White‐coat and masked hypertension diagnoses in chronic kidney disease patients

Abstract

The purpose of this study was to analyze which 24‐hour ambulatory blood pressure measurement (ABPM) parameters should be used on masked hypertension (MH) and white‐coat hypertension (WCH) diagnoses in chronic kidney disease (CKD) patients. Non‐dialysis CKD patients underwent 24‐hour ABPM examination between 01/27/2004 and 02/16/2012. They were followed from the 24‐hour ABPM to January/2014 in an observational study. The WCH definitions tested were as follows: (a) office blood pressure (BP) ≥ 140/90 mm Hg and daytime ABPM BP ≤ 135/85 mm Hg (old criterion); and (b) office BP ≥ 140/90 mm Hg and 24‐hour ABPM BP ≤ 130/80 mm Hg, daytime ABPM BP ≤ 135/85 mm Hg, and nighttime ABPM BP ≤ 120/70 mm Hg (new criterion). The MH definitions tested were as follows: (a) office BP < 140/90 mm Hg and daytime ABPM BP > 135/85 mm Hg (old criterion); and (b) office BP < 140/90 mm Hg and 24‐hour ABPM BP > 130/80 mm Hg or daytime ABPM BP > 135/85 mm Hg or nighttime ABPM BP > 120/70 mm Hg (new criterion). The two definitions' predictive capacity was compared, regarding both WCH and MH. Cardiovascular mortality was the primary and all‐cause mortality was the secondary outcome. Cox regression was adjusted to the variables: glomerular filtration rate, age, diabetes mellitus, and active smoking. There were 367 patients studied. The old criterion (exclusive mean daytime ABPM BP) was the only to distinguish sustained hypertension from WCH (adjusted HR: 3.730; 95% CI: 1.068‐13.029; P = .039), regarding all‐cause mortality. Additionally, the old criterion was the only one to distinguish normotension and MH, regarding cardiovascular mortality (adjusted HR: 7.641; 95% CI: 1.277‐45.738; P = .026). Therefore, WCH and MH definitions based exclusively on daytime ABPM BP values (old criterion) were able to better distinguish mortality in this studied CKD cohort.

1. INTRODUCTION

Based on office blood pressure (BP) measurements, 1.13 billion people worldwide were diagnosed hypertensive in 2015. ¹ Nevertheless, despite its importance, the office blood pressure measurement has its limitations. ³ , ⁴ There are two medical conditions whose diagnoses cannot be made only based on the office blood pressure. ³ One of them is known as white‐coat hypertension, referred to as an untreated condition in which office blood pressure measurements are elevated, but 24‐h ambulatory blood pressure measurement (ABPM), home blood pressure measurement (HBPM), or both show no abnormalities. ¹ , ² The other one is known as masked hypertension, characterized by a normal office blood pressure and hypertensive levels on 24‐h ABPM or 24‐h HBPM. ¹ , ² , ⁵ , ⁶ There are two different criteria to define white‐coat and masked hypertension. ¹ , ² , ⁷ , ⁸ , ⁹ , ¹⁰ The first one only considers the mean daytime blood pressure. In this review paper, it is referred to as the old criterion. The American Heart Association (AHA) adopts another cut point to the daytime blood pressure. ² The second criterion analyzes the mean values provided by three different evaluation periods: 24 hours, nighttime, and daytime. It is the new criterion. This second method is thought to be better in the evaluation of chronic kidney disease (CKD) patients. This belief is because CKD patients have nocturnal dipping less prominent. ⁴ , ¹¹ , ¹² The lack of nocturnal dipping is associated with more cardiovascular risk. ¹³ , ¹⁴ , ¹⁵

Despite the epidemiology mentioned above, there are few studies, based on hard end points, showing which one of these definitions is more predictive. ⁵ , ⁶ , ¹⁶ , ¹⁷ Therefore, the purpose of the current study was to verify which one of these definitions has greater predictive power, focused on CKD patients.

2. METHODS

In this longitudinal study, patients who underwent 24‐hour ABPM in the period between 1/27/2004 and 2/16/2012 were analyzed. Follow‐up was started at the time of 24‐hour ABPM and concluded in January 2014. The primary outcome was cardiovascular mortality; the secondary outcome was death from any cause. This study was approved by the local research ethics committee (number 13811019.4.0000.5411).

Age, sex, height, weight, body mass index (BMI), and smoking were recorded. Creatinine and routine urinalysis performed closest to 24‐hour ABPM (at a maximum of 3 months before or after the 24‐hour ABPM examination) were registered too. Glomerular filtration rate (GFR) was calculated by the formula of the Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI). ¹⁸ CKD was defined by “abnormalities of kidney structure (one or more of the following criteria: albuminuria; urine sediment abnormalities; electrolyte and other abnormalities due to tubular disorders; abnormalities detected by histology; structural abnormalities detected by imaging; history of kidney transplant) or function (GFR < 60 mL/min/1.73 m²), present for more than 3 months.” Five categories were established based on GFR ¹⁸ : G1: ≥90 mL/min/1.73 m²; G2: 60‐89 mL/min/1.73 m²; G3a: 45‐59 mL/min/1.73 m²; G3b: 30‐44 mL/min/1.73 m²; G4: 15‐29 mL/min/1.73 m²; and G5: <15 mL/min/1.73 m².

ABPMs were performed on weekdays with Spacelabs^® 90202 devices. ABPM standardization followed the “V Guidelines for ambulatory blood pressure monitoring (ABPM) and III Guidelines for home blood pressure monitoring (HBPM)”. ¹⁹ Mean and standard deviations for systolic BP (SBP), diastolic BP (DBP), and heart rate (24 hours, asleep, and waking) were obtained from device software. SBP and DBP sleep dipping were calculated.

The two definitions of white‐coat hypertension tested were as follows: (a) office blood pressure (BP) ≥ 140/90 mm Hg and daytime ABPM BP ≤ 135/85 mm Hg (old criterion); and (b) office BP ≥ 140/90 mm Hg and 24‐hour ABPM BP ≤ 130/80 mm Hg, daytime ABPM BP ≤ 135/85 mm Hg, and nighttime ABPM BP ≤ 120/70 mm Hg (new criterion). The masked hypertension definitions tested were as follows: (a) office BP < 140/90 mm Hg and daytime ABPM BP > 135/85 mm Hg (old criterion); and (b) office BP < 140/90 mm Hg and 24‐hour ABPM BP > 130/80 mm Hg or daytime ABPM BP > 135/85 mm Hg or nighttime ABPM BP > 120/70 mm Hg (new criterion).

Categorical variables were expressed in absolute numbers and percentages. Continuous variables were submitted to normality tests and expressed as mean ± standard deviation. Cox regression was performed considering death from cardiovascular causes as the primary and death from any cause as the secondary outcome. Cox analyses were adjusted to GFR, age, diabetes, and active smoking. Subanalyses were made to evaluate sex, proteinuria, CKD staging, and dipping status influence. Results were considered significant at P < .05.

3. RESULTS

There were 1308 examinations. Fifty‐two examinations were excluded because they were technically inappropriate. Also, 92 patients aged under 18, 18 transplanted kidney patients, and 149 repeated examinations were excluded too, totaling 259 unviable examinations. Another 119 examinations were excluded as well because patients lacked creatinine data or because they lacked urinalysis examination. In this last group, the glomerular filtration was greater than 60 mL/min/1.73 m2 and there were not kidney abnormalities, so the CKD definition was impossible. Still about this second set of patients, they were 49 ± 16.2 years old, 62 (52%) were male and none of them were on dialysis. At last, another 511 patients were excluded because they were not CKD patients or because they were on dialysis. There were no pregnant patients. Therefore, the final cohort comprised 367 patients (Figure 1), including 161 men (44%), with a mean ± SD age of 60 ± 16.0 years. Regarding ethnicity, 19 were African descendants (including mestizos) (5.2%) and one was Asian (0.3%) descendant. All other patients were Caucasian. Other significant clinical characteristics are expressed in Table 1.

FIGURE 1.

Final cohort selection

Table 1.

Clinical data of the cohort

Variables	All patients (n = 367)
GFR (mL/min/1.73 m2)	52 ± 27.8
24‐h SBP (mm Hg)	130 ± 18.0
24‐h DBP (mm Hg)	76 ± 11.6
Daytime SBP (mm Hg)	132 ± 17.9
Daytime DBP (mm Hg)	79 ± 12.1
Nighttime SBP (mm Hg)	125 ± 20.9
Nighttime DBP (mm Hg)	71 ± 12.5
Nighttime SBP dip (%)	5.4 ± 8.8
Nighttime DBP dip (%)	10.4 ± 9.82
Proteinuria a (g/24 h)	0.26 (0.15‐0.60)

Abbreviation: GFR, glomerular filtration rate.

^aMedian (first‐third quartile).

There were 35 deaths by all causes and 13 by cardiovascular disease. The two criteria were applied to distinguish white‐coat hypertension and sustained hypertension, according to the parameters obtained from 24‐hour ABPM. The old criterion had better performance in prediction, as it could predict better all‐cause deaths. Old criterion survival curves depicted on the graphic D (Figure 2) were compared with the curves based on the new criterion placed on graphic C (Figure 2). Nevertheless, graphics A and B (Figure 2) showed that neither new nor old criteria were able to predict death regarding cardiovascular mortality. Hazard ratios and confidence intervals in (Figure 2) were adjusted to GFR, age, diabetes, and active smoking.

FIGURE 2.

Survival curves of chronic kidney disease patients related to hypertension or white‐coat hypertension blood pressure classification; results of different classification criteria. A, cardiovascular mortality; patients divided by new criterion and B, cardiovascular mortality; patients divided by old criterion; C, all‐cause mortality; patients divided by new criterion and D, all‐cause mortality; patients divided by old criterion

Regarding masked hypertension and normotension, the old criterion predicted better cardiovascular mortality (graphic B Figure 3), when compared with the new criterion (graphic A Figure 3). Nevertheless, all‐cause mortality compared between the two criteria showed no significant statistical difference as shown in graphics C and D of Figure 3. Hazard ratios (HR) and confidence intervals (95% CI) in Figure 3 were adjusted to GFR, age, diabetes, and active smoking.

FIGURE 3.

Survival curves of chronic kidney disease patients related to normotension or masked hypertension blood pressure classification; results of different classification criteria. A, cardiovascular mortality; patients divided by new criterion and B, cardiovascular mortality; patients divided by old criterion; C, all‐cause mortality; patients divided by new criterion and D, all‐cause mortality; patients divided by old criterion

About the influence of gender, sex was also included as a confounding factor, along with age, diabetes, GFR, and smoking in a multiple Cox analysis. Once more, the old criterion predicted better cardiovascular mortality in masked hypertension definition (P = .024; HR: 55.071; 95% CI: 1.707‐1776.917) and all‐cause mortality in white‐coat hypertension delineation (P = .017; HR: 4.682; 95% CI: 1.322‐16.575).

The CKD staging was evaluated as a confounding factor in the substitution of GFR. There were 45 G1 patients, 53 G2 patients, 97 G3a patients, 89 G3b patients, 71 G4 patients, and 12 G5 patients. The results were similar to those obtained with GFR. A multiple Cox analysis was performed once more, including age, diabetes, and smoking as confounding factors. Old criterion showed better predictive value as it defined better white‐coat hypertension, regarding all‐cause mortality (P = .019; HR: 4.916; 95% CI: 1.294‐18.679). Old criterion was able again to outline more accurately masked hypertension, regarding cardiovascular mortality (P = .027; HR: 7.622; 95% CI: 1.253‐46.376).

Proteinuria was included as one adjusting factor as well. Old criterion remains with greater predictive power. In this case, a bivariate analysis was executed. White‐coat hypertension distinction in all‐cause mortality and masked hypertension definition in cardiovascular mortality were both obtained only by old criterion (P = .048; HR: 0.124; 95% CI: 0.016‐0.981 and P = .013; HR: 17.760; 95% CI: 1.846‐170.866, respectively).

The dipping status influence was also explored. Concerning the SBP pattern, there were 70 reverse dippers, 194 non‐dippers, 91 dippers, and 12 extreme dippers. Regarding DBP, there were 41 reverse dippers, 134 non‐dippers, 142 dippers, and 50 extreme dippers. Once again, one bivariate analysis was performed, including hypertension phenotypes and dipping status categorized into four profiles: dippers, non‐dippers, extreme dippers, and reverse dippers. The results were similar. Old criterion fulfilled better the imposed condition. It defined better white‐coat hypertension, in relation to all‐cause mortality (P = .024; HR: 4.327; 95% CI: 1.217‐15.388) and was more accurate at masked hypertension definition, regarding cardiovascular mortality (P = .021; HR: 8.732; 95% CI: 1.393‐54.723). The phenotypes characterization based on the new criterion obtained no statistical significance.

4. DISCUSSION

The purpose of this study was to verify which 24‐hour ABPM criteria predicted mortality of CKD patients. In the set studied, the old criterion (exclusive analysis of mean daytime blood pressure) showed better results when compared with the new one (analysis of mean 24‐hour, nighttime, and daytime blood pressure). In the differentiation between white‐coat and sustained hypertension, it predicted better all‐cause mortality. On the other hand, in the differentiation of masked hypertension and normotension, it predicted better cardiovascular outcomes.

Regarding the literature involved in this issue, the European Society of Cardiology/European Society of Hypertension ¹ and the American College of Cardiology/American Heart Association Task Force ² advocate the use of the new criterion on their definitions of masked and white‐coat hypertension. The rationale behind these definitions adopted by new guidelines is that the nighttime blood pressure has a better prognostic significance than diurnal blood pressure. ²⁰ However, no study verified the prognostic impact of these values of blood pressure in the subset of patients with discordant office/24‐hour ABPM blood pressure. ²¹ , ²² One Spanish study verified that white‐coat hypertension defined by the new criterion identified better cardiovascular risk factors among hypertensives. ²³ Nevertheless, this same study had not evaluated hard end points as the present review paper.

Paradoxically, nighttime BP, which is knowledgeable a good end point predictor, was consistently worse at discriminating the hypertension phenotypes in the current study. In order to explain this paradox, in the present cohort, the maximum 24‐hour SBP and the day period of its occurrence were recorded. It had a mean of 165 mm Hg during daytime and of 145 mm Hg during nighttime (P < .0001). The 24‐hour SBP shown in Table 1 was also divided into two. It had a mean of 132 mm Hg during daytime and of 125 mm Hg during nighttime (P < .0001). There were 75 reverse dippers in the cohort and 52 of them had the maximum 24‐hour SBP registered during daytime. It may be related to the final result of this study, but it is speculative.

In the current study, in the subset of CKD patients, the new criterion showed to be less predictive of cardiovascular or all‐cause mortality when compared with the old criterion. This study was one of the few to focus on these definitions and it brought unexpected results. This conclusion may have been because the patients were from a specific group: CKD patients.

There were 389 non‐CKD patients in the 24‐hour ABPM cohort studied, but there were only six deaths in this group. Only one of these deaths had a cardiovascular motivation (stroke). Four were due to infections and one was related to neoplasia. Therefore, it was not possible to include the non‐CKD as a control group, as the end point number (all‐cause and cardiovascular mortality) was too small.

Some limitations of this paper must be recognized. First, the studied patients were from a restricted area (surrounding the city of Botucatu, state of São Paulo). Second, these patients had a particular ethnic composition (only 19% of African and 0.3% of Asian descendants), which varied from the distribution worldwide. Third, the antihypertensive medication usage data were too scanty in our database, impeding it to be included as a confounding factor in the analysis. At last, the number of patients could be bigger; nevertheless, the set studied was a very specific one and it restricted the sample size. The end point outcomes were small at the non‐CKD patients, and this fact precluded further comparison. Despite these issues, this review paper was an innovative one, which debated a very important issue.

In conclusion, paradoxically, the masked and white‐coat hypertension CKD patients' classification by exclusive analysis of mean daytime ABPM blood pressure better predicted outcomes. It is important to point out that this was a hypothesis‐generating study and should be followed by other papers on this subject.

CONFLICT OF INTEREST

None.

da Silva HP, Gonçalves AB, Barretti P, et al. White‐coat and masked hypertension diagnoses in chronic kidney disease patients. J Clin Hypertens. 2020;22:1202–1207. 10.1111/jch.13924