Impact of Asleep and 24-Hour Blood Pressure Data on the Prevalence of Masked Hypertension by Race/Ethnicity

Abstract

BACKGROUND

We pooled ambulatory blood pressure monitoring data from 5 US studies, including the Jackson Heart Study (JHS), the Coronary Artery Risk Development in Young Adults (CARDIA) study, the Masked Hypertension Study, the Improving the Detection of Hypertension Study, and the North Carolina Masked Hypertension Study. Using a cross-sectional study design, we estimated differences in the prevalence of masked hypertension by race/ethnicity when out-of-office blood pressure (BP) included awake, asleep, and 24-hour BP vs. awake BP alone.

METHODS

We restricted the analyses to participants with office systolic BP (SBP) <130 mm Hg and diastolic BP (DBP) <80 mm Hg. High awake BP was defined as mean SBP/DBP ≥130/80 mm Hg, high asleep BP as mean SBP/DBP ≥110/65 mm Hg, and high 24-hour BP as mean SBP/DBP ≥125/75 mm Hg.

RESULTS

Among participants not taking antihypertensive medication (n = 1,292), the prevalence of masked hypertension with out-of-office BP defined by awake BP alone or by awake, asleep, or 24-hour BP was 34.5% and 48.7%, respectively, among non-Hispanic White, 39.7% and 67.6% among non-Hispanic Black, and 19.4% and 35.1% among Hispanic participants. After multivariable adjustment, non-Hispanic Black were more likely than non-Hispanic White participants to have masked hypertension by asleep or 24-hour BP but not awake BP (adjusted odds ratio [OR] 2.14 95% confidence interval [CI] 1.45–3.15) and by asleep or 24-hour BP and awake BP (OR 1.61; 95% CI 1.12–2.32) vs. not having masked hypertension.

CONCLUSIONS

Assessing asleep and 24-hour BP measures increases the prevalence of masked hypertension more among non-Hispanic Black vs. non-Hispanic White individuals.

blood pressure, hypertension, masked hypertension

Among individuals not taking antihypertensive medication, masked hypertension refers to blood pressure (BP) levels above the hypertension threshold when measured outside of the office while having BP levels below the hypertension threshold when measured in the office setting.¹ For individuals taking antihypertensive medication, this BP phenotype is called masked uncontrolled hypertension. Masked hypertension and masked uncontrolled hypertension are associated with subclinical cardiovascular disease (CVD), including left ventricular hypertrophy, and 2 times higher risk for CVD events compared with BP below the hypertension threshold both inside and outside of the office setting, a phenotype called sustained normotension for individuals not taking antihypertensive medication and sustained controlled BP for individuals taking antihypertensive medication.^2–4

In 2017, the American College of Cardiology (ACC)/American Heart Association (AHA) BP guideline recommended measuring BP outside of the office setting to screen for masked hypertension and masked uncontrolled hypertension.⁵ This guideline recommends using BP measurements while awake to assess for the presence of out-of-office hypertension. Ambulatory blood pressure monitoring (ABPM) is typically performed over a 24-hour period, including when people are asleep, and the current European guidelines recommend using awake, asleep, or 24-hour BP measurements to assess for the presence of out-of-office hypertension.⁶ Many adults without high awake BP have high asleep BP on ABPM.⁷ Including asleep BP in the definition of masked hypertension and masked uncontrolled hypertension may be particularly relevant for Black adults, a group with higher asleep BP compared with other racial/ethnic groups in the United States.^8,9

The goal of the present study was to determine the difference in the prevalence of masked hypertension and masked uncontrolled hypertension when out-of-office BP is defined using awake, asleep, and 24-hour BP vs. awake BP alone. In the primary analysis, we used BP thresholds to define high office, awake, asleep, and 24-hour BP from the 2017 ACC/AHA BP guideline.⁵ In a secondary analysis, we used BP thresholds from the 2018 European Society of Cardiology (ESC)/European Society of Hypertension (ESH) arterial hypertension guidelines.⁶ To conduct these analyses, we pooled data from 5 population-, community-, and practice-based studies conducted in the United States.

METHODS

Study participants

Details of the study design and methods of each cohort have been described previously and are provided in eMethods in the Supplement online.^10–16 Briefly, the Jackson Heart Study (JHS) is a community-based prospective cohort study designed to identify the reasons for the higher prevalence of CVD among Black adults and to find approaches for reducing this health disparity.¹¹ Between 2000 and 2004, JHS enrolled 5,306 noninstitutionalized Black adults aged ≥21 years from the Jackson, Mississippi metropolitan area. The current analysis was restricted to 930 JHS participants who had a complete (defined below) ABPM recording as part of their baseline study visit.¹⁰ The Coronary Artery Risk Development in Young Adults (CARDIA) study enrolled 5,115 Black and White adults, 18–30 years of age, at 4 field centers in 1985–1986.¹² Serial follow-up examinations were conducted 2, 5, 7, 10, 15, 20, 25, and 30 years after the Year 0 (baseline) examination. ABPM was conducted in a CARDIA ancillary study at the Year 30 examination and 740 participants had a complete recording.¹³ The Masked Hypertension Study was conducted at Stony Brook University and Columbia University in New York. Employees of these 2 universities, their medical schools and affiliated hospitals, and a private hedge fund management organization were enrolled and completed ABPM for this study.¹⁴ Between 2005 and 2012, 1,011 participants were enrolled in the study, and 731 had a complete ABPM recording. The Improving the Detection of Hypertension Study is a community-based study designed to compare the cost-effectiveness of different strategies to diagnose hypertension outside of the office. Between 2011 and 2013, 408 participants were enrolled and 376 had a complete ABPM recording.¹⁵ The North Carolina Masked Hypertension Study is a practice-based study that enrolled 420 adults from primary care clinics in North Carolina.¹⁶ Between 2012 and 2014, 272 participants had a complete ABPM recording. In each study, race/ethnicity was determined by self-report.

The institutional review boards of each study site approved the study protocol, and all participants gave written informed consent. The study protocol for the analysis of data from the 5 studies was approved by the institutional review boards at Duke University and the University of Alabama at Birmingham.

Data collection

A detailed description of data and specimen collection, and specimen processing for each study, is included in eMethods in the Supplement online. Data were collected using standardized protocols in accordance with the quality control procedures of each study (Supplementary Table S1 online).

Office and ABPM BP measurements

The assessment of office BP and ABPM in each study followed standardized protocols, which are summarized in Supplementary Table S2 online. For office BP measurements, trained research staff measured BP at least 2 times in participants’ right-arm (nondominant arm for in the Masked Hypertension Study and the Improving the Detection of Hypertension Study) brachial artery at ≥1-minute intervals after the participant had been seated in a quiet room for 5 minutes on a single occasion in the JHS and CARDIA and on multiple separate occasions in the other studies. The definitions of office BP within each cohort are summarized in Supplementary Table S2 online.

In all studies, an ABPM device was fitted on participants’ nondominant arms. ABPM devices were validated according to British Hypertension Society criteria,¹⁷ and calibrated/recalibrated prior to data collection. We defined awake and asleep periods during ABPM based on actigraphy for CARDIA, MHT, and IDH participants with self-report times used for JHS and NCMH study participants as actigraphy data were not collected in these studies. We have previously published data showing high agreement between awake and asleep BP when time that people were awake and asleep is defined using self-report and actigraphy.¹⁸ In the CARDIA, MHT, and IDH, 24-hour BP was calculated as a weighted average of the awake and asleep averages, with weight for asleep (WT_sleep) being the proportion of the 24 hours that the participant was asleep and the weight for awake being “1 − WT_sleep”. In the JHS and NCMH, we defined 24-hour BP as the average of all readings over the entire ABPM recording period. Participants were considered to have a complete ABPM recording if they had at least 70% of the total number planned readings over 24 hours, and ≥20 awake and ≥7 asleep systolic BP (SBP) and diastolic BP (DBP) measurements.¹⁹

For this analysis, we excluded participants who had high BP based on their office measurements because, by definition, they could not have masked hypertension. The remainder of the participants were categorized into 4 mutually exclusive groups:

• (1) participants without masked hypertension or masked uncontrolled hypertension (i.e., not having high awake, asleep, or 24-hour BP, consistent with sustained normotension or sustained controlled BP);

• (2) participants who had high awake BP and did not have high asleep or 24-hour BP;

• (3) participants who had high asleep or 24-hour BP and did not have high awake BP; and

• (4) participants who had high asleep or 24-hour BP and high awake BP.

In the primary analysis, we used the criteria in the 2017 ACC/AHA BP guideline to define high BP: office SBP/DBP ≥130/80 mm Hg; mean awake SBP/DBP ≥130/80 mm Hg; mean asleep SBP/DBP ≥110/65 mm Hg; and mean 24-hour SBP/DBP ≥125/75 mm Hg (Table 1, left panel).⁵ In the secondary analysis, we used criteria from the 2018 ESC/ESH arterial hypertension guidelines to define high BP: office SBP/DBP ≥140/90 mm Hg; mean awake SBP/DBP ≥135/85 mm Hg; mean asleep SBP/DBP ≥120/70 mm Hg; and mean 24-hour SBP/DBP ≥130/80 mm Hg (Table 1, right panel).⁹

Table 1.

Definitions of masked hypertension and masked uncontrolled hypertension

	2017 ACC/AHA BP guideline (primary analysis)		The 2018 ESC/ESH arterial hypertension guidelines (secondary analysis)
	Masked hypertension defined using awake BP only	Masked hypertension defined using awake, asleep, or 24-hour BP	Masked hypertension defined using awake BP only	Masked hypertension defined using awake, asleep, or 24-hour BP
Office-measured BP	SBP <130 mm Hg and DBP <80 mm Hg	SBP <130 mm Hg and DBP <80 mm Hg	SBP <140 mm Hg and BP <90 mm Hg	SBP <140 mm Hg and DBP <90 mm Hg
	AND all of the 3 below conditions	AND any of the 3 below conditions	AND all of the 3 below conditions	AND any of the 3 below conditions
Awake BP	SBP ≥130 mm Hg or DBP ≥80 mm Hg	SBP ≥130 mm Hg or DBP ≥80 mm Hg	SBP ≥135 mm Hg or DBP ≥85 mm Hg	SBP ≥135 mm Hg or DBP ≥85 mm Hg
	AND	OR	AND	OR
Asleep BP	SBP <110 mm Hg and DBP <65 mm Hg	SBP ≥110 mm Hg or DBP ≥65 mm Hg	SBP <120 mm Hg and DBP <70 mm Hg	SBP ≥120 mm Hg or DBP ≥70 mm Hg
	AND	OR	AND	OR
24-Hour BP	SBP <125 mm Hg and DBP <75 mm Hg	SBP ≥125 mm Hg or DBP ≥75 mm Hg	SBP <130 mm Hg and DBP <80 mm Hg	SBP ≥130 mm Hg or DBP ≥80 mm Hg

Abbreviations: ACC, American College of Cardiology; AHA, American Heart Association; BP, blood pressure; DBP, diastolic blood pressure; ESC, European Society of Cardiology; ESH, European Society of Hypertension; N/A, not applicable; SBP, systolic blood pressure. Masked hypertension and masked uncontrolled hypertension refer to individuals not taking and taking antihypertensive medication, respectively. These phenotypes use the same blood pressure thresholds.

Statistical analyses

Summary statistics for characteristics of participants, overall and by study cohort were calculated. We used pooled data from the 5 studies for the remainder of the analyses. The prevalence of masked uncontrolled hypertension has been reported to be higher than the prevalence of masked hypertension.²⁰ Therefore, all analyses were performed for participants taking and not taking antihypertensive medication, separately. Below, we describe the approach for participants not taking antihypertensive medication. Identical analyses were done for participants taking antihypertensive medication. All missing values were imputed with 10 datasets using chained equations (Supplementary Table S3 online).²¹ We calculated the prevalence and 95% confidence interval (CI) of masked hypertension using awake BP and awake, asleep, or 24-hour BP for the overall population and for non-Hispanic White, non-Hispanic Black, Hispanic individuals, and participants of other race/ethnicity, separately. To determine the statistical significance of differences in the prevalence of masked hypertension with out-of-office BP defined using awake, asleep, and 24-hour BP compared with using awake BP only by race/ethnicity, we used chi-square tests.

We used multinomial logistic regression models with those who had sustained normotension as the reference group to test whether the prevalence of masked hypertension differed among across race/ethnic groups when out-of-office BP was defined by (1) high awake BP without high asleep or 24-hour BP, (2) high asleep or 24-hour BP without high awake BP, and (3) high asleep or 24-hour BP and high awake BP after multivariable adjustment. Covariates in the regression models included age, sex, less than high school education, current smoking, alcohol intake, total and high-density lipoprotein cholesterol, body mass index, diabetes, glucose-lowering medication use, history of CVD, estimated glomerular filtration rate, urinary albumin-to-creatinine ratio, and office-measured SBP, DBP, and heart rate. We used total and high-density lipoprotein cholesterol, body mass index, estimated glomerular filtration rate, urinary albumin-to-creatinine ratio for adjustments as continuous variables. Covariates were selected a priori because they may differ by race/ethnicity and may be associated with BP.^5,22,23 We conducted 2 sensitivity analyses. First, since the JHS enrolled only non-Hispanic Black participants, we conducted the multinomial logistic regression models excluding this cohort. Second, we conducted the multinomial logistic regression models within each study, and pooled the results using standard meta-analysis methods (i.e., random effects). Statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC) and Stata version 16.1 (StataCorp LLC, College Station, TX). Statistical significance was defined by a 2-sided P value <0.05.

RESULTS

Results using BP thresholds in the 2017 ACC/AHA BP guideline

We included 480 JHS participants, 467 CARDIA study participants, 479 MHT participants, 213 IDH study participants, and 85 NCMHT study participants who had office BP <130/80 mm Hg (total sample size: n = 1,724; Supplementary Figures S1 and S2 and Table S4 online). The mean age of these participants was 50.2 (SD 12.1) years, 65.9% were women, and 38.3% reported being non-Hispanic White, 46.4% non-Hispanic Black, and 11.1% Hispanic. Demographic and clinical characteristics of the included participants are shown stratified by antihypertensive medication use in Table 2 and by race/ethnicity for those not taking and taking antihypertensive in Supplementary Tables S5 and S6 online, respectively.

Table 2.

Characteristics of participants taking and not taking antihypertensive medication

Characteristic	Antihypertensive medication use
	No (n = 1,292)	Yes (n = 432)
Age, years	47.2 (11.7)	59.0 (8.4)
Age category, %
<40	26.4	1.2
40–59	61.8	56.5
≥60	11.8	42.4
Male, %	35.9	28.7
Race/ethnicity, %
Non-Hispanic White	48.0	9.3
Non-Hispanic Black	31.6	90.7
Hispanic	14.8	0.0
Other	5.7	0.0
Less than high school, %	4.2	13.3
Current smoking, %	9.5	10.0
Alcohol intake, %
No alcohol	44.1	67.1
Light or moderate	48.3	28.9
Heavy	7.7	4.0
Body mass index, kg/m2	27.5 (5.6)	32.8 (6.9)
Total cholesterol, mg/dl	191.2 (37.3)	192.3 (42.2)
HDL cholesterol, mg/dl	58.7 (16.8)	55.4 (17.2)
Diabetes, %	6.2	36.1
Family history of hypertension, %	64.4	91.2
Glucose-lowering medication use, %	3.0	29.6
Lipid-lowering medication use, %	5.8	32.7
History of CVD, %	0.9	12.7
eGFR <60, ml/min/1.73 m2	7.4	8.6
ACR >30, mg/g	4.6	16.6
Systolic blood pressure, mm Hg
Office	111.4 (9.4)	115.7 (8.8)
Awake	121.6 (10.6)	125.6 (12.0)
Asleep	106.5 (11.5)	115.2 (14.6)
24-Hour	117.1 (10.4)	121.7 (12.0)
Diastolic blood pressure, mm Hg
Office	70.3 (6.3)	69.2 (6.5)
Awake	75.9 (7.0)	75.9 (8.8)
Asleep	62.2 (7.6)	65.9 (9.0)
24-Hour	71.9 (7.0)	71.8 (6.7)
Office heart rate, beats/min	67.8 (10.8)	68.5 (10.7)

Data are expressed as means (SD) or percentages. Participants in this table had an office systolic blood pressure <130 mm Hg and office diastolic blood pressure <80 mm Hg. Abbreviations: ACC, American College of Cardiology; ACR, urinary albumin-to-creatinine ratio; AHA, American Heart Association; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate; HDL, high-density lipoprotein; LDL, low-density lipoprotein.

The increase in the prevalence of masked hypertension and masked uncontrolled hypertension with out-of-office BP defined using awake, asleep, or 24-hour BP vs. awake BP only was larger for non-Hispanic Black compared with participants of other race/ethnic groups (Figure 1). Among untreated participants, less than 5% of each race/ethnicity group had high awake BP but did not have high asleep or 24-hour BP (Table 3). In contrast, 14.2%, 27.9%, 15.7%, and 12.3% of non-Hispanic White, non-Hispanic Black, Hispanic individuals, and participants of other race/ethnicity, respectively, had high asleep or 24-hour BP and did not have high awake BP. Among participants taking antihypertensive medication, 7.5% of non-Hispanic White and 2.6% of non-Hispanic Black individuals had high awake BP and did not have high asleep or 24-hour BP. Also, 17.5% of non-Hispanic White and 33.9% of non-Hispanic Black individuals had high asleep or 24-hour BP and did not have high awake BP. Among participants not taking and taking antihypertensive medication, over 35% of non-Hispanic White and non-Hispanic Black individuals had high asleep or 24-hour BP and high awake BP. The results for each study are shown in Supplementary Table S7 online.

Table 3.

Race/ethnicity-specific prevalence of masked hypertension and masked uncontrolled hypertension using awake, asleep, and 24-hour blood pressure and thresholds in the 2017 ACC/AHA blood pressure guideline

	Antihypertensive medication use
	No					Yes
	Non-Hispanic White (n = 620)	Non-Hispanic Black (n = 408)	Hispanic (n = 191)	Other (n = 73)	P value	Non-Hispanic White (n = 40)	Non-Hispanic Black (n = 392)	Hispanic (n = 0)	Other (n = 0)	P value
	Prevalence (95% confidence interval)					Prevalence (95% confidence interval)
Masked hypertension by high awake, asleep, or 24-hour BP	48.7 (44.8, 52.6)	67.6 (63.1, 72.2)	35.1 (28.3, 41.9)	37.0 (25.9, 48.1)	<0.001	62.5 (47.5, 77.5)	76.8 (72.6, 81.0)	—	—	0.050
Masked hypertension by high awake BPa	34.5 (30.8, 38.3)	39.7 (35.0, 44.5)	19.4 (13.8, 25.0)	24.7 (14.8, 34.5)	<0.001	45.0 (29.6, 60.4)	42.9 (38.0, 47.8)	—	—	0.794
Four mutually exclusive groups	Prevalence (95% confidence interval)					Prevalence (95% confidence interval)
No masked hypertension	51.3 (47.4, 55.2)	32.4 (27.8, 36.9)	64.9 (58.2, 71.7)	63.0 (51.9, 74.1)	<0.001	37.5 (22.5, 52.5)	23.2 (19.0, 27.4)	—	—	0.028
Masked hypertension by high awake BP only	4.0 (2.5, 5.6)	2.7 (1.1, 4.3)	1.6 (0.0, 3.3)	0.0 (0.0, 0.0)		7.5 (0.0, 15.7)	2.6 (0.8, 4.1)	—	—
Masked hypertension by high asleep or 24-hour BP but non-high awake BP	14.2 (11.5, 16.9)	27.9 (23.6, 32.3)	15.7 (10.6, 20.8)	12.3 (4.8, 19.9)		17.5 (5.7, 29.3)	33.9 (29.2, 38.6)	-	-
Masked hypertension by high awake BP and high asleep or 24-hour BP	30.5 (26.9, 34.1)	37.0 (32.3, 41.7)	17.8 (12.4, 23.2)	24.7 (14.8, 34.5)		37.5 (22.5, 52.5)	40.3 (35.5, 45.2)	—	—

Numbers in table are within-race/ethnic group percentages (95% confidence interval). Masked hypertension refers to participants not taking antihypertensive medication and masked uncontrolled hypertension refers to participants taking antihypertensive medication. Abbreviations: ABPM, ambulatory blood pressure; ACC, American College of Cardiology; AHA, American Heart Association; BP, blood pressure; DBP, diastolic blood pressure; SBP, systolic blood pressure.

^aMasked hypertension was defined by hypertensive awake BP, ignoring asleep and 24-hour BP. In all cohorts, we excluded participants who had office SBP/DBP ≥130/80 mm Hg. High awake BP was defined as mean SBP ≥130 mm Hg or DBP ≥80 mm Hg based on all awake readings. High asleep BP as mean SBP ≥110 mm Hg or DBP ≥65 mm Hg based on all asleep readings. High 24-hour BP as mean SBP ≥125 mm Hg or DBP ≥75 mm Hg over the entire ABPM period. There were only non-Hispanic White and non-Hispanic Black among participants taking antihypertensive medication.

Figure 1.

Prevalence of masked hypertension (panel a) and masked uncontrolled hypertension (panel b) by race/ethnicity among participants with office systolic blood pressure <130 mm Hg and office diastolic blood pressure <80 mm Hg. Each bar in panel (a) shows the prevalence of masked hypertension and in panel (b) shows the prevalence of masked uncontrolled hypertension. There were only non-Hispanic White and non-Hispanic Black among participants taking antihypertensive medication.

Among participants not taking antihypertensive medication, after multivariable adjustment and compared with participants without masked hypertension, non-Hispanic Black were more likely than non-Hispanic White individuals to have masked hypertension by asleep or 24-hour BP but not awake BP (adjusted odds ratio [OR] 2.14; 95% CI 1.45–3.15) and by awake and asleep or 24-hour BP (adjusted OR 1.61; 95% CI 1.12–2.32) (Table 4). Hispanic individuals and participants of other race/ethnicity did not differ significantly from non-Hispanic White individuals in the likelihood of having any of the masked hypertension phenotypes, though it appears that Hispanic individuals were less likely than non-Hispanic White individual to have high awake BP. Among participants taking antihypertensive medication and after multivariable adjustment, non-Hispanic Black were more likely than non-Hispanic White individuals to have masked hypertension by asleep or 24-hour BP but not awake BP (adjusted OR 3.50; 95% CI 1.25–9.77) and by awake and asleep or 24-hour BP (adjusted OR 1.91; 95% CI 0.77–4.75) vs. not having masked hypertension. Results for the analysis pooling ORs from the individual studies using meta-analytic techniques are shown in Supplementary Figures S3–S8 online. In the multinomial logistic regression models excluding the JHS participants, the ORs comparing Blacks to Whites for masked hypertension with out-of-office BP defined by asleep or 24-hour BP but not high awake BP and by high asleep or 24-hour BP and high awake BP were 1.78 (95% CI 1.09–2.93) and 1.92 (95% CI 1.23–3.00), respectively (Supplementary Table S8 online).

Table 4.

Adjusted odds ratios for masked hypertension (top panel) and masked uncontrolled hypertension (bottom panel) defined using blood pressure thresholds in the 2017 ACC/AHA blood pressure guideline

	Race–ethnicity
	Non-Hispanic White	Non-Hispanic Black	Hispanic	Other
Not taking antihypertensive medication
No masked hypertension	1 (ref)	1 (ref)	1 (ref)	1 (ref)
Masked hypertension by high awake BP only	1 (ref)	0.97 (0.44, 2.11)	0.33 (0.09, 1.01)	—
Masked hypertension by high asleep or 24-hour BP but non-high awake BP	1 (ref)	2.14 (1.45, 3.15)	1.01 (0.59, 1.70)	0.81 (0.36, 1.82)
Masked hypertension by high awake BP and high asleep or 24-hour BP	1 (ref)	1.61 (1.12, 2.32)	0.60 (0.36, 1.01)	0.93 (0.46, 1.88)
Taking antihypertensive medication
No masked uncontrolled hypertension	1 (ref)	1 (ref)	1 (ref)	1 (ref)
Masked uncontrolled hypertension by high awake BP only	1 (ref)	0.89 (0.14, 5.56)	—	—
Masked uncontrolled hypertension by high asleep or 24-hour BP but non-high awake BP	1 (ref)	3.50 (1.25, 9.77)	—	—
Masked uncontrolled hypertension by high awake BP and high asleep or 24-hour BP	1 (ref)	1.91 (0.77, 4.75)	—	—

Numbers in table are odds ratios (95% confidence interval). We excluded participants who had office SBP/DBP ≥130/80 mm Hg and participants who did not have a complete ABPM recording or had missing for antihypertensive medication use. High awake BP was defined as SBP ≥130 mm Hg or DBP ≥80 mm Hg while awake, high asleep BP as SBP ≥110 mm Hg or DBP ≥65 mm Hg while asleep and high 24-hour BP as SBP ≥125 mm Hg or DBP ≥75 mm Hg over the entire ABPM period. Adjusted odds ratios (95% CIs) for having masked hypertension or masked uncontrolled hypertension by race/ethnicity (reference: non-Hispanic White) defined using awake, asleep, or 24-hour BP vs. awake BP only were shown. Odds ratios were adjusted for age, sex, education less than high school, current smoking, alcohol intake, total and high-density lipoprotein cholesterol, body mass index, prevalent diabetes, glucose-lowering medication use, history of cardiovascular disease, estimated glomerular filtration rate, urinary albumin-to-creatinine ratio, and office-measured systolic BP, diastolic BP, and heart rate. Participants taking antihypertensive medication included non-Hispanic Whites and non-Hispanic Blacks only. Abbreviations: ABPM, ambulatory blood pressure monitoring; ACC, American College of Cardiology; AHA, American Heart Association; BP, blood pressure; CIs, confidence intervals; DBP, diastolic blood pressure; SBP, systolic blood pressure.

Results using BP thresholds in the 2018 ESC/ESH guidelines

We included 2,562 participants in the analysis using BP thresholds from the 2018 ESC/ESH guidelines to define masked hypertension (Supplementary Figures S9 and 10 online). The increase in the prevalence of masked hypertension and masked uncontrolled hypertension with out-of-office BP defined using awake, asleep, or 24-hour BP vs. awake BP only was larger for non-Hispanic Black compared with non-Hispanic White, Hispanics individuals, and participants of other race/ethnicity (Supplementary Table S9 and Figure S11 online). Among untreated participants and after multivariable adjustment, non-Hispanic Black were more likely than non-Hispanic White individuals to have masked hypertension when out-of-office BP was defined by high asleep or 24-hour BP but not high awake BP (adjusted OR 2.65; 95% CI 1.83–3.84) and by high awake and high asleep or 24-hour BP (adjusted OR 1.85; 95% CI 1.33–2.57) vs. not having masked hypertension (Table 5). Among participants taking antihypertensive medication and after multivariable adjustment, non-Hispanic Black individuals were less likely than non-Hispanic White individuals to have masked uncontrolled hypertension by high awake BP only vs. no masked hypertension (adjusted OR 0.29; 95% CI 0.07–0.79) and more likely, though not significantly so, to have masked uncontrolled hypertension by high asleep or 24-hour but not high awake BP (adjusted OR 1.89; 95% CI 0.76–4.71) and by high awake, asleep, or 24-hour BP (adjusted OR 1.57; 95% CI 0.72–3.41).

Table 5.

Adjusted odds ratios of masked hypertension and masked uncontrolled hypertension defined using blood pressure thresholds in the 2018 ESC/ESH arterial hypertension guidelines

	Antihypertensive medication use
	No				Yes
	Non-Hispanic White (n = 887)	Non-Hispanic Black (n = 603)	Hispanic (n = 300)	Other (n = 103)	Non-Hispanic White (n = 57)	Non-Hispanic Black (n = 612)	Hispanic (n = 0)	Other (n = 0)
Four mutually exclusive groups	Odds ratio (95% confidence interval)				Odds ratio (95% confidence interval)
No masked hypertension	1 (ref)	1 (ref)	1 (ref)	1 (ref)	1 (ref)	1 (ref)	1 (ref)	1 (ref)
Masked hypertension by high awake BP only	1 (ref)	1.35 (0.66, 2.77)	1.29 (0.52, 3.20)	0.46 (0.06, 3.65)	1 (ref)	0.29 (0.07, 0.79)	—	—
Masked hypertension by high asleep or 24-hour BP but non-high awake BP	1 (ref)	2.65 (1.83, 3.84)	1.00 (0.58, 1.72)	1.02 (0.46, 2.26)	1 (ref)	1.89 (0.76, 4.71)	—	—
Masked hypertension by high awake BP and high asleep or 24-hour BP	1 (ref)	1.85 (1.33, 2.57)	1.00 (0.63, 1.58)	0.58 (0.27, 1.26)	1 (ref)	1.57 (0.72, 3.41)	—	—

Numbers in table are odds ratios (95% confidence interval). We excluded participants who had office SBP/DBP ≥140/90 mm Hg and participants who did not have a complete ABPM recording or had missing for antihypertensive medication use. Hypertensive awake BP was defined as SBP ≥135 mm Hg or DBP ≥85 mm Hg while awake, hypertensive asleep BP as SBP ≥120 mm Hg or DBP ≥70 mm Hg while asleep and hypertensive 24-hour BP as SBP ≥130 mm Hg or DBP ≥80 mm Hg over the entire ABPM period. Adjusted odds ratios (95% CIs) for having masked hypertension by race/ethnicity (reference: non-Hispanic White) defined using awake, asleep, or 24-hour BP vs. awake BP only were shown. Odds ratios were adjusted for age, sex, education less than high school, current smoking, alcohol intake, total and high-density lipoprotein cholesterol, body mass index, prevalent diabetes, glucose-lowering medication use, history of cardiovascular disease, estimated glomerular filtration rate, albumin-to-creatinine ratio, and office-measured SBP, DBP, and heart rate. Participants taking antihypertensive medication included non-Hispanic Whites and non-Hispanic Blacks only. Abbreviations: ABPM, ambulatory blood pressure monitoring; ACC, American College of Cardiology; AHA, American Heart Association; BP, blood pressure; CIs, confidence intervals; DBP, diastolic blood pressure; ESC, European Society of Cardiology; ESH, European Society of Hypertension; SBP, systolic blood pressure.

Discussion

In the current pooled analysis of individual-level data from 5 population-, community-, and practice-based studies in the United States, the prevalence of masked hypertension and masked uncontrolled hypertension was substantially higher when using awake, asleep, and 24-hour BP to define out-of-office hypertension compared with using awake BP only. The differences noted in the prevalence of masked hypertension were larger among non-Hispanic Black individuals compared with other race/ethnic groups and the difference in the prevalence of masked uncontrolled hypertension was larger among non-Hispanic Black compared with non-Hispanic White individuals.

In the International Database of Ambulatory Blood Pressure in Relation to Cardiovascular Outcome (IDACO) study, which includes Asian, European, and South American community-based studies, the estimated prevalence of masked hypertension defined using the 2018 ESC/ESH guidelines was 13.4 % when out-of-office BP was defined as high awake BP without using asleep and 24-hour BP, whereas it was 19.6% when masked hypertension was defined using high awake, asleep, or 24-hour BP.⁴ However, the difference in the prevalence of masked hypertension when defined using awake, asleep, or 24-hour BP vs. awake BP alone among non-Hispanic Black and Hispanic individuals was not reported in the IDACO study since US adults were not included. Furthermore, the IDACO study defined high awake, asleep, and 24-hour BP using thresholds specified in the 2018 ESC/ESH guidelines. In the Spanish ABPM Registry, de la Sierra et al.,²⁴ reported the change in the prevalence of masked hypertension and masked uncontrolled hypertension when defined using the 2018 ESC/ESH arterial hypertension guidelines vs. the ACC/AHA BP guideline.² However, the study focused on changes in the prevalence of masked hypertension and masked uncontrolled hypertension according to the BP thresholds for hypertension while awake, asleep, and a 24-hour period among Hispanic participants. The current study extends prior knowledge by demonstrating that the increase in the prevalence of masked hypertension when out-of-office hypertension was defined as high awake, asleep, or 24-hour BP vs. high awake BP only may be larger among non-Hispanic Black compared with non-Hispanic White individuals.

Among untreated participants, less than 5% of each race/ethnicity group had high awake BP but did not have high asleep or 24-hour BP. Among participants taking antihypertensive medication, less than 8% of non-Hispanic White and Black individuals had high awake BP and did not have high asleep or 24-hour BP. The findings of masked hypertension or masked uncontrolled hypertension defined by awake BP only should be interpreted with caution because they are based on subgroups with small number of participants. The ORs for masked hypertension or masked uncontrolled hypertension defined by high asleep or 24-hour BP but non-high awake BP, which thresholds were based on the 2017 ACC/AHA blood pressure guideline or the 2018 ESC/ESH arterial hypertension guidelines, were 2 to 3 times higher for non-Hispanic Black vs. non-Hispanic White individuals. However, the ORs for masked hypertension defined by high awake BP only, which thresholds were based on the 2017 ACC/AHA blood pressure guideline, were similar for non-Hispanic Black and for Hispanic vs. non-Hispanic White individuals. Conversely, the ORs for masked uncontrolled hypertension defined by high awake BP only, which thresholds were based on the 2018 ESC/ESH arterial hypertension guidelines, were lower for non-Hispanic Black vs. non-Hispanic White individuals. However, the sample size of non-Hispanic White individuals with masked uncontrolled hypertension defined by high awake BP only was small (n = 3). Therefore, it is difficult to distinguish whether such discrepant results among treated vs. untreated individuals were a real effect or a random variation.

Both awake and asleep BP measurements may be important for assessing CVD risk among Black individuals. In a prior analysis of the JHS, one-third of participants without high awake BP had high asleep BP.²⁵ In the JHS, non-Hispanic Black participants with asleep SBP ≥110 mm Hg or DBP ≥65 mm Hg on ABPM had 2 times higher risk of CVD events, after adjustment for office-measured SBP and DBP.¹⁰ There is also a strong association between high asleep BP and an increased risk for CVD events, independent of awake BP.^10,26,27 These data suggest that using only awake BP to identify masked hypertension or masked uncontrolled hypertension may result in a failure to detect a high proportion of adults, especially non-Hispanic Black individuals, with higher CVD risk. The 2017 ACC/AHA BP guideline recommends self-measurement of BP using home devices (referred as to home BP monitoring) to evaluate BP outside of the clinic if ABPM is not available. The current study suggests that home BP monitoring may not be sufficient for identifying masked hypertension since it is generally not used to measure BP during sleep. Ultimately the goal is to eliminate racial disparities in hypertension. To achieve this, a multistep process is required, including identifying the mechanisms underlying higher asleep BP in non-Hispanic Black individuals and then developing interventions to address these.

Strengths of this study include the analysis of data from 5 well-characterized, community-based cohorts and the use of standardized data collection protocols. The studies had a substantial number of several race/ethnic groups facilitating the investigation of differences in masked hypertension and masked uncontrolled hypertension. Additionally, the large sample size in this pooled dataset allowed us to estimate the prevalence of masked hypertension and masked uncontrolled hypertension, separately. There are several potential limitations to the current analysis. Office BP was obtained on a single occasion, which may have resulted in the misclassification of high office BP. Study protocols were cohort specific. For example, the North Carolina Masked Hypertension Study used a different ABPM device from that used in the other 4 studies, and the intervals between BP readings on ABPM varied from 20 to 60 minutes across cohorts. Characteristics of participants differed by cohorts. Therefore, we did an analysis where the statistics were estimated for each study separately with the results pooled using standard meta-analysis methods. However, as indicated by the I-square statistic, the estimated effects varied across studies. Therefore, the results should be interpreted with caution. Furthermore, unmeasured characteristics for environmental exposures may represent confounders that we could not account for. Finally, self-reported race/ethnicity does not address differences in social determinants of health and aspects of structural racism that may underlie some of the differences in measured BP described here, and is not as accurate as direct assessment of individual genomic information to determine ancestry admixture.

The prevalence of masked hypertension and masked uncontrolled hypertension was each substantially higher when defined using awake, asleep, or 24-hour BP as measures of out-of-office BP compared with using awake BP only. The increase in prevalence was larger for non-Hispanic Blacks compared with non-Hispanic White individuals. Including asleep BP when defining masked hypertension can help identify individuals who have an increased CVD risk, especially among non-Hispanic Black individuals. Our findings may lead to updated guideline recommendations for the definition of masked hypertension, which may reduce the public health burden of CVD in the United States.

FUNDING

This study is supported by R01 HL144773-01 (PI: Yuichiro Yano and Paul Muntner) from the National Heart, Lung, and Blood Institute (NHLBI). CARDIA is conducted and supported by the NHLBI in collaboration with the University of Alabama at Birmingham (HHSN268201300025C and HHSN268201300026C), Northwestern University (HHSN268201300027C), University of Minnesota (HHSN268201300028C), Kaiser Foundation Research Institute (HHSN268201300029C), and Johns Hopkins University School of Medicine (HHSN268200900041C). CARDIA is also partially supported by the Intramural Research Program of the National Institute on Aging (NIA) and an intra-agency agreement between NIA and NHLBI (AG0005). The Jackson Heart Study (JHS) is supported and conducted in collaboration with Jackson State University (HHSN268201800013I), Tougaloo College (HHSN268201800014I), the Mississippi State Department of Health (HHSN268201800015I), and the University of Mississippi Medical Center (HHSN268201800010I, HHSN268201800011I, and HHSN268201800012I) contracts from the National Heart, Lung, and Blood Institute (NHLBI) and the National Institute on Minority Health and Health Disparities (NIMHD). This work was also supported by the NIH (R01 HL117323 and K24 HL125704) from the NHLBI, Bethesda, MD. MHT and IDH were both supported by NHLBI grant P01-HL047540 (PI: Joseph Schwartz). NCMT was funded by grant R01 HL098604 from the NHLBI with additional support provided by ULI RR025747 from the National Institutes of Health. This manuscript has been reviewed by CARDIA and JHS for scientific content. Paul Muntner received support through grant 15SFRN2390002 from the American Heart Association. Daichi Shimbo received support through grant K24-HL125704 from NHLBI. The views expressed in this manuscript are those of the authors and do not necessarily represent the views of the NHLBI; the National Institutes of Health; or the U.S. Department of Health and Human Services. NHLBI had no role in each of the following: design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

DISCLOSURE

The authors declared no conflict of interest.

This manuscript was sent to Guest Editor, Hillel W. Cohen, MPH, DrPH for editorial handling and final disposition.

DATA AVAILABILITY

Drs Yano, Muntner and Mr Poudel had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.