Forty studies (2,979 children) were included: 7 double-blind RCTs (342 children), 21 open-label studies (480 children), 2 naturalistic studies (n=64), 1 cross-sectional study (n=141), 1 retrospective descriptive study (1,674 children and adolescents), 6 retrospective chart reviews (n=270) and 2 case series (n=8).
Risperidone was evaluated in 19 studies, olanzapine in 12 studies, clozapine in 6 studies, quetiapine in 5 studies and ziprasidone in 2 studies.
Weight gain (34 studies).
The best evidence about risperidone came from 2 double-blind RCTs that reported significantly greater weight gain with risperidone compared with placebo at 5 and 8 weeks, respectively. Two longer term studies (1 open-label and 1 naturalistic study lasting 8 months and 1 year, respectively) also reported weight gain with risperidone. Comparative studies reported higher weight gain with AAs in comparison with typical antipsychotics, with the greatest weight gain associated with clozapine and olanzapine (1 RCT, 1 non randomised study and 1 cross-sectional study). Evidence about quetiapine was scarce (3 small studies with no control group). One of 2 studies of ziprasidone (a double-blind RCT) reported no difference in weight gain between ziprasidone and placebo at 8 weeks; the other (in children with excessive weight gain on other AAs) reported mixed results.
Glycaemia (8 studies).
Four uncontrolled studies reported no specific abnormalities of glycaemic control. One double-blind RCT reported a non statistically significant increase in blood glucose with olanzapine but not with risperidone or haloperidol, while 2 case series reported some hyperglycaemia with risperidone, quetiapine and olanzapine. One retrospective chart review reported a 0.9 per 100,000 risk of diabetes in adolescents taking olanzapine and a 0.0 per 10,000 risk in younger children. However, this study relied on spontaneous reports of side-effects and so its conclusions are limited.
Lipidaemia (5 studies).
One small open-label study and 1 chart review reported no change in cholesterol or triglycerides. Two small studies (1 case series and 1 open-label) reported increased cholesterol associated with olanzapine. One double-blind RCT reported a non statistically significant increase in low-density lipoprotein with olanzapine but not with risperidone or haloperidol.
Hyperprolactinaemia (14 studies).
Five studies (4 open-label and 1 case series) reported cases of hyperprolactinaemia associated with risperidone, whereas 2 double-blind RCTs reported non significant or limited increases in prolactin with risperidone. Three open-label comparative studies reported increased prolactin associated with haloperidol, clozapine and olanzapine. One retrospective chart review reported increased prolactin in younger children taking olanzapine. Two small open-label studies reported no change in prolactin with quetiapine. One double-blind RCT reported cases (5 out of 16 children) of transiently increased prolactin associated with ziprasidone.
Reversibility of MSEs.
One semi-naturalistic study reported weight loss after stopping risperidone. One open-label study reported a significant decrease in prolactin after reducing the dose of risperidone.