We report on a total of 10 studies including overall 16,165 subjects, investigating the association between HRE and future risk of developing aHT. These studies reported on a positive association between HRE and a higher risk of developing aHT as compared to individuals without HRE. The consistency of these findings were also confirmed in 17 out of 18 studies which were recently analysed in a systematic review studying the impact of HRE on the development of future aHT in normotensive individuals. According to this review, patients who manifested HRE during cardiopulmonary exercise testing (CPET) had a 1.4 to 4.2-fold higher risk of developing aHT [33].

Jae et al. went a step further and analysed the prognostic significance of blood pressure values analyzed as continuous variables with respect to the future development of aHT. They found that the most significant values for predicting future aHT were blood pressure values > 181 mm Hg or a relative SBP of 52 mm Hg during exercise testing (a relative SBP was defined as peak SBP minus resting SBP) [34]. These values are missed in the currently used definitions of HRE.

In the CARDIA study including 3741 young adults with no history of aHT, Manolio et al. also investigated the role of HRE and the development of aHT. These subjects underwent treadmill testing and performed a follow-up clinical visit after 5 years to evaluate for the development of aHT. HRE was defined as an increase of SBP > 210 mm Hg in men and > 190 mm Hg in women. 18% of all subjects met these criteria for HRE and, after a 5-years follow-up, presented with a 1.7-times higher likelihood to develop aHT as compared to subjects with normal SBP values during stress exercise. However, after stratification for race and gender, it turned out to be significant only in black women [22].

In 2000, Miyai et al. focused on 239 male subjects with high normal BP (SBP 130–139 mm Hg, diastolic blood pressure (DBP) 85–89 mm Hg). Subjects with any history of cardiovascular disease, stroke or diabetes or the usage of any drugs influencing BP were excluded. After a mean follow-up of 5.1 years, HRE turned out to be an independent risk factor for the development of aHT on multivariate analysis using the Cox proportional hazards survival model after adjusting for traditional risk factors (relative risk = 2.31, 95% confidence interval = 1.45 to 6.25) [35].

Yzaguirre et al. investigated the blood pressure thresholds for the development of late onset aHT in young patients (mean age 25.7 ± 11.1 years, 72% men) and investigated several cut-off values for HRE. The main differences between the six different definitions were whether the BP defining HRE was measured at maximum effort or at 100 watts. Statistically significant (odds ratio > 1 after adjustment for age) predictors were a maximal elevation of SBP values > 180 mm Hg at 100 Watts and DPB values > 95 mm Hg at peak exercise. Patients, whose values exceeded at least one of the two parameters were found to have a 70% increased risk of developing future aHT 20 years after the initial exercise testing [36].

In 2008, Farah et al. investigated 30 healthy individuals (no aHT, body mass index < 30 kg/m2 (BMI), normal lipid profile, no cardiovascular diseases and without known family history of hypertension), aged 20 to 64 years. A stress test using the Bruce protocol was performed and HRE was defined as a SBP > 200 mm Hg and an increase of DBP > 10 mm Hg during exercise testing. Additionally, another threshold, termed “hypertension in-exercise” (HIE), and defined as a SBP increase ≥ 200 mm Hg or DBP increase ≥ 100 mm Hg was investigated. Based on these parameters, the population was divided into 2 subgroups (17 normotensive vs. 13 hypertensive individuals). After two years of follow-up, individuals with HIE were more likely to develop aHT (defined as resting BP > 140/90 mm Hg) than their normotensive control group (77% vs. 6%, p < 0.001). Therefore, they concluded that abnormally high BP values during stress testing should be taken into account even in absence of aHT [37].

Odahara et al. published a prospective study in 2010 investigating the influence of potential risk factors for the development of aHT in 815 subjects with no cardiovascular disease and normal electrocardiogram results at baseline. HRE was defined as SBP > 250 mm Hg or DBP > 120 mm Hg. Over a mean follow-up of 7 years, 13.3% subjects developed aHT. An independent association between HRE and aHT was confirmed in multiple cox hazard analyses (hazard ratio: 2.26 (95% CI 1.34–3.73)) [38].

Sharabi et al. investigated 2783 healthy Israel Defense Force employees aged > 26 years in the setting of a periodic medical evaluation. Every employee aged > 39 years underwent an exercise stress testing with Bruce protocol. HRE was defined as an increase in SBP > 200 mm Hg or DBP > 100 mm Hg during exercise. After excluding women, subjects with known aHT and/or taking antihypertensive drugs, data from 190 male subjects (mean age 42.6 years) were analyzed. Out of the 190 subjects, 73 showed HRE. After a mean follow-up period of 5.7 years, patients with HRE had a 5.64 odds ratio (95% CI 1.40–22.30, p < 0.01) of developing aHT after adjustment for confounding factors [39].

Berger et al. investigated 7082 individuals (mean age 48 ± 9 years) subjects with > 4 consecutive exercise testing with BP measurement available and followed them up for 5 ± 3 years. Subjects with a previous diagnosis of aHT, on antihypertensive drugs, and history of cardiovascular disease were excluded. In a univariate Kaplan-Meier survival analysis, a cumulative probability (35%, p < 0.001) of future aHT for the highest quintile of SBP during exercise was found. Similar findings were described for DBP values. In a multivariate Cox proportional hazards regression modeling, subjects in the highest quartile of exercise SBP had a 2.58-fold (p < 0.001) risk of developing aHT. Each 5 mm Hg increase in exercise SBP and DBP were independently associated with, respectively, a 11% and 30% increased risk for new onset of aHT [40].

In contrast to the above mentioned studies, Lima et al. did not find any correlation between HRE and aHT and propose that HRE found in patients should rather be considered as an intermediate state between normotensive BP values and clinically established aHT [41].

To summarize, several studies convincingly show that HRE might be a possible intermediate condition in the development of aHT.

The primary patho-mechanism beyond the evolution from hypertension to HFpEF is based on the LaPlace’s law. Indeed, aHT results in a greater left ventricular pressure that triggers wall thickening in order to reduce wall stress. Consequently, left ventricular hypertrophy (LVH) develops. As a consequence, LVH leads to further myocardial impairments due to, among others, myocardial ischemia caused by a shortage of myocardial blood supply, especially in the midmyocardial layer [42, 43]. This, in turn, results in a fibrotic remodelling and stiffening of the ventricle, promoting elevated left ventricular (LV) filling pressures and therefore ensuing LV diastolic dysfunction [5, 44]. In support to the positive relationship between aHT and HFpEF, Redfield et al. showed that aHT clearly induced vascular remodelling and increased arterial stiffness in a subpopulation of patients particularly at risk for HFpEF (e.g. elderly and women) [45, 46]. Other factors, such as inflammation, might lead to coronary microvascular endothelial inflammation which might trigger LV hypertrophy [47, 48]. In summary, aHT plays an important role in the pathogenesis of HFpEF.

We report on a total of 6 studies including overall 1366 subjects, investigating the association between HRE and HFpEF. In the early 2000s, the definition of HFpEF as recently proposed in the current guidelines had not been developed, yet, and the condition we refer to HFpEF today, was referred to as diastolic heart failure. This needs to be taken into account when older studies investigating the association between HRE and HFpEF are analyzed.

As such, Kato et al. investigated 20 patients with diastolic heart failure and 20 age-matched (70 ± 4 years) hypertensive patients with LVH with no differences in BP values at rest and left ventricular (LV) mass. Despite similar relevant baseline characteristic, patients with diastolic heart failure had significantly higher SBP during exercise testing as compared to the control group. Mechanistically, it was suggested that ventricular and arterial stiffness, altered atrial constriction, different preload reserves and higher LV afterload in patients with HRE resulted in further deterioration of myocardial relaxation and LV filling [49].

Takamura et al. investigated 129 patients with coronary artery disease undergoing exercise test and divided them into three age-matched groups: HRE (defined as HRE with normal BP at rest), HRE + aHT, and a control group without normotensive values both at rest and during exercise. In this study, HRE was defined as SBP/DBP ≥ 210/105 mm Hg in males and ≥ 190/105 mm Hg in females, as measured after 6 min of exercise testing using the modified Bruce protocol. On echocardiography, subjects included in the groups HRE and HRE + aHT presented with significantly lower e’ values (e.g. early diastolic mitral annular velocity) and a significantly higher E/e’ values (which is a surrogate parameter reflecting LV filling pressure) [50].

In a previous, meta-analysis including 12 studies and 46.314 individuals which were followed-up for 15.2 ± 4.0 years, Schultz and al. found a 36% higher rate of adverse cardiovascular events (defined as nonfatal and fatal myocardial infarction, nonfatal or fatal cerebrovascular events or the development of coronary artery disease) and mortality in patients with HRE in comparison to subjects without HRE [51].

Lee et al. investigated the relevance of left atrial volume index (LAVI) in 118 hypertensive and 45 normotensive individuals and found it to be an independent predictor of HRE in hypertensive individuals but not in the normotensive control group after controlling for body mass index, age, gender and peak oxygen consumption [52].

Chung et al. investigated data from 797 patients (38% women, mean age 64 ± 10 years, 75% with a history of hypertension) and showed that the prevalence of HRE was higher in women than in men. Furthermore, patients with HRE presented with pathologic variations in LV mass index, LAVI, E/e’, end diastolic elastance and relative wall thickness, which are echocardiographic parameters of diastolic dysfunction and ventricular remodelling and known risk factors for the development of HFpEF [53].

Mottram et al. investigated 400 patients who initially presented with chest pain and underwent a cardiological work-up including exercise stress test, electro- or echocardiography and compared them to 17 gender- and age-matched controls [23]. Among the patients, 41 had HRE (defined as > 210/105 mm Hg in men; > 190/105 mm Hg in women), 22 presented with resting hypertension (HRE + aHT), and 19 were normotensive [23]. In this study, a high-normal resting BP (mean 136/86 mm Hg) was detected in the HRE-group. Such high-normal resting BP values represent a risk factors for the development of future aHT and, consequently, of HFpEF [54]. Controversially, no association was found between HRE and LV hypertrophy after correcting for confounding factors. The authors assumed that previous studies showing such correlations used less sensitive measurement methods to assess diastolic and systolic function [23].