Abstract
Purpose
Hemorrhoids (HEM) are the most common perianal disease, but current observational studies have yielded inconsistent results in investigating the risk factors. Our further exploration of the risk factors will help prevent the disease.
Patients and Methods
We conducted a two-sample bidirectional Mendelian randomization (MR) analysis using publicly available genome-wide association studies (GWAS) statistics from multiple consortia. The inverse-variance weighted (IVW) method was used for the primary analysis. We applied four complementary methods, including weighted median, weighted mode, MR-Egger regression, and Cochrane’s Q value, to detect and correct the effects of horizontal pleiotropy.
Results
Genetically determined constipation (OR = 0.97, 95% CI: 0.91–1.03, P = 0.28) and diarrhea (OR = 1.00, 95% CI: 0.99–1.01, P = 0.90) did not have a causal effect on HEM but stool frequency (OR = 1.28, 95% CI: 1.05–1.55, P = 0.01), waist-to-hip ratio adjusted for BMI (OR = 1.11, 95% CI: 1.06–1.64, P = 1.59×10-5), and order Burkholderiales (OR = 1.09, 95% CI = 1.04–1.14, p = 1.63×10-4) had a causal effect on. Furthermore, we found a significant causal effect of constipation on HEM in the reverse MR analysis (OR = 1.21, 95% CI: 1.13–1.28, P = 3.72×10-9). The results of MR-Egger regression, Weighted Median, and Weighted Mode methods were consistent with those of the IVW method. Horizontal pleiotropy was unlikely to distort the causal estimates, as indicated by the sensitivity analysis.
Conclusion
Our MR analysis reveals a causal association between stool frequency and waist-to-hip ratio with HEM, despite variations in results reported by observational studies. Unexpectedly, we found a relationship between the order Burkholderiales in the gut flora and HEM, although the mechanism is unclear.
Acknowledgments and Data Sharing Statement
The authors thank all investigators and participants for sharing genetic association estimates. The datasets generated and analyzed during the current study are available in the internet repository. The website addresses are as follows: IEU (https://gwas.mrcieu.ac.uk/);Finn Gen (https://www.finngen.fi/en);GWAS Catalog (https://www.ebi.ac.uk/gwas/);UK biobank (http://www.nealelab.is/uk-biobank);MiBioGen repository (https://mibiogen.gcc.rug.nl/);Original GWAS articles of Bonfiglio et al and Zheng et al reported.
Disclosure
The authors report no conflicts of interest in this work.