[Sources of error in measuring reference values]

For the solution of many problems it is essential to separate the reaction to a stimulus from the spontaneous circadian rhythm of the variable. But disturbing masking effects evoked by the actual activity level and after effects of earlier activities (masking effects) may obstruct such separations. Several protocols have been already described to reduce masking effects. They differ mainly in the quantity of the standardized activity. Those, which try to avoid masking due to sleep, have to use higher levels of activity to keep the subjects awake. Experimental comparisons of masking effects due to activity and due to sleep as well as to sleep deficits clarify that sleep is a less potent masking agent as often expected and may be allowed in routines designed to reduce masking effects. Masking effects due to sleep deficits and moderate activities are commonly underestimated and should be avoided as far as possible in such routines. Besides the circadian changes of reagibility, the extent of masking effects (even on body temperature) questions the usage of baseline measurements. Sleep can be investigated under reduced masking effects.