Prevalence and Clinical Characteristics including Patterns of Antihypertensive Drug Administration of the Different Home Blood Pressure Phenotypes in Treated Hypertensive Patients

Quality and quantity of home blood pressure (BP) control are important for optimizing hypertensive treatment. The prevalence and associated clinical characteristics of the different home blood pressure phenotypes in treated hypertensive patients were not elucidated. This study was conducted in Siriraj Hospital, Thailand from 2019 to 2020. We included treated hypertensive patients with ≥1 antihypertensive drug and had self-home BP measurement data. Both traditional (office BP < 140/90 mmHg and home BP < 130/80 mmHg) and new BP targets (office and home BP < 130/80 mmHg) were used for the classification of BP phenotypes. Home BP phenotypes consisted of controlled hypertension (all home BPs achieved home BP targets), isolated uncontrolled morning hypertension (MoHT) (only morning BP was above home BP targets), isolated uncontrolled evening hypertension (EHT) (only evening BP was above home BP targets), and combined morning-evening uncontrolled hypertension (MoEHT) (all home BPs were above home BP targets). Our study included 1,406 patients. The total mean age was 62.94 ± 13.97 years. There were 39.40% men. The prevalence of each home BP phenotype (by traditional BP target) was 55.76%, 12.66%, 7.40%, and 24.18% in controlled (home) hypertension, MoHT, EHT, and MoEHT, respectively. Classical BP control status was 35.21% well-controlled hypertension, 30.01% white-coat uncontrolled hypertension, 9.74% masked uncontrolled hypertension, and 25.04% sustained uncontrolled hypertension. The multivariable analysis showed the significantly associated factor of MoHT was the presence of previous cardiovascular disease (adjusted OR 5.54, 95% CI (2.02–15.22); p value = 0.001). Taking once-daily long-acting antihypertensive drugs in the morning were significantly associated with both EHT (adjusted OR 0.20, 95% CI (0.05–0.82); p value = 0.025) and MoEHT (adjusted OR 0.20, 95% CI (0.04–1.00); p value = 0.049). These results were consistent in groups classified by new home BP target <130/80 mmHg.


Introduction
Nowadays home blood pressure monitoring (HBPM) plays a role in the diagnosis and management of hypertension because home blood pressure (BP) is strongly associated with cardiovascular outcomes [1]. HBPM does not only improve hypertension awareness and adherence that result in better BP control [2,3] but it also classifes white-coat efect/white-coat hypertension and masked hypertension in addition to ofce BP measurement. Several current guidelines recommend using home BP as the target BP for the treatment of hypertension in order to improve the quality of BP control in hypertensive patients [4][5][6][7]. Previous studies demonstrated a positive association between morning BP or morning hypertension at home and cardiovascular disease, stroke [8,9], and chronic kidney disease [10]. Te Ohasama Study showed the predictive power of both morning and evening hypertension which were assessed by HBPM for stroke in hypertensive patients [11]. Patients with morning or evening or a combination of morning and evening hypertension might have diferent clinical characteristics. Some studies revealed that regular alcohol drinking [12], antihypertensive drug regimens, diabetes mellitus, male gender, and kidney disease afected home BP phenotypes [13]. Since the growing pieces of evidence have supported intensive BP control for the prevention of cardiovascular morbidities and mortality [14]. Some recent guidelines suggest reducing ofce and home BP to <130/80 mmHg [4,15], which is the new BP target. According to the difference between the traditional BP target (ofce BP target is <140/90 mmHg and home BP target is <135/85 mmHg) and the new BP target (both ofce and home BP targets are <130/ 80 mmHg), this study aimed to evaluate the prevalence of BP control status including diferent home BP phenotypes and to investigate whether clinical characteristics and treatment associated with the diferent home BP phenotypes in treated hypertensive patients.

Study Population.
A cross-sectional study was conducted at Siriraj Hospital, Mahidol University, Tailand. Te medical records of the hypertensive patients who were treated at the hypertension clinic from 1 January 2018 to 31 December 2020 were retrospectively reviewed. Te eligible participants were the patients who met all inclusion criteria: (1) age of at least 18 years, (2) diagnosed with hypertension and treated with at least one antihypertensive drug for at least 4 weeks, and (3) completed data of home BP in morning and evening times for at least 3 consecutive days and at least 12 home BP records (≥6 home BP records in each morning and evening) [16]. Patients with pregnancy, end-stage kidney disease, secondary hypertension, and unable measurement of BP at the brachial artery were excluded. Our study was approved by the Siriraj Institutional Review Board (SIRB) (COA no. Si 356/2022).

Blood Pressure Measurement.
Te physicians and welltrained nurses regularly generally informed hypertensive patients in the hypertension clinic on the standard method of HBPM and asked them to measure and record their home BPs twice, 1 minute apart each morning and evening times for 3 to 7 consecutive days before the next appointed visit. BP measurement was performed after at least 5 minutes of resting period in a sitting position on the chair with back support. Te appropriate-sized cuf was placed on their arms at the same level as the heart. Because all participants had antihypertensive drugs at diferent times, hence the participants were informed to measure their morning BP before taking antihypertensive drugs and before bedtime which was defned as evening BP. If some antihypertensive drugs were administrated before bedtime, evening BP was suggested to measure before taking these drugs. Te validated oscillometric BP devices for HBPM were Omron HEM-7130 and HEM 7211 (Omron Healthcare Co. Ltd.) because we practically provided the home BP devices to most patients in the hypertension clinic. Te results of the home BPs of the patients had been recorded in the medical records of Siriraj hospital. Te average home morning and evening BPs were calculated from all numbers of BPs each time. Average daytime home BP was calculated from the average of mean home morning and evening BP.
Ofce BP measurement followed the standard technique of accurate attended BP measurement as per international guideline's recommendation on the visit day. Te validated oscillometric BP devices (Omron HBP-300 and HBP-110; Omron Healthcare Co. Ltd.) were used for BP measurement in a sitting position after at least 3 to 5 minutes of rest. Te last two ofce BPs were calculated as average ofce BP.

Defnition of Diferent Home Blood Pressure Phenotypes
According to Home Blood Pressure Control. We classifed home BP phenotypes depending on morning and evening BP control into four following phenotypes: (1) controlled hypertension (all average morning and evening SBPs were <135 mmHg and DBPs were <85 mmHg); (2) isolated uncontrolled morning hypertension (MoHT) (Only average morning SBP was ≥135 mmHg and/or DBP was ≥85 mmHg but average evening SBP was <135 mmHg and DBP was <85 mmHg); (3) isolated uncontrolled evening hypertension (EHT) (Only average evening SBP was ≥135 mmHg and/or DBP was ≥85 mmHg but average morning SBP was <135 mmHg and DBP was <85 mmHg); and (4) combined morning-evening uncontrolled hypertension (MoEHT) (all average morning and evening SBPs were ≥135 mmHg and/ or DBPs were ≥85 mmHg).
In addition, we also used the new recommended BP target of the 2017 American College of Cardiology (ACC)/ American Heart Association (AHA) hypertension guideline [4] for classifying the following 4 home BP phenotypes: (1) controlled hypertension (all average morning and evening SBPs were <130 mmHg and DBPs were <80 mmHg); (2) isolated uncontrolled morning hypertension (MoHT) (Only average morning SBP was ≥130 mmHg and/or DBP was ≥80 mmHg but average evening SBP was <130 mmHg and DBP was <80 mmHg); (3) isolated uncontrolled evening hypertension (EHT) (Only average evening SBP was ≥130 mmHg and/or DBP was ≥80 mmHg but average morning SBP was <130 mmHg and DBP was <80 mmHg); and (4) combined morning-evening uncontrolled hypertension (MoEHT) (all average morning and evening SBPs were ≥130 mmHg and/or DBPs were ≥80 mmHg).

Defnition of Diferent Classical Blood Pressure Phenotypes According to Both Ofce and Home Blood Pressure Control.
Four classical BP phenotypes, which were well-controlled hypertension, white-coat uncontrolled hypertension (WCHT), masked uncontrolled hypertension (MHT), and sustained uncontrolled hypertension (SHT) were classifed according to traditional and new BP targets. Traditionally, the ofce BP target was <140/90 mmHg, and the home BP target was <135/85 mmHg. Te new BP target was <130/ 80 mmHg of both ofce and home BPs. Tus, wellcontrolled hypertension was defned as all ofce and average daytime home BPs achieved the BP targets. WCHT was the condition that had abnormally higher ofce BP than targeted ofce BPs while average daytime home BP still achieved the home BP targets. MHT was defned as ofce BP being within the ofce BP targets but home BP being above the home BP targets. Te defnition of SHT was the 2 International Journal of Hypertension condition that had both higher ofce and home BPs than the BP targets.

Antihypertensive Drugs Regimen.
To evaluate the pattern of prescribed antihypertensive drugs in routine clinical practice and the association between antihypertensive drug regimens and home BP phenotypes, we classifed antihypertensive drugs into two groups according to their pharmacokinetics (eg. duration of action). Long-acting antihypertensive drugs were defned as their duration of BP reduction was 24 hours or more. Short-acting antihypertensive drugs were defned as their duration of action was less than 24 hours. Furthermore, the regimen of medical treatment was divided into the following 3 groups: (1) only morning drug administration which was defned as taking a once-daily drug before noon (ante meridiem); (2) only evening drug administration which was defned as taking once-daily drug after noon (post meridiem); and (3) both morning and evening drug administration. Because of the efect of antihypertensive drugs' doses on BP values, trough-to-peak ratio, and blood pressure variability [17,18], we used the two steps for the calculation of the proportion of antihypertensive doses in each morning and evening times to 24 hours. In the frst step, the total antihypertensive therapeutic intensity score (total TIS) was calculated in the individuals. Te formula of total TIS [19] was shown in the following equation: total antihypertensive therapeutic intensity score � Actual da ily do se Recommende d maximal da ily do se . (1) Second step, the TIS of each morning and evening drug administration was divided by total TIS.

Data Collection.
Te information of all participants was collected from the medical records of Siriraj hospital. Baseline information consisted of age, gender, comorbidities including the previous history of cardiovascular disease, history of smoking and alcohol drinking, body weight, height, ofce, and home BPs, and detail of antihypertensive drug prescription was collected. Cardiovascular disease defnes as the presence of at least one of the following diseases: myocardial infarction, heart failure, ischemic and hemorrhagic stroke. Te mineralocorticoid receptor antagonist was only spironolactone due to its availability in Tailand. Te other class of antihypertensive drugs included centrally acting alpha-II agonists and direct vasodilators.

Statistical Analysis.
Te previous study reported that the proportion of BP control rate by achieving targeted BP at home was 57% [20]. We expected that the prevalence of wellcontrolled hypertension was 5% less than the study's prevalence so the expected prevalence was 52%. Te estimated sample sizes of at least 1,038 participants were required to detect the diference of 5% with 80% power using a 5%-level two-sided test.
Descriptive statistics were used for the analysis of baseline characteristics data. Continuous variables, such as age and body mass index, were expressed as mean ± standard deviation or median (interquartile range) depending on the data's distribution. Categorical variables were presented as numbers and percentages. One-way ANOVA (analysis of variance) and the Kruskal-Wallis test were used to compare the continuous data between all groups according to normal or non-normal distribution, respectively. Because multiple analyses were performed, the statistically signifcant level was adjusted by Bonferroni correction. Te Chi-square test was analyzed for the comparison of categorical variables. Te association between clinical characteristics including treatment and home BP phenotypes was analyzed by the univariable and multivariable multinomial logistic regression model. Because of using the two diferent thresholds of home BP targets (<140/90 and <130/80 mmHg) for the classifcation of home BP phenotypes, the results were separately analyzed and presented. Statistical analyses were performed by Stata Statistical Software Version 17. (StataCorp LLC, College Station, TX). A p value of less than 0.05 was considered statistically signifcant.

Baseline Characteristics and Home BP Phenotypes.
After screening 1,606 hypertensive patients with had completed home BP data, we excluded 167 hypertensive ones without pharmacologic treatment and 33 ones with incomplete home BP data. Tus, this study included 1,406 treated hypertensive patients. Te mean age of all groups was 62.94 ± 13.97 years. 554 (39.40%) men and 852 (60.60%) women were enrolled. Te mean body mass index (BMI) of all groups was 25.54 ± 4.52 kg/m 2 . Te patients with abnormally high BMI or overweight (BMI ≥ 23 kg/m 2 ) were 69.65%. Dyslipidemia (79.71%) was the most comorbidity and 23.16% of all patients had diabetes mellitus. A previous history of cardiovascular disease was found in 9.41% of all patients and the proportion of patients with chronic kidney disease (CKD) was 17.39%. Te mean number of antihypertensive drug classes was 2.13 ± 1.09 classes for total patients. Dihydropyridine calcium channel blocker (DHP-CCB) and angiotensin-II receptor blocker (ARB) were mainly used for the treatment of hypertension. Te average ofce BP was 142.25 ± 15.89/78.67 ± 11.94 mmHg and the average daytime home BP was 128.05 ± 11.65/ 76.10 ± 10.11 mmHg. Te detail of overall baseline characteristics was shown in Table 1.
Four home BP phenotypes were classifed by home BP target of <135/85 mmHg. Te diferent characteristics of International Journal of Hypertension   (Table 3). Te proportion of these four groups was changed when using the new threshold of ofce and home BP target of <130/80 mmHg. Te prevalence of well-controlled hypertension, WCHT, MHT, and SHT was 12.52%, 32.86%, 5.90%, and 48.72%, respectively (Supplementary Table 3). Using traditional BP targets (ofce BP target of <140/ 90 mmHg and home BP target of <135/85 mmHg (Table 4), the patients with controlled both morning and evening home BPs were mostly found in well-controlled hypertension (88.26%) and WCHT groups (82.23%) in spite of the fact that they were not found in MHTand SHTgroups. Tere was a gradually increased proportion of patients with MoHT from the group of well-controlled hypertension (7.27%), WCHT (9.95%), and MHT (17.52%) to SHT (21.59%). Te proportion of EHT was 4.44% in the group with wellcontrolled hypertension, 7.82% in WCHT, 13.14% in MHT, and 8.81% in SHT groups. MoEHT was only found in MHT (69.34%) and SHT (69.6%) groups. Supplementary Table 4 showed that the results by using a new BP target of <130/80 mmHg were the same as mentioned above.

Association of Clinical Factors and Home BP Phenotypes.
Te results of univariable and multivariable regression for determining the association of clinical factors and home BP phenotypes which were classifed by home BP target of <135/85 mmHg were shown in Tables 5 and 6, respectively. After ofce BP and average daytime home BP were adjusted in the multivariable regression model, only the presence of previous cardiovascular disease was signifcantly associated with MoHT (adjusted OR 5.54, 95% CI (2.02-15.22); p value � 0.001). Taking once daily long-acting antihypertensive drugs in the morning had a signifcant inverse association with both EHT (adjusted OR 0.20, 95% CI (0.05-0.82); p value � 0.025) and MoEHT (adjusted OR 0.20, 95% CI (0.04-1.00); p value � 0.049). Tese results were consistent in spite of reclassifying groups depending on the new home BP target of <130/80 mmHg (Supplementary Tables 5 and 6).

Discussion
Several present guidelines recommend lowering ofce and out-of-ofce BPs to each target BPs [4][5][6][7] due to out-ofofce BP better predicting future cardiovascular events than ofce BP [1]. Tis study showed the prevalence of BP control by using the diferent BP targets which were the traditional BP target (ofce BP of <140/90 mmHg and home BP of <130/80 mmHg) and the new BP target (both ofce and home BPS of <130/80 mmHg). Tis study did not only classify all treated hypertensive patients into diferent four BP phenotypes by using average ofce BP and average daytime home BP but the average morning BP and evening BP were used for allocating these patients into diferent 6 International Journal of Hypertension  [20] which demonstrated the change of prevalence of BP control when using diferent BP thresholds. BP variability is suggested to be taken into account in the treatment of hypertension because there have been supporting pieces of evidence of the cardiovascular prognostic power of BP variability including diurnal BP change [22][23][24]. In addition, Te Ohasama Study revealed the efect of morning and evening home BP on stroke [11]. It could be implied that uncontrolled home BP in either morning or evening was related to incident stroke and most home BPs should be achieved the goal BP. Considering the particular uncontrolled hypertension, we were able to divide it into 3 phenotypes in our study. Tey consisted of MoHT, EHT as well as MoEHT. MoEHT had the highest proportion among them. Te result also revealed the proportion of diferent home BP phenotypes in each classical BP phenotype. Our study showed the presence of MoHT and EHT in wellcontrolled hypertension and WCHT groups although achieving a home BP target is one criterion for wellcontrolled hypertension and WCHT. Using average daytime BP to represent home BP in this study is the reason that explains this fnding. Controlled (home) hypertension was commonly found in well-controlled hypertension and WCHT groups. On the contrary, MoEHT was commonly found in MHT and SHT groups.
Tere were several diferent characteristics between the 4 groups. Patients with MoEHT had the youngest age. Body mass index gradually increased from controlled hypertension, MoHT, EHT, to MoEHT. Some previous studies supported that overweight and obesity were strongly associated with poor BP control [25,26]. Chronic kidney disease was found in a higher proportion in uncontrolled hypertensive groups, especially EHT and MoEHT. Chronic kidney disease was signifcantly associated with uncontrolled BP [27]. Masked uncontrolled hypertension was also common in patients with chronic kidney disease and reduced glomerular fltration rate [27,28]. Te reason why the EHT had the highest proportion of chronic kidney disease remained unclear because most previous studies evaluated diurnal BP variation by 24-hour ambulatory BP monitoring and reported average daytime BP, average nighttime BP, and average 24-hour BP [29,30] and some studies investigated BP control rate by using mean overall home BP [31,32]. Guidelines for hypertension treatment recommend limiting alcohol intake because of its efect on BP control [4][5][6][7]. Te study showed that alcohol drinking was more commonly found in MoEHT than in the other groups. Te average ofce BP and average daytime BP in the MoEHT group were the highest among the four groups despite the fact that the MoEHT group had a higher number of antihypertensive drug classes and a higher proportion of taking mineralocorticoid receptor antagonist and peripheral alpha-I receptor blockers, which were not the main class of antihypertensive drugs and were used for add-on therapy. Te fnding demonstrated the inadequate hypertensive treatment of the patients in sustained uncontrolled   International Journal of Hypertension hypertensive groups. Te overall mean number of antihypertensive drug classes was 2.13 ± 1.09. It was concordant with the recommended combination therapy of antihypertensive drugs of present guidelines [5][6][7]. Uncontrolled hypertensive groups had a higher proportion of evening antihypertensive drug administration than controlled hypertensive ones. Te evening to total daily dose ratio of antihypertensive drugs in MoHT and MoEHT was signifcantly higher than controlled hypertension. Te long-acting antihypertensive drug was prescribed in higher proportion than the short-acting drug. It indicated that the prescribed antihypertensive drug regimen for the treatment of hypertension in routine practice followed the recommendation of recent guidelines [4][5][6][7]. Albuminuria is a surrogate marker of kidney damage and cardiovascular disease [33]. It was signifcantly presented in EHT and MoEHT compared to controlled hypertension because these two groups had a higher proportion of chronic kidney disease than the others. All aforementioned fndings indicated uncontrolled hypertensive groups, in particular MoEHT, had more severity of hypertension than ones with controlled BP. In addition, the analysis for investigation of the association of clinical factors and home BP phenotypes was performed by comparing each uncontrolled hypertensive group with the controlled hypertensive group. On the basis of multivariable multinomial logistic regression analysis with adjustment of average ofce and daytime home BPs, we separately discussed the association in each home's BP phenotypes. In the MoHT group, the multivariable analysis showed that a history of previous cardiovascular disease was the signifcantly independent associated factor. Even though the previous evidence supported that morning hypertension was strongly related to stroke [9,34,35], It could not show the causal efect of cardiovascular disease and MoHT in this study because of the limitation of the cross-sectional design. For EHT and MoEHT groups, the independently associated clinical factor was taking at least 1 long-acting antihypertensive drug once daily in the morning. Antihypertensive drugs were an important factor in home BP control since previous studies of chronotherapy of hypertension indicated that evening or bedtime administration of antihypertensive drugs improved morning BP control [36][37][38]. Because of the long duration of action of antihypertensive drugs that were taken in the morning, the evening home BP was controlled to achieve the BP target. Tis reason may explain the fnding of the inverse association between this factor and EHT. Furthermore, the clinical factor remained inversely associated with MoEHT. Patients with MoEHT had more severe hypertension and more numbers of comorbidities so a more complex drug regimen and polypharmacy might be necessary to control their diseases. Our results showed the gradually increased proportion of taking at least 1 longacting antihypertensive drug in both morning and evening from controlled hypertension, MoHT, EHT to MoEHT while once-daily therapy in the morning of at least 1 long-acting antihypertensive drug had the lowest proportion of prescription in the MoEHT group. Te reverse causality of the factor which is taking at least 1 long-acting antihypertensive drug once daily in the morning and MoEHT probably explained this inverse association between them.
Te strength of this study was the additional information on home BP control by consideration of the component of BP at each time. Te study also emphasized the diferent phenotypes of uncontrolled hypertension. Although EHT was a minor population in uncontrolled hypertensive groups similar to the Ohasama study [11] and had been usually ignored, this condition was not benign. Tis study also analyzed the efect of pattern antihypertensive drug administration and pharmacokinetics which might be an important factor for circadian BP change and control. However, the study had several limitations. First, this study design was cross-sectional. It limited the interpretation of the association of results and could not directly identify the causal relationship between factors and outcomes. Second, the defnition of evening antihypertensive drug administration included the variation of time to take medication after noon while evening BP was defned as measured BP before bedtime. Tus, evening BP might be afected by the peak efect of taking antihypertensive drugs after dinner in some patients. Finally, there were some efects of unmeasured confounding factors, for example, drug adherence, nonpharmacologic intervention, and daily activities, which might afect the association of clinical factors and home BP phenotypes.
In conclusion, the rate of well-controlled BP (both average ofce and daytime home BPs achieved ofce and home BP targets) remained low in treated hypertensive patients. EHT phenotype had the lowest proportion among uncontrolled hypertensive groups. Tere were several diferent clinical characteristics in diferent three phenotypes of uncontrolled home BPs. Te independent clinical factors associated with MHT were a previous history of cardiovascular disease. Taking at least 1 long-acting antihypertensive drug in the morning is signifcantly associated with EHT and MoEHT. However, further well-designed studies for investigating the efect of chronotherapy and pharmacokinetic properties of antihypertensive drugs on home blood pressure control and long-term cardiovascular outcome in each home BP phenotypes are required.

Data Availability
Te datasets of this study are available for the only investigators of this study. Te data are not publicly available. However, the authors consider providing these data upon reasonable request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.