QUESTION: ... I have some questions about the regression section inES-SDM. I have read your research paper Meta-analysis ofFunctional Magnetic Resonance Imaging Studies of Inhibition andAttention in Attention-deficit/Hyperactivity Disorder in《Arch GenPsychiatry》2012 ... I am unable to understand some contents inthat paper. It's reported "The meta-regression analysis withlong-term stimulant medication use showed that the percentage ofpatients on long term stimulant medication correlated significantlywith increasing activation in the right caudate tail (r = 0.233,permutation-derived P <0.001), so that medication-naive patientshad significantly reduced activation compared with healthy controls(z = 2.149, P<0.001) and with long-term medicated patients (z =1.646, P <0.001), who did not differ from each other (Figure3)." From Figure 3, I can only understand medication-naivepatients had significantly reduced activation compared withlong-term medicated patients, however, I can't figure out how tocome to the decision that medication-naive patients hadsignificantly reduced activation compared with healthy controls? Iam wondering whether you have done some other analyses except themeta-regression of ES-SDM?

ANSWER: It must be noted that ES-SDM returns three maps when you conduct aregression:
- ..._0: difference between medication-naive patients and healthycontrols.
- ..._1: difference between medicated patients and healthycontrols.
- ..._1m0: difference between ..._1 and _0, i.e. between medicatedpatients and medication-naive patients, i.e. the regression slope.
With ..._0 map you can know whether medication-naive patients hadsignificantly reduced activation compared with healthy controls.