Impact of Ambulatory Blood Pressure Monitoring on Reclassification of Hypertension Prevalence and Control in Older People in Spain

Abstract

Ambulatory blood pressure monitoring (ABPM) accurately classifies blood pressure (BP) status but its impact on the prevalence and control of hypertension is little known. The authors conducted a cross‐sectional study in 2012 among 1047 individuals 60 years and older from the follow‐up of a population cohort in Spain. Three casual BP measurements and 24‐hour ABPM were performed under standardized conditions. Approximately 68.8% patients were hypertensive based on casual BP (≥140/90 mm Hg or current BP medication use) and 62.1% based on 24‐hour ABPM (≥130/80 mm Hg or current BP medication use) (P=.009). The proportion of patients with treatment‐eligible hypertension who met BP goals increased from 37.4% based on the casual BP target to 54.1% based on the 24‐hour BP target (absolute difference, 16.7%; P<.01). These results were consistent across alternative BP thresholds. Therefore, compared with casual BP, 24‐hour ABPM led to a reduction in the proportion of older patients recommended for hypertension treatment and a substantial increase in the proportion of those with hypertension control.

High blood pressure (BP) poses a serious cardiovascular risk to the individual and a burden to the population.1, 2, 3 Casual BP determined in the office or clinic has been the standard of measurement for many decades. However, ambulatory BP monitoring (ABPM) can provide an estimate of the true, or mean, BP level and predicts clinical outcomes better than conventional BP measurements.4, 5, 6, 7

Although there are many studies on the epidemiology of hypertension using both in‐office and out‐of‐office BP measurements,8, 9, 10, 11, 12 the consequences of using ABPM for surveillance of high BP in the general population have not been fully explored. In fact, it has been recently proposed that ABPM should be included in the US National Health and Nutrition Examination Survey (NHANES) to more accurately classify hypertension prevalence and control.13 Furthermore, the National Institute for Health and Clinical Excellence (NICE) in the United Kingdom recommends the use of ABPM to confirm the diagnosis of hypertension for all patients with an office BP ≥140/90 mm Hg.14

The introduction of ABPM in the follow‐up of patients older than 60 years from the Spanish Nutrition and Cardiovascular Risk Survey (ENRICA)15 has provided us with the opportunity to examine how many cases would have a change in BP status by using this technology in this general population setting. Like some other population surveys,15, 16, 17 in this Spanish study, casual BP was taken at patients’ homes, thus providing a more accurate measurement of BP because of the reduction of the alert reaction.18 Given that home BP measurement is increasingly used18, 19 and is likely to keep rising in the next few years, especially for monitoring BP control, this is a scenario in which measuring the impact of using ABPM also makes sense. Thus, we estimated the impact of ABPM on the prevalence and control of BP in older adults in Spain, a population group in whom hypertension is more frequent and challenging. This information is relevant from both a clinic and public health viewpoint, since a more accurate classification of BP status would allow for quantifying overtreatment or undertreatment and obtaining a balanced view of the burden of hypertension in the population.

Methods

Study Design and Participants

We conducted a cross‐sectional analysis from the second wave of the Seniors‐ENRICA study, a cohort set up in 2008–2010 with individuals selected through random sampling of the population aged 60 years and older in Spain.15 This second wave was conducted in 2012 among 2519 individuals and included a phone interview on health status, lifestyle, and morbidity, as well as a home visit to record BP and anthropometry, habitual diet, and prescribed medication.

Because of logistic and cost reasons, ABPM was offered to 1698 individuals and performed in 1328 patients (response rate, 78.2%). Compared with participants without ABPM, those who underwent the measurement had similar age (71.8 years vs 71.7 years), proportion of men (47% vs 49%), education level (63% vs 61% with ≤ primary studies), mean body mass index (BMI; 27.8 kg/m² vs 27.5 kg/m²), proportion of diabetes (15.1% vs 16.1%), current smoking (11.0% vs 11.7%), and history of cardiovascular disease (CVD; 5.7% vs 4.5%).

Personnel involved in data collection received specific training in the study procedures. Study participants provided written consent and the study was approved by the clinical research ethics committee of the La Paz University Hospital in Madrid.

Study Variables

Study participants reported their age, educational level, and smoking status. Weight and height were measured in each patient under standardized conditions. BMI was calculated as weight in kilograms divided by height in meters squared. Obesity was defined as BMI ≥30 kg/m². Waist circumference was deemed to be located at the midpoint between the lowest rib and the iliac crest, and was measured with participants lightly clothed using a flexible, inelastic belt‐type tape. Abdominal obesity was defined as a waist circumference >102 cm in men and >88 cm in women. Information on physical activity was obtained with the questionnaire used in the European Prospective Investigation into Cancer and Nutrition (EPIC) study that combines physical activity at work and at leisure time,20 and physical inactivity was defined as being inactive or moderately inactive. Participants also reported if they had ever being diagnosed with CVD, diabetes, or hypertension. Medication use was collected by a face‐to‐face interview and verified against drug packaging during the home visits.

BP Measurement

BP was measured by certified examiners using standardized procedures and conditions.21 Casual BP was measured with validated automatic devices (Omron M6; Omron Healthcare, Lake Forest, IL) and appropriately sized cuffs. BP was determined three times at 2‐minute intervals after the patients rested for 5 minutes in a seated position. In the analyses, casual BP was calculated as the mean of the last two of the three readings.

Thereafter, 24‐hour ABPM was performed with a validated automated noninvasive oscillometric device (Microlife WatchBPO3 monitor; Microlife Corp, Widnau, Switzerland)22 programmed to register BP at 20‐minute intervals during the day and at 30‐minute intervals during the night for the 24‐hour period. Appropriate cuff sizes were used. The majority of registries were performed on working days and the patients were instructed to maintain their usual activities but keep the arm extended and immobile at the time of cuff inflation. The staff of the study returned to the patients’ homes for device removal the following day. Valid ABPM registries had to fulfill a series of pre‐established criteria, including 24‐hour duration and at least 70% successful recordings of systolic BP (SBP) and diastolic BP (DBP) during the day and night.23, 24 Daytime and nighttime periods were defined individually according to the patient's self‐reported time of going to bed and getting up.

Based on casual BP, hypertension was defined as mean SBP ≥140 mm Hg, DBP ≥90 mm Hg, or currently taking prescribed antihypertensive medication.23, 24, 25, 26 Treated hypertension was defined as an affirmative answer to the following questions: “Were you prescribed an antihypertensive medication by your physician?” and “Are you currently taking this BP medication?” Among patients with treated hypertension, BP control was defined as SBP <140 mm Hg and DBP <90 mm Hg, thresholds that were also used for identifying hypertension in untreated patients. We chose this threshold because it corresponds to the universal definition of hypertension and BP target for all ages at the time of the survey,25, 26 it is used in many population surveys,15, 16, 17, 27 and has a consensus ABPM equivalent (130/80 mm Hg for 24‐hour BP).23, 24, 25 Accordingly, ambulatory hypertension was defined as mean 24‐hour SBP ≥130 mm Hg, DBP ≥80 mm Hg, or taking current BP medication; this same value also corresponded to the treatment target. Treatment‐eligible hypertension was defined as either BP above target or BP at goal under drug treatment.

Statistical Analyses

A total of 1047 individuals with ≥70% valid ABPM readings and complete information on study variables were used for analysis (78.8% of all with available ABPM).

We first examined the relationship between casual and ambulatory BP through scatterplots supplemented with Bland‐Altman plots.28 We then classified individuals according to casual BP using two criteria: (1) BP categories (720 patients with hypertension and 327 with normotension), and (2) antihypertensive drug treatment (514 patients with treated hypertension and 533 untreated individuals; the latter including 206 untreated hypertensives and 327 normotensives). We then calculated the percentage (and 95% confidence intervals [CIs]) of hypertensive patients according to casual BP and according to ABPM. Likewise, we calculated the percentage of patients at BP goal according to casual BP and according to ABPM thresholds.23, 24, 25, 26 Although strictly speaking, only treated patients could meet BP goals, from a practical viewpoint untreated patients were included in this definition since they could also be within (rather than achieve) the normal BP range. Next, we calculated the percentage of patients reclassified from above casual BP goal to at 24‐hour BP goal, and those reclassified from at casual BP goal to above 24‐hour BP goal. Results were obtained for the total sample and also stratified by BP medication status. Lastly, we ran several sensitivity analyses to assess the robustness of the main results to alternative BP thresholds proposed by some authors for specific conditions. Thus, we used a casual BP threshold of 140/85 mm Hg, which some guidelines have recommended for people with diabetes.23 Despite BP at home being measured on only one occasion, we also used the BP threshold of 135/85 mm Hg suggested for definition of hypertension based on multiple BP readings at home over several days, and compared it with ambulatory daytime BP.18 We then made an additional sensitivity analysis using the recently proposed clinic BP goal of 150/90 mm Hg for older people,23, 29 and approaching its 24‐hour BP equivalent as 140/90 mm Hg.30

Data are presented as absolute frequencies and percentages for categorical variables and as mean±standard deviation for continuous variables. Differences in sample characteristics between groups were assessed with chi‐square test for categorical variables and Student t test for continuous data. McNemar's chi‐square test was used to compare the proportion of patients classified according to casual vs ABPM methods. Analyses were performed using SPSS version 21 (IBM, Armonk, NY), and statistical significance was set at P<.05.

Results

Sample Characteristics

Table 1 shows the participants’ characteristics. The mean age of the 1047 individuals was 71.7 years, 50.8% were men, mean BMI was 28.1 kg/m², 14.9% had diabetes, and 5.7% had previous CVD. Figure 1 displays the distribution of casual and ambulatory BP. Mean casual BP was 137.8/74.0 mm Hg, mean 24‐hour BP was 123.6/69.8 mm Hg, and mean daytime BP was 127.0/72.4 mm Hg (Table 1). Both SBP and DBP distributions were bell‐shaped. However, as expected in an older population, DBP figures contributed much less than SBP to elevated BP. For example, 47.9% of patients had casual SBP ≥140 mm Hg, 26.9% had 24‐hour SBP ≥130 mm Hg, and 23.2% had daytime SBP ≥135 mm Hg. The corresponding percentage of patients with casual DBP ≥90 mm Hg, 24‐hour DBP ≥80 mm Hg, and daytime DBP ≥85 mm Hg were 10.9%, 7.4%, and 4.5%, respectively.

Table 1.

Characteristics of the Study Participants

	All	Hypertension Status		Treatment Status
Variable	(N=1047)	Normotensive (n=327)	Hypertensive (n=720)	P Value	Untreated (n=533)	Treated (n=514)	P Value
Age, y	71.7±6.3	70.5±5.8	72.2±6.4	<.001	70.9±6.3	72.5±6.3	<.001
Men, %	50.8	48.0	52.1	.222	51.2	50.4	.788
≤Primary studies, %	61.4	61.8	61.3	.872	59.3	63.6	.150
Casual SBP, mm Hg	137.8±18.7	124.1±9.8	144.0±18.5	<.001	135.5±18.2	140.2±18.9	<.001
Casual DBP, mm Hg	74.0±10.3	70.1±8.2	75.8±10.6	<.001	74.1±10.2	73.9±10.4	.765
24‐h SBP, mm Hg	123.6±11.4	117.3±9.9	126.4±10.9	<.001	122.1±11.6	125.2±11.0	<.001
24‐h DBP, mm Hg	69.8±7.0	67.7±6.4	70.8±7.1	<.001	69.9±7.2	69.7±6.8	<.001
Daytime SBP, mm Hg	127.0±11.5	121.6±10.4	129.4±11.2	<.001	126.1±11.8	127.9±11.2	.012
Daytime DBP, mm Hg	72.4±7.5	70.7±6.9	73.2±7.6	<.001	72.8±7.7	72.0±7.2	.073
BMI, kg/m2	28.1±4.6	27.1±4.1	28.6±4.7	<.001	27.2±4.0	29.1±4.9	<.001
BMI ≥30 kg/m2, %	30.9	23.5	34.2	<.001	23.5	38.5	<.001
WC ≥102/88 cm, %	57.6	49.2	61.4	<.001	50.3	65.2	<.001
Physical inactivity, %	41.0	39.1	41.8	0.417	40.3	41.6	.670
Diabetes mellitus, %	14.9	10.4	16.9	<.001	10.9	19.1	<.001
Current smoking, %	11.1	11.6	10.8	.762	13.3	8.8	.06
Previous CVD, %	5.7	6.1	5.6	.718	4.1	7.4	.023

Abbreviations: BMI, body mass index; CVD, cardiovascular disease; DBP, diastolic blood pressure; SBP, systolic blood pressure; WC, waist circumference. For definition of hypertension and treatment status, see the Methods section.

Figure 1.

Distribution of casual, 24‐hour, and daytime systolic blood pressure (SBP).

Hypertensive patients were older, with a higher prevalence of obesity and diabetes than normotensive patients. In addition, treated hypertensive patients were older, with higher mean BMI and higher frequency of obesity, diabetes, and history of CVD than untreated patients (Table 1). Mean casual and ambulatory SBPs were higher in treated hypertensive patients than untreated participants (Table 1). However, compared with the 514 treated hypertensive patients, the 206 untreated hypertensive patients had higher mean casual and ambulatory BPs (data not shown). Among treated patients, 36% were taking angiotensin receptor blockers, 23.9% angiotensin‐converting enzyme inhibitors, 17.3% calcium channel blockers, 20% diuretics, 20.8% β‐blockers, and 4.7% α‐blockers. Overall, 56% of treated patients were on monotherapy, 31% were taking two drugs, and 13% were taking three or more drugs.

Relationship Between Casual and Ambulatory BPs

The scatter diagrams showed only moderate fitting to the regression line (Figure 2, panels a and b). Pearson correlation coefficients were also moderate: 0.62 (0.65 for untreated patients and 0.56 for treated patients) for the association between casual and 24‐hour SBP and 0.56 for the association between casual and daytime SBP. Corresponding coefficients for DBP were similar (data not shown). The Bland‐Altman plots of mean ambulatory SBP against the difference between casual and ambulatory SBPs showed a marked overestimation bias (Figure 2, panels c and d). The mean difference between casual and 24‐hour SBP was 14.3 mm Hg (13.5 in untreated patients and 15.1 in treated patients) and 4.17 mm Hg (4.18 in untreated and 4.16 in treated) for mean difference in casual vs 24‐hour DBP. The mean difference between casual and daytime SBP was 9.9 mm Hg (1.6 mm Hg for DBP).

Figure 2.

Scatterplots (panels a and b) and Bland‐Altman plots (panels c and d) for the association between casual and ambulatory systolic blood pressure (SBP). In panels a and b, solid circles represent treated patients and unfilled circles represent untreated patients. In panels c and d, the solid line indicates the mean SBP difference and dashed lines indicate the 95% limits of agreement (two standard deviations around the mean difference).

Prevalence of Hypertension According to BP Measurement

The prevalence of hypertension was 68.8% (95% CI, 66.0%–71.6%) based on only casual BP and 62.1% based on 24‐hour ABPM (95% CI, 59.2%–65.0%) (Table 2). The difference in hypertension prevalence between casual and 24‐hour BP generally remained across sociodemographic and cardiovascular risk factors such as education level, obesity, physical inactivity, diabetes, smoking, and previous CVD (data not shown).

Table 2.

Prevalence of Hypertension ^a According to Casual and 24‐Hour Ambulatory BP Among Older Adults From the General Population of Spain

Characteristics	No.	Based on Casual BP	Based on 24‐Hour BP	Absolute Difference	P Value
		Percentage (95% CI)	Percentage (95% CI)
Total	1047	68.8 (66.0–71.6)	62.1 (59.2–65.0)	−6.7	.009
Sex
Male	532	70.5 (66.6–74.4)	63.3 (59.2–67.4)	−7.2	.113
Female	515	67.0 (62.9–71.1)	60.8 (56.6–65.0)	−6.2	.138
Age, y
<70	470	63.2 (58.8–67.6)	54.7 (50.2–59.2)	−8.5	.101
≥70	577	73.3 (69.7–76.9)	68.1 (64.3–71.9)	−5.2	.325

Abbreviation: CI, confidence interval. ^aHypertension: casual blood pressure (BP) ≥140/90 mm Hg or taking current BP medication or 24‐hour BP ≥130/80 mm Hg or taking current BP medication.

BP Reclassification According to Treatment Status

A total of 206 (38.6%) untreated participants were above the casual BP normal value and 103 of these (19.3% of all patients) were above the 24‐hour normal value. Thus, 103 (19.3%) of all participants were above normal values according to casual BP measurement and would be reclassified as at normal BP with ABPM (Table 3). Similarly, 245 (47.7%) treated patients were above casual BP goal, and 121 of these (23.5% of all patients) were above the 24‐hour goal. Thus, 124 (24.1% of all patients) were above the goal according to casual BP and would be reclassified as at BP goal with ABPM (Table 3).

Table 3.

Reclassification of BP Control Status According to Ambulatory Monitoring by Treatment Status

	24‐Hour BP
Casual BP	<130/80 mm Hg	≥130/80 mm Hg	Total
Total	749 (71.5%)	298 (28.5%)	1047 (100.0%)
Untreated	397 (74.5%)	136 (25.5%)	533 (100.0%)
Treated	352 (68.5%)	162 (31.5%)	514 (100.0%)
	At casual and 24‐Hour BP Goal	Reclassified to Above 24‐Hour BP Goal	Not Hypertensive or at Casual BP Goal
<140/90 mmHg
Total	522 (49.9%)	74 (7.1%)	596 (56.9%)
Untreated	294 (55.2%)	33 (6.2%)	327 (61.4%)
Treated	228 (44.4%)	41 (8.0%)	269 (52.3%)
	Reclassified to at 24‐Hour BP Goal	Above Casual and 24‐Hour BP Goal	Above Casual BP Goal
≥140/90 mmHg
Total	227 (21.7%)	224 (21.4%)	451 (43.1%)
Untreated	103 (19.3%)	103 (19.3%)	206 (38.6%)
Treated	124 (24.1%)	121 (23.5%)	245 (47.7%)

Abbreviations: BP, blood pressure; treated, hypertensive patients taking current BP medication; untreated, normotensive and untreated hypertensive patients.

Likewise, 33 (6.2%) untreated participants had normal BP according to casual BP but would be reclassified as at above normal BP with ABPM, and 41 (8.0%) treated patients were at goal with casual BP and would be reclassified as at above BP goal with ABPM (Table 3).

Total reclassifications were 21.7% from above normal casual BP to at normal 24‐hour BP, and 7.1% from at normal casual BP to above normal 24‐hour BP (Table 3).

Casual and 24‐Hour BP Control

A total of 596 patients were within casual BP normal range, for either being normotensive or at treatment goal (56.9%; 95% CI, 53.9%–59.9%) and 749 were at 24‐hour BP normal range or goal (71.5%; 95% CI, 68.8%–74.2%; absolute difference, 14.6%; P<.001). A breakdown by treatment status showed 327 untreated patients at normal casual BP range (61.4%; 95% CI, 57.3%–65.5%, and 397 at normal 24‐hour BP range (74.5%; 95% CI, 70.8%–78.2%; difference, 13.1%; P<.001). On the other hand, 269 treated hypertensive patients were at casual BP goal (52.4%; 95% CI, 48.1%–56.7%), and 352 at 24‐hour BP goal (68.5%; 95% CI, 64.5%–72.5%; difference, 16.1%; P<.001).

Lastly, 522 (49.9%) of all patients (55.2% of untreated and 44.4% of treated) were within both normal casual and ambulatory BP ranges (Table 3).

Treatment‐Eligible Hypertension and BP Control

The 720 casual BP–based hypertensive patients and the 650 twenty‐four–hour BP–based hypertensive patients were eligible for treatment. BP control among these patients with treatment‐eligible hypertension increased from 37.4% (95% CI, 33.9%–40.9%) with casual BP target to 54.1% (95% CI, 50.3%–57.9%) with 24‐hour ABPM target (absolute difference, 16.7%; P<.01) (Figure 3).

Figure 3.

Number of patients eligible for medication treatment and percentage of patients with treatment‐eligible hypertension with blood pressure (BP) at goal, according to casual and 24‐hour BP. Medication‐eligible hypertension is defined as receiving treatment or above goal for each casual BP and 24‐hour BP target. Bars indicate 95% confidence intervals for percentage estimates of BP at goal; casual BP at goal, <140/90 mm Hg; 24‐hour BP at goal, <130/80 mm Hg.

Lastly, among patients with BP above goal with casual BP target, 206 (45.7%) were not receiving antihypertensive medication treatment. Among patients with BP above goal with the 24‐hour ABPM target, 136 (45.6%) were not treated with antihypertensive medication.

Sensitivity Analyses

Under the 140/85 mm Hg casual BP threshold only for people with diabetes, the main results remained materially unchanged (data not shown). Results were also similar when comparing casual BP (140/90 mm Hg threshold) with ambulatory daytime target (135/85 mm Hg) (Table 4). Interestingly, when using the 135/85 mm Hg threshold for both casual and ambulatory daytime BP, an even greater difference was obtained in the frequency of hypertension between both measurement methods (74.5% vs 60.6%, or 13.9%) as compared with the difference based on main targets (6.7%), BP control among treated‐eligible patients doubled (26.3% vs 59.7%), and the total reclassification proportions (15.9% and 2%) was moderately lower than those based on main targets (Table 4). Lastly, according to the casual BP goal of 150/90 mm Hg for older people (and 140/90 mm Hg as 24‐hour BP equivalent), the difference in hypertension frequency based on casual BP (58.9%) vs 24‐hour BP (52.0%) was practically identical to that based on main thresholds used (6.7%), and the total reclassification proportions were 26.0% and 3.6%, close to those from the main targets considered (Table 4); however, control among treatment‐eligible patients was much higher, thus a much larger number of patients would be seen as reaching target.

Table 4.

Impact of Different BP Thresholds on the Prevalence and Control of Hypertension

BP Criteria	Prevalence of Hypertension a	BP at Goal b	Reclassified to at Ambulatory BP Goal c	Reclassified to Above Ambulatory BP Goal d
Casual BP (140/90 mm Hg)	720 (68.8%)	269 (37.4%)	227 (21.7%)	74 (7.1%)
24‐h BP (130/80 mm Hg)	650 (62.1%)	352 (54.1%)
Casual BP (140/90 mm Hg)	720 (68.8%)	269 (37.4%)	266 (25.4%)	71 (6.8%)
Daytime BP (135/85 mm Hg)	635 (60.6%)	379 (59.7%)
Casual BP (135/85 mm Hg)	780 (74.5%)	205 (26.3%)	166 (15.9%)	21 (2.0%)
Daytime BP (135/85 mm Hg)	635 (60.6%)	379 (59.7%)
Casual BP (150/90 mm Hg)	617 (58.9%)	301 (48.8%)	272 (26.0%)	38 (3.6%)
24‐hour BP (140/90 mm Hg)	544 (52.0%)	462 (84.9%)

^aHypertension: casual blood pressure (BP) above goal or taking current BP medication or 24‐hour BP above goal or taking current BP medication. ^bPercentage of patients with treatment‐eligible hypertension with BP at goal according to casual and ambulatory BP. Medication‐eligible hypertension is defined as receiving treatment or above goal for each casual BP and ambulatory BP target. ^cReclassification of individuals from above goal with casual BP to at goal with ambulatory BP. ^dReclassification of individuals from at goal with casual BP to above goal with ambulatory BP.

Discussion

This contemporary population‐based study has comprehensively quantified the proportion of older patients potentially affected by ABPM targets. It has shown that the impact of using ABPM is appreciable. First, the prevalence of hypertension would be overestimated had BP status been assessed with casual BP instead of ABPM. Specifically, 6.7% or one in 15 hypertensive patients according to casual BP would not be considered hypertensive had ABPM been used, suggesting a considerable overdiagnosis when clinic BP is used alone. Extrapolating this 6.7% reduction to the older population of Spain in 2012,31 it would represent a reduction of approximately 700,000 older patients classified as needing BP medication (7.5 million with casual BP and 6.8 million with the 24‐hour BP targets). In addition, the percentage of patients with treatment‐eligible hypertension who met BP goals increased by 16.7% in absolute terms (one in six patients, or about 900,000 patients).

Also, there was a considerable gap between the percentage of hypertensive patients at BP goal with ambulatory vs casual BP (16.1% or one additional hypertensive actually controlled in six treated patients). This conveys an optimistic message to physicians because when BP is measured more accurately, the degree of BP control achieved is clearly higher. However, overall, only half of all patients were at BP goal according to both BP measurement techniques, which is important since some studies have shown that the ability to predict mortality was increased by the combination of in‐office and out‐of‐office BP values.32

Discordance Between Casual and Ambulatory BP: Clinical and Public Health Implications

Given the scarce concordance between casual and ambulatory BP, two main BP phenotypes emerge: white‐coat hypertension (only casual BP above goal) and masked hypertension (only 24‐hour BP above goal) among untreated patients, and “office or casual resistance” and masked uncontrolled hypertension among treated patients.4, 19, 23, 24, 25, 33 Nevertheless, we prefer to place emphasis on patient reclassification to better appreciate the impact of using ABPM. One in five untreated patients would be reclassified as not treatment‐eligible, a concept that is consistent with the generally benign prognosis and conservative therapeutic approach in patients with only elevated casual BP, at least in the absence of additional risk factors.23, 25, 34 Likewise, one in four treated patients would be reclassified as BP at goal, and thus might not need treatment intensification. Methodological differences aside, these proportions are moderately higher than those in other population‐based studies,8, 19, 23, 25 as corresponds to an older population.19, 23, 25 Interestingly, these proportions were lower than those obtained in Spanish older patients in clinical settings,35 probably reflecting the minimization of the alert reaction in our study. On the other hand, assuming that only elevated 24‐hour BP has a more serious prognosis and thus deserve more aggressive management,23, 25, 34 one in 16 untreated patients would be reclassified as treatment‐eligible, and one in 12 treated patients would be reclassified as possibly needing intensification of treatment. The proportion of isolated ambulatory hypertension cases in this study is at the lower end of the range reported in other population studies,8, 9, 12, 19, 23, 25 which is consistent with generally lower frequency in the elderly.19, 33, 36 Given the lack of clear indications for treatment of these discordant clinical entities,23, 25, 34, 37, 38 the therapeutic implications of reclassification have been presented only as an illustration of the potentially significant overtreatment or undertreatment if using only a more inaccurate technique of measuring BP.

Overall, the number of patients who were still considered to have above‐goal BP (28.5% of all patients) outnumbered the number of patients reclassified as at goal with ABPM target (21.7%), and almost half of patients with BP above goal with either casual or 24‐hour targets were untreated.

All of this supports the NICE statement on offering ABPM for the diagnosis of hypertension after an initial raised reading in the clinic. ABPM would reduce misdiagnosis, ensure that the right people were treated with antihypertensives, and reduce the number of patients treated for hypertension.14, 39 Although not universally accepted,24 this could save costs since the extra expenditures from ABPM are offset by cost savings from better targeted treatment.40 Furthermore, ABPM is well tolerated by patients, including the elderly.41, 42 Likewise, given these and other advantages of ABPM, some authors have proposed that the US Centers for Disease Control and Prevention include ABPM in the NHANES and that the US Food and Drug Administration require the use of ABPM as the gold standard for recording BP in randomized clinical trials.43 Yet, it is well known that ABPM is not available to most patients with hypertension and wonder whether it will not take a medicolegal challenge to make the technique universally available.

Study Strengths and Limitations

Given that this study was not strictly representative of the general older population of Spain, extrapolations should be interpreted with caution. Nevertheless, the baseline sociodemographic and clinical characteristics of the participants at the inception of the cohort were reasonably similar to those who did not participate (age, 68.6 vs 69.4 years; men, 48.1% vs 45.0%; mean casual BP, 139.7/77.5 mm Hg vs 139.2/77.1 mm Hg; mean BMI, 28.6 kg/m² vs 28.6 kg/m²; diabetes, 15.2% vs 18.0%; and previous CVD, 5.7% vs 5.9%).

Of note, as in some other population surveys,15, 16, 17 casual BP was not measured in the office but in patients’ homes, thus likely diminishing a reaction alarm and giving more realistic estimates. However, only a few BP readings were taken on a single occasion by observers who were not familiar to the patients. Certainly other methods such as automated office BP monitoring with the patient alone can also minimize anxiety‐related increases in BP,44 and some population studies have used office, home, and ABPM measurements.8, 9 Nevertheless, the present study was specifically focused on the direct comparison between the two out‐of‐office BPs individually, which may provide a more accurate classification of the BP status. Given that the present analyses suggest that casual BP measurements are a biased estimate of ambulatory BP, it is likely that having multiple BP measurements would not have minimized the beneficial impact of ABPM.

The issue of selecting BP threshold is far from settled. Nevertheless, despite our results being sensitive to different BP thresholds, their direction is consistent. Under all alternative thresholds, casual BP overestimates the true (ambulatory) hypertension prevalence and greatly underestimates the BP control among treatment‐eligible patients; therefore, upward and downward BP reclassification is noticeable. This suggests the potential benefit of ABPM as compared with casual BP.

Lastly, antihypertensive therapy was based on the participant's declaration and therefore may be imprecise. In addition, treatment adherence was not assessed, but it has been reported to be relatively high in Spain (68%).45

Conclusions

We estimated that the application of ABPM would potentially reduce the number of older persons for whom hypertension therapy would be recommended by approximately 7%. In addition, about 17% of treatment‐eligible patients would no longer be classified as having poorly controlled BP, but instead would be considered adequately managed. This surely makes the case for mandatory ABPM in this age group. However, even under the ABPM target, more than one fourth of older patients with hypertension still have uncontrolled BP, and approximately half of them remain untreated. Reclassification of patients based on ABPM illustrates the clinical and public health implications of targeting treatment in patients who have actually elevated BP and avoiding treatment in those who do not. This supports the usefulness of measuring BP with ABPM.46

J Clin Hypertens (Greenwich). 2015;17:453–461. DOI: 10.1111/jch.12525. © 2015 Wiley Periodicals, Inc.