I suppose you think that knowledge of the condition might have an effect on quality of sleep or the reporting of that quality. That's reasonable, but hardly a reason not to engage in the experiment. No experiment is perfect.