This cross-sectional study enrolled 409 patients with diabetes attending primary hospitals in the Wolaita zone between 6 March and 5 April 2023. The Wolaita zone is located in the southern region of Ethiopia, which is 380 km away from Ethiopia’s capital, Addis Ababa. In the study area, there exist eight primary hospitals providing services to all zonal districts’ population and other population groups near the Wolaita zone.

The study included adult patients with DM aged ≥18 years who were under follow-up at selected public PHC chronic outpatient departments and met the inclusion criteria. Eligible participants had a documented history of at least three consecutive follow-up visits at the study facility. Critically ill patients and pregnant women with DM were excluded during the data collection period.

The sample size (n) was calculated using the single population proportion formula, assuming an HTN prevalence of 59.5% among patients with T2DM of another study conducted at DMGH (31) with 95% confidence interval (CI) and 5% margin of error, and 10% of the population size was added to adjust the non-response rate of study participants and to estimate our study sample size. Thus, s a m p l e   s i z e = Z a 2 2 *  p 1 − p d 2   =   1.96 2 *   0.595 1 − 0.595 0.05 2 n   =   370 , and after adding 10% of the non-response rate, the sample used for final study was n = 407.

The study was conducted at seven randomly selected primary hospitals (Tebela, Bele awassa, Bodit, Bedesa, Gesuba, Bitana, and Bomb primary hospital) across the study setting. An SRS method was used to identify the study subjects and 81 adult patients with DM who were followed at Tebela primary hospital, 85 at Bele awassa primary hospital, 92 at Bodit primary hospital, 69 at Gesuba primary hospital, 48 at Bitana primary hospital, 41 at Bedesa primary hospital, and 65 at Bomb primary hospital. The patient’s medical registration number from each primary hospital HTN registration log book was first used to establish a sampling frame. Following that, each hospital received a proportionate share of the computed sample size. Next, a computer-generated specialized random sampling procedure was used to choose research participants from each of the chosen hospitals.

HTN: Participants were classified as hypertensive if they met either of the following criteria: (1) average systolic blood pressure (SBP) ≥140 mmHg and/or diastolic blood pressure (DBP) ≥90 mmHg, or (2) current use of antihypertensive medication (32). HTN staging followed standard classifications: prehypertension (120–139/80–89 mmHg), stage 1 (140–159/90–99 mmHg), stage 2 (160–179/100–109 mmHg), and hypertensive crisis (≥180/110 mmHg) (33). Current smokers were defined as individuals who reported cigarette use at least once during the month preceding the study (34). Poor glycemic control was operationally defined as having a hemoglobin A1c (HbA1c) level exceeding 7% (35).

Patients with diabetes were identified through medical records as individuals with an established diagnosis who were actively engaged in follow-up care at the participating healthcare facilities, as evidenced by possession of a medical record number.

Data were collected using a pretested interviewer-administered questionnaire adapted from relevant literature (31, 36). The study employed a mixed-method approach involving face-to-face interviews, physical examinations, and medical record reviews. Behavioral risk factors were evaluated using the WHO STEPwise surveillance methodology for chronic disease risk factors. Anthropometric measurements included weight, measured to the nearest 0.1 kg using a calibrated scale, and height, measured in meters using a stadiometer with participants standing erect on a level surface. Body mass index (BMI) was subsequently calculated as weight in kilograms divided by height in squared meters.

Blood pressure measurements were obtained using a manual mercury sphygmomanometer, with the left arm positioned at heart level. Participants rested for at least 10 min before measurement (extended to 30 min for those who consumed hot beverages like coffee). Three consecutive readings were taken, and the average of these measurements was used for analysis.

To ensure data quality, the questionnaire was pretested on 5% of the total sample size at a non-selected primary hospital, with reliability assessed using Cronbach’s alpha. Data collection was conducted by four nurses holding bachelor’s degrees, who received 1 day of training on fundamental data collection principles. Supervisors underwent additional training focused on ensuring data completeness, cross-checking entries, and error correction.

The collected data were cleaned using Epi-Data software version 4.6.0.2 before being exported to SPSS version 26 for statistical analysis. Initial analysis involved computation of descriptive statistics, with results presented as frequencies and percentages. Associations between explanatory and outcome variables were examined using bivariate and multivariable logistic regression with a stepwise approach. Statistical significance was determined by an adjusted odds ratio (AOR) with a 95% CI and a p-value of less than 0.05.

Data normality was assessed through visual inspection of Q-Q plots and histograms, with the Q-Q plot approximating the diagonal line and the histogram showing no skewness, confirming normal distribution. Multicollinearity among predictor variables was evaluated using tolerance values greater than 0.1 and variance inflation factors (VIFs) less than 10, indicating no significant collinearity. Model fitness was assessed using the Hosmer–Lemeshow goodness-of-fit test, which yielded a p-value of 0.501 (χ² = 1.44), confirming that the model was well-fitted to the data.

The Ethical Review Committee (ERC) of the College of Health Science, School of Public Health, Wachemo University performed a review and allowed the study to proceed (ref. no. wcu/000981/15). In addition, formal letters of permission were obtained from the Tebela town administrative health office, the Bele awassa town administrative health office, the Bodit town administrative health office, the Bedesa district health office, the Gesuba town administrative health office, the Bitana district health office, and the Bomb district health office. The letters of permission written from district/town health offices were submitted to the respective hospitals. The study’s objectives and methods were explained to each participant. Additionally, prior to participation, all study participants provided written informed permission. The study’s goal, the thorough review of their medical records, and the potential benefits of the research were also explained to the participants. The data gathered were used for the study’s objectives. All of the information gathered throughout the study was kept private, available only to the research team and utilized only for that reason. The participants’ medical record numbers were utilized instead of personal identifiers for the purpose of gathering data. The study was carried out in compliance with the applicable rules, laws, and Helsinki Declaration principles.

No patients or public groups were involved in the design, analysis, or interpretation of this study, and they were also not involved in the dissemination of the results.

The study included 407 out of 409 eligible patients with T2DM, yielding a response rate of 99.5%. The mean age of the participants was 57.1 years (± 9.91 SD), with the largest age group (35.6%, n = 145) being 51–60 years old. The cohort consisted of predominantly male (52.1%, n = 212) and urban-dwelling (85.7%, n = 349) individuals. Educationally, 41.0% (n = 167) had attained education beyond a college diploma, while 73.0% (n = 297) were married. Government employees constituted 39.6% (n = 161) of participants, and over half (52.8%, n = 215) reported a monthly income exceeding 5,000 Ethiopian birr (ETB). These characteristics collectively describe an urban-predominant, middle-aged population with relatively high educational attainment and income levels compared to national averages ( Table 1 ).

The study population had an average diabetes duration of 10.8 years (± 3.68 SD), ranging from 3 to 25 years. Glycemic control measurements showed a mean HbA1c of 8.2% (± 2.37 SD), with 54.1% (n = 220) of participants exhibiting poor glycemic control (HbA1c >7%). The average fasting blood glucose level was 152.2 mg/dL (± 45.8 SD), while mean blood pressure readings were 131.4/82.5 mmHg (± 16.7/9.0 SD) for systolic and diastolic pressures, respectively. Treatment patterns revealed that 39.8% (n = 162) of patients were on oral hypoglycemic monotherapy, while 37.6% (n = 153) received dual therapy for diabetes management. Anthropometric data indicated that 66.1% (n = 269) of participants fell within the normal BMI range (18.5–24.9 kg/m²). These metabolic parameters collectively demonstrate the clinical characteristics and management profile of the studied diabetic population. Approximately 281 (69.05%) study participants had no history of HTN in the family. Majority of study participants, 403 (99.0%), had no history of chronic kidney disease (CKD), 375 (92.1%) had no history of cardiac heart disease (CHD), 376 (92.4%) had no history of alcohol drinking, and 395 (97.1%) had no history of smoking. Greater than half, 219 (53.8%), of study participants had physical activity of less 3 days per week. Greater than two-thirds, 274 (67.3%), of participants had a history of diabetes for more than 10 years’ duration and 293 (72.0%) participants had a self-glucometer at home. Majority of respondents, 405 (99.5%), had a history of attending health education on DM at a health facility ( Table 2 ).

HTN was identified in 66.1% (95% CI: 59.9–70.3) of the participants with type 2 diabetes. From these, 68 (25.2%) were newly diagnosed at the time of data collection. The majority, 177 (43.5%), of patients with type 2 diabetes were prehypertensive, followed by stage 1 hypertensive, 138 (33.9%); 23 (5.7%) study participants were stage 2 hypertensive and 4 (1%) were in the hypertensive crisis category.

The association between potential predictors and HTN was examined through binary logistic regression analysis. Initial candidate variables considered for the model included age, marital status, family history of HTN, physical activity level, duration of T2DM, BMI, use of anti-diabetic medications, glycemic control status, and fasting blood glucose levels. In the final multivariate analysis, four factors emerged as statistically significant predictors of HTN: age, duration of T2DM, BMI, and family history of HTN.

Advanced age showed a significant association with HTN risk. After adjusting for all other variables in the model, participants aged over 60 years demonstrated 2.09 times greater odds of developing HTN compared to those younger than 50 years (AOR = 2.09; 95% CI: 1.02–4.28; p = 0.044). Elevated BMI also proved to be an important risk factor, with overweight or obese participants (BMI ≥ 25 kg/m²) having 4.19 times higher odds of HTN than those with normal BMI (AOR = 4.19; 95% CI: 2.10–8.33; p = 0.001).

The duration of diabetes showed a significant relationship with HTN risk. Patients with a T2DM duration of 10 years or more had 1.8 times greater odds of developing HTN compared to those with a shorter disease duration (AOR = 1.8; 95% CI: 1.05–3.03; p = 0.03). Most strikingly, family history of HTN emerged as the strongest predictor, with affected participants showing 11.7 times higher odds of developing HTN compared to those without such history (AOR = 11.73; 95% CI: 5.82–23.66; p = 0.001). The notably wide CI for family history may reflect limited precision due to the relatively small subgroup size (n = 126). This finding was further validated using Firth’s penalized regression analysis, which addresses small-sample bias, confirming the strong association (AOR = 10.41; 95% CI: 5.12–21.14) ( Table 3 ).

While this study identified four significant predictors of HTN (age, family history, BMI, and diabetes duration), several potential confounding variables remain unaccounted for in our analysis. These include medication adherence patterns, specific dietary factors (such as sodium intake or DASH diet compliance), and objective measures of physical activity. Such unmeasured variables may represent important modifiers of HTN risk that could influence the observed associations. Future research incorporating these additional parameters would provide a more comprehensive understanding of HTN risk factors in diabetic populations.