Page 63 - JCBP-3-3

Journal of Clinical and
Basic Psychosomatics Depression and chronic pancreatitis: MR study

between depression and CP as potential mediators. The on both phenotypes relative to the same allele. In addition,
genetic IVs for these potential mediating factors were a key assumption of MR design is that SNPs should affect
extracted from publicly available, large, and authoritative the outcomes solely through exposure. Therefore, we
GWAS databases. Detailed information is provided in examined whether each SNP associated with the exposure
Table S2. was also linked to common confounding factors (e.g.,
smoking, alcohol consumption, household income, and
2.2. MR BMI) through the Phenoscanner database (Table S3).
4,25
IVs were selected from relevant GWAS studies, and For each variant in the genetic instruments, the strength
the specific process is shown in Figure 2. The details are of the selected instrument was evaluated by calculating the
provided in the following sections. F-statistics using Equation I:

2.2.1. Two-sample bidirectional MR 2 × R −(N 2)
F = 26 (I)
We selected SNPs at the whole genome significance level 1 − R 2
(P < 5 × 10 ), while ensuring that the SNPs had relatively 2
−8
long physical distances and a low probability of linkage where R represents the proportion of variance in the
2
imbalance (R < 0.001 within 10,000 kb) to minimize the phenotype explained by each instrument, and N represents
correlation in GWAS P-values. When CP was used as the the sample size. F-statistics below 10 may indicate the
exposure factor, we applied a less stringent p-threshold presence of weak instrument bias.
−5
of 1 × 10 to obtain more SNPs for CP, as fewer than 10 To estimate of the total causal effect, we performed
independent SNPs met genome-wide significance and low two-sample MR analyses using inverse-variance weighted
linkage disequilibrium according to the above criteria. The (IVW), weighted median, MR-Egger regression, simple
minor allele frequency of the IVs for exposure factors was mode, and weighted mode methods. Among these, we
required to be >0.01. We excluded palindromic SNPs with applied the IVW method as the primary analysis. The
15
intermediate allele frequencies to ensure consistent effects weighted median method was used for further validation.

Figure 2. Criteria for selecting instrumental variable in two-sample, multivariable Mendelian randomization
Abbreviations: GWAS: Genome-wide association study; MAF: Minor allele frequency; MR: Mendelian randomization; MR-PRESSO: Mendelian
Randomization RESidual Sum and Outlier; SNPs: Single-nucleotide polymorphisms.

Volume 3 Issue 3 (2025) 57 doi: 10.36922/jcbp.5892