The study included consecutive patients with stable HF (more than 1 month without decompensation, clinical worsening, or treatment changes), who were enrolled if their physician considered their treatment optimized according to definitions in clinical practice guidelines.14 Based on the LVEF values, estimated using the biplane Simpson method, patients were classified as follows: HF with reduced ejection fraction (HFrEF) when LVEF was ≤ 40%, mid-range (HFmrEF) when 41% to 49%, and preserved (HFpEF) when ≥ 50%. The most recent LVEF value at the time of the study, always within the previous year, was taken as the reference. We excluded patients younger than 18 years, those with changes in their clinical status in the previous month, those requiring hypotensive medication adjustments in the medical visit, and those refusing to participate or sign the informed consent form.

Drugs used for BP control without an HF indication were considered other antihypertensives or additional antihypertensives.

Office BP measurement was performed with a validated semiautomated oscillometric device (OMRON M6 Comfort), with the patient in sitting position and after a 5-minute rest. The value used in the analysis was the mean of 2 determinations obtained 2 minutes apart. The measurements mainly used a standard-diameter cuff (25-35cm for office and 24-32cm for ambulatory testing), although occasionally a larger or smaller cuff was required, according to patient size. When there was a significant difference (>10mmHg) between the 2 determinations, a third was obtained and the discordant determination was eliminated. Within the first month after enrollment, patients underwent 24-hour ABPM with an automatic oscillometric device validated for this purpose (Mobil-O-Graph and Spacelabs). The cuff was placed on the nondominant arm and BP measurements were taken every 15 to 20 min during the day and every 20 to 30minutes at night. Diurnal and nocturnal periods were delineated based on information provided by the patient. Recordings with at least 70% of valid measurements were included in the analysis.

Based on the ABPM results, the following variables were obtained: mean 24-hour systolic BP (SBP), mean 24-hour diastolic BP (DBP), mean diurnal SBP, mean diurnal DBP, mean nocturnal SBP, and mean nocturnal DBP.

Patients were classified into the following groups based on office BP values and ABPM results:

• •

  Controlled HT (or no HT if there was no previous diagnosis): office SBP <140mmHg and DBP <90mmHg, with mean 24-hour SBP <130mmHg and mean 24-hour DBP <80 mmHg on ABPM.

• •

  Uncontrolled HT (or new HT diagnosis): office SBP ≥ 140mmHg and/or DBP ≥ 90mmHg, with mean 24-hour SBP ≥ 130mmHg and/or mean 24-hour DBP ≥ 80mmHg on ABPM.

• •

  White coat HT (white coat effect in patients previously diagnosed with HT): office SBP ≥ 140mmHg and/or DBP ≥ 90mmHg, with mean 24-hour SBP <130mmHg and mean 24-hour DBP <80mmHg on ABPM.

• •

  Masked HT (or uncontrolled masked HT in patients previously diagnosed with HT): office SBP <140mmHg and DBP <90mmHg, with mean 24-hour SBP ≥ 130mmHg and/or mean 24-hour DBP ≥ 80mmHg on ABPM.

Nocturnal ABPM patterns:

• •

  Dipper: mean SBP decrease ≥ 10% at night compared with the daytime value.

• •

  Extreme dipper: mean SBP decrease ≥ 20% at night compared with the daytime value.

• •

  Nondipper: mean SBP decrease 0% to 10% at night compared with the daytime value.

• •

  Riser (or reverse dipper): mean SBP increase at night compared with the daytime value.

Most of the population included (79%) had been diagnosed with HT before the study. To facilitate the analysis and data presentation, we decided to group patients showing analogous BP patterns on ABPM: no HT together with controlled HT, newly diagnosed HT with uncontrolled HT, white coat HT with white coat effect, and masked HT with uncontrolled masked HT. We will refer to each group by the HT nomenclature. Similarly, as there were very few patients with the extreme dipper nocturnal pattern, they were placed in the dipper pattern group.

Calculation of the total HT prevalence in our population included all forms of HT. The only patients excluded from the numerator were those with no history of HT and office and ABPM values within the normal range.

Quantitative variables are expressed as the mean±standard deviation or the median [interquartile range], and qualitative variables as the absolute and relative frequency. The chi-square or Fisher exact test was used to compare qualitative variables. ANOVA or the Kruskal-Wallis test and the Student t or Mann-Whitney U test, when appropriate, were used to compare quantitative variables. Baseline characteristics were compared between the various HT patterns. The normality of the distribution was evaluated using the Kolmogorov-Smirnov test.

Multivariate logistic regression analysis was performed to assess the factors associated with the HT patterns of interest (white coat, masked, and reverse dipper HT); results are reported as odds ratios (OR) and 95% confidence intervals (95%CI). The initial model included variables with a P value <.1 on univariate analysis. Variables from the initial model were eliminated using a backward stepwise procedure with the cutoff at P <.05. Data analyses were done with STATA, version 15.0 (Stata Corp., United States).

Within the period covering August 2017 to February 2021, 360 patients were recruited for the study and 266 were analyzed (figure 1 of the supplementary data). Baseline characteristics are shown in table 1. Mean age was 71.8±12 years, 177 (66.5%) were men, and 210 (79%) had a previous HT diagnosis. Ten additional hypertensive patients (3.8%) were diagnosed during the study, giving a total HT prevalence of 82.7%. Among the 266 patients included, 123 (46.2%) had HFrEF, 60 (22.6%) HFmrEF, and 83 (31.2%) HFpEF.

Office BP measurement yielded a median value of 120/70mmHg (105/63-133/80) and a mean of 120/72±20/12mmHg. The distribution of BP values in the population is shown in figure 2 of the supplementary data.

Among the 266 patients included, 181 (68%) had controlled HT, 27 (10.2%) uncontrolled HT, 28 (10.5%) white coat HT, and 30 (11.3%) masked HT. Fifty-five patients (20.7%) had elevated office BP values, of which 28 (50.9%) were due to white coat HT. Among the 211 (79.3%) patients with normal office BP values, 30 (14.2%) had masked HT (figure 1). Patients with uncontrolled HT were more often women and this group had a higher prevalence of diabetes mellitus (DM), HT, and HFpEF. Uncontrolled HT was associated with less frequent use of sacubitril-valsartan and aldosterone antagonists, and greater prescription of additional nonspecific antihypertensive agents for HF. Patients with white coat HT were older and more often women. Those with masked HT had a higher comorbidity burden (HT, DM, dyslipidemia, chronic obstructive pulmonary disease, and peripheral arterial disease), a greater prevalence of HFpEF, more frequent prescription of additional antihypertensives, and less sacubitril-valsartan use.

Figure 1.

Diurnal BP patterns based on office and ABPM measurements. ABPM, ambulatory blood pressure monitoring; BP, blood pressure.

(0.18MB).

In patients with elevated office BP, logistic regression analysis identified the following independent factors associated with fewer cases of white coat HT: ongoing treatment with angiotensin-converting enzyme inhibitors, angiotensin II receptor antagonists, and other antihypertensive drugs (table 2). Among those with normal office BP values, DM and treatment with other antihypertensives were associated with masked HT, whereas ongoing treatment with sacubitril-valsartan was related to a lower presence of this HT form (table 3).

The most frequent nocturnal pattern was the nondipper pattern (114 patients, 42.9%), followed by the dipper pattern (83, 31.2%) and reverse dipper (69, 25.9%) (figure 2). Baseline characteristics of these groups are shown in tables 1 and 2 of the supplementary data. The median night/day SBP ratio was 0.9 (0.9-1.0) and the mean night/day was 0.95±0.1. The median night/day DBP ratio was 0.9 (0.8-1.0) and the mean night/day was 0.9±0.1. The nondipper profile was associated with poorer functional class and greater presence of HFrEF. The reverse dipper pattern was observed in older patients, with a greater presence of HT, DM, and ivabradine use. DM and chronic obstructive pulmonary disease were independently associated with a higher risk of the reverse dipper pattern, whereas treatment with aldosterone antagonists showed an inverse association (table 4). There was no association between the office BP and the nocturnal BP pattern.

Figure 2.

Nocturnal blood pressure patterns on ambulatory blood pressure monitoring.

(0.13MB).

The SBP results were significantly lower in HFrEF than in the remaining HF groups for all the measurements performed (office, 24-hour, nocturnal, and diurnal) (table 3 and figure 3 of the supplementary data)

The HT patterns differed between the various HF groups, defined according to LVEF. Patients with HFrEF showed a higher percentage of controlled HT than the HFpEF group (76.4% vs 56.6%; P <.001) and a similar percentage of white coat HT (figure 3). In contrast, these patients had a lower prevalence of uncontrolled (5.7% vs 15.7%; P=.018) and masked (7.3% vs 16.9%; P=.033) HT than patients with HFpEF.

Figure 3.

Diurnal blood pressure patterns according to heart failure type

BP, blood pressure; HFmrEF, heart failure with mid-range ejection fraction; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction.

(0.22MB).

In general, there were no differences in nocturnal patterns between the LVEF groups (P=.17), and the reverse dipper pattern showed a similar prevalence (P=.84). A nonsignificant trend to greater presence of the dipper pattern was observed in HFmrEF (P=.12) (figure 4).

Figure 4.

Nocturnal blood pressure patterns according to heart failure type. HFmrEF, heart failure with mid-range ejection fraction; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction.

(0.22MB).

Despite the value of describing the prevalence and characteristics of BP patterns in a large HF population, our study has several limitations. This is a cross-sectional, observational study and is therefore subject to the particularities of this type of analysis. The cross-sectional design reflects patient characteristics at a specific moment in time but does not allow evaluation of BP, LVEF, or treatment over the course of their disease. Furthermore, although patients with the entire spectrum of LVEF values were included, they came from the HF units of 2 hospitals with a predominance of patients hospitalized in the previous months and a greater presence of HFrEF. Therefore, these units may not represent all HF patients in our area, and the results may not be applicable to other settings. Finally, the possible prognostic impact of the various BP patterns remains to be described.