Sleep polygraphy in schizophrenia: Methodological issues