256
Clinical methods and pathophysiology
Blood pressure measurement reliability among different racial-ethnic groups in a stroke prevention study Conrado J. Estolg, Philip M.W. Batha, Philip B. Gorelickb, Daniel Cottonc, Renee H. Martind, Michael A. Webere, Bjorn Dahloff and PRoFESS Publications Committee and PRoFESS Investigators Objective High blood pressure (BP) is commonly not diagnosed, and patients do not achieve target values when treated. Among 20 000 patients encompassing most races-ethnicities, we evaluated BP measurements and treatment response in a stroke prevention trial. Our goal was to identify BP measurement differences between clinical trial and patient determinations and among the racial-ethnic groups. Materials and methods A total of 20 332 patients with ischemic stroke were randomized to receive antiplatelet treatment and 80 mg of telmisartan versus placebo. BP measurements were obtained at the first clinic visit and then 1 and 3 months later and every 6 months thereafter. One week after the first clinic visit, patients were requested to report a BP measurement obtained elsewhere. Measurements at the trial clinics were obtained with the same electronic device. Statistical analysis was used to detect significant differences. Results The mean patient age was 66 years; 36% were women, and race-ethnicity comprised 58% Whites, 33% Asian, 4.9% Hispanic, and 4% Black. Overall, 74% of patients were hypertensive. BP varied between the raceethnicity groups, being highest in Hispanics (145/85) and lowest in Blacks (144/82). BP at visits clinic 1, nonclinic 1A, and clinic 2 were, respectively, 144/84, 137/80, and 139/81 mmHg, with the difference between visits 1–2 and
Introduction There are more than one billion people with hypertension (HTN) worldwide, and HTN is the most important modifiable vascular risk factor accounting for seven million deaths per year (13% of all deaths). Diagnosis and treatment of HTN are widely available. However, the rule of two-third applies with only two out of three people with HTN being diagnosed: two-third of these receive treatment and only two-third of these have their blood pressure (BP) controlled [1,2]. In a recent analysis, the prevalence of HTN among all people age 18 years or older during the period 2005–2008 was 30% and the lifetime risk of developing HTN at age 50 was 90% [3]. The population-attributable risk of HTN is 54% for stroke and 47% for coronary heart events; BP control has a direct relation with decreasing stroke risk [4,5]. Although variations in HTN severity and incidence have been reported among various world regions, studies have shown c 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins 1359-5237
visit 1A being significant. BPs were normal in 42% of the cases at visit 1A, and of these, only 44% were normal at visit 1 and 57.6% were normal on visit 2. Similar findings were noted for all race-ethnicity groups. Conclusion BP values varied among race-ethnicities and showed differences between clinic and patient measurements. This finding questions the reliability of self-reported BP and has implications for BP management in daily clinical practice. Blood Press Monit c 2014 Wolters Kluwer Health | Lippincott 19:256–262 Williams & Wilkins. Blood Pressure Monitoring 2014, 19:256–262 Keywords: blood pressure, hypertension, PRoFESS, race-ethnicity, stroke a
Nottingham City Hospital, Nottingham, UK, bSaint Mary’s Health Care, Grand Rapids, Michigan, cBoehringer Ingelheim Pharmaceuticals, Ridgefield, Connecticut, dDepartment of Biostatistics, Bioinformatics and Epidemiology, Medical University of South Carolina, Charleston, South Carolina, e910 Fifth Avenue, 6A, New York, NY, USA, fDepartment of Molecular and Clinical Medicine, Su Sahlgrenska, Go¨teborg, Sweden and gNeurological & Vascular Prevention Unit, Buenos Aires, Argentina Correspondence to Conrado J. Estol, MD, PhD, FAAN, Neurological & Vascular Prevention Unit, Av. Callao 875, C1023AAB Buenos Aires, Argentina Tel: + 54 11 4816 6733; fax: + 54 11 4816 6668; e-mail: [email protected] Received 3 November 2013 Revised 27 February 2014 Accepted 6 March 2014
that the same types of vascular risk factors similarly affect all populations across the world and that few risk factors account for the majority of new cardiovascular events [6–8]. However, differences in BP values and the reliability of BP control among different race-ethnicities have not been widely studied. Our aim was to compare BP measurements obtained at the PRoFESS (Prevention Regimen for Effectively Avoiding Secondary Strokes) trial clinics with one measurement reported by the patients for the detection of discrepancies to assess the reliability of BP measurements and search for differences among the multiple race-ethnicities studied.
Materials and methods We evaluated BP measurements in the 20 332 patients included in the secondary stroke prevention trial DOI: 10.1097/MBP.0000000000000045
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Blood pressure control reliability Estol et al. 257
PRoFESS, an international, multicenter, randomized, double-blind, 2 2 factorial trial comparing the combination of acetylsalicylic acid (aspirin) and extended-release dipyridamole versus clopidogrel, and the angiotensin receptor blocker telmisartan versus placebo. All patients received additional antihypertensive medication as indicated by the investigator. Details from the PRoFESS protocol design and overall results have been published previously [9,10]. From 2003 to 2006, PRoFESS enrolled 20 332 patients from 695 centers in 35 countries from North and Latin America, Occidental and Eastern Europe, Africa, and Australasia who had suffered a stroke within the previous 120 days. Patients with a hemorrhagic stroke on neuroimaging were excluded. Ethical approval was obtained in each participating country as per the local regulatory authorities’ requirements. All patients signed a written consent form in compliance with GCP regulations. The average patient follow-up was 2.5 years with a minimum of 18 months. After the baseline visit, patients were seen at 1, 3, and 6 months and then every 6 months until follow-up completion (visits 1, 2, 3, 4, etc., respectively). BP was measured with the same model electronic OMRON 705CP device (OMRON Corporation, Kyoto, Japan) in all 695 participating centers. Two measurements taken 2 min apart were averaged and the BP was recorded. At 1 week after the first clinic visit (visit 1), in what was defined as visit 1A, patients were requested to report to the investigator a BP value self-measured at home or obtained elsewhere (by a caregiver, the patient, a physician, a nurse, a healthcare worker, or others). Roughly 30% of PRoFESS patients were hospitalized at the baseline visit. In case patients were in the hospital at the seventh day after randomization, BP for visit 1A was obtained with the OMRON device before the intake of telmisartan. Patients were given the PRoFESS medication on visit 1, and thus half of the patients at visit 1A were receiving telmisartan 80 mg in addition to their baseline treatment. We separately compared the BP at visit 1A in the patients receiving placebo and telmisartan in relation with BP values obtained at visit 1 (no telmisartan) and subsequent clinic visits (50% of patients on telmisartan). In addition, we also compared the BP values excluding the patients who were hospitalized at the time of visit 1A. Data from PRoFESS provides strict and sequential BP measurements over 2–4 years in these patients. Race-ethnicity
As ‘ethnicity’ is a heterogeneous concept, difficult to define and strongly related to culture, language, geography, and other variables, we decided to classify patients under the more comprehensive race-ethnicity origin. The race-ethnicity for each patient was decided among the available options: European Whites, Black African, Caribbean Hispanic, Native Latin, South Asian, Chinese, Japanese, Malays, other Asian, Arab, Persian, and other.
Statistical analysis
Average systolic and diastolic BP levels were obtained in all racial-ethnic groups for all trial clinic visits and at visit 1A (outpatient/in-hospital visit at day 7) comparing clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A using paired t-tests. Furthermore, the percentage of normal BP levels (r 135/85 mmHg) for each visit was determined and then a comparison between rates of normal BPs at visits 1 and 2 and those at visit 1A was performed.
Results Among 20 332 patients enrolled in PRoFESS, 58% (11 697) were of European/White ethnicities, 33% (6660) were Asian, 4.9% (990) were Latin/Hispanic, and 4% (816) were Black African (94% African American). HTN was diagnosed in 74% of the patients. The average baseline BP reported in PRoFESS was 144/84 mmHg. Detailed baseline patient characteristics (age, sex, BMI, smoking, HTN, etc.) were reported previously [10]. Results of paired comparisons of BP measurements at clinic visits by the investigator and BP obtained by the patient elsewhere or in the hospital showed a significantly lower systolic and diastolic BP for visit 1A compared with visit 1. The mean systolic BP was 144 mmHg at visit 1 and 136.9 mmHg at visit 1A (P < 0.0001). The mean diastolic BP was 83.8 mmHg at visit 1 and 79.9 mmHg at visit 1A (P < 0.0001) (Tables 1 and 2). This comparison was repeated up to the fifth clinic visit. The comparison of systolic BP in visit 1A with clinic visit 2 was significantly lower for Whites and Latins, but not for Blacks and Asians. In visit 3, at 3 months, it was not significant for Blacks and all differences were significant when comparing visit 1A with visit 4 at 6 months. For diastolic BP measurements, the difference between visits 1, 2, 3, and 4 with visit 1A was significant (lower in 1A) except for Asians in clinic visit 2. The differences found would not be clinically meaningful if all values in visits 1 through 4 and visit 1A were normal or abnormal irrespective of the presence of a statistically significant difference in BP values. Therefore, we investigated whether the lower BPs in visit 1A were in the normal range or the abnormal range in parallel with BPs recorded at clinic visits 1 to 4. Considering all race-ethnic groups, 28.4% had normal BP values at visit 1, but in visit 1A the proportion of normal BP values increased to 42%. Table 3 shows the percentage of normal BP values for each visit (clinic and 1A) in each race-ethnic group. Table 4 shows the proportion of patients with normal BP at visits 1 and 2 of those who had normal BP at visit 1A. Compared with normal BP values in visit 1A, the percentage of normal BP measurements at clinic visit 1 ranged from 39.4 to 49.2%. The same comparison revealed that 52.8–59.2% of normal BPs at visit 1A were normal at visit 2. Even though the significant number of lower BP values at visit 1A persisted when comparisons were made with visits 2–4
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
258
Blood Pressure Monitoring 2014, Vol 19 No 5
Table 1
Systolic blood pressure levels at the clinic visits by race-ethnicity Visit 1 (baseline)
Ethnicity Asian (33.03%) European/Whites (58.01%) Black (African/ American) (4.05%) Latin Hispanic (4.91%) Mean BP
Visit 1A (1 week)
P*
Visit 2 (1 month)
Visit 3 (3 months)
P*
Visit 4 (6 months)
P*
Visit 5 (12 months)
P*
P*
144.5 (16) < 0.0001 136.9 (16.7) 137.4 (18.9) 0.1025 136.1 (18.3) 0.0024 135.7 (17.4) < 0.0001 136.3 (17.4) 0.0246 143.9 (16.6) < 0.0001 137 (18.3) 139 (19.5) < 0.0001 138.4 (19.4) < 0.0001 138.4 (19.1) < 0.0001 138.2 (18.9) < 0.0001 143.6 (18.2) < 0.0001 140.1 (20.2) 140.3 (23.6) 144.8 (17.7)
0.3323 141.1 (23.1)
< 0.0001 133.8 (19.6) 140.2 (23.4) < 0.0001 140.7 (23)
0.3456 141.5 (22.2)
0.0258 139.5 (21.6)
0.9227
< 0.0001 139.4 (21.8) < 0.0001 140.0 (22.7) < 0.0001
144.1 (16.5) < 0.0001 136.9 (17.9) 138.6 (19.7) < 0.0001 137.8 (19.5) < 0.0001 137.7 (18.9) < 0.0001 137.7 (18.8) < 0.0001
*P-values compare values at clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A (7 days) using paired t-tests. There are 20 comparisons on each table, and thus only P-values < 0.0025 are considered statistically significant using the Bonferroni adjustment. BP, blood pressure.
Table 2
Diastolic blood pressure levels at the clinic visits by race-ethnicity in the placebo and the telmisartan groups
Ethnicity
Visit 1
Asian (33.03%) European/Whites (58.01%) Black (African/American) (4.05%) Latin Hispanic (4.91%) Mean BP
P*
Visit 1A
Visit 2
P*
Visit 3
P*
Visit 4
P*
Visit 5
P*
84.9 (9.9) < 0.0001 81.40 (10.0) 81.3 (11) 0.6095 80.5 (10.4) < 0.0001 80.5 (10.2) < 0.0001 80.5 (10.1) < 0.0001 83.2 (10.7) < 0.0001 79.18 (11.0) 80.1 (11.6) < 0.0001 80.7 (11.5) < 0.0001 80.6 (11.5) < 0.0001 80.5 (11.4) < 0.0001 82.3 (10.7) < 0.0001 80.02 (12.0) 81 (13) 0.0216 81.4 (12.8) 0.0151 81.9 (12.7) 0.0028 80.6 (12.6) 0.3618 85 (11) < 0.0001 78.51 (11.1) 82.1 (13.2) < 0.0001 82.4 (13) < 0.0001 81.4 (12.5) < 0.0001 81.9 (12.8) < 0.0001 83.8 (10.5) < 0.0001 79.94 (10.7) 81.1 (11.6) < 0.0001 80.8 (11.3) < 0.0001 80.7 (11.2) < 0.0001 80.6 (11.1) < 0.0001
*P-values compare values at clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A (7 days) using paired t-tests. There are 20 comparisons on each table, and thus only P-values < 0.0025 are considered statistically significant using the Bonferroni adjustment. BP, blood pressure.
Table 3
Normal blood pressure levels at the clinic visits by race-ethnicity
Ethnicity
Visit 1 [N (%)]
Asian (33.03%) European/Whites (58.01%) Black (African/American) (4.05%) Latin Hispanic (4.91%) Total
Table 4
1747 3463 248 276 5734
(26.2) (29.6) (30.4) (27.9) (28.4)
Visit 1A [N (%)] 2798 4937 264 479 8478
Visit 2 [N (%)]
(42) (42.2) (32.3) (48.4) (42)
2765 4571 310 380 8026
(41.5) (39) (37.1) (38.4) (39.8)
Visit 3 [N (%)] 2913 4617 259 366 8155
(43.7) (39.5) (31.7) (36.1) (40.4)
Visit 4 [N (%)] 2884 4530 241 366 8021
(43.3) (38.7) (29.5) (36.9) (39.8)
Visit 5 [N (%)] 2743 4414 235 346 7738
(41.2) (37.7) (28.8) (34.9) (38.4)
Comparison of normals at visit 1A with normals at visits 1 and 2
Ethnicity Asian (33.03%) European/Whites (58.01%) Black (African/American) (4.05%) Latin Hispanic (4.91%) Total
Visit 1A normals Of those normal at visit 1A, how many normal at visit 1? Of those normal at visit 1A, how many normal at visit 2? [N (%)] [N (%)] [N (%)] 2798 (42) 4937 (42.2) 264 (32.3)
1161 (41.5) 2242 (45.4) 130 (49.2)
1659 (59.3) 2817 (57) 155 (58.7)
479 (48.4) 8478 (42)
189 (39.5) 3722 (43.9)
253 (52.8) 4884 (57.6)
when 50% of patients were also receiving telmisartan, to avoid potential bias from patients not receiving telmisartan in visit 1, we performed the BP comparisons among visits including only patients in the placebo group. Results are presented in Tables 5 and 6 for systolic and diastolic BPs, respectively. Differences between trial visits 1 and 2 and the BP provided by patients at visit 1A continued to be significant with higher values recorded at PRoFESS clinics. Of note, the mean BP for all groups continued to be lower at visit 1A until trial visit 4. The difference between the clinic and the reported BPs was the largest for the Latin Hispanic group; the difference was significant for diastolic but not for systolic
BP after visit 1 in Blacks, and the difference was significant in Asians, but, surprisingly and unlike for all other groups, from visit 3 onwards, systolic and diastolic BPs were lower at clinic visits compared with the BP reported by patients at visit 1A. Also, to decrease bias, we repeated the comparisons avoiding inclusion of the patients who were admitted in the hospital at the time of visit 1A (B30% of all patients) as the BP in these patients does not reflect the typical outpatient BP measurement context. Results are presented in Tables 7 and 8 for systolic and diastolic BPs, respectively. Mean BP measurements including all race-ethnic groups were significantly higher in clinic visits 1 and 2 compared with
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Blood pressure control reliability Estol et al.
Table 5
259
Systolic blood pressure levels at the clinic visits by race-ethnicity in the placebo group
Ethnicity
Pa
Visit 1
Asian (N = 3328) European/Whites (N = 5870) Black (African/ American) (N = 409) Latin Hispanic (N = 496) Mean BP
Visit 1A
Visit 2
Pb
Visit 3
Pb
Visit 4
Pb
Visit 5
Pb
144.4 (16.1) < 0.0001 138.9 (16.3) 140 (18.2) 0.0008 138.5 (17.5) 0.4337 137.8 (16.8) 144.1 (16.8) < 0.0001 139.7 (17.9) 141.7 (19) < 0.0001 141.1 (19.1) < 0.0001 140.8 (18.6)
0.0016 138.3 (16.9) 0.0001 140.2 (18.5)
0.1439 0.0534
144.3 (18.5)
0.0714 142.2 (21.3)
0.6803
143.9 (17.7)
0.0205 142.3 (19.9) 143.7 (23.7)
0.0723 143.5 (22)
0.2488 143.9 (22.1)
< 0.0001 135.8 (19.4) 142.3 (22.6) < 0.0001 142.2 (22.4) < 0.0001
141 (20.9) < 0.0001 141.8 (23.5) < 0.0001
144.2 (16.7) < 0.0001 139.3 (17.6) 141.2 (19.2) < 0.0001 140.4 (18.9) < 0.0001 139.9 (18.4)
0.0069 139.7 (18.4)
0.0511
P-values compare values at clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A (7 days) using paired t-tests. There are 20 comparisons on each table, and thus only P-values < 0.0025 are considered statistically significant using the Bonferroni adjustment. BP, blood pressure. a Visit 1 versus 1A (paired t-test). b Visit 2, 3, 4, 5 versus 1A (paired t-test).
Table 6
Diastolic blood pressure levels at the clinic visits by race-ethnicity in the placebo group
Ethnicity
Visit 1
Asian (N = 3328) European/Whites (N = 5870) Black (African/American) (N = 409) Latin Hispanic (N = 496) Mean BP
P*
Visit 1A
Visit 2
P*
Visit 3
85 (10) < 0.0001 82.5 (10) 82.8 (10.8) 0.1757 83 (10.7) < 0.0001 80.5 (10.7) 82.3 (11.5) < 0.0001 82.2 (10.9)
0.0205 80.8 (12)
82.7 (13)
0.0037
81.9 (10) 82 (11.4) 82.3 (12.7)
P*
Visit 4
P*
Visit 5
P*
0.0038 81.6 (9.9) < 0.0001 81.5 (9.9) < 0.0001 < 0.0001 81.8 (11.2) < 0.0001 81.6 (11.3) < 0.0001 0.0462
83 (12.6)
0.0091 81.9 (12.4)
0.2773
84.4 (10.8) < 0.0001 79.5 (11) 83.3 (12.5) < 0.0001 83.30 (12.96) < 0.0001 82.1 (11.9) 0.0009 83.3 (12.9) < 0.0001 83.8 (10.5) < 0.0001 81.1 (10.6) 82.5 (11.3) < 0.0001 82 (11.1) < 0.0001 81.8 (10.9) < 0.0001 81.6 (11) 0.0010
*P-values compare values at clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A (7 days) using paired t-tests. There are 20 comparisons on each table, and thus only P-values < 0.0025 are considered statistically significant using the Bonferroni adjustment. BP, blood pressure.
Table 7
Systolic blood pressure levels at the clinic visits by race-ethnicity in the placebo group for only those not hospitalized
Ethnicity
Visit 1
Asian (N = 2085) European/Whites (N = 3897) Black (African/ American) (N = 327) Latin Hispanic (N = 432) Mean BP
Pa
Visit 1A
Visit 2
Pb
Visit 3
Pb
Visit 4
Pb
Visit 5
Pb
143.3 (16.4) < 0.0001 138.7 (16.6) 139.1 (18.2) 0.3244 137.6 (17.1) 143.5 (17.3) < 0.0001 139.6 (18.2) 141.3 (19) < 0.0001 140.1 (19.1)
0.0166 137.2 (17.2) 0.0011 138.1 (17.3) 0.0960 139.9 (18.5) 0.3214 139 (18)
0.2524 0.1001
144.4 (18.4)
0.4569 143.5 (22.7) 0.2424 141.2 (20.6)
0.8365
0.0799 142.4 (20.3) 143.5 (23.9)
0.2048 142.9 (21.7)
143.7 (18.1) < 0.0001 136.4 (19.3) 143.1 (22.6) < 0.0001 142.5 (22.5) < 0.0001 141.2 (21.1) 0.0002 142.1 (23.8) < 0.0001 143.5 (17.1)
< 0.0001 139.3 (17.9) 140.8 (19.3) < 0.0001 139.6 (18.9)
0.1065 139.3 (18.5) 0.6654
139 (18.3)
0.5726
P-values compare values at clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A (7 days) using paired t-tests. There are 20 comparisons on each table, and thus only P-values < 0.0025 are considered statistically significant using the Bonferroni adjustment. BP, blood pressure. a Visit 1 versus 1A (paired t-test). b Visits 2, 3, 4, 5 versus 1A (paired t-test).
Table 8
Diastolic blood pressure levels at the clinic visits by race-ethnicity in the placebo group for only those not hospitalized
Ethnicity Asian (N = 2085) European/Whites (N = 3897) Black (African/American) (N = 327) Latin Hispanic (N = 432) Mean BP
Visit 1
P*
Visit 1A
Visit 2
P*
Visit 3
P*
Visit 4
P*
Visit 5
84.9 (10.1) < 0.0001 82.6 (10) 82.5 (10.8) 0.7912 81.7 (9.9) 0.0006 81.3 (9.9) < 0.0001 81.5 (10) 82.6 (10.7) < 0.0001 80 (11) 81.6 (11.4) < 0.0001 81 (11.2) < 0.0001 80.8 (11) 0.0009 80.5 (11.1) 82.4 (10.8) 0.0799 80.8 (11.8) 82.1 (13.1) 0.0360 81.7 (12.7) 0.2583 82.8 (12.9) 0.0392 81.4 (12.3) 84.4 (11) < 0.0001 79.9 (10.9) 83.7 (12.4) < 0.0001 83.5 (12.7) < 0.0001 82.2 (11.9) 83.4 (10.6) < 0.0001 80.8 (10.7) 82 (11.3) < 0.0001 81.4 (11) 0.0003 81.1 (10.8)
P* 0.0003 0.0819 0.6318
0.0050 83.7 (12.7) < 0.0001 0.1455 81.1 (11) 0.3911
*P-values compare values at clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A (7 days) using paired t-tests. There are 20 comparisons on each table, and thus only P-values < 0.0025 are considered statistically significant using the Bonferroni adjustment. BP, blood pressure.
the BP at visit 1A, but this difference was not significant at clinic visit 3 and beyond. Again, the Latin Hispanic group revealed the largest differences between clinic BP
and BPs from visit 1A. In Blacks, differences were significant for some of the visits in both systolic and diastolic values and Asians again had significantly lower
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
260
Blood Pressure Monitoring 2014, Vol 19 No 5
values at clinic visit 3 and beyond for both systolic and diastolic values. Table 9 shows the differences found between the Hispanic groups from Argentina and those classified as ‘other countries’ (the majority from Brasil, Mexico, and the USA). Differences in systolic and diastolic BPs are significant (P < 0.0001), although no significant difference was found in the prevalence of left ventricular hypertrophy. Table 10 shows BP differences among the three largest Asian groups (South East Asia, Chinese, and ‘other’). Again, systolic and diastolic BP differences were significant (P < 0.0001), whereas left ventricular hypertrophy did not show significant differences.
Discussion We found that across all race-ethnicities, most BP measurements obtained outside the PRoFESS clinics 1 week after the first trial visit (visit 1A) were significantly lower than the first two clinic determinations (1 week before and 3 weeks after the measurement ‘elsewhere’ at visit 1A). This finding was consistent when evaluating the entire PRoFESS cohort, when separately comparing patients in the placebo and the telmisartan groups and when excluding patients who were hospitalized at the time of the first two visits (1 and 1A). In all groups (except for the diastolic BP in the nonhospitalized and the placebo groups in Asians during visit 2), the mean systolic and diastolic BPs were significantly lower in visit 1A compared with visits 1 and 2 and lower in general when compared with subsequent clinic visits. Our study included all major race-ethnicities around the world. It can be speculated that the differences we found within the Latin and the Asian race-ethnic groups could be secondary to genetic or socio-economic factors particular to the different regions where these populations live (Tables 9 and 10). For example, Latins living in a medium-income country such as Argentina are likely to have, on average, less education and a more limited access to adequate medical services compared with Latins living in the USA. Table 9
Different issues that explain poor control of BP include inadequate medical and public education to recognize this as a major risk factor for death worldwide, a high degree of noncompliance reaching 50%, use of subtherapeutic doses of antihypertensive medication, neglect in recognizing slight BP elevations as important and inadequate BP measurements [11]. We focused our analysis on the latter and thus we aimed to detect significant differences between BPs measured by trial physicians with the standard electronic machine selected for PRoFESS and those from visit 1A, which encompassed BPs obtained by patients outside the trial clinic. Our hypothesis was that outside a strictly controlled environment, BP measurements, including those made by medical and paramedical personnel, frequently provide falsely normal or only slightly elevated values. Unreliable measurements by patients are frequent, and in a study on individuals who were unaware that their BP measurements were stored in an electronic BP machine, the authors found the report of fictional BP values made up by study subjects and the report of the closest to normal value after the patients repeatedly measured their BP [12]. Use of aneroid and mercury devices for BP determinations in visit 1A may account for some of the differences found between trial clinic and visit 1A measurements as available data support the use of electronic devices based on their greater reliability and reproducibility of BP measurements [13]. Interestingly, excluding patients who were hospitalized, 7806 (57.7%) of the BP measurements at visit 1A were made by healthcare professionals not related to the PRoFESS trial and 5324 (39.3%) were taken by patients or their caregivers. Also, most patients reported a BP value from a single measurement, which should yield a higher result compared with the average of two determinations (where the second value is usually lower than the first) obtained at the PRoFESS clinics for all patients. Rather than assuming that BPs provided by patients from visit 1A were falsely low, it could be argued that BP measurements at PRoFESS centers were elevated (relative to visit 1A values) due to the ‘white coat effect’.
Differences between Hispanics from Argentina and those from other countries (most from North America)
Blood pressure and LVH Systolic BP Diastolic BP LVH
Latin Hispanic (4.91%)
Argentina (N = 404)
Other countries (N = 586)
144.8 (17.7) 85 (11) 157 (17.3%)
148.6 (17.7) 88.2 (10.9) 16.8%
142.2 (17.3) 82.8 (10.6) 17.6%
BP, blood pressure; LVH, left ventricular hypertrophy.
Table 10
Differences among Asians from different regions
Blood pressure and LVH Systolic BP Diastolic BP LVH
Asian (All) (33.03%)
South Asian (N = 1712)
Chinese (N = 3666)
Other Asian (N = 1282)
144.5 (16) 84.9 (9.9) 952 (14.3%)
139.9 (15.0) 84.5 (9.0) 14.2%
146.9 (15.9) 85.1 (10.2) 13.9%
143.7 (16.1) 85.1 (10.4) 17.1%
BP, blood pressure; LVH, left ventricular hypertrophy.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Blood pressure control reliability Estol et al. 261
However, numerous studies have proven this notion wrong by providing clear evidence that the rate of cardiovascular events in patients whose BP was increased in response to stress was the same as in those with persistent HTN in 24-h Holter studies [14–20]. A recent study conducted in 11 Australian centers on 8575 patients with a goal of identifying ambulatory BP equivalents of clinic BP for the diagnosis of HTN found small BP differences between clinic and ambulatory values [21]. These findings support the notion that large differences in BP cannot be attributed to a ‘white coat’ effect and are rather due to abnormally elevated BP in the clinic (or falsely low determinations elsewhere). Owing to our large sample size, differences in variables are likely to be statistically significant, and for this reason, it is important to establish the clinical relevance of these differences. Data from the Framingham and other studies have shown that most vascular events occur with slight increases in BP and that even a moderate decrease in BP causes a significant reduction in cardiovascular events [22–26]. The fact that more than 75% of cardiovascular events occur in ‘low-risk and intermediate-risk’ groups attests the importance of rigorous control of all and every degree of vascular risk factors [27]. A recent meta-analysis involving 70 664 patients revealed a statistically significant (relative risk 0.78; 95% confidence interval 0.71–0.86; P < 0.000001) 22% stroke risk reduction with BP treatment in people with pre-HTN (i.e. those with BP in the range of 120–139/80–89 mmHg) [28]. The study continued to be significant when using a lower baseline BP (< 130/85). Aforementioned data sustain that even BP values that are slightly elevated result in a significant number of vascular events in the long term, and thus strict BP measurements to detect these small variations are warranted. To evaluate the potential confounding effects of our large sample size, we investigated the differences in BP between clinic visits and visit 1A in a smaller group of PRoFESS patients from two high-recruiting sites in Argentina. Among 190 patients, BP was higher than 140/90 mmHg in 86% of the patients at visit 1 (and 61% had a BP greater than 160/100 mmHg), in 79% of them at clinic visit 2, in 83% of them at clinic visit 3, in 68% of them at visit 4, and in 72% of them at visit 5. In contrast, only 38% reported HTN at visit 1A (measurements outside the clinic), none with BP greater than 160/100 mmHg and 44% with values lower than 120/80 mmHg. The analysis of this restricted sample of patients reveals, as expected, wider differences than those observed in the entire 20 332 PRoFESS cohort (C. Estol, unpublished observation).
part of the PRoFESS treatment. However, further comparisons with patients at visits 2–4, of whom 50% were also receiving telmisartan, also confirmed a significantly lower BP at visit 1A. We noted that of a total of 8478 normal BP measurements at visit 1A, only 57.6% BPs were normal at visit 2. Again, this suggests that there were a significant proportion of falsely normal BP values at visit 1A. This finding persisted with most visits from 1 through 4 (the latter 6 months after visit 1) when we compared BP results (Tables 3 and 4). The fact that by visit 4 (at 6 months from visit 1), all BP measurements (except in the Asian group) were higher compared with visit 1A has important clinical implications regarding the limitations of BP control, especially in the context of the strict frame of a clinical trial. Among other reasons, it may be that falsely low or normal ambulatory BP recordings mislead patients and physicians to believe that BP is controlled [29]. A significant number of trials have failed to report positive outcomes attributed to ‘minimal or no differences’ in BP between the treatment and the placebo groups. It may be that patients in a secondary stroke prevention clinical trial could feel under ‘pressure’ to report lower BPs or simply that common mistakes in routine BP measurements (underinflation of cuff in patients with high BP, rounding-down of BP results, etc.) result in falsely low values. BP values that were persistently higher than guideline recommendations could be explained by the fact that PRoFESS was not a true BP-lowering study, but rather an add-on antihypertensive drug and antiplatelet study, and as such, investigators were not motivated enough to strictly achieve target BP values. The recent SPS3 trial clearly reflects the significant resources needed and difficulties encountered in achieving a stringent target BP value [30]. Our findings of unreliable BP measurements are likely to be greater in regular clinical practice. In PRoFESS, the complexity of BP management became apparent with BP values that were falsely low when measured outside the trial setting and BP that persisted elevated in sequential clinic visits despite medical treatment.
Acknowledgements This work was supported by Boehringer Ingelheim Pharmaceuticals Inc. (BIPI). BIPI was given the opportunity to check the data used in the manuscript for factual accuracy only. Conflicts of interest
A main objective in our analysis was to determine whether the greater proportion of ‘normal’ BP values at visit 1A were truly normal. When compared with BP values only from visit 1 (1 week earlier), a greater proportion of normal values at visit 1A could be expected because 50% of the patients were given telmisartan as
Dr Estol and Ms Hebert Martin have nothing to disclose. Dr Bath receives consulting fees from Boehringer Ingelheim. Dr Gorelick receives consulting fees from AstraZenica/Quintiles, Bayer, Janssen, Takeda, Roche/ Parexel, Shire, Brainsgate, and D-Pharm and lecture fees from Boehringer Ingelheim. Mr Cotton is an employee of
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
262 Blood Pressure Monitoring 2014, Vol 19 No 5
Boehringer Ingelheim. Dr Weber receives consulting fees from Novartis. In relation to what could be perceived as conflicts of interests Dr. Dahlo¨f reports personal fees from Boehringer Ingelheim for committee meetings and travel, during the conduct of the study; personal fees for consultancy from Novartis, Pfizer, Servier, Boehringer Ingelheim, Bayer, MSD, BMS, AstraZeneca, Vicore Pharma, personal fees for expert witness from Novartis, personal fees for lectures from Novartis, Boehringer Ingelheim, Vicore Pharma, Pfizer, MSD, board member and stock owner Mintage Scientific, board member and stock owner Cereno Scientific, outside the submitted work.
13
14
15 16
17 18
19
References 1
Wolf-Maier K, Cooper RS, Banegas JR, Giampaoli S, Hense HW, Joffres M, et al. Hypertension prevalence and blood pressure levels in 6 European countries, Canada, and the United States. JAMA 2003; 289:2363–2369. 2 Muldrow PJ. Detection and control of hypertension in the population: the United States experience. Am J Hypertens 1998; 11:744–746. 3 Keenan N, Rosendorf K. Prevalence of hypertension and controlled hypertension, United States 2005–2008. MMWR 2011; 60 (Suppl):94–97. 4 Burt VL, Whelton P, Roccella EJ, Brown C, Cutler JA, Higgins M, et al. Prevalence of hypertension in the US adult population. Results from the third National Health and Nutrition Examination Survey, 1988–1991. Hypertension 1995; 25:305–311. 5 Hajjar I, Kotchen T. Trends in prevalence, awareness, treatment, and control of hypertension in the United States, 1988–2000. JAMA 2003; 290:199–206. 6 Eastern Stroke Coronary Heart Disease Collaborative Research. Blood pressure, cholesterol, and stroke in eastern Asia. Lancet 1998; 352:1801–1807. 7 Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case–control study. Lancet 2004; 364:937–952. 8 Stamler J, Stamler R, Neaton JD, Wentworth D, Daviglus MI, Garside D, et al. Low risk-factor profile and long-term cardiovascular and noncardiovascular mortality and life expectancy. JAMA 1999; 282:2012–2018. ˆ unpuu S, Lawton WA, et al. 9 Yusuf S, Diener HC, Sacco RL, Cotton D, O Telmisartan to prevent recurrent stroke and cardiovascular events. N Engl J Med 2008; 359:1225–1237. 10 Sacco RL, Diener HC, Yusuf S, Cotton D, Ounpuu S, Lawton WA, et al. Aspirin and extended release dipyridamole versus clopidogrel for recurrent stroke. N Engl J Med 2008; 359:1238–1251. 11 Spence JD. Antihypertensive drugs and prevention of atherosclerotic stroke. Stroke 1986; 17:808–810. 12 Landert M. Manipulation of blood pressure self-monitoring values. Praxis (Bern 1994) 2003; 23:1075–1080.
20
21
22 23 24 25
26
27
28
29
30
Pickering TG, Hall JE, Appel LJ, Falkner BE, Graves J, Hill MN, et al. AHA Scientific Statement. Recommendations for blood pressure measurement in humans and experimental animals. Hypertension 2005; 45 (Pt 1):142–161. Mancia G, Grassi G, Pomidossi G, Gregorini L, Bertinieri G, Paratiet G, et al. Effects of blood-pressure measurement by the doctor on patient’s blood pressure and heart rate. Lancet 1983; 322:695–698. Markopoulos N. Isolated clinic hypertension is not an innocent phenomenon. Am J Hypertens 1999; 12:245–250. Qurush AI, Fared M, Sure K, Mohammad Y, Gateman LR, Nelson Hopkins L, et al. Isolated and borderline isolated systolic hypertension relative to longterm risk and type of stroke, a 20-year follow-up of the National Health and Nutrition Survey. Stroke 2002; 33:2781–2788. Glen SK, Elliott HL, Curio JL, Lee KR, Reid JL. White coat hypertension as a cause of cardiovascular dysfunction. Lancet 1996; 348:654–657. Jennings JR, Lamarck TW, Everson-Rose SA, Kaplan GA, Zanuck SB, Salome JT, et al. Exaggerated blood pressure responses during mental stress are prospectively related to enhanced carotid atherosclerosis in middle-aged Finnish men. Circulation 2004; 110:2198–2203. Alderman MH. Blood pressure reactivity predicts myocardial infarction among treated hypertensive patients. J Clan Epidemiology 1990; 43: 859–866. Mania G, Boboli M, Hachette R, Mayotte F, Quarti-Trevano F, Polo Friz H, et al. Long term risk of sustained hypertension in white-coat or masked hypertension. Hypertension 2009; 54:226–232. Head G, Mihailidou AS, Duggan KA, Beilin LJ, Berry N, Brown MA, et al. Definition of ambulatory blood pressure targets for diagnosis and treatment of hypertension in relation to clinic blood pressure: prospective cohort study. BMJ 2010; 340:1104–1111. Kannel WB, Wolf PA. Framingham study insight on the hazards of elevated blood pressure. JAMA 2008; 300:2545–2547. Staessen JA, Kuztnesova T, Stolarz K. Hypertension prevalence and stroke mortality across populations. JAMA 2003; 289:2420–2422. Friday G, Alter M, Lai S. Control of hypertension and risk of stroke recurrence. Stroke 2002; 33:2652–2657. Rodgers A, MacMahon S, Gamble G, Slattery J, Sandercock P, Warlow C. Blood pressure and risk of stroke in patients with cerebrovascular disease. BMJ 1996; 313:147–153. Rashid PA, Leonardi Bee J, Bath PM. Blood pressure reduction and the secondary prevention of stroke and other vascular events: a systematic review. Stroke 2003; 34:2741–2749. Muntendam P, McCall C, Sanz J, Falk E, Fuster V. High-Risk Plaque Initiative. The Bio-image Study: novel approaches to risk assessment in the primary prevention of atherosclerotic cardiovascular disease. Am Heart J 2010; 160:49–57. Sipahi I, Swaminathan A, Natesan V, Debanne SM, Simon DI, Fang JC. Effect of antihypertensive therapy on incident stroke in cohorts with prehypertensive blood pressure levels: a meta-analysis of randomized controlled trials. Stroke 2012; 43:432–440. Joffres MR, Ghadirian P, Fodor JG, Petrasovits A, Chockalingam A, Hamet P. Awareness, treatment and control of hypertension in Canada. Am J Hypertens 1997; 10:1102–1110. Benavente OL, White CL, Pearce L, Pergola P, Roldan A, Benavente MF, et al. The Secondary Prevention of Small Subcortical Strokes (SPS3) study. Int J Stroke 2011; 6:164–175.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Clinical methods and pathophysiology
Blood pressure measurement reliability among different racial-ethnic groups in a stroke prevention study Conrado J. Estolg, Philip M.W. Batha, Philip B. Gorelickb, Daniel Cottonc, Renee H. Martind, Michael A. Webere, Bjorn Dahloff and PRoFESS Publications Committee and PRoFESS Investigators Objective High blood pressure (BP) is commonly not diagnosed, and patients do not achieve target values when treated. Among 20 000 patients encompassing most races-ethnicities, we evaluated BP measurements and treatment response in a stroke prevention trial. Our goal was to identify BP measurement differences between clinical trial and patient determinations and among the racial-ethnic groups. Materials and methods A total of 20 332 patients with ischemic stroke were randomized to receive antiplatelet treatment and 80 mg of telmisartan versus placebo. BP measurements were obtained at the first clinic visit and then 1 and 3 months later and every 6 months thereafter. One week after the first clinic visit, patients were requested to report a BP measurement obtained elsewhere. Measurements at the trial clinics were obtained with the same electronic device. Statistical analysis was used to detect significant differences. Results The mean patient age was 66 years; 36% were women, and race-ethnicity comprised 58% Whites, 33% Asian, 4.9% Hispanic, and 4% Black. Overall, 74% of patients were hypertensive. BP varied between the raceethnicity groups, being highest in Hispanics (145/85) and lowest in Blacks (144/82). BP at visits clinic 1, nonclinic 1A, and clinic 2 were, respectively, 144/84, 137/80, and 139/81 mmHg, with the difference between visits 1–2 and
Introduction There are more than one billion people with hypertension (HTN) worldwide, and HTN is the most important modifiable vascular risk factor accounting for seven million deaths per year (13% of all deaths). Diagnosis and treatment of HTN are widely available. However, the rule of two-third applies with only two out of three people with HTN being diagnosed: two-third of these receive treatment and only two-third of these have their blood pressure (BP) controlled [1,2]. In a recent analysis, the prevalence of HTN among all people age 18 years or older during the period 2005–2008 was 30% and the lifetime risk of developing HTN at age 50 was 90% [3]. The population-attributable risk of HTN is 54% for stroke and 47% for coronary heart events; BP control has a direct relation with decreasing stroke risk [4,5]. Although variations in HTN severity and incidence have been reported among various world regions, studies have shown c 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins 1359-5237
visit 1A being significant. BPs were normal in 42% of the cases at visit 1A, and of these, only 44% were normal at visit 1 and 57.6% were normal on visit 2. Similar findings were noted for all race-ethnicity groups. Conclusion BP values varied among race-ethnicities and showed differences between clinic and patient measurements. This finding questions the reliability of self-reported BP and has implications for BP management in daily clinical practice. Blood Press Monit c 2014 Wolters Kluwer Health | Lippincott 19:256–262 Williams & Wilkins. Blood Pressure Monitoring 2014, 19:256–262 Keywords: blood pressure, hypertension, PRoFESS, race-ethnicity, stroke a
Nottingham City Hospital, Nottingham, UK, bSaint Mary’s Health Care, Grand Rapids, Michigan, cBoehringer Ingelheim Pharmaceuticals, Ridgefield, Connecticut, dDepartment of Biostatistics, Bioinformatics and Epidemiology, Medical University of South Carolina, Charleston, South Carolina, e910 Fifth Avenue, 6A, New York, NY, USA, fDepartment of Molecular and Clinical Medicine, Su Sahlgrenska, Go¨teborg, Sweden and gNeurological & Vascular Prevention Unit, Buenos Aires, Argentina Correspondence to Conrado J. Estol, MD, PhD, FAAN, Neurological & Vascular Prevention Unit, Av. Callao 875, C1023AAB Buenos Aires, Argentina Tel: + 54 11 4816 6733; fax: + 54 11 4816 6668; e-mail: [email protected] Received 3 November 2013 Revised 27 February 2014 Accepted 6 March 2014
that the same types of vascular risk factors similarly affect all populations across the world and that few risk factors account for the majority of new cardiovascular events [6–8]. However, differences in BP values and the reliability of BP control among different race-ethnicities have not been widely studied. Our aim was to compare BP measurements obtained at the PRoFESS (Prevention Regimen for Effectively Avoiding Secondary Strokes) trial clinics with one measurement reported by the patients for the detection of discrepancies to assess the reliability of BP measurements and search for differences among the multiple race-ethnicities studied.
Materials and methods We evaluated BP measurements in the 20 332 patients included in the secondary stroke prevention trial DOI: 10.1097/MBP.0000000000000045
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Blood pressure control reliability Estol et al. 257
PRoFESS, an international, multicenter, randomized, double-blind, 2 2 factorial trial comparing the combination of acetylsalicylic acid (aspirin) and extended-release dipyridamole versus clopidogrel, and the angiotensin receptor blocker telmisartan versus placebo. All patients received additional antihypertensive medication as indicated by the investigator. Details from the PRoFESS protocol design and overall results have been published previously [9,10]. From 2003 to 2006, PRoFESS enrolled 20 332 patients from 695 centers in 35 countries from North and Latin America, Occidental and Eastern Europe, Africa, and Australasia who had suffered a stroke within the previous 120 days. Patients with a hemorrhagic stroke on neuroimaging were excluded. Ethical approval was obtained in each participating country as per the local regulatory authorities’ requirements. All patients signed a written consent form in compliance with GCP regulations. The average patient follow-up was 2.5 years with a minimum of 18 months. After the baseline visit, patients were seen at 1, 3, and 6 months and then every 6 months until follow-up completion (visits 1, 2, 3, 4, etc., respectively). BP was measured with the same model electronic OMRON 705CP device (OMRON Corporation, Kyoto, Japan) in all 695 participating centers. Two measurements taken 2 min apart were averaged and the BP was recorded. At 1 week after the first clinic visit (visit 1), in what was defined as visit 1A, patients were requested to report to the investigator a BP value self-measured at home or obtained elsewhere (by a caregiver, the patient, a physician, a nurse, a healthcare worker, or others). Roughly 30% of PRoFESS patients were hospitalized at the baseline visit. In case patients were in the hospital at the seventh day after randomization, BP for visit 1A was obtained with the OMRON device before the intake of telmisartan. Patients were given the PRoFESS medication on visit 1, and thus half of the patients at visit 1A were receiving telmisartan 80 mg in addition to their baseline treatment. We separately compared the BP at visit 1A in the patients receiving placebo and telmisartan in relation with BP values obtained at visit 1 (no telmisartan) and subsequent clinic visits (50% of patients on telmisartan). In addition, we also compared the BP values excluding the patients who were hospitalized at the time of visit 1A. Data from PRoFESS provides strict and sequential BP measurements over 2–4 years in these patients. Race-ethnicity
As ‘ethnicity’ is a heterogeneous concept, difficult to define and strongly related to culture, language, geography, and other variables, we decided to classify patients under the more comprehensive race-ethnicity origin. The race-ethnicity for each patient was decided among the available options: European Whites, Black African, Caribbean Hispanic, Native Latin, South Asian, Chinese, Japanese, Malays, other Asian, Arab, Persian, and other.
Statistical analysis
Average systolic and diastolic BP levels were obtained in all racial-ethnic groups for all trial clinic visits and at visit 1A (outpatient/in-hospital visit at day 7) comparing clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A using paired t-tests. Furthermore, the percentage of normal BP levels (r 135/85 mmHg) for each visit was determined and then a comparison between rates of normal BPs at visits 1 and 2 and those at visit 1A was performed.
Results Among 20 332 patients enrolled in PRoFESS, 58% (11 697) were of European/White ethnicities, 33% (6660) were Asian, 4.9% (990) were Latin/Hispanic, and 4% (816) were Black African (94% African American). HTN was diagnosed in 74% of the patients. The average baseline BP reported in PRoFESS was 144/84 mmHg. Detailed baseline patient characteristics (age, sex, BMI, smoking, HTN, etc.) were reported previously [10]. Results of paired comparisons of BP measurements at clinic visits by the investigator and BP obtained by the patient elsewhere or in the hospital showed a significantly lower systolic and diastolic BP for visit 1A compared with visit 1. The mean systolic BP was 144 mmHg at visit 1 and 136.9 mmHg at visit 1A (P < 0.0001). The mean diastolic BP was 83.8 mmHg at visit 1 and 79.9 mmHg at visit 1A (P < 0.0001) (Tables 1 and 2). This comparison was repeated up to the fifth clinic visit. The comparison of systolic BP in visit 1A with clinic visit 2 was significantly lower for Whites and Latins, but not for Blacks and Asians. In visit 3, at 3 months, it was not significant for Blacks and all differences were significant when comparing visit 1A with visit 4 at 6 months. For diastolic BP measurements, the difference between visits 1, 2, 3, and 4 with visit 1A was significant (lower in 1A) except for Asians in clinic visit 2. The differences found would not be clinically meaningful if all values in visits 1 through 4 and visit 1A were normal or abnormal irrespective of the presence of a statistically significant difference in BP values. Therefore, we investigated whether the lower BPs in visit 1A were in the normal range or the abnormal range in parallel with BPs recorded at clinic visits 1 to 4. Considering all race-ethnic groups, 28.4% had normal BP values at visit 1, but in visit 1A the proportion of normal BP values increased to 42%. Table 3 shows the percentage of normal BP values for each visit (clinic and 1A) in each race-ethnic group. Table 4 shows the proportion of patients with normal BP at visits 1 and 2 of those who had normal BP at visit 1A. Compared with normal BP values in visit 1A, the percentage of normal BP measurements at clinic visit 1 ranged from 39.4 to 49.2%. The same comparison revealed that 52.8–59.2% of normal BPs at visit 1A were normal at visit 2. Even though the significant number of lower BP values at visit 1A persisted when comparisons were made with visits 2–4
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
258
Blood Pressure Monitoring 2014, Vol 19 No 5
Table 1
Systolic blood pressure levels at the clinic visits by race-ethnicity Visit 1 (baseline)
Ethnicity Asian (33.03%) European/Whites (58.01%) Black (African/ American) (4.05%) Latin Hispanic (4.91%) Mean BP
Visit 1A (1 week)
P*
Visit 2 (1 month)
Visit 3 (3 months)
P*
Visit 4 (6 months)
P*
Visit 5 (12 months)
P*
P*
144.5 (16) < 0.0001 136.9 (16.7) 137.4 (18.9) 0.1025 136.1 (18.3) 0.0024 135.7 (17.4) < 0.0001 136.3 (17.4) 0.0246 143.9 (16.6) < 0.0001 137 (18.3) 139 (19.5) < 0.0001 138.4 (19.4) < 0.0001 138.4 (19.1) < 0.0001 138.2 (18.9) < 0.0001 143.6 (18.2) < 0.0001 140.1 (20.2) 140.3 (23.6) 144.8 (17.7)
0.3323 141.1 (23.1)
< 0.0001 133.8 (19.6) 140.2 (23.4) < 0.0001 140.7 (23)
0.3456 141.5 (22.2)
0.0258 139.5 (21.6)
0.9227
< 0.0001 139.4 (21.8) < 0.0001 140.0 (22.7) < 0.0001
144.1 (16.5) < 0.0001 136.9 (17.9) 138.6 (19.7) < 0.0001 137.8 (19.5) < 0.0001 137.7 (18.9) < 0.0001 137.7 (18.8) < 0.0001
*P-values compare values at clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A (7 days) using paired t-tests. There are 20 comparisons on each table, and thus only P-values < 0.0025 are considered statistically significant using the Bonferroni adjustment. BP, blood pressure.
Table 2
Diastolic blood pressure levels at the clinic visits by race-ethnicity in the placebo and the telmisartan groups
Ethnicity
Visit 1
Asian (33.03%) European/Whites (58.01%) Black (African/American) (4.05%) Latin Hispanic (4.91%) Mean BP
P*
Visit 1A
Visit 2
P*
Visit 3
P*
Visit 4
P*
Visit 5
P*
84.9 (9.9) < 0.0001 81.40 (10.0) 81.3 (11) 0.6095 80.5 (10.4) < 0.0001 80.5 (10.2) < 0.0001 80.5 (10.1) < 0.0001 83.2 (10.7) < 0.0001 79.18 (11.0) 80.1 (11.6) < 0.0001 80.7 (11.5) < 0.0001 80.6 (11.5) < 0.0001 80.5 (11.4) < 0.0001 82.3 (10.7) < 0.0001 80.02 (12.0) 81 (13) 0.0216 81.4 (12.8) 0.0151 81.9 (12.7) 0.0028 80.6 (12.6) 0.3618 85 (11) < 0.0001 78.51 (11.1) 82.1 (13.2) < 0.0001 82.4 (13) < 0.0001 81.4 (12.5) < 0.0001 81.9 (12.8) < 0.0001 83.8 (10.5) < 0.0001 79.94 (10.7) 81.1 (11.6) < 0.0001 80.8 (11.3) < 0.0001 80.7 (11.2) < 0.0001 80.6 (11.1) < 0.0001
*P-values compare values at clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A (7 days) using paired t-tests. There are 20 comparisons on each table, and thus only P-values < 0.0025 are considered statistically significant using the Bonferroni adjustment. BP, blood pressure.
Table 3
Normal blood pressure levels at the clinic visits by race-ethnicity
Ethnicity
Visit 1 [N (%)]
Asian (33.03%) European/Whites (58.01%) Black (African/American) (4.05%) Latin Hispanic (4.91%) Total
Table 4
1747 3463 248 276 5734
(26.2) (29.6) (30.4) (27.9) (28.4)
Visit 1A [N (%)] 2798 4937 264 479 8478
Visit 2 [N (%)]
(42) (42.2) (32.3) (48.4) (42)
2765 4571 310 380 8026
(41.5) (39) (37.1) (38.4) (39.8)
Visit 3 [N (%)] 2913 4617 259 366 8155
(43.7) (39.5) (31.7) (36.1) (40.4)
Visit 4 [N (%)] 2884 4530 241 366 8021
(43.3) (38.7) (29.5) (36.9) (39.8)
Visit 5 [N (%)] 2743 4414 235 346 7738
(41.2) (37.7) (28.8) (34.9) (38.4)
Comparison of normals at visit 1A with normals at visits 1 and 2
Ethnicity Asian (33.03%) European/Whites (58.01%) Black (African/American) (4.05%) Latin Hispanic (4.91%) Total
Visit 1A normals Of those normal at visit 1A, how many normal at visit 1? Of those normal at visit 1A, how many normal at visit 2? [N (%)] [N (%)] [N (%)] 2798 (42) 4937 (42.2) 264 (32.3)
1161 (41.5) 2242 (45.4) 130 (49.2)
1659 (59.3) 2817 (57) 155 (58.7)
479 (48.4) 8478 (42)
189 (39.5) 3722 (43.9)
253 (52.8) 4884 (57.6)
when 50% of patients were also receiving telmisartan, to avoid potential bias from patients not receiving telmisartan in visit 1, we performed the BP comparisons among visits including only patients in the placebo group. Results are presented in Tables 5 and 6 for systolic and diastolic BPs, respectively. Differences between trial visits 1 and 2 and the BP provided by patients at visit 1A continued to be significant with higher values recorded at PRoFESS clinics. Of note, the mean BP for all groups continued to be lower at visit 1A until trial visit 4. The difference between the clinic and the reported BPs was the largest for the Latin Hispanic group; the difference was significant for diastolic but not for systolic
BP after visit 1 in Blacks, and the difference was significant in Asians, but, surprisingly and unlike for all other groups, from visit 3 onwards, systolic and diastolic BPs were lower at clinic visits compared with the BP reported by patients at visit 1A. Also, to decrease bias, we repeated the comparisons avoiding inclusion of the patients who were admitted in the hospital at the time of visit 1A (B30% of all patients) as the BP in these patients does not reflect the typical outpatient BP measurement context. Results are presented in Tables 7 and 8 for systolic and diastolic BPs, respectively. Mean BP measurements including all race-ethnic groups were significantly higher in clinic visits 1 and 2 compared with
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Blood pressure control reliability Estol et al.
Table 5
259
Systolic blood pressure levels at the clinic visits by race-ethnicity in the placebo group
Ethnicity
Pa
Visit 1
Asian (N = 3328) European/Whites (N = 5870) Black (African/ American) (N = 409) Latin Hispanic (N = 496) Mean BP
Visit 1A
Visit 2
Pb
Visit 3
Pb
Visit 4
Pb
Visit 5
Pb
144.4 (16.1) < 0.0001 138.9 (16.3) 140 (18.2) 0.0008 138.5 (17.5) 0.4337 137.8 (16.8) 144.1 (16.8) < 0.0001 139.7 (17.9) 141.7 (19) < 0.0001 141.1 (19.1) < 0.0001 140.8 (18.6)
0.0016 138.3 (16.9) 0.0001 140.2 (18.5)
0.1439 0.0534
144.3 (18.5)
0.0714 142.2 (21.3)
0.6803
143.9 (17.7)
0.0205 142.3 (19.9) 143.7 (23.7)
0.0723 143.5 (22)
0.2488 143.9 (22.1)
< 0.0001 135.8 (19.4) 142.3 (22.6) < 0.0001 142.2 (22.4) < 0.0001
141 (20.9) < 0.0001 141.8 (23.5) < 0.0001
144.2 (16.7) < 0.0001 139.3 (17.6) 141.2 (19.2) < 0.0001 140.4 (18.9) < 0.0001 139.9 (18.4)
0.0069 139.7 (18.4)
0.0511
P-values compare values at clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A (7 days) using paired t-tests. There are 20 comparisons on each table, and thus only P-values < 0.0025 are considered statistically significant using the Bonferroni adjustment. BP, blood pressure. a Visit 1 versus 1A (paired t-test). b Visit 2, 3, 4, 5 versus 1A (paired t-test).
Table 6
Diastolic blood pressure levels at the clinic visits by race-ethnicity in the placebo group
Ethnicity
Visit 1
Asian (N = 3328) European/Whites (N = 5870) Black (African/American) (N = 409) Latin Hispanic (N = 496) Mean BP
P*
Visit 1A
Visit 2
P*
Visit 3
85 (10) < 0.0001 82.5 (10) 82.8 (10.8) 0.1757 83 (10.7) < 0.0001 80.5 (10.7) 82.3 (11.5) < 0.0001 82.2 (10.9)
0.0205 80.8 (12)
82.7 (13)
0.0037
81.9 (10) 82 (11.4) 82.3 (12.7)
P*
Visit 4
P*
Visit 5
P*
0.0038 81.6 (9.9) < 0.0001 81.5 (9.9) < 0.0001 < 0.0001 81.8 (11.2) < 0.0001 81.6 (11.3) < 0.0001 0.0462
83 (12.6)
0.0091 81.9 (12.4)
0.2773
84.4 (10.8) < 0.0001 79.5 (11) 83.3 (12.5) < 0.0001 83.30 (12.96) < 0.0001 82.1 (11.9) 0.0009 83.3 (12.9) < 0.0001 83.8 (10.5) < 0.0001 81.1 (10.6) 82.5 (11.3) < 0.0001 82 (11.1) < 0.0001 81.8 (10.9) < 0.0001 81.6 (11) 0.0010
*P-values compare values at clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A (7 days) using paired t-tests. There are 20 comparisons on each table, and thus only P-values < 0.0025 are considered statistically significant using the Bonferroni adjustment. BP, blood pressure.
Table 7
Systolic blood pressure levels at the clinic visits by race-ethnicity in the placebo group for only those not hospitalized
Ethnicity
Visit 1
Asian (N = 2085) European/Whites (N = 3897) Black (African/ American) (N = 327) Latin Hispanic (N = 432) Mean BP
Pa
Visit 1A
Visit 2
Pb
Visit 3
Pb
Visit 4
Pb
Visit 5
Pb
143.3 (16.4) < 0.0001 138.7 (16.6) 139.1 (18.2) 0.3244 137.6 (17.1) 143.5 (17.3) < 0.0001 139.6 (18.2) 141.3 (19) < 0.0001 140.1 (19.1)
0.0166 137.2 (17.2) 0.0011 138.1 (17.3) 0.0960 139.9 (18.5) 0.3214 139 (18)
0.2524 0.1001
144.4 (18.4)
0.4569 143.5 (22.7) 0.2424 141.2 (20.6)
0.8365
0.0799 142.4 (20.3) 143.5 (23.9)
0.2048 142.9 (21.7)
143.7 (18.1) < 0.0001 136.4 (19.3) 143.1 (22.6) < 0.0001 142.5 (22.5) < 0.0001 141.2 (21.1) 0.0002 142.1 (23.8) < 0.0001 143.5 (17.1)
< 0.0001 139.3 (17.9) 140.8 (19.3) < 0.0001 139.6 (18.9)
0.1065 139.3 (18.5) 0.6654
139 (18.3)
0.5726
P-values compare values at clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A (7 days) using paired t-tests. There are 20 comparisons on each table, and thus only P-values < 0.0025 are considered statistically significant using the Bonferroni adjustment. BP, blood pressure. a Visit 1 versus 1A (paired t-test). b Visits 2, 3, 4, 5 versus 1A (paired t-test).
Table 8
Diastolic blood pressure levels at the clinic visits by race-ethnicity in the placebo group for only those not hospitalized
Ethnicity Asian (N = 2085) European/Whites (N = 3897) Black (African/American) (N = 327) Latin Hispanic (N = 432) Mean BP
Visit 1
P*
Visit 1A
Visit 2
P*
Visit 3
P*
Visit 4
P*
Visit 5
84.9 (10.1) < 0.0001 82.6 (10) 82.5 (10.8) 0.7912 81.7 (9.9) 0.0006 81.3 (9.9) < 0.0001 81.5 (10) 82.6 (10.7) < 0.0001 80 (11) 81.6 (11.4) < 0.0001 81 (11.2) < 0.0001 80.8 (11) 0.0009 80.5 (11.1) 82.4 (10.8) 0.0799 80.8 (11.8) 82.1 (13.1) 0.0360 81.7 (12.7) 0.2583 82.8 (12.9) 0.0392 81.4 (12.3) 84.4 (11) < 0.0001 79.9 (10.9) 83.7 (12.4) < 0.0001 83.5 (12.7) < 0.0001 82.2 (11.9) 83.4 (10.6) < 0.0001 80.8 (10.7) 82 (11.3) < 0.0001 81.4 (11) 0.0003 81.1 (10.8)
P* 0.0003 0.0819 0.6318
0.0050 83.7 (12.7) < 0.0001 0.1455 81.1 (11) 0.3911
*P-values compare values at clinic visits 1 (baseline), 2 (1 month), 3 (3 months), and 4 (6 months) with visit 1A (7 days) using paired t-tests. There are 20 comparisons on each table, and thus only P-values < 0.0025 are considered statistically significant using the Bonferroni adjustment. BP, blood pressure.
the BP at visit 1A, but this difference was not significant at clinic visit 3 and beyond. Again, the Latin Hispanic group revealed the largest differences between clinic BP
and BPs from visit 1A. In Blacks, differences were significant for some of the visits in both systolic and diastolic values and Asians again had significantly lower
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
260
Blood Pressure Monitoring 2014, Vol 19 No 5
values at clinic visit 3 and beyond for both systolic and diastolic values. Table 9 shows the differences found between the Hispanic groups from Argentina and those classified as ‘other countries’ (the majority from Brasil, Mexico, and the USA). Differences in systolic and diastolic BPs are significant (P < 0.0001), although no significant difference was found in the prevalence of left ventricular hypertrophy. Table 10 shows BP differences among the three largest Asian groups (South East Asia, Chinese, and ‘other’). Again, systolic and diastolic BP differences were significant (P < 0.0001), whereas left ventricular hypertrophy did not show significant differences.
Discussion We found that across all race-ethnicities, most BP measurements obtained outside the PRoFESS clinics 1 week after the first trial visit (visit 1A) were significantly lower than the first two clinic determinations (1 week before and 3 weeks after the measurement ‘elsewhere’ at visit 1A). This finding was consistent when evaluating the entire PRoFESS cohort, when separately comparing patients in the placebo and the telmisartan groups and when excluding patients who were hospitalized at the time of the first two visits (1 and 1A). In all groups (except for the diastolic BP in the nonhospitalized and the placebo groups in Asians during visit 2), the mean systolic and diastolic BPs were significantly lower in visit 1A compared with visits 1 and 2 and lower in general when compared with subsequent clinic visits. Our study included all major race-ethnicities around the world. It can be speculated that the differences we found within the Latin and the Asian race-ethnic groups could be secondary to genetic or socio-economic factors particular to the different regions where these populations live (Tables 9 and 10). For example, Latins living in a medium-income country such as Argentina are likely to have, on average, less education and a more limited access to adequate medical services compared with Latins living in the USA. Table 9
Different issues that explain poor control of BP include inadequate medical and public education to recognize this as a major risk factor for death worldwide, a high degree of noncompliance reaching 50%, use of subtherapeutic doses of antihypertensive medication, neglect in recognizing slight BP elevations as important and inadequate BP measurements [11]. We focused our analysis on the latter and thus we aimed to detect significant differences between BPs measured by trial physicians with the standard electronic machine selected for PRoFESS and those from visit 1A, which encompassed BPs obtained by patients outside the trial clinic. Our hypothesis was that outside a strictly controlled environment, BP measurements, including those made by medical and paramedical personnel, frequently provide falsely normal or only slightly elevated values. Unreliable measurements by patients are frequent, and in a study on individuals who were unaware that their BP measurements were stored in an electronic BP machine, the authors found the report of fictional BP values made up by study subjects and the report of the closest to normal value after the patients repeatedly measured their BP [12]. Use of aneroid and mercury devices for BP determinations in visit 1A may account for some of the differences found between trial clinic and visit 1A measurements as available data support the use of electronic devices based on their greater reliability and reproducibility of BP measurements [13]. Interestingly, excluding patients who were hospitalized, 7806 (57.7%) of the BP measurements at visit 1A were made by healthcare professionals not related to the PRoFESS trial and 5324 (39.3%) were taken by patients or their caregivers. Also, most patients reported a BP value from a single measurement, which should yield a higher result compared with the average of two determinations (where the second value is usually lower than the first) obtained at the PRoFESS clinics for all patients. Rather than assuming that BPs provided by patients from visit 1A were falsely low, it could be argued that BP measurements at PRoFESS centers were elevated (relative to visit 1A values) due to the ‘white coat effect’.
Differences between Hispanics from Argentina and those from other countries (most from North America)
Blood pressure and LVH Systolic BP Diastolic BP LVH
Latin Hispanic (4.91%)
Argentina (N = 404)
Other countries (N = 586)
144.8 (17.7) 85 (11) 157 (17.3%)
148.6 (17.7) 88.2 (10.9) 16.8%
142.2 (17.3) 82.8 (10.6) 17.6%
BP, blood pressure; LVH, left ventricular hypertrophy.
Table 10
Differences among Asians from different regions
Blood pressure and LVH Systolic BP Diastolic BP LVH
Asian (All) (33.03%)
South Asian (N = 1712)
Chinese (N = 3666)
Other Asian (N = 1282)
144.5 (16) 84.9 (9.9) 952 (14.3%)
139.9 (15.0) 84.5 (9.0) 14.2%
146.9 (15.9) 85.1 (10.2) 13.9%
143.7 (16.1) 85.1 (10.4) 17.1%
BP, blood pressure; LVH, left ventricular hypertrophy.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Blood pressure control reliability Estol et al. 261
However, numerous studies have proven this notion wrong by providing clear evidence that the rate of cardiovascular events in patients whose BP was increased in response to stress was the same as in those with persistent HTN in 24-h Holter studies [14–20]. A recent study conducted in 11 Australian centers on 8575 patients with a goal of identifying ambulatory BP equivalents of clinic BP for the diagnosis of HTN found small BP differences between clinic and ambulatory values [21]. These findings support the notion that large differences in BP cannot be attributed to a ‘white coat’ effect and are rather due to abnormally elevated BP in the clinic (or falsely low determinations elsewhere). Owing to our large sample size, differences in variables are likely to be statistically significant, and for this reason, it is important to establish the clinical relevance of these differences. Data from the Framingham and other studies have shown that most vascular events occur with slight increases in BP and that even a moderate decrease in BP causes a significant reduction in cardiovascular events [22–26]. The fact that more than 75% of cardiovascular events occur in ‘low-risk and intermediate-risk’ groups attests the importance of rigorous control of all and every degree of vascular risk factors [27]. A recent meta-analysis involving 70 664 patients revealed a statistically significant (relative risk 0.78; 95% confidence interval 0.71–0.86; P < 0.000001) 22% stroke risk reduction with BP treatment in people with pre-HTN (i.e. those with BP in the range of 120–139/80–89 mmHg) [28]. The study continued to be significant when using a lower baseline BP (< 130/85). Aforementioned data sustain that even BP values that are slightly elevated result in a significant number of vascular events in the long term, and thus strict BP measurements to detect these small variations are warranted. To evaluate the potential confounding effects of our large sample size, we investigated the differences in BP between clinic visits and visit 1A in a smaller group of PRoFESS patients from two high-recruiting sites in Argentina. Among 190 patients, BP was higher than 140/90 mmHg in 86% of the patients at visit 1 (and 61% had a BP greater than 160/100 mmHg), in 79% of them at clinic visit 2, in 83% of them at clinic visit 3, in 68% of them at visit 4, and in 72% of them at visit 5. In contrast, only 38% reported HTN at visit 1A (measurements outside the clinic), none with BP greater than 160/100 mmHg and 44% with values lower than 120/80 mmHg. The analysis of this restricted sample of patients reveals, as expected, wider differences than those observed in the entire 20 332 PRoFESS cohort (C. Estol, unpublished observation).
part of the PRoFESS treatment. However, further comparisons with patients at visits 2–4, of whom 50% were also receiving telmisartan, also confirmed a significantly lower BP at visit 1A. We noted that of a total of 8478 normal BP measurements at visit 1A, only 57.6% BPs were normal at visit 2. Again, this suggests that there were a significant proportion of falsely normal BP values at visit 1A. This finding persisted with most visits from 1 through 4 (the latter 6 months after visit 1) when we compared BP results (Tables 3 and 4). The fact that by visit 4 (at 6 months from visit 1), all BP measurements (except in the Asian group) were higher compared with visit 1A has important clinical implications regarding the limitations of BP control, especially in the context of the strict frame of a clinical trial. Among other reasons, it may be that falsely low or normal ambulatory BP recordings mislead patients and physicians to believe that BP is controlled [29]. A significant number of trials have failed to report positive outcomes attributed to ‘minimal or no differences’ in BP between the treatment and the placebo groups. It may be that patients in a secondary stroke prevention clinical trial could feel under ‘pressure’ to report lower BPs or simply that common mistakes in routine BP measurements (underinflation of cuff in patients with high BP, rounding-down of BP results, etc.) result in falsely low values. BP values that were persistently higher than guideline recommendations could be explained by the fact that PRoFESS was not a true BP-lowering study, but rather an add-on antihypertensive drug and antiplatelet study, and as such, investigators were not motivated enough to strictly achieve target BP values. The recent SPS3 trial clearly reflects the significant resources needed and difficulties encountered in achieving a stringent target BP value [30]. Our findings of unreliable BP measurements are likely to be greater in regular clinical practice. In PRoFESS, the complexity of BP management became apparent with BP values that were falsely low when measured outside the trial setting and BP that persisted elevated in sequential clinic visits despite medical treatment.
Acknowledgements This work was supported by Boehringer Ingelheim Pharmaceuticals Inc. (BIPI). BIPI was given the opportunity to check the data used in the manuscript for factual accuracy only. Conflicts of interest
A main objective in our analysis was to determine whether the greater proportion of ‘normal’ BP values at visit 1A were truly normal. When compared with BP values only from visit 1 (1 week earlier), a greater proportion of normal values at visit 1A could be expected because 50% of the patients were given telmisartan as
Dr Estol and Ms Hebert Martin have nothing to disclose. Dr Bath receives consulting fees from Boehringer Ingelheim. Dr Gorelick receives consulting fees from AstraZenica/Quintiles, Bayer, Janssen, Takeda, Roche/ Parexel, Shire, Brainsgate, and D-Pharm and lecture fees from Boehringer Ingelheim. Mr Cotton is an employee of
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
262 Blood Pressure Monitoring 2014, Vol 19 No 5
Boehringer Ingelheim. Dr Weber receives consulting fees from Novartis. In relation to what could be perceived as conflicts of interests Dr. Dahlo¨f reports personal fees from Boehringer Ingelheim for committee meetings and travel, during the conduct of the study; personal fees for consultancy from Novartis, Pfizer, Servier, Boehringer Ingelheim, Bayer, MSD, BMS, AstraZeneca, Vicore Pharma, personal fees for expert witness from Novartis, personal fees for lectures from Novartis, Boehringer Ingelheim, Vicore Pharma, Pfizer, MSD, board member and stock owner Mintage Scientific, board member and stock owner Cereno Scientific, outside the submitted work.
13
14
15 16
17 18
19
References 1
Wolf-Maier K, Cooper RS, Banegas JR, Giampaoli S, Hense HW, Joffres M, et al. Hypertension prevalence and blood pressure levels in 6 European countries, Canada, and the United States. JAMA 2003; 289:2363–2369. 2 Muldrow PJ. Detection and control of hypertension in the population: the United States experience. Am J Hypertens 1998; 11:744–746. 3 Keenan N, Rosendorf K. Prevalence of hypertension and controlled hypertension, United States 2005–2008. MMWR 2011; 60 (Suppl):94–97. 4 Burt VL, Whelton P, Roccella EJ, Brown C, Cutler JA, Higgins M, et al. Prevalence of hypertension in the US adult population. Results from the third National Health and Nutrition Examination Survey, 1988–1991. Hypertension 1995; 25:305–311. 5 Hajjar I, Kotchen T. Trends in prevalence, awareness, treatment, and control of hypertension in the United States, 1988–2000. JAMA 2003; 290:199–206. 6 Eastern Stroke Coronary Heart Disease Collaborative Research. Blood pressure, cholesterol, and stroke in eastern Asia. Lancet 1998; 352:1801–1807. 7 Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case–control study. Lancet 2004; 364:937–952. 8 Stamler J, Stamler R, Neaton JD, Wentworth D, Daviglus MI, Garside D, et al. Low risk-factor profile and long-term cardiovascular and noncardiovascular mortality and life expectancy. JAMA 1999; 282:2012–2018. ˆ unpuu S, Lawton WA, et al. 9 Yusuf S, Diener HC, Sacco RL, Cotton D, O Telmisartan to prevent recurrent stroke and cardiovascular events. N Engl J Med 2008; 359:1225–1237. 10 Sacco RL, Diener HC, Yusuf S, Cotton D, Ounpuu S, Lawton WA, et al. Aspirin and extended release dipyridamole versus clopidogrel for recurrent stroke. N Engl J Med 2008; 359:1238–1251. 11 Spence JD. Antihypertensive drugs and prevention of atherosclerotic stroke. Stroke 1986; 17:808–810. 12 Landert M. Manipulation of blood pressure self-monitoring values. Praxis (Bern 1994) 2003; 23:1075–1080.
20
21
22 23 24 25
26
27
28
29
30
Pickering TG, Hall JE, Appel LJ, Falkner BE, Graves J, Hill MN, et al. AHA Scientific Statement. Recommendations for blood pressure measurement in humans and experimental animals. Hypertension 2005; 45 (Pt 1):142–161. Mancia G, Grassi G, Pomidossi G, Gregorini L, Bertinieri G, Paratiet G, et al. Effects of blood-pressure measurement by the doctor on patient’s blood pressure and heart rate. Lancet 1983; 322:695–698. Markopoulos N. Isolated clinic hypertension is not an innocent phenomenon. Am J Hypertens 1999; 12:245–250. Qurush AI, Fared M, Sure K, Mohammad Y, Gateman LR, Nelson Hopkins L, et al. Isolated and borderline isolated systolic hypertension relative to longterm risk and type of stroke, a 20-year follow-up of the National Health and Nutrition Survey. Stroke 2002; 33:2781–2788. Glen SK, Elliott HL, Curio JL, Lee KR, Reid JL. White coat hypertension as a cause of cardiovascular dysfunction. Lancet 1996; 348:654–657. Jennings JR, Lamarck TW, Everson-Rose SA, Kaplan GA, Zanuck SB, Salome JT, et al. Exaggerated blood pressure responses during mental stress are prospectively related to enhanced carotid atherosclerosis in middle-aged Finnish men. Circulation 2004; 110:2198–2203. Alderman MH. Blood pressure reactivity predicts myocardial infarction among treated hypertensive patients. J Clan Epidemiology 1990; 43: 859–866. Mania G, Boboli M, Hachette R, Mayotte F, Quarti-Trevano F, Polo Friz H, et al. Long term risk of sustained hypertension in white-coat or masked hypertension. Hypertension 2009; 54:226–232. Head G, Mihailidou AS, Duggan KA, Beilin LJ, Berry N, Brown MA, et al. Definition of ambulatory blood pressure targets for diagnosis and treatment of hypertension in relation to clinic blood pressure: prospective cohort study. BMJ 2010; 340:1104–1111. Kannel WB, Wolf PA. Framingham study insight on the hazards of elevated blood pressure. JAMA 2008; 300:2545–2547. Staessen JA, Kuztnesova T, Stolarz K. Hypertension prevalence and stroke mortality across populations. JAMA 2003; 289:2420–2422. Friday G, Alter M, Lai S. Control of hypertension and risk of stroke recurrence. Stroke 2002; 33:2652–2657. Rodgers A, MacMahon S, Gamble G, Slattery J, Sandercock P, Warlow C. Blood pressure and risk of stroke in patients with cerebrovascular disease. BMJ 1996; 313:147–153. Rashid PA, Leonardi Bee J, Bath PM. Blood pressure reduction and the secondary prevention of stroke and other vascular events: a systematic review. Stroke 2003; 34:2741–2749. Muntendam P, McCall C, Sanz J, Falk E, Fuster V. High-Risk Plaque Initiative. The Bio-image Study: novel approaches to risk assessment in the primary prevention of atherosclerotic cardiovascular disease. Am Heart J 2010; 160:49–57. Sipahi I, Swaminathan A, Natesan V, Debanne SM, Simon DI, Fang JC. Effect of antihypertensive therapy on incident stroke in cohorts with prehypertensive blood pressure levels: a meta-analysis of randomized controlled trials. Stroke 2012; 43:432–440. Joffres MR, Ghadirian P, Fodor JG, Petrasovits A, Chockalingam A, Hamet P. Awareness, treatment and control of hypertension in Canada. Am J Hypertens 1997; 10:1102–1110. Benavente OL, White CL, Pearce L, Pergola P, Roldan A, Benavente MF, et al. The Secondary Prevention of Small Subcortical Strokes (SPS3) study. Int J Stroke 2011; 6:164–175.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
