Colorectal cancer (CRC) is the world's third most diagnosed cancer and the second highest in mortality (1,2). Studies have demonstrated that screening for CRC reduces CRC mortality and is cost-effective if not cost-saving (3,4). Since the introduction of screening, improvements have been made to guidelines globally to specifically target the high-risk groups, and it has been established that these risk-tailored screening programs have improved incidence and outlook of CRC (5,6). However, the incidence of CRC at young ages is rising, and so, it becomes increasingly important to identify specific CRC risk factors for young adults so that high-risk individuals are screened earlier or at the very least made aware of their risk. Most studies investigating the CRC risk factors were interested in CRC cases in all ages (7). As a result, it is unknown how risk factors and their estimates differ in different age groups.
Family history of CRC has been recognized as an established risk factor of CRC (8–10). Diabetes mellitus (DM) and CRC share some risk factors, including obesity and sedentary lifestyle and metabolic disturbances (11–13). Meta-analyses of epidemiological studies have suggested that a diagnosis of DM is also associated with an increased risk of CRC (14,15). However, DM has yet to be established as a recognized CRC risk factor (16). The magnitude of associations between DM, family history of CRC, and increased risk of CRC, a possible interaction between these associations, and effect of age at diagnosis on these associations are not clear. We aimed to answer the abovementioned questions with the world's largest study of its kind using several Swedish register-based nationwide data sets with a long-term follow-up. We particularly intended to evaluate how personal history of DM and family history of CRC affect the risk of sporadic and familial CRC at different ages.
MATERIAL AND METHODS
The Swedish family cancer data sets were used in this study, which encompass a multitude of registered data in Sweden from 1958 to 2015. The data sets were combined through the linkage of records from the Multigeneration Register, Swedish Cancer Registry, national censuses, and Death Registry using lifetime unique national registration numbers assigned to individuals residing in Sweden. The Swedish Cancer Registry, which was established in 1958, contains detailed cases of cancer diagnostic reports by physicians. Cancer records are reported in accordance with the International Classifications of Diseases (ICD) revisions 7 through 10. The Multigeneration Register contains genealogical information of Swedish residents who were born after 1931 (offspring) and their parents. Further linkage established with the national census includes information on socioeconomic status, residential area, migration, and other demographic statistics. All these culminate into the largest study data set of its kind, containing about 13 million individuals with genealogy information and data on over 160,000 CRC patients.
Using national identification numbers, we linked the family cancer data sets to data from the Swedish inpatient and outpatient registers. Inpatient (hospital) records from 1964 to 2015 and outpatient records from 2001 to 2015 for diabetic patients were available. Data on DM were extracted based on ICD codes (ICD-7:260; ICD-8:250; ICD-9:250; ICD-10:E10, E11, E13, and E14). Information on the DM type was available since 1997. Patients with both type 1 and type 2 DM diagnoses were excluded from the analysis by type of DM. Those with pregnancy-related and malnutrition-related diabetes were also excluded from this study. We also established the family history of CRC in first-degree relatives (FDRs; parents, siblings, or children). Both DM personal history and CRC family history were treated as time-dependent (time-varying or dynamic) variables in our risk calculations, that is, individuals were treated as cases of DM from the year in which they received a diagnosis and were recorded as noncases until that point. Similarly, an individual was considered having an FDR with CRC from the year in which the relative was diagnosed. Considering risk groups with exposures of interest as dynamic (time-dependent) variables allowed standardized incidence ratios (SIRs) to reflect possible risk estimates in real clinical settings (17). Because the disease onset and family history are not established at birth and change throughout the life, these dynamic (time-varying) variables were considered as such in our analyses.
Starting follow-up time in the study was defined for each individual as the most recent of birth year, immigration year, or 1964, which is the start of the inpatient register. The end of the follow-up was characterized as the earliest of CRC diagnosis date, death year, emigration year, or 2015, which is the year of last entry of all data sets used. The risk of CRC by DM history and CRC family history was quantified by SIRs. Stratified analyses were conducted according to the age at CRC diagnosis (<50 and ≥50 years). SIRs were expressed as ratios between the observed and expected CRC cases. We calculated the expected number of cases using strata-specific person-years in people with DM and/or family history of CRC multiplied by strata-specific incidence rates in those without a DM diagnosis or family history of CRC. All SIRs were adjusted for the following covariates: sex, age, calendar period (in 5-year increments from 1964 onward except for the first and the last periods, i.e., 1964–1969 and 2000–2015), socioeconomic status, and residential area. As a sensitivity analysis, we adjusted SIRs for having ever been hospitalized for obesity, alcoholism, or chronic obstructive pulmonary disorder. We additionally conducted a sensitivity analysis by the exclusion of patients with inflammatory bowel diseases (IBDs), which are established CRC risk factors (18), to ensure that they did not confound our risk estimates.
We calculated cumulative risks to determine the risk of various groups for different age intervals i.e. the risk of developing CRC for individuals with/without DM from birth to age 29, 39, and so on. All statistical analyses were conducted using SAS version 9.4 (by SAS Institute, Cary, NC). Our data were pseudonymized with no risk of identification of patients, and thus, no informed consent was needed. The study protocol was approved by the Lund Regional Ethics Committee (2012/795).
A total of 12,614,256 individuals with available genealogy information were included in this study. Among them 559,375 (4.4%) diabetic patients were identified and 101,135 (18%) of them were diagnosed before the age of 50 years. A total of 162,226 (1.3%) patients with CRC were identified and included in the study. The median age of CRC diagnosis was 75 years for those with DM diagnosis age greater than or equal to 50 years, 59 years for those with DM diagnosis age greater than 50 years (51 years for those with DM diagnosis age greater than 30 years, 44 years for those with DM diagnosis age greater than 20 years), and 71 years for those without DM or family history of CRC (Table 1). About 21% of all CRC cases in the database occurred before the age of 60 years.
SIR for sporadic CRC
Compared with those without family history of CRC and without personal history of DM, diabetic patients diagnosed before the age of 50 years without family history of CRC showed a 1.9-fold (95% CI: 1.5–2.3) increased risk of sporadic CRC diagnosed before the age of 50 years (Table 1). The SIRs stratified by age at CRC diagnosis showed that for the same group, the increased risk of CRC at/after age 50 years was 1.3-fold (95% CI: 1.2–1.4). DM diagnosis at all ages showed higher risk of sporadic CRC diagnosed before age 50 years than that of sporadic CRC diagnosed at or after age 50 years, except for DM diagnosis before age 20 years (SIR = 1.2, 95% CI: 0.7–1.9). Highest risk of sporadic CRC was found when DM was diagnosed at age 40–49 years, and CRC was diagnosed before age 50 years (SIR = 3.6, 95% CI: 2.8–4.5).
SIR for familial CRC
Having 1 FDR with CRC without a personal history of DM diagnosis was associated with a 2.4-fold increased risk of CRC diagnosed before age 50 years (95% CI: 2.2–2.6) and 1.6-fold increased risk of CRC diagnosed at/after age 50 years (Table 1). Having both 1 FDR with CRC and a personal history of DM diagnosed before age 50 years was associated with a 2.3-fold increased risk of CRC (95% CI: 1.7–2.9), higher than the individual risk of those with 1 FDR with CRC (SIR = 1.6, 95% CI: 1.6–1.7) or of those with DM diagnosis before age 50 years (SIR = 1.4, 95% CI: 1.3–1.5). The risk of CRC diagnosed before age 50 years for those who have both 1 FDR with CRC and a personal history of DM diagnosed before age 50 years was 6.9-fold increased (95% CI: 3.8–11) and the increased risk for CRC diagnosed at/after age 50 years was 1.9-fold (95% CI: 1.4–2.5). The highest risk of CRC diagnosis before age 50 years among diabetic patients was observed for those with 1 FDR with CRC and a DM diagnosis at age 40–49 years, with about 12-fold increase compared with risk in those without DM and family history of CRC (95% CI: 4.7–24). Further adjustments for history of hospitalization for obesity, alcoholism, and chronic obstructive pulmonary disorder did not affect the risk estimates substantially (data not shown).
Absolute (cumulative) risk
Using absolute (cumulative) risk, we showed that the risk of developing CRC before age 50 years in the general population was 0.2% (95% CI: 0.2%–0.2%; Table 2). For diabetic patients diagnosed before age 50 years who had no family history of CRC in the same age interval, the risk was 0.4% (95% CI: 0.3%–0.4%), and for those who had no DM diagnosis and 1 FDR with CRC, the risk was 0.5% (95% CI: 0.5%–0.5%). For developing sporadic CRC before age 50 years, the highest risk observed was for those with both a DM diagnosis at age 40–49 years (0.5%, 95% CI: 0.3%–0.7%).
SIRs by type of DM
When we limited our study to period 1997–2015 to stratify the analyses by definite cases of type 1 and type 2 DM, we found that those with type 2 DM had 3.5-fold elevated risk of sporadic CRC before age 50 years (95% CI 2.3–5.1), but we did not find elevated risk of sporadic CRC before age 50 years in those with type 1 DM (SIR = 1.0, 95% CI 0.6–1.7; see Table S1, Supplementary Digital Content 1, http://links.lww.com/AJG/B533). The association of both types of DM with late-onset CRC (diagnosed at/after 50) was similarly weak and not statistically significant. Type 2 diabetic patients who also had family history of CRC showed an 18-fold increased risk of early-onset CRC (95% CI 5.9–42), whereas type 1 diabetic patients with a similar family history showed an 8.6-fold increased risk of early-onset CRC (95% CI 2.3–21).
SIRs by CRC subsite
Those with a family history of CRC but without a personal history of diabetes had only 2.4-fold increased risk of early-onset CRC. After stratification of CRC into proximal colon, distal colon, and rectal cancers, we did not find significant difference in the risk association (95% CIs overlapped; see Supplementary Table S2, Supplementary Digital Content 1, http://links.lww.com/AJG/B533). SIRs for diabetic patients were highest for early-onset distal colon cancer (SIR = 2.3, 95% CI: 1.7–3.2), followed by the proximal colon cancer (SIR = 1.8, 95% CI: 1.4–2.4) and rectal cancer (SIR = 1.8, 95% CI: 1.3–2.0) diagnosed before age 50 years. SIR for cancers in all 3 regions were higher before age 50 years, except for those with DM diagnosed before age 30 years with the smallest samples size.
SIRs after the exclusion of patients with IBD
We also excluded patients with IBD from our data set and found no substantial changes in our risk estimates (see Table S3, Supplementary Digital Content 1, http://links.lww.com/AJG/B533).
Establishing the world's largest study of its kind by using several Swedish register-based nationwide data sets with long-term follow-ups, we found novel information about associations between DM, family history of CRC, and increased risk of CRC. Having DM was associated with increased risk of CRC in the magnitude close to having a history of CRC in 1 FDR. Associations of both DM and family history of CRC with increased CRC risk were most prominent in young adults and in concordant age groups.
We showed that having family history of CRC was associated with an increased risk of CRC before age 50 years, 2.4 times that of the population without a family history of CRC. Similarly, having a DM diagnosis by age 50 years also was associated with almost 2-fold increased risk of CRC than the population, with those diagnosed after age 40 having an even higher risk at 3.6-fold. Furthermore, we observed that a combined personal history of DM and a family history of CRC was associated with a higher risk of CRC than each group individually at nearly 7 times higher risk of CRC in those without such histories. The substantial increase in CRC risk in presence of DM and family history of CRC suggests an interaction between these 2 factors, which warrants further investigation.
Our finding regarding elevated risk of early-onset CRC in diabetic patients could be even more important when considering that early-onset CRC might be a unique subset of CRC (19). Besides CRC family history, male sex, and having rare genetic syndromes such as hereditary nonpolyposis CRC or IBD have especially been associated with early-onset CRC (20–23). The aforementioned risk factors have been mentioned in the CRC screening guidelines, whereas DM has not. A meta-analysis of the risk of CRC in diabetic patients demonstrated a positive association between these 2 conditions (SIR 1.21, 95% CI: 1.02–1.42 (14)). However, even the meta-analysis found a disparity in the observed associations. Certain studies found the association of DM and CRC with a risk ratio as high as 2.05 (95% CI: 1.69–2.48) and others as low as 1.05 (95% CI: 0.94–1.18), whereas our risk estimate was in between with an 1.6-fold increased lifetime risk of CRC in diabetic patients diagnosed at any age. Variation in these estimates is likely because of the differences in cohort size, age at diagnosis of DM and CRC, management of CRC family history, and DM medication. Studies have also demonstrated that prediabetes, the precursor to type 2 DM, has association with increased risk of CRC, suggesting that the lifestyle factors and glycemic disturbances associated with type 2 DM may be mediating the association between DM and CRC risk (24). The association between family history and CRC has been well documented with a risk ratio for having any FDRs with CRC being 1.87 (95% CI: 1.68–2.09) as reported by a meta-analysis, whereas those with only 1 diagnosed FDR had a risk ratio 1.76 (25). Nonetheless, to date, no study has investigated the association between risk of CRC and different combinations of personal history of DM at different ages and family history of CRC.
In light of increasing CRC incidence among younger adults, studies have investigated whether the starting age of CRC screening for the average-risk population should be lowered (26,27), and several countries have adjusted their screening guidelines toward a lower starting age (28,29). There are potential issues with this strategy such as higher costs of screening because more people in the population are eligible, and still, there will be people with a higher risk at younger ages than the new benchmark starting age of mass screening who are overlooked (30). We found that individuals with DM before age 50 years received CRC diagnoses at earlier ages (median age of 59 years) than the general population (age 71 years) or even those without DM but with a family history of CRC (age 65 years). When also acknowledging that the average time for an adenomatous polyp to develop into CRC is greater than 10 years (31), screening diabetic patients before age 50 years (recommended starting age of CRC screening in most countries) should be considered. However, further studies are warranted to find out the underlying mechanism of our findings and to elucidate underlying the risk factors of DM that at the same time strongly predispose to CRC. Common risk factors of DM and CRC such as higher BMI and sedentary lifestyle seem to play important roles in this association. It has been also suggested that the abnormally high levels of insulin and glucose may create an environment in the colon and rectum that promotes the development and growth of CRC (32–35). The prevalence of DM in Sweden is roughly 4.5% compared with around 9% in the United States (36). Average BMI in Sweden is 25.8 as of 2014, whereas in the United States, it is 28.8 (37). Because DM is more common in the United States and accompanied with other CRC risk factors such as high BMI, CRC screening for diabetic patients may be even more important in the United States, particularly given our finding that type 2 DM is associated with a pronounced risk of early-onset CRC.
One of the advantages of our study was using the world's largest nationwide register-based family-disease data sets with long-term follow-ups. The structures of these linked data sets mitigated selection and recall biases. For example, the information on family history, which was established by linking the cancer registry and genealogy data set, allowed for accurate information about family history avoiding the bias of self-reported family history and information bias. Inclusion of information on age at DM diagnosis enabled us to exclude the possibility of a reverse association between CRC and DM. Knowing the year of DM diagnosis and CRC diagnosis in the relatives enabled us to allocate person-years precisely based on the dynamic nature of such histories. Analysis by subsite of CRC showed that all regions of the colon and rectum were associated with increased risk of CRC, internally validating the association between DM and CRC risk (see Supplementary Table S2, Supplementary Digital Content 1, http://links.lww.com/AJG/B533).
Our method of analyses for the purpose of this study is superior to that in most register-based cohort studies that traditionally used personal/familial history independent from the time of diagnosis (38,39). Our findings, which are according to the dynamic method of personal and family history assessment, provide risk estimates that are more applicable to the real-world scenarios and can be used in risk prediction and counseling for screening. Another strength of this study is the consideration of absolute cumulative risk. This is a more in-depth look at CRC incidence in the population as compared to just the use of relative risk measures, such as SIRs or hazard ratios, used by most population-based studies (14,15).
We did not have information on the type of DM for those who were diagnosed with DM before 1997. Therefore, we performed a sensitivity analysis limited to the follow-up period 1997–2015 to find out the association between type of DM and risk of early- and late-onset CRC. In those who had no family history of CRC, we found a positive association between personal history of type 2 DM and increased risk of early-onset CRC (diagnosed before age 50 years), but not with personal history of type 1 DM (see Supplementary Table S1, Supplementary Digital Content 1, http://links.lww.com/AJG/B533). Because type 1 DM happens at younger ages, most patients with type 1 DM who were diagnosed in the past 2 decades (1997–2015) did not reach the age of CRC cancer by end of our study (2015). If we classify DM diagnoses before age 30 years as type 1 [a reliable criterion with proven strong predictive value (40)], type 1 diabetic patients also tended to have 40% higher risk of early-onset proximal colon cancer (SIR = 1.4, 95% CI 0.9–2.1). Further information on other possible limitations of this study including lack of detailed data on DM treatment with insulin or metformin, end-organ manifestations of DM, colonoscopy data, and lifestyle factors can be found in the eDiscussion (see Supplementary Digital Content 1, http://links.lww.com/AJG/B533). In brief, it is quite unlikely that the association we found between increased risk of early-onset CRC and DM was confounded by lack of abovementioned data, and if any, their effects were toward the dilution of the association (see eDiscussion, Supplementary Digital Content 1, http://links.lww.com/AJG/B533).
A study, which compared the adenoma detection rate among cohorts of patients aged 40–49 years with DM and without DM found that DM is associated with about 3-fold higher risk of colorectal adenomas in patients aged 40–49 years (34). This proposes a likely real association between DM and early-onset colorectal tumors (and tumor risk advancement by DM) rather than a detection bias, which has been further discussed in the Supplementary eDiscussion.
The present study provided evidence that individuals with a personal history of DM are at an increased risk for CRC, in particular early-onset CRC. We found that young diabetic patients even without family history of CRC are at an increased risk for CRC in a similar magnitude to individuals with a family history of CRC. These findings warrant further studies on harms, benefits, and cost-effectiveness of CRC screening in diabetic patients at earlier ages than in the general population, especially in those with type 2 diabetes.
CONFLICTS OF INTEREST
Guarantor of the article: Mahdi Fallah, MD, PhD.
Specific author contributions: M.F. and E.K.: Equal contribution. M.F. and E.K.: Study concept and design. K.S., J.S., M.F., and E.K.: Acquisition of data. U.A.K., Y.T., and M.F.: Statistical analysis. U.A.K., E.K., and M.F.: Interpretation of data. U.A.K., E.K., M.F.: Drafting of the manuscript. All co-authors: Critical revision of the manuscript for important intellectual content. U.A.K. and M.F.: Obtained funding. E.K. and M.F.: Study supervision.
Financial support: U.A.K. has been supported by a scholarship from the Helmholtz Association of German Research Centers. Y.T. has been supported by the China Scholarship Council. Funders 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 decision to submit the manuscript for publication.
Potential competing interests: None to report.
WHAT IS KNOWN
- ✓ CRC incidence has been increasing prominently in young adults below the age of screening.
- ✓ People with a family history of CRC are at increased risk later in life.
WHAT IS NEW HERE
- ✓ In diabetic patients, the risk of CRC was elevated at all ages, especially for CRC before age 50 years.
- ✓ The risk of developing CRC before age 50 years in diabetic patients was similar to having a family history of CRC.
- ✓ Diabetic patients with a family history of CRC have roughly 7-fold elevated risk of CRC before age 50 years.
- ✓ Significantly elevated risk of CRC in young adults with diabetes makes diabetic patients ideal candidates for risk-adapted CRC screening.
1. Arnold M, Sierra MS, Laversanne M, et al. Global patterns and trends in colorectal cancer incidence and mortality. Gut 2017;66:683–91.
2. Bray F, Ferlay J, Soerjomataram I, et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018;68:394–424.
3. Lansdorp-Vogelaar I, Knudsen AB, Brenner H. Cost-effectiveness of colorectal cancer screening. Epidemiol Rev 2011;33:88–100.
4. Patel SS, Kilgore ML. Cost effectiveness of colorectal cancer screening strategies. Cancer Control 2015;22:248–58.
5. American Cancer Society. Cancer Facts & Figures 2018 American Cancer Society: Atlanta, 2018.
6. Cairns SR, Scholefield JH, Steele RJ, et al. Guidelines for colorectal cancer screening and surveillance in moderate and high risk groups (update from 2002). Gut 2010;59:216–7.
7. Johnson CM, Wei C, Ensor JE, et al. Meta-analyses of colorectal cancer risk factors. Cancer Causes Control 2013;24:1207–22.
8. Henrikson NB, Webber EM, Goddard KA, et al. Family history and the natural history of colorectal cancer: Systematic review. Genet Med 2015;17:702–12.
9. Butterworth AS, Higgins JP, Pharoah P. Relative and absolute risk of colorectal cancer for individuals with a family history: A meta-analysis. Eur J Cancer 2006;42:216–27.
10. Frank C, Fallah M, Sundquist J, et al. Population landscape of familial cancer. Sci Rep 2015;5:12891.
11. Edwards BK, Ward E, Kohler BA, et al. Annual report to the nation on the status of cancer, 1975-2006, featuring colorectal cancer trends and impact of interventions (risk factors, screening, and treatment) to reduce future rates. Cancer 2010;116:544–73.
12. Mokdad AH, Ford ES, Bowman BA, et al. Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. JAMA 2003;289:76–9.
13. Slattery ML. Physical activity and colorectal cancer. Sports Med 2004;34:239–52.
14. Guraya SY. Association of type 2 diabetes mellitus and the risk of colorectal cancer: A meta-analysis and systematic review. World J Gastroenterol 2015;21:6026–31.
15. Larsson S, Orsini Nicola, Wolk A. Diabetes mellitus and risk of colorectal cancer: A meta-analysis. J Natl Cancer Inst 2005;97:1679–87.
16. Schreuders EH, Ruco A, Rabeneck L, et al. Colorectal cancer screening: A global overview of existing programmes. Gut 2015;64:1637–49.
17. Mukama T, Kharazmi E, Sundquist K, et al. Familial risk of breast cancer by dynamic, accumulative, and static definition of family history. Cancer [Epub ahead of print March 10, 2020.]
18. Keller DS, Windsor A, Cohen R, et al. Colorectal cancer in inflammatory bowel disease: Review of the evidence. Tech Coloproctol 2019;23:3–13.
19. Silla IO, Rueda D, Rodriguez Y, et al. Early-onset colorectal cancer: A separate subset of colorectal cancer. World J Gastroenterol 2014;20:17288–96.
20. Ballester V, Rashtak S, Boardman L. Clinical and molecular features of young-onset colorectal cancer. World J Gastroenterol 2016;22:1736–44.
21. Gausman V, Dornblaser D, Anand S, et al. Risk factors associated with early-onset colorectal cancer. Clin Gastroenterol Hepatol [Epub ahead of print October 14, 2019.]
22. Triantafillidis JK, Nasioulas G, Kosmidis PA. Colorectal cancer and inflammatory bowel disease: Epidemiology, risk factors, mechanisms of carcinogenesis and prevention strategies. Anticancer Res 2009;29:2727–37.
23. Mauri G, Sartore-Bianchi A, Russo AG, et al. Early-onset colorectal cancer in young individuals. Mol Oncol 2019;13:109–31.
24. Zhou XH, Qiao Q, Zethelius B, et al. Diabetes, prediabetes and cancer mortality. Diabetologia 2010;53:1867–76.
25. Mehraban Far P, Alshahrani A, Yaghoobi M. Quantitative risk of positive family history in developing colorectal cancer: A meta-analysis. World J Gastroenterol 2019;25:4278–91.
26. Brenner H, Zwink N, Ludwig L, et al. Should screening colonoscopy be offered from age 50? Results from a statewide pilot project, and from a randomized intervention study. Dtsch Arztebl Int 2017;114:94–100.
27. Peterse EFP, Meester RGS, Siegel RL, et al. The impact of the rising colorectal cancer incidence in young adults on the optimal age to start screening: Microsimulation analysis I to inform the American Cancer Society colorectal cancer screening guideline. Cancer 2018;124:2964–73.
28. McFerran E, Kee F, Coleman HG. Colorectal cancer screening: Surely FIT for us too. Frontline Gastroenterol 2019;10:445–6.
29. Walter LC, Wender RC, Church TR, et al. Colorectal cancer screening for average-risk adults: 2018 guideline update from the American cancer society. CA Cancer J Clin 2018;68:250–81.
30. Liang PS, Allison J, Ladabaum U, et al. Potential intended and unintended consequences of recommending initiation of colorectal cancer screening at age 45 years. Gastroenterology 2018;155:950–4.
31. Atkin WS, Saunders BP. Surveillance guidelines after removal of colorectal adenomatous polyps. Gut 2002;51:v6–9.
32. Khaw KT, Wareham N, Bingham S, et al. Preliminary communication: Glycated hemoglobin, diabetes, and incident colorectal cancer in men and women: A prospective analysis from the European prospective investigation into cancer-Norfolk study. Cancer Epidemiol Biomarkers Prev 2004;13:915–9.
33. Siddiqui AA, Maddur H, Naik S, et al. The association of elevated HbA1c on the behavior of adenomatous polyps in patients with type-II diabetes mellitus. Dig Dis Sci 2008;53:1042–7.
34. Vu HT, Ufere N, Yan Y, et al. Diabetes mellitus increases risk for colorectal adenomas in younger patients. World J Gastroenterol 2014;20:6946–52.
35. Atkinson MA, Eisenbarth GS, Michels AW. Type 1 diabetes. Lancet 2014;383:69–82.
36. Bullard KM, Cowie CC, Lessem SE, et al. Prevalence of diagnosed diabetes in adults by diabetes type—United States, 2016. MMWR Morb Mortal Wkly Rep 2018;67:359–61.
37. Mendis S, Davis S, Norrving B. Organizational update the world Health organization global status report on noncommunicable diseases 2014; one more landmark step in the combat against stroke and vascular disease. Stroke 2015;46:E121–E122.
38. Goldgar DE, Easton DF, Cannon-Albright LA, et al. Systematic population-based assessment of cancer risk in first-degree relatives of cancer probands. J Natl Cancer Inst 1994;86:1600–8.
39. Schoen RE, Razzak A, Yu KJ, et al. Incidence and mortality of colorectal cancer in individuals with a family history of colorectal cancer. Gastroenterology 2015;149:1438–45 e1.
40. Mollazadegan K, Kugelberg M, Montgomery SM, et al. A population-based study of the risk of diabetic retinopathy in patients with type 1 diabetes and celiac disease. Diabetes Care 2013;36:316–21.