Blood pressure response profiles during graded exercise predict new-onset hypertension
A Carlen, T Lindow, V Elmberg, L Brudin, M Ekstrom, K Hedman, N CauwenberghsAbstract
Background
An exaggerated increase in systolic blood pressure (SBP) during exercise has been linked to the development of hypertension. Previous studies have typically relied on peak SBP or two-point slopes to characterize the SBP response to exercise, overlooking the potentially relevant information embedded in the full SBP time-series.
Purpose
To identify distinct responses from the complete SBP times-series recorded during maximal graded exercise and evaluate their relevance for prediction of future hypertension.
Methods
We analysed SBP recordings from 3,486 patients free from hypertension at rest who underwent maximal cycle ergometry (mean age, 50.6 years; 45% women). Incident hypertension was ascertained through nationwide registries including inpatient and outpatient medical diagnoses as well as dispensed antihypertensive medication. Group-based trajectory modelling (GBTM) was applied on the full SBP time-series to identify exercise SBP response profiles, grouping patients with similar exercise SBP trends. We determined multivariable-adjusted associations between future hypertension and the identified SBP responses. Additionally, stratified analyses were performed according to prehypertensive status (with prehypertension defined as SBP ≥130 mmHg or diastolic blood pressure ≥85 mmHg).
Results
In both men and women, GBTM identified three distinct SBP response profiles, reflecting low, mid and high SBP responses (Figure 1). Over a median follow-up period of 7.9 years (IQR: 5.5-10.9 years), 171 men (8.8%) and 169 women (10.9%) developed hypertension. Incidence of hypertension was tenfold higher in the high as in the low SBP responders (e.g. event rates in women: 27.8 for high vs 2.5 for low SBP responders). The higher risk in the high SBP responders was independent of traditional risk factors, including resting SBP (adjusted hazard ratio relative to average risk (HRadj): 1.26 [1.00-1.57] in men, 1.59 [1.22-2.07] in women). In the whole cohort, patients with prehypertension and a high SBP response exhibited the greatest risk for future hypertension (e.g. HRadj in women: 1.92 [1.48-2.49]), whereas those with normal resting blood pressure and low SBP response had the lowest risk (e.g. HRadj in women: 0.51 [0.35-0.74]). Of note, high SBP response was independently associated with future hypertension among patients without prehypertension at rest (e.g. HRadj in women: 2.17 [1.49-3.17]), but not in those with prehypertension (Figure 2).
Conclusion(s)
Time-series analysis of exercise SBP revealed distinct response profiles with incremental value to resting blood pressure for predicting future hypertension. These findings suggest that integrative interpretation of exercise SBP may refine early risk stratification for hypertension, particularly in patients without prehypertension at rest.Figure 1Figure 2