Assessment of Construct Validity of Recovery Measures through Theoretical Formulations as Applied to Cardiovascular Parameters

Abstract

The recovery period is generally excluded under stress testing, despite its importance, because recovery measures lack information on construct validity. The smaller the carryover effect, the higher the construct validity. Theoretical formulae were derived for three recovery measures (i.e., total carryover [TCO], mean recovery rate [MRR], and area under the curve [AUC]) to assess to what extent this criterion is fulfilled. The cardiovascular parameters (particularly blood pressure [BP] and heart rate [HR]) were targeted. After applying this criterion to theoretical formulae and previous experimental findings, the three recovery measures were ordered as follows: MRR for BP (highly valid) > AUC for both BP and HR (moderately valid) > MRR for HR and TCO for both BP and HR (slightly valid). The unexpected finding of MRR for HR was due to the invisible carryover effect of vagal rebound on HR. The implication of the sharp contrast in MRR for BP versus HR is discussed, indicating their qualitative differences as cardiovascular parameters.