This community-based cross-sectional study was conducted in Durame town, southern Ethiopia, from July to September 2022. The town had an estimated population of 52,513 living in three kebeles (the smallest administrative unit in Ethiopia). There is one general hospital, three health centers, nine clinics, one health post, and nine drug stores (10).

All randomly selected adults aged ≥18 years who lived in the study area for at least six months without a history of hypertension and did not visit any health facility for a blood pressure check in the past year were included in the study. Adults who were unable to respond to the interview questions and/or take physical measurements were excluded from the study. Data on socio-demographic characteristics, family history of hypertension, behavioral characteristics, physical activity, dietary practices, and health-seeking behaviors were collected using a pre-tested questionnaire.

After completing the interview, the data collectors measured the participants' weight and height. The weight was measured by using a portable digital weight scale and recorded to the nearest 0.1 kg. The height was measured by using a portable Stadiometer and recorded to the nearest 0.1 cm. The participants were instructed to stand on the footplate with their backs to the Stadiometer, their legs straight, their arms at their sides, and their shoulders relaxed, and touch the Stadiometer at their heels, buttocks, upper back, and head. Then, the headpiece was lowered until it touched the crown of the head firmly, compressing the hair.

Blood pressure was measured using a standard mercury sphygmomanometer with a cuff size that covers two-thirds of the upper arm. No exercise, smoking, coffee, or tea were allowed 30 min before the measurement. The participants were instructed to sit on the chair with their legs uncrossed, feet on the floor, back supported, arm supported on the desk to keep the upper arm at the same level as the heart, and an empty bladder guaranteed. The upper arm was uncovered two inches above the elbow crease. A cuff is placed on the brachial artery in the upper arm. Three blood pressure measurements were taken 10 min apart in a sitting position. Finally, the average of the three readings was taken to determine the blood pressure status of the respondents.

The questionnaire was translated into Amharic, the official language, and back to English to maintain consistency in meaning. The interview questionnaire was pilot tested on 26 adults living in Teza Agara that had characteristics nearly similar to Durame town in order to identify the interview questions and physical measurements that needed adjustments. The data collectors and supervisors received intensive training for two days on how to obtain ethical consent, conduct interviews, measure weight, height, and blood pressure, and keep records. The collected data were checked for consistency and completeness on a daily basis.

Epi Info version 7 was used to determine the minimum sample size of 526 by taking the 28.8% expected proportion of undiagnosed hypertension (9), 5% margin of error, 95% confidence level, 1.5 design effects, and 10% non-response rate.

Multistage sampling was used to recruit the study participants. In the first stage, four ketenes (villages in the kebeles) were selected from a pool of 10 ketenes by using the lottery method. The town administrative office provided the list of households in each selected ketene, which was used as a sampling frame. At the second stage, a total of 526 samples were proportionally allocated to each selected ketene (i.e., 100 for Arada ketene, 144 for Teza ketene, 139 for Arada Girige ketene, and 143 for Abonsa ketene). At the last stage, the households with the eighth interval in the sampling frame of each ketene were selected for the data collection. If more than one eligible adult was present in the selected household, one household member was selected randomly using the lottery method.

The data were entered into EpiData version 4.6 and analyzed on SPSS version 26. Frequency and percentage were used to present categorical data, whereas mean and standard deviation were used to present continuous data. All independent variables were fitted separately into a bivariate logistic regression model to estimate the degree of association with undiagnosed hypertension among adults. Then, variables with a p-value  <  0.25 were exported to a multivariable logistic regression model to control confounders. The model's fitness was checked by the Hosmer-Lemeshow goodness-of-fit test. A multicollinearity test was also done by examining the standard error (SE) for the b coefficient. A SE > 2 indicates the presence of multicollinearity among independent variables. In this case, SE was less than 2 for all variables. The statistical significance was declared by using an odds ratio (OR) with a 95% confidence interval (CI).

Undiagnosed hypertension: refers to a systolic blood pressure of ≥140 mmHg or a diastolic blood pressure of ≥90 mmHg and/or both (1), without a prior history of hypertension, and having not visited any health facility for a blood pressure check in the past year. The average of three sphygmomanometer readings taken 10 min apart in a sitting position was used to determine the blood pressure status.

Prehypertension: refers to a systolic blood pressure of 120–139 mmHg or a diastolic pressure of 80–89 mmHg and/or both (1).

Normal blood pressure: refers to a systolic blood pressure between 90 and 120 mmHg and a diastolic blood pressure between 60 and 80 mmHg (1).

Body mass index: refers to a quotient of weight divided by height squared (kg/m²).

Drank too much alcohol: refers to adults who drank over 330 ml of beer (5% alcohol), 330 to 750 ml of wine (20%–25% alcohol), or 200–250 ml of local alcohol, like tella (8%–15% alcohol), tej (10%–13% alcohol), or arake (40%–45% alcohol), per week (1).

Family history of hypertension: refers to a history of hypertension in the father, mother, or both as reported by the respondent.

Low consumption of fruits and/or vegetables: refers to adults who ate fruits and/or vegetables less than five times per day. One serving is equivalent to one apple, orange, or banana, or three tablespoons of cooked vegetables (1).

Physical inactivity: refers to adults who undertake less than 150 min of moderate-intensity physical activity per week, less than 75 min of vigorous-intensity physical activity per week, or a combination of both (11).

Street foods: refers to foods and beverages that were prepared and sold by vendors in streets and public places for consumption.

A total of 516 adults participated in the study, making the response rate 98.1%. Of whom, 50.8% were males. The mean age with a standard deviation was 34 ± 11. Nearly half, 49.6%, of the participants belonged to the age group of 30–49 years. The majority, 82.0%, of the participants were married. About 47.3% of the participants had a high school or higher certificate. Regarding occupation, the proportion of government/self-employed and farmers was 37.8% and 33.3%, respectively. About half, 49.6%, of the participants earned a monthly income of less than 2,300 birr (Table 1).

The majority, 93.4%, of the participants did not use any tobacco products. A nearly similar proportion, 94.4%, of them did not smoke cigarettes in the past 30 days. About three-fourths of the participants did not drink alcohol during the data collection. Beer was a favourite type of alcoholic drink for nearly half, 52.0%, of the participants. The majority, 81.2%, of the participants did not drink too much alcohol. The proportion of those who did not chew khat was 85.7% (Table 2).

All adults who participated in this study heard about hypertension. The main source of information about hypertension was health institutions, which accounted for 60.7%. The major cause of hypertension was reported to be stress (52.8%), followed by high salt and fat consumption (30%). More than two-thirds, or 70%, of the participants reported that hypertension could not be prevented. The majority, 60.3%, of the participants did not report that hypertension could cause serious health complications. Less than one-third, or 30.8%, of the participants reported that regular measurement of blood pressure was important to prevent hypertension (Table 3).

The mean systolic blood pressure with a standard deviation was 113 ± 23 mmHg, and the mean diastolic blood pressure with a standard deviation was 76 ± 10 mmHg. The distribution of blood pressure was observed to be higher than these among participants in the age group of 30–49 years, attended primary school and below, had a family history of hypertension, physically inactive, drank too much alcohol, drank tej, areke, wine and beer, chewed khat, and consumed street foods regularly (Table 4). The prevalence of undiagnosed hypertension was 14% (95% CI: 11.2–17.1%), whereas the prevalence of prehypertension was 10.1% (95% CI: 7.8–12.8%).

Age, occupation status, family history of hypertension, drinking too much alcohol, chewing khat, physical inactivity, low consumption of fruits and/or vegetables, body mass index, consuming street foods, and healthcare for hypertensive symptoms without serious illness showed association with undiagnosed hypertension in the bivariate analysis at p-value < 0.25. Family history of hypertension [Adjust OR (AOR)  =  6.9, 95% CI: (3.62, 13.27)], drinking too much alcohol [AOR = 5.7, 95% CI: (2.97, 10.75)], physical inactivity [AOR = 2.5, 95% CI: (1.34, 4.73)], consuming street foods regularly [AOR = 2.8, 95% CI: (1.28, 6.01)], and seeking healthcare for hypertensive symptoms without serious illness [AOR = 2.4, 95% CI: (1.28, 4.56)] remained significant in the multivariable logistic regression model (Table 5).