highly significantp < 0.025
A p-curve analysis was conducted on 16 statistically significant results ( p < 0.05), 13 of which were highly significant ( p < 0.025).
A p-curve analysis was conducted on 16 statistically significant results ( p < 0.05), 13 of which were highly significant ( p < 0.025).
To address statistical dependency between effect sizes arising from potentially overlapping cohorts, a three-level random-effects model was employed with the aforementioned clusters, resulting in a summary effect of 1.40 mg/dL (MD = 1.40 mg/dL, 95% CI: −2.81 to 5.61, p = 0.50), which similarly did not reach statistical significance ( Figure 3 ).