Figure 1
Flowchart of the study process to evaluate the association between waist circumference (WC) and coronary artery disease (CAD) using observational and Mendelian randomization (MR) analyses.
Table 1
Baseline characteristics and results after propensity score matching.
| LEVEL | CORONARY ARTERY DISEASE | NON-CORONARY ARTERY DISEASE | P | |
|---|---|---|---|---|
| N | 155 | 155 | ||
| BMXWAIST (mean (SD)) | 124.50 (14.52) | 99.36 (12.65) | <0.001 | |
| RIDAGEYR (mean (SD)) | 68.08 (14.59) | 68.12 (14.63) | 0.978 | |
| RIAGENDR (%) | Male | 118 (76.1) | 118 (76.1) | 1 |
| Female | 37 (23.9) | 37 (23.9) | ||
| RIDRETH1 (%) | Mexican American | 18 (11.6) | 19 (12.3) | 0.999 |
| Other Hispanic | 7 (4.5) | 8 (5.2) | ||
| Non-Hispanic Black | 10 (6.5) | 10 (6.5) | ||
| Non-Hispanic White | 114 (73.5) | 112 (72.3) | ||
| Other Race | 6 (3.9) | 6 (3.9) |
Table 2
Data description of waist circumference and coronary artery disease.
| PHENOTYPE | POPULATION | SAMPLE SIZE | DATE | SNPS | ACCESS ADDRESS | |
|---|---|---|---|---|---|---|
| Exposure | Waist circumference | European | 462,166 | 2018 | 9,851,867 | https://gwas.mrcieu.ac.uk/datasets/ukb-b-9405/ |
| Outcome | Coronary artery disease | 42,096 | 2015 | 8,597,751 | https://gwas.mrcieu.ac.uk/datasets/ebi-a-GCST003116/ |
Table 3
The results about MR.
| EXPOSURE | OUTCOME | METHOD | nSNP | BETA | SE | p-VALUE |
|---|---|---|---|---|---|---|
| Waist circumference | Coronary Artery disease | MR Egger | 340 | 0.6342 | 0.1354 | 4.064e–06 |
| Weighted median | 340 | 0.5183 | 0.07244 | 8.345e–13 | ||
| Inverse variance weighted | 340 | 0.497 | 0.04743 | 1.084e–25 | ||
| Weighted mode | 340 | 0.572 | 0.1467 | 0.000116 |
Figure 2
This figure presents the results of the leave-one-out sensitivity analysis, where each SNP is sequentially removed from the analysis to assess its individual influence on the overall causal estimate. The x-axis represents the causal effect estimates of waist circumference (WC) on coronary artery disease (CAD) after removing each SNP, while the y-axis represents the SNPs. The consistency of the causal effect across different iterations of SNP exclusion indicates that no single SNP disproportionately influences the overall results. This supports the robustness of the causal estimate and minimizes concerns about outlier SNPs driving the observed association.
Figure 3
The funnel plot visualizes the distribution of individual SNP effects on the association between WC and CAD. The x-axis represents the effect size (log odds ratio) of each SNP, and the y-axis shows the standard error of these estimates. The symmetry of the plot suggests that directional pleiotropy is unlikely to be a significant concern in this analysis. A symmetric distribution indicates that the observed effects are not biased by pleiotropic SNPs, supporting the validity of the causal inference.
Figure 4
This forest plot displays the individual causal effects of each SNP on CAD risk, with the x-axis representing the estimated effect sizes and confidence intervals for each SNP. Most SNPs show a positive association between WC and CAD, with confidence intervals overlapping, supporting a consistent direction of effect. This consistency across SNPs reinforces the overall finding that increased WC is causally associated with a higher risk of CAD.
Figure 5
The scatter plot compares the SNP-specific causal estimates across different MR methods, including inverse variance weighted (IVW), MR-Egger, and weighted median. The x-axis represents the WC effect size, while the y-axis shows the CAD effect size. The alignment of the SNP-specific estimates across different MR methods suggests consistency in the direction and magnitude of the association between WC and CAD. The close overlap of the IVW and weighted median lines supports the robustness of the results, indicating that the causal relationship remains consistent regardless of the MR method used.