Plasma beta-endorphin levels were measured in 13 autistic children, ag
ed 3.67 to 11.67 years at the end of treatment (naltrexone, haloperido
l, pimozide, or placebo) and in 5 of the 13 children also at baseline.
Baseline plasma beta-endorphin levels were lower than those reported
in the literature. There was a strong correlation between plasma beta-
endorphin levels and severity of sterotypies in all children. Naltrexo
ne did not seem to have a specific effect on plasma beta-endorphin lev
els; short-term haloperidol treatment was associated with an increase,
whereas long-term haloperidol treatment seemed to have a depressive e
ffect on plasma beta-endorphin levels, which rose after withdrawal of
haloperidol.