Dear all,

I would appreciate some advice please!

I have performed a meta-analysis of 69 studies looking at the association between a risk factor and disease which appears to significant (RR 1.36%, p<0.001) but there is significant heterogeneity between studies (I2 77%; p<0.001).

From my meta-regression analysis, I am having trouble fully explaining the heterogeneity. Of 10 covariates/potential confounders, only 1 appears to significantly influence heterogeneity (p=0.004) bu the adjust R2 is only 13.7%, so it appears not account for much of the variance. There are clearly other subgroup differences but none that are statistically signficant.

Do you have any suggestions for some further tests that I could do to try to explain the heterogeneity?

Best wishes,