Thresholds for Ambulatory Blood Pressure Among African ......Aug 10, 2017 · 2470 June 20/27, 2017...

June 20/27, 2017 Circulation. 2017;135:2470–2480. DOI: 10.1161/CIRCULATIONAHA.116.0270512470

ORIGINAL RESEARCH ARTICLE

Editorial, see p 2481

BACKGROUND: Ambulatory blood pressure (BP) monitoring is the reference standard for out-of-clinic BP measurement. Thresholds for identifying ambulatory hypertension (daytime systolic BP [SBP]/diastolic BP [DBP] ≥135/85 mm Hg, 24-hour SBP/DBP ≥130/80 mm Hg, and nighttime SBP/DBP ≥120/70 mm Hg) have been derived from European, Asian, and South American populations. We determined BP thresholds for ambulatory hypertension in a US population-based sample of African American adults.

METHODS: We analyzed data from the Jackson Heart Study, a population-based cohort study comprised exclusively of African American adults (n=5306). Analyses were restricted to 1016 participants who completed ambulatory BP monitoring at baseline in 2000 to 2004. Mean SBP and DBP levels were calculated for daytime (10:00 am‒8:00 pm), 24-hour (all available readings), and nighttime (midnight‒6:00 am) periods, separately. Daytime, 24-hour, and nighttime BP thresholds for ambulatory hypertension were identified using regression- and outcome-derived approaches. The composite of a cardiovascular disease or an all-cause mortality event was used in the outcome-derived approach. For this latter approach, BP thresholds were identified only for SBP because clinic DBP was not associated with the outcome. Analyses were stratified by antihypertensive medication use.

RESULTS: Among participants not taking antihypertensive medication, the regression-derived thresholds for daytime, 24-hour, and nighttime SBP/DBP corresponding to clinic SBP/DBP of 140/90 mm Hg were 134/85 mm Hg, 130/81 mm Hg, and 123/73 mm Hg, respectively. The outcome-derived thresholds for daytime, 24-hour, and nighttime SBP corresponding to a clinic SBP ≥140 mm Hg were 138 mm Hg, 134 mm Hg, and 129 mm Hg, respectively. Among participants taking antihypertensive medication, the regression-derived thresholds for daytime, 24-hour, and nighttime SBP/DBP corresponding to clinic SBP/DBP of 140/90 mm Hg were 135/85 mm Hg, 133/82 mm Hg, and 128/76 mm Hg, respectively. The corresponding outcome-derived thresholds for daytime, 24-hour, and nighttime SBP were 140 mm Hg, 137 mm Hg, and 133 mm Hg, respectively, among those taking antihypertensive medication.

CONCLUSIONS: On the basis of the outcome-derived approach for SBP and regression-derived approach for DBP, the following definitions for daytime, 24-hour, and nighttime hypertension corresponding to clinic SBP/DBP ≥140/90 mm Hg are proposed for African American adults: daytime SBP/DBP ≥140/85 mm Hg, 24-hour SBP/DBP ≥135/80 mm Hg, and nighttime SBP/DBP ≥130/75 mm Hg, respectively.

Thresholds for Ambulatory Blood Pressure Among African Americans in the Jackson Heart Study

© 2017 American Heart Association, Inc.

Correspondence to: Joseph Ravenell, MD, MS, New York University School of Medicine, Department of Population Health, 227 East 30th Street, 6th Floor, #637, New York, NY 10016. E-mail [email protected]

Sources of Funding, see page 2479

Key Words: ambulatory blood pressure monitoring ◼ African American adults ◼ hypertension

Joseph Ravenell, MD, MSDaichi Shimbo, MDJohn N. Booth III, MSDaniel F. Sarpong, PhDCharles Agyemang, MPH,

PhDDanielle L. Beatty Moody,

PhDMarwah Abdalla, MD,

MPHTanya M. Spruill, PhDAmanda J. Shallcross, ND,

MPHAdam P. Bress, PharmD,

MSPaul Muntner, PhDGbenga Ogedegbe, MD,

MS, MPH

by guest on August 6, 2017

http://circ.ahajournals.org/D

ownloaded from

by guest on A

ugust 6, 2017http://circ.ahajournals.org/

Dow

nloaded from



ownloaded from

by guest on A


Dow

nloaded from



ownloaded from

by guest on A


Dow

nloaded from



ownloaded from

by guest on A


Dow

nloaded from



ownloaded from

by guest on A


Dow

nloaded from



ownloaded from

mailto:[email protected]

http://circ.ahajournals.org/











Ambulatory Blood Pressure in African Americans


Circulation. 2017;135:2470–2480. DOI: 10.1161/CIRCULATIONAHA.116.027051 June 20/27, 2017 2471

The 2015 US Preventive Services Task Force recom-mendation on screening for high blood pressure (BP) suggests using out-of-clinic BP measurements

to confirm diagnoses of hypertension made in the clinic setting.1 Ambulatory BP monitoring (ABPM) is the most commonly recommended approach for out-of-office BP measurement.2,3 In addition to confirming the diagnosis of hypertension, ABPM can be used to determine BP control among individuals taking antihypertensive medication as well as identify BP phenotypes, including masked and noc-turnal hypertension.3,4 Several guidelines, scientific state-ments, and position papers have proposed BP thresholds for identifying ambulatory hypertension.3,5–8 These thresh-olds were primarily derived from population-based studies of European, Japanese, and South American populations.

African Americans, a group with high risk for cardio-vascular disease (CVD) outcomes, have higher mean daytime and nighttime ambulatory BP compared with whites.9–11 The applicability of ABPM thresholds derived in European, Japanese, and South American popula-tions to African American adults is unclear. In the current study, we used data from the Jackson Heart Study (JHS), a population-based cohort comprised exclusively of Afri-can American adults to determine ambulatory daytime, 24-hour, and nighttime BP thresholds that correspond to different clinic BP thresholds. In addition, we identified

BP thresholds on ABPM that provide similar probability of CVD or all-cause mortality events as these clinic BP thresholds.

METHODSStudy PopulationThe JHS is a population-based prospective cohort study designed to identify CVD risk factors among African American adults.12 The JHS enrolled 5306 noninstitutionalized African American adults, ≥21 years of age, between 2000 and 2004. Participants were recruited from the Atherosclerosis Risk in the Community site in Jackson, MS, and a representative sample of urban and rural Jackson, Mississippi, metropolitan tricounty (Hinds, Madison, and Rankin Counties) residents, volunteers, randomly selected individuals, and secondary family members.12,13 The current analysis was restricted to 1148 JHS participants who completed ABPM as part of their baseline study visit (visit 1 2000–2004). Participants (n=132) who did not meet the International Database on ABPM in relation to Cardiovascular Outcomes criteria for valid ABPM (n=102; described later) or who were missing informa-tion on clinic BP or antihypertensive medication use (n=30) were excluded from the analyses. After these exclusions were applied, the final analytic sample size was 1016 participants.

The institutional review boards of the University of Mississippi Medical Center, Jackson State University, and Tougaloo College approved the JHS protocol. The current anal-ysis of JHS data was approved by the institutional review board at the University of Alabama at Birmingham. All participants provided written informed consent.

Data CollectionA detailed description of the methodology and data-collection procedures for the baseline visit in the JHS have been described elsewhere.13–17 Briefly, data were collected during an in-home interview, by clinic examination, and by ABPM. ABPM was an optional procedure that participants were invited to complete. Of relevance to the current analysis, during the in-home inter-view, self-reported information was collected on sociodemo-graphics, health behaviors (eg, alcohol consumption, current smoking), prior diagnoses of comorbid conditions, and anti-hypertensive medication use (see Methods in online-only Data Supplement). During the clinic examination, trained technicians measured height, weight, and BP; collected blood and urine samples; and conducted 2-dimensional echocardiography. Body mass index was calculated as weight in kilograms divided by height in meters squared. Using specimens collected dur-ing the study visit, low-density lipoprotein and high-density lipoprotein cholesterol, serum creatinine, urine albumin and creatinine, fasting serum glucose, and glycohemoglobin were measured. Estimated glomerular filtration rate was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation,18 and albumin-to-creatinine ratio was also calculated. Diabetes mellitus was defined by a fasting blood glucose ≥126 mg/dL, glycohemoglobin ≥6.5%, or a prior diagnosis of diabe-tes mellitus with the use of glucose-lowering medication.

Clinic BP MeasurementClinic BP was measured by trained staff using a random zero sphygmomanometer (Hawksley and Sons Ltd.) during the baseline

Clinical Perspective

What Is New?• Ambulatory blood pressure monitoring (ABPM) is the

most commonly recommended approach for out-of-office BP measurement.

• Several guidelines, scientific statements, and posi-tion papers have proposed BP thresholds for identi-fying ambulatory hypertension on ABPM.

• These thresholds were primarily derived from pop-ulation-based studies of European, Japanese, and South American populations.

• We determined BP thresholds on ABPM for ambula-tory hypertension in a US population-based sample of African American adults not taking and taking anti-hypertensive medication.

• The ambulatory BP thresholds identified for African American adults were higher than those from pub-lished recommendations.

What Are the Clinical Implications?• The use of these ABPM thresholds in African American

adults will lead to a lower prevalence of daytime, 24-hour, and nighttime hypertension, compared with using ABPM thresholds from published recommendations.

• Therefore, these findings have important clinical implications for the diagnosis and treatment of hypertension in African American adults.



ownloaded from


Ravenell et al


study visit. Each participant’s right arm circumference was mea-sured to determine the appropriate cuff size. After participants had rested for ≥5 minutes in an upright position with their back and arms supported, feet flat on the floor, and legs uncrossed, 2 BP measurements in the right arm, separated by 1 minute, were recorded. The mean of these clinic BP measurements was cal-culated. Quality control was conducted by the JHS Coordinating Center by monitoring digit preference for each staff member and by comparing the mean BP level measured within and between study staff. A BP comparability substudy was conducted, in which BP was measured simultaneously, using a Y connector, by random zero sphygmomanometer and an Omron HEM-907XL device, a semiautomated device. As described in Methods and Figure I in the online-only Data Supplement and in prior analyses of the JHS, the random-zero BP measurements were calibrated to the semi-automated device using robust regression.13,17 Elevated clinic BP was defined as mean clinic systolic BP (SBP) ≥140 mm Hg or mean clinic diastolic BP (DBP) ≥90 mm Hg.

Ambulatory BP MonitoringAfter completing the clinic examination, participants were given the opportunity to complete ABPM for 24 hours. Participants who consented were fitted with an ABPM device (Spacelabs 90207, Spacelabs) on their nondominant arm. BP was recorded every 20 minutes. After 24 hours, the device was removed and data were downloaded onto a computer and processed with Medifacts International’s Medicom software International Database on ABPM in relation to Cardiovascular Outcomes criteria were used to define a complete ABPM mea-surement, which is defined as ≥10 daytime (10:00–20:00) and ≥5 nighttime (00:00–06:00) SBP and DBP measurements.19 The mean SBP and DBP levels were calculated for the daytime, nighttime, and 24-hour periods, separately.

CVD Events and All-Cause MortalityThe adjudication procedures for CVD events and all-cause mortal-ity have been described previously.20 Briefly, living participants or their proxies were contacted annually by telephone to assess potential CVD events and vital status. Hospital lists for discharges with specific diagnosis criteria were also obtained from the Jackson, Mississippi, tricounty area hospitals. Death certificates were requested from the Mississippi State Department of Health for JHS participants as needed. When a CVD-related hospital-ization or death was identified, medical records were retrieved and abstracted. Trained clinicians adjudicated events following published guidelines using the information available about the cir-cumstance surrounding an event.20 For the current analysis, defi-nite or probable CVD events (ie, coronary heart disease, nonfatal myocardial infarction, or acute coronary heart disease death or stroke defined as noncarotid embolic or thrombotic brain infarc-tion, brain hemorrhage, or subarachnoid hemorrhage) and all-cause mortality were available through December 31, 2012.

Statistical AnalysesCharacteristics of the participants included in the analytic sample (n=1016), the sample who underwent ABPM but were excluded from the analysis (n=132), and non-ABPM sample (n=4158) were calculated. All subsequent analyses were weighted to the age and sex distribution of African American adults using data from the 2010 US Census.21 Characteristics were calculated for the overall

Table 1. Characteristics of Jackson Heart Study Participants With Ambulatory Blood Pressure Monitoring at Baseline Included in the Current Analysis

VariableOverall

(n=1016)

Taking Antihypertensive Medication

No (n=441)Yes

(n=575)

Demographic characteristic

Age, y 50.5±11.2 47.1±11.2 55.7±9.7

Female sex, % 55.2 49.6 64.0

Education <high school, %

11.8 8.9 16.5

Clinical characteristic

Body mass index, kg/m2 31.2±6.6 30.2±7.6 32.6±5.5

Diabetes mellitus, % 17.2 7.8 32.4

LDL cholesterol, mg/dL

124.7±35.7 126.2±43.3 122.2±28.1

HDL cholesterol, mg/dL

51.0±13.6 50.9±15.1 51.1±12.4

Estimated glomerular filtration rate <60 mL/min/1.73 m2, %

4.7 1.1 10.2

Albumin to creatinine ratio ≥ 30 mg/g, %

9.5 5.3 16.6

Health behavior

Alcohol use, %

Nondrinker 51.3 45.6 60.4

Moderate drinker 43.5 48.6 35.5

Heavy drinker 5.1 5.8 4.1

Currently smoking, % 13.0 13.7 11.9

Clinic blood pressure, mm Hg

Systolic 124.4±14.8 121.6±17.2 128.7±11.9

Diastolic 76.1±8.3 76.2±9.6 75.8±7.1

Ambulatory blood pressure, mm Hg

Daytime systolic 127.2±12.8 124.8±13.6 130.9±11.4

Daytime diastolic 79.5±9.0 79.3±9.9 79.9±8.3

24-hour systolic 123.5±12.9 120.7±13.6 127.9±11.5

24-hour diastolic 75.4±8.7 74.8±9.5 76.2±8.0

Nighttime systolic 117.5±14.6 114.2±15.1 122.7±13.1

Nighttime diastolic 68.8±9.5 67.7±10.4 70.5±8.7

Characteristics are weighted to the age-sex distribution for the African American adult population using 2010 US census data. Unweighted characteristics for the overall sample (n=1016) are reported in Table I in the online-only Data Supplement. Values indicate mean±SD, unless otherwise indicated. HDL indicates high-density lipoprotein cholesterol; and LDL, low-density lipoprotein cholesterol.



ownloaded from





sample and participants taking and not taking antihypertensive medication, separately. We used 3 approaches used in previous ABPM studies to identify thresholds for ambulatory hypertension: distribution-derived, regression-derived, and outcome-derived.22–31 All of these analyses were conducted stratified by antihyperten-sive medication use. The 3 approaches described for determining daytime BP thresholds were repeated for 24-hour and, separately, nighttime BP. All analyses were conducted using SAS version 9.4 (SAS Institute).

Distribution-Derived ApproachThe goal of the distribution-derived approach is to identify percen-tiles of the BP distribution on ABPM for a population.22,24,27,30–32 The distributions of clinic and daytime SBP, and separately

clinic and daytime DBP, were calculated. The 90th, 95th, and 99th percentiles of the daytime SBP and DBP distributions were identified using quantile regression. The 95% confidence inter-val was calculated by computing a Huber sandwich estimate using a local estimate of the sparsity function. The percentage of participants with clinic SBP ≥140 mm Hg and clinic DBP ≥90 mm Hg was calculated for those with daytime SBP and DBP greater than or equal to these BP percentiles.

Regression-Derived ApproachThe goal of the regression-derived approach is to identify the BP levels on ABPM that correspond to specific clinic BP lev-els.22,23,26,29 Using the intercept and beta coefficient from a linear regression model with clinic SBP as the outcome and daytime

Table 2. Percentage of Participants With Elevated Clinic SBP and DBP Among Those With Ambulatory SBP and DBP at or Above the 90th, 95th, and 99th Percentiles, Stratified by Antihypertensive Medication Use

Systolic Blood Pressure Diastolic Blood Pressure

Ambulatory Threshold

(95% CI), mm Hg

% With Elevated Clinic BP

(≥140 mm Hg)

Ambulatory Threshold

(95% CI), mm Hg

% With Elevated Clinic BP

(≥90 mm Hg)

Not taking antihypertensive medication

Daytime BP

90th percentile 138 (135–140) 47.6% 91 (90–92) 24.9%

95th percentile 144 (137–151) 55.9% 94 (93–95) 27.8%

99th percentile 158 (134–182) 100.0% 99 (98–99) 55.1%

24-hour BP

90th percentile 133 (130–136) 48.2% 86 (84–87) 33.2%

95th percentile 140 (134–146) 72.2% 88 (85–91) 65.1%

99th percentile 152 (123–181) 100.0% 95 (91–100) 44.5%

Nighttime BP

90th percentile 129 (125–133) 41.7% 80 (77–83) 32.0%

95th percentile 137 (132–142) 50.6% 84 (81–88) 51.5%

99th percentile 147 (113–181) 83.7% 93 (89–97) 53.8%

Taking antihypertensive medication

Daytime BP

90th percentile 149 (145–152) 41.1% 93 (92–93) 7.8%

95th percentile 155 (150–159) 40.7% 99 (98–100) 0%

99th percentile 163 (153–173) 48.0% 104 (102–106) 0%

24-hour BP

90th percentile 146 (142–150) 45.6% 90 (86–93) 7.7%

95th percentile 152 (147–156) 53.7% 95 (92–98) 2.0%

99th percentile 164 (153–175) 66.8% 103 (99–107) 0%

Nighttime BP

90th percentile 144 (140–148) 49.0% 84 (80–88) 4.7%

95th percentile 149 (141–157) 48.3% 91 (87–94) 4.1%

99th percentile 166 (153–178) 42.1% 101 (96–106) 0%

Results are weighted to the age-sex distribution for the African American adult population using 2010 US census data. BP indicates blood pressure; and CI, confidence interval.



ownloaded from


Ravenell et al


SBP as the independent variable, the level of daytime SBP and 95% confidence interval corresponding to clinic BP level of 140 mm Hg were identified. Distribution curves were plotted to show the percentage of participants with clinic SBP ≥140 mm Hg and with daytime SBP at or above the thresholds from ABPM cor-responding to clinic SBP of 140 mm Hg. Daytime SBP levels corresponding to clinic SBP of 120 mm Hg, 130 mm Hg, and 160 mm Hg were also determined. We used the same approach to determine daytime DBP levels corresponding with clinic DBP of 80 mm Hg, 85 mm Hg, and 100 mm Hg.

Outcome-Derived ApproachThe goal of the outcome-derived approach is to identify the threshold for ambulatory BP that corresponds to the same probability of an event associated with a clinic BP level (eg, SBP of ≥140 mm Hg).25,28 Higher clinic SBP was statistically

significantly associated with the composite of CVD or all-cause mortality (P<0.001), but clinic DBP was not (P=0.121). Therefore, thresholds for ambulatory SBP but not DBP thresh-olds were calculated. Thresholds for daytime SBP yielding similar 5-year predicted probability of a CVD event or all-cause mortality associated with clinic SBP ≥140 mm Hg were first calculated. To do this, we performed Cox regression with the composite outcome of CVD or mortality and clinic SBP as the independent variable. We identified the 5-year predicted prob-ability of CVD or mortality for a clinic BP level of 140 mm Hg. Next, we conducted a Cox regression model with the outcome of CVD/all-cause mortality and daytime SBP as the indepen-dent variable. From this latter model, we determined the day-time SBP value that corresponded to the 5-year predicted probability of the outcome for a clinic SBP level of 140 mm Hg (ie, from the first Cox model described earlier). A bootstrap with 1000 data sets was used to calculate a 95% confidence interval for the daytime SBP yielding a similar 5-year predicted probability of the composite CVD/all-cause mortality outcome. Daytime SBP levels yielding similar 5-year predicted probability of an outcome associated with clinic SBP ≥120 mm Hg, ≥130 mm Hg, and ≥160 mm Hg were also calculated.

Comparisons With Recommended Ambulatory BP ThresholdsIn scientific statements, guidelines, and position papers, ambu-latory hypertension has been defined as having daytime SBP/DBP ≥135/85 mm Hg, 24-hour SBP/DBP ≥130/80 mm Hg, and nighttime SBP/DBP ≥120/70 mm Hg.3,5–8 We graphed the thresholds identified using the distribution-derived approach (95th percentile of BP), regression-derived approach (corre-sponding to clinic SBP/DBP of 140/90 mm Hg), and outcome-derived approach (corresponding to clinic SBP of 140 mm Hg) and contrasted them with these previously published thresholds.

RESULTSCharacteristics of the JHS participants included in the analytic sample (n=1016), the sample who underwent ABPM but were excluded from the analysis (n=132) and non-ABPM sample (n=4158) are shown in Table I in the online-only Data Supplement. The mean age±SD of par-ticipants in the analytic sample weighted to the 2010 US African American adult population was 50.5±11.2 years; 55.2% were female and 38.9% were taking an-tihypertensive medication (Table 1). Participants taking antihypertensive medication were older and more likely to be female, had less than a high school education, dia-betes mellitus, estimated glomerular filtration rate <60 mL/min/1.73 m2, urine albumin-to-creatinine ratio ≥30 mg/g, and did not drink alcohol compared with those not taking antihypertensive medication. In addition, mean body mass index was higher, whereas low-density lipo-protein cholesterol was lower among those taking versus not taking antihypertensive medication. Clinic, daytime, nighttime, and 24-hour SBP were each higher among par-ticipants taking versus not taking antihypertensive medi-cation, whereas no differences were present for DBP.

Table 3. Daytime, 24-Hour, and Nighttime SBP and DBP Thresholds Corresponding to Clinic SBP of 160, 140, 130, and 120 mm Hg and Clinic DBP of 100, 90, 85, and 80 mm Hg Among Participants Not Taking and Taking Antihypertensive Medication

Daytime SBP/DBP

24-Hour SBP/DBP

Nighttime SBP/DBP


Clinic SBP, mm Hg

160 144 (142–146) 140 (138–142) 133 (130–136)

140 134 (133–135) 130 (129–131) 123 (121–125)

130 129 (128–130) 125 (124–126) 118 (117–119)

120 124 (123–125) 120 (119–121) 113 (112–114)

Clinic DBP, mm Hg

100 90 (88–92) 86 (83–88) 77 (74–79)

90 85 (84–87) 81 (80–82) 73 (71–74)

85 83 (82–84) 79 (78–80) 71 (70–72)

80 81 (80–82) 77 (76–77) 69 (68–70)


Clinic SBP, mm Hg

160 143 (141–146) 142 (139–144) 137 (134–140)

140 135 (134–137) 133 (132–134) 128 (126–129)

130 131 (130–132) 128 (127–129) 123 (122–124)

120 127 (126–129) 124 (123–125) 119 (117–120)

Clinic DBP, mm Hg

100 89 (87–92) 86 (84–88) 80 (77–82)

90 85 (84–87) 82 (80–83) 76 (74–77)

85 83 (82–85) 80 (79–81) 74 (73–75)

80 82 (81–82) 78 (77–79) 72 (71–73)

Data represent ambulatory blood pressure thresholds (95% CI) corresponding to specified clinic SBP/DBP levels. Results are weighted to the age-sex distribution for the African American adult population using 2010 US census data. CI indicates confidence interval; DBP, diastolic blood pressure; and SBP, systolic blood pressure.



ownloaded from





Distribution-Derived Approach to Determine Ambulatory BP ThresholdsAmong participants not taking antihypertensive medica-tion, the 90th, 95th, and 99th percentile thresholds for daytime SBP were 138 mm Hg, 144 mm Hg, and 158 mm Hg, respectively (Table 2). Among participants at or above the 90th, 95th, and 99th percentile thresholds for daytime SBP, 47.6%, 55.9%, and 100.0% had clinic SBP ≥140 mm Hg, respectively. Among participants not taking antihypertensive medication, the 90th, 95th, and 99th percentile thresholds for daytime DBP was 91 mm Hg, 94 mm Hg, and 99 mm Hg, respectively. Among participants at or above the 90th, 95th, and 99th per-centile thresholds for daytime DBP, 24.9%, 27.8%, and 55.1% had clinic DBP ≥90 mm Hg, respectively. The 90th, 95th, and 99th percentile thresholds for daytime, 24-hour, and nighttime SBP and DBP were higher for participants taking versus not taking antihypertensive medication.

Regression-Derived Approach to Determine Ambulatory BP ThresholdsAmong participants not taking antihypertensive medica-tion, daytime, 24-hour, and nighttime thresholds cor-responding to clinic SBP/DBP of 140/90 mm Hg were

134/85 mm Hg, 130/81 mm Hg, and 123/73 mm Hg, respectively (Table 3). A higher percentage of partici-pants had daytime SBP ≥134 mm Hg, 24-hour SBP ≥130 mm Hg, and nighttime SBP ≥123 mm Hg than clinic SBP ≥140 mm Hg (Figure 1). Also, a higher percentage of participants had daytime DBP ≥85 mm Hg, 24-hour DBP ≥81 mm Hg, and nighttime DBP ≥73 mmHg than clinic DBP ≥90 mm Hg. Similar results were observed among participants taking antihypertensive medication (Figure II in the online-only Data Supplement). Daytime, 24-hour, and nighttime thresholds corresponding to clinic SBP/DBP of 120/80 mm Hg, 130/85 mm Hg, and 160/100 mm Hg among participants not taking antihypertensive medication are shown in Table 3 and Figure 2. Daytime, 24-hour, and nighttime SBP thresholds corresponding to clinic SBP of 120 mm Hg, 130 mm Hg, and 140 mm Hg were higher among participants taking versus those not taking antihypertensive medication (Table 3 and Figure III in the online-only Data Supplement). Among partici-pants taking versus those not taking antihypertensive medication, SBP thresholds corresponding to clinic SBP of 160 mm Hg were similar for the daytime and 24-hour period and higher for the nighttime period. Daytime, 24-hour, and nighttime DBP thresholds were similar among participants taking versus not taking antihypertensive medication.

Figure 1. Daytime, 24-hour, and nighttime blood pressure thresholds corresponding to a clinic systolic blood pressure/diastolic blood pressure threshold of 140/90 mm Hg determined using the regression-derived approach in participants not taking antihypertensive medication. Dash line represents the distribution of ambulatory blood pressure. Solid line represents the distribution of clinic blood pressure. Light gray areas indicate the participants with daytime, 24-hour, and nighttime blood pressure at or above the thresholds corre-sponding to clinic systolic blood pressure/diastolic blood pressure of 140/90 mm Hg. Dark gray areas indicate the participants with clinic systolic blood pressure/diastolic blood pressure at or above the thresholds of 140/90 mm Hg.



ownloaded from


Ravenell et al


Outcome-Derived Approach to Determine Ambulatory BP ThresholdsThere were 165 events (n=80 CVD events and n=85 deaths) over a median 10.8 years of follow-up. The 5-year predicted probability of a CVD or an all-cause mortality event was higher at increasing levels of clinic SBP (Table 4). Among participants not taking antihyper-tensive medication, daytime, 24-hour, and nighttime SBP thresholds corresponding to clinic SBP of 140 mm Hg were 138 mmHg, 134 mm Hg, and 129 mm Hg, respec-tively. Among participants taking antihypertensive medi-cation, daytime, 24-hour, and nighttime SBP thresholds corresponding to clinic SBP of 140 mm Hg were 140 mm Hg, 137 mm Hg, and 133 mm Hg, respectively. Among participants taking versus not taking antihyper-tensive medication, daytime and 24-hour SBP thresh-olds were higher for clinic SBP thresholds of 160 mm Hg and similar for clinic SBP thresholds of 120 mm Hg and 130 mm Hg. Last, among participants taking versus not taking antihypertensive medication, nighttime SBP

thresholds were higher for clinic SBP thresholds of 120 mm Hg, 130 mm Hg, and 160 mm Hg.

Comparisons With Published RecommendationsThresholds defined by the 95th percentile of the popula-tion distribution and the outcome-derived approach were higher than those from published recommendations3,5–8 for daytime, 24-hour, and nighttime BP among participants not taking and taking antihypertensive medication (Figure 3 and Figure IV in the online-only Data Supplement). Thresholds determined using the regression-derived approach were similar to published recommendations3,5–8 for daytime and 24-hour BP and higher for nighttime BP among participants not taking and taking antihypertensive medication.

DISCUSSIONIn the present study, we identified BP thresholds for am-bulatory hypertension in a US population-based sample of African American adults not taking and taking antihy-

144134129124

130120 140 16080

180

Day

time

SBP,

mm

Hg

100 220

140130125120

130120 140 16080

180

24-h

our

SBP,

mm

Hg

100 220

133123118113

130120 140 16080

180

Nig

htti

me

SBP

, mm

Hg

100 220

Clinic Systolic Blood Pressure, mmHg

90

85

83

81

80 85 1009040

100

Day

time

DB

P, m

mH

g

50 105

86

81

79

77

80 85 1009040

100

24-h

our

DB

P, m

mH

g

50 105

77

73

71

69

80 85 1009040

100

Nig

httim

e D

BP

, mm

Hg

50 105

Clinic Diastolic Blood Pressure, mmHg

Figure 2. Daytime, 24-hour, and nighttime SBP and DBP thresholds corresponding to a clinic SBP/DBP threshold of 120/80 mm Hg, 130/85 mm Hg, 140/90 mm Hg, and 160/100 mm Hg determined using the regression-derived approach in participants not taking antihypertensive medication. Light gray area represents 95% confidence interval bands. Dark gray area represents 95% prediction bands. DBP indicates diastolic blood pressure; and SBP, systolic blood pressure.



ownloaded from





pertensive medication using the distribution-, regression-, and outcome-derived approaches. The ambulatory BP thresholds identified for African American adults were

higher than those from published recommendations, which were based on data from European, Japanese, and SouthAmerican populations.3,5–8

Figure 3. Daytime, 24-hour, and nighttime SBP and DBP thresholds among participants not taking antihypertensive medication compared with published recommendations3,5–8 of blood pressure thresholds for ambulatory hyperten-sion: daytime SBP/DBP ≥135/85 mm Hg, 24-hour SBP/DBP ≥130/80 mm Hg, and nighttime SBP/DBP ≥120/70 mm Hg. Lines represent recommended systolic blood pressure and diastolic blood pressure thresholds. DBP indicates diastolic blood pressure; and SBP, systolic blood pressure.



ownloaded from


Ravenell et al


Using the outcome-derived approach correspond-ing to a clinic SBP threshold of 140 mm Hg and re-gression-derived approach corresponding to a clinic DBP threshold of 90 mm Hg, we recommend defin-ing daytime hypertension as mean daytime SBP/DBP ≥140/85 mm Hg, 24-hour hypertension as mean 24-hour SBP/DBP ≥135/80 mm Hg, and nighttime hypertension as mean nighttime SBP/DBP ≥130/75 mmHg in African American adults. The use of these BP thresholds, which are higher than those from published recommendations, will lead to a lower prevalence of daytime, 24-hour, and nighttime hypertension among African American adults. Consequently, the prevalence of white coat hypertension (ie, clinic hypertension with-out ambulatory hypertension) will increase and masked hypertension (ie, absence of clinic hypertension with ambulatory hypertension) will decrease.

An early method to derive ambulatory BP thresholds was the distribution-derived approach, commonly using the 95th percentile.22,24,27,30–32 In prior studies, thresh-olds using the distribution-derived approach have varied substantially from 1 another.24,27,30–32 These divergent findings likely reflect differences in population charac-teristics, including the prevalence of clinic hypertension and whether the population was composed of individuals not taking and taking antihypertensive medication. An-other issue with the distribution-derived approach is the assumption that a fixed percentage of the population has ambulatory hypertension. Given these limitations, the ambulatory BP thresholds we recommend do not take into consideration the thresholds determined using the distribution-derived approach.

There are several strengths of the current study. Few prior studies have conducted ABPM among African Ameri-can adults, a population disproportionately affected by BP-related CVD compared with other racial/ethnic groups in the United States. The relatively large sample size with ABPM in the JHS allowed us to determine ABPM thresholds among participants taking and not taking antihypertensive medication, separately, and for a wide range of clinic BP levels. Despite these strengths, the results should be in-terpreted in the context of possible limitations. Only a sub-set of JHS participants underwent ABPM. Further, 2 clinic BP readings were obtained for each participant. The BP thresholds may have differed if ≥3 BP measurements were obtained at the baseline visit or serial clinic BP measure-ments were averaged across multiple visits. Also, clinic BP was measured using a random-zero sphygmomanometer among JHS participants. However, these measurements were calibrated to an oscillometric device. Further, the ABPM cuff was placed on the participant’s nondominant arm to minimize the effect of daily activities on BP mea-surement (eg, writing, picking up objects), whereas clinic BP was measured in the right arm. Therefore, the present results may have been affected by interarm differences in BP measurements. Last, data were available only for Afri-can American adults. ABPM thresholds for whites, Hispan-ics, and Asians in the United States need to be identified among the same population to determine whether ambula-tory BP thresholds differ by race/ethnicity.

In conclusion, the current study provides BP thresholds for identifying ambulatory hypertension in a US population-based sample of African American adults not taking and taking antihypertensive medication. The following ambula-

Table 4. Daytime, 24-Hour, and Nighttime SBP Thresholds Yielding an Equivalent 5-Year Predicted Probability of the Composite Outcome of Cardiovascular Disease or All-Cause Mortality at Clinic SBP Thresholds of 160 mm Hg, 140 mm Hg, 130 mm Hg, and 120 mm Hg

Clinic SBP Threshold, mm Hg

5-Year Predicted Probability of

Cardiovascular Disease/All-Cause

Mortality, %Daytime SBP (95%

CI), mm Hg24-Hour SBP (95%

CI), mm HgNighttime SBP (95%

CI), mm Hg


160 10.2 (4.4‒21.1) 148 (140–157) 144 (136–152) 140 (132–152)

140 4.3 (2.0–8.3) 138 (130–146) 134 (126–142) 129 (120–139)

130 2.7 (1.3–5.4) 133 (124–141) 129 (121–137) 123 (114–132)

120 1.8 (0.8–3.7) 128 (118–137) 124 (116–133) 117 (108–127)


160 9.0 (5.4–13.8) 152 (143–166) 148 (140–159) 145 (136–156)

140 5.9 (4.2 – 8.0) 140 (129–151) 137 (127–146) 133 (123–142)

130 4.8 (3.3–6.5) 134 (120–144) 131 (120–140) 127 (116–137)

120 3.8 (2.5–5.6) 127 (109–138) 125 (112–135) 121 (107–131)

CI, confidence interval; and SBP, systolic blood pressure. Results are weighted to the age-sex distribution for the African American adult population using 2010 US census data.



ownloaded from





tory hypertension definitions corresponding to clinic SBP/DBP ≥140/90 mmHg are proposed: daytime hypertension defined as daytime SBP/DBP ≥140/85 mm Hg, 24-hour hy-pertension defined as 24-hour SBP/DBP ≥135/80 mm Hg, and nighttime hypertension defined as nighttime SBP/DBP ≥130/75 mm Hg. These thresholds differ from previously published recommendations, suggesting that there may be racial differences in thresholds for ambulatory BP.

SOURCES OF FUNDINGThe Jackson Heart Study is supported and conducted in collabo-ration with Jackson State University (N01-HC-95170), University of Mississippi Medical Center (N01-HC-95171), and Touglaoo Col-lege (N01-HC-95172) and contracts HHSN268201300046C, HHSN268201300047C, HHSN268201300048C, HH-SN268201300049C, and HHSN268201300050C from the Na-tional Heart, Lung, and Blood Institute and the National Institute on Minority Health and Health Disparities at the National Institutes of Health. This work was also supported by the National Insti-tutes of Health (HL047540, HL117323, HL117323-02S2, K24-HL125704) from the National Heart, Lung, and Blood Institute, 5KL2TR001065 from the National Center for Advancing Transla-tional Sciences, Bethesda, MD, and 15SFRN2390002 from the American Heart Association. The funding sources played no role in the design, conduct, analyses, or reporting of the present study.

DISCLOSURESDr Muntner received an institutional grant from Amgen Inc. Dr Shimbo is a consultant for Abbott Vascular and Novartis Phar-maceuticals Corporation. Dr Bress received an institutional grant from Novartis. The other authors report no conflicts of interest.

AFFILIATIONSFrom Department of Population Health, New York University School of Medicine (J.R., T.M.S., A.J.S., G.O.); Department of Medicine, Columbia University, New York (D.S., M.A.); Depart-ment of Epidemiology, University of Alabama at Birmingham (J.N.B., P.M.); Center for Minority Health & Health Disparities Research and Education, Xavier University of Louisiana, New Orleans (D.F.S.); Department of Public Health, Amsterdam Medical Centre, University of Amsterdam, The Netherlands (C.A.); Department of Psychology, University of Maryland, Balti-more County (D.L.B.M.); and Division of Health System Innova-tion and Research, Department of Population Health Sciences, University of Utah School of Medicine, Salt Lake City (A.P.B.).

FOOTNOTESReceived December 21, 2016; accepted April 10, 2017.

The online-only Data Supplement is available with this arti-cle at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/ CIRCULATIONAHA.116.027051/-/DC1.

Continuing medical education (CME) credit is available for this article. Go to http://cme.ahajournals.org to take the quiz.

Circulation is available at http://circ.ahajournals.org.

REFERENCES 1. Siu AL; U.S. Preventive Services Task Force. Screening for high

blood pressure in adults: U.S. Preventive Services Task Force rec-ommendation statement. Ann Intern Med. 2015;163:778–786. doi: 10.7326/M15-2223.

2. Piper MA, Evans CV, Burda BU, Margolis KL, O’Connor E, Whit-lock EP. Diagnostic and predictive accuracy of blood pressure screening methods with consideration of rescreening intervals: a systematic review for the U.S. Preventive Services Task Force. Ann Intern Med. 2015;162:192–204. doi: 10.7326/M14-1539.

3. Shimbo D, Abdalla M, Falzon L, Townsend RR, Muntner P. Role of ambulatory and home blood pressure monitoring in clinical prac-tice: a narrative review. Ann Intern Med. 2015;163:691–700. doi: 10.7326/M15-1270.

4. Pickering TG, Shimbo D, Haas D. Ambulatory blood-pressure monitoring. N Engl J Med. 2006;354:2368–2374. doi: 10.1056/NEJMra060433.

5. Mancia G, Fagard R, Narkiewicz K, Redon J, Zanchetti A, Böhm M, Christiaens T, Cifkova R, De Backer G, Dominiczak A, Galderisi M, Grobbee DE, Jaarsma T, Kirchhof P, Kjeldsen SE, Laurent S, Manolis AJ, Nilsson PM, Ruilope LM, Schmieder RE, Sirnes PA, Sleight P, Viigimaa M, Waeber B, Zannad F, Redon J, Dominiczak A, Narkiewicz K, Nilsson PM, Burnier M, Viigimaa M, Ambrosioni E, Caufield M, Coca A, Olsen MH, Schmieder RE, Tsioufis C, van de Borne P, Zamorano JL, Achenbach S, Baumgartner H, Bax JJ, Bueno H, Dean V, Deaton C, Erol C, Fagard R, Ferrari R, Hasdai D, Hoes AW, Kirchhof P, Knuuti J, Kolh P, Lancellotti P, Linhart A, Ni-hoyannopoulos P, Piepoli MF, Ponikowski P, Sirnes PA, Tamargo JL, Tendera M, Torbicki A, Wijns W, Windecker S, Clement DL, Coca A, Gillebert TC, Tendera M, Rosei EA, Ambrosioni E, Anker SD, Bauersachs J, Hitij JB, Caulfield M, De Buyzere M, De Geest S, Derumeaux GA, Erdine S, Farsang C, Funck-Brentano C, Gerc V, Germano G, Gielen S, Haller H, Hoes AW, Jordan J, Kahan T, Koma-jda M, Lovic D, Mahrholdt H, Olsen MH, Ostergren J, Parati G, Perk J, Polonia J, Popescu BA, Reiner Z, Rydén L, Sirenko Y, Stanton A, Struijker-Boudier H, Tsioufis C, van de Borne P, Vlachopoulos C, Volpe M, Wood DA. 2013 ESH/ESC guidelines for the manage-ment of arterial hypertension: the Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). Eur Heart J. 2013;34:2159–2219. doi: 10.1093/eurheartj/eht151.

6. Pickering TG, Hall JE, Appel LJ, Falkner BE, Graves J, Hill MN, Jones DW, Kurtz T, Sheps SG, Roccella EJ. Recommendations for blood pressure measurement in humans and experimental animals: part 1: blood pressure measurement in humans: a state-ment for professionals from the Subcommittee of Professional and Public Education of the American Heart Association Council on High Blood Pressure Research. Circulation. 2005;111:697–716. doi: 10.1161/01.CIR.0000154900.76284.F6.

7. Pickering TG, White WB; American Society of Hypertension Writ-ing Group. ASH position paper: home and ambulatory blood pres-sure monitoring: when and how to use self (home) and ambula-tory blood pressure monitoring. J Clin Hypertens (Greenwich). 2008;10:850–855. doi: 10.1111/j.1751-7176.2008.00043.x.

8. O’Brien E, Parati G, Stergiou G, Asmar R, Beilin L, Bilo G, Clement D, de la Sierra A, de Leeuw P, Dolan E, Fagard R, Graves J, Head GA, Imai Y, Kario K, Lurbe E, Mallion JM, Mancia G, Mengden T, Myers M, Ogedegbe G, Ohkubo T, Omboni S, Palatini P, Redon J, Ruilope LM, Shennan A, Staessen JA, vanMontfrans G, Verdecchia P, Waeber B, Wang J, Zanchetti A, Zhang Y; European Society of Hypertension Working Group on Blood Pressure Monitoring. European Society of Hypertension position paper on ambulatory blood pressure monitoring. J Hypertens. 2013;31:1731–1768. doi: 10.1097/HJH.0b013e328363e964.

9. Muntner P, Lewis CE, Diaz KM, Carson AP, Kim Y, Calhoun D, Yano Y, Viera AJ, Shimbo D. Racial differences in abnormal am-



ownloaded from


Ravenell et al


bulatory blood pressure monitoring measures: results from the Coronary Artery Risk Development in Young Adults (CARDIA) study. Am J Hypertens. 2015;28:640–648. doi: 10.1093/ajh/hpu193.

10. Knox SS, Hausdorff J, Markovitz JH; Coronary Artery Risk Devel-opment in Young Adults Study. Reactivity as a predictor of subse-quent blood pressure: racial differences in the Coronary Artery Risk Development in Young Adults (CARDIA) study. Hypertension. 2002;40:914–919.

11. Profant J, Dimsdale JE. Race and diurnal blood pressure patterns: a review and meta-analysis. Hypertension. 1999;33:1099–1104.

12. Sempos CT, Bild DE, Manolio TA. Overview of the Jackson Heart Study: a study of cardiovascular diseases in African American men and women. Am J Med. Sci. 1999;317:142–146.

13. Abdalla M, Booth JN 3rd, Seals SR, Spruill TM, Viera AJ, Diaz KM, Sims M, Muntner P, Shimbo D. Masked hypertension and incident clinic hypertension among blacks in the Jackson Heart Study. Hypertension. 2016;68:220–226. doi: 10.1161/HYPERTENSIO-NAHA.115.06904.

14. Jackson Heart Study Protocol Manuals. April 2003 and Septem-ber 2005. https://www.jacksonheartstudy.org/Research/Study-Design. Accessed March 8, 2017.

15. Carpenter MA, Crow R, Steffes M, Rock W, Heilbraun J, Evans G, Skelton T, Jensen R, Sarpong D. Laboratory, reading center, and coordinating center data management methods in the Jackson Heart Study. Am J Med. Sci. 2004;328:131–144.

16. Taylor HA Jr, Wilson JG, Jones DW, Sarpong DF, Srinivasan A, Gar-rison RJ, Nelson C, Wyatt SB. Toward resolution of cardiovascular health disparities in African Americans: design and methods of the Jackson Heart Study. Ethn Dis. 2005;15(4 Suppl 6):S6–S4.

17. Diaz KM, Booth JN 3rd, Seals SR, Abdalla M, Dubbert PM, Sims M, Ladapo JA, Redmond N, Muntner P, Shimbo D. Physical activity and incident hypertension in African Americans: the Jackson Heart Study. Hypertension. 2017;69:421–427. doi: 10.1161/HYPER-TENSIONAHA.116.08398.

18. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J; CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150:604–612.

19. Thijs L, Hansen TW, Kikuya M, Björklund-Bodegård K, Li Y, Dolan E, Tikhonoff V, Seidlerová J, Kuznetsova T, Stolarz K, Bianchi M, Richart T, Casiglia E, Malyutina S, Filipovsky J, Kawecka-Jaszcz K, Nikitin Y, Ohkubo T, Sandoya E, Wang J, Torp-Pedersen C, Lind L, Ibsen H, Imai Y, Staessen JA, O’Brien E; IDACO Investigators. The International Database of Ambulatory Blood Pressure in relation to Cardiovascular Outcome (IDACO): protocol and research perspec-tives. Blood Press Monit. 2007;12:255–262.

20. Keku E, Rosamond W, Taylor HA Jr, Garrison R, Wyatt SB, Richard M, Jenkins B, Reeves L, Sarpong D. Cardiovascular disease event classification in the Jackson Heart Study: methods and proce-dures. Ethn Dis. 2005;15(4 Suppl 6):S6–S62.

21. U.S. Census Bureau, 2010 Census, Current Population Survey, Annual Social and Economic Supplement. March 2010. https://www.census.gov/population/race/data/ppl-ba10.html. Accessed March 21, 2017.

22. Hansen TW, Kikuya M, Thijs L, Li Y, Boggia J, Björklund-Bodegârd K, Torp-Pedersen C, Jeppesen J, Ibsen H, Staessen JA. Diagnostic thresholds for ambulatory blood pressure moving lower: a review based on a meta-analysis-clinical implications. J Clin Hypertens (Greenwich). 2008;10:377–381.

23. Head GA, Mihailidou AS, Duggan KA, Beilin LJ, Berry N, Brown MA, Bune AJ, Cowley D, Chalmers JP, Howe PR, Hodgson J, Ludbrook J, Mangoni AA, McGrath BP, Nelson MR, Sharman JE, Stowasser M; Ambulatory Blood Pressure Working Group of the High Blood Pressure Research Council of Australia. Definition of ambulatory blood pressure targets for diagnosis and treatment of hyperten-sion in relation to clinic blood pressure: prospective cohort study. BMJ. 2010;340:c1104.

24. Imai Y, Nagai K, Sakuma M, Sakuma H, Nakatsuka H, Satoh H, Minami N, Munakata M, Hashimoto J, Yamagishi T. Ambula-tory blood pressure of adults in Ohasama, Japan. Hypertension. 1993;22:900–912.

25. Kikuya M, Hansen TW, Thijs L, Björklund-Bodegård K, Kuznetsova T, Ohkubo T, Richart T, Torp-Pedersen C, Lind L, Ibsen H, Imai Y, Staessen JA; International Database on Ambulatory blood pressure monitoring in relation to Cardiovascular Outcomes Investigators. Diagnostic thresholds for ambulatory blood pressure monitoring based on 10-year cardiovascular risk. Circulation. 2007;115:2145–2152. doi: 10.1161/CIRCULATIONAHA.106.662254.

26. Mancia G, Sega R, Bravi C, De Vito G, Valagussa F, Cesana G, Zanchetti A. Ambulatory blood pressure normality: results from the PAMELA study. J Hypertens. 1995;13(12 Pt 1):1377–1390.

27. O’Brien E, Murphy J, Tyndall A, Atkins N, Mee F, McCarthy G, Staessen J, Cox J, O’Malley K. Twenty-four-hour ambulatory blood pressure in men and women aged 17 to 80 years: the Allied Irish Bank Study. J Hypertens. 1991;9:355–360.

28. Ohkubo T, Imai Y, Tsuji I, Nagai K, Ito S, Satoh H, Hisamichi S. Ref-erence values for 24-hour ambulatory blood pressure monitoring based on a prognostic criterion: the Ohasama Study. Hyperten-sion. 1998;32:255–259.

29. Schettini C, Bianchi M, Nieto F, Sandoya E, Senra H. Ambulatory blood pressure: normality and comparison with other measure-ments. Hypertension Working Group. Hypertension. 1999;34(4 Pt 2):818–825.

30. Staessen J, Bulpitt CJ, Fagard R, Mancia G, O’Brien ET, Thijs L, Vyncke G, Amery A. Reference values for ambulatory blood pres-sure: a population study. J Hypertens Suppl. 1991;9:S320–S321.

31. Staessen JA, O’Brien ET, Amery AK, Atkins N, Baumgart P, De Cort P, Degaute JP, Dolenc P, De Gaudemaris R, Enstrom I, Fa-gard R, Gosse P, Gourlay S, Hayashi H, Imai Y, James G, Kawasaki T, Kuschnir E, Kuwajima I, Lindholm L, Liu L, Macor F, Mancia G, McGrath B, Middeke M, Ming J, Omboni S, Otsuka K, Palatini P, Parati G, Pieper C, Verdecchia P, Zachariah P, Zhang W. Am-bulatory blood pressure in normotensive and hypertensive sub-jects: results from an international database. J Hypertens Suppl. 1994;12:S1–S12.

32. Wiinberg N, Høegholm A, Christensen HR, Bang LE, Mikkelsen KL, Nielsen PE, Svendsen TL, Kampmann JP, Madsen NH, Bentzon MW. 24-h ambulatory blood pressure in 352 normal Danish sub-jects, related to age and gender. Am J Hypertens. 1995;8(10 Pt 1):978–986. doi: 10.1016/0895-7061(95)00216-2.



ownloaded from


Bress, Paul Muntner and Gbenga OgedegbeDanielle L. Beatty Moody, Marwah Abdalla, Tanya M. Spruill, Amanda J. Shallcross, Adam P.

Joseph Ravenell, Daichi Shimbo, John N. Booth III, Daniel F. Sarpong, Charles Agyemang,Heart Study

Thresholds for Ambulatory Blood Pressure Among African Americans in the Jackson

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 2017 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/CIRCULATIONAHA.116.027051

2017;135:2470-2480; originally published online April 20, 2017;Circulation.

http://circ.ahajournals.org/content/135/25/2470World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circ.ahajournals.org/content/suppl/2017/04/20/CIRCULATIONAHA.116.027051.DC1Data Supplement (unedited) at:

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer this process is available in the

click Request Permissions in the middle column of the Web page under Services. Further information aboutOffice. Once the online version of the published article for which permission is being requested is located,

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the EditorialCirculationin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:



ownloaded from

http://circ.ahajournals.org/content/135/25/2470

http://circ.ahajournals.org/content/suppl/2017/04/20/CIRCULATIONAHA.116.027051.DC1

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/


1

Supplemental Methods Data Collection and Clinical Covariates

During clinical examination, trained African American interviewers administered standardized questionnaires to assess selected demographic and behavior characteristics: age, sex, education, socioeconomic status, alcohol consumption and, current smoking. Education was measured as the highest level of schooling completed and classified into two categories within this study as “less than high school” or “greater than high school”. Current smoking was defined by affirmative responses to the questions “Have you smoked more than 400 cigarettes in your lifetime?” and “Do you now smoke cigarettes?” Daily alcohol consumption was assessed from a validated food frequency questionnaire,1 and in our study defined as “none”: 0 drinks/week, “moderate” consumption: 1-14 and 1-7 alcoholic drinks/week for men and women respectively, and “heavy” consumption: >14 and >7 alcoholic drinks/week for men and women respectively.2

Participants were asked to bring any medications taken within 2 weeks prior to the baseline examination to the clinic visit and were transcribed verbatim. Medication coding was performed by a pharmacist using the Medispan dictionary and classified into categories according to the Therapeutic Classification System. Antihypertensive medication use was defined by self-report. Participants were asked to avoid caffeine, eating, heavy physical activity, smoking, and alcohol intake for 12 hours prior to the clinic examination. During the clinical examination, weight and height were measured for each participant. Body mass index was calculated as the weight in kilograms divided by height in meters squared (kg/m2). Fasting blood samples were collected according to standardized procedures and processed at two central laboratories (University of Mississippi Medical Center and the University of Minnesota).1 Total and high-density lipoprotein (HDL) cholesterol was quantified by an oxidase method and low-density lipoprotein (LDL) cholesterol was calculated using the Freidewald equation. Estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation.3 Reduced eGFR was defined as <60 ml/min/1.73 m2. Urinary albumin to creatinine ratio was calculated. Diabetes was defined as a fasting (≥8 hours) serum glucose ≥126 mg/dL or hemoglobin A1c ≥6.5% or use of insulin or oral hypoglycemic medications within 2 weeks prior to the clinic examination.

Blood Pressure Comparability Study

As the Jackson Heart Study (JHS) began transitioning from using a random-zero sphygmomanometer (RZS) to an Omron HEM-907XL (Omron Healthcare Inc., Lake Forest, Il.), an automatic oscillatory device (AOD), a Blood Pressure Comparability Study was conducted during Visit 2. The purpose of the Blood Pressure Comparability Study was to allow the JHS Coordinating Center (JHS CC) to calibrate the BP measurements of participants with only RZS readings; this includes all participants at clinic exam Visit 1 and part of the cohort at clinic exam Visit 2. Of the 5,301 JHS participants enrolled at Visit 1, 5,280 had RZS readings only. At Visit 2, 1,966 had RZS reading only, 113 had AOD readings only, and 2,115 had both RZS and AOD readings.

2

Participants who had RZS and AOD readings comprised the Blood Pressure Comparability Study. At Visit 3, 3,814 had AOD readings only.

At Visit 2, two technicians measured each participant’s blood pressure (BP) simultaneously with the RZS and AOD using a Y connector and a double-headed stethoscope. Each technician was blinded to the other’s readings. The RZS was used to determine the pulse obliteration pressure using the standard RZS equipment. Next, the technician replaced the RZS cuff with the AOD cuff. The AOD was set to manual mode and used to inflate and deflate the BP cuff; the maximum value corresponding to the RZS in use was used to determine the peak inflation value for the AOD. During the deflation period, both technicians listened for the onset of the 1st and 5th Korotkoff phase sounds and recorded the corresponding systolic and diastolic BP. This was repeated twice per participant; thus, participants had a total of four BP readings: two from the RZS and two from the AOD. For analysis purposes, we averaged the two device-specific measurements. Once exploratory data analysis began, the JHS CC discovered a large amount of noise in the data; Supplemental Figure 1 below shows the observed noise for systolic BP and diastolic BP, respectively. The level of noise observed in the JHS data is not unexpected and has been shown elsewhere in the literature.4-6 However, due to such noise, the JHS CC decided to investigate several calibration methods beyond normal ordinary least squared (OLS) regression. The following 5 methods were explored:

1. Ignoring the change. This method completely ignores that the JHS changed blood pressure machines. If an AOD measure is present at clinic exam Visit 2, its value is used. Otherwise, the RZS measure is used.

2. Ordinary least squares. This method predicts a participant’s AOD measure by creating a prediction equation using ordinary OLS regression, where the participant’s RZS measure is the only predictor.

3. Deming regression. Considered the gold standard for calibration techniques, this method is similar to OLS but is an errors-in-variables regression model that accounts for error both in the dependent and independent variables.

4. Adjustment via average difference. Rather than adjust using a prediction equation, this method adjusts all participants by the observed average difference between the AOD and the RZS.

5. Modeling the difference between AOD and RZS as a function of RZS. This method utilizes the difference between AOD and RZS measurements as the outcome and the participant’s RZS measure as the predictor; that is, this models the difference between machines as a function of the participant’s RZS measure. Rather than normal OLS methods, this prediction model is created using robust regression.7

Several example models were compared across the different calibration methods. In particular, the JHS CC examined the resulting regression coefficients and standard errors; results across calibration methods were similar. Based on the shape of the data, the JHS CC felt most comfortable calibrating the data using a function of the RZS measurement. Thus, the JHS CC recommended that JHS researchers use the calibration equation created by method 5 (modeling the difference between AOD and

3

RZS as a function of RZS). The following equations resulted from method 5 and were used to predict SBP and DBP for the current study:

= + −AOD RZS RZSSBP 11.0 0.1*SBPSBP

= + −AOD RZS RZSDBP 10.3 0.2 *DBPDBP

4

Supplemental References

1. Carpenter MA, Crow R, Steffes M, Rock W, Heilbraun J, Evans G, Skelton T, Jensen R and Sarpong D. Laboratory, reading center, and coordinating center data management methods in the Jackson Heart Study. Am J Med Sci. 2004;328:131-144. 2. Dietary Guidelines for Americans Washington, D.C: US Government Printing Office (December 2010). http://www.health.gov/dietaryguidelines/dga2010/DietaryGuidelines2010.pdf Accessed March 8, 2017. 3. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF, 3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J and Ckd EPI. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150:604-612. 4. Ostchega Y, Nwankwo T, Sorlie PD, Wolz M and Zipf G. Assessing the validity of the Omron HEM-907XL oscillometric blood pressure measurement device in a National Survey environment. J Clin Hypertens (Greenwich). 2010;12:22-28. 5. Ostchega Y, Zhang G, Sorlie P, Hughes JP, Reed-Gillette DS, Nwankwo T and Yoon S. Blood pressure randomized methodology study comparing automatic oscillometric and mercury sphygmomanometer devices: National Health and Nutrition Examination Survey, 2009-2010. Natl Health Stat Report. 2012:1-15. 6. Stang A, Moebus S, Mohlenkamp S, Dragano N, Schmermund A, Beck EM, Siegrist J, Erbel R, Jockel KH and Heinz Nixdorf Recall Study Investigative G. Algorithms for converting random-zero to automated oscillometric blood pressure values, and vice versa. Am J Epidemiol. 2006;164:85-94. 7. Hampel F, Ronchetti E, Rousseeuw P and Stahel W. The approach based on influence functions. New York: Wiley-Interscience; 2005.

5

Supplemental Table 1. Characteristics of the Jackson Heart Study participants included in the analytical sample, the sample who underwent ABPM but were excluded from the analysis, and non-ABPM sample.

Analytic sample

Excluded from analytic sample*

Non-ABPM sample

n=1,016 n=132 n=4,158 Demographic Characteristics

Age, years 59.2 ± 10.9 58.4 ± 12.5 54.3 ± 13.1 Female sex, % 68.2% 68.9% 62.2% Education < high school, % 19.0% 26.5% 20.5%

Clinical Characteristics Body mass index, kg/m2 31.1 ± 6.4 32.9 ± 6.5 31.9 ± 7.4 Diabetes, % 24.5% 29.0% 21.0% LDL cholesterol, mg/dL 125.8 ± 35.7 126.5 ± 36.8 126.9 ± 36.8 HDL cholesterol, mg/dL 54.0 ± 15.0 52.9 ± 15.3 51.2 ± 14.5 Estimated glomerular filtration rate < 60 ml/min/1.73 m2, % 7.3% 12.1% 8.0% Albumin to creatinine ratio ≥ 30 mg/g, % 10.6% 12.9% 13.2%

Health Behaviors Alcohol use, % Non-drinker 63.2% 59.8% 57.3% Moderate drinker 34.4% 35.6% 39.1% Heavy drinker 2.4% 4.5% 3.7% Current Smoking, % 10.2% 17.7% 13.8%

Clinic blood pressure, mmHg Systolic 127.6 ± 15.9 128.5 ± 14.6 127.4 ± 17.2 Diastolic 74.4 ± 8.5 74.4 ± 8.3 76.1 ± 8.8

Antihypertensive medication use, % 56.6% 64.2% 47.5% Characteristics for the analytic sample weighted to the 2010 US Census distribution of age and sex for African American adults are reported in Table 1. Numbers in the table are percentages or mean ± standard deviation. ABPM: ambulatory blood pressure monitoring. HDL: high-density lipoprotein cholesterol. LDL: low-density lipoprotein cholesterol. *These participants underwent ABPM and either did not meet the IDACO criteria (n=102; see Methods section for a description of these criteria) or had missing data on clinic BP or antihypertensive medication use (n=30).

Supplemental Figures and Figure Legends

Supplemental Figure 1. Comparisons of random-zero sphygmomanometer and autonomic oscillatory device (Omron HEM-907XL) measurements for systolic blood pressure (left panel) and diastolic blood pressure (right panel).

7

Supplemental Figure 2. Daytime, 24-hour, and nighttime blood pressure thresholds corresponding to a clinic systolic blood pressure/diastolic blood pressure threshold of 140/90 mmHg determined using the regression-derived approach in participants taking antihypertensive medication. Dash line represents the distribution of ambulatory blood pressure. Solid line represents the distribution of clinic blood pressure. Light gray shaded regions indicate the participants with daytime, 24-hour, and nighttime blood pressure at or above the thresholds corresponding to clinic systolic blood pressure/diastolic blood pressure of 140/90 mmHg. Dark gray shaded regions indicate the participants with clinic systolic blood pressure/diastolic blood pressure at or above the thresholds of 140/90 mmHg.

135

140

Perc

entag

e (%

)

85

90

133

140

Perc

entag

e (%

)

82

90

128

140

Systolic blood pressure, mmHg

Perc

entag

e (%

)

80 100 120 140 160 180 200

76

90

Diastolic blood pressure, mmHg

40 60 80 100 120

8

Supplemental Figure 3. Daytime, 24-hour, and nighttime blood pressure thresholds corresponding to a clinic systolic blood pressure/diastolic blood pressure threshold of 120/80 mmHg, 130/85 mmHg, 140/90 mmHg, and 160/100 mmHg determined using the regression-derived approach in participants taking antihypertensive medication. Light gray area represents 95% confidence interval bands. Dark gray area represents 95% prediction bands. DBP: diastolic blood pressure; SBP: systolic blood pressure.

143135131127

130120 140 16080

180

Day

time

SBP,

mm

Hg

100 200

142133128124

130120 140 16080

180

24-h

our S

BP, m

mH

g

100 200

137128123119

130120 140 16080

180

Nig

httim

e SB

P, m

mH

g

100 200

Clinic Systolic Blood Pressure, mmHg

89

85

83

82

80 85 1009040

100

Day

time

DBP

, mm

Hg

50 105

86

82

80

78

80 85 1009040

100

24-h

our D

BP, m

mH

g

50 105

80

76

74

72

80 85 1009040

100

Nig

httim

e D

BP

, mm

Hg

50 105

Clinic Diastolic Blood Pressure, mmHg

9

Supplemental Figure 4. Daytime, 24-hour, and nighttime blood pressure thresholds estimated among participants taking antihypertensive medication compared to published recommendations of blood pressure thresholds for ambulatory hypertension: daytime systolic blood pressure/diastolic blood pressure ≥ 135/85 mmHg, 24-hour systolic blood pressure/diastolic blood pressure ≥ 130/80 mmHg, and nighttime systolic blood pressure/diastolic blood pressure ≥ 120/70 mmHg. Blue lines represent recommended systolic blood pressure and diastolic blood pressure thresholds. DBP: diastolic blood pressure; SBP: systolic blood pressure.

Supplemental Figure 3

Daytime SBP Daytime DBP

24-hour SBP 24-hour DBP

Nighttime SBP Nighttime DBP

110

115

120

125

130

135

140

145

150

155

160 Distribution-derived approach

(95th percentile)

Regression-derived approach

Outcome-derived approach

6065707580859095

100105110115120125130135140

Distribution-derived approach

(95th percentile)


110

115

120

125

130

135

140

145

150

155

160 Distribution-derived

approach (95th percentile)



6065707580859095

100105110115120125130135140


(95th percentile)


110

115

120

125

130

135

140

145

150

155

160


(95th percentile)



6065707580859095

100105110115120125130135140


(95th percentile)


Thresholds for Ambulatory Blood Pressure Among African ......Aug 10, 2017 · 2470 June 20/27, 2017...

Documents

Transcript of Thresholds for Ambulatory Blood Pressure Among African ......Aug 10, 2017 · 2470 June 20/27, 2017...