Association between remnant cholesterol and advanced colorectal adenoma in individuals under 50 years of age
Article information
Abstract
Background/Aims
The incidence of early-onset colorectal cancer is rising globally. While metabolic dysregulation is a known risk, the specific contribution of lipid profiles to colorectal carcinogenesis in young adults remains unclear. Remnant cholesterol (RC) has emerged as a significant cardiovascular risk factor, but its association with young-onset colorectal neoplasia is not fully elucidated.
Methods
We conducted a retrospective, cross-sectional study of 4,100 asymptomatic individuals under 50 years of age undergoing screening colonoscopy. RC was calculated as total cholesterol minus high-density lipoprotein cholesterol and low-density lipoprotein cholesterol. We used multivariate logistic regression to assess the independent association between RC levels and the presence of young-onset adenoma (YOA) and advanced YOA.
Results
Patients with YOA had significantly higher RC levels than controls (17.65 ± 14.52 mg/dL vs. 14.69 ± 11.14 mg/dL; P< 0.001). In adjusted multivariate analysis, RC was independently associated with YOA risk (odds ratio [OR], 1.011; 95% confidence interval [CI], 1.003–1.019; P= 0.005). Notably, for advanced YOA, RC remained the only lipid parameter that retained statistical significance in the multivariable model (OR, 1.021; 95% CI, 1.004–1.038; P= 0.015), whereas traditional lipid markers were not significant.
Conclusions
Higher RC levels were independently associated with YOA and advanced YOA in this cohort of asymptomatic individuals under 50 years of age. Among the lipid parameters examined, RC remained statistically significant in the multivariable model for advanced YOA.
INTRODUCTION
The incidence of colorectal cancer (CRC) among individuals younger than 50 years of age, termed early-onset CRC (EOCRC), has been rising globally at an alarming rate [1]. This epidemiological shift presents a profound public health challenge, as EOCRC often presents at more advanced stages and exhibits distinct, aggressive clinicopathological features compared to late-onset CRC [1,2]. Given that most CRCs arise from precursor adenomas, the early identification and removal of young-onset adenomas (YOAs) are critical strategies for interrupting the adenoma-carcinoma sequence [3]. Although current guidelines have lowered the screening age to 45 years, a significant number of younger adults remain at risk, underscoring the need for precise risk stratification markers to identify those who would benefit from earlier or more intensive surveillance [4,5].
Metabolic dysregulation, characterized by obesity, insulin resistance, and dyslipidemia (DL), is a well-established risk factor for CRC [6]. Specifically, elevated triglyceride (TG) levels have been identified as an independent risk factor for colorectal polyp recurrence, particularly in cases of advanced YOA [7].
However, the specific contributions of lipid fractions to colorectal carcinogenesis remain unclear. Although low-density lipoprotein cholesterol (LDL-C) is the primary target for cardiovascular risk reduction, its association with colorectal adenoma remains inconsistent [8,9]. This suggests that traditional lipid profiles may not fully capture the atherogenic and potentially carcinogenic metabolic burdens in younger populations [9]. Furthermore, routine lipid panels often overlook the complex interplay between lipid metabolism and systemic inflammation, which is highly prevalent in young adults with westernized dietary habits [10,11].
Remnant cholesterol (RC), defined as the cholesterol content of TG-rich lipoproteins, has recently emerged as a significant risk factor for residual cardiovascular risk and all-cause mortality, independent of LDL-C [12]. Unlike fasting TG, which primarily reflect inert energy storage, RC represents the biologically active, cholesterol-enriched particles that can directly penetrate the endothelial and mucosal barriers. Once localized within tissues, these remnant particles are readily engulfed by macrophages, driving foam cell formation and creating a highly pro-inflammatory microenvironment [13,14]. Biologically, these remnant lipoproteins induce low-grade systemic inflammation and oxidative stress, mechanisms that share pathophysiology with tumor promotion [15,16]. Despite this plausibility, data regarding the specific association between RC and the risk of colorectal neoplasia in young adults are scarce.
We hypothesized that RC may be associated with young-onset colorectal adenomatous lesions and could provide additional information beyond conventional lipid parameters. Therefore, we investigated the association between serum RC levels and YOA, particularly advanced YOA, in asymptomatic individuals younger than 50 years undergoing screening colonoscopy.
METHODS
1. Study Design and Population
This retrospective, cross-sectional study analyzed data from asymptomatic individuals younger than 50 years of age who underwent screening colonoscopy at the Health Screening Center of Dong-A University Hospital. The study period spanned from January 2018 to December 2024. Initially, we identified 6,025 eligible participants for this study. To ensure a homogeneous study population and accurate metabolic analysis, we applied the following exclusion criteria: (1) a history of prior colonoscopy (n=1,523); (2) a family history of CRC in a first-degree relative (n=346); (3) serum TG levels exceeding 500 mg/dL (n =43), as extreme hypertriglyceridemia precludes the reliable calculation of RC; and (4) missing clinical data (n =13). After applying these criteria, 4,100 participants were included in the final analysis (Fig. 1). In addition, no patients in the final study population were newly diagnosed with CRC at the index colonoscopy.
Flowchart of the study population selection. A total of 6,025 individuals aged <50 years who underwent screening colonoscopy at Dong-A University Hospital between 2018 and 2024 were initially assessed. After excluding 1,925 subjects based on predefined criteria, including prior colonoscopy history (n=1,523), family history of colorectal cancer in first-degree relatives (n=346), high serum triglyceride levels (n=43), and missing data (n=13), a final cohort of 4,100 participants was included in the analysis.
2. Definitions of Variables
Clinical data, including age, sex, and medical history, were retrieved from electronic medical records and self-reported questionnaires. Body mass index (BMI) was calculated as weight in kilograms divided by height squared in meters (kg/m2). Lifestyle factors, including smoking status and alcohol consumption, were evaluated using a standardized questionnaire. Smoking status was stratified as never, former, or current smoker. Comorbidities such as hypertension (HTN), diabetes mellitus, and DL were defined by a prior physician diagnosis or current use of relevant medications.
Venous blood samples were collected following an overnight fast of ≥ 8 hours on the same day as colonoscopy as part of the health screening evaluation. Serum concentrations of total cholesterol (TC), TG, high-density lipoprotein cholesterol (HDL-C), LDL-C, and fasting glucose were quantified using automated enzymatic assays. RC was calculated as TC–HDL-C–LDL-C. Although this calculated value is closely related to TG/5 when LDL-C is derived using the Friedewald equation in fasting samples, RC is conceptually intended to reflect the cholesterol content of TG-rich remnant lipoproteins rather than TG quantity itself.
YOA was defined as the presence of at least one histologically confirmed adenoma (tubular, villous, villo-tubular, or serrated) in individuals younger than 50 years of age. Advanced YOA was defined as an adenoma presenting with at least one of the following high-risk features: a diameter of 10 mm or larger, the presence of a villous component, or high-grade dysplasia.
3. Primary Endpoint
The primary endpoint of this study was to investigate the independent association between serum RC levels and the presence of YOA, compared with a control group without adenomas. Additionally, we evaluated the association between RC and the risk of advanced YOA to assess the potential of RC as a metabolic marker for the severity of colorectal neoplasia.
4. Statistics and Data Analysis
Statistical analyses were performed using SPSS software (version 26.0; IBM Corp., Armonk, NY, USA) and R software (version 4.5.3; R Foundation for Statistical Computing, Vienna, Austria). Continuous variables are expressed as means±standard deviations, and categorical variables as frequencies and percentages. Differences in baseline characteristics between the groups were compared using the Student t-test for continuous variables and the chi-square test for categorical variables. To identify independent risk factors associated with YOA and advanced YOA, univariate and multivariate logistic regression analyses were performed. Variables with a P-value of less than 0.05 in the univariate analysis were included in the multivariate models. Because RC was calculated as TC−HDL-C−LDL-C, it was not entered into the same multivariable model together with its component lipid variables, in order to avoid structural multicollinearity. Accordingly, RC was evaluated in a separate multivariable model from conventional lipid parameters. In addition, pairwise correlation analyses and collinearity diagnostics were performed to assess potential multicollinearity among the covariates included in the final models. Results are reported as odds ratios (ORs) with 95% confidence intervals (CIs). All P-values were two-sided, and a P-value of less than 0.05 was considered to indicate statistical significance.
5. IRB Approval
The study protocol was approved by the Institutional Review Board (IRB) of Dong-A University Hospital (IRB No. DAUHIRB-25-117) and was conducted in accordance with the Declaration of Helsinki. The requirement for informed consent was waived by the IRB due to the retrospective nature of the study and the use of anonymized clinical data.
RESULTS
1. Baseline Characteristics of the Study Population
The baseline characteristics of the study population are presented in Table 1. The mean age of the patients in the YOA group was significantly higher than that of the control group (41.42±5.55 years vs. 38.51±6.37 years; P<0.001). The proportion of males was also higher in the case group (66.7% vs. 53.5%; P<0.001). There were also significant differences between the 2 groups with respect to BMI, fasting glucose levels, and levels of TC, HDL-C, LDL-C (all P<0.001). Additionally, the level of RC was significantly higher in the YOA group than in the control group (17.65 ±14.52 mg/dL vs. 14.69 ±11.14 mg/dL; P<0.001). Similarly, the TG level was significantly elevated in the case group compared with the control group (129.59 ± 83.26 mg/dL vs. 111.87 ± 70.41 mg/dL; P<0.001). Regarding lifestyle factors and medical history, participants with YOA were more likely to be current smokers (26.3% vs. 18.4%; P<0.001) and to consume alcohol (52.4% vs. 47.8%; P=0.017) than those in the control group. The prevalence of HTN and DL was also significantly higher in the case group (P<0.001 and P=0.015, respectively), whereas the prevalence of diabetes mellitus did not differ significantly between the 2 groups.
Baseline Characteristics of the Study Population
2. Univariate and Multivariate Analyses of YOAs
In the univariate logistic regression analysis, age, male sex, higher BMI, and higher levels of fasting glucose were significantly associated with the presence of YOA (P<0.05 for all comparisons). Lipid profiles, including levels of TC, LDL-C, TG, and RC, were also significantly higher in patients with YOA than in controls. Variables that were significant in the univariate analysis were entered into the multivariable logistic regression model (Table 2). Age was a significant predictor of YOA (adjusted OR, 1.082; 95% CI, 1.067–1.097; P <0.001). Male sex was also strongly associated with an increased risk (adjusted OR, 1.614; 95% CI, 1.310–1.987; P<0.001). Among lifestyle factors and comorbidities, current smoking (adjusted OR, 1.250; 95% CI, 1.003–1.558; P=0.047) and a history of HTN (adjusted OR, 1.417; 95% CI, 1.046–1.918; P=0.024) remained significant in the multivariate model. In the multivariable model, RC remained significantly associated with YOA (adjusted OR, 1.011; 95% CI, 1.003–1.019; P=0.005). TC and LDL-C were evaluated in separate multivariable models and were both significantly associated with YOA. In additional analyses, pairwise correlation and collinearity diagnostics did not indicate problematic multicollinearity in the final multivariable model for YOA (Supplementary Tables 1 and 2).
Univariate and Multivariate Analyses Related to Young-Onset Adenoma
3. Univariate and Multivariate Analyses of Advanced YOAs
We further analyzed risk factors associated with advanced YOA. In the univariate analysis, older age (OR, 1.095; 95% CI, 1.038–1.156; P=0.001) and HTN (OR, 4.421; 95% CI, 2.102–9.294; P<0.001) were significantly associated with advanced YOA. Among lipid parameters, TC (P=0.043), RC (P=0.005), and DL (P= 0.032) were significant, whereas LDL-C (P= 0.157) and TG (P=0.106) were not. In the multivariable analysis, age (adjusted OR, 1.080; 95% CI, 1.022–1.141; P=0.006) and HTN (adjusted OR, 3.123; 95% CI, 1.432–6.810; P=0.004) remained significantly associated with advanced YOA. Among lipid parameters, TC and DL were no longer significant, whereas RC remained significantly associated with advanced YOA (adjusted OR, 1.021; 95% CI, 1.004–1.038; P=0.015) (Table 3). In additional analyses, collinearity diagnostics did not indicate problematic multicollinearity in the final multivariable model for advanced YOA (Supplementary Table 2).
Univariate and Multivariate Analyses Related to Advanced Young-Onset Adenoma
4. Additional Analysis Using Sex-Specific Low HDL-C Thresholds
In an additional analysis using sex-specific HDL-C thresholds, low HDL-C was defined as HDL-C <40 mg/dL in men and <50 mg/dL in women. Because RC was calculated from conventional lipid components, sex-specific low HDL-C was evaluated in a separate model rather than simultaneously with RC. In these additional analyses, low HDL-C was not independently associated with either YOA or advanced YOA (Supplementary Table 3).
DISCUSSION
In this study of 4,100 asymptomatic individuals under 50 years of age, higher RC levels were independently associated with YOA. In addition, RC remained significantly associated with advanced YOA in the multivariable analysis, whereas TC and LDL-C were not. These findings suggest that RC may reflect a metabolic profile associated with YOA and advanced YOA in this population.
The incidence of EOCRC is rising globally, necessitating more precise risk stratification strategies beyond age and family history [17]. While metabolic syndrome and insulin resistance are established risk factors for colorectal neoplasia, the specific contribution of individual lipid components has remained debated [18-20]. Our findings suggest that the association between RC and YOA is maintained independent of LDL-C [9]. In our analysis of YOA, although TG lost statistical significance after adjustment, RC remained significant. This suggests that RC may reflect aspects of TG-rich lipoprotein metabolism that are not fully captured by TG levels alone. This distinction may be relevant given the inconsistent evidence linking traditional lipid profiles to colorectal neoplasia. RC may therefore warrant further evaluation as a metabolic marker in this context [21,22]. Recent studies have suggested a link between RC and colorectal carcinogenesis [23,24]. In a 2024 study combining cross-sectional and Mendelian randomization analyses, RC was associated with CRC risk [23]. In addition, a more recent prospective cohort study reported that elevated RC was associated with a higher long-term risk of overall cancer, including CRC, and that this association was more pronounced in the presence of systemic inflammation [24]. In this context, our findings extend the literature by focusing on adenomatous precursor lesions, particularly YOA and advanced YOA, in asymptomatic individuals younger than 50 years.
A possible explanation for our findings is that RC reflects the cholesterol carried within TG-rich remnant lipoproteins, which may provide information distinct from that captured by TG, TC, or LDL-C alone. These remnant particles may be more closely related to low-grade inflammation, oxidative stress, and insulin resistance, all of which are biologically relevant to colorectal neoplasia [25,26]. Chronic low-grade inflammation, mediated by cytokines such as tumor necrosis factor-α and interleukin-6, is a known driver of colorectal tumorigenesis [27-29]. Furthermore, RC is closely linked to insulin resistance [30]. Hyperinsulinemia increases the bioavailability of insulin-like growth factor-1, which promotes cell proliferation and inhibits apoptosis in colonic epithelial cells [31,32]. Our finding that RC is associated with advanced YOA—lesions with a higher potential for malignant transformation—provides biological plausibility for a possible link between remnant lipoproteins and neoplastic development through these metabolic and inflammatory pathways. This is consistent with the broader understanding that metabolic dysregulation, including insulin resistance and chronic systemic inflammation, may contribute to early-onset colorectal carcinogenesis [33].
A distinguishing feature of our study is the association observed specifically in the advanced YOA group. In this subgroup, HTN and RC were the only significant modifiable risk factors. While LDL-C is the primary target for cardiovascular prevention, its role in CRC is inconsistent [9]. Our data showed that LDL-C was not a significant predictor for advanced lesions in the multivariate analysis. This may suggest that, in young adults, RC may reflect metabolic characteristics associated with advanced colorectal neoplasia beyond those captured by conventional lipid parameters, although this interpretation should be made cautiously given the interdependence among lipid measures. This highlights the potential relevance of specific lipid subfractions, such as RC, which may provide additional insight into metabolic perturbations associated with early-onset colorectal neoplasia beyond broadly measured lipid parameters [34,35].
Our study has several methodological strengths. We enforced strict exclusion criteria, specifically removing subjects with extreme hypertriglyceridemia ( >500 mg/dL). This approach reduced the potential influence of extreme metabolic outliers, addressing a common limitation of retrospective studies evaluating lipid parameters. Furthermore, the inclusion of a large cohort of strictly asymptomatic young adults undergoing screening colonoscopy minimizes the selection bias often associated with diagnostic colonoscopy studies. Additionally, adjustment for a wide range of potential confounders—including lifestyle factors, comorbidities, and traditional lipid markers—strengthened the robustness of the observed association between RC and advanced YOA.
However, there are limitations to consider. First, this was a single-center, retrospective study, which limits the ability to infer causality and generalize findings to other ethnic groups. Prospective, multicenter studies with direct RC measurements are needed to validate these findings and further investigate the underlying biological mechanisms. Second, because the study population was restricted to individuals younger than 50 years of age, we were unable to directly compare the association between RC and colorectal neoplasia across age groups. Therefore, the present findings should not be interpreted as demonstrating that this association is unique to YOA. Third, RC was calculated rather than directly measured. Although calculated RC is a widely accepted surrogate in large epidemiological studies, direct quantification could provide greater precision. In addition, because RC was derived from conventional lipid components, it is intrinsically interrelated with TC, HDL-C, and LDL-C. Therefore, although we used separate multivariable models to reduce structural multicollinearity, comparisons between RC and conventional lipid parameters should be interpreted cautiously. Similarly, DL history may partly reflect the underlying lipid abnormalities captured by RC. In addition, while we adjusted for smoking and alcohol, we could not account for dietary habits or physical activity, which are unmeasured confounders sharing a link with both lipid levels and adenoma risk. Finally, the number of advanced YOA cases was relatively small, although the association with RC remained statistically significant.
In conclusion, higher RC levels were independently associated with YOA, particularly advanced YOA, in asymptomatic individuals younger than 50 years. Our findings suggest that RC may provide additional information beyond traditional lipid markers in this specific population. Further studies are needed to determine whether RC may have a role in risk stratification for colorectal neoplasia in young adults.
Notes
Funding Source
The authors received no financial support for the research, authorship, and/or publication of this article.
Conflict of Interest
No potential conflict of interest relevant to this article was reported.
Data Availability Statement
Data analyzed in this study are available from the corresponding author upon reasonable request.
Author Contributions
Conceptualization: Lee JY. Data curation: Joo K. Formal analysis: Joo K, Lee JY. Funding acquisition: Lee JY. Investigation; Methodology: Joo K, Lee JY. Project administration: Lee JH, Lee JY. Resources; Software: Joo K, Lee JY. Supervision; Validation: Lee JH. Visualization: Joo K, Lee JY. Writing–original draft: Joo K, Lee JY. Writing–review & editing: all authors. Approval of final manuscript: all authors.
Supplementary Material
Supplementary materials are available at the Intestinal Research website (https://www.irjournal.org).
Supplementary Table 1.
Pairwise Correlations among Candidate Variables Included in the Multivariable Analyses
Supplementary Table 2.
Collinearity Diagnostics for the Final Multivariable Models for YOA and Advanced YOA
Supplementary Table 3.
Additional Analyses Using Sex-Specific Low HDL-C Categories for YOA and Advanced YOA