Skip to main content
Download PDF

The association between prescription drugs and colorectal cancer prognosis: a nationwide cohort study using a medication-wide association study

BMC Cancer volume 23, Article number: 643 (2023) Cite this article

Abstract

Background

With the availability of health insurance claim data, pharmacovigilance for various drugs has been suggested; however, it is necessary to establish an appropriate analysis method. To detect unintended drug effects and to generate new hypotheses, we conducted a hypothesis-free study to systematically examine the relationship between all prescription nonanticancer drugs and the mortality of colorectal cancer patients.

Methods

We used the Korean National Health Insurance Service-National Sample Cohort database. A total of 2,618 colorectal cancer patients diagnosed between 2004 and 2015 were divided into drug discovery and drug validation sets (1:1) through random sampling. Drugs were classified using the Anatomical Therapeutic Chemical (ATC) classification system: 76 drugs classified as ATC level 2 and 332 drugs classified as ATC level 4 were included in the analysis. We used a Cox proportional hazard model adjusted for sex, age, colorectal cancer treatment, and comorbidities. The relationship between all prescription nonanticancer drugs and the mortality of colorectal cancer patients was analyzed, controlling for multiple comparisons with the false discovery rate.

Results

We found that one ATC level-2 drug (drugs that act on the nervous system, including parasympathomimetics, addictive disorder drugs, and antivertigo drugs) showed a protective effect related to colorectal cancer prognosis. At the ATC level 4 classification, 4 drugs were significant: two had a protective effect (anticholinesterases and opioid anesthetics), and the other two had a detrimental effect (magnesium compounds and Pregnen [4] derivatives).

Conclusions

In this hypothesis-free study, we identified four drugs linked to colorectal cancer prognosis. The MWAS method can be useful in real-world data analysis.

Peer Review reports

Introduction

Colorectal cancer is the third most common cancer and the second leading cause of cancer death worldwide [1]. In addition, in high-HDI (human development index) countries, the survival rate of colorectal cancer has steadily increased due to early detection and improved treatment outcomes. As a result, the prevalence of colorectal cancer is steadily increasing [2]. Lifestyle factors such as drinking, smoking, obesity, and low physical activity are risk factors that act in common not only for colorectal cancer but also for noncancer chronic diseases. Chronic diseases such as hypertension, diabetes, and hyperlipidemia often accompany colorectal cancer [3, 4]. Because it is limited to detecting rare adverse effects through clinical trials, it is necessary to reveal unintended effects of drugs through pharmacovigilance after drug marketing [5]. Pharmacovigilance after drug marketing is mainly performed by monitoring reports of adverse events, but rare events with relatively long induction periods, such as cancer occurrence or cancer-caused death, are difficult to monitor [6]. In cases such as these, pharmacoepidemiological studies are helpful, as are representative examples of the relationship between noncancer chronic disease medications, including metformin (a treatment for type 2 diabetes), statins (a treatment for hyperlipidemia), and aspirin (which is known to be effective in preventing cardiovascular disease), and the survival prognosis of colorectal cancer patients [7,8,9,10,11,12]. However, these studies are hypothesis-driven, and if selective reporting occurs, biased results may be obtained [13, 14]. Recently, as real-world data such as health insurance claim data become available, hypothesis non-specific studies using the concept of hypothesis-free data-driven approach are being conducted. In the field of pharmacoepidemiology, which systematically analyzes the relationship between various drugs and outcomes, this type of study, modeled after the genome-wide association study (GWAS), is known as a medication-wide association study (MWAS) or a drug-wide association study (DWAS) [15,16,17,18,19,20]. Hypothesis-free studies prevent selective reporting of results by correcting for multiple comparisons in the analysis and reporting it transparently [13, 14, 20]. We analyzed the relationship between prescription drugs and mortality in patients diagnosed with colorectal cancer using the Korean National Health Insurance Service-National Sample Cohort database (KNHIS-NSC).

Materials and methods

Database

This study was conducted using data from individuals enrolled in the KNHIS-NSC from 2002 to 2015 [21]. The KNHIS-NSC database consists of claim data regarding the use of medical institutions such as hospitals and pharmacies by Korean National Health Insurance subscribers. All medical institutions in Korea are obliged to participate in the Korean National Health Insurance system. Therefore, claim data include data from all hospitals and pharmacies in Korea. These data, consisting of systematic stratified random sampling of 1,025,340 people (approximately 2% of the Korean population), were released for research purposes. The database includes demographic information such as sex, age, date of death, cause of death, income decile, and treatment date; diagnosis information based on the Korean Standard Classification of Diseases (KCD) reflecting the International Standard Classification of Diseases (ICD); records of inpatient and outpatient usage and related prescriptions; and health-examination results [21].

Study sample

Colorectal cancer patients were defined as a case with at least one of the KCD codes C18, C19, and C20 who underwent diagnostic tests, such as colonoscopy, or received colorectal cancer treatment, such as surgery, chemotherapy, or radiation therapy [9, 22]. The date of the first treatment was set as time zero (i.e., study entry). To include only newly diagnosed colorectal cancer patients, a wash-out period of 2 years was set; therefore, patients diagnosed with colorectal cancer in 2002 and 2003 were excluded.

Cross validation

Cross validation was performed to ensure the reproducibility of the results. The selected participants were randomly divided into drug discovery and drug validation sets in a 1:1 ratio. Then, sequential analysis was performed using the two-stage method [23]. The selection process for the study population is depicted in Fig. 1.

Fig. 1
figure 1

Study flow diagram

Full size image

Exposure

After drugs classified as anticancer drugs according to the Korean Ministry of Food and Drug Safety were excluded, all drug prescription data were converted to their Anatomical Therapeutic Chemical (ATC) Classification System codes. The ATC system is a drug classification system administered by the World Health Organization that classifies drugs into five levels. The ATC codes were analyzed at ATC levels 2 (therapeutic group) and 4 (chemical subgroup) [24]. In pharmacoepidemiological studies, if prevalent users are included, selection bias may occur. This is because prevalent users survive while taking drugs and may have improved health-related behaviors [25,26,27,28]. To prevent such selection bias, nonusers can be compared with incident users who had no history of the target drug use before study entry and started taking the target drugs after study entry [27, 28]. However, this method can also lead to confounding by indication (i.e., treatment is generally given to a patient with a specific condition) [27]. To solve this problem, we defined drug users as patients who did not take certain drugs for 1–1.5 years before study entry and then took drugs for the past 1 year before study entry (i.e., past 1-year incident users) (Fig. 2).

Fig. 2
figure 2

Schematic diagram of the definitions of drug users and nonusers

Full size image

Outcomes

We used the outcomes of all-cause mortality and colorectal cancer-specific mortality. Follow-up started at the date of the first treatment for colorectal cancer and terminated at whichever came first, the date of death or the date of the end of follow-up of the KNHIS-NSC (December 31, 2015).

Covariates

We adjusted for the following covariates: sex, age, and Charlson Comorbidity Index (CCI). The treatment received by the patient, such as colonoscopy, surgery, chemotherapy, and radiation therapy, was included as a proxy variable for colorectal cancer staging as used in previous studies. We categorized the study population based on cancer treatment, considering that patients with early-stage cancer typically undergo surgery alone, while those with advanced-stage cancer often receive palliative therapy such as chemotherapy or radiotherapy without undergoing surgery [9, 22]. To adjust for patients who did not receive any treatment after colorectal cancer diagnosis (i.e., terminal stage), undergoing colonoscopy only was included as a covariate for treatment.

Statistical analysis

We performed a survival analysis using the Cox proportional hazards model. Because hypothesis-free studies analyze the relationship between many drugs and outcomes at once, type I errors (false positives) increase due to multiple comparisons. In this study, we controlled for type I errors using the false discovery rate (FDR). The FDR represents the proportion of falsely rejected hypotheses in multiple comparison [29]. Significant signals were first detected in the drug discovery set with an FDR of 5%, and then the result was subsequently verified with a p value threshold of 0.05 in the drug validation set (Fig. 3). The results of each drug’s verification were displayed using a volcano plot, which is a type of scatter plot. The x-axis was represented by the base-2 logarithm of the HR, while the y-axis was represented by the negative logarithm of the FDR-adjusted p value.

Fig. 3
figure 3

Summary of the analytic method

Full size image

Results

Study characteristics

After excluding 213 patients who received colorectal cancer treatment in 2002 and 2003, which was set as the wash-out period, a total of 2,618 colorectal cancer patients met the eligibility criteria and were included in the study (Fig. 1). A total of 1,309 patients were included in the drug discovery set and the drug validation set. The characteristics of the study subjects are presented in Table 1. Overall, the distributions of the general characteristics of the two datasets were similar except for the higher proportion of males in the discovery set.

Table 1 General characteristics of the colorectal cancer patients included in the discovery and validation sets
Full size table

Drugs classified as ATC level 2

A total of 76 drugs classified as ATC level 2 were included in the analysis. A volcano plot of detected signals for all-cause mortality is presented in Fig. 4. Only one signal (N07: Other nervous system drugs) was found in the drug discovery set based on the preset 5% FDR. This drug had a protective effect that was reproduced in the drug validation set. Figure 5 displays the signal search results for colorectal cancer-specific mortality. Other nervous system drugs (N07) repeatedly exhibited significant signals. One signal (J01: Antibacterials for systemic use, protective effect) was additionally discovered in the drug discovery set but not reproduced in the validation set. The full results are presented in Supplementary Table 1.

Fig. 4
figure 4

Volcano plot demonstrating the hazard ratios (HRs) and FDR-adjusted p values for the association between drugs classified as ATC level 2 and all-cause mortality

Full size image
Fig. 5
figure 5

Volcano plot demonstrating the hazard ratios (HRs) and FDR-adjusted p values for the association between drugs classified as ATC level 2 and colorectal cancer-specific mortality

Full size image

Drugs classified as ATC level 4

A total of 332 drugs classified as ATC level 4 were included in the analysis. As shown in Fig. 6, for all-cause mortality, ten drugs exceeded the threshold in the drug discovery set. Four of these drugs also produced significant results in the drug validation set. Two drugs exerted a protective effect (N07AA: anticholinesterases, N01AH: opioid anesthetics), and two drugs exerted a detrimental effect (magnesium compounds (A02AA) and Pregnen [4] derivatives (G03DA)). For colorectal cancer-specific mortality, a total of 9 drugs were significant in the drug discovery set, and 4 of these drugs were also significant in the validation set. The drugs that were significant in all-cause mortality were equally significant in colorectal cancer-specific mortality (Fig. 7). The full results are presented in Supplementary Table 2.

Fig. 6
figure 6

Volcano plot demonstrating the hazard ratios (HRs) and FDR-adjusted p values for the association between drugs classified as ATC level 4 and all-cause mortality

Full size image
Fig. 7
figure 7

Volcano plot demonstrating the hazard ratios (HRs) and FDR-adjusted p values for the association between drugs classified as ATC level 4 and colorectal cancer-specific mortality

Full size image

Results for drugs previously reported to be associated with colorectal cancer prognosis

Next, we assessed whether metformin, statins, and aspirin, which have been reported to be related to the survival of colorectal cancer, showed a significant relationship with mortality in our study [7,8,9,10,11,12]. None of these drugs showed significant results even without FDR correction. Moreover, for the hazard ratio, the direction of the effect of metformin and statins was to increase mortality (metformin (A10BA): HR for all-cause mortality, 1.26; 95% CI, 0.71–2.24; HR for colorectal cancer-specific mortality, 1.48; 95% CI, 0.79–2.79; statin (C10AA): HR for all-cause mortality, 1.3; 95% CI, 0.85–1.98; HR for colorectal cancer-specific mortality, 1.05; 95% CI, 0.6–1.83). However, aspirin (B01AC) tended to decrease mortality (HR for all-cause mortality, 0.9; 95% CI, 0.57–1.40; HR for colorectal cancer-specific mortality, 0.84; 95% CI, 0.49–1.45) (Supplementary Table 2).

Discussion

This study aimed to discover a new hypothesis related to the survival of colorectal cancer patients using a hypothesis-free method. In the analysis, a total of 4 drugs were identified to be related to the prognosis of colorectal cancer patients.

Anticholinesterases (N07AA) and opioid anesthetics (N01AH) exerted protective effects against both all-cause mortality and colorectal cancer-specific mortality. To the best of our knowledge, no epidemiological studies have reported a relationship between these two drugs and cancer. Although no in vitro or in vivo studies have found that anticholinesterases are directly related to colorectal cancer, several studies have shown that acetylcholine acts on colorectal cancer muscarinic receptor (MR) signaling to promote the proliferation of colorectal cancer cells [30,31,32]. Opioid anesthetics may be related to surgery under general anesthesia. Therefore, rather than opioid anesthetics themselves affecting the prognosis of colorectal cancer patients, it is possible that health-related behaviors after general anesthesia for surgery acted as a confounder. Two other drugs (Pregnen [4] derivatives and magnesium compounds) had a detrimental effect on the prognosis of colorectal cancer patients. Epidemiological studies have shown that progesterone from Pregnen [4] derivatives increases mortality in colorectal cancer patients. In the Women’s Health Initiative (WHI) trial, the incidence of colorectal cancer was lower in the group that took estrogen plus progestin than in the group that took the placebo, but the stage of cancer was more advanced at the time of diagnosis [33]. A follow-up study also showed that the hormone group exhibited more advanced stages of colorectal cancer and, although it was not significantly different, that the colorectal cancer mortality rate was also higher [34]. Magnesium compounds (A02AA) were a significant signal in colorectal cancer-specific mortality. No studies have found that magnesium compounds are related to the prognosis of colorectal cancer patients. Indeed, many epidemiological studies have reported that magnesium compounds lower the incidence of colorectal cancer [35,36,37]. Magnesium compounds may have different effects before and after the onset of colorectal cancer. Alternatively, although magnesium protects against colorectal cancer, magnesium intake in patients diagnosed with colorectal cancer may also have residual confounders that offset its protective effect. To interpret this signal, further studies on the effects of magnesium in patients with colorectal cancer are needed.

While metformin, statins, and aspirin were previously reported to be associated with the prognosis of colorectal cancer patients [7,8,9,10,11,12], this result was not reproduced in our study. These three drugs did not significantly influence colorectal cancer patient mortality even when the p values were not FDR corrected. Moreover, only aspirin had an effect in the predicted direction (to decrease mortality). These differences are thought to be due to differences in study design. The hypothesis-free method has the advantage of including all covariates in the model simultaneously and deriving results for various exposures. However, it also has several disadvantages, such as the lack of hypothesis-specific model selection and subgroup analysis. For example, studies that analyzed the prognosis of patients with metformin and colorectal cancer were conducted by limiting the subjects to diabetic patients or including patients’ adherence to metformin in the model [9, 11]. Additionally, the results can be influenced by the drug user definition (i.e., prevalent vs. incident users). In pharmacoepidemiological studies, it is recommended that only subjects whose exposure started at the time of follow-up be included in the study to prevent prevalent user bias [27, 28]. To avoid prevalent user bias and confounding by indication, the subjects were limited to past 1-year incident users in our study. In a study examining the relationship between statin use and cancer mortality, statin exposure and cancer-specific mortality were inversely associated in prevalent users, but this association disappeared in incident users; moreover, the direction of the effect was changed such that it increased cancer-specific mortality [38]. In addition, since the claim data used in this study reflects the real world, it should be considered that the subjects might have used the medications without control in various clinical environments.

The strength of this study is that it attempted to identify drugs related to the prognosis of cancer patients with a hypothesis-free method. Previous MWASs or DWASs have focused on drugs related to the incidence of cancer [18,19,20]. In particular, it was proposed to define drug users as past year incident users, thereby minimizing prevalent user bias and confounding by indications. In addition, the past year serves as a landmark time that provides effective control over the immortal-time bias. Additionally, a new hypothesis was presented as in the previously hypothesis-free studies; specifically, in this study, we found a total of 4 signals related to colorectal cancer prognosis. Presenting the signals in this manner can prevent selective reporting, as all results are systematically derived under the control of multiple comparisons. Further studies are needed to determine whether these associations are causal. The limitations of this study are as follows. First, since this study is an observational study, it cannot be free from the issue of validity due to comparability [39]. In the hypothesis-free approach, since many hypotheses are tested at once, if there is a problem with the validity of the study, it can lead to more serious issues. Unlike clinical trials that randomly assign individuals to take or not take the drug, basic differences between drug users and nonusers in hypothesis-free studies may generate comparability problems [40]. This problem is more serious when the indication of the drug is related to the patient’s prognosis. To solve this problem, we restricted the definition of drug use to 1 year before receiving colorectal cancer treatment and adjusted for the type of treatment received by colorectal cancer patients as well as the CCI. In a study that used the same data as this study, the type of treatment received by colorectal cancer patients was reported to be a useful proxy variable for cancer staging [9, 22]. Second, although the KNHIS-NSC data are representative of the Korean population, they were not collected for research purposes but are real-world data from health insurance claims. Therefore, information bias may occur. However, in the case of cancer, the accuracy of the diagnostic code is relatively high, and the risk of misclassification was reduced through the operational definition, including the type of colorectal cancer treatment [41].

Conclusion

In this study, we found four signals related to prognosis in colorectal cancer patients through a hypothesis-free method. In addition, we observed conflicting results for statin, metformin, and aspirin compared to those of previous studies. The MWAS approach can be useful for pharmacovigilance in rare events with a long induction period and for proposing a new hypothesis.

Data Availability

The NHIS-NSC data used to support the findings of this study are only available to authorized personnel and can be accessed from the website of the NHIS (https://nhiss.nhis.or.kr) after completing the application process and receiving approval (http://nhiss.nhis.or.kr/bd/ab/bdaba021eng.do).

Abbreviations

ATC:

Anatomical therapeutic chemical classification system

CCI:

Charlson comorbidity index

DWAS:

Drug-wide association study

FDR:

False discovery rate

GWAS:

Genome-wide association study

HDI:

Human development index

HR:

Hazard ratio

ICD:

International standard classification of diseases

KCD:

Korean standard classification of diseases

KNHIS-NSC:

Korean national health insurance service-national sample cohort

MR:

Muscarinic receptor

MWAS:

Medication-wide association study

WHI:

Women’s health initiative

References

  1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and Mortality Worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209–49. https://doi.org/10.3322/caac.21660

    Article  PubMed  Google Scholar 

  2. Rawla P, Sunkara T, Barsouk A. Epidemiology of colorectal cancer: incidence, mortality, survival, and risk factors. Prz Gastroenterol. 2019;14(2):89–103. https://doi.org/10.5114/pg.2018.81072

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Sturmer T, Buring JE, Lee IM, Gaziano JM, Glynn RJ. Metabolic abnormalities and risk for colorectal cancer in the physicians’ health study. Cancer Epidemiol Biomarkers Prev. 2006;15(12):2391–7. https://doi.org/10.1158/1055-9965.EPI-06-0391

    Article  PubMed  PubMed Central  Google Scholar 

  4. Tran TT, Gunathilake M, Lee J, Kim J. Association between metabolic syndrome and its components and incident colorectal cancer in a prospective cohort study. Cancer. 2022;128(6):1230–41. https://doi.org/10.1002/cncr.34027

    Article  CAS  PubMed  Google Scholar 

  5. Dal Pan GJ, Lindquist M, Gelperin K. Postmarketing spontaneous Pharmacovigilance Reporting Systems. In: Strom BL, Kimmel SE, Hennessy S, editors. Textbook of Pharmacoepidemiology. Hoboken, NJ: John Wiley & Sons; 2021. pp. 113–35. https://doi.org/10.1002/9781119701101.ch7

    Chapter  Google Scholar 

  6. Laporte JR. Fifty years of pharmacovigilance - Medicines safety and public health. Pharmacoepidemiol Drug Saf. 2016;25(6):725–32. https://doi.org/10.1002/pds.3967

    Article  PubMed  Google Scholar 

  7. Rothwell PM, Wilson M, Elwin CE, Norrving B, Algra A, Warlow CP, et al. Long-term effect of aspirin on colorectal cancer incidence and mortality: 20-year follow-up of five randomised trials. Lancet (London England). 2010;376(9754):1741–50. https://doi.org/10.1016/S0140-6736(10)61543-7

    Article  CAS  PubMed  Google Scholar 

  8. Voorneveld PW, Reimers MS, Bastiaannet E, Jacobs RJ, van Eijk R, Zanders MMJ, et al. Statin Use after diagnosis of Colon cancer and patient survival. Gastroenterology. 2017;153(2):470–9e4. https://doi.org/10.1053/j.gastro.2017.05.011

    Article  CAS  PubMed  Google Scholar 

  9. Choe S, Lee J, Park JW, Jeong SY, Cho YM, Park BJ, et al. Prognosis of patients with Colorectal Cancer with Diabetes according to Medication Adherence: a Population-Based Cohort Study. Cancer Epidemiol Biomarkers Prev. 2020;29(6):1120–7. https://doi.org/10.1158/1055-9965.EPI-19-1455

    Article  PubMed  Google Scholar 

  10. Li Y, He X, Ding Y, Chen H, Sun L. Statin uses and mortality in colorectal cancer patients: an updated systematic review and meta-analysis. Cancer Med. 2019;8(6):3305–13. https://doi.org/10.1002/cam4.2151

    Article  PubMed  PubMed Central  Google Scholar 

  11. Cao X, Wu Y, Wang J, Liu K, Wang X. The Effect of Metformin on Mortality among Diabetic Cancer Patients: a systematic review and Meta-analysis. JNCI cancer spectrum. 2017;1(1):pkx007. https://doi.org/10.1093/jncics/pkx007

    Article  PubMed  PubMed Central  Google Scholar 

  12. Xiao S, Xie W, Fan Y, Zhou L. Timing of aspirin use among patients with Colorectal Cancer in Relation to Mortality: a systematic review and Meta-analysis. JNCI cancer spectrum. 2021;5(5):pkab067. https://doi.org/10.1093/jncics/pkab067

    Article  PubMed  PubMed Central  Google Scholar 

  13. Ioannidis JP. Why most published research findings are false. PLoS Med. 2005;2(8):e124. https://doi.org/10.1371/journal.pmed.0020124

    Article  PubMed  PubMed Central  Google Scholar 

  14. Ioannidis JP. Why most discovered true associations are inflated. Epidemiol (Cambridge Mass). 2008;19(5):640–8. https://doi.org/10.1097/EDE.0b013e31818131e7

    Article  Google Scholar 

  15. Harpaz R, DuMouchel W, Shah NH, Madigan D, Ryan P, Friedman C. Novel data-mining methodologies for adverse drug event discovery and analysis. Clin Pharmacol Ther. 2012;91(6):1010–21. https://doi.org/10.1038/clpt.2012.50

    Article  CAS  PubMed  Google Scholar 

  16. Ryan PB, Madigan D, Stang PE, Schuemie MJ, Hripcsak G. Medication-wide association studies. CPT: pharmacometrics & systems pharmacology. 2013;2(9):e76. https://doi.org/10.1038/psp.2013.52

    Article  CAS  Google Scholar 

  17. Ryan PB, Madigan D, Stang PE, Overhage JM, Racoosin JA, Hartzema AG. Empirical assessment of methods for risk identification in healthcare data: results from the experiments of the Observational Medical Outcomes Partnership. Stat Med. 2012;31(30):4401–15. https://doi.org/10.1002/sim.5620

    Article  PubMed  Google Scholar 

  18. Pottegård A, Friis S, Christensen R, Habel LA, Gagne JJ, Hallas J. Identification of Associations between prescribed Medications and Cancer: a Nationwide Screening Study. EBioMedicine. 2016;7:73–9. https://doi.org/10.1016/j.ebiom.2016.03.018

    Article  PubMed  PubMed Central  Google Scholar 

  19. Støer NC, Botteri E, Thoresen GH, Karlstad Ø, Weiderpass E, Friis S, et al. Drug Use and Cancer Risk: a drug-wide Association Study (DWAS) in Norway. Cancer Epidemiol Biomarkers Prev. 2021;30(4):682–9. https://doi.org/10.1158/1055-9965.EPI-20-1028

    Article  PubMed  Google Scholar 

  20. Patel CJ, Ji J, Sundquist J, Ioannidis JP, Sundquist K. Systematic assessment of pharmaceutical prescriptions in association with cancer risk: a method to conduct a population-wide medication-wide longitudinal study. Sci Rep. 2016;6:31308. https://doi.org/10.1038/srep31308

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Lee J, Lee JS, Park SH, Shin SA, Kim K. Cohort Profile: the National Health Insurance Service-National Sample Cohort (NHIS-NSC), South Korea. Int J Epidemiol. 2017;46(2):e15. https://doi.org/10.1093/ije/dyv319

    Article  PubMed  Google Scholar 

  22. Lee J, Choe S, Park JW, Jeong SY, Shin A. The risk of Colorectal Cancer after Cholecystectomy or Appendectomy: a Population-based Cohort Study in Korea. J Prev Med Public Health. 2018;51(6):281–8. https://doi.org/10.3961/jpmph.18.105

    Article  PubMed  PubMed Central  Google Scholar 

  23. Kang G, Liu W, Cheng C, Wilson CL, Neale G, Yang JJ, et al. Evaluation of a two-step iterative resampling procedure for internal validation of genome-wide association studies. J Hum Genet. 2015;60(12):729–38. https://doi.org/10.1038/jhg.2015.110

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Liang H, Hu B, Chen L, Wang S. Aorigele. Recognizing novel chemicals/drugs for anatomical therapeutic chemical classes with a heat diffusion algorithm. Biochim Biophys Acta Mol Basis Dis. 2020;1866(11):165910. https://doi.org/10.1016/j.bbadis.2020.165910

    Article  CAS  PubMed  Google Scholar 

  25. Lund JL, Richardson DB, Stürmer T. The active comparator, new user study design in pharmacoepidemiology: historical foundations and contemporary application. Curr Epidemiol Rep. 2015;2(4):221–8. https://doi.org/10.1007/s40471-015-0053-5

    Article  PubMed  PubMed Central  Google Scholar 

  26. Danaei G, Tavakkoli M, Hernán MA. Bias in observational studies of prevalent users: lessons for comparative effectiveness research from a meta-analysis of statins. Am J Epidemiol. 2012;175(4):250–62. https://doi.org/10.1093/aje/kwr301

    Article  PubMed  PubMed Central  Google Scholar 

  27. Fu EL, van Diepen M, Xu Y, Trevisan M, Dekker FW, Zoccali C, et al. Pharmacoepidemiology for nephrologists (part 2): potential biases and how to overcome them. Clin kidney J. 2021;14(5):1317–26. https://doi.org/10.1093/ckj/sfaa242

    Article  PubMed  Google Scholar 

  28. Ray WA. Evaluating medication effects outside of clinical trials: new-user designs. Am J Epidemiol. 2003;158(9):915–20. https://doi.org/10.1093/aje/kwg231

    Article  PubMed  Google Scholar 

  29. Benjamini Y, Hochberg Y. Controlling the false Discovery rate: a practical and powerful Approach to multiple testing. J Royal Stat Soc Ser B (Methodological). 1995;57(1):289–300. https://doi.org/10.1111/j.2517-6161.1995.tb02031.x

    Article  Google Scholar 

  30. Xie G, Raufman JP. Muscarinic receptor signaling and colon cancer progression. J cancer metastasis Treat. 2016;2:195–200. https://doi.org/10.20517/2394-4722.2016.05

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Xie G, Cheng K, Shant J, Raufman JP. Acetylcholine-induced activation of M3 muscarinic receptors stimulates robust matrix metalloproteinase gene expression in human colon cancer cells. Am J Physiol Gastrointest liver Physiol. 2009;296(4):G755–63. https://doi.org/10.1152/ajpgi.90519.2008

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Cheng K, Samimi R, Xie G, Shant J, Drachenberg C, Wade M, et al. Acetylcholine release by human colon cancer cells mediates autocrine stimulation of cell proliferation. Am J Physiol Gastrointest liver Physiol. 2008;295(3):G591–7. https://doi.org/10.1152/ajpgi.00055.2008

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Chlebowski RT, Wactawski-Wende J, Ritenbaugh C, Hubbell FA, Ascensao J, Rodabough RJ, et al. Estrogen plus progestin and colorectal cancer in postmenopausal women. N Engl J Med. 2004;350(10):991–1004. https://doi.org/10.1056/NEJMoa032071

    Article  CAS  PubMed  Google Scholar 

  34. Simon MS, Chlebowski RT, Wactawski-Wende J, Johnson KC, Muskovitz A, Kato I, et al. Estrogen plus progestin and colorectal cancer incidence and mortality. J Clin oncology: official J Am Soc Clin Oncol. 2012;30(32):3983–90. https://doi.org/10.1200/JCO.2012.42.7732

    Article  Google Scholar 

  35. Lin J, Cook NR, Lee IM, Manson JE, Buring JE, Zhang SM. Total magnesium intake and colorectal cancer incidence in women. Cancer Epidemiol Biomarkers Prev. 2006;15(10):2006–9. https://doi.org/10.1158/1055-9965.EPI-06-0454

    Article  CAS  PubMed  Google Scholar 

  36. Larsson SC, Bergkvist L, Wolk A. Magnesium intake in relation to risk of colorectal cancer in women. JAMA. 2005;293(1):86–9. https://doi.org/10.1001/jama.293.1.86

    Article  CAS  PubMed  Google Scholar 

  37. Folsom AR, Hong CP. Magnesium intake and reduced risk of colon cancer in a prospective study of women. Am J Epidemiol. 2006;163(3):232–5. https://doi.org/10.1093/aje/kwj037

    Article  PubMed  Google Scholar 

  38. Emilsson L, García-Albéniz X, Logan RW, Caniglia EC, Kalager M, Hernán MA. Examining Bias in Studies of Statin Treatment and Survival in patients with Cancer. JAMA Oncol. 2018;4(1):63–70. https://doi.org/10.1001/jamaoncol.2017.2752

    Article  PubMed  Google Scholar 

  39. Szklo M, Nieto FJ. Identifying Noncausal Associations:Confouding Epidemiology beyond the basics. 4th ed. Burlington, Massachusetts: Jones & Bartlett Learning; 2019. 175–207.

    Google Scholar 

  40. Hernán MA, Methods of Public Health Research - Strengthening Causal Inference from Observational Data. N Engl J Med. 2021;385(15):1345–8. https://doi.org/10.1056/NEJMp2113319

    Article  PubMed  Google Scholar 

  41. Kim JA, Yoon S, Kim LY, Kim DS. Towards actualizing the value potential of Korea Health Insurance Review and Assessment (HIRA) data as a resource for Health Research: strengths, Limitations, applications, and strategies for optimal use of HIRA data. J Korean Med Sci. 2017;32(5):718–28. https://doi.org/10.3346/jkms.2017.32.5.718

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

Not Applicable.

Funding

This work was supported by the Bisa Research Grant of Keimyung University in 2022 and the National Research Foundation of Korea (2017R1A2B4009233). The funding organizations had no role in the design or conduct of this research.

Author information

Authors and Affiliations

  1. Department of Preventive Medicine, Keimyung University School of Medicine, 1095 Dalgubeol-daero, Dalseo- gu, Daegu, 42601, Korea

    Hyeong-Taek Woo

  2. Department of Preventive Medicine, Seoul National University College of Medicine, Seoul, Korea

    Hyeong-Taek Woo & Aesun Shin

  3. Department of Surgery, Seoul National University College of Medicine, Seoul, Korea

    Seung-Yong Jeong

  4. Cancer Research Institute, Seoul National University, Seoul, Korea

    Seung-Yong Jeong & Aesun Shin

Authors
  1. Hyeong-Taek Woo
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Seung-Yong Jeong
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Aesun Shin
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Conception and design: HT, SY, AS. Development of methodology: HT, AS. Acquisition of data: HT, AS. Analysis and interpretation of data: HT, SY, AS. Writing, review, and/or revision of the manuscript: HT, AS. Administrative, technical, or material support: AS. Study supervision: SY, AS.

Corresponding author

Correspondence to Hyeong-Taek Woo.

Ethics declarations

Ethics and consent to participate

This study was approved by the Institutional Review Board (IRB) of Seoul National University Hospital, Korea (IRB No. E-2007-005-1137). The Institutional Review Board (IRB) of Seoul National University Hospital, Korea waived the requirement for informed consent. All methods were performed in accordance with relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

Aesun Shin is an editorial board member (Associate Editor) of BMC Cancer. The other authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Woo, HT., Jeong, SY. & Shin, A. The association between prescription drugs and colorectal cancer prognosis: a nationwide cohort study using a medication-wide association study. BMC Cancer 23, 643 (2023). https://doi.org/10.1186/s12885-023-11105-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12885-023-11105-9

Keywords

BMC Cancer

ISSN: 1471-2407

Contact us