Risk Factors Associated With Young-Onset Colorectal Adenomas and Cancer: A Systematic Review and Meta-Analysis of Observational Research

Abstract

Highlights

Introduction

Colorectal cancer (CRC) is a heterogeneous disease of the colon and rectum that typically develop from adenomatous polyps or adenomas.¹ In 2018 it was reported that of the 18.1 million new cases of cancer, CRC accounted for 10.2% and among 9.6 million cancer-related deaths, CRC accounted for 9.2%.² CRC has negative impacts on physical health,³ life expectancy,⁴ finances,⁵ and psychosocial health.⁶

Generally CRC has been considered a disease of older adults, with the Canadian Task Force on Preventive Health Care recommending screening for adults aged 50 years and older, and the American Cancer Society recommending screening for adults aged 45 years and older.^7,8 However, there is growing evidence that the risk of young-onset CRC and adenoma (yCRAC) in adults under the age of 50 years is increasing. In 2019, using data from Cancer Incidence in Five Continents, Siegel et al. showed increasing risk of yCRAC particularly in high-income countries with cross-sectional age-standardized CRC incidence for adults 20-49 years ranging from 3.5 (95% confidence interval [CI] 3.2 to 3.9) per 100,000 in India to 12.9 (95% CI 12.6-13.3) per 100,000 in Korea, with the average annual percentage change (AAPC) as high as 4.2 (95% CI 3.4-5.0).⁹

Despite these data on increasing risk of yCRAC, reasons for this increase have not been systematically evaluated. With respect to CRC at any age at diagnosis, certain modifiable risk factors have been found to be predictive of CRC onset. In their 2017 review, Kerr et al. reported obesity as a risk factor for CRC¹⁰ which was supported by Avgerinos et al. who in 2019 proposed potential biological mechanisms linking obesity to CRC, including abnormalities in signaling pathways, the role of insulin-like growth factors, sex hormones, and adipocytokines, chronic low-grade inflammation, and alterations in circadian rhythms and nutrient intake.¹¹ More recently, a 2019 review by Murphy et al.¹² reported that obesity, smoking, and alcohol consumption are predictors of CRC onset. However, no review to date has examined which risk factors are related to yCRAC, especially in light of aforementioned increasing risk. Thus, our objective was to conduct a systematic review and meta-analysis of epidemiological studies examining risk factors associated with yCRAC.

Methods

Results

Study Selection and Characteristics

Initial searches identified 989 articles, with an additional article identified through backward reference searching (see Figure 1). Once duplicates were removed, 499 articles underwent title and abstract screening, and 27 articles underwent full-text screening. Nine articles were included in the narrative synthesis, and 6 articles were included in the meta-analysis. Table 2 summarizes study characteristics including quality assessment scores which ranged from 6 to 8, indicating good quality.¹⁷ Most studies (n = 5) examined cases under the age of 50 years,^18-22 one study examined cases under the age of 49 years,²³ one study examined cases under the age of 40 years,²⁴ one study examined cases among those aged 25-42 years²⁵ and one study examined cases in those aged 20-29 years and 30-39 years separately.²⁶ With respect to yCRAC outcomes evaluated, 7 studies defined any diagnosis of yCRAC to be any patient with either an adenoma or cancer.^19,21-26 Lee et al. 2019¹⁸ included any patient with a diagnosis of a sessile serrated adenoma, thus this study is included only in the narrative synthesis and not the meta-analysis, as it is not necessarily equivalent to the criterion for yCRAC used in other studies. Six studies^19-22,24,26 assessed advanced yCRAC as outcome with 5 of these^20-22,24,26 providing a definition of an adenoma more than 10 mm in diameter or with villous or tubulovillous structure, or dysplasia or cancer. Kim et al. (2019)¹⁹ used less a less inclusive criterion for advanced cancer, as they only considered tumors with a diameter greater than 10 mm.

Open in a separate window

Figure 1.

PRISMA diagram for systematic literature search.

Table 2.

Characteristics of Included Studies.

Author, year	Country	Study Design	Age range (year)	Sample size	Sex, N (%) female	yCRAC outcome
						Any yCRAC	Advanced yCRAC	Quality Score
Rosato, 2013	Italy/ Switzerland	Case control	< 45	1,690	787 (46.6%)	Any adenoma or cancer	Not examined	6
Lee, 2016	Korea	Cross-sectional	< 50	1, 776	505 (28.4%)	Any adenoma or cancer	Any one of the following: 10 mm diameter tubulovillous or villous structure high grade dysplasia	6
Kwak, 2016	Korea	Cross-sectional	20-39	4, 286	1,214 (28.3%)	Any adenoma or cancer	Any one of the following: 10 mm diameter tubulovillous or villous structure high grade dysplasia	6
Kim, 2016	Korea	Cross-sectional	<50	59,782	17,138 (28.7%)	Not examined	Any one of the following: 10 mm diameter tubulovillous or villous structure high grade dysplasia	8
Jung, 2017	Korea	Cross-sectional	30-49	57,635	16,866 (29.3%)	Cancer or adenoma	Any one of the following: 10 mm diameter tubulovillous or villous structure high grade dysplasia	7
Kim, 2018	Korea	Cross-sectional	<50	41,702	11,933 (28.6%)	Any component of villous histology, high-grade dysplasia, or carcinoma	Cancer ≥10 mm in diameter	8
Kim, 2019	Korea	Cross-sectional	20-29a 30-39a	72,356	Not reported	Any CRC or adenoma	Any one of the following: 10 mm diameter tubulovillous or villous structure high grade dysplasia	6
Lee, 2019	Korea	Cross-sectional	30-49	13,618	5,981 (43.9%)	Serile sessile polyp	Not examined	N/A
Liu, 2019	United States	Cohort	25-42	85,256	85,256 (100%)	Any diagnosis, self-report confirmed by medical records	Not examined	6

Open in a separate window

^a Age groups assessed separately.

Meta-Analysis

Six studies^{19,21,22,24-26} were included in the meta-analysis and we calculated pooled OR’s to describe the association between the following risk factors and both any diagnosis of yCRAC and advanced yCRAC: 1) cigarette smoking, 2) alcohol consumption; 3) obesity (BMI of 25 or greater); 4) elevated blood glucose; 5) elevated blood pressure; and 6) elevated blood triglycerides . Of note, in one study by Kim et al. (2019),²⁶ ORs were reported separately for their 20-29 year old cohort and the 30-39 year old cohort, and as such, we separately entered corresponding ORs for each age group into relevant meta-analyses.

Lifestyle Factors (Smoking, Alcohol)

Among lifestyle factors, smoking was the most commonly evaluated risk factor in 5 studies, and all studies considered a participant a current smoker if they reported consuming cigarettes on a regular basis.^{19,21,22,24,26} Three studies examined the association between smoking and any yCRAC diagnosis, and resulted in a pooled OR of 1.69 (95% CI 1.44-1.99).^22,24,26 Likewise, 4 studies examined the association between smoking and an advanced yCRAC diagnosis, and the pooled OR was 1.56 (95% CI 1.23-1.98).^19,21,22,26 Alcohol consumption was examined in 4 studies.^19,22,24,26 Researchers used different criteria to define excessive alcohol consumption including more than 20 grams per day,²⁶ more than 40grams per day,²⁴ more than 140 grams per week,²² and consumption of more than 4 times per week.¹⁹ Three studies examined the association between alcohol consumption and yCRAC diagnosis,^22,24,26 and the pooled OR was 1.48 (95% CI 1.40-1.57). Three studies examined the association between alcohol consumption and advanced yCRAC diagnosis,^19,22,26 and resulted in a pooled OR of 1.46 (95% CI 1.27-1.69). These results are presented in Figure 2, displaying Forest Plots for each pooled OR.

Open in a separate window

Figure 2.

Pooled Odds Ratios (ORs) between lifestyle factors and clinical risk factors and any yCRAC diagnosis. 2a: smoking; 2b: alcohol consumption; 2c: obesity; 2d: elevated blood glucose; 2e: elevated blood pressure; 2f: elevated triglycerides.

Clinical Risk Factors (Obesity, Elevated Blood Glucose, Elevated Blood Pressure, Elevated Triglycerides)

Researchers in the 6 studies included in the meta-analysis^{19,21,22,24-26} also examined risk factors associated with yCRAC and advanced yCRAC that were determined by clinical measurements by trained interviewers. With respect to obesity, studies examined the association between elevated BMI and yCRAC (n = 4 studies)^22,24-26 and advanced yCRAC (n = 4 studies).^19,21,22,26 All studies applied having a BMI equal to or greater than 25 kg/m² as the criterion for obesity,^19,21,24,26 except for one study,²⁵ which used a BMI equal to or greater than 23 kg/m² as the obesity criterion. When we pooled the ORs, the pooled OR for obesity and any yCRAC diagnosis was 1.45 (95% CI 1.37-1.52) and the pooled OR for obesity and advanced yCRAC diagnosis was 1.26 (95% CI 1.04-1.54) (see Figure 3). When examining the association between elevated blood glucose and any yCRAC diagnosis or an advanced yCRAC diagnosis, the pooled OR for elevated blood glucose and any yCRAC diagnosis was 1.69 (95% CI 1.27-2.25) and the pooled OR for elevated blood glucose and advanced yCRAC diagnosis was 1.60 (95% CI 1.30-1.96).All studies^19,24,26 defined an elevated blood glucose level as a fasting blood glucose level at or above 100 mg/dL. Three studies evaluated elevated blood pressure and its association with any yCRAC diagnosis^22,24,26 and 3 studies evaluated the association between elevated blood pressure and an advanced yCRAC diagnosis.^21,22,26 Authors also used variable criteria for determining elevated blood pressure: (1) a systolic blood pressure of 130 mmHg or greater and/or a diastolic blood pressure of 85 mmHg or greater, or using antihypertensive drugs^24,26; (2) a systolic blood pressure of 140 mmHg or greater and/or a diastolic blood pressure of 90 or greater²²; or (3) the authors did not specify the criterion for elevated blood pressure.²¹ When we pooled the ORs, the pooled OR for any yCRAC diagnosis was 1.56 (95% CI 1.31-1.86) and the pooled OR for an advanced yCRAC diagnosis was 1.51 (95% CI 1.14-2.01).Finally, 3 studies examined the association between elevated triglycerides and any yCRAC diagnosis,^22,24,26 and 3 studies examined the association with an advanced yCRAC diagnosis.^21,22,26 The criterion for elevated triglycerides was a measured value of 150 mg/dL,^22,24,26 or the authors did not specify a minimum value.²¹ The pooled OR for elevated triglycerides and any yCRAC diagnosis was 1.51 (95% CI 1.41-1.62), and the pooled OR for elevated triglycerides and an advanced yCRAC was 1.60 (95% CI 1.07-2.38).

Open in a separate window

Figure 3.

Pooled Odds Ratios (ORs) between lifestyle factors and clinical factors and advanced yCRAC diagnosis. 3a: smoking; 3b: alcohol consumption; 3c: obesity; 3d: elevated blood glucose level; 3e: elevated blood pressure; 3f: elevated triglycerides.

Discussion

Our systematic review and meta-analysis has demonstrated that several modifiable lifestyle indicators, (smoking, alcohol consumption), clinical factors (obesity/elevated BMI, elevated blood glucose, elevated blood pressure, elevated triglycerides), and chronic conditions (type 2 diabetes, metabolic syndrome) are all significant predictors of both yCRAC diagnoses and advanced yCRAC diagnoses. Specifically, pooled ORs for the association between lifestyle factors of smoking (OR = 1.69, 95% CI 1.44-1.99) and alcohol consumption (OR = 1.48, 95% CI 1.40-1.57) and any yCRAC, with similar associations between these factors and an advanced yCRAC diagnosis, provide confirmatory evidence that these established risk factors for CRC¹² also play a role in the risk of yCRAC. We additionally showed associations between clinical factors and any diagnosis of yCRAC, including: obesity (OR = 1.45, 95% CI 1.37-1.52); elevated blood glucose (OR = 1.69, 95% CI 1.27-2.25); elevated blood pressure (OR = 1.56, 95% CI 1.31-1.86); and elevated triglycerides (OR = 1.51,95% CI 1.41-1.62), with similar results for the associations between clinical factors and an advanced yCRAC diagnosis. Finally, we found significant associations between yCRAC and advanced yCRAC for chronic conditions: type 2 diabetes and advanced yCRAC diagnosis (OR = 1.60, 95% CI 1.32-1.95), and metabolic syndrome and any yCRAC diagnosis (OR = 1.56, 95% CI 1.44-1.68). Thus, this systematic review provides evidence that certain lifestyle and clinical risk factors for CRC, and chronic conditions, are associated with yCRAC. Key in our review was the consideration of chronic conditions and its association with yCRAC, indicating that health professionals should be aware of conditions increasing patients’ risk of yCRAC. It is important to note that some clinical factors, for example hypertension, elevated blood glucose levels, and obesity, may not automatically indicate presence of a clinically significant health condition requiring treatment. However, these factors, when present in a patient presenting with other risk factors for yCRAC, may necessitate further clinical investigation or observation.

Other researchers have also found that these risk factors are associated with CRC incidence in the general population, not specific to those under the age of 50 years. Kerr et al.¹⁰ and Avreginos et al.¹¹ both found that obesity is related to CRC diagnosis in older populations, and in their review, Murphy et al.¹² report that obesity, smoking, and alcohol consumption are associated with CRC diagnoses in all age groups. One biological mechanism proposed by Avreginos et al. is that obesity is associated with a higher fat diet, microbiome alteration, and subclinical inflammation, which contributes to greater risk of developing polyps and eventually colorectal cancer. However, even with CRC in the general population, there has been less focus on the contribution of clinical risk factors (elevated blood glucose, blood pressure, and triglycerides) and chronic disease (type 2 diabetes, metabolic syndrome) to the risk and burden of CRC.

Given that the incidence and prevalence of yCRAC has been shown to be increasing in multiple countries,²⁷ it is important for health professionals, especially primary care providers, to be aware of the consequences of the presence of these factors in a patient, and ultimately CRC screening program recommendations⁷ may need to consider revising their screening recommendations for those under the age of 50 years whose clinical risk factors and/or chronic conditions place them at increased risk of yCRAC. For example, the American Cancer Society recently lowered their recommended screening age from 50 years to 45 years, to assist in reducing the number of yCRAC cases that are not identified through routine screening.²⁸ Our systematic review raises the possibility that certain groups of individuals under the age at 50 years may be at increased risk of yCRAC, and thus screening programs may need to be modified to address these higher-risk groups. Additionally, there is a need for health professionals to be better aware of the possibility of yCRAC in their patients when patients present with yCRAC symptoms and investigate these cases accordingly. For example, our group recently published a patient-oriented, multi-method study indicating that yCRAC is frequently misdiagnosed in symptomatic patients under the age of 50, leading to delays in diagnosis and treatment.²⁹ Thus, through both modifying population-based screening programs, and educating health professionals as to the risk of yCRAC in patients under 50, especially for those in higher-risk groups identified in this systematic review, the burden of this disease in those under the age of 50 years can hopefully be reduced.

This review and meta-analysis is, to our knowledge, the first to examine risk factors for CRC specifically in those under the age of 50. While there were a limited number of published observational studies meeting study inclusion criteria, and while authors used differing criteria for both yCRAC and the risk factors for yCRAC that we examined, this study still makes a potentially important contribution given the paucity of research specific to yCRAC and the emerging nature of the field. We also collaborated with an information specialist in order to develop a comprehensive literature search across multiple databases and a detailed search strategy. We also used MOOSE reporting guidelines for our review and meta-analysis, which gives further credibility to our results.³⁰ There were some limitations associated with this study. First, we only searched for peer-reviewed journal articles, and may have overlooked unpublished data, such as government reports. Second, we excluded studies that examined yCRAC risk factors in those at increased risk of yCRAC, for example those individuals with genetic disorders that place them at high risk of developing CRC at an early age. This exclusion criterion was to help elucidate the associations of lifestyle and clinical factors and chronic diseases with yCRAC diagnosis. Finally, few published observational studies were available, and there is a large heterogeneity, both of the study samples and the criteria used to define both yCRAC and the risk factors in the study. However, given the emerging nature of the yCRAC research field, and the growing clinical need to address the increasing incidence of yCRAC worldwide,²⁷ this study still provides a potentially important contribution to the field, and can assist in developing future research in this emerging field.

Conclusions

We have demonstrated that lifestyle factors, clinical factors, and chronic conditions are all associated with yCRAC diagnosis, and warrant further attention. While some factors (i.e. smoking, alcohol consumption, obesity) may be modified through behavior change interventions, other risk factors (i.e. elevated blood glucose, blood pressure, and triglycerides) and chronic conditions (i.e. type 2 diabetes, metabolic syndrome) may be more difficult to address through these interventions. These findings suggest that there may need to be more rigorous screening by primary care providers for these risk factors, given their association with yCRAC. Eventually, populations with these elevated risk factors may ultimately require more stringent screening for colorectal cancer in individuals under age 50, in order to reduce the burden of this disease and allow for earlier treatment and better prognosis.

Supplemental Material

Acknowledgments

Abbreviations

Footnotes

References