The potential impact of antipsychotics on lipids in schizophrenia: is there enough evidence to confirm a link?