Is schizophrenia associated with brain volume changes independently of medication?

Further to my previous post, I have been looking at the 2010 study by Jo Moncrieff & Jonathan Leo on the effects of antipsychotic drugs on brain volume. I had not looked at this study before my previous post and perhaps I was being unfair concluding there needed to be more systematic review of brain scan studies of patients with psychosis or schizophrenia prior to them receiving antipsychotic drugs, because Jo and Jonathan have already done it. In her book, The Bitterest Pills, Jo mentioned this joint paper, which was published by Robin Murray, as editor of Psychological Medicine, "despite opposition from most of the five referees" (p.157).

The authors managed to find 21 "drug-naive" studies published between 1995 and April 2009 of patients who had had no more than 4 weeks of antipsychotic treatment. They conclude:-

Most studies of drug-naive patients examined here did not report or detect differences in total brain volume, global grey-matter volume or CSF volumes between patients and controls, including three studies of untreated patients with long-term illness. These results are particularly remarkable, given the difficulty of selecting a comparable control group in these studies. The results suggest that the brain changes found in some first-episode studies may also be attributable to drug treatment, especially because some studies suggest that structural changes may occur after only short periods of treatment.

They reinforced this conclusion in response to correspondence:-

As we showed in our systematic review, a large majority of studies with drug-naive patients with psychosis or schizophrenia have not found any differences in global brain or grey-matter volumes, or in total CSF or ventricular volumes between patients and controls (Moncrieff & Leo, 2010). Although some of these studies reported differences in the volumes of specific structures, such as the thalamus and the caudate nuclei, others found no differences and multiple testing suggests some of the results may be false positives.

I'm left wondering exactly what the evidence is for Robin Murray's view, mentioned in my previous post, that "subtle brain changes [are] present at onset of schizophrenia".