1 Introduction

Hypertension is the worldwide leading cause for mortality and disability. It is a risk factor for cardiovascular disease, stroke, and chronic kidney disease [1]. In 2017, around 10.4 million deaths were attributed to hypertension around the world [2]. Majority of the burden of hypertension is concentrated in the low-and-middle income countries, including the South Asian nations [3, 4]. In the South Asian nation of Bangladesh, around 28% of the adult population had hypertension as per the latest Health and Demographic Survey 2017–18 [5]. Ahmed et al. [6] estimated that every three in five Bangladeshi hypertensive adults (59.9%) aged ≥ 35 years were unaware about their blood pressure status, while 55% of women was not aware of their hypertension. Undiagnosed hypertension can lead to major macrovascular events (e.g., coronary artery disease, myocardial infarction, congestive heart failure, stroke, and peripheral vascular disease) and microvascular events (e.g., retinopathy, nephropathy, and neuropathy), which in turn may lead to death [1, 7].

To develop effective interventions to prevent and control hypertension among Bangladeshi adults, updated knowledge on the prevalence and risk factors of undiagnosed hypertension among this age group is necessary. Also, it is necessary to find out the inequitable distribution of hypertension among different groups of people within the larger population for targeted preventive efforts. The aim of this study was to the inequalities in undiagnosed hypertension among adults (aged ≥ 18 years) in Bangladesh using the nationally representative sample of Bangladesh Demographic and Health Survey 2017–18 (BDHS 2017–18). Although previous study by Khanam et al. [9] utilized BHDS 2017–18 data to assess the prevalence, risk factors, and inequality of hypertension, the inequalities in undiagnosed hypertension were not assessed. This study will fill up this knowledge gap and inform the policy makers about the inequalities in undiagnosed hypertension using the latest nationally representative survey dataset [8].

2 Methods

2.1 Study Design

The secondary data of BDHS 2017–18 was analyzed in this study. BDHS 2017–18 was a nationally representative survey implemented by Mitra and Associates between October 2017 and October 2018 [9]. This data was analyzed between March 2021 and July 2021. The final report of BDHS 2017–18 was published previously which described the methodology, sample size calculation, data collection tools, and process [9]. A two-staged stratified cluster sampling was followed. At the first stage, a list of enumeration areas (EAs) was identified using 2011 Population and Housing Census of Bangladesh. The EAs were considered as primary sampling unit and on average each EA had 120 households. Total 675 EAs were selected based on probability proportional to their size (urban: 250 EAs; rural: 425 EAs). Then from each EAs, 30 households were randomly selected. Overall, 20,250 households were selected and one-fourth of them were selected randomly for biomarker testing, and anthropometric measurement [8]. Total 14,704 individuals (8013 women and 6691 men) were eligible for biomarker measurement. In this study, the target population was individuals aged ≥ 18 years with hypertension.

2.2 Outcome Variable

The main outcome of interest was hypertension status, which was dichotomized into diagnosed and undiagnosed hypertension. A LIFE SOURCE® UA-767 Plus BP monitor with size adjusted measuring cuffs was used for blood pressure measurement. The blood pressure was measured by two trained health technician (one male and one female) per each study team. The supervisor of the study team supervised the measurement procedure. In addition to that, external quality control team from Mitra and Associates and National Institute of Population Research and Training (NIPORT) monitored the quality of the data [9].

Three measurements were taken in ten minutes intervals between measurements. The first measurement was discarded and the average of the second and third measurements was considered as the final blood pressure. Joint National Committee 7 (JNC7) guideline was used to classify hypertension. Hypertension was defined as a systolic blood pressure (SBP) more than 140 mmHg or diastolic blood pressure (DBP) more than 90 mmHg or using blood pressure lowering medication at the time of the survey irrespective of blood pressure level. If a hypertensive respondent had SBP ≥ 140 mm Hg or DBP ≥ 90 mmHg and was never told by the health professionals to lower/control blood pressure, then he/she was considered as having undiagnosed hypertension [10].

2.3 Independent Variables

The following independent variables were considered: age (18–29, 30–49, 50–69, ≥ 70 years); education (no education, primary, secondary, college and higher education); body mass index (BMI) (underweight/normal, overweight/obese); household wealth index (poorest, poorer, middle, richer, richest); place of residence (rural, city corporation, urban), and division of residence (Dhaka, Barisal, Chattogram, Khulna, Rajshahi, Rangpur, Sylhet, Mymensingh). The information on these variables except BMI was collected using questionnaire.

Lightweight, electronic SECA 878 (with digital display and boards) was used to measure height and weight. BMI was measured by dividing the participants’ weight by height squared. The unit of BMI was kilogram/meter2 (kg/m2). Asia-specific BMI cut-off was used for categorization. BMI was categorized into: (1) underweight/normal BMI (< 23.0 kg/m2); (2) overweight/obesity (≥ 23.0 kg/m2) [11].

Information on household asset such as construction materials, type of sanitation facilities, and water source, use of electricity, health services and other amenities. Then wealth index was calculated using principal component analysis. Finally, the wealth index was categorized into quintiles [9, 12,13,14].

2.4 Statistical Analysis

At first, descriptive analysis was conducted, and the findings were presented in frequency and percentage. Then bivariate analysis was performed to find out the prevalence of undiagnosed hypertension across the covariates. Chi-square test was performed to find out the differences between the categories. Bivariate logistic regression was conducted to find out the individual association between the covariates and undiagnosed hypertension. Finally, multivariable logistic regression was fitted for identifying the overall association between the covariates and undiagnosed hypertension. Both crude odds ratio (COR) and adjusted odds ratio (AOR), with 95% confidence interval (CI) were reported. The survey weight of BDHS 2017–18 were adjusted while performing all analyses.

In order to determine the inequalities in undiagnosed hypertension among Bangladeshi adults, concentration index, with standard error (SE) was reported. At first, the households were ranked based on wealth index score. Then, a concentration curve was plotted to measure the distribution of undiagnosed hypertension.

Distribution of undiagnosed hypertension was measured by plotting a concentration curve representing the cumulative proportion of undiagnosed hypertension in Y-axis and cumulative proportions of the population in X-axis. The cumulative proportion of undiagnosed hypertension was plotted in Y-axis, and the cumulative proportions of the population was plotted in X-axis. In case of no inequality (the prevalence of undiagnosed hypertension is same across the wealth quintiles), the concentration curve and the diagonal will coincide. On the other hand, the deviation of concentration curve from the diagonal implies inequalities in the prevalence [15,16,17].

The concentration index was defined as twice the area between the concentration curve and the diagonal. The concentration index ranges from − 1 to + 1 [16, 18]. If the value is positive, then prevalence of undiagnosed hypertension is higher among the richer wealth quintiles, and a negative value implies more concentration of undiagnosed hypertension among the poorer wealth quintiles. All analyses were performed using STATA V 16.1 [19].

2.5 Ethical Consideration

The institutional review board at ICF (IRB: FWA00000845) and the Bangladesh Medical Research Council (IRB: BMRC/NREC/2016–2019/324) approved the study protocol of BDHS 2017-18. Prior to data collection, informed consent was taken from the participants.

3 Results

In total the weighted data of 3196 hypertensive adults were included in the study. The characteristics of the participants with hypertension and with prevalence of undiagnosed hypertension is presented in Table 1. Majority of the participants were aged between 30 and 49 years, female, were educated up to primary level, were overweight/obese, belonged to the richest wealth quintile, residing in the rural area, and in Dhaka division. The prevalence of undiagnosed hypertension was significantly higher among the 18–29 years old (66.3%, compared to other age categories, P < 0.001), males (57.4%, compared to females, P < 0.001), participants who received education up to college and higher (56.7%, compared to other education categories, P = 0.003), participants with underweight/normal BMI (53.9%, compared to overweight/obese, P = 0.001), poorest wealth quintile (57.3%, compared to other quintiles, P = 0.002), and residence in Sylhet division (57.8%, compared to residents of other division, P = 0.009). Although not statistically significant, the prevalence was higher among those residing in rural area (51.3%, compared to urban residents, P = 0.092).

Table 1 Characteristics of Bangladeshi adult population with hypertension and with prevalence of undiagnosed hypertension, (n = 3196)
Full size table

The results from the multivariable logistic regression model, showing factors associated with undiagnosed hypertension among Bangladeshi adults is shown in Table 2. In the final model age, BMI, gender, and wealth index, were found significantly associated with undiagnosed hypertension. Participants aged 30–49 years were 36% less likely to have undiagnosed hypertension (AOR: 0.64; 95% CI 0.47–0.87) compared to 18–29 years. The odds were 70% and 74% less among the individuals aged 50–69 years (AOR: 0.30; 95% CI 0.22–0.41) and ≥ 70 years (AOR: 0.34; 95% CI 0.23–0.50) compared to 18–29 years. Females were 45% less likely to have undiagnosed hypertension than the males (AOR: 0.55; 95% CI 0.46–0.66). Overweight/Obese adults were less likely to have their hypertension undiagnosed (AOR: 0.74; 95% CI 0.62–0.89) compared to those with underweight and normal BMI. The adults who belonged to the richest wealth quintiles had 39% lower odds of having undiagnosed hypertension compared to those who were from the poorest wealth quintile (AOR: 0.61; 95% CI 0.44–0.86). Education, place and division of residence were not associated with undiagnosed hypertension.

Table 2 Factors associated with undiagnosed hypertension among Bangladeshi adult population
Full size table

The socioeconomic inequalities in undiagnosed hypertension among adults is shown in Table 3. Overall, concentration index showed that a disproportionate prevalence of undiagnosed hypertension towards the poorest quintile (concentration index: − 0.103; SE: 0.024). The poor (Q1)-to-rich (Q5) ratio was 1.3 indicating that the prevalence of undiagnosed hypertension was higher among the poorest individuals. The inequalities in undiagnosed hypertension was higher among 50–69 years old age group (concentration index: − 0.154; SE: 0.040), males (concentration index: − 0.116; SE: 0.035), those who were educated up to secondary education (concentration index: − 0.193; SE: 0.047), overweight/obese individuals (concentration index: − 0.114; SE: 0.030), urban area (concentration index: − 0.102; SE: 0.034), Barisal division (concentration index: − 0.141; SE: 0.062).

Table 3 Socioeconomic inequalities in undiagnosed hypertension among Bangladeshi adult population, (n = 3196)
Full size table

4 Discussion

This study aimed to find out the inequalities in undiagnosed hypertension among adults in Bangladesh. The study found that every one in two Bangladeshi adults had undiagnosed hypertension. Participants aged 30–49 years, 50–69 years, and ≥ 70 years, female gender, being overweight/obese, and belonging to the richest wealth quintiles had lower odds of having undiagnosed hypertension. In most of the categories, the concentration index was significantly in negative direction, which indicates that the inequality is more concentrated in the poor socio-economic group.

Participants aged 30–49 years, 50–69 years, and ≥ 70 years had lower odds of having undiagnosed hypertension. This finding is similar to the results of a previous study done in the context of Nepal [20]. Adults of this age group are in more proximity of the health system than those of younger age group. The risk of non-communicable diseases including cardiovascular disease increases after the age of 35–40 years leading to the more frequent visit of healthcare providers [21]. As a result, the chance of having undiagnosed hypertension decreases with advancing age [6].

Similarly, females were less likely to have undiagnosed hypertension. Females are usually in more proximity of the health system than their male counterparts. Pregnant women during their antenatal visit undergo blood pressure check-up. As a result, they are less likely to have undiagnosed hypertension [20].

Similarly, overweight and obesity is an established risk factor of hypertension and other non-communicable diseases. As a result, overweight and obese adults come in frequent contact with healthcare providers and get their blood pressure measured. As a result, the probability of having undiagnosed hypertension decreases [22, 23].

Participants belonging to households from the richest wealth status had lower likelihood of having undiagnosed hypertension. This is consistent with previous studies conducted in South Asia [20, 24]. Also, we found that the participants from the lower socio-economic group were disproportionately affected by undiagnosed hypertension. This might be due to the greater access of adults from the richest wealth quintiles to the health system [25, 26]. Economic reasons (both direct and indirect medical cost) hinder the individuals from the poorer wealth quintiles from care seeking [27, 28].

We did not find any association between undiagnosed hypertension and educational status, place and division of residence. Previously Ahmed et al. [6] found less educational attainment was associated with undiagnosed hypertension. The discrepancy was mainly due to time and study population as Ahmed et al. utilized data from BDHS 2011 and their target population was adults aged ≥ 35 years. Similarly, the concentration index was almost similar among the urban (− 0.101) and rural area (− 0.102). Also, we did not find any significant differences at the divisional level. The public health promotion program aiming to address undiagnosed hypertension should focus on individuals irrespective of their educational status, place and division of residence.

In Bangladesh, the out-of-pocket expenditure (OOPE) for health remains very high and the proportion is increasing day by day. OOPE in total health expenditure increased from 55.9% in 1997 to 63.3% in 2012, as estimated by national surveys [29]. According to the recent World Bank data, the proportion has risen to 73.87% in 2018 [30]. Khan et al. [31] estimated that around 14.2% households face catastrophic health expenditure (CHE), and the poorest households are disproportionately affected by the CHE. Data from a nationally representative survey showed that presence of a chronic disease (which includes noncommunicable disease) was associated with OOPE in Bangladesh [32].

The higher burden of undiagnosed hypertension among the adult individuals is a big challenge for the health system of Bangladesh. Government of Bangladesh has recently endorsed ‘Multisectoral Action Plan for Prevention and Control of Noncommunicable Diseases 2018–2025’. It aims to reduce the premature mortality due to noncommunicable diseases by 25% within 2025 [33]. The high burden of undiagnosed hypertension among Bangladeshi adults needs to be addressed in order to achieve the aforementioned goal. Bangladesh Health Facility Survey 2017 found that in spite of having instruments for blood pressure measurement (i.e., stethoscope, blood pressure apparatus, adult scale) in more than 85% of the facilities, only 10% of them have at least trained staff to provide services related to hypertension. Health facility strengthening with trained staffs must be implemented in order to reduce the burden of undiagnosed hypertension [34].

Even in case of diagnosed hypertension, the control of blood pressure is necessary. For example, Tocci et al. [35] found that approximately one-third of the diagnosed hypertensives in Italy do not have their blood pressure controlled. In Russia, more than half of the diagnosed hypertensives had uncontrolled hypertension. Pre-existing comorbidities including cardiovascular diseases increased the likelihood of having uncontrolled hypertension [36]. Future studies should investigate the determinants of uncontrolled hypertension in Bangladeshi adults using BDHS 2017–18 dataset.

We followed the JNC7 guidelines for the definition of hypertension since the BDHS 2017–18 report followed this guideline to define hypertension [9]. This will help the policy makers to interpret the findings in the context of BDHS 2017–18 report. Future studies should compare the findings with 2017 American College of Cardiology/American Heart Association hypertension guideline, which used a different cut-off for defining hypertension [37].

This study has some key strengths. BDHS 2017–18 utilized a nationally representative sample. Therefore, the findings of the study can be generalized to the target population. The possibility of measurement error was minimized by utilizing standardized questionnaires and calibrated tools. However, the limitations of the study warrant discussion. First, due to cross-sectional nature of the study, the temporal association between the exposure and the outcome could not be established. Blood pressure was measured on the same day, whereas longitudinal measurement is recommended [10]. This might induce white coat hypertension [38]. Also, blood pressure was measured in one arm, whereas the guidelines recommend measurement in both arms [39]. However, this study depicts the inequity in undiagnosed hypertension concentrated among the adults from the poorest wealth quintile and help us to identify the predictors of undiagnosed hypertension in the study population.

5 Conclusions

Undiagnosed hypertension is common among Bangladeshi adults. Younger adults, males, individuals with underweight/normal BMI, residents from poor socio-economic status are more prone to suffer from hypertension which go undiagnosed. Targeted intervention to check regular blood pressure among the vulnerable group needs to be adopted to treat this insidious disease and prevent further complications and comorbidities.