Is Isolated Diastolic Hypertension an Important Phenotype?

Abstract

Purpose of Review

Isolated diastolic hypertension (IDH) is a frequent hypertension phenotype. We review IDH pathophysiology, risk stratification, and therapeutic decisions.

Recent Findings

Recent guidelines lowering blood pressure cutoff levels have increased IDH prevalence and likely decreased associated cardiovascular risk. Long-term cardiovascular risk and pharmacological intervention in IDH are controversial. Narrow pulse pressure and other physiological and epidemiological characteristics are shared with a systodiastolic hypertension (SDH) subgroup. We propose that IDH be incorporated into a broader category, predominantly diastolic hypertension (PDH), defined by pulse pressure ≤ 45 mmHg and includes IDH and SDH with a narrow pulse pressure.

Summary

IDH-PDH is associated with cardiovascular risk in the long term, especially in young patients. Standardization of the IDH definition and population may contribute to future research to understand genetics, pathophysiology, and eventually therapy in this important subgroup of hypertensive patients.

Introduction

Blood pressure (BP) is a cyclic phenomenon that is coupled with the phases of the cardiac cycle. The cardiac ejection phase (systole) is associated with maximal blood pressure peak systolic blood pressure (SBP), whereas the relaxing or filling cardiac phase (diastole) is coupled with diastolic blood pressure (DBP), or minimal pressure. Classically, SBP results from cardiac output (CO) and vessel mechanical properties, while DBP is a consequence of potential energy accumulated in the vessel due to the elastic components of the vessel. Sustained increase in BP is called hypertension, the main risk factor for myocardial infarction, stroke, and vascular diseases, and is the primary contributor to disease burden globally [1]. Several global, national, and regional initiatives are underway to improve BP control and reduce the associated burden of disease [2].

American College of Cardiology–American Heart Association (ACC–AHA) 2017 clinical guidelines for high-BP management define hypertension as SBP and/or DBP steadily ≥ 130 and 80 mmHg, respectively [3••]. Other global and regional guidelines maintain the limits of < 140/90 mmHg SBP/DBP [4••, 5, 6]. Most hypertensive patients have both SBP and DBP elevated; however, subgroups of patients may only have one above normal, called isolated systolic hypertension (ISH) or isolated diastolic hypertension (IDH), respectively. These subgroups have been impacted by the new guideline cutoffs, and thus, there are new groups of IDH/ISH patients in different parts of the world. IDH and ISH in younger patients have different population distribution, pathogenic mechanisms, and risk, and eventually may constitute an important opportunity to explore different causes of hypertension that remain inconclusive for > 90% of hypertensive patients [7, 8]. Additionally, patients can be classified according to pulse pressure (PP), i.e., the difference between SBP and DBP. Epidemiological data have shown that elevated or wide PP, especially in elderly people, may have a poor prognosis. Much less interest has been shown in patients with narrow PP, where the majority has IDH or predominantly diastolic hypertension (PDH) [9], because of the apparent low CV risk, mainly driven by the young age of these patients. In this review, we explore the prevalence, etiology, prognosis, and treatment of IDH and review the importance of classifying patients according to PP.

Epidemiology of IDH

IDH constitutes the most frequent hypertensive phenotype in individuals younger than 50 years. In the USA, 46.9% of untreated hypertensive patients younger than 50 years present isolated DBP > 90 mmHg according to the NHANES III study [10]. Prevalence increases in those aged < 40 years, where IDH is present in > 55% of untreated patients, with a general prevalence of 2.6% [10, 11]. Overall, peak IDH prevalence is between 30 and 49 years old [12]. In the general population aged > 18 years, IDH prevalence has been reported between 2.5 and 7% [12–15, 16•, 17]. IDH prevalence in hypertensive patients decreases with age, being < 15% in the fifth and sixth decades and almost nonexistent in untreated elderly patients [10]. In children younger than 18 years, the prevalence of IDH is 1.9%, and around 13.1% of hypertensive children have IDH, whereas the overwhelming majority (76%) have ISH [18•].

However, the last ACC–AHA guidelines [3••], which lowered the diastolic cutoff to 80 mmHg, increased IDH prevalence, making this phenotype more common in clinical practice [19•]. A recent study demonstrated that the number of patients with IDH—DBP ≥ 80 mmHg with SBP < 130—had increased significantly to 6.5% from 1.3% under the JNC7 guidelines [20••]. Table 1 shows prevalence and the potential increment of different BP phenotypes after the recent guidelines. Another recent study in the Chinese population showed an increase of unmedicated IDH from 7.79 to 24.72% using ACC–AHA vs JNC7 guidelines (< 140/90) [21•]. Also, a study from Brazil on a general unmedicated population showed a ninefold increase from 3.9 to 34.1% [22]. In the KNHANES IV–VII survey in the South Korean population, Cho et al. reported a 3.4-fold increase from 5.2 to 17.9% [19•].

Table 1.

Prevalence of different blood pressure phenotypes among adults aged 18–39 and ACC/AHA 2017 guidelines impact

	ISH	IDH	SDH
Overall prevalence	1.57	2.66	0.93
Male	2.23	4.09	1.42
Female	0.92	1.25	0.46
ACC/AHA 2017 increment (%)	−10	4.3	30.5

IDH isolated diastolic hypertension; SDH systodiastolic hypertension; ISH isolated systolic hypertension

Modified from [10]

IDH is slightly more frequent in males than females [12]; however, it is far more equally distributed than the ISH phenotype in youth, which is mainly present in males [23]. As such, IDH prevalence is 6.2–8.7% in males and 2.9–6.6% in females, while 65% of IDH patients are male [12, 24, 25]. Overweight and obesity are highly associated with this phenotype. Particularly central obesity, linked with other components of metabolic syndrome, is frequently associated with IDH in young people [12, 14, 16•]. However, other studies have shown very low overweight/obesity prevalence in young IDH patients [26, 27]. It is unclear whether the association with obesity is because it advances the development of hypertension in an individual prone to hypertension or if obesity stamps a pulse pressure pattern on these predisposed individuals. We favor the latter, as obesity is associated with an increase in all hypertensive categories (IDH, ISH, and SDH) [24]. Studies showing socioeconomic characteristics of the IDH population reveal that in general, IDH patients tend to have higher education and, in some reports, more income than other hypertension subtypes [11, 16•].

Some reports have shown that awareness and treatment of hypertension are very low among IDH patients [10]. This may be explained by several factors, including the general concept that “SBP is all that matters” in terms of cardiovascular (CV) risk and that usually these patients are young and hypertension mostly asymptomatic [28•]. Awareness among IDH patients (46.8%) in the NHANES III cohort was significantly lower than ISH (58.4%) and SDH (67.2%) patients [10]. Data from the China PEACE Million Persons Project showed that awareness among IDH patients was 10.3% and that 86.1% of these were without treatment [16•].

The Framingham Heart Study showed that new-onset IDH (using the < 140/90 definition) derives mainly from normotensive and high-normal patients, and it is uncommon that an ISH or SDH patient transforms to an IDH phenotype [24]. On the contrary, patients with IDH have a hazard ratio of 23 for converting to SDH in 10 years compared with normotensives [24]. Figure 1 shows the potential evolution of this phenotype. In this longitudinal cohort study, changes in BMI, cholesterol, and plasma glucose were predictors of IDH development in previously normotensive or normal-high-BP patients, adjusted to baseline characteristics. The Harvest study in young adults confirmed that once a diagnosis of IDH is obtained by either clinic or ambulatory BP, those patients have higher odds than normotensives of being persistently hypertensive after around 7 years of follow-up [29].

Fig. 1.

Natural history of the blood pressure phenotype. IDH and some SDH with narrow pulse pressure are classified as a PDH. Those patients may evolve to SDH. These patients can remain as an IDH or evolve to ISH. Recent guidelines lowered cutoff levels and IDH may be associated with low cardiovascular risk; however, young patients have long-term increased cardiovascular risk. ISH corresponds to elderly patients. Modified from [9]. IDH, isolated diastolic hypertension; SDH, systodiastolic hypertension; ISH, isolated systolic hypertension; PDH, predominantly diastolic hypertension

Pulse Pressure and Predominantly Diastolic Hypertension

PP represents the pulsatile component of systemic BP, and the steady component is represented by mean arterial pressure. In general, normal PP is considered to be 50 mmHg [30]. IDH patients have narrow PP, indicating that DBP is the main contributor to mean arterial pressure. In the SDH group, there are narrow, normal, and wide pulse pressure patients, and the only reason they are grouped together is the need for guidelines to set limits on antihypertensive medication initiation. The arbitrary limit of IDH and SDH separates patients with a narrow pulse pressure phenotype that otherwise may be similar. As such, some SDH patients with narrow PP most likely share epidemiological characteristics and pathophysiological contributors with IDH [9]. The same patient will maintain his/her pathophysiology of narrow PP even if trespassing the arbitrary IDH–SDH boundary, and will not be subject to the geographical interpretation of clinical trials and guidelines. Furthermore, the ability to maintain the same phenotypic category may allow longitudinal follow-up of these patients and perhaps individualization if this subgroup is more prone to different patterns of organ damage, e.g., proteinuria or ventricular hypertrophy.

We have proposed that hypertensive patients with narrow PP, including IDH and some SDH with PP < 45 mmHg, may be classified as having PDH, in order to help identify genetic and physiopathogenic mechanisms [9]. Some studies have revealed hemodynamic and demographic differences among young hypertensive patients with different PP. The Enigma study compared ISH patients with young essential hypertensives. The essential hypertensives had PP < 50, were equally distributed with regard to sex, and have different hemodynamic profiles from those with wide PP, who were predominantly male [23]. Some studies have shown that PP is highly familial and associated with such genetic markers as telomere length [31]. Therefore, we believe that PDH patients may be a special “cluster” with some IDH characteristics in common, no matter which definition is used for SBP and DBP (Fig. 1).

Hemodynamic Profiles in IDH and PDH

There is a tendency to simplify the hemodynamic patterns behind hypertensive pulse-pressure phenotypes. It is generally considered that ISH patients have increased arterial stiffness and/or high cardiac output (CO). On the other hand, IDH patients classically have been thought to have high peripheral vascular resistance (PVR) [23].

In addition, hyperdynamic phenotypes with high CO have been reported to be more prevalent in younger age or early hypertension stages, with a transition to decreased CO and high PVR in older age [32, 33]. However, that conclusion derives from cross-sectional studies, where no longitudinal data have been collected. According to our knowledge, there has been only one study, following 16 patients for 10 years, where this transitioned pattern was observed [34]. Some obvious contradictions arise from these “simplified” hemodynamic models. Would IDH patients be expected to have an early hyperdynamic phenotype [32, 34] because of younger age? Or should they have low CO and high PVR? [23].

BP phenotypes, i.e., ISH, IDH, or SDH, depend on arterial content and continent characteristics. In regression models where the outcome is PP, sex (male), stroke volume, pulse-wave velocity (PWV), and augmentation index (AI) are all independent determinants of PP [23]. In other words, blood volume, flow (stroke volume), and arterial compliance (PWV and augmentation index) are the main determinants of the phenotype.

In a small study, we reported PDH patients with expected high PVR (“vasotonic patients”), but also surprisingly others with high CO and low PVR (“hyperdynamic patients”) [26]. Contrary to the simplified hemodynamic model, it is possible to find IDH-PDH patients with high CO and low PVR and others with low CO and high PVR. Similar observations have been reported in ISH patients [23]. While some patients have the classic pattern of high CO and low PVR, there are others with low CO and high PVR and intermediate phenotypes. As such, it is very difficult to predict the hemodynamic phenotype behind a PP phenotype obtained on peripheral BP measurement. Moreover, other clinical characteristics, such as heart rate, body mass index, or sex, cannot predict the hemodynamic phenotype with precision [26].

According to these studies, it is possible to find ISH or IDH patients with a hyperdynamic profile where high CO and low PVR are present. The difference between these hyperdynamic ISH and IDH patients is in arterial compliance: in hyperdynamic IDH-PDH patients, high arterial compliance is found [26]. Less intuitive is the situation where low CO and high PVR are present in an ISH or IDH patient. In those ISH patients with low CO, increased arterial stiffness or an alteration in wave reflection may explain the BP phenotype. However, according to the Enigma study, up to 11% of ISH patients may have low CO and low PWV, indicating that other variables may also be participating, such as total cross-sectional arterial area [35].

Other studies in young hypertensive patients have pointed out different hemodynamic patterns according to sex or body weight. Females and patients with obesity tend to have higher PVR [36•, 37], which may explain IDH-PDH hemodynamic pattern diversity, because almost half are male and overweight/obese according to different series [14, 26]. Future studies may use these covariates to adjust and reconcile peripheral BP and hemodynamic phenotypes.

Is IDH an Important Phenotype?

IDH is a well-accepted and easy-to-define phenotype used to divide hypertensive patients into different categories. Unfortunately, this phenotype has not increased our knowledge of the etiology of hypertension or helped with therapeutic recommendations. Moreover, recent analyses using the ACC–AHA BP thresholds have stated that IDH is not associated with CV risk [20••, 38••], and thus, IDH may not be an important or interesting phenotype at all. Also, unfortunately, it is the consequence of expert opinion interpreting results of major hypertension trials that were performed in patients generally aged > 55 years. As most patients with IDH are in younger age brackets, the systolic and diastolic hypertensive cutoff levels derived from these clinical trials may not be adequate for these younger patients. Moreover, as the results of these trials are open to interpretation, IDH does not mean the same in different parts of the world, as occurs today with European and American guidelines [4••, 39]. Therefore, IDH is a moving target subject to the appearance and interpretation of new trials that may move SBP and DBP levels in different directions. Some investigators have questioned the validity of IDH as a relevant phenotype [9, 40].

Searching for subgroups within essential hypertension patients is clearly necessary for understanding etiology and pathophysiology. Therefore, IDH has been proposed as a subgroup of interest. However, dividing hypertensive phenotypes by boundaries based on interpretation of clinical trials has certainly not helped in achieving more knowledge and understanding the cause of hypertension in these patients. Some associations have been made, but no clear progress in understanding the etiology and pathophysiology of these patients [20••, 41]. A more reasonable approach may be to group patients according to their PP pattern, as most likely patients with narrow PP have a distinct pathophysiology compared to patients with wide PP or predominantly systolic hypertension [23, 27].

Cardiovascular Risk and IDH Prognosis

In the absence of prospective studies performed on IDH patients, guidelines suggest using the same therapeutic BP levels as those for the general population. Some authors have studied risk and the association with CV events as an alternative to recommend guidance for pharmacological treatment [20••]. However, BP threshold definitions must be considered, because decreasing SBP and DBP levels (ACC–AHA guidelines) has increased patient prevalence in the IDH category, who are now at less risk than those categorized as IDH in the past [21•]. Study analysis must distinguish those including general population versus younger patients (< 50 years old) and those truly IDH-PDH unmedicated, or if the PP pattern is a result of pharmacological therapy.

Many analyses of IDH and CVD risk prior to ACC/AHA 2017 guidelines have been performed and summarized previously [40, 42]. A study is worth mentioning because of their extended follow-up. In young patients (< 50 years old) with 31 years of follow-up, IDH was found to increase the risk of CV disease (CVD) compared to normal BP and ISH in men, but less than ISH in women [43].

There are few studies that have used a BP cutoff < 130/80 mmHg, and some have shown increased CVD risk [44•, 45•], while others have failed to do so [38••, 46•]. An analysis from the UK Biobank found an association of IDH with CVD risk using the European Society of Cardiology definition, but no association using the ACC/AHA criteria [38••]. However, another analysis of the UK Biobank in unmedicated general population found an association of IDH with CV risk using both the ACC/AHA definition and National Institute for Health and Care Excellence guidelines [44•]. The association was mainly driven by myocardial infarction incidence and CV death. Recently, Wu et al. published a prospective study that compared CV outcomes and IDH in a community-based Chinese population [21•]. The study included patients without antihypertensive medication therapy in the general population that were followed for a 10-year period. The study found that IDH defined by JNC7 criteria had a higher incidence of myocardial infarction, cerebral hemorrhage, and total CVD, while IDH had a higher incidence of cerebral hemorrhage and total CVD by the ACC/AHA definition. Mean age was 48 years for the latter, and 80% of the patients were men.

A recent study from South Korea on young patients (< 40 years old) not taking antihypertensive medication at baseline found an association between CV risk and stage 1 IDH, ISH, and SDH. IDH and ISH had similar CV risk, and both were less than SDH [47•]. This analysis has been questioned because it was based on a South Korean healthcare program and BP may not have been measured with a standardized technique. Also, adjudication from hospitalization codes may not be accurate [48]. However, the accuracy of hospitalization codes against adjudication from medical records was > 90% for myocardial infarction and > 80% for stroke based on multiple validation studies in South Korea [49]. In addition, in a study from Japan, IDH stage 1 (DBP 80–89) and stage 2 (DBP > 90) showed increased CV events (myocardial infarction, angina pectoris, and stroke) [45•]. On a different note, a recent analysis of the government-managed National Health Insurance Corporation, which insures around 97% of South Koreans, found that IDH and ISH were equally associated with chronic kidney disease, but had less risk than SDH in young patients [50•].

A recent cohort and meta-analysis showed association of IDH with incident systolic hypertension, but lack of association with subclinical CVD (coronary artery calcium score and CVD events) [46•]. Analysis of the ARIC study showed no association of IDH with CV outcomes (incident atherosclerotic CVD and heart failure) or chronic kidney disease using the ACC/AHA definition [20••]. This analysis was done on the general population (mean age 49.6 years). Adjustments were performed to consider antihypertensive medication use, but it is difficult to know in what category these patients would be if they were unmedicated. A sensitivity analysis using NHANES data (median age 40 years, age range ≥ 20 years) and the CLUE II cohort (median age 42 years) also showed a lack of association with CV risk [20••].

A review of the major studies evaluating IDH association with CVD risk has been published recently, where stratifying risk according to patient age was proposed. In young patients (< 40 years old), IDH may add CVD risk, but in older patients, the influence is weak [51••].

Treatment of IDH

To date, no claims have been made for therapy in IDH that go beyond treatment for the general population. No one has showed interest in performing long-term therapeutic studies in these patients, most likely because they are at low CV risk and long follow-up will be needed. PDH patients have a different pathophysiology than other phenotypes, and thus clinical trials within this group of young patients should be performed to determine longitudinally the correct time to start antihypertensive medications. The pursuit of therapeutic thresholds should separate narrow PP hypertension from ISH in young patients, as clearly these two subgroups do not have the same risk [43].

Based on the high risk of transition to any form of systolic hypertension and the associated CVD risk in young IDH patients, there is no doubt as to the importance of periodic BP evaluation and initiation of lifestyle improvements in these patients [24, 46•]. Although no special antihypertensive therapy has been proposed for IDH, based on the predictors of hypertension progression, we would strongly recommend decreasing body mass index, treating any associated dyslipidemia, and incorporating a Mediterranean-style diet, in order to limit salt consumption and favor natriuresis with a potassium-rich diet [52].

For high CVD risk IDH patients, trials are needed with different antihypertensive medications, as some mechanisms of action may have better efficacy in this group. We predict that given vasoconstriction and increased PVR are important hallmarks in this phenotype, vasodilators may have an advantage over other therapeutic agents. However, as stated previously, the hemodynamic phenotype is highly heterogeneous in IDH-PDH patients, and the use of noninvasive hemodynamic evaluation may have a place in guiding the therapeutic approach in these patients. A subanalysis of the Progress study [53] showed that the use of perindopril–indapamide was beneficial for secondary prevention in IDH patients. On average, this combination decreased SBP and DBP by 6.2 and 2.3 mmHg, respectively, in IDH patients, confirming the clinical observation that DBP is more difficult to decrease than SBP.

Certainly, the PDH subgroup would not benefit from the suggestions by several authors that only SBP matters [28]. These recommendations are based on therapeutic results from older patients. This suggestion may or may not be true for general population therapeutic decisions, but this would never help to increase awareness of different hypertensive phenotypes. We suggest the need for different target BP levels for isolated/predominantly diastolic hypertension and isolated/predominantly systolic hypertension in young patients, as they clearly have different long-term risks [43].

Conclusion

IDH is an established hypertensive phenotype that is reported very often in the literature. Unfortunately, its definition has been subject to interpretation and has changed over time. Furthermore, it may continue to change as new trials are reported. A lot is known about this phenotype, but no clear advancements in hypertensive pathophysiology have been made. Maybe a more reasonable and long-term sustainable phenotype is PDH, which includes IDH regardless of which definition is used and SDH with narrow PP. It makes sense that an individual will have the same initial pathophysiology, even as hypertension gets worse. We propose that hypertensive patients categorized according to their PP and PDH be a phenotype that is reported and studied for genetics, pathophysiology, and pharmacological therapy.