Baseline prevalence of high blood pressure and its predictors in a rural adult population of Bangladesh: Outcome from the application of WHO PEN interventions

Abstract This cross‐sectional study estimated the prevalence of high blood pressure (BP) and examined its predictors at baseline following protocol 1 (actions 1 and 2) of World Health Organization (WHO) Package of Essential Noncommunicable Disease (PEN) Interventions in a selected rural area of Bangladesh. A total of 11 145 adults (both sex and age ≥ 18 years) completed both the questionnaire and clinical measurements at the household and community clinics, respectively. We defined high BP as systolic BP ≥ 120 mmHg or diastolic BP ≥ 80 mmHg, prehypertension (pre‐HTN) as systolic BP 120–139 mmHg or diastolic BP 80–89 mmHg, and hypertension (HTN) as systolic BP ≥ 140 mmHg or diastolic BP ≥ 90 mmHg and/or anti‐hypertensive drug intake for the raised BP. The prevalence of high BP was 51.2% (pre‐HTN, 25.3%; HTN, 25.9%). Among them, the proportion of pre‐HTN was higher among men (28.7%) while HTN was higher among women (27.4%). Other than fast food intake (pre‐HTN, OR: 1.110, P = .063) and women sex (HTN, OR: 1.236, P < .001), the pre‐HTN and HTN had higher odds for having same predictors as follows: age ≥ 40 years, family history of HTN, physical inactivity, central obesity, generalized obesity, and diabetes. In conclusion, the application of WHO PEN protocol 1 detected one‐fourth of the rural adult population had pre‐HTN and HTN respectively, and the common significant predictors of those were the age, family history of HTN, physical inactivity, generalized obesity, and diabetes.


INTRODUCTION
Raised blood pressure (BP) or hypertension is a major risk factor for noncommunicable diseases (NCD) like coronary artery disease (CAD), chronic kidney disease, and stroke. 1 Among the NCD risk factors, raised BP attributed the highest proportion of global death compared to other NCD risk factors. 2 Again, more than half (53%) of the cardiovascular disease (CVD) related mortality was contributed by hypertension. Thus, it is considered as the most prevalent NCD risk factor which is highly preventable. 3,4 Although HTN is mostly preventable, globally 22% of the adult population had raised BP in 2015 and 28% of this burden was distributed among less developed or low-income countries. 5 The prevalence of raised BP varied across the regions of the World Health Organization and South-East Asia positioned (SEA) third place (25%) next to Africa (27%) and the Eastern Mediterranean region (26%). 6 A systematic review and meta-analysis conducted among the countries of the South Asian Association for Regional Cooperation (SAARC) reported that most of the studies of this region had inconsistent data on HTN and its risk factors. 7 Hence, accurate and consistent estimation of HTN prevalence and its risk factors is crucial for prevention and management, especially in developing countries with low resources.
Bangladesh is currently in the third stage of epidemiological transition which is characterized by shifting from acute infectious and deficiency diseases to NCDs. The wave of this transition has also influenced the livelihood and health status of the rural population reported as an increase in chronic diseases. 8 About 73% of the Bangladeshi population is living in rural areas where the population, healthcare infrastructure, and healthcare provider characteristics are very different compared to the urban setting. 9 A study reported that most of the rural population had to depend on homeopathic physicians, local nurses, hospital compounders, the sellers of medical store, etc, during their medical emergencies. 10 The same study also stated that strong beliefs on superstitions encouraged them to go to local Herbalists, Hakim or Kabiraj for a natural cure, who have limited knowledge on modern medical sciences. 10 Regarding HTN, the evidence demonstrated that graduate doctors (MBBS and specialized) diagnosed only 53.5% of the HTN in rural areas of Bangladesh, and the remaining portion was diagnosed by unqualified healthcare providers. Again 26% of the diagnosed population discontinued their medication for HTN. 11 These findings indicate that identification of HTN among the rural population of Bangladesh is inadequate and getting appropriate care and follow-up is unlikely.
Previous studies reported that the prevalence of pre-HTN and HTN among the rural Bangladeshi population ranged from 20.3% to 41.5% 12-15 and 6.9%-30.6%, [12][13][14][15][16] respectively. However, we have not found any study that evaluated the distribution of high BP based on the data collected following the WHO Package of Essential Noncommunicable Disease Interventions (WHO PEN). 17

Study design, setting, and population
This cross-sectional study was conducted at baseline during the WHO PEN implementation in the Dhangara Union of Raiganj Upazila under the Sirajganj district, a rural area of the Rajshahi division ( Figure S1).

Data collection procedure
A face-to-face interview was conducted at the household level using a mobile app developed based on a semi-structured pre-tested questionnaire. A group of well-trained male and female data collectors (students of health sciences perusing BSc/MSc) interviewed the participants, measured their physical parameters, and assessed glycaemic status both at households and CCs respectively. The pre-tested questionnaire was mainly prepared and adapted as per the STEP-wise approach to Surveillance (STEPS) of NCD risk factors survey questionnaire of the World Health Organization (WHO, Version 3.2) 18 and the "Action1 (ask about)" and "Action 2 (Assess)" sections of the WHO PEN protocol 1 (supporting document 2). 17 We modified both the WHO STEPS and The "measurements" part was prepared based on STEP 2 (physical measurements) and STEP 3 (biochemical measurements) of the STEPS survey questionnaire, and the "Action 2 (assess)" section of WHO PEN.
In the CC, we used the "measurement" part of the drafted questionnaire to record anthropometric measurements, blood pressure, and capillary blood glucose levels.

Definition of high BP and its risk factors
High BP is defined as systolic BP ≥ 120 mmHg or diastolic BP ≥ 80 mmHg. 20 The pre-HTN within this range was defined as systolic BP 120 mmHg to 139 mmHg or diastolic BP 80 mmHg to 89 mmHg. 19 Again, the cut-off for HTN was systolic BP ≥ 140 mmHg or diastolic BP ≥ 90 mmHg 20 and/or documented anti-hypertensive drug intake for the raised BP. Here, new cases were those diagnosed for the first time in the study, and old cases were those previously diagnosed by a health professional and/or documented anti-hypertensive drug intake for the raised BP.
The current tobacco use, current alcohol consumption, inadequate fruit and/vegetable intake, low physical activity, and diabetes were categorized as per the definition used in the "Noncommunicable disease risk factors survey Bangladesh 2010." 19 Generalized obesity and central obesity were categorized as per the definition of WHO and the International Diabetic Federation respectively. 21,22 The added salt intake was defined as taking dietary salt during eating a meal. 23 The fast food intake was defined as "food that can be prepared quickly and easily and is sold in restaurants and snack bars as a quick meal or to be taken out." 24,25 The details of the definitions are added as a Supporting document 3.

Quality assurance
To maintain the quality control of the study, we applied several measures: (i) pre-testing of the drafted questionnaire to detect any inconsistency, unclear wording, or unusually longer time taken to administer, and finalized after appropriate modification; (ii) pre-survey training of the team members including investigators, supervisor, data collectors to outline the rationale of the study, and the procedures and potential difficulties associated with data collection; (iii) strict monitoring at the field level to closely monitor the data collection through field coordinators. Besides, for monitoring of the total system, a separate dashboard was created that was regularly monitored by the investigators (iv), the app was tested in the field for data synchronization and integration with the existing primary health care system, (v) use of show cards for a better understanding of the related risk factors of high BP, (vi) use of robust equipment for physical and biochemical measurements.

Ethical consideration
Data were collected after informed written consent was obtained. Ethical approval to conduct the study was taken from the Ethical Review

Statistical analysis
The data were entered in the pre-designed Microsoft office excel format which was later imported into the statistical software Statistical Product and Service Solutions version 20.0 for Windows (SPSS, Inc., Chicago, IL, USA). Initially, we checked the data for incompleteness, inconsistency, missing data, coding errors, and any outliers. A total of 11 244 adults visited the community clinics and provided their measurements. However, based on the initial checking, 11 145 remained for analysis and reporting. To assess the distribution of high BP and its risk factors, categorical variables were presented using frequency and percentages, and the continuous variable was presented using mean and standard deviation. We also used error bars to represent the overall proportion of high BP, new and old cases of HTN. These descriptive analyses were applied to achieve the first objective of this study (distribution of high BP) according to age, sex, and risk factors.
To identify the predictors of pre-HTN and HTN, we used multinomial logistic regression analysis. For this, first, we checked the assumptions of regression analysis that included multicollinearity, outlier, normality, linearity, homoscedasticity, and the independence of observations. We did not find any violation of these assumptions. To identify the variables eligible to be included in the multinomial logistic regression analysis, univariate analyses were computed and a P-value of ≤.25

3.2
Age-and sex-specific distribution of high BP among the study population Overall, 51.2% of participants had high BP without any sex differences ( Figure 1). Among the participants with high BP, the overall proportion of pre-HTN (25.3%) and HTN (25.9%) were similar ( Table 2). The sex-wise distribution showed that the proportion of individuals with pre-HTN was higher among men while HTN was higher among women (Table 2). Again, among the hypertensive population, the proportion of new and old cases were 14% and 11.8%, respectively. However, the proportion of old cases of HTN was more prevalent among women (15.9%) compared to men (12.8%) (Figure 2). According to age-wise distribution, the proportion of individuals with pre-HTN was higher than the HTN below the age of 50 years and after the age of 50 years, this relationship reversed.
The proportion of people with HTN gradually increased with advancing age and became maximum (48.9%) after the age of 60 years (Table 2).

Risk factor-specific distribution of high BP among the study population
Inadequate fruit/vegetable intake (68.5%) and added salt intake (62.8%) were the highly prevalent risk factors of high BP among the study population (Table 3).
Here (Table 3), the risk factors-wise distribution showed sex-specific differences for pre-HTN, but not for the HTN. In men, pre-HTN was higher among those with generalized obesity (32%), consumed fast food (31.9%), did less physical activity (30.2%), and had a family history of HTN (30.9%). In women, pre-HTN was higher among the participants who had a history of alcohol intake at least once in their lifetime (31.8%) and generalized obesity (30.8%). However, in both sexes, the prevalence of hypertension was higher among the participants with diabetes (men, 50.1%; women, 55.6%) and generalized obesity (men, 48%; women, 45.7%) ( Table 3).

Predictors of high BP among the study population
The univariate analysis shows highly significant (P < .001) association for most of the risk factors of high BP (Table S1). Based on the univariate analysis, multinomial logistic regression was applied in Table 4 that shows the predictors of high BP (pre -HTN and HTN) among the study population. Other than fast food intake (pre-HTN,  and Democratic People's Republic of Korea (14.8%). 32,33 The proportion of HTN detected in this study is higher than the recent global data (17.5%), 34 and lower than the previous data (31.1%) of the year 2000-2010. 35 In our study, the proportion of HTN was higher among women which is also consistent with other rural studies of Bangladesh 13,14 and evidence from an Iranian study. 36 Regarding detection of old and new cases, STEPS survey Bangladesh 2010 detected more old cases (18.6%), and STEPS survey Bangladesh 2018 detected a similar proportion (13.7%) of old cases of HTN as the current study. 19,29 In this study, unhealthy dietary habit (inadequate fruit/vegetable intake and added salt intake) was detected as the highly prevalent risk factors of high BP. In this regard, a recent review also reported the unhealthy diet and high sodium intake as the main culprit of sustain high BP. 35 The distribution of pre-HTN risk factors shows notable differences between sexes. In men, the prevalence of pre-HTN was higher among the participants who had generalized obesity, consumed fast food, a family history of HTN, and inadequate physical activity. Previous studies also reported a high prevalence of pre-HTN among the participants with generalized obesity, 37,38 family history of HTN, 39,40 and physical inactivity. 41 We have not found any study that reported the burden of pre-HTN concerning fast food consumption to compare our findings. In women, pre-HTN was higher among those with a history of alcohol intake and who were generally obese. These findings are also consistent with a study conducted among Tanzanian rural women for alcohol intake and generalized obesity. 42 Although pre-HTN shows the sex-specific difference in its distribution, HTN shows a high prevalence among the participants of both genders with diabetes and generalized obesity. This finding is consistent with a rural study of Bangladesh. 15 The current study identifies some non-modifiable risk factors as the predictors of pre-HTN and HTN. These include age and family history Among the behavioral risk factors, fast food intake showed higher odds of pre-HTN compared to having normal BP. Globally, data of association between fast food consumption and high BP in the adult is scanty. In this regard, an epidemiological study of Bangladesh reported a high percentage of fast food consumption (82%) as the risk factors of obesity and HTN among university students. 48 However, we have not found any straightforward evidence that reported a significant association between fast food consumption and high BP. Interestingly, in children, no association was found between fast food intake and high BP which contradicts our study. 49 Inadequate physical activity is another behavioral risk factor that showed higher odds of pre-HTN compared to having normal BP. A prior study in a rural setting of Bangladesh also found a significant association between physical inactivity and HTN prevalence. 16 Another Brazilian study also reported a higher prevalence of HTN among the participants who were physically inactive than those who were active. 50 Generally, the proportion of developing high blood pressure is 30-50% higher among those who are physically inactive. 51

LIMITATIONS
The study findings should be interpreted in light of some limitations.

CONCLUSION
Half of the selected rural adult population of Bangladesh has high blood pressure. Again, one-fourth of them have pre-HTN and HTN.
The equal burden of pre-HTN and HTN in a rural setting is alarming as it will increase the total healthcare expenditure (direct/indirect) in near future among the rural deprived population. Again, as their healthseeking behavior is poor, they rarely visit the nearby health facilities to identify themselves as at risk of high BP and subsequent complications. Moreover, as an economically poor community, they have limited access to get the advanced health services to mitigate the risk and manage the related complications. All of these demands a costeffective protocol like WHO PEN to include these high-risk populations under a unique screening program. Here, the identified people with pre-HTN are the target of future intervention to reduce the subsequent HTN and its complications. Besides, we recommend implementing the WHO PEN protocol 1 in other regions of the country and various settings to generate comparable data and assess its effectiveness that will help to adopt the new protocol within the existing health care system.

CONFLICT OF INTEREST
Authors have no conflicts of interest to disclose.