randomized double-blind trials. In a 3-week trial  and a 4-week trial , olanzapine was superior to placebo. In a 4-week study , olanzapine appeared to have antimanic effects comparable to those of lithium. In patients with acute or mixed mania, olanzapine was superior to divalproex sodium and as effective as haloperidol. Olanzapine was generally well tolerated but was associated with significant weight gain.
In the series of articles mentioned earlier (in the section on clozapine), Volavka et al.  reported the results of a 14-week randomized double-blind trial comparing clozapine, olanzapine, risperidone and haloperidol in 157 partially responsive but residually symptomatic patients with schizophrenia. They found that the three SGAMs produced statistically significant reductions in symptoms from baseline, while haloperidol did not, but only clozapine and olanzapine were significantly better than haloperidol. However, the doses of olanzapine and risperidone were higher than usual (olanzapine mean, 30.4 mg/day; risperidone mean, 11.4 mg/day). Although the greater efficacy of the SGAMs was statistically significant, it was clinically modest. EPSs were more common in patients treated with haloperidol and risperidone, and weight gain was greater with clozapine and olanzapine treatment. Using data from the same study, Citrome et al.  examined treatment effects on aggressive behavior. They found that clozapine was significantly more effective in reducing hostility and aggression than haloperidol, while olanzapine and risperidone were not.
Conley and Mahmoud  reported a study of 377 patients with schizophrenia or schizoaffective disorder who were randomized to double-blind treatment with risperidone or olanzapine for 8 weeks. The mean (sd) doses were 12.4 mg/day for olanzapine and 4.8 mg/day for risperidone; 77.2% of the olanzapine-treated patients and 71.8% of the risperidone-treated patients completed the study. Both treatments produced significant symptom reductions; these did not differ in an end-point analysis, but did differ at week 8, favouring risperidone with respect to positive symptoms and anxiety/depression factors on the Positive and Negative Sympsom Scale (PANSS). EPS symptoms were comparable between treatments, but olanzapine was associated with more weight gain
Perry et al.  examined the relationship between plasma levels of olanzapine and the therapeutic response. They found a threshold level of 23.2 ng/ml above which a response was more likely. This suggests the potential benefit of higher doses of olanzapine to maximize the likelihood of exceeding the therapeutic threshold.
Weight gain and disturbances of the glucose and lipid metabolism have been the focus of a number of reports about the safety of olanzapine. Wetterling , in his review, calculated a monthly body weight gain of 2.3 kg. However, Kinon et al.  found that weight gain associated with olanzapine treatment occurred early and reached a plateau after approximately 9 months of treatment. Eder et al. , who also reported a significant weight gain with olanzapine treatment, found this attributable mainly to an increase in body fat rather than lean body mass. Serum levels of leptin also increased two- to threefold over the 8-week study period. Considerable weight gain in adolescent patients treated with olanzapine has also been found by Kumra et al. , Haapasalo-Pesu and Saarijarvi ,