Lower peak systolic blood pressure during exercise testing predicts higher risk of all-cause mortality even when accounting for exercise capacity and other confounders

Abstract Background The risk associated with an exaggerated systolic blood pressure (SBP) response during exercise testing is controversial. In part, this may be due to the confounding effect of exercise capacity, as achieving a high workload is related to both better survival and higher peak SBP. Purpose We aimed to determine the prognostic value of the SBP response at clinical exercise testing, with and without accounting for exercise capacity and other important confounders, for predicting all-cause mortality. Methods We analysed data from 10,245 subjects aged 18–85 years (54% male), referred for bicycle ergometer exercise testing at a single hospital between 2005 and 2015. Resting SBP, peak SBP, peak Watt as well as the first and last SBP and corresponding Watt during exercise were retrieved. The slope of the increase in SBP per increment in Watt between the first and last measurement during exercise (the SBP/Watt-slope) was calculated. Survival status was determined from the National Cause of Death register up until April 2019. Cox proportional hazard ratios for all-cause mortality were calculated in males and females for peak SBP and the SBP/Watt-slope, and natural cubic spline models (with three knots) were computed to explore the risk of death based on a continuum of each SBP measure. The models were adjusted for age, SBP at rest, a diagnosis of diabetes, hypertension, baseline cardiovascular disease (ICD10 Chapter IX) and the use of cardiac and/or antihypertensive medications. In additional analyses, peak SBP and the SBP/Watt-slope were adjusted for peak Watt or exercise capacity (% predicted), respectively. Results Over a median follow-up of 7.9 (IQR 5.7) years, 890 (8.7%) subjects died. Compared to survivors, subjects who died were older (69±10 vs 57±14 years, p