[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure.
Distribution of Ever Participation in Lower Endoscopy at the Baseline Interview
Distribution of Ever Participation in Lower Endoscopy at the Baseline Interview

Colorectal cancer test participation among participants not diagnosed with incident colorectal cancer during the study period and who were eligible for colorectal cancer screening at the baseline interview. All comparisons were significant at P < .05. Additional baseline characteristics by colorectal cancer test participation are given in eTable 1 and eTable 2 in the Supplement.

Table 1.  
Selected Baseline Characteristics of Southern Community Cohort Study Participants, 2002-2009, by CRC Incidence and Vital Statusa
Selected Baseline Characteristics of Southern Community Cohort Study Participants, 2002-2009, by CRC Incidence and Vital Statusa
Table 2.  
Associations Between Lower Endoscopy and Colorectal Cancer Incidence by Timing of CRC Testing Among Southern Community Cohort Study Participants Eligible for CRC Screening at the Baseline Interview
Associations Between Lower Endoscopy and Colorectal Cancer Incidence by Timing of CRC Testing Among Southern Community Cohort Study Participants Eligible for CRC Screening at the Baseline Interview
Table 3.  
Associations Between CRC Testing Between the Baseline and Follow-up Interviews With CRC Incidence and Mortality Among Southern Community Cohort Study Participants Eligible for CRC Screening at the Follow-up Interview
Associations Between CRC Testing Between the Baseline and Follow-up Interviews With CRC Incidence and Mortality Among Southern Community Cohort Study Participants Eligible for CRC Screening at the Follow-up Interview
Table 4.  
Associations Between Lower Endoscopy and CRC Mortality by Timing of CRC Testing and Sex Among Southern Community Cohort Study Participants Eligible for CRC Screening at the Baseline Interview
Associations Between Lower Endoscopy and CRC Mortality by Timing of CRC Testing and Sex Among Southern Community Cohort Study Participants Eligible for CRC Screening at the Baseline Interview
1.
American Cancer Society.  Cancer Facts & Figures 2018. Atlanta, GA: American Cancer Society; 2018.
2.
American Cancer Society.  Colorectal Cancer Facts & Figures 2017-2019. Atlanta, GA: American Cancer Society; 2017.
3.
Elmunzer  BJ, Hayward  RA, Schoenfeld  PS, Saini  SD, Deshpande  A, Waljee  AK.  Effect of flexible sigmoidoscopy-based screening on incidence and mortality of colorectal cancer: a systematic review and meta-analysis of randomized controlled trials.  PLoS Med. 2012;9(12):e1001352. doi:10.1371/journal.pmed.1001352PubMedGoogle Scholar
4.
Weissfeld  JL, Schoen  RE, Pinsky  PF,  et al; PLCO Project Team.  Flexible sigmoidoscopy in the randomized Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial: added yield from a second screening examination.  J Natl Cancer Inst. 2012;104(4):280-289. doi:10.1093/jnci/djr549PubMedGoogle ScholarCrossref
5.
Nishihara  R, Wu  K, Lochhead  P,  et al.  Long-term colorectal-cancer incidence and mortality after lower endoscopy.  N Engl J Med. 2013;369(12):1095-1105. doi:10.1056/NEJMoa1301969PubMedGoogle ScholarCrossref
6.
Shaukat  A, Mongin  SJ, Geisser  MS,  et al.  Long-term mortality after screening for colorectal cancer.  N Engl J Med. 2013;369(12):1106-1114. doi:10.1056/NEJMoa1300720PubMedGoogle ScholarCrossref
7.
Shroff  J, Thosani  N, Batra  S, Singh  H, Guha  S.  Reduced incidence and mortality from colorectal cancer with flexible-sigmoidoscopy screening: a meta-analysis.  World J Gastroenterol. 2014;20(48):18466-18476. doi:10.3748/wjg.v20.i48.18466PubMedGoogle ScholarCrossref
8.
Lin  JS, Piper  MA, Perdue  LA,  et al.  Screening for colorectal cancer: updated evidence report and systematic review for the US Preventive Services Task Force.  JAMA. 2016;315(23):2576-2594. doi:10.1001/jama.2016.3332PubMedGoogle ScholarCrossref
9.
Rex  DK, Johnson  DA, Anderson  JC, Schoenfeld  PS, Burke  CA, Inadomi  JM; American College of Gastroenterology.  American College of Gastroenterology guidelines for colorectal cancer screening 2009 [published correction appears in Am J Gastroenterol 2009;104(6):1613].  Am J Gastroenterol. 2009;104(3):739-750. doi:10.1038/ajg.2009.104PubMedGoogle ScholarCrossref
10.
Franks  P, Muennig  P, Lubetkin  E, Jia  H.  The burden of disease associated with being African-American in the United States and the contribution of socio-economic status.  Soc Sci Med. 2006;62(10):2469-2478. doi:10.1016/j.socscimed.2005.10.035PubMedGoogle ScholarCrossref
11.
Williams  DR, Mohammed  SA, Leavell  J, Collins  C.  Race, socioeconomic status, and health: complexities, ongoing challenges, and research opportunities.  Ann N Y Acad Sci. 2010;1186:69-101. doi:10.1111/j.1749-6632.2009.05339.xPubMedGoogle ScholarCrossref
12.
Grubbs  SS, Polite  BN, Carney  J  Jr,  et al.  Eliminating racial disparities in colorectal cancer in the real world: it took a village.  J Clin Oncol. 2013;31(16):1928-1930. doi:10.1200/JCO.2012.47.8412PubMedGoogle ScholarCrossref
13.
Phelan  JC, Link  BG, Diez-Roux  A, Kawachi  I, Levin  B.  “Fundamental causes” of social inequalities in mortality: a test of the theory.  J Health Soc Behav. 2004;45(3):265-285. doi:10.1177/002214650404500303PubMedGoogle ScholarCrossref
14.
Doubeni  CA, Corley  DA, Zauber  AG.  Colorectal cancer health disparities and the role of US law and health policy.  Gastroenterology. 2016;150(5):1052-1055. doi:10.1053/j.gastro.2016.03.012PubMedGoogle ScholarCrossref
15.
Hall  WJ, Chapman  MV, Lee  KM,  et al.  Implicit racial/ethnic bias among health care professionals and its influence on health care outcomes: a systematic review.  Am J Public Health. 2015;105(12):e60-e76. doi:10.2105/AJPH.2015.302903PubMedGoogle ScholarCrossref
16.
Signorello  LB, Hargreaves  MK, Steinwandel  MD,  et al.  Southern community cohort study: establishing a cohort to investigate health disparities.  J Natl Med Assoc. 2005;97(7):972-979.PubMedGoogle Scholar
17.
Signorello  LB, Hargreaves  MK, Blot  WJ.  The Southern Community Cohort Study: investigating health disparities.  J Health Care Poor Underserved. 2010;21(1)(suppl):26-37. doi:10.1353/hpu.0.0245PubMedGoogle ScholarCrossref
18.
US Preventive Services Task Force.  Screening for colorectal cancer: recommendation and rationale.  Ann Intern Med. 2002;137(2):129-131. doi:10.7326/0003-4819-137-2-200207160-00014PubMedGoogle ScholarCrossref
19.
US Preventive Services Task Force.  Screening for colorectal cancer: US Preventive Services Task Force recommendation statement.  Ann Intern Med. 2008;149(9):627-637. doi:10.7326/0003-4819-149-9-200811040-00243PubMedGoogle ScholarCrossref
20.
Levin  B, Lieberman  DA, McFarland  B,  et al; American Cancer Society Colorectal Cancer Advisory Group; US Multi-Society Task Force; American College of Radiology Colon Cancer Committee.  Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology.  Gastroenterology. 2008;134(5):1570-1595. doi:10.1053/j.gastro.2008.02.002PubMedGoogle ScholarCrossref
21.
Winawer  S, Fletcher  R, Rex  D,  et al; Gastrointestinal Consortium Panel.  Colorectal cancer screening and surveillance: clinical guidelines and rationale-Update based on new evidence.  Gastroenterology. 2003;124(2):544-560. doi:10.1053/gast.2003.50044PubMedGoogle ScholarCrossref
22.
Murff  HJ, Peterson  NB, Fowke  JH,  et al.  Colonoscopy screening in African Americans and whites with affected first-degree relatives.  Arch Intern Med. 2008;168(6):625-631. doi:10.1001/archinte.168.6.625PubMedGoogle ScholarCrossref
23.
Hall  IJ, Tangka  FKL, Sabatino  SA, Thompson  TD, Graubard  BI, Breen  N.  Patterns and trends in cancer screening in the United States.  Prev Chronic Dis. 2018;15:E97. doi:10.5888/pcd15.170465PubMedGoogle Scholar
24.
Doubeni  CA, Laiyemo  AO, Reed  G, Field  TS, Fletcher  RH.  Socioeconomic and racial patterns of colorectal cancer screening among Medicare enrollees in 2000 to 2005.  Cancer Epidemiol Biomarkers Prev. 2009;18(8):2170-2175. doi:10.1158/1055-9965.EPI-09-0104PubMedGoogle ScholarCrossref
25.
National Center for Health Statistics.  Health, United States, 2011: With Special Feature on Socioeconomic Status and Health. Hyattsville, MD: National Center for Health Statistics; 2012.
26.
Peterson  NB, Murff  HJ, Fowke  JH,  et al.  Use of colonoscopy and flexible sigmoidoscopy among African Americans and whites in a low-income population.  Prev Chronic Dis. 2008;5(1):A28.PubMedGoogle Scholar
27.
Bernardo  BM, Gross  AL, Young  G,  et al.  Predictors of colorectal cancer screening in two underserved U.S. populations: a parallel analysis.  Front Oncol. 2018;8:230. doi:10.3389/fonc.2018.00230PubMedGoogle ScholarCrossref
28.
Cohen  SS, Murff  HJ, Signorello  LB, Blot  WJ.  Obesity and colorectal cancer screening among black and white adults.  Cancer Causes Control. 2012;23(5):709-716. doi:10.1007/s10552-012-9940-yPubMedGoogle ScholarCrossref
29.
Laiyemo  AO, Doubeni  C, Pinsky  PF,  et al.  Race and colorectal cancer disparities: health-care utilization vs different cancer susceptibilities.  J Natl Cancer Inst. 2010;102(8):538-546. doi:10.1093/jnci/djq068PubMedGoogle ScholarCrossref
30.
Reiter  PL, Katz  ML, Oliveri  JM, Young  GS, Llanos  AA, Paskett  ED.  Validation of self-reported colorectal cancer screening behaviors among Appalachian residents.  Public Health Nurs. 2013;30(4):312-322. doi:10.1111/phn.12038PubMedGoogle ScholarCrossref
31.
Shokar  NK, Vernon  SW, Carlson  CA.  Validity of self-reported colorectal cancer test use in different racial/ethnic groups.  Fam Pract. 2011;28(6):683-688. doi:10.1093/fampra/cmr026PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    Public Health
    December 20, 2019

    Association of Race and Socioeconomic Status With Colorectal Cancer Screening, Colorectal Cancer Risk, and Mortality in Southern US Adults

    Author Affiliations
    • 1Vanderbilt Epidemiology Center, Vanderbilt-Ingram Cancer Center, Division of Epidemiology, Department of Medicine, Vanderbilt University School of Medicine, Nashville, Tennessee
    • 2Department of Population Health Sciences, School of Medicine and Public Health, University of Wisconsin, Madison
    • 3University of Wisconsin Carbone Cancer Center, Madison
    • 4International Epidemiology Field Station, Vanderbilt Institute for Clinical and Translational Research, Nashville, Tennessee
    • 5Department of Medicine, Vanderbilt University Medical Center, Nashville, Tennessee
    JAMA Netw Open. 2019;2(12):e1917995. doi:10.1001/jamanetworkopen.2019.17995
    Key Points español 中文 (chinese)

    Question  What are the associations between colorectal cancer screening modalities with colorectal cancer risk and mortality among African American individuals and individuals with low socioeconomic status, who more often face barriers to screening and experience colorectal cancer health disparities?

    Findings  In this cohort study of 47 596 adults, colorectal cancer screening was significantly associated with reduced colorectal cancer risk and mortality. Use of colonoscopy was associated with a 45% reduced incidence of colorectal cancer, use of sigmoidoscopy with a 34% reduced incidence, and stool-based tests with a 25% reduced incidence compared with no screening.

    Meaning  The large colorectal cancer disparities experienced by individuals with low socioeconomic status and African American individuals may be lessened by improving access to and uptake of colorectal cancer screening.

    Abstract

    Importance  Colorectal cancer (CRC) screening is rarely studied in populations who may face additional barriers to participate in cancer screening, such as African American individuals and individuals with low socioeconomic status (SES).

    Objective  To examine the associations of CRC screening and modalities with CRC incidence and mortality by race and SES.

    Design, Setting, and Participants  This cohort study used data from the Southern Community Cohort Study, which enrolled more than 85 000 participants from community health centers or stratified random sampling of the general population in 12 states in the southeastern United States. The present study included data from cohort members who were eligible for CRC screening as recommended by expert organizations based on age and family history. Participants completed questionnaires from 2002 to 2009 and were contacted again from 2008 to 2012. Linkages to state cancer registries and the National Death Index as of December 31, 2016, identified incident CRC and vital status. Data analysis was performed from January 1, 2018, to October 30, 2019.

    Main Outcomes and Measures  Incident CRC (n = 632) and mortality (n = 10 003). Cox proportional hazards regression models evaluated associations between screening modalities and CRC risk and mortality. Information on fecal occult blood test use was only obtained on the follow-up questionnaire. Self-identified race was measured as African American/black, white, or other, and SES was defined by household income.

    Results  This study included 47 596 participants (median baseline age, 54 years [interquartile range, 10 years]; 32 185 [67.6%] African American; 28 884 [60.7%] female; and 26 075 [54.8%] with household income <$15 000). A total of 24 432 participants (63.9%) had never undergone CRC testing at baseline. The CRC testing assessed at baseline and follow-up interviews was associated with significant CRC risk reduction (hazard ratio [HR], 0.55; 95% CI, 0.44-0.70 for ever colonoscopy at baseline). Results were similar in analyses stratified by race (African American: HR, 0.65; 95% CI, 0.50-0.85; white: HR, 0.44; 95% CI, 0.27-0.70) and household income (<$15 000: HR, 0.63; 95% CI, 0.46-0.86, ≥$15 000: HR, 0.49; 95% CI, 0.35-0.69). Ever sigmoidoscopy at baseline was associated with CRC risk reduction (HR, 0.66; 95% CI, 0.51-0.87), and undergoing fecal occult blood test in the interval between baseline and follow-up interview was associated with CRC risk reduction (HR, 0.75; 95% CI, 0.57-0.98). Inverse associations were also observed between CRC mortality and receipt of colonoscopy (HR for women, 0.39; 95% CI, 0.21-0.73; HR for men, 0.69; 95% CI, 0.40-1.18) and sigmoidoscopy (HR for women, 0.37; 95% CI, 0.16-0.85; HR for men, 0.82; 95% CI, 0.46-1.47); however, the association did not extend to fecal occult blood test (HR for women, 1.02; 95% CI, 0.62-1.70; HR for men, 1.03; 95% CI, 0.55-1.93).

    Conclusions and Relevance  In this study, CRC test rates were low among African American individuals and those with low SES. The findings suggest that screening, particularly with colonoscopy, is significantly associated with reduced risk of CRC and mortality. The CRC disparities experienced by individuals with low SES and African American individuals may be lessened by improving access to and uptake of CRC screening.

    Introduction

    Colorectal cancer (CRC) is the fourth most commonly diagnosed cancer in the United States and second-leading cause of cancer death.1 Colorectal cancer outcomes vary by race and socioeconomic status (SES), with African American individuals and individuals with low SES having higher incidence and mortality than other groups. Most CRCs arise from noncancerous, adenomatous polyps.2 Colorectal cancer screening programs aim to intervene in the cancer progression process by the detection and subsequent removal of polyps before they progress to cancer and by early detection of CRCs in asymptomatic individuals. The importance of CRC screening by visual- and stool-based screening modalities for reducing CRC incidence and mortality has been demonstrated in randomized clinical trials and cohort studies.3-8 Cohort studies provide opportunities to evaluate the effectiveness of CRC screening in targeted populations. A previous cohort study5 of data from follow-up of nurses and other health care professionals reported 40% to 56% lower CRC incidence among participants screened through lower endoscopy (colonoscopy or sigmoidoscopy) compared with participants who had never undergone lower endoscopy screening, and a 68% reduction in colorectal cancer mortality among participants who had undergone screening colonoscopy. Stool-based screening programs, such as fecal occult blood tests (FOBTs) and fecal immunochemical tests, have also been shown to be effective for detecting colorectal polyps and cancers.8 Randomized clinical trials of stool-based screening modalities have found reduced CRC incidence and mortality among patients with abnormal stool-based screening findings who underwent diagnostic colonoscopy and polypectomy.8,9

    The effectiveness of CRC screening has been rarely studied in African American individuals or populations with low SES, who may face additional barriers to participate in cancer screening or receive less than optimal medical care. Specifically, difference in screening participation is hypothesized to contribute to disparities in CRC outcomes.10-12 Barriers that may contribute to the effectiveness of CRC screening among African American patients or patients with low SES include inadequate insurance coverage, lack of a medical home, less access to financial resources, less access to transportation, and schedule availability.13,14 In addition, African American individuals report lower quality of health care and may be subject to implicit racial bias of their health care practitioners.15 The Southern Community Cohort Study (SCCS) is a landmark prospective cohort study designed to ensure the cohort’s makeup is mostly African American and of low SES to address questions related to cancer disparities associated with SES and race. The SCCS cohort provided a unique opportunity to investigate CRC test patterns and associations with CRC risk and mortality in underserved racial and socioeconomic populations. We evaluated the associations between CRC test modalities and incident CRC and mortality.

    Methods
    Study Population

    This cohort study used data from the SCCS (detailed descriptions of the prospective SCCS have been previously published).16,17 In brief, the SCCS enrolled more than 85 000 adult participants from 2002 to 2009 in 12 states in the southeastern United States. Participants were aged 40 to 79 years at enrollment and English speaking. Most participants were enrolled at community health centers, which are health care organizations that primarily work with medically underserved populations. The remaining cohort participants were enrolled using an identical mailed questionnaire sent to stratified random samples of residents in the same 12 states. At baseline study entry, all participants completed questionnaires to obtain information on cancer screening participation, lifestyle habits, personal and familial medical history, and demographic characteristics, including self-identified race and sex. Participants self-identified their race according to categories defined by the investigators. The SCCS was approved by institutional review boards at Vanderbilt University Medical Center and Meharry Medical College. All participants provided written informed consent, and data were deidentified. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

    Ascertainment of information on the cohort subsequent to baseline enrollment has been obtained through passive and active methods. Participants were contacted again from 2008 to 2012 to obtain updated information on health history and lifestyle exposures via telephone or mailed paper questionnaire. A total of 34 516 participants (72.5%) in this study completed the follow-up interview. The median time between baseline and follow-up interview was 4.7 years (interquartile range [IQR], 1.9 years). Linkages to state cancer registries and the National Death Index through December 31, 2016, acquired information on site-specific cancer incidence and cause-specific mortality. Incident colon or rectal cancer were defined by International Classification of Diseases for Oncology codes C180 to C189, C199, and C209 and CRC mortality by International Statistical Classification of Diseases and Related Health Problems, Tenth Revision codes C18 to C20.

    Analytical Data Set: Participant Eligibility Information

    Participants eligible for inclusion in the study were without diagnosis of any cancer (not including nonmelanoma skin cancer) or colorectal adenoma at or before study entry. Participants also must have been eligible for CRC screening using the screening recommendations for the period of enrollment into the SCCS (2002-2009) from the US Preventive Services Task Force18,19 or the joint recommendation from the American Cancer Society, the US Multi-Society Task Force on CRC, and the American College of Radiology.20,21 The CRC screening recommendations applied to adults 50 years or older at study entry and adults 40 years or older at study entry who had a family history of CRC in a first-degree relative. A second group of participants included those who became eligible for inclusion at the follow-up interview by meeting the eligibility criterion of age of 50 years or older or a family history of CRC at the follow-up interview. We identified 38 211 participants eligible for CRC screening at study enrollment and 47 596 participants eligible for CRC screening at their follow-up interview.

    Exposure Assessment: CRC Test Participation Before Baseline Enrollment and Between Baseline and Follow-up Interviews

    We examined the associations between CRC outcomes and use of the following CRC test modalities: colonoscopy, sigmoidoscopy, or stool-based screening. At baseline and follow-up interviews, participants were asked about their use of colonoscopy or sigmoidoscopy. In the calendar period between baseline and follow-up interviews, use of stool-based screening modalities became more common in the United States; thus, during the follow-up interview, participants were asked if they had an FOBT since joining the study.

    We conducted stratified analysis during which we categorized participants into groups by available follow-up interview information, CRC status, and vital status (eFigure in the Supplement). Associations with CRC incidence were first evaluated by assessing associations between ever participation in CRC screening at baseline and follow-up interviews, and analyses were separately conducted to quantify the associations of colonoscopy, sigmoidoscopy, and FOBT with CRC incidence and mortality.

    Statistical Analysis

    Frequency distributions of participant characteristics were tabulated by CRC diagnosis and vital status and by CRC test participation. We evaluated the association between race and participation in colonoscopy using multivariable logistic regression. Hazard ratios (HRs) and 95% CIs were estimated using Cox proportional hazards regression models for the associations between CRC test variables and CRC incidence, CRC mortality, and all-cause mortality. We evaluated associations with non-CRC mortality and found similar results as analyses for all-cause mortality; thus, we present results only for all-cause mortality. Age was used as the time scale. In analyses using data from the baseline interview, entry time was defined as age at study entry. For CRC risk models, exit time was defined as age at CRC diagnosis, age at loss to follow-up, or December 31, 2016. For mortality models, exit time was defined as age at death, age at loss to follow-up, or December 31, 2016. In analyses in which the exposure was CRC testing in the interval between baseline and follow-up interview, entry time was defined as the participant’s age at the median of the interval between baseline and follow-up interview because we were unable to determine the date at which the participant underwent CRC testing. Participants who contributed no follow-up time were not included in statistical models. When CRC incidence was identified from a participant’s death certificate, the date of CRC onset was set to the date of death (n = 45). Participants with missing cause of death (n = 85) were excluded from CRC mortality analyses. We evaluated the proportional hazards assumption graphically and considered it met.

    Statistical models included the following variables selected a priori as potential confounders because of known and hypothesized associations with CRC screening and CRC outcomes: self-identified race (African American, white, or other), sex, enrollment source (community health center or non–community health center), health insurance status (yes or no), annual household income (<$15 000, $15 000-24 999, $25 000-49 999, or ≥$50 000), educational level (<9 years, 9-11 years, high school, some college, or college graduate and beyond), body mass index (<18.5, 18.5-24.9, 25.0-29.9, 30.0-34.9, or ≥35.0 [calculated as weight in kilograms divided by square of height in meters]), smoking status (never, former, or current), alcohol intake (women: 0, ≤1 drink, or >1 drink per day; men: 0, ≤2, or >2 drinks per day), and family history of CRC diagnosis in a first-degree relative (yes, no, or unknown). Missing covariate data (<2% of the variable) were set to sex- and race-specific medians or modes.

    We assessed the possibility of differences in the association between CRC screening and CRC outcomes by race, sex, and household income. Possible interactions between CRC testing and factors of interest were assessed by likelihood ratio tests to compare main outcomes models with and without the addition of cross-product terms. Statistical analyses were performed using SAS statistical software, version 9.4 (SAS Institute Inc) from January 1, 2018, to October 30, 2019. A 2-sided P < .05 was considered to be statistically significant.

    Results

    This study included 47 596 participants (median baseline age, 54 years [IQR, 10.0 years]; 32 185 [67.6%] African American; 28 884 [60.7%] female; and 26 075 [54.8%] with household income <$15 000). During 478 275 person-years of follow-up, we observed 632 incident CRC cases and 204 deaths from CRC. The mean (SD) follow-up time from baseline to end of study or death was 10.6 (3.0) years (median, 11.3 years; IQR, 3.7 years). Participants who subsequently developed CRC were older and more likely to be African American, have lower household income, and have a history of smoking cigarettes than those who remained CRC free at the end of the follow-up (Table 1). The mean (SD) age at CRC diagnosis in the study was 63.3 (8.2) years, with 12 patients (1.9%) younger than 50 years at diagnosis. Patients with CRC who died during the study were more likely than patients alive at the end of study to be older at baseline, to be male, to have less educational attainment, to have a lower household income, and to report current smoking at baseline. Patients who died during the study were also less likely to report a family history of CRC and were more often diagnosed with a more advanced-stage cancer than the patients alive at the end of study.

    Of the participants eligible for screening at the baseline interview, 13 371 (35.0%) reported undergoing CRC tests at their baseline interview; however, the rate varied by modality, SES, smoking status, and race (Figure and eTable 1 and eTable 2 in the Supplement). At baseline, colonoscopy was more commonly used than sigmoidoscopy (10 413 [27.4%] vs 5978 [15.7%]), and 3020 participants (7.9%) had used both modalities. Participants enrolled in the study at a community health service were less likely to have undergone lower endoscopy than participants enrolled through the general population (10 622 [32.4%] vs 2749 [50.2%]). The CRC test rates were highest among participants with an annual household income of $50 000 or greater (1989 [55.5%]). The rate of lower endoscopy before baseline was particularly low for individuals who were uninsured (2926 [22.0%]) or current smokers (3359 [25.5%]).

    Similar to a previous report22 using initial SCCS data, at the baseline interview, colonoscopy rates were lower among African American individuals than among white individuals (6230 [24.5%] vs 3643 [32.7%], P < .001), but sigmoidoscopy rates were similar (3928 [15.5%] vs 1764 [15.8%], P = .54). Age may have been associated with the observed racial/ethnic differences in colonoscopy rates because test use increased with age at entry and, on average, white participants were slightly older at baseline than African American participants (median age, 57 years [IQR, 9 years] vs 55 years [IQR, 9 years]). However, the odds of colonoscopy at baseline were lower among African American participants than among white participants even with adjustment for age (odds ratio, 0.71; 95% CI, 0.68-0.75). In the interval between baseline and the follow-up interview, colonoscopy (12 047 [36.1%]) and FOBT (12 039 [35.0%]) were more commonly reported than sigmoidoscopy (3215 [9.4%]), and the frequency did not vary by race.

    Ever colonoscopy at baseline was associated with lower CRC risk (HR, 0.55; 95% CI, 0.44-0.70) and ever sigmoidoscopy at baseline was associated with CRC risk reduction (HR, 0.66; 95% CI, 0.51-0.87) (Table 2). Results were similar in analyses stratified by race (African American: HR, 0.65; 95% CI, 0.50-0.85; white: HR, 0.44; 95% CI, 0.27-0.70) and household income (<$15 000: HR, 0.63; 95% CI, 0.46-0.86, ≥$15 000: HR, 0.49; 95% CI, 0.35-0.69). Similarly, the inverse association between lower endoscopy before the baseline and CRC risk was consistent in all strata in analyses that were stratified by sex (men: HR, 0.64; 95% CI, 0.47-0.88; women: HR, 0.48; 95% CI, 0.36-0.63; P = .09 for interaction), household income (<$15 000: HR, 0.59; 95% CI, 0.45-0.78; ≥$15 000: HR, 0.51; 95% CI, 0.37-0.69; P = .39 for interaction), and race (white: HR, 0.48; 95% CI, 0.31-0.72; African American: HR, 0.59; 95% CI, 0.46-0.75; P = .34 for interaction) (eTable 3 in the Supplement presents incidence results for African American participants). The inverse associations between lower endoscopy and CRC risk were also observed in analyses stratified by the availability of follow-up interview data; among the participants eligible at baseline interview, lower endoscopy before baseline was associated with a 46% (95% CI, 33%-56%) decreased risk of CRC, and among the participants eligible at the follow-up interview, participation in lower endoscopy in the interval between baseline and the follow-up interview was associated with a 48% (95% CI, 31%-61%) decreased risk of CRC.

    Analysis of the association between FOBT screening in the interval between baseline and the follow-up interview revealed an inverse association (HR, 0.75; 95% CI, 0.57-0.98), although the association was less prominent than the association observed for colonoscopy (HR, 0.53; 95% CI, 0.40-0.71) and sigmoidoscopy (HR, 0.53; 95% CI, 0.32-0.86) for the same interval (Table 3).

    A high mortality rate was found among participants diagnosed with incident CRC; 315 participants (49.8%) died during the study period, with most deaths among those diagnosed with stage IV disease or missing stage at diagnosis (Table 1). The association between lower endoscopy before baseline and CRC mortality did not vary by household income or race (eTable 4 in the Supplement). However, we observed sex-specific differences in the association between lower endoscopy and CRC mortality (P = .003 for interaction). Women who had lower endoscopy before baseline had a reduced risk of CRC mortality (HR, 0.34; 95% CI, 0.19-0.60), and the inverse association between lower endoscopy before baseline and CRC mortality was consistent in analyses stratified by the availability of follow-up interview data (Table 4). Among men, ever use of lower endoscopy was not associated with CRC mortality (HR, 0.70; 95% CI, 0.43-1.14). When the data were stratified by the availability of follow-up interview data, the association became more apparently null in all subgroups, likely because of small sample size and the changing composition of the reference group.

    Among participants who were only screened between baseline and the follow-up interview, there was no association between lower endoscopy use and death from CRC among women (HR, 0.97; 95% CI, 0.49-1.91), whereas there was an association between lower endoscopy and increased CRC mortality among men (HR, 3.70; 95% CI, 1.62-8.46). In the interval between baseline and follow-up interviews, FOBT was not associated with CRC mortality in men or women (Table 3).

    We also evaluated associations with all-cause mortality and did not observe outcome modification by sex. Among participants eligible for CRC screening at baseline interview, lower endoscopy before baseline was associated with null or modest increased mortality (eTable 5 in the Supplement). Compared with participants who had never been tested at the follow-up interview, there was reduced all-cause mortality among participants who underwent colonoscopy only in the interval between baseline and the follow-up interview and participants who underwent lower endoscopy before baseline and between baseline and the follow-up interview (eTable 6 in the Supplement).

    Discussion

    In this cohort primarily composed of African American individuals and individuals with low SES, CRC test participation was low. At baseline, 63.9% of the cohort had never undergone CRC testing despite meeting criteria for screening recommendations based on age and family history of CRC. For comparison, during the same period, the US national rate of colorectal test use was approximately 50%.23 In addition, CRC test rates were lowest among the groups who most often experience adverse CRC outcomes: African American individuals, individuals without health insurance, and those who smoke. Our study results suggest that efforts in public health may improve CRC screening rates with implementation strategies that focus on these high-risk groups and that aim to lower barriers to CRC screening. These barriers include limited access to gastroenterologists to perform screening or diagnostic colonoscopy and cost-prohibitive out-of-pocket expenses.24

    Although CRC test rates were low, we did observe strong inverse associations between CRC test modalities and CRC incidence. Our results using data from traditionally underserved populations are in line with previous studies3-8 that found that CRC screening by visual- and stool-based screening modalities was associated with reduced CRC incidence and mortality. Specifically, the US Preventive Services Task Force asserts that the CRC test modalities examined in the present study (colonoscopy, sigmoidoscopy, and FOBT) have differing abilities to detect CRC, but all have been shown to be associated with reduced CRC incidence and mortality in the US general population.8 Compared with the cohort data from our study that used information from individuals who primarily have low SES and are African American, data from a cohort that primarily includes individuals with higher SES who are white indicated a similar association of lower endoscopy use and CRC incidence. A report5 from the Nurses’ Health Study and the Health Professionals Follow-up Study found that lower endoscopy was associated with reduced CRC incidence (HR, 0.44; 95% CI, 0.38-0.52).

    We also observed an inverse association between CRC testing and CRC mortality, although the association was most apparent among women. Limited sample size in our sex-specific analyses may have contributed to wide CIs and lack of statistically significant findings among men. Among men, use of lower endoscopy between baseline and the follow-up interview was associated with increased CRC mortality, raising the possibility that these participants may have undergone lower endoscopy as a diagnostic test rather than as a screening test. Similar to a previous studies,8 we did not find consistent associations between CRC testing and all-cause mortality.

    Of importance, the risk-lowering benefits of CRC screening associated with CRC risk and mortality did not vary by race or household income. Our findings suggest that increasing uptake of CRC screening among African American individuals may reduce the sizable racial disparity in CRC incidence and mortality. Factors hypothesized to contribute to the racial disparities in CRC outcomes include differences in the prevalence of risk factors25 and CRC screening.10-12 In general, African American individuals, individuals with lower household income, and uninsured individuals are less likely to be concordant with CRC screening guidelines.22,26-28 In fact, access to and use of health care may play a more important role in CRC racial disparities than other risk factors, as evidenced by data from the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial and others.2,14,29 For instance, in PLCO Cancer Screening Trial participants randomized to sigmoidoscopy screening, the screening test was equally likely to detect abnormal colorectal findings among African American and white individuals; however, African American participants were less likely than their white counterparts to follow up for additional diagnostic tests after the abnormal sigmoidoscopy findings.29 The inability of patients to afford insurance cost sharing and the logistical difficulties associated with transportation and scheduling are hypothesized to contribute to the lower rates of follow-up health care after abnormal CRC screening findings.2,14,29

    Strengths and Limitations

    This study has a number of strengths. The SCCS is a large, prospective cohort study with systematic follow-up to identify incident CRC diagnoses. The cohort covers an underserved at-risk population that is seldom included in large numbers in other investigations.

    This study has limitations. Primarily, we used a crude measure of yes or no to classify screening exposure, instead of determining whether participants received screening under expert-recommended intervals (ie, every 10 years for colonoscopy). In addition, we used self-reported CRC tests instead of the criterion standard of medical records. Prior studies30,31 suggest that participants self-report procedures at higher rates than the rates measured via review of health records. However, despite the limitations of potential overreporting of the exposure and lack of information on adherence to screening interval, we were able to detect inverse associations between self-reported CRC tests and CRC outcomes.

    Another limitation of the study is the inability to distinguish between lower endoscopies performed for screening vs diagnostic purposes. Our CRC screening exposure variable may include diagnostic tests for individuals already exhibiting symptoms of CRC, which may have attenuated the results in the CRC incidence analysis and contributed to the CRC mortality associations observed with lower endoscopy use between baseline and the follow-up interview among men. A possible future direction of this work may be to incorporate data from health records or health care claims to verify the timing and indication for the colorectal test encounter.

    Conclusions

    This study revealed associations between use of lower endoscopy or FOBT and reduced CRC risk and a significant association between lower endoscopy and CRC mortality risk reduction among women. The lowest CRC testing rates were found among individuals who lacked insurance, had low SES, were current smokers, or were African American. Our results may be applicable to other underserved populations, particularly groups with low SES as defined by educational level and income. Promotion of CRC screening targeted at these high-risk groups may improve CRC disparities. In addition, lowering financial barriers and increasing access to CRC screening for the uninsured and underinsured may increase participation in screening and ultimately reduce CRC racial and socioeconomic disparities in incidence and mortality.

    Back to top
    Article Information

    Accepted for Publication: October 31, 2019.

    Published: December 20, 2019. doi:10.1001/jamanetworkopen.2019.17995

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2019 Warren Andersen S et al. JAMA Network Open.

    Corresponding Author: Shaneda Warren Andersen, PhD, Department of Population Health Sciences, School of Medicine and Public Health, University of Wisconsin, 610 Walnut St, WARF Office Building, Ste 1007B, Madison, WI 53726 (snandersen@wisc.edu).

    Author Contributions: Dr Warren Andersen had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Warren Andersen, Lipworth, Murff, Zheng.

    Acquisition, analysis, or interpretation of data: Warren Andersen, Blot, Lipworth, Steinwandel, Zheng.

    Drafting of the manuscript: Warren Andersen.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Warren Andersen, Steinwandel.

    Obtained funding: Blot, Zheng.

    Administrative, technical, or material support: Warren Andersen, Zheng.

    Supervision: Warren Andersen, Murff.

    Conflict of Interest Disclosures: Dr Warren Andersen reported receiving grants from the National Cancer Institute during the conduct of the study. Dr Blot reported receiving grants from the National Institutes of Health during the conduct of the study. No other disclosures were reported.

    Funding/Support: This study was supported by the University of Wisconsin at Madison, Office of Vice Chancellor for Research and Graduate Education with funding from the Wisconsin Alumni Research Foundation and grant R00 CA207848 from the National Cancer Institute (Dr Warren Andersen). The Southern Community Cohort Study (SCCS) is funded by grant R01 CA92447 (Drs Blot and Zheng) from the National Cancer Institute at the National Institutes of Health, including special allocations from grant 3R01 CA092447-08S1 from the American Recovery and Reinvestment Act.

    Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.

    Disclaimer: The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control or Prevention (CDC), the Mississippi Cancer Registry, or the West Virginia Cancer Registry.

    Additional Information: Data from the SCCS used in this study were provided by the Alabama Statewide Cancer Registry, Kentucky Cancer Registry, Tennessee Department of Health, Office of Cancer Surveillance, Florida Cancer Data System, North Carolina Central Cancer Registry, North Carolina Division of Public Health, Georgia Comprehensive Cancer Registry, Louisiana Tumor Registry, Mississippi Cancer Registry, South Carolina Central Cancer Registry, Virginia Department of Health, Virginia Cancer Registry, and the Arkansas Department of Health Cancer Registry. The Arkansas Central Cancer Registry is fully funded by a grant from National Program of Cancer Registries of the CDC. Data on SCCS cancer cases from Mississippi were collected by the Mississippi Cancer Registry, which participates in the National Program of Cancer Registries of the CDC. Data on SCCS cancer cases from West Virginia were provided by the West Virginia Cancer Registry.

    References
    1.
    American Cancer Society.  Cancer Facts & Figures 2018. Atlanta, GA: American Cancer Society; 2018.
    2.
    American Cancer Society.  Colorectal Cancer Facts & Figures 2017-2019. Atlanta, GA: American Cancer Society; 2017.
    3.
    Elmunzer  BJ, Hayward  RA, Schoenfeld  PS, Saini  SD, Deshpande  A, Waljee  AK.  Effect of flexible sigmoidoscopy-based screening on incidence and mortality of colorectal cancer: a systematic review and meta-analysis of randomized controlled trials.  PLoS Med. 2012;9(12):e1001352. doi:10.1371/journal.pmed.1001352PubMedGoogle Scholar
    4.
    Weissfeld  JL, Schoen  RE, Pinsky  PF,  et al; PLCO Project Team.  Flexible sigmoidoscopy in the randomized Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial: added yield from a second screening examination.  J Natl Cancer Inst. 2012;104(4):280-289. doi:10.1093/jnci/djr549PubMedGoogle ScholarCrossref
    5.
    Nishihara  R, Wu  K, Lochhead  P,  et al.  Long-term colorectal-cancer incidence and mortality after lower endoscopy.  N Engl J Med. 2013;369(12):1095-1105. doi:10.1056/NEJMoa1301969PubMedGoogle ScholarCrossref
    6.
    Shaukat  A, Mongin  SJ, Geisser  MS,  et al.  Long-term mortality after screening for colorectal cancer.  N Engl J Med. 2013;369(12):1106-1114. doi:10.1056/NEJMoa1300720PubMedGoogle ScholarCrossref
    7.
    Shroff  J, Thosani  N, Batra  S, Singh  H, Guha  S.  Reduced incidence and mortality from colorectal cancer with flexible-sigmoidoscopy screening: a meta-analysis.  World J Gastroenterol. 2014;20(48):18466-18476. doi:10.3748/wjg.v20.i48.18466PubMedGoogle ScholarCrossref
    8.
    Lin  JS, Piper  MA, Perdue  LA,  et al.  Screening for colorectal cancer: updated evidence report and systematic review for the US Preventive Services Task Force.  JAMA. 2016;315(23):2576-2594. doi:10.1001/jama.2016.3332PubMedGoogle ScholarCrossref
    9.
    Rex  DK, Johnson  DA, Anderson  JC, Schoenfeld  PS, Burke  CA, Inadomi  JM; American College of Gastroenterology.  American College of Gastroenterology guidelines for colorectal cancer screening 2009 [published correction appears in Am J Gastroenterol 2009;104(6):1613].  Am J Gastroenterol. 2009;104(3):739-750. doi:10.1038/ajg.2009.104PubMedGoogle ScholarCrossref
    10.
    Franks  P, Muennig  P, Lubetkin  E, Jia  H.  The burden of disease associated with being African-American in the United States and the contribution of socio-economic status.  Soc Sci Med. 2006;62(10):2469-2478. doi:10.1016/j.socscimed.2005.10.035PubMedGoogle ScholarCrossref
    11.
    Williams  DR, Mohammed  SA, Leavell  J, Collins  C.  Race, socioeconomic status, and health: complexities, ongoing challenges, and research opportunities.  Ann N Y Acad Sci. 2010;1186:69-101. doi:10.1111/j.1749-6632.2009.05339.xPubMedGoogle ScholarCrossref
    12.
    Grubbs  SS, Polite  BN, Carney  J  Jr,  et al.  Eliminating racial disparities in colorectal cancer in the real world: it took a village.  J Clin Oncol. 2013;31(16):1928-1930. doi:10.1200/JCO.2012.47.8412PubMedGoogle ScholarCrossref
    13.
    Phelan  JC, Link  BG, Diez-Roux  A, Kawachi  I, Levin  B.  “Fundamental causes” of social inequalities in mortality: a test of the theory.  J Health Soc Behav. 2004;45(3):265-285. doi:10.1177/002214650404500303PubMedGoogle ScholarCrossref
    14.
    Doubeni  CA, Corley  DA, Zauber  AG.  Colorectal cancer health disparities and the role of US law and health policy.  Gastroenterology. 2016;150(5):1052-1055. doi:10.1053/j.gastro.2016.03.012PubMedGoogle ScholarCrossref
    15.
    Hall  WJ, Chapman  MV, Lee  KM,  et al.  Implicit racial/ethnic bias among health care professionals and its influence on health care outcomes: a systematic review.  Am J Public Health. 2015;105(12):e60-e76. doi:10.2105/AJPH.2015.302903PubMedGoogle ScholarCrossref
    16.
    Signorello  LB, Hargreaves  MK, Steinwandel  MD,  et al.  Southern community cohort study: establishing a cohort to investigate health disparities.  J Natl Med Assoc. 2005;97(7):972-979.PubMedGoogle Scholar
    17.
    Signorello  LB, Hargreaves  MK, Blot  WJ.  The Southern Community Cohort Study: investigating health disparities.  J Health Care Poor Underserved. 2010;21(1)(suppl):26-37. doi:10.1353/hpu.0.0245PubMedGoogle ScholarCrossref
    18.
    US Preventive Services Task Force.  Screening for colorectal cancer: recommendation and rationale.  Ann Intern Med. 2002;137(2):129-131. doi:10.7326/0003-4819-137-2-200207160-00014PubMedGoogle ScholarCrossref
    19.
    US Preventive Services Task Force.  Screening for colorectal cancer: US Preventive Services Task Force recommendation statement.  Ann Intern Med. 2008;149(9):627-637. doi:10.7326/0003-4819-149-9-200811040-00243PubMedGoogle ScholarCrossref
    20.
    Levin  B, Lieberman  DA, McFarland  B,  et al; American Cancer Society Colorectal Cancer Advisory Group; US Multi-Society Task Force; American College of Radiology Colon Cancer Committee.  Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology.  Gastroenterology. 2008;134(5):1570-1595. doi:10.1053/j.gastro.2008.02.002PubMedGoogle ScholarCrossref
    21.
    Winawer  S, Fletcher  R, Rex  D,  et al; Gastrointestinal Consortium Panel.  Colorectal cancer screening and surveillance: clinical guidelines and rationale-Update based on new evidence.  Gastroenterology. 2003;124(2):544-560. doi:10.1053/gast.2003.50044PubMedGoogle ScholarCrossref
    22.
    Murff  HJ, Peterson  NB, Fowke  JH,  et al.  Colonoscopy screening in African Americans and whites with affected first-degree relatives.  Arch Intern Med. 2008;168(6):625-631. doi:10.1001/archinte.168.6.625PubMedGoogle ScholarCrossref
    23.
    Hall  IJ, Tangka  FKL, Sabatino  SA, Thompson  TD, Graubard  BI, Breen  N.  Patterns and trends in cancer screening in the United States.  Prev Chronic Dis. 2018;15:E97. doi:10.5888/pcd15.170465PubMedGoogle Scholar
    24.
    Doubeni  CA, Laiyemo  AO, Reed  G, Field  TS, Fletcher  RH.  Socioeconomic and racial patterns of colorectal cancer screening among Medicare enrollees in 2000 to 2005.  Cancer Epidemiol Biomarkers Prev. 2009;18(8):2170-2175. doi:10.1158/1055-9965.EPI-09-0104PubMedGoogle ScholarCrossref
    25.
    National Center for Health Statistics.  Health, United States, 2011: With Special Feature on Socioeconomic Status and Health. Hyattsville, MD: National Center for Health Statistics; 2012.
    26.
    Peterson  NB, Murff  HJ, Fowke  JH,  et al.  Use of colonoscopy and flexible sigmoidoscopy among African Americans and whites in a low-income population.  Prev Chronic Dis. 2008;5(1):A28.PubMedGoogle Scholar
    27.
    Bernardo  BM, Gross  AL, Young  G,  et al.  Predictors of colorectal cancer screening in two underserved U.S. populations: a parallel analysis.  Front Oncol. 2018;8:230. doi:10.3389/fonc.2018.00230PubMedGoogle ScholarCrossref
    28.
    Cohen  SS, Murff  HJ, Signorello  LB, Blot  WJ.  Obesity and colorectal cancer screening among black and white adults.  Cancer Causes Control. 2012;23(5):709-716. doi:10.1007/s10552-012-9940-yPubMedGoogle ScholarCrossref
    29.
    Laiyemo  AO, Doubeni  C, Pinsky  PF,  et al.  Race and colorectal cancer disparities: health-care utilization vs different cancer susceptibilities.  J Natl Cancer Inst. 2010;102(8):538-546. doi:10.1093/jnci/djq068PubMedGoogle ScholarCrossref
    30.
    Reiter  PL, Katz  ML, Oliveri  JM, Young  GS, Llanos  AA, Paskett  ED.  Validation of self-reported colorectal cancer screening behaviors among Appalachian residents.  Public Health Nurs. 2013;30(4):312-322. doi:10.1111/phn.12038PubMedGoogle ScholarCrossref
    31.
    Shokar  NK, Vernon  SW, Carlson  CA.  Validity of self-reported colorectal cancer test use in different racial/ethnic groups.  Fam Pract. 2011;28(6):683-688. doi:10.1093/fampra/cmr026PubMedGoogle ScholarCrossref
    ×