Title: Diagnostic accuracy of different exercise blood pressure metrics in identifying hypertension on 24-h ambulatory blood pressure monitoring in athletes
Authors: Janssens, K
Foulkes, SJ
D’Ambrosio, P
Mitchell, AM
Rowe, SJ
Bekhuis, Y
Spencer, L
Parr, EB
Head, GA
Heidbuchel, H
Claessen, G
La Gerche, A
Issue Year: 2026
Publisher SPRINGERNATURE
Series JOURNAL OF HUMAN HYPERTENSION:
Abstract Exercise blood pressure (BP) metrics have been promoted as a means of identifying latent or mild hypertension in athletes. We evaluated the diagnostic accuracy of four exercise BP metrics to diagnose hypertension, defined by 24-h ambulatory BP monitoring (ABPM) as mean >= 130/80 mmHg, daytime >= 135/85 mmHg and/or nighttime >= 120/70 mmHg. Fifty-four endurance-trained athletes (48 [IQR 24-58] years, 67% male) underwent exercise testing with serial BP measurements and 24-h ABPM. Exaggerated exercise BP (EEBP) was defined as exceeding a threshold for any of the four metrics: 1) maximal systolic BP (SBPmax) using published cut-offs; SBPmax >= 220 mmHg and >= 200 mmHg in males and females, respectively, 2) SBP/Workload-slope, 3) SBPmax/Workload-ratio, 4) SBP at 2 Watts/kg. Receiver operating characteristic (ROC) curve analysis and Youden's Index determined optimal cut-offs and diagnostic performance. Twelve of 54 (22%) athletes met criteria for hypertension. SBPmax was reasonably sensitive for identifying hypertension (83% identified) but demonstrated poor specificity (62% of non-hypertensives misclassified), with an area under the ROC-curve (AUC) of 0.63. An EEBP cut-off of SBP > 176 mmHg at 2 Watts/kg had the best diagnostic performance with 100% sensitivity, moderate specificity of 62%, and AUC of 0.79. Other metrics, incorporating exercise BP and workload, SBPmax/Workload-ratio and SBP/Workload-slope, had moderate diagnostic utility (AUC = 0.71 and 0.67, respectively). In endurance athletes, exercise BP metrics demonstrated modest and variable diagnostic accuracy for identifying hypertension on 24-h ABPM. Assessment of SBP at a relative submaximal workload provided acceptable diagnostic accuracy while reducing overdiagnosis associated with published SBPmax thresholds.
URI: https://publications.svi.edu.au/publications/10064
Other Identifiers 10.1038/s41371-025-01089-3
Publication type Article