Article ID: CJ-20-0972
Background: Rising blood pressure (BP) in the morning, known as the morning BP surge (MBPS), is known to pose a risk for cardiovascular events in hypertensive individuals. It was not known whether the MBPS was associated with a worse prognosis in patients with heart failure (HF) with a reduced (HFrEF) or preserved (HFpEF) ejection fraction.
Methods and Results: We performed a prospective, observational cohort study of hospitalized HF patients who underwent ambulatory BP monitoring (ABPM). The MBPS was calculated by subtracting the mean systolic BP (SBP) during the 1 h that included the lowest sleep BP from the mean SBP during the 2 h after waking. The MBPS group was defined as the top decile of MBPS (>40 mmHg). In all, 456 hospitalized HF patients (mean [±SD] age 68±13 years, 63.9% male) were followed-up for a median of 1.67 years. There were 90 events (16.3 per 100 person-years) of the composite outcome (all-cause mortality and worsening HF) in the HFrEF group, compared with 53 events (19.6 per 100 person-years) in the HFpEF group. Multivariate Cox regression analysis showed that MBPS was a significant predictor of outcome (hazard ratio 2.84, 95% confidence interval 1.58–5.10, P<0.01) in the HFrEF but not HFpEF group.
Conclusions: MBPS was found to be a novel predictor of worsening HF in patients with HFrEF.
With the aging of the population, the number of heart failure (HF) patients continues to increase.1,2 HF is a progressive disease with a poor prognosis, despite advances in HF treatment. The reported 5-year mortality rate of HF is approximately 50%.3 Patients with HF are classified primarily as having either HF with reduced ejection fraction (HFrEF) or HF with preserved ejection fraction (HFpEF), with the prognoses of these 2 types of HF reported being similar.4,5
Editorial p ????
Blood pressure (BP) is a leading risk factor for cardiovascular mortality and morbidity in not only hypertensive individuals, but also in those with HF. Several hypertension guidelines state that BP measured in an out-of-office setting, such as ambulatory BP monitoring (ABPM), is necessary for the diagnosis and treatment of hypertension.6 Various types of BP variability that can be detected by ABPM are also recognized as potential risk factors independent of mean BP levels. Nevertheless, the current HF guidelines only recommend adequate BP control regardless of HFrEF or HFpEF status.7 This absence of a more precise guideline is due to the lack of data for the management of BP in HF patients using ABPM.
Cardiovascular events include increases in BP in the early morning period.8 Thus, it has been investigated whether an increase in BP during the morning poses a potential risk of cardiovascular events. Several studies demonstrated that an excessive morning BP surge (MBPS) assessed by ABPM is associated with the incidence of cardiovascular events in hypertensive patients.9 Although the MBPS may present a risk of worse outcomes in HF patients, there has been no study about this association. In addition, differences in prognostic factors have been described for HFrEF and HFpEF, namely low systolic BP (SBP), high heart rate, and low body mass index (BMI) in HFrEF and atrial fibrillation in HFpEF.10 Thus, we hypothesized that the prognostic effect of MBPS may also differ between HFrEF and HFpEF. To test this hypothesis, we investigated whether MBPS assessed by ABPM performed in-hospital after HF treatment is associated with a worse prognosis based on HFrEF or HFpEF status.
The design of the present observational prospective cohort study has been described previously.11 Briefly, 536 symptomatic HF patients who were hospitalized due to HF were enrolled. After excluding 80 patients with incomplete data (Supplementary Figure 1), 456 patients were entered into the present analysis. All baseline examinations were performed after HF symptoms had improved and stabilized. MBPS was calculated using the BP values measured by non-invasive ABPM (TM-2430; A&D, Tokyo, Japan) and was calculated as described previously12 by subtracting the lowest nocturnal BP (1-h mean of the 3 BP readings centered on the lowest night-time reading) from the morning BP (2-h mean of 4 30-min BP readings just after waking up). Patients were also stratified according to the extent of their MBPS. In the present study, the top decile of MBPS (90th percentile) was used to define the MBPS group, as in previous studies.12,13 The nocturnal BP fall (%) was calculated as follows: (awake SBP − sleep SBP) / awake SBP × 100. Patients were classified into different groups based on the nocturnal BP fall as follows: (1) riser BP pattern, nocturnal BP fall <0%; (2) non-dipper BP pattern, nocturnal BP fall 0–10%; (3) dipper BP pattern, nocturnal BP fall 10–20%; and (4) extreme-dipper BP pattern, nocturnal BP fall >20%.14
Patients were followed-up for a median of 1.67 years. The primary endpoint of this study was the “composite outcome”, which consisted of all-cause mortality and hospitalization due to worsening HF. The first occurrence of the primary endpoint was included in the analysis. The number of study subjects, number of outcomes, and the follow-up duration in the present study differed from those in our previous report because the required data and the endpoints were partially different.
Statistical AnalysesData are expressed as the mean±SD, percentages, or as the median with interquartile range (IQR) for continuous variables, or as counts with percentages for categorical variables. Unpaired t-tests were used to test the significance of differences between the MBPS and non-MBPS groups among HFrEF and HFpEF patients. The Chi-squared test was used to compare proportions. Kaplan-Meier survival curves with log-rank statistics were used to test the significance of differences between the HFrEF and HFpEF groups. Univariate and multivariate Cox regression analyses were used to investigate the prognostic factors in HFrEF and HFpEF with adjustment for covariates. Univariate analysis was used in Model 1. Age, sex, and BMI were entered into multivariate analysis in Model 2 and a preliminary analysis was performed to select the variables to be entered into Model 3. Hematocrit, creatinine, sodium, B-type natriuretic peptide (BNP), left ventricular mass index, casual diastolic BP (DBP), 24-h DBP, awake DBP, β-blocker use, the use of antiplatelet drugs, ischemic heart disease (IHD), and non-IHD, which were significantly associated with the outcome in univariate analysis for patients with HFrEF, were entered into the multivariate model using the stepwise method in patients with HFrEF. Creatinine, sodium, and BNP were selected. Model 2 plus these selected variables were used in Model 3 in patients with HFrEF. Hematocrit, fasting glucose, BNP, the use of antiplatelet drugs, casual DBP, 24-h SBP, 24-h DBP, 24-h pulse rate (PR), awake SBP, awake DBP, awake PR, and sleep DBP, which were significantly associated with the outcome in univariate analysis for patients with HFpEF, were entered into the multivariate model using the stepwise method in patients with HFpEF. BNP, 24-h PR, and awake DBP were selected. Model 2 plus these selected variables except non-significant variables (i.e., sex, and BMI) were used in Model 3 in patients with HFpEF. Two-tailed P<0.05 was considered significant.
First, we investigated the association between MBPS and outcome in all patients with HF. During the median follow-up of 1.67 years, there were 143 events of the composite outcome (17.3 per 100 person-years) among patients with HF. All-cause mortality occurred in 50 patients, and hospitalization due to worsening HF occurred in 105 patients (12 patients were duplicated). The mean MBPS for all patients was 21±15 mmHg. The distribution of MBPS values is shown in Figure 1. The cut-off value for identifying the top decile (the MBPS group; n=44) was 40 mmHg. When patients with HF were divided into MBPS and non-MBPS groups, the use of angiotensin receptor blockers (ARBs), awake SBP, awake pulse pressure (PP), the prevalence of the extreme dipper pattern, and the prevalence of the dipper pattern were significantly higher in the MBPS than non-MBPS group. The prevalence of the non-dipper pattern and that of the riser pattern was significantly lower in the MBPS than non-MBPS group (Supplementary Tables 1,2).
Distribution of the morning blood pressure surge (MBPS). In the present study, values in the 90th percentile of MBPS were defined as the presence of MBPS.
We performed Kaplan-Meier analysis for the composite endpoint, as well as all-cause mortality and hospitalization due to worsening HF separately, in all 456 subjects according to the presence or absence of the MBPS. The MBPS group had a significantly higher incidence of the composite endpoint and hospitalization due to worsening HF than the non-MBPS group, but not a significantly higher incidence of all-cause mortality (Supplementary Figure 2). Univariate and multivariate Cox regression analyses were used to investigate the prognostic factors for the composite endpoint. In the univariate analysis (Model 1), MBPS was a significant predictor for the incidence of the endpoints in all patients with HF (hazard ratio [HR] 1.67, 95% confidence interval [CI] 1.04–2.68, P=0.03). In the multivariate Cox regression analysis, after adjusting for age, sex, and BMI, MBPS was no longer a significant predictor for the endpoints (Model 2). A multivariate Cox regression analysis was performed adjusting for confounding factors (Model 2 plus creatinine, BNP, 24-h PR, and awake DBP, which were selected in the preliminary analysis), and MBPS remained a significant predictor of the endpoints (Model 3; HR 1.94, 95% CI 1.17–3.22, P=0.01; Supplementary Table 3).
To investigate whether there was a significant interaction between left ventricular ejection fraction (LVEF) and the prognostic value of MBPS, patients with HF were divided into HFrEF and HFpEF groups using a cut-off value of LVEF 50%: there were 297 patients with HFrEF and 159 patients with HFpEF. During the median follow-up of 1.67 years, there were 90 events of the composite outcome (16.3 per 100 person-years) among patients with HFrEF. Of these 90 events, 66 were HF events and 33 were all-cause deaths (9 patients were duplicated). Among patients with HFpEF, there were 53 events of the composite outcome (19.6 per 100 person-years). Of these 53 events, 39 were HF events and 17 were all-cause deaths (3 patients were duplicated).
Each of the HFrEF and HFpEF patients were assigned to either the MBPS or non-MBPS group, and the characteristics of the groups were investigated (Table 1). MBPS was observed in 25 of 297 patients with HFrEF (8.4%) and in 19 of 159 patients with HFpEF (11.9%). Among patients with HFrEF, the mean age, the percentage of males, and BMI were similar in the MBPS and non-MBPS groups. The percentage of ARB use was higher in the MBPS than non-MBPS group. Awake PP, the value of MBPS, the prevalence of the extreme dipper BP pattern, and the prevalence of the dipper BP pattern were significantly higher in the MBPS than non-MBPS group. Sleep DBP, the prevalence of the non-dipper BP pattern, and the prevalence of the riser BP pattern were significantly lower in the MBPS than non-MBPS group (Table 2). Among patients with HFpEF, the prevalence of New York Heart Association Class III/IV and the relative wall thickness were significantly higher in the MBPS than non-MBPS group. The value of MBPS, the prevalence of the extreme dipper BP pattern, and the prevalence of the dipper BP pattern were significantly lower in the MBPS than non-MBPS group. The prevalence of the non-dipper BP pattern was significantly higher in the MBPS than non-MBPS group.
| HFrEF (n=297) | HFpEF (n=159) | |||||
|---|---|---|---|---|---|---|
| MBPS (n=25) | Non-MBPS (n=272) | P value | MBPS (n=19) | Non-MBPS (n=140) | P value | |
| Age (years) | 70±13 | 66±13 | 0.13 | 71±13 | 72±11 | 0.75 |
| Male sex | 18 (72.0) | 191 (70.2) | 0.53 | 11 (57.9) | 70 (50.7) | 0.37 |
| BMI (kg/m2) | 21.7±3.5 | 22.8±4.7 | 0.26 | 24.7±5.6 | 23.0±5.3 | 0.20 |
| NYHA Class III/IV | 7 (29.2) | 67 (24.7) | 0.39 | 10 (52.6) | 27 (19.4) | <0.01 |
| Casual SBP (mmHg) | 115±22 | 118±23 | 0.62 | 133±27 | 125±22 | 0.13 |
| Casual DBP (mmHg) | 69±13 | 72±15 | 0.33 | 73±16 | 69±13 | 0.20 |
| Casual PP (mmHg) | 47±19 | 50±24 | 0.57 | 60±23 | 60±26 | 0.92 |
| Casual PR (beats/min) | 73±15 | 72±15 | 0.86 | 74±13 | 71±14 | 0.53 |
| Underlying heart disease | ||||||
| IHD | 13 (52.0) | 109 (40.1) | 0.17 | 3 (15.8) | 27 (19.3) | 0.50 |
| Non-IHD | 12 (48.0) | 163 (59.9) | 0.17 | 16 (84.2) | 113 (80.7) | 0.50 |
| History of hypertension | 21 (84.0) | 190 (69.9) | 0.10 | 13 (68.4) | 99 (70.7) | 0.51 |
| Recurrent HF | 8 (32.0) | 76 (27.9) | 0.41 | 5 (26.3) | 25 (17.9) | 0.27 |
| Diabetes | 11 (44.0) | 107 (39.3) | 0.40 | 7 (36.8) | 49 (35.0) | 0.53 |
| Atrial fibrillation | 7 (28.0) | 66 (24.3) | 0.42 | 5 (26.3) | 46 (33.1) | 0.38 |
| Cardiovascular drugs | ||||||
| CCBs | 7 (28.0) | 66 (24.3) | 0.42 | 10 (52.6) | 67 (47.9) | 0.44 |
| ACEi | 10 (40.0) | 148 (54.4) | 0.12 | 7 (36.8) | 44 (31.4) | 0.41 |
| ARBs | 14 (56.0) | 83 (30.5) | 0.01 | 9 (47.4) | 70 (50.0) | 0.51 |
| β-blockers | 18 (72.0) | 215 (79.0) | 0.28 | 13 (68.4) | 80 (57.1) | 0.25 |
| α-blockers | 1 (4.0) | 5 (1.8) | 0.41 | 3 (15.8) | 11 (7.9) | 0.22 |
| Diuretics | 23 (92.0) | 247 (90.8) | 0.60 | 15 (78.9) | 122 (87.1) | 0.26 |
| Antiplatelet drugs | 13 (52.0) | 142 (52.2) | 0.57 | 6 (31.6) | 51 (36.4) | 0.44 |
| Anticoagulant drugs | 11 (44.0) | 143 (52.6) | 0.27 | 7 (36.8) | 66 (47.1) | 0.28 |
| Creatinine (mmol/L) | 1.2±0.6 | 1.2±0.7 | 0.61 | 1.2±0.6 | 1.2±0.9 | 0.86 |
| Hematocrit (%) | 38.0±6.4 | 39.7±6.7 | 0.23 | 35.1±7.6 | 36.6±6.8 | 0.37 |
| Fasting glucose (mmol/L) | 150±85 | 129±47 | 0.27 | 133±47 | 128±51 | 0.65 |
| HbA1c (%) | 6.3±2.5 | 6.1±1.7 | 0.55 | 5.4±1.6 | 5.8±1.4 | 0.33 |
| eGFR (mL/min/1.73 m2) | 54±29 | 58±26 | 0.44 | 53±29 | 57±43 | 0.69 |
| Sodium (mEq/L) | 138±5 | 138±3 | 0.99 | 138±3.0 | 138±11 | 0.78 |
| Potassium (mEq/L) | 4.5±0.6 | 4.5±2.5 | 0.80 | 4.3±0.4 | 4.4±0.5 | 0.26 |
| BNPA (pg/mL) | 295.0 (92.1–776.7) | 284.6 (138.0–559.5) | 0.85 | 99.6 (33.3–414.9) | 173.5 (75.1–274.5) | 0.30 |
| ECG-LVH | 6 (25.0) | 78 (28.8) | 0.45 | 4 (21.1) | 41 (29.3) | 0.33 |
| CTR (%) | 55±10 | 56±6 | 0.86 | 59±8 | 56±7 | 0.11 |
| LVDd (mm) | 58±11 | 58±9 | 0.83 | 47±9 | 49±8 | 0.45 |
| RWT | 0.36±0.10 | 0.38±0.13 | 0.45 | 0.58±0.20 | 0.48±0.12 | 0.046 |
| LVMI (g/m2) | 158±61 | 163±46 | 0.61 | 152±77 | 144±50 | 0.65 |
| LVEF (%) | 33.7±8.4 | 33.1±9.7 | 0.75 | 61.1±9.2 | 61.5±8.8 | 0.84 |
| 3% ODI | 13.0±13.7 | 13.9±12.3 | 0.72 | 12.6±9.7 | 12.0±12.6 | 0.85 |
| 3% ODI ≥5 | 16 (66.7) | 188 (71.2) | 0.40 | 14 (73.7) | 84 (61.8) | 0.23 |
Unless indicated otherwise, data are given as the mean±SD or as n (%). AGeometric mean with the range in parentheses. ACEi, angiotensin converting enzyme inhibitor; ARBs, angiotensin receptor blockers; BMI, body mass index; BNP, B-type natriuretic peptide; CTR, cardiothoratic ratio; DBP, diastolic blood pressure; ECG-LVH, left ventricular hypertrophy verified by electrocardiography; eGFR, estimated glomerular filtration rate; HF, heart failure; HFpEF, HF with preserved ejection fraction; HFrEF, HF with reduced ejection fraction; IHD, ischemic heart disease; LVDd, left ventricular diastolic diameter; LVEF, left ventricular ejection fraction; LVM, left ventricular mass; MBPS, morning blood pressure surge; NYHA, New York Heart Association; ODI, oxygen desaturation index; PP, pulse pressure; PR, pulse rate; RWT, relative wall thickness; SBP, systolic blood pressure.
| HFrEF (n=297) | HFpEF (n=159) | |||||
|---|---|---|---|---|---|---|
| MBPS (n=25) |
Non-MBPS (n=272) |
P value | MBPS (n=19) |
Non-MBPS (n=140) |
P value | |
| 24-h SBP (mmHg) | 119±18 | 117±18 | 0.61 | 122±31 | 123±17 | 0.88 |
| 24-h DBP (mmHg) | 68±7 | 70±11 | 0.28 | 73±16 | 69±13 | 0.16 |
| 24-h PP (mmHg) | 51±15 | 47±13 | 0.11 | 49±23 | 55±13 | 0.26 |
| 24-h PR (beats/min) | 70±10 | 69±11 | 0.60 | 69±11 | 66±11 | 0.33 |
| Awake SBP (mmHg) | 124±18 | 118±18 | 0.11 | 132±22 | 124±17 | 0.07 |
| Awake DBP (mmHg) | 71±7 | 71±10 | 0.89 | 75±14 | 69±9 | 0.09 |
| Awake PP (mmHg) | 54±16 | 47±12 | 0.02 | 57±13 | 55±13 | 0.47 |
| Awake PR (beats/min) | 71±10 | 70±11 | 0.57 | 70±11 | 68±11 | 0.37 |
| Sleep SBP (mmHg) | 108±20 | 113±21 | 0.24 | 118±24 | 120±20 | 0.64 |
| Sleep DBP (mmHg) | 62±8 | 68±12 | 0.02 | 68±18 | 65±10 | 0.46 |
| Sleep PP (mmHg) | 46±16 | 45±15 | 0.88 | 49±13 | 55±15 | 0.14 |
| Sleep PR (beats/min) | 68±12 | 67±12 | 0.59 | 66±11 | 64±11 | 0.34 |
| MBPS (mmHg) | 53±10 | 17±11 | <0.01 | 52±10 | 19±11 | <0.01 |
| Pattern of nocturnal BP fall | ||||||
| Extreme dipper BP pattern | 7 (28.0) | 7 (2.6) | <0.01 | 3 (15.8) | 2 (1.4) | 0.01 |
| Dipper BP pattern | 9 (36.0) | 51 (18.8) | 0.04 | 10 (52.6) | 36 (26.1) | 0.02 |
| Non-dipper BP pattern | 8 (32.0) | 152 (56.1) | 0.02 | 3 (15.8) | 56 (40.5) | 0.03 |
| Riser BP pattern | 1 (4.0) | 61 (22.5) | 0.02 | 3 (15.8) | 44 (31.9) | 0.12 |
Unless indicated otherwise, data are given as the mean±SD or as n (%). BP, blood pressure. Other abbreviations as in Table 1.
Among patients with HFrEF, the incidence of the endpoints was higher among those with MBPS than in the non-MBPS group (56.0% vs. 29.3%; P<0.01). However, among patients with HFpEF, the incidence of the endpoints was similar between the MBPS and non-MBPS groups (31.6% vs. 33.8%; P=0.54). Kaplan-Meier analyses of the endpoints revealed that, among patients with HFrEF, the incidence of the endpoints was significantly higher in the MBPS than non-MBPS group (P<0.01); in contrast, among patients with HFrEF, the incidence of the endpoints was similar between the MBPS and non-MBPS groups (Figure 2).
Kaplan-Meier curves for the composite endpoint according to the presence or absence of the morning blood pressure surge (MBPS) in patients with (A) heart failure with reduced ejection fraction (HFrEF; n=297) and (B) heart failure with preserved ejection fraction (HFpEF; n=159). There was a significantly higher incidence of endpoints in MBPS than non-MBPS patients in the HFrEF group, but not in the HFpEF group.
Univariate and multivariate Cox regression analyses were used to investigate the prognostic factors for the endpoints, with the results summarized in Table 3. In the univariate analysis (Model 1), MBPS was a significant predictor for the incidence of the endpoints in patients with HFrEF (HR 2.38, 95% CI 1.34–4.20, P<0.01). In the multivariate Cox regression analysis, after adjusting for age, sex, and BMI (Model 2), MBPS was a significant predictor for the endpoints (HR 2.16, 95% CI 1.21–3.85, P<0.01). In multivariate Cox regression analysis adjusted for confounding factors (Model 2 plus creatinine, sodium, and BNP, which were selected in the preliminary analysis), MBPS remained a significant predictor for the endpoints (HR 2.84, 95% CI 1.58–5.10, P<0.01; Model 3; Table 3). When the value of MBPS was entered into the model as a continuous variable, instead of MBPS or non-MBPS, MBPS was no longer a significant predictor for the endpoints (data not shown).
| Variable | Model 1 | Model 2 | Model 3 | |||
|---|---|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | |
| HFrEF (n=297) | ||||||
| MBPS (yes=1, no=0) | 2.38 (1.34–4.20) | <0.01 | 2.16 (1.21–3.85) | <0.01 | 2.84 (1.58–5.10) | <0.01 |
| Age | – | 1.02 (0.99–1.04) | 0.06 | 1.01 (0.99–1.03) | 0.44 | |
| Male sex (yes=1, no=0) | – | 0.97 (0.61–1.55) | 0.91 | 0.69 (0.43–1.11) | 0.13 | |
| BMI | – | 0.93 (0.88–0.99) | 0.02 | 0.94 (0.89–0.99) | 0.046 | |
| Creatinine | – | – | 1.72 (1.31–2.26) | <0.01 | ||
| Sodium | – | – | 0.90 (0.85–0.95) | <0.01 | ||
| BNPA | – | – | 1.50 (1.18–1.90) | <0.01 | ||
| HFpEF (n=159) | ||||||
| MBPS (yes=1, no=0) | 0.92 (0.39–2.15) | 0.84 | 0.97 (0.41–2.28) | 0.97 | 0.96 (0.34–2.71) | 0.94 |
| Age | – | 1.05 (1.02–1.09) | <0.01 | 1.06 (1.02–1.10) | <0.01 | |
| Male sex (yes=1, no=0) | – | 0.99 (0.57–1.71) | 0.96 | – | ||
| BMI | – | 0.95 (0.89–1.02) | 0.13 | – | ||
| BNPA | – | – | 1.73 (1.26–2.39) | <0.01 | ||
| 24-h PR | – | – | 0.95 (0.92–0.98) | <0.01 | ||
| Awake DBP | – | – | 0.92 (0.88–0.95) | <0.01 | ||
AGeometric mean. CI, confidence interval; HR, hazard ratio. Other abbreviations as in Table 1.
As indicated in Table 3, univariate and multivariate Cox regression analysis revealed that, in patients with HFpEF, MBPS was not a predictor for the incidence of the endpoints in any model.
This was a cohort study of hospitalized HF patients with ABPM. To the best of our knowledge, this is the first study to show the prognostic impact of the MBPS in patients with HFrEF. Although the BP level was low in the patients with HFrEF in this series, our analyses revealed that hemodynamic abnormality early in the morning made the outcome of HFrEF worse.
Causes of MBPS in HF PatientsThe causes of MBPS in HF patients remain unknown. It was previously shown that MBPS was associated with the extreme dipper BP pattern and excessive BP decline at night.15 Previous studies of hypertensive patients showed that aging, hypertension, diabetes, inflammation, alcohol intake, smoking, physical stress, psychological stress, and poor sleep quality increase the incidence of MBPS.9,16 In addition, neurohormonal abnormalities of the renin-angiotensin system and sympathetic nervous system were associated with mechanisms of the MBPS.17 Impaired baroreflex sensitivity (BRS) was observed in HF patients, and this condition leads to activation of the sympathetic nervous system. This mechanism could be the main cause of MBPS. In fact, a significant association has been reported between MBPS and BRS.18,19 In these reports, low BRS was associated with a large MBPS. Although there are no data regarding the association between MBPS and BRS in HF patients, BRS may play an important role in the incidence of MBPS in these patients.
MBPS Is an Indicator of Worsening HFAs noted above, MBPS is associated with low BRS.18,19 Low BRS activates the sympathetic nervous system and worsens the prognosis in HF patients.20 HF patients with low BRS have a poorer prognosis than HF patients without a low BRS.21 The reason for this could be that the lower BRS, the higher the activity of the sympathetic nervous system in HF patients. Therefore, the MBPS in HF is an indicator of low BRS, sympathetic nervous system activation, and a poor outcome.
It has not been established whether MBPS is associated with a poor prognosis only in patients with HFrEF. One of the reasons for this is the difference in BRS between HFrEF and HFpEF. It was reported that patients with HFrEF had a significantly lower BRS than patients with HFpEF.22 Because of the different patient backgrounds and comorbidities, the significance of BRS could differ between patients with HFrEF and HFpEF. Therefore, low BRS and MBPS could have a strong impact on the outcomes of patients with HFrEF.
Hemodynamic Effects of MBPSThe hemodynamic aberrations that occur due to MBPS could make the prognosis of HF worse. However, the hemodynamic stress of MBPS has not been completely clarified. MBPS is accompanied by increased left ventricular afterload early in the morning, and this affects the hemodynamics of the left ventricle. In the present patient population, MBPS was a prognostic factor only in patients with HFrEF and not in those with HFpEF. This difference can be explained by the difference in the left ventricular pressure-volume loops between patients with HFrEF and HFpEF.23–25 In patients with HFrEF, left ventricular contractility is decreased, and the end-systolic pressure-volume line is tilted and shifted to the right. Therefore, in individuals with HFrEF, the heart cannot eject blood into the aorta at high pressure as a result of the MBPS. Decreased stroke volume cannot maintain sufficient blood flow, resulting in increasing left ventricular end-diastolic pressure and volume, and an increase in left atrial pressure. Finally, lung congestion would occur as a result of an afterload mismatch in which stroke volume cannot be maintained while afterload increases.
In patients with HFpEF, the left ventricular pressure-volume loop is shifted up and to the left. BP elevation due to MBPS increases arterial elastance and decreases stroke volume. Both increasing heart rate and increasing preload to compensate for reduced stroke volume cause further deterioration of left ventricular diastolic function and an increase in left atrial pressure. In addition to this mechanism, a nocturnal fluid shift from the peripheral extremities to the heart increases left arterial pressure. Together, these effects would lead to lung congestion.
When the transient increase in afterload due to MBPS occurs, the stroke volume would decrease in both HFrEF and HFpEF patients, but adverse outcomes tended to occur only in patients with HFrEF in the present study. This discrepancy could be explained by left ventricular contractility. Because end-systolic elastance is decreased and the left ventricular pressure-volume loop is shifted to the right in patients with HFrEF, the stroke volume could be easily reduced in HFrEF compared with HFpEF even if the MBPS value is the same. Therefore, MBPS could have an adverse effect only in HFrEF.
Timing of BP Elevation and Different Outcomes Between HFrEF and HFpEFThis study showed that MBPS worsened the outcome of HFrEF. In contrast, a riser BP pattern (i.e., nocturnal BP exceeding awake BP) worsened the outcome of HFpEF in our previous study.11 Both MBPS and a riser BP pattern are transient BP elevations from night to morning, and it is unclear why the difference of the timing of the BP elevation causes different outcomes between patients with HFrEF and HFpEF. This difference could depend on the extent and duration of the BP elevation.
The BP of the MBPS in HF patients with MBPS was almost 50 mmHg in the present study, and the duration of the morning surge was 2 h. MBPS, even if only a transient BP elevation, could be sufficient to exacerbate outcomes in patients with HFrEF because of the aforementioned mechanism in patients with HFrEF.
Conversely, a riser BP pattern is one of the abnormal circadian BP rhythms in which nocturnal BP exceeds awake BP. The mechanism of the riser BP pattern could be a nocturnal fluid shift from the extremities to the heart. The BP value of nocturnal BP elevation was only 7 mmHg in our previous study in patients with HFpEF and a riser BP pattern.11 However, this BP elevation could continue throughout the night. This long-lasting overload could easily lead to high left atrial pressure because left ventricular stiffness is increased in patients with HFpEF, and could result in lung congestion. In contrast, the adverse effect of the riser BP pattern on hemodynamics may be only slight in patients with HFrEF because the extent of the BP elevation is small and the heart of HFrEF patients is not stiffer than that of patients with HFpEF.
Study LimitationsThis study has some limitations. First, the ABPM was performed during hospitalization, and ABPM in a hospital setting may not always represent an individual’s normal daily life and sleep habits. However, earlier studies showed that ABPM had prognostic significance even during hospitalization.26 Second, the doses of antihypertensive medications and the timing of drug administration could influence MBPS. However, multiple drugs are usually prescribed to HF patients. Third, the definition of HFpEF was based on the cut-off value of LVEF, and we did not diagnose diastolic HF using echocardiographic parameters of diastolic function. Fourth, the diagnosis of HF was not based on specified guidelines or Framingham criteria, but rather was made by cardiologists according to previous and current HF treatment guidelines.7 Fifth, the incidence of the outcome could be high in recurrent HF patients with HFrEF. One of the possible causes of such a relationship could be an intolerance or non-responsiveness to β-blockers, but this has not been investigated. Sixth, although the cut-off value of MBPS obtained from receiver operating characteristic analysis was not a risk of the composite outcome, this value of MBPS was lower than the value of the top decile of MBPS in both the present and previous studies.12,13 Therefore, exaggerated MBPS may be the only parameter with prognostic power in patients with HF, even if the value of MBPS in the present study was lower than that of previous studies. Further research is required to investigate the pathological threshold of MBPS in patients with HF. Seventh, 80 of 536 subjects were excluded because of incomplete data. The results of the present study could contain selection and indication biases.
In conclusion, MBPS was a prognostic factor for adverse outcomes in patients with HFrEF, but not in patients with HFpEF. ABPM could be an important tool for hemodynamic risk stratification for HFrEF.
PerspectivesMBPS is a phenotype of abnormal neurohumoral systems in HF patients. Interventions for MBPS have not been performed in HF patients. It is difficult to intervene in MBPS in HF patients because many factors are associated with MBPS. One potential strategy to reduce MBPS is drug chronotherapy. Other interventions targeting possible causes of MBPS are treatments for sleep apnea syndrome and alcohol abstinence. Further investigations of MBPS in HF are needed.
The authors gratefully acknowledge Kimiyo Saito, Chisato Mikogai, Hideko Taguchi, Kaori Kobayashi, Mika Kunimatsu, and Miki Sato for the coordination and data management of this study, Ayako Okura for editorial assistance, and Hiroshi Kanegae for statistical assistance.
This study did not receive any specific funding.
K.K. has received research grants from Omron Healthcare and A&D Co., and is a member of Circulation Journal’s Editorial Team. The other authors report no conflicts of interest.
The institutional review boards of Jichi Medical University School of Medicine and the other 4 participating hospitals approved this study (疫12-52).
The deidentified participant data will not be shared.
Please find supplementary file(s);
http://dx.doi.org/10.1253/circj.CJ-20-0972
