Abstract
Background
Cholecystectomy is widely used to treat gallbladder disease, but its link to colorectal cancer (CRC) remains controversial. This study aimed to assess the association between cholecystectomy and CRC by combining cross-sectional analysis with Mendelian randomization (MR).
Methods
We analyzed 12,490 adults from the National Health and Nutrition Examination Survey (NHANES) 2017–2023 for the association between cholecystectomy and CRC. Weighted logistic regression models with progressive adjustments were applied: Model 1 was unadjusted, Model 2 accounted for sociodemographic factors, while Model 3 was further controlled for lifestyle, diet, and comorbidities. Subgroup and sensitivity analyses were conducted to evaluate the results’ robustness. MR analysis further complemented the observational analysis and evaluated potential causality.
Results
In the NHANES sample, cholecystectomy was not significantly associated with CRC after full adjustment [Model 3: odds ratio (OR) =2.06; 95% confidence interval (CI): 0.93–4.55; P=0.07], although crude and partially adjusted models showed positive associations (Model 1: OR =3.77, 95% CI: 2.08–6.83, P<0.001; Model 2: OR =2.35, 95% CI: 1.23–4.51, P=0.01). The association remained non-significant across multiple sensitivity analyses and was consistent across population subgroups. MR analysis further indicated no causal link between cholecystectomy and CRC risk.
Conclusions
No statistically significant overall association was observed between cholecystectomy and CRC in our analysis.
Keywords: Cholecystectomy, colorectal cancer (CRC), National Health and Nutrition Examination Survey (NHANES), Mendelian randomization (MR)
Highlight box.
Key findings
• No statistically significant overall association was observed between cholecystectomy and colorectal cancer (CRC) in National Health and Nutrition Examination Survey and Mendelian randomization (MR) analyses.
What is known and what is new?
• Previous studies on cholecystectomy and CRC risk have reported inconsistent findings.
• This study combines population-based data and MR, and offers new evidence against a causal link between cholecystectomy and CRC.
What is the implication, and what should change now?
• Clinicians can proceed with cholecystectomy when indicated, with no need to have undue concern for increasing CRC risk.
Introduction
Due to the high prevalence of gallbladder diseases, cholecystectomy has become one of the most performed abdominal surgeries. It is widely endorsed in clinical guidelines as the standard treatment for various gallbladder conditions, including symptomatic gallstones, gallbladder polyps, and cholecystitis (1-3). Although being generally safe and effective for symptom relief, cholecystectomy induces continuous bile flow into the duodenum, bypassing the gallbladder’s meal-related regulation (4). This alteration may prolong colonic exposure to primary bile acids and increase the formation of secondary bile acids such as deoxycholic acid (DCA). Some experimental studies indicate that DCA may promote colorectal carcinogenesis via Wnt pathway activation and oxidative DNA damage (5,6). Concurrently, post-cholecystectomy bile acid dysregulation may alter gut microbiota composition, thereby increasing tauroursodeoxycholic acid production (7) and potentially creating a pro-carcinogenic microenvironment (4,8-10).
Despite these plausible biological mechanisms, epidemiological studies are yet to reach a consistent conclusion regarding the association between cholecystectomy and colorectal cancer (CRC) (11). Several investigations suggest a heightened risk of CRC following cholecystectomy (12-15), while other studies report no significant correlation (16-20). Heterogeneity in study designs and covariate adjustment may underlie these inconsistencies. Several lifestyle factors, including smoking (21), alcohol consumption (22), sedentary behavior (23-25), dietary patterns (26,27), or inadequate Vitamin C intake (28), have been increasingly recognized as relevant to CRC risk. However, many prior studies did not fully adjust for these factors, which could have biased their observed associations.
The National Health and Nutrition Examination Survey (NHANES) is an ongoing cross-sectional study designed to assess the nutritional status and emerging health concerns of the United States population (29), which provides detailed information on demographics, lifestyle, diet, and health conditions. Its rich data facilitate the analysis of the cholecystectomy-CRC association with more comprehensive control of potential confounders. MR is an approach that leverages genetic variants as proxies for exposures to evaluate potential causal effects on outcomes (30-33). As genetic variants are allocated randomly at conception and tend to be independent of environmental factors or personal habits, MR reduces susceptibility to confounding and reverse causality (34). These inherent properties make MR often considered analogous to randomized controlled trials and useful for strengthening findings from observational research.
Given the widespread use of cholecystectomy and the rising incidence of CRC, clarifying whether cholecystectomy contributes causally to CRC risk has important clinical implications. A positive association could provide a rationale for postoperative surveillance, while a null finding may help alleviate patient concerns about an inevitable cancer risk after surgery. Considering the inconsistent findings and limited adjustment for confounders in previous research, our study aims to provide new evidence from large-scale NHANES data and two-sample MR analyses to shed further light on the association between cholecystectomy and CRC. We present this article in accordance with the STROBE and STROBE-MR reporting checklists (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1493/rc).
Methods
Study participants in the NHANES
This cross-sectional survey included participants from the continuous NHANES cycles spanning from March 2017 to August 2023, which represented a nationally representative sample. These specific cycles were chosen because participants completed the cholecystectomy questionnaire during that period. Between March 2017 and August 2023, a total of 27,493 individuals participated in NHANES, including 15,560 from the 2017–2020 cycle and 11,933 from the 2021–2023 cycle. Participants with missing cholecystectomy data (n=10,465), missing CRC data (n=11), or missing weight data (n=4,527) were excluded. Ultimately, 12,490 adults aged ≥20 years were included. A detailed flowchart was shown in Figure 1. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.
Figure 1.
Flowchart of participant selection process. NHANES, National Health and Nutrition Examination Survey.
Definition and assessment of CRC and cholecystectomy in NHANES
Self-reported cancer and gallbladder surgery were recorded in medical conditions within NHANES. CRC status was identified based on whether participants reported being told by a healthcare professional that they had cancer, followed by specification of the cancer type. Those who indicated colon or rectal cancer were classified as having CRC (35,36). The history of cholecystectomy was determined by asking participants if they had ever undergone gallbladder removal surgery, with affirmative responses indicating prior cholecystectomy.
Covariates in NHANES
This study included covariates previously associated with or hypothesized to be associated with CRC. These covariates were age, sex, race/ethnicity, educational attainment, poverty income ratio (PIR), marital status, smoking status, alcohol consumption, body mass index (BMI), sedentary time, sleep duration, diabetes, hypertension, total energy intake, and intakes of Vitamin C, protein, sugar, fat, and dietary fiber. The specific definitions and categorizations of these covariates were described in Table S1 (37-40).
Data sources for MR studies
The summary statistics from the largest genome-wide association study (GWAS) on cholecystectomy were available on the “Open GWAS” platform. In this study, the exposure was clearly defined as “cholecystectomy/gallbladder removal”. The GWAS dataset comprises 18,319 individuals who underwent cholecystectomy and 444,614 controls. CRC genetic summary statistics were obtained from FinnGen DF12, comprising 11,790 cases and 378,749 controls, and the open GWAS ID for this data is “finngen_R12_C3_ COLORECTAL_EXALLC”. All GWAS datasets used in the analysis were approved by the respective Ethics Committees and were publicly accessible. The sources of MR data were detailed in Table S2.
MR design
MR analysis adhered to three fundamental assumptions: (I) relevance: genetic variants must demonstrate significant association with cholecystectomy; (II) independence: variants must remain unassociated with confounding variables; and (III) exclusion restriction: variants influence CRC exclusively through cholecystectomy. Only single-nucleotide polymorphisms (SNPs) available in both exposure and outcome datasets were retained. To construct valid instruments, we identified SNPs linked to cholecystectomy (P<5×10−8), followed by exclusion of those with weak strength (F-statistic <10) or linkage disequilibrium (r2>0.001 within a 10,000 kb window). The causal impact of cholecystectomy on CRC was subsequently evaluated via MR methods.
Statistical analysis
In the NHANES analysis, sample weights for each observation were calculated using the recommended first-day dietary recall sample weights (WTDRD1) from NHANES, covering March 2017 through August 2023, with adjustments for clustering and stratification arising from NHANES’ complex sampling design (i.e., 1/2.625 * WTDRD1). Baseline characteristics were evaluated using the Rao-Scott Chi-squared test for categorical variables and unadjusted linear regression for continuous data. Categorical variables were reported as frequencies accompanied by weighted percentages, whereas continuous variables were presented as weighted means with corresponding standard deviations (SDs).
Survey-weighted logistic regression served as the main analytical strategy to examine the link between cholecystectomy and CRC, taking potential confounders into account. Three models with progressively varied covariate adjustments were analyzed: Model 1 was unadjusted; Model 2 included adjustments for age, sex, race/ethnicity, educational attainment, PIR, and marital status; while Model 3 incorporated additional adjustments for alcohol consumption, smoking status, BMI, sedentary time, sleep duration, diabetes, hypertension, total energy intake, Vitamin C intake, protein intake, sugar intake, fat intake, and fiber intake. Missing values for covariates were handled through multiple imputation by chained equations with the “mi” package in R. To explore subpopulations potentially affected by sociodemographic disparities, stratified analyses were performed according to age group, sex, race/ethnicity, educational attainment, PIR, and marital status. Interaction terms between cholecystectomy and each stratification variable were included, and the corresponding P values were used to assess the significance of effect modification. We also conducted sensitivity analyses as follows: (I) a complete-case analysis excluding missing data to verify consistency with the imputed results; (II) propensity score matching (PSM) to reduce group-size imbalance and balance the covariates between CRC and non-CRC groups (41,42), with covariate balance assessed by standardized mean differences (SMDs) and P values (SMD <0.1 considered strict balance, 0.1–0.2 moderate balance, and P>0.05 indicating no statistically significant difference) (43-46); and (III) a simplified model using the imputed data, adjusting for key confounders to prevent over-fitting and assess the robustness of the findings.
The two-sample MR analysis utilized inverse variance weighting (IVW) as its principal analytical strategy. Robustness of IVW estimates was examined through MR-Egger, weighted median, maximum likelihood, and weighted mode. Directional pleiotropy was evaluated using the MR-Egger intercept. Horizontal pleiotropic outliers were detected using the MR-PRESSO outlier test. Heterogeneity across SNPs was evaluated with Cochran’s Q test. Finally, leave-one-out analysis was conducted to assess if any single variant influenced the results.
All statistical analyses and data visualizations were performed using R software (version 4.4.2). Two-sided tests were applied, with significance set at P<0.05.
Results
Population characteristics of study subjects by CRC status
The characteristics of study participants were shown in Table 1. The study involved 12,490 participants [age: mean ± SD, 48.00±17.39 years; 6,633 (51.67%) females; 5,720 (61.87%) non-Hispanic White individuals]. Participants were categorized into 102 with CRC (weighted 1%) and 12,388 without CRC (weighted 99%). Compared to individuals without CRC, those with CRC were generally older, more likely to be widowed, divorced, or separated, and had higher rates of alcohol use, smoking, and comorbidities such as diabetes or hypertension. They also had lower intakes of Vitamin C and protein and were more likely to have undergone cholecystectomy.
Table 1. Demographic and clinical characteristics of participants without or with colorectal cancer.
| Characteristic | Total† | Colorectal cancer status | P§ | ||
|---|---|---|---|---|---|
| Overall (n=12,490; 100%)‡ | Non-CRC (n=12,388; 99%)‡ | CRC (n=102; 1%)‡ | |||
| Age (years) | 12,490 | 48.00 (17.39) | 48.00 (17.36) | 68.00 (13.89) | <0.001 |
| Age group | 12,490 | <0.001 | |||
| <65 | 8,888.00 (77.91) | 8,855.00 (78.16) | 33.00 (39.64) | ||
| ≥65 | 3,602.00 (22.09) | 3,533.00 (21.84) | 69.00 (60.36) | ||
| Sex | 12,490 | 0.97 | |||
| Female | 6,633.00 (51.67) | 6,581.00 (51.67) | 52.00 (51.98) | ||
| Male | 5,857.00 (48.33) | 5,807.00 (48.33) | 50.00 (48.02) | ||
| Race/ethnicity | 12,490 | 0.40 | |||
| Non-Hispanic White | 5,720.00 (61.87) | 5,669.00 (61.82) | 51.00 (69.74) | ||
| Non-Hispanic Black | 2,623.00 (11.48) | 2,601.00 (11.48) | 22.00 (11.81) | ||
| Other Hispanic | 1,254.00 (8.73) | 1,246.00 (8.76) | 8.00 (4.41) | ||
| Mexican American | 1,188.00 (7.68) | 1,180.00 (7.70) | 8.00 (4.24) | ||
| Other race—including multi-racial | 1,705.00 (10.24) | 1,692.00 (10.25) | 13.00 (9.80) | ||
| Educational attainment | 12,490 | 0.11 | |||
| Below high school | 1,892.00 (9.60) | 1,868.00 (9.58) | 24.00 (12.57) | ||
| High school | 2,820.00 (25.93) | 2,791.00 (25.86) | 29.00 (37.13) | ||
| Above high school | 7,778.00 (64.46) | 7,729.00 (64.56) | 49.00 (50.30) | ||
| Poverty income ratio | 12,490 | 0.50 | |||
| ≤1 | 2,108.00 (12.81) | 2,089.00 (12.81) | 19.00 (13.11) | ||
| >1 and <4 | 6,865.00 (51.31) | 6,803.00 (51.27) | 62.00 (58.21) | ||
| ≥4 | 3,517.00 (35.88) | 3,496.00 (35.92) | 21.00 (28.68) | ||
| Marital status | 12,490 | <0.001 | |||
| Married or living with partner | 7,065.00 (60.65) | 7,017.00 (60.72) | 48.00 (49.05) | ||
| Never married | 2,439.00 (20.39) | 2,425.00 (20.46) | 14.00 (9.87) | ||
| Widowed or divorced or separated | 2,986.00 (18.96) | 2,946.00 (18.81) | 40.00 (41.08) | ||
| Alcohol consumption | 12,490 | 0.01 | |||
| Never | 1,503.00 (9.86) | 1,478.00 (9.77) | 25.00 (23.07) | ||
| Mild | 5,657.00 (43.94) | 5,623.00 (43.98) | 34.00 (38.27) | ||
| Moderate | 2,637.00 (21.98) | 2,619.00 (22.03) | 18.00 (12.92) | ||
| Heavy | 2,693.00 (24.22) | 2,668.00 (24.21) | 25.00 (25.75) | ||
| Smoking status | 12,490 | 0.02 | |||
| Never | 7,249.00 (60.10) | 7,203.00 (60.15) | 46.00 (53.06) | ||
| Former | 3,167.00 (24.51) | 3,121.00 (24.41) | 46.00 (39.59) | ||
| Now | 2,074.00 (15.39) | 2,064.00 (15.44) | 10.00 (7.35) | ||
| Body mass index (kg/m2) | 12,490 | 0.13 | |||
| <18.5 | 164.00 (1.32) | 163.00 (1.33) | 1.00 (0.67) | ||
| ≥18.5 and <25 | 2,972.00 (24.61) | 2,955.00 (24.67) | 17.00 (14.73) | ||
| ≥25 and <30 | 3,950.00 (32.25) | 3,914.00 (32.16) | 36.00 (45.19) | ||
| ≥30 | 5,404.00 (41.82) | 5,356.00 (41.84) | 48.00 (39.41) | ||
| Sedentary time (hours) | 12,490 | 0.20 | |||
| <4 | 3,551.00 (25.26) | 3,527.00 (25.31) | 24.00 (17.57) | ||
| 4–6 | 4,572.00 (37.79) | 4,537.00 (37.81) | 35.00 (34.40) | ||
| >6 | 4,367.00 (36.95) | 4,324.00 (36.88) | 43.00 (48.03) | ||
| Sleep duration (hours) | 12,490 | 0.80 | |||
| <7 | 2,989.00 (22.12) | 2,967.00 (22.09) | 22.00 (25.47) | ||
| 7–9 | 8,121.00 (68.20) | 8,058.00 (68.23) | 63.00 (63.84) | ||
| >9 | 1,380.00 (9.68) | 1,363.00 (9.68) | 17.00 (10.69) | ||
| Diabetes | 12,490 | 3,267.00 (20.92) | 3,218.00 (20.72) | 49.00 (52.05) | <0.001 |
| Hypertension | 12,490 | 5,732.00 (38.37) | 5,660.00 (38.15) | 72.00 (71.98) | <0.001 |
| Total energy intake (kcal) | 12,490 | 1,940.00 (969.34) | 1,943.00 (969.26) | 1,820.00 (970.91) | 0.055 |
| Dietary Vitamin C intake (mg/d) | 12,490 | 49.50 (86.81) | 49.80 (86.88) | 33.20 (73.10) | 0.040 |
| Dietary protein intake (g/d) | 12,490 | 71.66 (41.41) | 71.72 (41.47) | 63.08 (29.00) | 0.03 |
| Dietary sugar intake (g/d) | 12,490 | 84.90 (72.79) | 84.97 (72.74) | 76.92 (80.95) | 0.80 |
| Dietary fat intake (g/d) | 12,490 | 78.77 (48.28) | 78.82 (48.31) | 69.67 (42.23) | 0.07 |
| Dietary fiber intake (g/d) | 12,490 | 14.30 (11.01) | 14.30 (11.03) | 14.40 (6.95) | 0.50 |
| Gallbladder surgery | 12,490 | 1,462.00 (10.95) | 1,428.00 (10.82) | 34.00 (31.36) | <0.001 |
†, n (unweighted); ‡, weighted means with associated standard deviations for continuous variables; number and corresponding weighted proportions for categorical variables; §, design-based Kruskal-Wallis test; Pearson’s χ2: Rao & Scott adjustment. CRC, colorectal cancer.
Observational relationships between cholecystectomy and CRC in the NHANES cohort
In Table 2, cholecystectomy was significantly associated with CRC in Model 1 [odds ratio (OR) =3.77; 95% confidence interval (CI): 2.08–6.83; P<0.001] and Model 2 (OR =2.35; 95% CI: 1.23–4.51; P=0.01). However, after further adjustment for all covariates in Model 3, this association became non-significant (OR =2.06; 95% CI: 0.93–4.55; P=0.07).
Table 2. Weighted logistic regression models assessing the association between cholecystectomy and colorectal cancer.
| Cholecystectomy | Model 1 | Model 2 | Model 3 | |||||
|---|---|---|---|---|---|---|---|---|
| OR (95% CI) | P | OR (95% CI) | P | OR (95% CI) | P | |||
| No | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||||
| Yes | 3.77 (2.08–6.83) | <0.001 | 2.35 (1.23–4.51) | 0.01 | 2.06 (0.93–4.55) | 0.07 | ||
Model 1: unadjusted. Model 2: adjusted for age (continuous), sex, race/ethnicity, marital status, poverty income ratio, and educational attainment. Model 3: adjusted for Model 2 plus alcohol consumption, smoking status, body mass index, sedentary time, sleep duration, diabetes, hypertension, total energy intake, dietary Vitamin C intake, dietary protein intake, dietary sugar intake, dietary fat intake, and dietary fiber intake. CI, confidence interval; OR, odds ratio.
Subgroup and sensitivity analysis in the NHANES cohort
Subgroup analyses revealed no significant interactions across demographic and socioeconomic strata (all P values for interaction >0.05, Table 3), indicating that the link between cholecystectomy and CRC did not vary meaningfully across these groups. Sensitivity analyses yielded stable findings. When participants with missing data (n=4,242) were excluded (Table 4), the estimated association remained consistent with the main analysis (Table 2). In the PSM analysis, covariates were generally balanced between groups (Table S3), and multivariable logistic regression was applied post-matching to reduce residual confounding (47). The association between cholecystectomy and CRC remained non-significant (Table 4). In addition, the simplified model adjusting for key confounders using imputed data also produced comparable results (Table 4). These analyses supported the internal consistency and robustness of the association estimates.
Table 3. Subgroup analysis investigating the association between cholecystectomy and colorectal cancer.
| Characteristic | OR (95% CI) | P | P for interaction |
|---|---|---|---|
| Age group (years) | 0.67 | ||
| <65 | 3.05 (0.74–12.47) | 0.11 | |
| ≥65 | 1.98 (0.99–3.97) | 0.05 | |
| Sex | 0.62 | ||
| Female | 1.48 (0.59–3.72) | 0.36 | |
| Male | 3.47 (1.32–9.12) | 0.02 | |
| Race/ethnicity | 0.67 | ||
| Non-Hispanic White | 2.12 (0.94–4.81) | 0.07 | |
| Non-Hispanic Black | 1.47 (0.26–8.35) | 0.64 | |
| Other Hispanic | 2.20 (0.51–9.53) | 0.27 | |
| Other race—including multi-racial | 3.52 (1.17–10.56) | 0.03 | |
| Educational attainment | 0.09 | ||
| Below high school | 8.31 (1.84–37.59) | 0.01 | |
| High school | 0.72 (0.20–2.62) | 0.58 | |
| Above high school | 4.09 (1.73–9.69) | 0.004 | |
| Poverty income ratio | 0.90 | ||
| ≤1 | 6.05 (1.66–22.10) | 0.01 | |
| >1 and <4 | 1.51 (0.63–3.57) | 0.32 | |
| ≥4 | 4.45 (0.62–31.77) | 0.12 | |
| Marital status | 0.16 | ||
| Never married | 0.37 (0.01–9.41) | 0.51 | |
| Married or living with partner | 3.19 (1.41–7.22) | 0.01 | |
| Widowed or divorced or separated | 1.77 (0.62–5.10) | 0.26 |
Each stratification was adjusted for age (continuous), sex, race/ethnicity, marital status, poverty income ratio and educational attainment, alcohol consumption, smoking status, body mass index, sedentary time, sleep duration, diabetes, hypertension, total energy intake, dietary Vitamin C intake, dietary protein intake, dietary sugar intake, dietary fat intake, and dietary fiber intake. CI, confidence interval; OR, odds ratio.
Table 4. Sensitivity analysis of the association between cholecystectomy and colorectal cancer.
| Cholecystectomy | Complete-data analysis† | Propensity score matching‡ | Simplified model using imputed data§ | |||||
|---|---|---|---|---|---|---|---|---|
| OR (95% CI) | P | OR (95% CI) | P | OR (95% CI) | P | |||
| No | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||||
| Yes | 2.15 (0.66–6.98) | 0.20 | 1.82 (0.86–3.93) | 0.12 | 2.14 (0.78–5.86) | 0.13 | ||
†, excluded participants with missing values in any covariates from Model 3 (n=4,242), resulting in a final analytic sample of 8,248 from the original 12,490. Adjusted for all variables included in Model 3; ‡, matching for age, sex, race/ethnicity, marital status, poverty income ratio, educational attainment, alcohol consumption, smoking status, sedentary time, and body mass index. Further adjusted for all variables included in Model 3 using multivariable logistic regression; §, based on multiple imputation dataset; adjusted for key covariates to avoid overfitting, including age, sex, race/ethnicity, marital status, poverty income ratio, educational attainment, alcohol consumption, smoking status, and body mass index. CI, confidence interval; OR, odds ratio.
Causal associations between cholecystectomy and CRC in MR
In the MR analysis, cholecystectomy was treated as the exposure and CRC as the outcome. To evaluate potential horizontal pleiotropy and outlier influence, we applied the MR-PRESSO test, which identified three outlier SNPs (rs11543269, rs2727270, and rs8112972) (Table S4). After excluding these outliers, 39 SNPs were retained as valid instrumental variables in subsequent MR analyses.
No indication of directional pleiotropy was detected based on the MR-Egger intercept (P=0.22), supporting the validity of these instrumental variables. However, residual heterogeneity still persisted, as demonstrated by Cochran’s Q statistics for both MR-Egger (P=1.44×10−3; I2=45.1%) and IVW (P=9.78×10−4; I2=46.3%) (Table S5). Due to observed heterogeneity, the random effects IVW model was prioritized for primary inference.
The random effects IVW analysis revealed no evidence of cholecystectomy’s causal effect on CRC (OR =1.98; 95% CI: 0.61–6.47; Table 5). Similarly, additional MR methods, including MR-Egger (OR =0.94; 95% CI: 0.18–4.93), weighted median (OR =1.34; 95% CI: 0.41–4.44), maximum likelihood (OR =2.03; 95% CI: 0.85–4.87), and weighted mode (OR =1.25; 95% CI: 0.40–3.93), also failed to provide statistically significant evidence of a causal effect (Table 5). The causal effect estimates from various MR methods were shown in Figure 2A, and the forest plot of individual SNP effects was presented in Figure 2B. Leave-one-out analysis (Figure 2C) indicated that no single SNP materially influenced the overall effect estimate. The funnel plot’s symmetry (Figure 2D) suggested little visual evidence of directional pleiotropy. These results supported the robustness of the MR estimates.
Table 5. MR estimation of causal effect of cholecystectomy on colorectal cancer risk.
| Method | OR | 95% CI | P |
|---|---|---|---|
| Inverse variance weighted | 1.98 | 0.61–6.47 | 0.26 |
| MR Egger | 0.94 | 0.18–4.93 | 0.94 |
| Weighted median | 1.34 | 0.41–4.44 | 0.63 |
| Maximum likelihood | 2.03 | 0.85–4.87 | 0.11 |
| Weighted mode | 1.25 | 0.40–3.93 | 0.71 |
CI, confidence interval; MR, Mendelian randomization; OR, odds ratio.
Figure 2.
MR plots for the relationship of cholecystectomy with CRC. (A) Scatter plot of five MR methods. Points represent individual SNPs and regression lines correspond to the different methods. (B) Forest plot of individual and combined SNPs MR-estimated effect sizes. (C) Leave-one-out sensitivity analysis, with estimates calculated after sequentially removing each SNP. (D) Funnel plot of SNP effects, showing symmetry and no evidence of directional pleiotropy. CRC, colorectal cancer; IV, instrumental variable; MR, Mendelian randomization; SE, standard error; SNP, single-nucleotide polymorphism.
Discussion
In this study, we integrated nationally representative data from NHANES (2017–2023) with MR analysis to explore whether cholecystectomy is associated with CRC. After controlling for comprehensive covariates, no significant overall link was observed between cholecystectomy and CRC. The MR analysis further provided support by revealing no evidence of a causal link.
Over the past few decades, the potential link between cholecystectomy and CRC has remained controversial. In 2017, a meta-analysis of 10 studies reported an increased risk of CRC following cholecystectomy (14), whereas a more recent meta-analysis of 14 studies found no significant overall association (20). Numerous observational studies have also explored this association, yet their results have been inconsistent. A case-control study found that cholecystectomy was not a notable contributor to CRC, irrespective of sex, familial predisposition, tumor location, or mismatch repair status (17). Similarly, a Korean cohort study observed no long-term increase in CRC risk after applying a post-surgical lag period, suggesting that earlier positive findings may have been influenced by bias (19). However, another Korean population-based study presented contrasting results, showing increased CRC risk post-cholecystectomy, especially in females (13). Additionally, a cohort study from Taiwan reported an elevated risk of both colorectal and gastric cancers within 5 years post-cholecystectomy, which persisted beyond that period (15). Variability in outcomes may arise from inconsistencies in research frameworks and the extent of covariate adjustment. In fact, most previous studies did not adequately control for key confounding variables, particularly lifestyle-related factors such as diet, sedentary behavior, or BMI (13,15,19,48). Our study partially overcomes these limitations by leveraging NHANES data, which enables more rigorous adjustment for potential confounders.
Notably, in our study, the association between cholecystectomy and CRC was significant in unadjusted and partially adjusted models, but became non-significant after adjustment for a comprehensive set of covariates (P=0.07). The final P value did not reach the conventional threshold for statistical significance. However, the result should be interpreted with caution, as it is marginal and may indicate a residual association. Larger or prospective studies are needed to determine whether this trend reflects a true association or arises from limited statistical power. Nevertheless, our findings still indicate that the positive associations reported in earlier studies may have been driven by insufficient adjustment for key factors such as health behaviors, diet, and metabolic conditions.
Regarding sex-specific effects, prior studies reported inconsistent findings, with some suggesting elevated CRC risk after cholecystectomy among females (13,14). However, our subgroup analyses found no evidence of effect modification by sex or other key demographic factors. This discrepancy may partly stem from differences in sample representativeness or the extent of covariate adjustment. The limited number of CRC cases in our sample may also have reduced the statistical power to detect subtle subgroup effects. To enhance the reliability of our results, we carried out a range of sensitivity checks, including exclusion of missing data, PSM, and simplified covariate adjustment. All approaches yielded consistent null findings.
Nonetheless, observational findings are inevitably influenced by residual confounding and reverse causation. To address this, we employed MR as an independent approach. This analysis likewise revealed no indication of a causal effect of cholecystectomy on CRC. Two previous MR studies had already examined this association, with results consistent with ours (16,49). Compared with these earlier studies, our MR analysis utilized the CRC GWAS dataset from the latest FinnGen release (DF12), which had a larger sample size (n=390,538), enhancing the statistical power and robustness of the results.
From a clinical perspective, our overall null findings are reassuring. For patients with clear surgical indications, cholecystectomy should not be delayed or withheld out of excessive concern for increased CRC risk. In our study, after accounting for modifiable risk factors such as lifestyle behaviors, dietary patterns, and metabolic comorbidities, cholecystectomy showed no statistically significant association with a higher prevalence of CRC. This finding may also imply the importance of post-cholecystectomy lifestyle and dietary management. However, since the P value was marginal, subtle risks in certain patient groups cannot be entirely excluded. The association between cholecystectomy and CRC requires further validation. One key advantage of our study lies in the utilization of NHANES, a nationally representative dataset containing detailed information on numerous potential risk variables, allowing for comprehensive covariate adjustment. Moreover, we employed multiple complementary analytical approaches, including sensitivity analyses, subgroup analyses, and MR. The consistency of findings across these methods further enhances the credibility and robustness of our conclusions.
However, there are certain limitations in this study. Firstly, we did not perform subgroup analyses by tumor location due to the limited number of CRC cases in the NHANES dataset, although this figure is consistent with the actual CRC incidence of 1% in American society (50). This is noteworthy, as some prior studies have proposed that cholecystectomy may be more closely linked to malignancies in the proximal colon (12,13,51,52). Secondly, the relatively small number of CRC cases compared with controls may have limited statistical power and introduced some instability in the estimates, although we performed PSM sensitivity analysis to alleviate this concern (41,42). Thirdly, both cholecystectomy and CRC diagnoses were self-reported by participants rather than verified from medical records, which may introduce recall bias or misclassification. Such reporting errors could affect the accuracy of case ascertainment and potentially bias the observed associations. Fourthly, NHANES lacks timing data for diagnoses, limiting the assessment of latency effects. Although MR mitigates reverse causation, it may not fully capture the complex biological effects induced by surgery. Lastly, as our analyses were primarily based on U.S. and European populations, the applicability of our findings to other ethnic groups remains uncertain.
Conclusions
In conclusion, our study found no overall link between cholecystectomy and CRC. Future research should focus on prospective studies with extended follow-up periods, adequately powered subgroup analyses, and thorough confounder adjustment to elucidate individual variability in CRC risk after cholecystectomy.
Supplementary
The article’s supplementary files as
Acknowledgments
None.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.
Footnotes
Reporting Checklist: The authors have completed the STROBE and STROBE-MR reporting checklists. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1493/rc
Funding: This work was granted by the National Natural Science Foundation of China (Nos. 82173316 and 82473092).
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1493/coif). The authors have no conflicts of interest to declare.
References
- 1.Agresta F, Campanile FC, Vettoretto N, et al. Laparoscopic cholecystectomy: consensus conference-based guidelines. Langenbecks Arch Surg 2015;400:429-53. 10.1007/s00423-015-1300-4 [DOI] [PubMed] [Google Scholar]
- 2.Pisano M, Allievi N, Gurusamy K, et al. 2020 World Society of Emergency Surgery updated guidelines for the diagnosis and treatment of acute calculus cholecystitis. World J Emerg Surg 2020;15:61. 10.1186/s13017-020-00336-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.van Dijk AH, de Reuver PR, Besselink MG, et al. Assessment of available evidence in the management of gallbladder and bile duct stones: a systematic review of international guidelines. HPB (Oxford) 2017;19:297-309. 10.1016/j.hpb.2016.12.011 [DOI] [PubMed] [Google Scholar]
- 4.Housset C, Chrétien Y, Debray D, et al. Functions of the Gallbladder. Compr Physiol 2016;6:1549-77. 10.1002/cphy.c150050 [DOI] [PubMed] [Google Scholar]
- 5.Fu T, Coulter S, Yoshihara E, et al. FXR Regulates Intestinal Cancer Stem Cell Proliferation. Cell 2019;176:1098-1112.e18. 10.1016/j.cell.2019.01.036 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Yao Y, Li X, Xu B, et al. Cholecystectomy promotes colon carcinogenesis by activating the Wnt signaling pathway by increasing the deoxycholic acid level. Cell Commun Signal 2022;20:71. 10.1186/s12964-022-00890-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Tang B, Li S, Li X, et al. Cholecystectomy-related gut microbiota dysbiosis exacerbates colorectal tumorigenesis. Nat Commun 2025;16:7638. 10.1038/s41467-025-62956-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ren X, Xu J, Zhang Y, et al. Bacterial Alterations in Post-Cholecystectomy Patients Are Associated With Colorectal Cancer. Front Oncol 2020;10:1418. 10.3389/fonc.2020.01418 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Xu F, Chen R, Zhang C, et al. Cholecystectomy Significantly Alters Gut Microbiota Homeostasis and Metabolic Profiles: A Cross-Sectional Study. Nutrients 2023;15:4399. 10.3390/nu15204399 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ma Y, Qu R, Zhang Y, et al. Progress in the Study of Colorectal Cancer Caused by Altered Gut Microbiota After Cholecystectomy. Front Endocrinol (Lausanne) 2022;13:815999. 10.3389/fendo.2022.815999 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lorestani P, Faghihkhorasani F, Saffar H, et al. The Impact of Cholecystectomy on Colorectal Cancer Risk: A Comprehensive Review on Risk Factors and the Association. Curr Gastroenterol Rep 2025;27:59. 10.1007/s11894-025-01008-z [DOI] [PubMed] [Google Scholar]
- 12.Schernhammer ES, Leitzmann MF, Michaud DS, et al. Cholecystectomy and the risk for developing colorectal cancer and distal colorectal adenomas. Br J Cancer 2003;88:79-83. 10.1038/sj.bjc.6600661 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kim SB, Kim KO, Kim TN. Prevalence and Risk Factors of Gastric and Colorectal Cancer after Cholecystectomy. J Korean Med Sci 2020;35:e354. 10.3346/jkms.2020.35.e354 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Zhang Y, Liu H, Li L, et al. Cholecystectomy can increase the risk of colorectal cancer: A meta-analysis of 10 cohort studies. PLoS One 2017;12:e0181852. 10.1371/journal.pone.0181852 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Chen YK, Yeh JH, Lin CL, et al. Cancer risk in patients with cholelithiasis and after cholecystectomy: a nationwide cohort study. J Gastroenterol 2014;49:923-31. 10.1007/s00535-013-0846-6 [DOI] [PubMed] [Google Scholar]
- 16.Chen L, Fan Z, Sun X, et al. Associations of cholecystectomy with the risk of colorectal cancer: a Mendelian randomization study. Chin Med J (Engl) 2023;136:840-7. 10.1097/CM9.0000000000002612 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Shang J, Reece JC, Buchanan DD, et al. Cholecystectomy and the risk of colorectal cancer by tumor mismatch repair deficiency status. Int J Colorectal Dis 2016;31:1451-7. 10.1007/s00384-016-2615-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Altieri A, Pelucchi C, Talamini R, et al. Cholecystectomy and the risk of colorectal cancer in Italy. Br J Cancer 2004;90:1753-5. 10.1038/sj.bjc.6601721 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Lee J, Choe S, Park JW, et al. The Risk of Colorectal Cancer After Cholecystectomy or Appendectomy: A Population-based Cohort Study in Korea. J Prev Med Public Health 2018;51:281-8. 10.3961/jpmph.18.105 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Yu L, Liu W, Yan Y, et al. No association between cholecystectomy and risk of colorectal cancer: a meta-analysis of cohort studies. Int J Colorectal Dis 2023;38:179. 10.1007/s00384-023-04463-0 [DOI] [PubMed] [Google Scholar]
- 21.Bai X, Wei H, Liu W, et al. Cigarette smoke promotes colorectal cancer through modulation of gut microbiota and related metabolites. Gut 2022;71:2439-50. 10.1136/gutjnl-2021-325021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Jin EH, Han K, Shin CM, et al. Sex and Tumor-Site Differences in the Association of Alcohol Intake With the Risk of Early-Onset Colorectal Cancer. J Clin Oncol 2023;41:3816-25. 10.1200/JCO.22.01895 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kerr J, Anderson C, Lippman SM. Physical activity, sedentary behaviour, diet, and cancer: an update and emerging new evidence. Lancet Oncol 2017;18:e457-71. 10.1016/S1470-2045(17)30411-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Doubeni CA, Major JM, Laiyemo AO, et al. Contribution of behavioral risk factors and obesity to socioeconomic differences in colorectal cancer incidence. J Natl Cancer Inst 2012;104:1353-62. 10.1093/jnci/djs346 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Carethers JM, Doubeni CA. Causes of Socioeconomic Disparities in Colorectal Cancer and Intervention Framework and Strategies. Gastroenterology 2020;158:354-67. 10.1053/j.gastro.2019.10.029 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Rodriguez PM, Cruz NI, Gonzalez CI, et al. The effect of a high fat diet on the incidence of colonic cancer after cholecystectomy in mice. Cancer 1988;62:727-9. 10.1002/1097-0142(19880815)62:4<727::aid-cncr2820620414>3.0.co;2-f [DOI] [PubMed] [Google Scholar]
- 27.Xu F, Yu Z, Liu Y, et al. A High-Fat, High-Cholesterol Diet Promotes Intestinal Inflammation by Exacerbating Gut Microbiome Dysbiosis and Bile Acid Disorders in Cholecystectomy. Nutrients 2023;15:3829. 10.3390/nu15173829 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Yun J, Mullarky E, Lu C, et al. Vitamin C selectively kills KRAS and BRAF mutant colorectal cancer cells by targeting GAPDH. Science 2015;350:1391-6. 10.1126/science.aaa5004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kabbany MN, Conjeevaram Selvakumar PK, Watt K, et al. Prevalence of Nonalcoholic Steatohepatitis-Associated Cirrhosis in the United States: An Analysis of National Health and Nutrition Examination Survey Data. Am J Gastroenterol 2017;112:581-7. 10.1038/ajg.2017.5 [DOI] [PubMed] [Google Scholar]
- 30.Sekula P, Del Greco M F, Pattaro C, et al. Mendelian Randomization as an Approach to Assess Causality Using Observational Data. J Am Soc Nephrol 2016;27:3253-65. 10.1681/ASN.2016010098 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Boehm FJ, Zhou X. Statistical methods for Mendelian randomization in genome-wide association studies: A review. Comput Struct Biotechnol J 2022;20:2338-51. 10.1016/j.csbj.2022.05.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Skrivankova VW, Richmond RC, Woolf BAR, et al. Strengthening the Reporting of Observational Studies in Epidemiology Using Mendelian Randomization: The STROBE-MR Statement. JAMA 2021;326:1614-21. 10.1001/jama.2021.18236 [DOI] [PubMed] [Google Scholar]
- 33.Zhang YM, Zhou XJ, Shi SF, et al. Homocysteine and IgA nephropathy: observational and Mendelian randomization analyses. Chin Med J (Engl) 2020;133:277-84. 10.1097/CM9.0000000000000613 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Wang M, Jian Z, Yuan C, et al. Coffee Consumption and Prostate Cancer Risk: Results from National Health and Nutrition Examination Survey 1999-2010 and Mendelian Randomization Analyses. Nutrients 2021;13:2317. 10.3390/nu13072317 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Gao W, Jin L, Li D, et al. The association between the body roundness index and the risk of colorectal cancer: a cross-sectional study. Lipids Health Dis 2023;22:53. 10.1186/s12944-023-01814-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Yang F, Liang H, Rosenthal RJ, et al. The significant interaction between age and diabetes mellitus for colorectal cancer: Evidence from NHANES data 1999-2016. Prim Care Diabetes 2021;15:518-21. 10.1016/j.pcd.2021.02.006 [DOI] [PubMed] [Google Scholar]
- 37.Dos Santos M, Ferrari G, Lee DH, et al. Association of the "Weekend Warrior" and Other Leisure-time Physical Activity Patterns With All-Cause and Cause-Specific Mortality: A Nationwide Cohort Study. JAMA Intern Med 2022;182:840-8. 10.1001/jamainternmed.2022.2488 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Li Y, Di X, Liu M, et al. Association between daily sitting time and kidney stones based on the National Health and Nutrition Examination Survey (NHANES) 2007-2016: a cross-sectional study. Int J Surg 2024;110:4624-32. 10.1097/JS9.0000000000001560 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Hirshkowitz M, Whiton K, Albert SM, et al. National Sleep Foundation's updated sleep duration recommendations: final report. Sleep Health 2015;1:233-43. 10.1016/j.sleh.2015.10.004 [DOI] [PubMed] [Google Scholar]
- 40.Rattan P, Penrice DD, Ahn JC, et al. Inverse Association of Telomere Length With Liver Disease and Mortality in the US Population. Hepatol Commun 2022;6:399-410. 10.1002/hep4.1803 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Iwagami M, Shinozaki T. Introduction to Matching in Case-Control and Cohort Studies. Ann Clin Epidemiol 2022;4:33-40. 10.37737/ace.22005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Cenzer I, Boscardin WJ, Berger K. Performance of matching methods in studies of rare diseases: a simulation study. Intractable Rare Dis Res 2020;9:79-88. 10.5582/irdr.2020.01016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Kibuchi E, Sturgis P, Durrant GB, et al. The Efficacy of Propensity Score Matching for Separating Selection and Measurement Effects Across Different Survey Modes. J Surv Stat Methodol 2024;12:764-89. [Google Scholar]
- 44.Stuart EA. Matching methods for causal inference: A review and a look forward. Stat Sci 2010;25:1-21. 10.1214/09-STS313 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Liu ZP, Su XX, Chen LF, et al. Long- and short-term outcomes for resectable gallbladder carcinoma patients treated with curative-intent laparoscopic versus open resection: a multicenter propensity score-matched comparative study. Hepatobiliary Surg Nutr 2024;13:788-800. 10.21037/hbsn-23-518 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Pompili M, Saviano A, de Matthaeis N, et al. Long-term effectiveness of resection and radiofrequency ablation for single hepatocellular carcinoma ≤3 cm. Results of a multicenter Italian survey. J Hepatol 2013;59:89-97. 10.1016/j.jhep.2013.03.009 [DOI] [PubMed] [Google Scholar]
- 47.Nguyen TL, Collins GS, Spence J, et al. Double-adjustment in propensity score matching analysis: choosing a threshold for considering residual imbalance. BMC Med Res Methodol 2017;17:78. 10.1186/s12874-017-0338-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Ábrahám S, Németh T, Benkő R, et al. Evaluating the distribution of the locations of colorectal cancer after appendectomy and cholecystectomy. World J Surg Oncol 2020;18:94. 10.1186/s12957-020-01861-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Zhu Y, Shen L, Huo Y, et al. Gallstones, cholecystectomy, and cancer risk: an observational and Mendelian randomization study. Front Med 2025;19:79-89. 10.1007/s11684-024-1111-5 [DOI] [PubMed] [Google Scholar]
- 50.Siegel RL, Wagle NS, Cercek A, et al. Colorectal cancer statistics, 2023. CA Cancer J Clin 2023;73:233-54. 10.3322/caac.21772 [DOI] [PubMed] [Google Scholar]
- 51.Shao T, Yang YX. Cholecystectomy and the risk of colorectal cancer. Am J Gastroenterol 2005;100:1813-20. 10.1111/j.1572-0241.2005.41610.x [DOI] [PubMed] [Google Scholar]
- 52.Ekbom A, Yuen J, Adami HO, et al. Cholecystectomy and colorectal cancer. Gastroenterology 1993;105:142-7. 10.1016/0016-5085(93)90019-9 [DOI] [PubMed] [Google Scholar]