MR analysis and sensitivity analysis
We primarily employed the inverse variance-weighted (IVW) method to
assess the impact of the telomere length on the percentage of different
immune cells(15). Additionally, we utilized four supplementary analyses
to confirm the results, including weighted mode, weighted median, simple
mode, and MR-Egger regression(16-18). After conducting the MR analysis,
we performed sensitivity analyses including heterogeneity and pleiotropy
assessment. Heterogeneity among SNP-specific causal estimates was
evaluated using meta-analysis methods. Therefore, Cochran’s Q statistic
was calculated to measure heterogeneity between the estimated effects of
individual telomere length associated SNPs on outcomes(19), the Pvalue greater than 0.05 indicating a lack of heterogeneity. To
investigate whether the telomere length associated SNPs exhibit
pleiotropy, the MR-Egger regression test was employed to identify
potential pleiotropy, with a P value greater than 0.05 for the
MR-Egger intercept indicating a lack of horizontal pleiotropy(18).
Additionally, MR-PRESSO (Mendelian Randomization Pleiotropy RESidual Sum
and Outlier) analysis was conducted to detect wide-ranging horizontal
pleiotropy in the causal relationship(20). Furthermore, leave-one-out
analysis was employed to identify relevant SNPs that may influence the
causal effects and identify potential outliers. The relationship was
quantified by calculating odds ratios (OR) with 95% confidence
intervals (CI). Scatter plots and funnel plots were generated to provide
a clear visualization.