Increased risk of colorectal cancer in young males with higher cardiovascular risk: A nationwide population-based cohort study

Abstract

BACKGROUND

Although the link between cardiovascular disease (CVD) and various cancers is well-established, the relationship between CVD risk and colorectal cancer (CRC) remains underexplored.

AIM

To elucidate the relationship between CVD risk scores and CRC incidence.

METHODS

In this population-based cohort study, participants from the 2009 National Health Checkup were followed-up until 2020. The cardiovascular (CV) risk score was calculated as the sum of risk factors (age, family history of coronary artery disease, hypertension, smoking status, and high-density lipoprotein levels) with high-density lipoprotein (≥ 60 mg/dL) reducing the risk score by one. The primary outcome was incidence of newly diagnosed CRC.

RESULTS

Among 2526628 individuals, 30329 developed CRC during a mean follow-up of 10.1 years. Categorized by CV risk scores (0, 1, 2, and ≥ 3). CRC risk increased with higher CV risk scores after adjusting for covariates [(hazard ratio = 1.155, 95% confidence interval: 1.107-1.205) in risk score ≥ 3, P < 0.001]. This association was exclusively observed in males, most notably in the younger cohort (< 50 years) and was more pronounced in individuals not using statins. Moreover, even in participants without diabetes, a higher CV risk was associated with an increased CRC risk.

CONCLUSION

Increased CV risk scores were significantly associated with higher CRC risk, especially among males, younger populations, and non-statin users. Thus, males with a higher CV risk score, even at a younger age, are recommended to control their risk factors and undergo individualized CRC screening.

Key Words: Colorectal cancer; Cardiovascular disease; Cohort study; Age; Gender

Core Tip: While studies have reported an association between cardiovascular disease (CVD) and colorectal cancer (CRC), large-scale investigations on the incidence of CRC based on the risk of CVD are lacking. This study benefits from being a nationwide population-based cohort study with long-term follow-up data spanning 10 years, which is a significant strength. This study highlights that increased cardiovascular risk scores are significantly associated with higher CRC risk, especially among males and younger populations. Thus, males with a higher cardiovascular risk score, even at a younger age, are recommended to control their risk factors and undergo individualized CRC screening.

INTRODUCTION

Cardiovascular diseases (CVD) and cancer are the two leading causes of death worldwide[1]. A recent cohort study reported an increased risk of cancer in patients with CVD[2]. Other cohort studies comparing atherosclerotic and non-atherosclerotic CVD highlighted a higher risk of cancer in individuals with atherosclerotic CVD[3]. A recent prospective cohort study demonstrated an association between traditional CVD risk factors, such as age, male sex, smoking, and an increased risk of future cancer[4]. Moreover, another prospective cohort study revealed that meeting the criteria for ideal cardiovascular (CV) health metrics [including no smoking, body mass index (BMI) < 25 kg/m², a healthy diet, regular physical activity, total cholesterol < 200 mg/dL, blood pressure < 120/80 mmHg, and fasting glucose < 100 mg/dL] was associated with a reduced incidence of cancer[5].

CVD and cancer share similar risk factors, such as obesity and diabetes mellitus (DM). They are believed to involve common biological mechanisms, such as inflammation, oxidative stress, and reactive oxygen species[6]. However, the precise mechanisms underlying the relationship between CVD and cancer have not been fully elucidated.

Colorectal cancer (CRC) is the third most commonly diagnosed cancer and the second leading cause of cancer-related deaths worldwide[7]. Several studies have evaluated the association between CVD and CRC. A previous meta-analysis reported an association between ischemic heart disease and colorectal neoplasms (odds ratio = 1.87)[8]. Another cohort study revealed a 1.34-fold higher risk of colon cancer in patients with atherosclerotic CVD[2]. However, large-scale studies investigating the incidence of CRC based on CVD risk factors are relatively scarce. This study aimed to investigate the association between risk scores for quantifying CVD risk and CRC incidence.

MATERIALS AND METHODS

National Health Insurance Service database

South Korea provides universal medical coverage through the National Health Insurance Service (NHIS). In addition, the NHIS offers regular health checkups and cancer screenings. Adults aged 40 and above can receive free health checkups every two years. Individuals bear only 10% of the cost of screening for gastric, liver, colorectal, breast, cervical, and lung cancers[9]. These health checkups comprised anthropometric data measured by licensed medical staff, self-administered lifestyle questionnaires (e.g., smoking, alcohol consumption, and regular exercise), previous medical history, blood tests, and urine tests. All data within the NHIS are managed by the National Health Insurance Corporation, encompassing the demographic information of patients, records of health checkups, hospital visits, diagnostic details, prescribed medications, treatments administered, and dates of death, constituting a comprehensive repository of various types of medical information[10].

Study population

A flowchart of the enrollment process for this study is shown in Figure 1. Among those who underwent the NHIS health checkup in 2009, 40% (n = 4234415) were randomly sampled, resulting in 4233321 individuals aged 20-90 years. Individuals with incomplete medical records (n = 1646622) or a history of cancer (n = 38874) were excluded from the study. Individuals diagnosed with CRC or who died for any reason (n = 21197) within one year of follow-up were also excluded because of the potential presence of preexisting CRC. Finally, 2526628 individuals were included in the study cohort.

Figure 1 Flow chart of enrollment process in this study.

The study protocol was approved by the Ethics Committee of Soongsil University (IRB No. SSU-202007-HR-236-01) and was conducted per the Declaration of Helsinki. The requirement for patient consent was waived, because of the retrospective nature of the study.

CV risk evaluation and related variables

Risk assessment for CVD was conducted by scoring various factors based on the “2018 Guidelines for the Management of Dyslipidemia in Korea”, published by the Korean Society of Lipid and Atherosclerosis[11]. The major risk factors other than low-density lipoprotein cholesterol (LDL-C) are as follows: (1) Age: Male ≥ 45 years and female ≥ 55 years; (2) Family history of premature coronary artery disease (CAD): CAD < 55 years for males and < 65 years for females among parents and siblings; (3) Hypertension: Systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg, or taking anti-hypertensive medications; (4) Smoking; and (5) Low high-density lipoprotein cholesterol (HDL-C) (< 40 mg/dL).

The risk score was calculated as the sum of individual risk factors, with high HDL-C (≥ 60 mg/dL) considered a protective factor, resulting in a deduction of 1 point from the total sum of the risk factors.

LDL-C level is an important risk factor for CVD. However, when individuals take lipid-lowering agents, their LDL-C levels may decrease, potentially leading to an inaccurate reflection of risk. Therefore, LDL-C levels were not included in the risk score but were analyzed separately.

Smoking status is surveyed based on participant smoking status at the time of the study (i.e., nonsmoker, past smoker, and current smoker). Alcohol intake was classified into non-drinkers, moderate drinkers (averaging < 30 g of alcohol per day), and heavy drinkers (≥ 30 g per day). Regular exercise was defined as moderate exercise for at least five days per week or vigorous exercise for at least three days per week. Hypertension was defined as having a blood pressure ≥ 140/90 mmHg or taking anti-hypertensive medications. Dyslipidemia was defined as a total cholesterol level ≥ 240 mg/dL after a 12-hour fast or the use of lipid-lowering agents. Statin use was defined as statin use at the time of data collection. DM was characterized by fasting glucose level ≥ 126 mg/dL or using glucose-lowering agents.

Study outcome

The cohort was followed up until the date of outcome, date of death, or December 31, 2020. The study outcome was newly diagnosed CRC. All clinical events were identified using the diagnostic codes and claims from the NHIS database. CRC was assessed using the International Classification of Diseases 10^th revision codes C18-C20 and the registration code for cancer V193.

Statistical analysis

Continuous variables following a normal distribution were presented as mean ± SD, whereas those not following a normal distribution were presented as median (interquartile range). Categorical variables are presented as frequencies (percentages). For comparisons between groups, analysis of variance was used for continuous variables, and the χ² test was used for categorical variables. The incidence rate of CRC was expressed as the number of events divided by 1000 person-years. A multivariate Cox proportional hazards model was used to assess the risk of CRC based on the CV risk score. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated without adjustment (Model 1); after adjustment for sex and age (Model 2); after adjustment for sex, age, alcohol consumption, regular exercise, BMI, DM, previous history of coronary artery disease or stroke, and aneurysm of the abdominal aorta (Model 3); and after adjustment for Model 3 and statin use (Model 4). Statistical significance was set at P < 0.05 significant. All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, United States).

RESULTS

Patient characteristics

We divided the total study population of 2526628 individuals into groups based on CV risk scores. The CV risk score was evaluated on a scale of 0-5 based on the number of corresponding risk factors. However, because of the relatively small number of individuals with risk scores of 3, 4, and 5, they were combined into a single group for analysis. The demographic and clinical characteristics according to the CV risk score are presented in Supplementary Table 1, and a comparison of the clinical characteristics and CV risk factors by sex is presented in Supplementary Table 2.

Association between CRC risk and CV risk score

The average follow-up duration was 10.13 ± 1.23 years (median 10.31 years, lower quartile: 10.11 years, upper quartile: 10.56 years), during which 30329 individuals were newly diagnosed with CRC. Figure 2 presents the Kaplan-Meier survival curve for CRC incidence, showing that as the CV risk score increased, the CRC incidence also increased. This pattern was consistent for both males and females.

Figure 2 Kaplan-Meier survival curve for colorectal cancer incidence. As the cardiovascular risk score increased, the incidence of colorectal cancer also increased. This pattern was consistent for both males and females.

The incidence rate of CRC increased significantly with higher CV risk scores (P for trend < 0.001), as shown in Table 1. After adjusting for various risk factors (including sex, age, alcohol consumption, regular exercise, BMI, DM, history of percutaneous coronary intervention or coronary artery bypass graft, myocardial infarction, stroke, abdominal aortic aneurysm, and statin use), the analysis revealed that individuals with a risk score of ≥ 3 had an HR of 1.155 (95%CI: 1.107-1.205) for CRC, signifying an elevated risk.

Table 1 Risk of colorectal cancer according to cardiovascular risk score.

	Risk score	n	Events	Person years	IRa	Model 1 HR (95%CI)	Model 2 HR (95%CI)	Model 3 HR (95%CI)	Model 4 HR (95%CI)
Total	0	1125969	7898	11544364	0.68	1 (Ref.)	1 (Ref.)	1 (Ref.)	1 (Ref.)
	1	775642	9612	7856740	1.22	1.788 (1.736-1.842)	0.949 (0.919-0.980)	0.936 (0.906-0.966)	0.938 (0.908-0.968)
	2	451302	8563	4492913	1.91	2.788 (2.704-2.874)	1.049 (1.014-1.086)	1.022 (0.987-1.058)	1.027 (0.992-1.064)
	≥ 3	173715	4256	1697936	2.51	3.671 (3.536-3.810)	1.188 (1.139-1.238)	1.147 (1.100-1.197)	1.155 (1.107-1.205)
	P for trend					< 0.001	< 0.001	< 0.001	< 0.001
Male	0	426624	2512	4363981	0.58	1 (Ref.)	1 (Ref.)	1 (Ref.)	1 (Ref.)
	1	522624	5920	5289246	1.12	1.944 (1.856-2.037)	1.135 (1.082-1.191)	1.116 (1.064-1.170)	1.117 (1.065-1.171)
	2	325451	6093	3238870	1.88	3.269 (3.120-3.424)	1.283 (1.222-1.346)	1.241 (1.182-1.304)	1.247 (1.187-1.309)
	≥ 3	138611	3462	1353225	2.56	4.449 (4.226-4.684)	1.421 (1.346-1.499)	1.356 (1.284-1.433)	1.367 (1.294-1.444)
	P for trend					< 0.001	< 0.001	< 0.001	< 0.001
Female	0	699345	5386	7180383	0.75	1 (Ref.)	1 (Ref.)	1 (Ref.)	1 (Ref.)
	1	253018	3692	2567494	1.44	1.917 (1.839-1.999)	0.987 (0.941-1.035)	0.968 (0.923-1.015)	0.967 (0.922-1.015)
	2	125851	2470	1254042	1.97	2.629 (2.506-2.757)	1.019 (0.963-1.077)	0.984 (0.929-1.042)	0.984 (0.929-1.042)
	≥ 3	35104	794	344712	2.30	3.078 (2.857-3.316)	1.093 (1.008-1.185)	1.049 (0.966-1.139)	1.049 (0.966-1.139)
	P for trend					< 0.001	0.074	0.640	0.650

^aColorectal cancer incidence rate per 1000 person-years.

Model 1: Not adjusted; Model 2: Adjusted for sex and age; Model 3: Adjusted for sex, age, alcohol consumption, regular exercise, body mass index, diabetes mellitus, and previous history of percutaneous coronary intervention or coronary artery bypass graft, myocardial infarction, stroke, and aneurysm of abdominal aorta; Model 4: Adjusted for sex, age, alcohol consumption, regular exercise, body mass index, diabetes mellitus, previous history of percutaneous coronary intervention or coronary artery bypass graft, myocardial infarction, stroke, aneurysm of abdominal aorta, and statin use; IR: Incidence rate; HR: Hazard ratio; CI: Confidence interval.

Upon stratification by sex, the pattern of increased CRC risk with rising CV risk scores was more pronounced among males (P for trend < 0.001). However, no significant association was observed in females (P for trend 0.650).

Risk of CRC according to CV risk score subdivided by age

We divided the participants into age groups: Those < 50 years and those ≥ 50 years, to analyze CRC risk according to CV risk score (Figure 3). In both age groups, there was a tendency for CRC risk to increase with higher CV risk scores, with the younger age group showing a higher risk of CRC compared to the older age group (HR = 1.460, 95%CI: 1.312-1.625) in risk score ≥ 3 of < 50 years; (HR = 1.086, 95%CI: 1.034-1.140) in risk score ≥ 3 of ≥ 50 years (P for interaction < 0.001) (Table 2).

Figure 3 Colorectal cancer incidence rate by cardiovascular risk score subdivided by age. In both age groups, colorectal cancer (CRC) risk tended to increase with higher cardiovascular risk scores, with the younger age group showing a higher risk of CRC than the older age group (P for interaction < 0.001). In males, there was a consistent trend of increasing CRC risk, with higher cardiovascular risk scores observed in both younger and older age groups.

Table 2 Risk of colorectal cancer according to cardiovascular risk score when subdivided by age.

Gender	Age	Risk score	n	Events	Person years	IRa	Model 1 HR (95%CI)	Model 2 HR (95%CI)	Model 3 HR (95%CI)	Model 4 HR (95%CI)
Total	< 50 years	0	893565	4125	9184868	0.45	1 (Ref.)	1 (Ref.)	1 (Ref.)	1 (Ref.)
		1	433525	2079	4447577	0.47	1.041 (0.987-1.097)	1.043 (0.984-1.105)	1.027 (0.969-1.089)	1.028 (0.970-1.090)
		2	161618	1090	1653286	0.66	1.468 (1.374-1.570)	1.177 (1.090-1.270)	1.147 (1.062-1.238)	1.151 (1.066-1.243)
		≥ 3	42156	469	428824	1.09	2.437 (2.215-2.682)	1.506 (1.355-1.673)	1.448 (1.301-1.610)	1.460 (1.312-1.625)
		P for trend					< 0.001	< 0.001	< 0.001	< 0.001
	≥ 50 years	0	232404	3773	2359496	1.60	1 (Ref.)	1 (Ref.)	1 (Ref.)	1 (Ref.)
		1	342117	7533	3409163	2.21	1.382 (1.329-1.437)	0.976 (0.937-1.016)	0.963 (0.925-1.003)	0.965 (0.927-1.005)
		2	289684	7473	2839626	2.63	1.647 (1.583-1.712)	1.045 (1.003-1.089)	1.017 (0.976-1.061)	1.022 (0.980-1.065)
		≥ 3	131559	3787	1269112	2.98	1.868 (1.786-1.954)	1.122 (1.070-1.176)	1.080 (1.029-1.134)	1.086 (1.034-1.140)
		P for trend					< 0.001	< 0.001	< 0.001	< 0.001
	P for interaction						< 0.001	< 0.001	< 0.001	< 0.001
Male	< 50 years	0	380916	1284	3914571	0.33	1 (Ref.)	1 (Ref.)	1 (Ref.)	1 (Ref.)
		1	361592	1587	3708820	0.43	1.305 (1.212-1.404)	1.044 (0.969-1.125)	1.023 (0.949-1.103)	1.024 (0.950-1.103)
		2	152047	1024	1554993	0.66	2.008 (1.850-2.180)	1.142 (1.047-1.246)	1.101 (1.008-1.202)	1.105 (1.012-1.206)
		≥ 3	41490	463	422036	1.10	3.348 (3.010-3.723)	1.398 (1.248-1.566)	1.319 (1.176-1.481)	1.333 (1.188-1.497)
		P for trend					< 0.001	< 0.001	< 0.001	< 0.001
	≥ 50 years	0	45708	1228	449410	2.73	1 (Ref.)	1 (Ref.)	1 (Ref.)	1 (Ref.)
		1	161032	4333	1580426	2.74	1.003 (0.942-1.069)	1.004 (0.942-1.070)	1.004 (0.942-1.070)	1.005 (0.944-1.071)
		2	173404	5069	1683877	3.01	1.102 (1.035-1.173)	1.089 (1.023-1.159)	1.075 (1.010-1.145)	1.080 (1.015-1.151)
		≥ 3	97121	2999	931189	3.22	1.180 (1.104-1.261)	1.165 (1.090-1.245)	1.137 (1.062-1.216)	1.146 (1.071-1.226)
		P for trend					< 0.001	< 0.001	< 0.001	< 0.001
	P for interaction						< 0.001	< 0.001	< 0.001	< 0.001
Female	< 50 years	0	512649	2841	5270298	0.54	1 (Ref.)	1 (Ref.)	1 (Ref.)	1 (Ref.)
		1	71933	492	738757	0.67	1.235 (1.122-1.359)	1.027 (0.933-1.131)	1.015 (0.921-1.118)	1.015 (0.921-1.118)
		2	9571	66	98293	0.67	1.244 (0.975-1.588)	0.932 (0.730-1.190)	0.907 (0.710-1.161)	0.907 (0.709-1.161)
		≥ 3	666	6	6788	0.88	1.650 (0.743-3.665)	1.123 (0.504-2.502)	1.086 (0.487-2.422)	1.085 (0.486-2.422)
		P for trend					< 0.001	0.891	0.851	0.850
	≥ 50 years	0	186696	2545	1910086	1.33	1 (Ref.)	1 (Ref.)	1 (Ref.)	1 (Ref.)
		1	181085	3200	1828737	1.75	1.314 (1.248-1.385)	1.007 (0.953-1.064)	0.990 (0.937-1.046)	0.990 (0.936-1.046)
		2	116280	2404	1155749	2.08	1.563 (1.479-1.653)	1.070 (1.007-1.137)	1.034 (0.972-1.100)	1.034 (0.971-1.100)
		≥ 3	34438	788	337923	2.33	1.754 (1.620-1.900)	1.153 (1.060-1.254)	1.106 (1.015-1.205)	1.105 (1.014-1.204)
		P for trend					< 0.001	< 0.001	0.022	0.024
	P for interaction						0.256	0.218	0.339	0.338

^aColorectal cancer incidence rate per 1000 person-years.

Model 1: Not adjusted; Model 2: Adjusted for sex and age; Model 3: Adjusted for sex, age, alcohol consumption, regular exercise, body mass index, diabetes mellitus, and previous history of percutaneous coronary intervention or coronary artery bypass graft, myocardial infarction, stroke, and aneurysm of abdominal aorta; Model 4: Adjusted for sex, age, alcohol consumption, regular exercise, body mass index, diabetes mellitus, previous history of percutaneous coronary intervention or coronary artery bypass graft, myocardial infarction, stroke, aneurysm of abdominal aorta, and statin use; IR: Incidence rate; HR: Hazard ratio; CI: Confidence interval.

We analyzed according to age group stratified by sex. In males, there was a consistent trend of increasing CRC risk, with higher CV risk scores observed in both younger and older age groups. In females, however, an increased risk of CRC was observed only in those ≥ 50 years with a risk score ≥ 3.

Risk of CRC according to CV risk score subdivided by statin use

We analyzed the CRC risk according to the CV risk score stratified by statin use. In the statin non-user group, there was a significant increase in CRC risk with higher CV risk scores (P for trend < 0.001). However, no clear associations were observed in the statin user group. This pattern was more pronounced in males, whereas in the statin non-user group, the CRC risk increased by 1.37 times (P for trend < 0.001). In females, there was no association between the CV risk score and CRC risk, regardless of statin use (Table 3).

Table 3 Risk of colorectal cancer according to cardiovascular risk score when subdivided by statin use.

		Risk score	n	Events	Person years	IRa	Model 1 HR (95%CI)	Model 2 HR (95%CI)	Model 3 HR (95%CI)
Total	Statin (-)	0	1095789	7504	11237006	0.67	1 (Ref.)	1 (Ref.)	1 (Ref.)
		1	721199	8615	7310810	1.18	1.765 (1.711-1.820)	0.944 (0.913-0.976)	0.932 (0.902-0.964)
		2	388601	7217	3874060	1.86	2.791 (2.702-2.883)	1.061 (1.023-1.100)	1.035 (0.998-1.074)
		≥ 3	138158	3366	1352131	2.49	3.735 (3.587-3.890)	1.200 (1.147-1.255)	1.160 (1.108-1.214)
		P for trend					< 0.001	< 0.001	< 0.001
	Statin (+)	0	30180	394	307358	1.28	1 (Ref.)	1 (Ref.)	1 (Ref.)
		1	54443	997	545930	1.83	1.425 (1.268-1.602)	0.991 (0.880-1.117)	0.972 (0.862-1.095)
		2	62701	1346	618852	2.17	1.698 (1.518-1.900)	1.003 (0.892-1.127)	0.974 (0.865-1.096)
		≥ 3	35557	890	345805	2.57	2.010 (1.785-2.263)	1.145 (1.010-1.297)	1.103 (0.972-1.253)
		P for trend					< 0.001	0.010	0.046
	P for interaction						< 0.001	0.036	0.071
Male	Statin (-)	0	418317	2395	4280007	0.56	1 (Ref.)	1 (Ref.)	1 (Ref.)
		1	500104	5515	5065266	1.09	1.946 (1.854-2.041)	1.144 (1.090-1.202)	1.126 (1.072-1.182)
		2	293142	5368	2922285	1.84	3.283 (3.129-3.445)	1.306 (1.242-1.373)	1.264 (1.201-1.329)
		≥ 3	113465	2807	1109737	2.53	4.525 (4.285-4.779)	1.443 (1.363-1.528)	1.374 (1.297-1.456)
		P for trend					< 0.001	< 0.001	< 0.001
	Statin (+)	0	8307	117	83974	1.39	1 (Ref.)	1 (Ref.)	1 (Ref.)
		1	22520	405	223980	1.81	1.298 (1.057-1.594)	0.889 (0.723-1.093)	0.870 (0.707-1.070)
		2	32309	725	316586	2.29	1.644 (1.352-1.998)	0.954 (0.783-1.163)	0.928 (0.760-1.132)
		≥ 3	25146	655	243488	2.69	1.931 (1.586-2.351)	1.107 (0.907-1.350)	1.067 (0.873-1.305)
		P for trend					< 0.001	0.003	0.011
	P for interaction						< 0.001	0.004	0.006
Female	Statin (-)	0	677472	5109	6956999	0.73	1 (Ref.)	1 (Ref.)	1 (Ref.)
		1	221095	3100	2245544	1.38	1.880 (1.798-1.965)	0.962 (0.914-1.012)	0.946 (0.899-0.995)
		2	95459	1849	951776	1.94	2.647 (2.510-2.792)	1.001 (0.941-1.065)	0.973 (0.913-1.036)
		≥ 3	24693	559	242394	2.31	3.148 (2.885-3.435)	1.069 (0.973-1.174)	1.033 (0.939-1.136)
		P for trend					< 0.001	0.408	0.868
	Statin (+)	0	21873	277	223384	1.24	1 (Ref.)	1 (Ref.)	1 (Ref.)
		1	31923	592	321950	1.84	1.484 (1.287-1.712)	1.150 (0.992-1.333)	1.130 (0.975-1.311)
		2	30392	621	302266	2.05	1.659 (1.440-1.912)	1.140 (0.980-1.327)	1.109 (0.951-1.293)
		≥ 3	10411	235	102317	2.30	1.856 (1.559-2.208)	1.241 (1.033-1.491)	1.199 (0.996-1.443)
		P for trend					< 0.001	0.047	0.126
	P for interaction						< 0.001	0.521	0.534

^aColorectal cancer incidence rate per 1000 person-years.

Model 1: Not adjusted; Model 2: Adjusted for sex and age; Model 3: Adjusted for sex, age, alcohol consumption, regular exercise, body mass index, diabetes mellitus, and previous history of percutaneous coronary intervention or coronary artery bypass graft, myocardial infarction, stroke, and aneurysm of abdominal aorta.

Risk of CRC according to CV risk score, stratified by DM status

The relationship between the CV risk score and CRC risk was analyzed by dividing the participants based on the presence of DM. In the group without DM, the CRC risk was 1.16 times higher when the CV risk score was ≥ 3. In males without DM, the risk increased to 1.37 times. However, there was no increase in CRC risk in females, regardless of DM status (Supplementary Table 3).

DISCUSSION

This nationwide cohort study utilized health check-up data to analyze the risk of CRC based on CV risk scores over a more than 10-year follow-up period. As the CV risk score increased, there was a corresponding increase in the risk of CRC by up to 1.16 times. This association was exclusively observed in males, with the CRC risk being 1.37 times higher in individuals with a risk score of 3 or higher. A recent meta-analysis reported that unfavorable CV risk factors, including obesity [relative risk (RR) = 1.31], current smoking (RR = 1.20), diabetes (RR = 1.25), and hypertension (RR = 1.07), were associated with an increased risk of CRC[12]. In another cohort study analyzing the relationship between a healthy heart score[13] comprising diet, physical activity, alcohol intake, smoking, body weight, and total and cause-specific mortality, it was reported that in the highest CVD risk group, colon cancer mortality was 1.5 times higher compared to the lowest CVD risk group[14]. In another meta-analysis investigating the association between healthy lifestyle factors and CRC risk, an inverse relationship was observed, indicating a lower risk of CRC with a healthier lifestyle[15]. However, there was no significant association between the risk of CRC and CV risk score in women. This might be due to the higher proportion of females in the lower CV risk score category and a higher proportion of males as the risk score increased, suggesting a potential impact of sex disparity among the CV risk score groups included in this study. Nonetheless, there was a sex-related difference in the relationship between the CV risk score and CRC risk.

Sex differences are believed to play a role in colorectal carcinogenesis[16]. These differences encompass two aspects: Sexual dimorphism, which involves biological variances in hormones and genes, and gender disparities, which reflect non-biological distinctions in societal attitudes and behaviors[16]. According to the Global Cancer Observatory, the global incidence of CRC is higher in males than in females worldwide[17]. A recent large population-based study on CRC reported sex differences in incidence, patient characteristics, and tumor characteristics[18]. Combining previous reports with our findings, it is reasonable to consider that there may be sex-based differences in the risk factors for CRC.

The analysis, stratified by age group with 50 years as the threshold, revealed a more pronounced increase in CRC risk with escalation of the CV risk score among individuals under 50 years of age. Specifically, in those under 50, when the risk score was ≥ 3, the CRC risk increased by over 1.5 times. Upon stratification by sex, this trend was more significant in males than females. According to previous reports, the overall incidence of CRC has declined since implementation of CRC screening programs in the 1990s[19]. However, there has been a rapid increase in the incidence of early-onset CRC (EO-CRC) in individuals under the age of 50[20,21]. Recent studies have highlighted various aspects of EO-CRC, including its epidemiology, risk factors, pathophysiology, and clinical characteristics[20,22]. The age group analysis in our study was not based on the onset age of CRC; therefore, the results do not specifically address EO-CRC. However, it is conceivable that a considerable proportion of CRC cases in younger age groups included EO-CRC. Given the elevated risk of CRC associated with high CV risk scores in younger individuals, proactive colon cancer screening is advisable, especially among those with elevated CV risk scores.

Dyslipidemia is a major risk factor for CVD. Moreover, recent meta-analyses have reported that dyslipidemia increases the risk of CRC[23]. We analyzed the relationship between CV risk score and LDL-C levels, which showed no consistent trend. However, as the CV risk score increased, the prevalence of dyslipidemia also increased at rates of 12.3%, 17.6%, 23.5%, and 28.3%. According to the guidelines for dyslipidemia management in Korea[11], statin therapy is determined based on the individual’s CV risk. Target LDL-C concentrations varied by risk group, with very high-risk groups aiming for < 70 mg/dL, high-risk groups for < 100 mg/dL, moderate-risk groups for < 130 mg/dL, and low-risk groups for < 160 mg/dL. Statin therapy was initiated if LDL-C levels exceeded these thresholds. Therefore, LDL-C levels are significantly influenced by statin use, making it challenging to assess LDL-C levels as CV or CRC risk factors. Therefore, we analyzed the relationship between CV risk scores and CRC risk based on statin use. Our analysis revealed a significant increase in CRC risk, with higher CV risk scores in the non-statin group, particularly in males. However, in the statin group, there was no increase in CRC risk, even with higher CV risk scores. The group using statins may have varying individual levels of CVD risk, and statin therapy itself may influence CVD risk, leading to potential bias in the analysis of CV risk scores and CRC risk. Furthermore, the effect of statins on CRC incidence cannot be disregarded. A meta-analysis suggested a modest reduction in CRC risk with statin use (RR = 0.90)[24]. However, long-term use of statins may even increase the risk of colon cancer (HR = 1.85 for > 15 years)[25], Our study analyzed data based on statin use status, without information on the duration of statin therapy; further analyses regarding the impact of statins could not be conducted.

DM is a well-recognized risk factor for CVD and CRC[26-28]. We analyzed whether the relationship between CV risk scores and CRC risk differed according to the presence or absence of DM. In our results, even in the absence of DM, a higher CV risk score was associated with an increased risk of CRC, with the risk increasing by 1.4 times in males without DM. In patients with DM, there was also a trend of increasing CRC risk, with higher CV risk scores. However, because DM can independently influence both CV and CRC risk, the impact attributed to CV risk may be offset.

According to a recent study, the incidence of cancers, including CRC, has been reported to be higher in patients with heart failure with preserved ejection fraction[29]. Additionally, there have been reports of an association between fatty liver disease and CRC[30,31]. Although our study did not include analyses on heart failure or fatty liver disease, future research investigating the association between metabolic dysfunction and CRC incidence would be valuable.

CRC is the second leading cause of cancer-related deaths; however, early detection and intervention can reduce mortality[32]. United States preventive services task force recommends screening for CRC starting at the age of 50 years in average-risk, asymptomatic adults[32]. Recently, there has been growing interest in personalized CRC screening, which involves the evaluation of risk factors such as lifestyle, environmental factors, and genetic variants[33-35]. Our findings suggest that individuals under 50 years of age with high CV risk scores may have an increased risk of developing CRC. Therefore, CRC screening may be beneficial in these cases.

This study benefits from being a nationwide population-based cohort study with long-term follow-up data spanning 10 years, which is a significant strength. Our study did not utilize well-known risk assessment tools such as the Framingham risk score (FRS) or the pooled cohort equation (PCE). This can be considered a limitation of the present study. However, it is worth noting that the FRS or PCE may overestimate atherosclerotic CVD risk in Koreans[36,37]. Therefore, we opted to evaluate the CV risk using the scoring system developed by the Korean Society of Lipid and Atherosclerosis, considering its relevance to the Korean population. Conducting a future study comparing and analyzing these three assessment tools would be beneficial. Another limitation of this study was the inclusion of patients with coronary artery disease, ischemic stroke, and aneurysms of the abdominal aorta, which categorizes them as having a high risk of CVD. These patients were also assessed using the CV risk score, and some were included in the lower-risk score groups. However, statistical adjustments were made to minimize the impact of these factors when evaluating the correlation between CRC and the CV risk score. Therefore, the influence of these high-risk patients on the association between CRC and the CV risk score was mitigated to some extent. Another limitation is that our data were extracted using international classification of diseases codes to determine the occurrence of CRC during the follow-up period. Although it is estimated that most CRCs are diagnosed through endoscopy, precise data on the proportion of cases diagnosed via this method cannot be obtained, and analysis of tumor location or pathological grading is not feasible. Further research is required to analyze the tumor location and pathological grading. Another point to consider is that given the elevated CRC risk in patients with precancerous gastric conditions, individuals with more CV and metabolic factors may also have a higher risk of gastric precancerous lesions, potentially influencing CRC risk indirectly[38]. However, as this study did not include data on endoscopic results or Helicobacter pylori infection, we were unable to analyze these. It would be beneficial to analyze this finding in future follow-up studies.

CONCLUSION

In conclusion, this study underscores an increased risk of CRC associated with higher CV risk scores, especially in males, the younger individuals, and those not on statin therapy, even those without diabetes. Thus, males with a higher CV risk score, even at a younger age, may benefit from effective interventions targeting their risk factors, which could also help prevent CRC. Additionally, individualized CRC screening approaches based on risk could be beneficial.