Harvard psychiatrist says Kennedy was “not accurate” in claiming his keto studies “cured” schizophrenia, cautioning that evidence remains preliminary and patients should not stop medications.