Colorectal cancer screening test exposure patterns in US adults 45 to 49 years of age, 2019-2021

Abstract

Several organizations now recommend that individuals at average risk for colorectal cancer (CRC) begin screening at 45 rather than 50 years of age. We present contemporary estimates of CRC screening in newly eligible adults aged 45 to 49 years between 2019 and 2021. Nationally representative prevalence estimates and population number screened were estimated based on the National Health Interview Survey. A logistic regression model assessed CRC screening prevalence differences by survey year and sociodemographic characteristics. In 2021, 19.7%—that is, fewer than 4 million of the eligible 19 million adults aged 45 to 49 years—were up-to-date on CRC screening. Screening was lowest in those who were uninsured (7.6%), had less than a high school diploma (15.4%), and Asian (13.1%). Additionally, fecal occult blood test and/or fecal immunochemical testing was underused, with only 2.4% (<460 000 people) reporting being up-to-date with screening using this modality in 2021. CRC screening in eligible young adults remains low. Concerted efforts to improve screening are warranted, particularly in underserved populations.

In 2018, the American Cancer Society (ACS) lowered the recommended age for individuals at average risk for colorectal cancer (CRC) to begin screening from 50 to 45 years based on evidence of rising CRC risk at younger ages and simulation modeling showing that the benefits of earlier screening outweighed the harms (1-3). Studies published since the ACS guideline update have reported reductions in CRC incidence (4,5) as well as the cost-effectiveness associated with earlier screening (6). Additionally, recommending screening in those aged 45 to 49 years has modestly affected colonoscopy screening volume, while lesion detection rates, including for advanced adenomas, in those first-time screeners at average risk and aged 45 to 49 years were comparable to the rates found in first-time screeners at average risk and aged 50 to 54 years (7). In 2021, the United States Preventive Services Task Force (USPSTF) and the US Multi-Society Task Force on Colorectal Cancer followed suit, lowering the recommended age to begin screening to 45 years.

One study published shortly after the ACS guideline change in 2018 showed an increase in past-year CRC screening in those aged 45 to 49 years from quarter 1 to quarter 4 of 2018 (8). A later study, contrarily, found no increase in past-year CRC screening by quarter in 2019 (9). Both studies, however, highlighted the low rate of screening up to 2019. Herein, we extended these findings by examining changes in screening prevalence in this newly eligible age group with contemporary estimates through 2021.

Data for CRC screening in individuals aged 45 to 49 years were obtained from the 2019 and 2021 National Health Interview Survey (NHIS), a nationally representative cross-sectional household survey of noninstitutionalized adults. We limited our analysis to the 2019 and 2021 surveys because cancer screening questions are included only biennially and because the NHIS was heavily redesigned in 2019, precluding comparison with surveys from previous years. Although the survey is primarily conducted face to face, 62.8% of interviews in 2021 were conducted over the telephone because of the COVID-19 pandemic compared with 34.3% in 2019. The outcome of interest was self-reported up-to-date CRC screening using 1 of the following modalities: 1) fecal occult blood test (FOBT) or fecal immunochemical test (FIT) in the past year, 2) multitarget stool DNA-FIT test in the past 3 years 3) sigmoidoscopy or computed tomography colonography in the past 5 years, or 4) colonoscopy in the past 10 years. Multitarget stool DNA-FIT testing questions were asked only among those who responded “yes” to an FOBT and/or FIT test (Supplementary Table 1, available online) in 2019, so we restricted multitarget stool DNA-FIT testing to those who also received an FOBT and/or FIT test in the 2021 up-to-date CRC screening estimate for consistency across survey years. Additionally, where provided separately, FOBT and/or FIT prevalence did not include multitarget stool DNA-FIT testing. Screening prevalence and population number screened were estimated based on 2019 and 2021 data. Prevalence differences in screening with associated 95% confidence intervals (CIs) between survey years and across sociodemographic variables were calculated using a logistic regression model adjusted for age, sex, social construction of racialized and ethnic groups, and education level. All analyses were weighted to be nationally representative and accounted for complex survey sampling.

Up-to-date CRC screening prevalence in those aged 45 to 49 years remained stable at 20.8% (95% CI = 18.9% to 23.0%) in 2019 and 19.7% (95% CI = 17.8% to 21.6%) in 2021, colonoscopy remained stable between 19.5% (95% CI = 17.6% to 21.5%) and 17.8% (95% CI = 16.0% to 19.8%), whereas the rate of FOBT and/or FIT testing increased from 1.2% (95% CI = 0.8% to 1.9%) to 2.4% (95% CI = 1.8% to 3.3%) (adjusted prevalence difference = 1.3%, 95% CI = 0.3% to 2.3%) (Table 1). In pooled 2019-2021 data, up-to-date CRC screening prevalence varied by sociodemographic characteristics (Table 2). Up-to-date CRC prevalence ranged from 13.1% (95% CI = 9.3% to 18.2%) in non-Hispanic Asian people to 22% (95% CI = 20.2% to 23.9%) in non-Hispanic White people (adjusted prevalence difference = ‒9%, 95% CI = ‒13.7% to ‒4.3%); from 15.4% (95% CI = 11.3% to 20.7%) in people with less than a high school diploma to 23% (95% CI = 20.9% to 25.2%) in those with a college degree (adjusted prevalence difference = ‒6.4%, 95% CI = ‒11.7% to ‒1.1%); and from 7.6% (95% CI = 5.5% to 10.4%) in uninsured people to 21.4% (95% CI = 19.8% to 23.1%) in privately insured people (adjusted prevalence difference = ‒12.4%, 95% CI = ‒16.1% to ‒8.7%). The prevalence of up-to-date CRC screening ranged from 17.1% (95% CI = 14.7% to 19.8%) in the western United States and 18.9% (95% CI = 16.2% to 22.0%) in the Midwest to 20.9% (95% CI = 18.7% to 23.3%) in the South and 24.7% (95% CI = 20.9% to 28.9%) in the northeast United States.

Table 1.

Up-to-date colorectal cancer screening prevalence among individuals aged 45 to 49 years, 2019 and 2021

	2019			2021			Change from 2019 to 2021
	Screened	Unscreened	Prevalence	Screened	Unscreened	Prevalence	Adjusted prevalence differencea
	No. (95% CI)	No. (95% CI)	% (95% CI)	No. (95% CI)	No. (95% CI)	% (95% CI)	% points (95% CI)
CRC screening	4 171 400 (3 688 404 to 4 654 397)	15 846 367 (14 818 481 to 16 874 252)	20.8 (18.9, to 23.0)	3 729 005 (3 313 434 to 4 144 576)	15 244 087 (14 245 205 to 16 242 968)	19.7 (17.8 to 21.6)	‒1.0 (‒3.7 to 1.7)
Colonoscopy	3 890 607 (3 435 540 to 4 345 673)	16 077 954 (15 042 865 to 17 113 043)	19.5 (17.6 to 21.5)	3 358 939 (2 956 943 to 3 760 935)	15 492 327 (14 484 931 to 16 499 722)	17.8 (16.0 to 19.8)	‒1.5 (‒4.2 to 1.2)
FOBT and/or FIT	236 521 (131 626 to 341 416)	19 629 410 (18 442 959 to 20 815 861)	1.2 (0.8 to 1.9)	455 831 (317 306 to 594 356)	18 434 233 (17 315 781 to 19 552 685)	2.4 (1.8 to 3.3)	1.3 (0.3 to 2.3)

a

Prevalence estimates and prevalence differences were survey weighted. Prevalence differences were adjusted for age, sex, racial and ethnic group, and education level. CRC screening was defined as FOBT or FIT in the past year, multitarget stool DNA-FIT testing in the past 3 years among those who received an FOBT and/or FIT test, sigmoidoscopy, or computed tomography colonography in the past 5 years, or colonoscopy in the past 10 years. Colonoscopy was defined as colonoscopy in the past 10 years. FOBT and/or FIT was defined as FOBT or FIT, not including multitarget stool DNA-FIT testing, in the past year. CI = confidence interval; CRC = colorectal cancer; FIT = fecal immunochemical test; FOBT = fecal occult blood test.

Table 2.

Up-to-date colorectal cancer screening prevalence and adjusted prevalence differences, by sociodemographic characteristics, among individuals aged 45 to 49 years, 2019-2021

	CRC screening		Colonoscopy		FOBT and/or FIT
	Prevalencea	Change from 2019 to 2021, adjusted prevalence differencea	Prevalencea	Change from 2019 to 2021, adjusted prevalence differencea	Prevalencea	Change from 2019 to 2021, adjusted prevalence differencea
Characteristics	% (95% CI)	% points (95% CI)	% (95% CI)	% points (95% CI)	% (95% CI)	% points (95% CI)
Sex
Female	21.6 (19.7 to 23.6)	2.6 (‒0.1 to 5.3)	20.1 (18.2 to 22.1)	2.7 (0.0 to 5.4)	2.0 (1.4 to 2.7)	0.3 (‒0.7 to 1.3)
Male	18.9 (16.9 to 21.0)	[Referent]	17.2 (15.4 to 19.3)	[Referent]	1.6 (1.1 to 2.3)	[Referent]
Race and ethnicity
Asian	13.1 (9.3 to 18.2)	‒9.0 (‒13.7 to ‒4.3)	11.7 (8.2 to 16.5)	‒9.0 (‒13.5 to ‒4.5)	—b	—
Black	23.7 (19.6 to 28.3)	2.2 (‒2.5 to 6.9)	20.9 (17.1 to 25.2)	0.7 (‒3.8 to 5.2)	—	—
Hispanic	16.1 (13.0 to 19.7)	‒3.9 (‒8.2 to 0.4)	14.3 (11.4 to 17.9)	‒4.1 (‒8.2 to 0.0)	2.0 (1.2 to 3.3)	0.2 (‒1.0 to 1.4)
White	22.0 (20.2 to 23.9)	[Referent]	20.7 (18.9 to 22.5)	[Referent]	1.5 (1.1 to 2.1)	[Referent]
Education
College degree	23.0 (20.9 to 25.2)	[Referent]	21.6 (19.6 to 23.8)	[Referent]	1.6 (1.1 to 2.4)	[Referent]
Some college	21.9 (19.3 to 24.8)	‒1.7 (‒5.2 to 1.8)	20.0 (17.5 to 22.8)	‒2.2 (‒5.7 to 1.3)	2.3 (1.6 to 3.4)	0.6 (‒0.6 to 1.8)
High school diploma	17.0 (14.3 to 20.1)	‒6.1 (‒9.6 to ‒2.6)	15.9 (13.2 to 18.9)	‒5.8 (‒9.3 to ‒2.3)	—	—
Less than high school	15.4 (11.3 to 20.7)	‒6.4 (‒11.7 to ‒1.1)	12.6 (8.9 to 17.6)	‒7.8 (‒12.9 to ‒2.7)	—	—
Income
≥200% of the federal poverty level	20.9 (19.4 to 22.5)	[Referent]	19.4 (18.0 to 20.9)	[Referent]	1.5 (1.1 to 2.0)	[Referent]
100%-199% of the federal poverty level	17.7 (14.2 to 21.9)	‒0.6 (‒5.3 to 4.1)	16.5 (13.1 to 20.7)	0.2 (‒4.7 to 5.1)	2.7 (1.6 to 4.5)	1.0 (‒0.6 to 2.6)
<100% of the federal poverty level	19.2 (14.5 to 25.0)	1.6 (‒4.7 to 7.9)	16.4 (12.0 to 22.1)	0.7 (‒5.6 to 7.0)	—	—
Immigration status
US citizen	22.9 (21.3 to 24.7)	[Referent]	21.3 (19.7 to 23.0)	[Referent]	1.8 (1.4 to 2.4)	[Referent]
In the United States <10 y	14.1 (8.3 to 23.1)	‒7.8 (‒15.8 to 0.2)	10.6 (5.9 to 18.5)	‒9.8 (‒16.7 to ‒2.9)	—	—
In the United States ≥10 y	13.0 (10.3 to 16.3)	‒9.0 (‒13.7 to ‒4.3)	11.5 (9.0 to 14.6)	‒8.6 (‒12.9 to ‒4.3)	1.7 (0.9 to 3.0)	‒0.3 (‒2.1 to 1.5)
Insurance type
Private	21.4 (19.8 to 23.1)	[Referent]	19.8 (18.3 to 21.5)	[Referent]	1.7 (1.3 to 2.2)	[Referent]
Uninsured	7.6 (5.5 to 10.4)	‒12.4 (‒16.1 to ‒8.7)	6.4 (4.6 to 9.1)	‒11.7 (‒15.2 to ‒8.2)	—	—
Medicaid and other public insurancec	22.2 (18.2 to 26.9)	2.2 (‒2.7 to 7.1)	20.6 (16.6 to 25.2)	2.8 (‒2.1 7.7)	—	—
US region
Northeast	24.7 (20.9 to 28.9)	[Referent]	24.1 (20.5 to 28.3)	[Referent]	—	[Referent]
Midwest	18.9 (16.2 to 22.0)	‒6.3 (‒11.0 to ‒1.6)	17.5 (14.9 to 20.5)	‒7.1 (‒11.8 to ‒2.4)	—	—
South	20.9 (18.7 to 23.3)	‒3.9 (‒8.6 to 0.8)	19.2 (17.1 to 21.5)	‒4.8 (‒9.3 to ‒0.3)	2.0 (1.3 to 2.9)	—
West	17.1 (14.7 to 19.8)	‒6.5 (‒11.4 to ‒1.6)	14.8 (12.5 to 17.4)	‒8.3 (‒13.0 to ‒3.6)	2.3 (1.5 to 3.4)	—

a

Prevalence estimates and prevalence differences were survey weighted. Prevalence differences were adjusted for age, sex, racial and ethnic group, and education level. CRC screening was defined as FOBT or FIT in the past year, multitarget stool DNA-FIT testing in the past 3 years among those who received an FOBT and/or FIT test, sigmoidoscopy, computed tomography colonography in the past 5 years, or colonoscopy in the past 10 years. Colonoscopy was defined as colonoscopy in the past 10 years. FOBT and/or FIT was defined as FOBT or FIT, not including multitarget stool DNA-FIT testing, in the past year. CI = confidence interval; CRC = colorectal cancer; FIT = fecal immunochemical test; FOBT = fecal occult blood test.

b

Estimates were unstable if the relative SE was 30% or more or the denominator was less than 50.

c

“Medicaid and other public insurance” did not include Medicare because participants were younger than 65 years of age.

Three years after the ACS guideline update, fewer than 4 million of the 19 million eligible US adults aged 45 to 49 years were up-to-date for CRC screening in 2021, with no change from 2019. This finding translates to about 15 million eligible adults who have not been screened. Of concern are the racial, ethnic, and socioeconomic disparities evident in CRC screening uptake in this younger population. FOBT and/or FIT testing remains underutilized, with just 2.4% (<460 000 people) reporting being up-to-date with this modality in 2021.

Low CRC screening rates among individuals aged 45 to 49 years are perhaps related to cancer screening declines during the COVID-19 pandemic (10). Also, influential organizations such as the American Association of Family Physicians and American College of Physicians do not currently recommended initiation at age 45 years. More likely, low rates are possibly related to the USPSTF not updating its recommendations until 2021. The USPSTF recommendations are linked to insurance coverage, a crucial determinant of cancer screening referral and utilization (11).

One of the desired outcomes of reducing the age to begin screening from 50 to 45 years is to alleviate some of the disparate CRC incidence and mortality in historically underserved racial and ethnic groups. CRC incidence and mortality rates, irrespective of age, are higher in Alaska Native, American Indian, and Black people than in White, Hispanic, and Asian American or Pacific Islander people (12). CRC mortality rates in Alaska Native people are particularly disproportionate—nearly 3 times the rate in White people (12). It is encouraging, therefore, that Black people aged 45 to 49 years had CRC screening prevalence similar to White people, although prevalence remains low in both groups. Consistent with CRC screening in those aged 50 to 75 years (13), however, we found that prevalence was lowest in Asian people, uninsured people, and individuals with low socioeconomic status. The exception to this pattern was the comparable up-to-date rates in dual-eligible, Medicaid, and other publicly insured individuals vs privately insured people, which is suggestive of spillover effects from expanded health-care access during the COVID-19 pandemic (14,15). Public health insurance expansions may therefore offer the greatest benefit for lower-resourced and underscreened groups, including uninsured and low-income individuals, who have the lowest screening rates.

Although evidence of interventions that effectively improve CRC screening specifically in younger adults is lacking, low rates can potentially be improved by using FOBT and/or FIT testing, which remains underutilized in this population and offers an economical and convenient alternative to time-intensive, invasive colonoscopy screening. For FOBT and/or FIT testing to be part of a viable population-based CRC screening strategy, policy changes must ensure that follow-up diagnostic colonoscopies after positive FOBT and/or FIT tests be considered part of the screening continuum and not incur cost barriers. A notable example is the 2022 US Department of Labor–Health and Human Services–Treasury joint statement, in accordance with section 2713 of the Public Health Service Act and regulations, requiring that insurance plans and issuers cover colonoscopy—without cost-sharing—after a positive noninvasive FOBT and/or FIT based or direct visualization test (16).

Apart from structural policy changes, patient-level efforts, such as mailed FIT outreach programs in underserved groups of eligible adults aged 45 to 49 years, have been effective in improving initial FOBT and/or FIT testing, along with time to follow-up colonoscopy when contacted by a patient navigator for scheduling (17,18). Health system interventions to improve clinician-patient communication regarding CRC screening are also likely to be impactful, as more than 70% of all adults overdue for screening in 2019-2021 did not receive a screening recommendation from their clinician, despite having attended a wellness visit in the past year (19).

Despite pooling data from the 2019 and 2021 NHIS survey years to estimate prevalence in sociodemographic groups with small sample sizes, we could not estimate prevalence among Alaska Native and American Indian people, who have disproportionately higher risk for CRC burden (12). The NHIS is self-reported, and response bias is a concern; additionally, responses may not align with those recorded in electronic health records (20). The 2021 NHIS recommended interview modalities (in person and telephone) differed from the 2019 survey (in person), but interviews had largely returned to in person by May 2021. We could not distinguish between screening and diagnostic exams for up-to-date CRC screening prevalence because the NHIS collects the indication only for colonoscopy and no other screening modality. Among respondents 45 to 49 years of age, 8% reported that the “main reason” for colonoscopy was follow-up of an earlier test; 43% reported that it was “because of a problem.”

Among those aged 45 to 49 years, reported CRC screening remained low in 2021, with little improvement since 2019. About 15 million eligible individuals aged 45 to 49 years were not up-to-date as of 2021. Concerted efforts at the patient, clinician, health system, and societal levels are needed to improve CRC screening rates among newly eligible adults aged 45 to 49 years, particularly among those in underserved racial, ethnic, and socioeconomic groups.

Data availability

The data that support the findings of this study are publicly available at NHIS (cdc.gov).

Author contributions

Jessica Star, MA, MPH (Conceptualization; Formal analysis; Methodology; Validation; Writing—original draft; Writing—review & editing), Rebecca L. Siegel, MPH (Conceptualization; Writing—review & editing), Adair K. Minihan, MPH (Writing—review & editing), Robert A. Smith, PhD (Writing—review & editing), Ahmedin Jemal, DVM, PhD (Conceptualization; Writing—review & editing), Priti Bandi, PhD (Conceptualization; Supervision; Writing—review & editing).

Funding

All authors were employed by the American Cancer Society, which receives grants from private and corporate foundations, including foundations associated with companies in the health sector, for research outside the submitted work.

Conflicts of interest

The authors have no competing interests to report.

Acknowledgements

The funders had no role in the design of the study; the collection, analysis, or interpretation of the data; or the writing of the manuscript and decision to submit it for publication.