Colorectal Cancer Incidence Trends by Age, Stage, and Racial/Ethnic Group in California, 1990–2014

This article is featured in Highlights of This Issue, p. 983

In the United States, colorectal cancer is the third most common malignancy and the second and third leading cause of cancer-related death among men and women, respectively (1). Over the past 20 years, there has been a substantial decline in the incidence of colorectal cancer, largely attributable to the introduction of screening for adults aged between 50 and 75 years and a reduction in modifiable risk factors such as smoking (2). However, recent evidence suggests that the incidence of early-onset colorectal cancer (i.e., in adults aged less than 50 years) is rising (3–6), prompting debate about the most effective starting age for colorectal cancer screening, and a call for investigation into the risks and benefits of screening before age 50 (6–10).

In the United States, colorectal cancer disproportionately affects the non-Hispanic (NH) black population, who consistently experience one of the highest incidence rates of all racial/ethnic groups. Over the period 2009 to 2013, the incidence of colorectal cancer among NH black men was 58 per 100,000 compared with 46 among NH white men, 43 among Hispanic men, and 38 among Asian American and Pacific Islander (AAPI) men (11). Although AAPI populations combined have a comparatively low incidence of colorectal cancer, ethnicities within this classification have very different cancer profiles. In 2004 to 2008, the incidence of colorectal cancer among Japanese men was 67 per 100,000, compared with 23 among Indian and Pakistani Asians (12). There is also evidence to suggest that colorectal cancer incidence rates in these populations are increasing, both in young adults (6, 13) and among those eligible for screening (12, 14).

California has the largest and most diverse racial/ethnic population in the United States: in 2016, the population was 39% Hispanic, 38% NH white, 15% AAPI, and 6% NH black. Using data from the California Cancer Registry (CCR), we examined trends in incidence of colorectal cancer among detailed racial/ethnic groups by age and stage over the 25-year period of 1990 to 2014, with a specific focus on young adults.

The CCR (www.ccrcal.org), a statewide population-based cancer surveillance system comprising three SEER registries, has collected detailed information on patients diagnosed with cancer in California since 1988. Data from the CCR were obtained on all cases of colorectal cancer diagnosed in adults aged 20 years and older from 1990 through 2014 (N = 349,176). Age group at diagnosis was defined as 20 to 49 years, 50 to 74 years (screening age), and 75 years or older. In supplementary analyses, tumor subsite was stratified as right colon (C18.0-C18.4), left colon (C18.5-C18.7), and rectum (C19.9, C20.9).

Stage at diagnosis was based on the Surveillance Epidemiology and End Results (SEER) Summary Stage (15) and categorized as: early stage (in situ and localized tumors) and late stage (regional and distant tumors). Race/ethnicity was obtained from medical records and categorized as: NH white, NH black, Hispanic, Chinese, Japanese, Filipino, Korean, South Asian (Indian and Pakistani), Vietnamese, and other Southeast Asian (Thai, Hmong, Cambodian, Laotian). Hispanic and Asian ethnicity were additionally coded based on birthplace and surnames using the validated North American Association of Central Cancer Registries (NAACCR) Hispanic Identification Algorithm (16) and Asian Pacific Islander Identification Algorithm (17). For Asian American groups, trends in colorectal cancer incidence in California were compared with national trends for the same time period (1990–2014) using specialized SEER incidence data for detailed Asian/Pacific Islander groups in SEER 9 registries (18).

Annual cancer incidence rates (per 100,000 population) for each strata defined by sex, age, race/ethnicity, and stage at diagnosis were calculated and age-adjusted to the 2000 U.S. standard population using the SEER*Stat software (version 8.3.4; ref. 19). A minimum of 5 incident cases per strata were required for incidence rate calculations. Annual population counts were estimated using linear interpolation and extrapolation of 1990, 2000, and 2010 Census counts. Incidence rates by sex, age, race/ethnicity, and stage are presented for the most recent 5-year period (2010–2014), and incidence rate ratios (IRR) reported for each racial/ethnic group compared with NH whites. Trends in incidence over the 25-year period of 1990 to 2014 are presented by sex, age, race/ethnicity, and stage. Joinpoint regression models (20) were used to characterize the magnitude and direction of incidence trends. The simplest Joinpoint models were fitted (zero joinpoints, linear trend), with one annual percentage rate change reported for the entire period (21). To accommodate for the small number of cases in some strata, 3-year cumulative incidence rates were used to improve the stability of annual percent change statistics.

Of the 68,884 incident cases of colorectal cancer diagnosed in California in the period 2010 to 2014, 10.9% (n = 7,572) occurred in adults aged less than 50 years, but the proportion varied substantially by racial/ethnic group. Approximately 21% of South Asians and 20% of Southeast Asians were younger than 50 years at diagnosis, compared with 8.6% of NH whites and Japanese. Patients younger than 50 years were more likely to present with late-stage disease: approximately 66% of early-onset colorectal cancers were classified as late-stage (regional or distant disease) at diagnosis, compared with 54% of cases in adults aged 50 years and older.

The incidence of colorectal cancer varied dramatically across age and racial/ethnic groups (Table 1). In the 20- to 49-year age group, there was no difference in incidence between NH blacks and NH whites for either early- or late-stage disease (Table 2). In Hispanics, the incidence of early-onset colorectal cancer (early and late stage) was 20% to 30% lower than in NH whites, and in several Asian American groups (Korean and South Asian men and women, and Chinese women), the incidence of late-stage colorectal cancer was 30% to 60% lower.

In the screening age group (50–74 years), the high incidence of colorectal cancer in NH blacks was particularly evident. Compared with NH whites, NH black men had 50% higher incidence of early-stage colorectal cancer [IRR, 1.5; 95% confidence interval (CI) 1.4–1.6] and 40% higher incidence of late-stage disease (IRR, 1.4; 95% CI, 1.3–1.5), whereas NH black women had 60% higher incidence of early-stage (IRR, 1.6; 95% CI, 1.4–1.7) and 40% higher incidence of late-stage colorectal cancer (IRR, 1.4; 95% CI, 1.3–1.5).

Over the period 1990 to 2014, the incidence of early-onset colorectal cancer increased in nearly all racial/ethnic groups in California, and was most pronounced for regional/distant stage. Increases were statistically significant in NH white men (by +0.9% every 3 years for early stage; +1.9% for late stage) and Hispanic men (+1.6% for early stage; +2.2% for late stage; Table 3; Figs. 1A and B). Similarly, the incidence of early-onset colorectal cancer in women increased significantly in NH whites (by +1.9% every 3 years for early stage; +2.4% for late stage), Hispanics (+2.1% for early stage; +0.7% for late stage), and Southeast Asians (+2.7% for late stage; Table 3; Fig. 1C and D). At the national level, as represented by SEER data, the incidence of early-onset colorectal cancer increased significantly in South Asian, Filipino, Vietnamese, and other Southeast Asian men (Supplementary Fig. S1A and S1B), and South Asian women (Supplementary Fig. S1C and S1D). There were statistically significant increases in incidence for all subsites (left colon, right colon, and rectum), with the largest increases seen for left colon and rectal cancer in NH whites and Hispanics (Supplementary Table S1).

There were large declines in the incidence of colorectal cancer in adults aged 50 and older in California over the period 1990 to 2014. For screening-aged adults, significant decreases in the incidence of early-stage colorectal cancer were seen for NH white men (by −1.9% every 3 years), NH white women (−1.3%), and Japanese men (−2.2%; Table 3; Supplementary Fig. S2A and S2C). More widespread declines were found for late-stage colorectal cancer, with significant decreases for NH white, NH black, Hispanic, Chinese, and Japanese men and women, by approximately −1% to −3% every 3 years. Declines in incidence in adults aged 50 and older were most notable for left colon and rectal cancers (Supplementary Table S1). In contrast, the incidence of early-stage colorectal cancer increased in Vietnamese women by +1.5% every 3 years, and the incidence of late-stage colorectal cancer increased in Southeast Asian men by +4.1% every 3 years, and in Southeast Asian women by +2.7% every 3 years (Table 3; Supplementary Fig. S2B–S2D). The increasing trend in colorectal cancer incidence for Vietnamese and other Southeast Asian women of screening age was also seen at a national level (Supplementary Fig. S3C and S3D), along with an increasing trend for Vietnamese and South Asian men (Supplementary Fig. S3A and S3B).

We leveraged high-quality contemporary data from the CCR to investigate trends in the incidence of colorectal cancer over a 25-year period (1990–2014) by race/ethnicity, including seven Asian American ethnic groups, with attention to examining trends among young adults. In California, the most populous and racially/ethnically diverse state in the United States, we found increases in the incidence of advanced stage early-onset colorectal cancer among nearly all racial/ethnic groups, with statistically significant increases for NH white and Hispanic adults. Although the incidence of colorectal cancer among adults of screening age has declined in most racial/ethnic groups, we observed increasing incidence among Vietnamese and other Southeast Asian groups (Thai, Hmong, Cambodian, and Laotian). The high incidence of colorectal cancer in screening-age and older NH blacks persisted over this 25-year period, although there is evidence of a decline in these age groups.

Reports of rising colorectal cancer incidence among young adults in the United States, as well as among vulnerable populations such as immigrants (22), are mounting. In our study, statistically significant increases in the incidence of early-onset colorectal cancer were limited to NH white and Hispanic populations, the two largest racial/ethnic groups in California. Although the number of colorectal cancer cases in young adults remains small compared with older adults, long-term projections suggest that by 2030, 11% of colon cancers and 23% of rectal cancers in the United States will occur in individuals younger than 50 years (11). The rising incidence of early-onset colorectal cancer does not appear to be unique to the United States (23), with similar patterns reported in Canada (24), Australia (25, 26), and Asia (27, 28). The underlying causes of the increasing incidence in this age group are unclear, although poor diet, sedentary lifestyles, and increasing rates of obesity are likely to be contributing (29).

There is robust evidence that effective screening strategies can reduce the incidence of and mortality from colorectal cancer through early diagnosis and removal of precancerous lesions (adenomatous polyps; refs. 30–32), and eliminate racial/ethnic disparities (33). The U.S. Preventive Services Task Force (34) currently recommends that average-risk adults aged 50 to 75 years have either a high-sensitivity fecal occult blood test, fecal immunochemical test (FIT), or multitargeted stool DNA test (FIT-DNA) every year; flexible sigmoidoscopy every 5 years, or every 10 years with FIT every year; or colonoscopy every 10 years. Except for those at increased risk of the disease (e.g., those with inflammatory bowel disease, history of familial adenomatous polyposis or hereditary nonpolyposis, a previous adenomatous polyp, or a family history of colorectal cancer), adults under 50 are not included in these guidelines. Some reports suggest that screening initiated at age 45 years is cost effective (35), and could achieve a more favorable balance between benefit (life-years gained) and burden (36), but the debate on the risks and benefits of screening before age 50 for asymptomatic average-risk adults continues.

For adults aged 50 to 75 years, colorectal cancer screening recommendations have been in place in some form since the mid-1990s, but uptake has been lower than expected (22, 37) and varies considerably by race/ethnicity (22, 37–38). In 2008, only 55% of the U.S. population was up-to-date with colorectal cancer screening (39). Although there has been some improvement over time, a report by the U.S. Centers for Diseases Control and Prevention (CDC) found that by 2015, only 62% of screening eligible adults were up-to-date with colorectal cancer screening, still short of the Healthy People 2020 target of 70.5% (40).

Yet, we found widespread declines in the incidence of colorectal cancer among screening-aged adults in most racial/ethnic groups in California commensurate with improved screening uptake over time. Two notable exceptions were in screening-aged Vietnamese and other Southeast Asian men and women, for whom there has been an increase in colorectal cancer incidence. In non-Vietnamese Southeast Asians, the increase of 3% to 4% every 3 years is striking; by 2000 to 2014, the incidence of late-stage colorectal cancer in men was higher than in NH black men. The reasons behind this increase are unclear, and could be partially explained by a low uptake of screening (39), in addition to increases in colorectal cancer risk factors including obesity. Klabunde and colleagues identified a number of population groups with consistently low uptake of colorectal cancer screening, such as recent immigrants (those who have been in the United States for less than 10 years), and those with poor education, low income, and lack of health insurance (22), but the relevance of these factors to Southeast Asian populations is difficult to assess, due to the lack of information specific to this Asian American ethnic group.

The incidence of colorectal cancer in NH blacks aged 50 and over was consistently high throughout the study period. In addition to a high prevalence of lifestyle-related risk factors, it has been postulated that a low uptake of screening in this population could be the underlying cause of this trend, although data from the California Health Interview Survey (41) and CDC Behavioral Risk Factor Surveillance System (42) report equivalent screening rates for NH black and NH white populations (78% and 72%, respectively, in 2014; ref. 42). Although our finding of declining incidence in NH blacks of screening age is a positive one, the faster rate of decline in NH whites means black/white disparities persist. In 2017, in response to the higher incidence and younger mean age of colorectal cancer onset in African Americans, the U.S. Multi-Society Task Force on Colorectal Cancer recommended beginning screening at age 45 in this population (43). It is worth noting that in our study, the incidence of early-onset colorectal cancer in 2010 to 2014 was similar for NH black and NH white populations, and even higher for some Asian American ethic groups.

Our study has several limitations that are inherent to cancer registry data. We did not have access to important information on family history of colorectal cancer or other lifestyle-related risk factors (obesity, smoking, diet, physical activity), clinical data such as microsatellite instability, or genetic data (KRAS mutation). We also lacked data on Hispanic country of origin, nativity, and immigration history. It has been suggested that the acculturation of recent immigrants may result in a higher prevalence of risk factors for colorectal cancer, including obesity, lack of physical activity, smoking, and alcohol consumption (44, 45). We found increases in the incidence of early-onset colorectal cancer in most racial/ethnic groups in California, but due to the small number of cases in some strata, especially for the AAPI groups, increases were only statistically significant for NH white and Hispanic groups. To assess whether the incidence of colorectal cancer is increasing in AAPI groups at the national level, we examined SEER incidence data for AAPI groups over the period 1990 to 2014 (18), and found statistically significant increases in colorectal cancer incidence for South Asian men and women, and Filipino, Vietnamese, and other Southeast Asian men aged under 50, and non-significant increases for Japanese, Chinese, and Koreans.

In summary, our results emphasize the need to investigate the causes of the rising incidence rates of colorectal cancer among young adults and among Southeast Asians of screening age, as well as the persistently high incidence among screening age and older NH blacks. The heterogeneity of incidence rates across Asian American ethnic groups reinforces the need to disaggregate cancer statistics for this population. Improvements in colorectal cancer awareness among patients and physicians, the promotion and availability of recommended screening test, and the early initiation of screening for high-risk populations are crucial to decrease the burden of colorectal cancer.

No potential conflicts of interest were disclosed.

Conception and design: J. Yang, M. Somsouk, L. Le Marchand, S.L. Gomez, S. Shariff-Marco

Development of methodology: J. Yang, S. Shariff-Marco

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S.L. Gomez, S. Shariff-Marco

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): L. Ellis, M. McKinley, J. Yang, M. Somsouk, I. Cheng, S. Shariff-Marco, R. Abrahão

Writing, review, and/or revision of the manuscript: L. Ellis, M. McKinley, J. Yang, M. Somsouk, I. Cheng, S. Shariff-Marco, R. Abrahão

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M. McKinley, S. Shariff-Marco

Study supervision: S.L. Gomez, S. Shariff-Marco

The collection of cancer incidence data used in this study was supported by the California Department of Public Health, the Centers for Disease Control and Prevention's National Program of Cancer Registries, and the NCI's Surveillance, Epidemiology and End Results Program. The ideas and opinions expressed herein are those of the author(s) and do not necessarily reflect the opinions of the State of California Department of Public Health, the NCI, and the Centers for Disease Control and Prevention or their Contractors and Subcontractors. This work was supported by the NCI's Surveillance, Epidemiology and End Results Program (contract HHSN261201000140C awarded to the Cancer Prevention Institute of California); the NCI-funded Stanford Cancer Institute (P30 CA124435 awarded to Stanford University); and the Center for Disease Control and Prevention [Prevention Research Center Cooperative Agreement Number U48DP004998 (SIP 14–013) awarded to the University of California, San Francisco]. The funders had no role in the design or conduct of the study; the collection, management, analysis, or interpretation of the data; the preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.

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.