[dankai]

Look at the sample sizes and the demographics (https://imgur.com/a/LG5vVMF) of the groups the study (https://moffittcaspi.trinity.duke.edu/sites/moffittcaspi.tri...) compares. Also look at the publication bias of the first author (https://scholar.google.com/citations?view_op=list_works&hl=e...).

[SubiculumCode]

if you mean the very small mean iq difference at baseline, that is hardly a significant concern, here especially in a longitudinal design. As far as publication bias, do you mean expert in his/her field?