence rates by SEER Summary stage (localized, regional, distant, or unstaged). Distant stage is defined by SEER as cancer that has metastasized to other organs or remote lymph nodes (19). We also examined race/ethnicity differences in rates of prostate cancer incidence, distant stage disease, and mortality by region and age group.

All analyses were performed in SEER*Stat, version 8.3.5 (20). Rates per 100,000 men were age-adjusted to the 2000 U.S. standard population with 19 age groups (census P25-1130; ref. 21). Population denominators are specific to race, ethnicity, and sex, and were obtained from SEER (22). Rate ratios (RR) and 95% confidence intervals (95% CI) were calculated using the Tiwari method to compare differences in rates by race/ethnicity (23, 24). Rate differences were considered statistically significant at P < 0.05. Data were suppressed for any cell with less than 16 cases because of poor reliability and to maintain confidentiality of subjects (18). Data for figures have been submitted as Supplementary Tables.

Table 1 presents age-adjusted rates of prostate cancer incidence, distant stage diagnosis, and mortality. From 2012 through 2015, 723,269 men were newly diagnosed with prostate cancer (age-adjusted rate 101.52 per 100,000 men). Of these cases, 65.6% occurred in men less than 70 years of age, although rates increased with age. Forty-four percent of distant stage diagnoses occurred in men less than 70 years of age. Compared with NHW men, rates for each indicator (incidence, distant stage diagnosis, and mortality) were higher for NHB men and lower for NHAPI men.

Incidence rates were highest in the Northeast for NHW, NHB, and Hispanic men (Fig. 1; Supplementary Table S1). For NHAIAN men, incidence was lower in the Northeast and West and higher in the Midwest and South, whereas for NHAPI men rates were similar across most regions. Compared with NHW men, rates for NHB men were higher across all regions. Hispanics had lower rates than NHW men in all regions except for the Northeast, where they had higher rates. NHAPI and NHAIAN men had lower rates in all regions compared with NHW men.

In general, rates of distant stage diagnosis were lower in the South for all races/ethnicities (Fig. 2; Supplementary Table S2). Despite lower overall incidence in most regions, rates of distant stage disease among Hispanic men were as high (Midwest and West) or higher (Northeast and South) than those of NHW men. Similarly, NHAIAN men had lower incidence rates compared with NHW men in all regions, but similar (Midwest and South) or higher (West) rates of distant stage diagnosis. NHB men had at least a twofold rate of distant stage diagnosis across all regions compared with NHW men.

Mortality rates in the West were among the highest for all races/ethnicities, except for NHB men, for whom rates were similar across regions (Fig. 3; Supplementary Table S3). NHB men had the highest rates in each region, with the largest difference compared with NHW men in the South and smallest in the West. Hispanics had significantly lower mortality rates across all regions except the Northeast, where the rate was similar to NHW men. The lowest mortality rate among Hispanic men was in the Midwest. Mortality rates among NHAPI men were significantly lower than NHW men in each region, whereas rates for NHAIAN men were not significantly different from NHW men for any region.

Table 2 presents age-adjusted rates (rates adjusted within each age-group to account for differences in age distribution by race/ethnicity) of prostate cancer incidence, distant stage diagnosis, and mortality by age and race/ethnicity. NHB men had the highest rates for each indicator across all age categories, whereas NHAPI men had the lowest rates for each indicator across all age categories. For distant stage diagnosis, large differences in rates were observed between NHB men and NHW men among ages <55 and between 55 and 69 years (RR 2.97 and RR 2.91, respectively). NHB men constituted 28.7% and 24.1% of men diagnosed with distant stage prostate cancer among those aged <55 years and aged 55 to 69 years, respectively. Hispanics had a similar rate of distant stage diagnosis compared with NHW men among men <55 years, but a higher rate among men aged 55 to 69 years, despite lower overall incidence in this age group. Prostate cancer mortality was lower for Hispanic men in this age group than for NHW men, however.

This analysis of USCS data and mortality data from 2012 to 2015 revealed regional and age-group differences in prostate cancer incidence, distant stage diagnosis, and mortality by race/ethnicity. For Hispanic and NHAIAN men, prostate cancer incidence, distant stage diagnosis, and mortality varied by region, while NHB and NHAPI men consistently had the highest and lowest rates, respectively, across regions. Importantly, our study builds on the growing body of evidence by providing region-specific data for Hispanic, NHAPI, and NHAIAN populations; highlights regional differences in mortality by race/ethnicity; and examines distant stage prostate cancer diagnoses by age and race/ethnicity. As previous data have shown regional differences in prostate cancer incidence and mortality between NHB and NHW men, our article also examines such regional differences since the 2012 change in USPSTF guidelines for other race/ethnicity groups.

NHB men had a greater rate than white men of incident prostate cancer, being diagnosed with advanced stage disease, and dying from prostate cancer across all regions. Multiple reasons for black–white disparities in prostate cancer incidence have been explored. For differences in incidence, some have suggested a possible role of genetic and hormonal factors, such as circulating PSA and testosterone concentrations (2, 7, 25–27). Environmental factors such as diet and gene–environment interactions have also been suggested (2, 28). Prostate cancer mortality disparities have been examined with differences in treatment based on socioeconomic factors implicated (9, 29).

In contrast, NHAPI men had the lowest rate of developing prostate cancer, advanced stage disease, and dying from prostate cancer in every region and age group examined. Previous studies have postulated a possible role of hormonal, genetic, and dietary factors in lower prostate cancer incidence rates in certain NHAPI groups (30–34). However, despite the favorable prostate cancer indicators, prostate cancer remains the most commonly diagnosed cancer among NHAPI men (35). Notably, the finding that NHAPI men had lower rates of distant stage diagnosis than NHW men differed from previous reports, although differences could be attributable in part to variations across studies, such as in subpopulations reported (36–38).

Overall, our results showing decreased prostate cancer incidence and mortality but higher distant stage rates in Hispanic men compared to NHW men are consistent with previous reports (7–9, 39). However, it is uncertain why Hispanic men had increased incidence of prostate cancer in the Northeast compared with NHW men, and why rates of distant stage disease were at least as high as NHW men for each region despite lower incidence for most regions. Although we did not examine this directly, one possible factor that may contribute to these findings is differences in subpopulations across regions. Previous evidence examining cancer mortality rates for 2008 to 2012 in Florida found that cancer mortality varied across Hispanic ethnic groups (40). Cancer screening use, including PSA testing, has also been reported to vary among these groups (41). Furthermore, Hispanic subgroups differ across geographic regions. For example, among Hispanic groups with a population of 1 million or more in 2010, most Dominicans and Puerto Ricans and 37% of South Americans resided in the Northeast, compared with fewer than 20% of other Hispanic subgroups (42). Dominicans have also been reported to have a higher percentage of black race than other Hispanic subgroups, at least in Florida (40). Increased participation in PSA screening results in higher disease detection rates. In many states, where estimates are available, recent PSA screening data show that Hispanics have lower PSA screening rates than NHW men overall (43). Further analysis and disaggregation of Hispanic subgroups in the Northeast and other regions will help to elucidate these findings further.

Our findings of decreased prostate cancer incidence but not mortality for NHAIAN men compared with NHW men are consistent with a previous study comparing their prostate cancer incidence and mortality rates with NHW men (44). Factors that have been suggested to contribute to this pattern include lower socioeconomic status, decreased access to care, and increased rate of exposure to prostate cancer mortality risk factors, such as cigarette smoking (44–46). Similar to Hispanics, NHAIAN men showed a pattern of lower incidence rates but similar or higher rates of distant stage diagnosis than NHW men in each region reported. Differences across regions could reflect, in part, differences in subpopulations (44). Whether differences in PSA screening may have contributed to these patterns is unclear. Lower rates of PSA testing have been reported for Hispanic and NHAIAN men than NHW men nationally (45–48). Findings of lower incidence but similar or higher distant stage disease may have implications for men and their healthcare providers as they discuss whether to be screened.

Although the Northeast had the highest overall incidence and distant stage diagnosis rates, it also had the lowest mortality rate, whereas the West had among the highest mortality rates despite having the lowest incidence rate. These findings raise questions about possible geographic differences in screening, access to treatment, and/or quality of care. Across all regions, mortality rates for NHB men were generally at least twice those of any other race/ethnicity group, consistent with previous evidence that rates are highest for NHB men (7–9). For Hispanic men, despite having similar or higher distant stage diagnosis rates in all regions, mortality rates were lower than NHW men in all regions except the Northeast where rates were similar. One factor possibly contributing to these inconsistencies between prostate cancer diagnoses and mortality could be that our cross-sectional study is not capturing the temporality of the relationship between prostate cancer diagnosis and mortality, with many men surviving for years after being diagnosed with prostate cancer (5). These regional differences in mortality, especially as they relate to prostate cancer incidence and distant stage diagnosis, provide interesting points for further study of possible differences in access to care and treatment patterns by region. Future monitoring of patterns will be important to determine whether observed differences in prostate cancer incidence and distant stage diagnosis by race/ethnicity and region lead to differences in mortality rates for these groups.

Overall, distant stage diagnosis rates were higher among NHB and Hispanic men than NHW men. By age, increases were more evident for NHB men <55 years and those 55 to 69 years old, although they had higher rates in all three age groups. A previous study concluded that prostate cancer may grow more aggressively in black than in white men, which could contribute to the increased distant stage diagnosis rate (49). In addition, Hispanic men had a higher rate of distant stage diagnosis compared with NHW men among men ages 55 to 69 years, despite lower overall incidence in this age group. The USPSTF recommends that men aged 55 to 69 years have discussions with medical providers regarding the risks and benefits of being screened. Such discussions may be informed by findings of higher rates of distant stage at diagnosis for some groups such as NHB and Hispanic men.

Strengths of this study include high-quality data from a large, nationwide dataset that enables examination of differences by race/ethnicity, region, and age. As stated previously, one limitation to this cross-sectional study is that we are not able to capture temporality between prostate cancer diagnosis and mortality. Another limitation to this study is that the dataset does not include information about PSA screening for race/ethnic groups. It is possible that differences in PSA screening rates underlie some of the disease indicator patterns presented. Previous evidence indicates that PSA screening rates among white and black men are relatively similar by region, although more recent data would be informative (2). We were unable to explore patterns among Hispanics as PSA screening data for Hispanic men are unavailable for a large number of states. Studies have also reported that Hispanic, NHAPI, and NHAIAN men have lower PSA screening rates than NHW men (45–48, 50). Also, counts for NHAPI/NHAIAN were smaller compared with other racial/ethnic groups so caution should be taken in interpretation of data for these populations. In addition, many AIAN self-identify as AIAN race and Hispanic ethnicity or multiracial subjecting them to potential misclassification (51). Finally, heterogeneity exists within race/ethnicity groups, and variation in rates within such groups may be masked (52).

Our findings demonstrate continued black–white disparities by geographic region in prostate cancer incidence, stage at diagnosis, and deaths while providing region-specific data for Hispanic, NHAPI, and NHAIAN men. These findings may inform conversations between men and their health care providers about PSA screening, by identifying not only men at increased risk for developing prostate cancer, but also those more likely to be diagnosed with distant stage disease or to die from it. Efforts to identify factors that drive differences in developing prostate cancer across regions and populations may prove useful to better understand the extent to which differences in screening and access to quality care may contribute to differences in outcomes. Furthermore, our findings highlight a role for additional disparities research, especially for groups relatively less studied such as Hispanic and NHAIAN men.

No potential conflicts of interest were disclosed.

The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

Conception and design: S. Iyengar, I.J. Hall, S.A. Sabatino

Development of methodology: S. Iyengar, S.A. Sabatino

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S. Iyengar

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S. Iyengar

Writing, review, and/or revision of the manuscript: S. Iyengar, I.J. Hall, S.A. Sabatino

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): S. Iyengar, I.J. Hall

The authors thank Jessica King for guidance in use of SEER*Stat and orientation to USCS data.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.