Lieberman et al1 reported a significant decrease in whole-brain gray matter in patients diagnosed with a first episode of schizophrenia spectrum disorders treated with haloperidol but not in those treated with olanzapine. Although Lieberman et al took into consideration several possible confounding factors, weight gain induced by neuroleptics was not accounted for. In a previous study, using a sample overlapping with the present one, it was shown that patients treated with olanzapine gained significantly more weight compared with those treated with haloperidol.2 In view of some evidence showing that brain ventricles and brain volume could be affected by the overall body weight3 and nutritional status,4 controlling for weight gain may be important. This may be relevant since patients treated with haloperidol had, on average, a longer duration of illness, which may be associated with more protracted poor nutritional habits. In fact, if we consider only measurements from weeks 12, 24, and 52 (Figure 2 of the Lieberman et al article), the slopes of whole-brain gray matter decrease in both groups are parallel, indicating that the dynamic change of gray matter volume is similar over time. The difference of gray matter in week 12 between the 2 groups may reflect a general nutritional difference rather than a specific effect of olanzapine on counteracting any putative intrinsic pathological process related to schizophrenia. In addition, the fact that patients treated with olanzapine gained gray matter at week 12 is not compatible with the hypothesis of slowing a putative atrophic effect of the disease but rather with other mechanisms of action of olanzapine that may be related to its effects on body mass. In fact, a formal test for differences in the course of gray matter loss over time could have been provided in the form of treatment × time of assessment interaction (excluding week 104 because of small sample sizes).
Joober R, Schmitz N, Malla A, Sengupta S, Karma S. Is Olanzapine a Brain-Sparing Medication?. Arch Gen Psychiatry. 2006;63(11):1292. doi:10.1001/archpsyc.63.11.1292