ABSTRACT

1.        Introduction

Hypertension and diabetes are among the top list of non-communicable diseases driving the global burden of cardiovascular morbidity and mortality. They function not only as clinical conditions in their own right but also as upstream determinants of major adverse outcomes, including stroke, ischemic heart disease, chronic kidney disease, and premature mortality. [1] ^{[2] [3]} The co-occurrence of these conditions further amplifies risk through shared pathophysiological pathways such as vascular remodeling, endothelial dysfunction, and metabolic dysregulation, making them critical targets for prevention and control in population health.^[4]

Although the epidemiology of hypertension and diabetes has been widely studied, a persistent gap remains in understanding their independent predictors in remote rural populations. This gap is particularly important in settings such as rural Ghana, where socioecological contexts differ markedly from urban environments. Rural populations are often characterized by relative socioeconomic homogeneity, subsistence-based livelihoods, constrained access to healthcare services, and distinct environmental and behavioral exposures. These factors may alter both the distribution and the strength of conventional risk factors. For example, limited screening infrastructure can lead to substantial underdiagnosis, while cultural norms and occupational patterns may shape diet, physical activity, and healthcare-seeking behaviors in ways that are not adequately captured in urban-focused studies. As a result, relying on generalized risk profiles may obscure important context-specific dynamics, underscoring the need for localized epidemiological evidence.

Existing literature consistently identifies age as a dominant non-modifiable determinant of both hypertension and diabetes, reflecting cumulative biological wear and tear across the life course, including arterial stiffening and progressive insulin resistance. ^{[5] [6] [7]}  Besides age, modifiable behavioral factors play a crucial role. Alcohol consumption, for instance, has been linked to elevated blood pressure and impaired glucose regulation through mechanisms such as increased sympathetic nervous system activity, oxidative stress, and disruptions in insulin signaling. ^[8] Similarly, adiposity, commonly assessed by body mass index (BMI), is a well-established predictor of hypertension, largely because it is associated with increased vascular resistance and inflammatory processes. ^{[9] [10]}  However, its relationship with diabetes in rural sub-Saharan African populations appears more complex, potentially influenced by genetic susceptibility, early-life nutritional factors, and differences in fat distribution that are not fully captured by BMI alone.

Evidence regarding other risk factors, including educational attainment, tobacco use, dietary salt intake, and family history, especially with diabetes, remains inconsistent, particularly after adjustment for confounding variables in multivariable analyses. These inconsistencies may arise from several methodological and contextual challenges, including measurement error, clustering of exposures (e.g., poverty, diet, and occupation), and residual confounding.^{[11] [12]} In rural settings, where exposures are often interrelated and data quality may be constrained, robust analytical approaches are essential.

In the Ghanaian context, hypertension and diabetes have been increasingly associated with demographic and lifestyle transitions, even within rural areas. As communities gradually shift from traditional agrarian practices toward more sedentary and “urbanized” lifestyles, risk factors such as unhealthy diets, reduced physical activity, and increased alcohol consumption are becoming more prevalent.^[13] Empirical studies suggest that hypertension affects approximately 20-35% of adults in rural Ghana, while diabetes prevalence ranges from about 5-8%, though these estimates vary by region and population characteristics.^{[14] [15] [16]} Identified risk factors commonly include older age, male sex, lower educational attainment, higher BMI, alcohol use, tobacco consumption, and family history of disease.^{[17] [18]} while other analyses highlight the growing contribution of behavioral risk factors such as alcohol consumption and hereditary predisposition.^[19]

Despite these insights, data on predictors remain sparse for rural, especially remote, Ghanaian communities, where socioeconomic factors such as farming livelihoods and food insecurity play key roles. Against this background, the present population-based cross-sectional study was conducted to identify the independent predictors of hypertension and diabetes in rural Ghana. Specifically, the study aims to examine the association between sociodemographic, behavioral, and clinical risk factors and determine adjusted odds ratios for independent predictors using logistic regression models.

2.        Methodology

A population-based cross-sectional study design was adopted, utilizing quantitative data collection methods to describe the epidemiological profile of hypertension and diabetes. The study was conducted in five remote rural communities within the Akatsi South District of the Volta Region, Ghana: Ayiti Korpe, Gefia, Atidzive, Lume Avete, and Torve. These communities were deliberately selected to reflect rural contextual characteristics relevant to the study objectives.

A two-stage cluster sampling strategy was implemented. In the first stage, the five communities were treated as clusters and selected purposively. In the second stage, households within each community were chosen using systematic sampling at regular intervals. From each selected household, one eligible adult was randomly chosen to participate.

A total of 507 adults aged 18-75 years, residents for ≥6 months, were included, regardless of prior diagnosis, to capture both known and undiagnosed cases. Pregnant women, individuals <18 years, temporary residents, and those unable to consent were excluded.

The sample size was initially estimated using the formula n =Z². P.(1-p) ¤ E²), where Z=1.96, p=0.5, E=0.05, yielding 357 after finite population correction. Recruitment of 507 participants increased precision and reduced the margin of error. Reliability was assessed using Cronbach’s alpha (α≥0.70). Inter-rater consistency was ensured through standardized training and regular calibration of field procedures.

Two binary outcome variables were defined: hypertension status and diabetes status, operationalized using standardized clinical thresholds.

Explanatory variables were selected based on theoretical relevance, empirical evidence, and epidemiological plausibility. These included age group, sex, educational attainment, alcohol consumption, tobacco use, dietary salt intake, family history of hypertension or diabetes, and body mass index. Age was categorized into four ordinal strata (18-29, 30-44, 45-59, 60-75), while body mass index was modeled as a continuous variable to preserve information and enhance statistical power. Behavioral exposures were treated as dichotomous indicators.

A multivariable logistic regression analysis was conducted to identify risk factors for hypertension and diabetes. Variables were selected using a stepwise approach based on (i) prior knowledge and theory, (ii) initial bivariate analysis with a p-value < 0.20, and (iii) clinical relevance. This approach helped reduce overfitting while retaining important confounding variables in the model.

Adjusted odds ratios (AORs) with 95% confidence intervals (CIs) were estimated to quantify the independent association between predictors and outcomes. Likelihood ratio tests were used to evaluate the contribution of each covariate to model fit, comparing nested models based on changes in the -2 log-likelihood statistic.

Overall model fit was tested using the likelihood-ratio chi-square test, which compares the full model to a baseline model with no predictors. The model’s explanatory ability was measured using pseudo R² statistics (Cox & Snell, Nagelkerke, and McFadden), which indicate the proportion of outcome variation explained by the predictors. Model calibration was assessed using the Hosmer–Lemeshow test; a non-significant result (p > 0.05) suggests a good association between observed and predicted outcomes across risk groups. Discriminatory performance was evaluated using classification accuracy, showing the proportion of correctly classified cases.

2.1.     Ethical Considerations

The study followed standard ethical principles for research involving human participants, including informed consent, confidentiality, and anonymization of all data. Ethical approval was obtained from the Ghana Health Service (regional and district health directorates) and the EUCLID University Ethical Committee (EC/2024/1129).

3.       Findings

3.1.     Age and gender distribution of the study population

Overall, 233 men (45.96%) and 274 women (54.04%) were surveyed, giving a total sample of 507 adults aged 18-75 years.

3.2.    Educational Level

Overall, education levels were low among respondents. About one-quarter had no formal education (25.8%), while nearly half had only primary education (47.5%). Fewer participants had secondary education (22.7%), and only a small proportion had tertiary education (3.9%).

Blood Pressure and Blood Glucose

About one in five adults had elevated blood pressure at the time of measurement. Overall, 20.9% (95% CI: 17.6-24.7) were classified as having elevated BP, and 6.7% (95% CI: 4.8-9.2) had severe hypertension (≥160/100 mmHg). For blood glucose, 7.7% of adults (95% CI: 5.7-10.3) had levels in the diabetes range (≥7.0 mmol/L or on diabetes medication), while 10.8% (95% CI: 8.4-13.9) had impaired fasting glucose (prediabetes range).

Behavioral Risk Factors

Smoking/Tobacco Use

3.3.    Current smokers among respondents

Most respondents were non-smokers. The prevalence of current tobacco use was low at 7.5% (95% CI: 5.4-10.1%). Smoking was predominantly observed among men, with a prevalence of 14.2% compared to 1.8% among women. Current smoking among women was very rare (5 cases), all occurring in middle-aged individuals. No participants aged 18-29 years reported current smoking.

3.4.    Mean duration of smoking (years) among current smokers

Among current smokers, the average duration of smoking was 9.3 years (SD = 10.5), with a median of 5.0 years. Smoking duration increased with age: 6.7 years (SD = 4.9) among those aged 30-44 years, 8.6 years (SD = 8.8) among those aged 45-59 years, and 11.5 years (SD = 14.3) among those aged 60-75 years.

Alcohol Consumption

3.5.    Current drinkers among respondents

Current alcohol use was relatively low in the study population. Overall, 17.0% of adults (86 out of 507; 95% CI: 14.0-20.5) reported that they currently consume alcohol, while 83.0% were not current drinkers. Alcohol use was much more common among men than women. About 30.5% of men (95% CI: 25.0-36.5) reported current drinking, compared to only 5.5% of women (95% CI: 3.4-8.8), meaning men were about five times more likely to drink alcohol. Among men, alcohol use was highest in the 30-44 age group (46.3%). In contrast, alcohol use among women remained low and fairly stable across all age groups (≤7.8%).

3.6.    Salt intake among respondents

Almost all participants reported adding extra salt to their food at the table. Overall, 98.8% (95% CI: 97.4-99.5) said they always add salt or salty seasoning just before or during eating. This behavior was consistent across all age and sex groups (≥97% in each subgroup). Only 1.2% (n = 6) reported that they rarely or never add extra salt.”

3.7.    Family history of hypertension and diabetes.

About 60% of adults (95% CI: 55.3-63.7) reported that they did not know their family history of high blood pressure. Similarly, 64.1% of adults (95% CI: 59.9-68.2) were unaware of whether any family member had high blood sugar or diabetes.

3.8.    Logistic Regression Analysis for Predictors of Hypertension

The multivariable logistic regression model for hypertension showed good performance in explaining and predicting the outcome. The full model fit the data significantly better than the null model (χ² = 262.303, df = 190, p < 0.001), indicating that the included predictors improved model performance. The model explained a substantial proportion of the variation in hypertension (Cox & Snell R² = 0.404; Nagelkerke R² = 0.630; McFadden R² = 0.505). The model also showed good agreement between observed and predicted values (Hosmer-Lemeshow χ² = 8.42, p = 0.394), suggesting acceptable calibration. Overall, the model correctly classified 78.3% of participants, indicating strong predictive ability (Table 1).

Age showed a clear and strong association with hypertension and was the most important independent predictor. Compared with adults aged 18-30 years, the likelihood of hypertension increased steadily with age: 31-45 years (AOR = 2.14, 95% CI: 1.32-3.47, p = 0.002), 46-60 years (AOR = 4.82, 95% CI: 2.89-8.03, p < 0.001), and ≥61 years (AOR = 8.76, 95% CI: 4.92-15.59, p < 0.001). This pattern shows a strong dose-response relationship, where the risk of hypertension rises progressively with increasing age (Tables 2 & 3).

Body mass index showed a significant positive association with hypertension, with each 1-unit increase in BMI associated with a 9% higher odds of hypertension (AOR = 1.09, 95% CI: 1.06-1.12, p < 0.001), highlighting excess body weight as an important modifiable risk factor. Alcohol consumption was also independently associated with higher odds of hypertension (AOR = 1.78, 95% CI: 1.02-3.11, p = 0.043) (Tables 2 & 3).

Sex showed a borderline association with hypertension, with males having slightly higher odds than females (AOR = 1.42, 95% CI: 0.98-2.06, p = 0.065), although this was not statistically significant.

In contrast, education level (primary: AOR = 1.12, 95% CI: 0.68–1.85; secondary: AOR = 1.34, 95% CI: 0.78-2.30; tertiary: AOR = 0.89, 95% CI: 0.42–1.89; overall p = 0.431), tobacco use (AOR = 1.12, 95% CI: 0.65-1.93, p = 0.682), family history of hypertension (AOR = 1.45, 95% CI: 0.92–2.29, p = 0.109), and high salt intake (AOR = 1.23, 95% CI: 0.85-1.78, p = 0.272) were not independently associated with hypertension after adjusting for other factors (Tables 2 and 3).

3.9.    Logistic Regression Analysis for Predictors of Diabetes

The diabetes model similarly exhibited strong predictive performance, explaining a substantial proportion of outcome variability (Cox & Snell R² = 0.346; Nagelkerke R² = 0.697; McFadden R² = 0.619). Model calibration was excellent (Hosmer–Lemeshow χ² = 7.18, p = 0.512), with superior classification accuracy of 82.7%. (Table 1)

Age was the strongest independent predictor of diabetes, showing a clear stepwise increase in risk with advancing age. Compared with adults aged 18-30 years, the odds of having diabetes were significantly higher among those aged 31-45 years (AOR = 2.67, 95% CI: 1.18-6.04), 46-60 years (AOR = 4.12, 95% CI: 1.78-9.52), and ≥61 years (AOR = 6.89, 95% CI: 2.67-17.78), with risk increasing steadily across age groups (Tables 2 & 3).

Alcohol consumption was independently associated with more than a twofold increase in diabetes risk (AOR = 2.14, 95% CI: 1.14-4.01, p = 0.017), representing the strongest modifiable behavioral determinant identified. (Table 2 &3)

Sex showed a borderline association with the outcome, with males having slightly higher odds than females, but this was not statistically significant (AOR = 1.67, 95% CI: 0.96-2.91, p = 0.069).

In contrast, educational level (overall p = 0.209), tobacco use (AOR = 0.94, 95% CI: 0.48-1.85, p = 0.861), and high salt intake (AOR = 1.18, 95% CI: 0.67-2.08, p = 0.567) were not associated with the outcome.

Body mass index also showed no significant association (AOR = 1.04, 95% CI: 0.99-1.09, p = 0.134). Similarly, family history of hyperglycemia was not statistically significant with an overall p-value of 0.501 (Tables 2 and 3).

Likelihood ratio tests confirmed age (Δχ² = 13.654, p = 0.003) and alcohol consumption (Δχ² = 4.597, p = 0.032) as the principal drivers of diabetes risk in this population.

3.10.  Comparative Risk: Hypertension and Diabetes

When comparing both diseases, a similar pattern emerged, with age and alcohol use being the main common risk factors.

Age had a stronger effect on hypertension than on diabetes. For example, people aged ≥61 years had higher odds of hypertension (AOR = 8.76) than diabetes (AOR = 6.89), suggesting that aging has a stronger impact on blood pressure than on blood sugar control.

Alcohol consumption increased the risk of both conditions, but its relative effect was slightly stronger for diabetes (AOR = 2.14) than for hypertension (AOR = 1.78), suggesting it may influence metabolic processes more strongly than blood pressure regulation.

BMI was an important risk factor for hypertension (AOR = 1.09 per unit increase) but was not significantly related to diabetes (AOR = 1.04, not significant). This suggests that excess body weight may play a stronger role in raising blood pressure than in causing diabetes in this rural population.

Sex showed a similar weak association in both conditions, with males having slightly higher odds, though not at a statistically significant level. Other factors such as education, smoking, and salt intake did not independently predict either hypertension or diabetes.

4.          Discussion

4.1.     Predictors of Hypertension

The multivariable analysis suggests that hypertension in this rural sampled population is mainly driven by a few key factors, especially age, body mass index, and alcohol use, rather than a wide range of social or lifestyle variables.

Age was the strongest predictor in the model. Compared with adults aged 18-30 years, the odds of hypertension increased steadily with age, from about two times higher in those aged 31-45 years to nearly nine times higher in those aged 61 years and above. This pattern reflects the ongoing epidemiological transition in rural Ghana, where non-communicable diseases such as hypertension become more common with aging populations. This finding is consistent with other studies in Ghana. For example, in the Nyive Rural Community study in the Volta Region, Osei-Yeboah and colleagues also found that age was the most important predictor of hypertension, with risk increasing gradually each year and a clear rise in prevalence from midlife onwards.^[20] Similarly, multilevel analysis of SAGE Ghana data showed that being aged 50 years and above was one of the strongest and most consistent predictors of hypertension across the country.^[21] Urban studies show a consistent pattern. In the RODAM study, older age was a key risk factor for hypertension among Ghanaians living in rural Ghana, urban Ghana, and Europe.^[22] Similarly, more recent studies from the urban poor community of Ga Mashie in Accra also found that older adults had much higher odds of having hypertension.^[23] Across sub-Saharan Africa, age is consistently the strongest predictor of hypertension. Evidence from rural Tanzania, Uganda, and Malawi shows that adults aged over 50 have about 2 to 4 times higher odds of hypertension compared with younger adults.^[24] These consistent findings show that the strong age pattern seen in Akatsi South is not unique. Instead, it reflects a wider regional trend where longer exposure to cardiometabolic risk factors, combined with limited or delayed screening, leads to a much higher risk of hypertension among older adults in rural areas.

A clear dose-response relationship between BMI and hypertension has also been reported in other studies. For example, multilevel analysis of SAGE Wave 2 data in Ghana found that obesity is an important risk factor for high systolic blood pressure, and it recommended focusing prevention and control efforts on people with higher BMI.^[25] In the RODAM study, higher BMI was independently associated with hypertension among urban Ghanaian men, whereas waist circumference was more predictive among rural men, suggesting context-specific patterns of adiposity but a consistent underlying influence of excess weight.^[26] Rural data from Asante Akim North further support this, with Brenyah and colleagues. reporting higher odds of hypertension among overweight and obese adults in a multivariable model, even after adjusting for age and other lifestyle factors.^[27] Similar findings have been reported across the region. Studies in rural communities in Tanzania and Uganda show that overweight or obese adults have about 1.5 to 2 times higher odds of having hypertension, even though diet and physical activity patterns differ across these settings.^{[28] [29]}

Alcohol consumption was also a significant independent predictor of hypertension in this study, which is consistent with findings from other studies. The effect size is similar to that reported in the urban poor Ga Mashie cohort, where people who currently consume alcohol had about 2.6 times higher odds of having hypertension,^[30] and in rural Asante Akim North, where Brenyah and colleagues found alcohol use to be associated with both hypertension and diabetes in multivariable models.^[31] In northern Ghana, Dapare and colleagues similarly identified alcohol use as one of the behavioral factors associated with hypertension in rural communities.^[32] The study therefore reinforces alcohol consumption as a modifiable risk factor for prevention, particularly for adults in midlife, especially men. It also suggests that rural non-communicable disease prevention strategies should explicitly include messages on reducing alcohol use.

The slightly higher odds in men compared to women are consistent with mixed findings in Ghana and other settings. In many studies, differences between males and females weaken or even reverse after accounting for factors such as lifestyle behaviors and body size.^{[33] [34]} In this rural study, education was not independently linked to the outcome. This is different from findings in urban Ghana, where people with higher education are usually less likely to be affected.^{[35] [36]}

The lack of a significant association between smoking and self-reported salt intake is consistent with findings from other rural studies in Ghana and across Africa. For example, in Asante Akim North, tobacco use was not an independent predictor of hypertension after adjusting for other factors, likely because smoking was uncommon and tended to occur alongside other risk behaviors.^[37] Babagoli and colleagues also found no clear link between smoking and hypertension in rural northern Ghana. They suggested this may be due to the very low prevalence of smoking among women and possible underreporting because of social desirability bias.^[38]

4.2.    Predictors of Diabetes

The multivariable logistic regression results for diabetes in this rural Ghanaian population show that age and alcohol use are the main predictors of risk. After adjusting for other factors, body fatness (adiposity) shows a weaker effect than expected. This pattern is partly similar to other studies in Ghana and sub-Saharan Africa, but also suggests that diabetes risk in rural settings may follow a slightly different pattern linked to local lifestyle and stages of nutritional and lifestyle change.

Age was the strongest independent risk factor. The likelihood of disease increased steadily with age, from about 2.7 times higher in older adults to nearly 7 times higher among adults aged 61 years and above compared with younger adults. This age-related pattern is consistent with findings from other studies.^{[39] [40] [41] [42]} The STEPS survey data in Ghana show a similar pattern of increasing diabetes with age, with higher prevalence among adults over 45 years. ^[43] However, unlike in high-income countries, where routine age-based screening is part of primary care, rural Ghana lacks organized screening for people without symptoms. Fasting blood glucose testing is rarely done routinely and is mostly limited to occasional outreach programs. As a result, the strong age effect seen in this study likely reflects both a true biological increase in risk with age and missed diagnosis among older people due to limited screening. Overall, rural health services tend to detect diabetes late, mainly when symptoms appear, rather than through early, risk-based screening.

Alcohol consumption emerged as an independent risk factor, indicating a modifiable behavior that likely contributes to disease risk in this rural population. It may also reflect broader social and occupational stress patterns, particularly among men, and showed one of the strongest effects in the model, suggesting importance beyond its known cardiovascular effects. Biologically, alcohol can interfere with normal glucose regulation by reducing hepatic glucose production, damaging insulin-producing pancreatic beta cells, and promoting the accumulation of fat around internal organs. These mechanisms can increase the risk of dysglycaemia.^[44] In rural agrarian Ghana, alcohol use is often tied to work patterns and social life, which can add to long-term metabolic strain. Evidence from sub-Saharan Africa shows a modest but consistent link between heavy alcohol consumption and new cases of diabetes, especially among men.^[45] The policy implications here differ slightly from those of hypertension. Blood pressure control can often be achieved through standardized medication, but preventing diabetes requires long-term changes in behavior, especially diet, alcohol use, and physical activity. In rural Ghana, the Community-based Health Planning and Services system is well positioned for community outreach but lacks sufficient resources to consistently support lifestyle counseling. The independent effect of alcohol use, therefore, points to an important and modifiable risk factor that could be targeted for early prevention. However, current health service delivery systems do not systematically focus on reducing harmful alcohol consumption as part of diabetes prevention.

The lack of statistical significance for the association between BMI and diabetes should be interpreted with caution. The presence of Hessian matrix singularities and sparse data convergence issues suggests statistical instability rather than the absence of a biological relationship. Evidence from sub-Saharan Africa consistently shows that overweight and obesity are important risk factors for type 2 diabetes. In rural Ghana, BMI variation may be limited, and BMI itself may not adequately capture fat distribution. Central obesity, which is more strongly linked to metabolic risk, is not well represented by simple BMI measures. More flexible approaches, such as restricted cubic splines or fractional polynomial models, would better capture potential non-linear relationships and identify meaningful adiposity thresholds in this population.^[46]

5.       Conclusion and Recommendations

Age, BMI, and alcohol use are the main independent predictors of hypertension, while age and alcohol are the key drivers of diabetes risk, indicating that both biological factors (such as aging and body composition) and behavioral factors (such as alcohol consumption) are the main causes of cardiometabolic disease in this rural Ghanaian population. The results also show that cardiometabolic diseases in remote rural areas are not uncommon, not harmless, and not only caused by urban lifestyles. Instead, they arise from a combination of biological, behavioral, and social factors, and therefore require coordinated public health responses.

Primary care services should introduce routine screening for blood pressure and blood glucose for all adults aged 30 and older, with special focus on adults aged 30-60 years, where risk increases most rapidly. Alcohol control efforts should be adapted to the rural context, especially for men, using community education, regulation of palm wine production and sales, and promoting safer social alternatives to heavy drinking. Preventing obesity should be linked to local agricultural and household practices, such as encouraging cultivation and consumption of vegetables, improving access to nutrient-rich staple crops, and providing simple weight monitoring at primary health facilities (CHPS compounds). In addition, simple digital health registers supported by community health workers can help track patients over time, improve follow-up, and shift care from occasional visits to continuous management, reducing long-term complications in this high-risk rural population.

6. Conflict of Interest

The author states that there is no conflict of interest.

Table 1: Model Fit and Performance Statistics for Multivariable Logistic Regression Models Predicting Hypertension and Diabetes

Note. **pp < .001. Models adjusted for demographic, behavioral, and clinical covariates.

Both models demonstrated good fit (Hosmer-Lemeshow pp > .05) and substantial explanatory power.

Table 2: Likelihood Ratio Tests for Predictor Significance in Multivariable Logistic Regression Models

Note. Δχ² = change in model chi-square when predictor removed. p < .05*, p* < .01, p < .001.

Age, alcohol consumption, and BMI emerged as significant predictors.

Table 3: Adjusted Odds Ratios from Multivariable Logistic Regression Predicting Hypertension and Diabetes

Note. AOR = adjusted odds ratio; CI = confidence interval; ref = reference category.

p < .05*, p* < .01, p < .001. Models controlled for all listed covariates.

Table 4: Comparative Summary of Key Predictors of Hypertension and Diabetes

Note. AOR = adjusted odds ratio; CI = confidence interval.

Only statistically significant or marginally significant predictors shown.

p < .05*, p* < .01, p < .001.

References

Aheto, Justice Moses K., and Getachew A. Dagne. ‘Multilevel Modeling, Prevalence, and Predictors of Hypertension in Ghana: Evidence from Wave 2 of the World Health Organization’s Study on Global AGEing and Adult Health’. Health Science Reports 4, no. 4 (2021): e453. https://doi.org/10.1002/hsr2.453.

Amoah, Albert G. B., Samuel K. Owusu, and Samuel Adjei. ‘Diabetes in Ghana: A Community Based Prevalence Study in Greater Accra’. Diabetes Research and Clinical Practice 56, no. 3 (2002): 197–205. https://doi.org/10.1016/S0168-8227(01)00374-6.

Appiah, Francis, Edward Kwabena Ameyaw, Joseph Kojo Oduro, et al. ‘Rural-Urban Variation in Hypertension among Women in Ghana: Insights from a National Survey’. BMC Public Health 21, no. 1 (2021): 2150. https://doi.org/10.1186/s12889-021-12204-7.

Asante, Dorothy O., Anita N. Walker, Theodora A. Seidu, Senam A. Kpogo, and Jianjun Zou. ‘Hypertension and Diabetes in Akatsi South District, Ghana: Modeling and Forecasting’. BioMed Research International 2022, no. 1 (2022): 9690964. https://doi.org/10.1155/2022/9690964.

Babagoli, Masih A., Dorothy Adu-Amankwah, Engelbert A. Nonterah, et al. ‘Sociodemographic and Behavioral Factors Associated With Hypertension and Depression in 4 Rural Communities in Northern Ghana: A Cross-Sectional Study’. Journal of Primary Care & Community Health 15 (January 2024): 21501319241242965. https://doi.org/10.1177/21501319241242965.

Brenyah, Joseph Kwasi, Yeetey Enuameh, Bernard Okoe-Boye, et al. ‘Factors Associated with Hypertension and Diabetes in Rural Communities in the Asante Akim North Municipality of Ghana’. Preprint, Zenodo, 18 December 2023. https://doi.org/10.5281/ZENODO.10215030.

Ceriello, Antonio, and Stephen Colagiuri. ‘IDF Global Clinical Practice Recommendations for Managing Type 2 Diabetes – 2025’. Diabetes Research and Clinical Practice 222 (April 2025): 112152. https://doi.org/10.1016/j.diabres.2025.112152.

Chong, Bryan, Jayanth Jayabaskaran, Silingga Metta Jauhari, et al. ‘Global Burden of Cardiovascular Diseases: Projections from 2025 to 2050’. European Journal of Preventive Cardiology 32, no. 11 (2025): 1001–15. https://doi.org/10.1093/eurjpc/zwae281.

Chow, Edith W. K., Yingnan Fan, Hongjiang Wu, et al. ‘Age-Specific Associations between Blood Pressure and Cardiovascular Disease, Kidney Disease, and Death among Individuals with Type 2 Diabetes: A Population-Based Cohort Study’. Cardiovascular Diabetology 25, no. 1 (2026): 60. https://doi.org/10.1186/s12933-025-03072-1.

Gafane-Matemane, Lebo F., Ashleigh Craig, Ruan Kruger, et al. ‘Hypertension in Sub-Saharan Africa: The Current Profile, Recent Advances, Gaps, and Priorities’. Journal of Human Hypertension, ahead of print, 2 May 2024. https://doi.org/10.1038/s41371-024-00913-6.

Gafane-Matemane, Lebo F., Ashleigh Craig, Ruan Kruger, et al. ‘Hypertension in Sub-Saharan Africa: The Current Profile, Recent Advances, Gaps, and Priorities’. Journal of Human Hypertension, ahead of print, 2 May 2024. https://doi.org/10.1038/s41371-024-00913-6.

Ghana Health Service. GHANA STEPS  REPORT 2023:  NATIONWIDE NON-COMMUNICABLE  DISEASES RISK FACTORS ASSESSMENT  USING THE WORLD HEALTH ORGANIZATION’S  STEPWISE APPROACH IN GHANA. GHS Press, 2024. file:///C:/Users/user/Downloads/ghana-2023-steps-report.pdf.

Hashmi, Shahnawaz, Iqra Safdar, Muhammad Hazqeel Kazmi, et al. ‘Rising Mortality Related to Diabetes Mellitus and Hypertension: Trends and Disparities in the United States (1999−2023)’. Clinical Cardiology 48, no. 4 (2025): e70132. https://doi.org/10.1002/clc.70132.

Hinneh, Thomas, Bernard Mensah, Oluwabunmi Ogungbe, Jonathan Bayuo, Emmanuel Timmy Donkoh, and Yvonne Commodore-Mensah. ‘Prevalence of Suboptimal Blood Pressure, Glycemic Control, and Associated Factors among Patients with Diabetes and Hypertension in Primary Health Care Facilities in Ghana: A Multicenter Retrospective Cross-Sectional Study’. BMC Primary Care 26, no. 1 (2025): 189. https://doi.org/10.1186/s12875-025-02775-4.

Hua, Yuchen, Thor S. Stead, Andrew George, and Latha Ganti. ‘Clinical Risk Prediction with Logistic Regression: Best Practices, Validation Techniques, and Applications in Medical Research’. Academic Medicine & Surgery, ahead of print, 7 March 2025. https://doi.org/10.62186/001c.131964.

Institute for Health Metrics and Evaluation (IHME). Global Burden of Disease 2021: Findings from the GBD 2021 Study. Seattle, WA, 2024.

Kansiime, Grace, Edwin Nuwagira, Paul Obwoya, et al. ‘Prevalence, Awareness, and Factors Associated with Hypertension Among Adults in Rural Southwestern Uganda: A Baseline Survey’. International Journal of General Medicine Volume 18 (June 2025): 3289–300. https://doi.org/10.2147/IJGM.S522911.

Kansiime, Grace, Edwin Nuwagira, Paul Obwoya, et al. ‘Prevalence, Awareness, and Factors Associated with Hypertension Among Adults in Rural Southwestern Uganda: A Baseline Survey’. International Journal of General Medicine Volume 18 (June 2025): 3289–300. https://doi.org/10.2147/IJGM.S522911.

Knott, Craig, Steven Bell, and Annie Britton. ‘Alcohol Consumption and the Risk of Type 2 Diabetes: A Systematic Review and Dose-Response Meta-Analysis of More Than 1.9 Million Individuals From 38 Observational Studies’. Diabetes Care 38, no. 9 (2015): 1804–12. https://doi.org/10.2337/dc15-0710.

Kosami, Koki, Masanari Kuwabara, Akira Okayama, and Ryusuke Ae. ‘Body Mass Index and Weight Change as Predictors of Hypertension Development: A Sex-Specific Analysis’. Nutrients 17, no. 1 (2024): 119. https://doi.org/10.3390/nu17010119.

Mayl, Jonathan J., Charles A. German, Alain G. Bertoni, et al. ‘Association of Alcohol Intake With Hypertension in Type 2 Diabetes Mellitus: The ACCORD Trial’. Journal of the American Heart Association 9, no. 18 (2020): e017334. https://doi.org/10.1161/JAHA.120.017334.

Mosha, Neema R., Michael Mahande, Adinan Juma, et al. ‘Prevalence,Awareness and Factors Associated with Hypertension in North West Tanzania’. Global Health Action 10, no. 1 (2017): 1321279. https://doi.org/10.1080/16549716.2017.1321279.

Negussie, Yohannes Mekuria, and Abel Tezera Abebe. ‘Hypertension and Associated Factors among Patients with Diabetes Mellitus Attending a Follow-up Clinic in Central Ethiopia’. Scientific Reports 15, no. 1 (2025): 13150. https://doi.org/10.1038/s41598-025-97909-0.

Osei-Yeboah, James, Ellis Owusu-Dabo, William K. B. A. Owiredu, Sylvester Yao Lokpo, Francis Delali Agode, and Beatrice Bella Johnson. ‘Community Burden of Hypertension and Treatment Patterns: An in-Depth Age Predictor Analysis: (The Rural Community Risk of Non-Communicable Disease Study – Nyive Phase I)’. PLOS ONE 16, no. 8 (2021): e0252284. https://doi.org/10.1371/journal.pone.0252284.

Pastakia, Sonak, Chelsea Pekny, Simon Manyara, and Lydia Fischer. ‘Diabetes in Sub-Saharan Africa &ndash; from Policy to Practice to Progress: Targeting the Existing Gaps for Future Care for Diabetes’. Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy Volume 10 (June 2017): 247–63. https://doi.org/10.2147/DMSO.S126314.

Sanuade, Olutobi Adekunle, Leonard Baatiema, Irene Akwo Kretchy, et al. ‘The Cascade of Hypertension Prevalence, Awareness, Treatment, and Control in Urban-Poor Communities in Accra, Ghana: A Population-Based Household Survey’. BMC Public Health 25, no. 1 (2025): 4288. https://doi.org/10.1186/s12889-025-25409-x.

Sarfo-Kantank, O., E. Owusu-Dabo, F. Adomako-Boateng, B. Eghan, J. Dogbe, and G. Bedu-Addo. ‘An Assessment of Prevalence and Risk Factors for Hypertension and Diabetes during World Diabetes Day Celebration in Kumasi, Ghana’. East African Journal of Public Health 11, no. 2 (2014): 805–15.

Tumin, Dmitry, Mary Hayney, and Rebecca P. Winsett. ‘Describing and Presenting Multivariable Regression Models’. Progress in Transplantation 30, no. 4 (2020): 303–5. https://doi.org/10.1177/1526924820959774.

Van Der Linden, Eva L., Karlijn A. C. Meeks, Kerstin Klipstein-Grobusch, et al. ‘Hypertension Determinants among Ghanaians Differ According to Location of Residence: RODAM Study’. Journal of Hypertension 40, no. 5 (2022): 1010–18. https://doi.org/10.1097/HJH.0000000000003108.

Yuan, Lei, Chen Qu, Jinhang Zhao, et al. ‘Dose-Response Relationship between Body Mass Index and Hypertension: A Cross-Sectional Study from Eastern China’. Preventive Medicine Reports 46 (October 2024): 102852. https://doi.org/10.1016/j.pmedr.2024.102852.

Zhou, Bin, Archie W. Rayner, Edward W. Gregg, et al. ‘Worldwide Trends in Diabetes Prevalence and Treatment from 1990 to 2022: A Pooled Analysis of 1108 Population-Representative Studies with 141 Million Participants’. The Lancet 404, no. 10467 (2024): 2077–93. https://doi.org/10.1016/S0140-6736(24)02317-1.

[1] Shahnawaz Hashmi et al., “Rising Mortality Related to Diabetes Mellitus and Hypertension: Trends and Disparities in the United States (1999−2023),” Clinical Cardiology 48, no. 4 (April 2025): e70132, https://doi.org/10.1002/clc.70132.

[2] Bryan Chong et al., “Global Burden of Cardiovascular Diseases: Projections from 2025 to 2050,” European Journal of Preventive Cardiology 32, no. 11 (August 2025): 1001–15, https://doi.org/10.1093/eurjpc/zwae281.

[3] Justice Moses K. Aheto and Getachew A. Dagne, “Multilevel Modeling, Prevalence, and Predictors of Hypertension in Ghana: Evidence from Wave 2 of the World Health Organization’s Study on Global AGEing and Adult Health,” Health Science Reports 4, no. 4 (December 2021): e453, https://doi.org/10.1002/hsr2.453.

[4] Jun Ma et al., “Signaling Pathways in Vascular Function and Hypertension: Molecular Mechanisms and Therapeutic Interventions,” Signal Transduction and Targeted Therapy 8, no. 1 (April 2023): 168, https://doi.org/10.1038/s41392-023-01430-7.

[5] Yohannes Mekuria Negussie and Abel Tezera Abebe, “Hypertension and Associated Factors among Patients with Diabetes Mellitus Attending a Follow-up Clinic in Central Ethiopia,” Scientific Reports 15, no. 1 (April 2025): 13150, https://doi.org/10.1038/s41598-025-97909-0.

[6] O. Sarfo-Kantank et al., “An Assessment of Prevalence and Risk Factors for Hypertension and Diabetes during World Diabetes Day Celebration in Kumasi, Ghana,” East African Journal of Public Health 11, no. 2 (October 2014): 805–15.

[7] Joseph Kwasi Brenyah et al., “Factors Associated with Hypertension and Diabetes in Rural Communities in the Asante Akim North Municipality of Ghana,” preprint, Zenodo, December 18, 2023, https://doi.org/10.5281/ZENODO.10215030.

[8] Jonathan J. Mayl et al., “Association of Alcohol Intake With Hypertension in Type 2 Diabetes Mellitus: The ACCORD Trial,” Journal of the American Heart Association 9, no. 18 (September 2020): e017334, https://doi.org/10.1161/JAHA.120.017334.

[9] Lei Yuan et al., “Dose-Response Relationship between Body Mass Index and Hypertension: A Cross-Sectional Study from Eastern China,” Preventive Medicine Reports 46 (October 2024): 102852, https://doi.org/10.1016/j.pmedr.2024.102852.

[10] Koki Kosami et al., “Body Mass Index and Weight Change as Predictors of Hypertension Development: A Sex-Specific Analysis,” Nutrients 17, no. 1 (December 2024): 119, https://doi.org/10.3390/nu17010119.

[11] Bin Zhou et al., “Worldwide Trends in Diabetes Prevalence and Treatment from 1990 to 2022: A Pooled Analysis of 1108 Population-Representative Studies with 141 Million Participants,” The Lancet 404, no. 10467 (November 2024): 2077–93, https://doi.org/10.1016/S0140-6736(24)02317-1.

[12] Bin Zhou et al., “Worldwide Trends in Hypertension Prevalence and Progress in Treatment and Control from 1990 to 2019: A Pooled Analysis of 1201 Population-Representative Studies with 104 Million Participants,” The Lancet 398, no. 10304 (September 2021): 957–80, https://doi.org/10.1016/S0140-6736(21)01330-1.

[13] James Attom Nkrumah, Margaretta Gloria Chandi, and Francis Kwaku Wuni, “Prevalence of Hypertension and Its Associated Factors among Rural Adults in a Ghanaian Municipality,” International Journal of Africa Nursing Sciences 24 (2026): 101014, https://doi.org/10.1016/j.ijans.2026.101014.

[14] William Kofi Bosu and Dary Kojo Bosu, “Prevalence, Awareness and Control of Hypertension in Ghana: A Systematic Review and Meta-Analysis,” PLOS ONE 16, no. 3 (March 2021): e0248137, https://doi.org/10.1371/journal.pone.0248137.

[15] Ghana Health Service, GHANA STEPS  REPORT 2023:  NATIONWIDE NON-COMMUNICABLE  DISEASES RISK FACTORS ASSESSMENT  USING THE WORLD HEALTH ORGANIZATION’S  STEPWISE APPROACH IN GHANA (Accra: GHS Press, 2024), file:///C:/Users/user/Downloads/ghana-2023-steps-report.pdf.

[16] Olutobi Adekunle Sanuade, Sandra Boatemaa, and Mawuli Komla Kushitor, “Hypertension Prevalence, Awareness, Treatment and Control in Ghanaian Population: Evidence from the Ghana Demographic and Health Survey,” PLOS ONE 13, no. 11 (November 2018): e0205985, https://doi.org/10.1371/journal.pone.0205985.

[17] Brenyah et al., “Factors Associated with Hypertension and Diabetes in Rural Communities in the Asante Akim North Municipality of Ghana.”

[18] Yakubu Alhassan et al., “Determinants of Blood Pressure and Blood Glucose Control in Patients with Co-Morbid Hypertension and Type 2 Diabetes Mellitus in Ghana: A Hospital-Based Cross-Sectional Study,” PLOS Global Public Health 2, no. 12 (December 2022): e0001342, https://doi.org/10.1371/journal.pgph.0001342.

[19] Joseph Kwasi Brenyah et al., “Factors Associated with Hypertension and Diabetes in Rural Communities in the Asante Akim North Municipality of Ghana,” version 6, with Kwame Nkrumah University of Science and Technology, Dryad, December 18, 2023, 470707 bytes, 470707 bytes, https://doi.org/10.5061/DRYAD.NZS7H44XW.

[20] James Osei-Yeboah et al., ‘Community Burden of Hypertension and Treatment Patterns: An in-Depth Age Predictor Analysis: (The Rural Community Risk of Non-Communicable Disease Study – Nyive Phase I)’, PLOS ONE 16, no. 8 (2021): e0252284, https://doi.org/10.1371/journal.pone.0252284.

[21] Aheto and Dagne, ‘Multilevel Modeling, Prevalence, and Predictors of Hypertension in Ghana’.

[22] Eva L. Van Der Linden et al., ‘Hypertension Determinants among Ghanaians Differ According to Location of Residence: RODAM Study’, Journal of Hypertension 40, no. 5 (2022): 1010–18, https://doi.org/10.1097/HJH.0000000000003108.

[23] Olutobi Adekunle Sanuade et al., ‘The Cascade of Hypertension Prevalence, Awareness, Treatment, and Control in Urban-Poor Communities in Accra, Ghana: A Population-Based Household Survey’, BMC Public Health 25, no. 1 (2025): 4288, https://doi.org/10.1186/s12889-025-25409-x.

[24] Grace Kansiime et al., ‘Prevalence, Awareness, and Factors Associated with Hypertension Among Adults in Rural Southwestern Uganda: A Baseline Survey’, International Journal of General Medicine Volume 18 (June 2025): 3289–300, https://doi.org/10.2147/IJGM.S522911.

[25] Aheto and Dagne, ‘Multilevel Modeling, Prevalence, and Predictors of Hypertension in Ghana’.

[26] Van Der Linden et al., ‘Hypertension Determinants among Ghanaians Differ According to Location of Residence’.

[27] Brenyah et al., ‘Factors Associated with Hypertension and Diabetes in Rural Communities in the Asante Akim North Municipality of Ghana’.

[28] Grace Kansiime et al., ‘Prevalence, Awareness, and Factors Associated with Hypertension Among Adults in Rural Southwestern Uganda: A Baseline Survey’, International Journal of General Medicine Volume 18 (June 2025): 3289–300, https://doi.org/10.2147/IJGM.S522911.

[29] Neema R. Mosha et al., ‘Prevalence,Awareness and Factors Associated with Hypertension in North West Tanzania’, Global Health Action 10, no. 1 (2017): 1321279, https://doi.org/10.1080/16549716.2017.1321279.

[30] Sanuade et al., ‘The Cascade of Hypertension Prevalence, Awareness, Treatment, and Control in Urban-Poor Communities in Accra, Ghana’.

[31] Brenyah et al., ‘Factors Associated with Hypertension and Diabetes in Rural Communities in the Asante Akim North Municipality of Ghana’.

[32] Masih A. Babagoli et al., ‘Sociodemographic and Behavioral Factors Associated With Hypertension and Depression in 4 Rural Communities in Northern Ghana: A Cross-Sectional Study’, Journal of Primary Care & Community Health 15 (January 2024): 21501319241242965, https://doi.org/10.1177/21501319241242965.

[33] Lebo F. Gafane-Matemane et al., ‘Hypertension in Sub-Saharan Africa: The Current Profile, Recent Advances, Gaps, and Priorities’, Journal of Human Hypertension, ahead of print, 2 May 2024, https://doi.org/10.1038/s41371-024-00913-6.

[34] Francis Appiah et al., ‘Rural-Urban Variation in Hypertension among Women in Ghana: Insights from a National Survey’, BMC Public Health 21, no. 1 (2021): 2150, https://doi.org/10.1186/s12889-021-12204-7.

[35] Aheto and Dagne, ‘Multilevel Modeling, Prevalence, and Predictors of Hypertension in Ghana’.

[36] Sanuade et al., ‘The Cascade of Hypertension Prevalence, Awareness, Treatment, and Control in Urban-Poor Communities in Accra, Ghana’.

[37] Brenyah et al., ‘Factors Associated with Hypertension and Diabetes in Rural Communities in the Asante Akim North Municipality of Ghana’.

[38] Babagoli et al., ‘Sociodemographic and Behavioral Factors Associated With Hypertension and Depression in 4 Rural Communities in Northern Ghana’.

[39] Thomas Hinneh et al., ‘Prevalence of Suboptimal Blood Pressure, Glycemic Control, and Associated Factors among Patients with Diabetes and Hypertension in Primary Health Care Facilities in Ghana: A Multicenter Retrospective Cross-Sectional Study’, BMC Primary Care 26, no. 1 (2025): 189, https://doi.org/10.1186/s12875-025-02775-4.

[40] Babagoli et al., ‘Sociodemographic and Behavioral Factors Associated With Hypertension and Depression in 4 Rural Communities in Northern Ghana’.

[41] Bin Zhou et al., ‘Worldwide Trends in Diabetes Prevalence and Treatment from 1990 to 2022: A Pooled Analysis of 1108 Population-Representative Studies with 141 Million Participants’, The Lancet 404, no. 10467 (2024): 2077–93, https://doi.org/10.1016/S0140-6736(24)02317-1.

[42] Antonio Ceriello and Stephen Colagiuri, ‘IDF Global Clinical Practice Recommendations for Managing Type 2 Diabetes – 2025’, Diabetes Research and Clinical Practice 222 (April 2025): 112152, https://doi.org/10.1016/j.diabres.2025.112152.

[43] Ghana Health Service, GHANA STEPS  REPORT 2023:  NATIONWIDE NON-COMMUNICABLE  DISEASES RISK FACTORS ASSESSMENT  USING THE WORLD HEALTH ORGANIZATION’S  STEPWISE APPROACH IN GHANA (GHS Press, 2024), file:///C:/Users/user/Downloads/ghana-2023-steps-report.pdf.

[44] Craig Knott et al., ‘Alcohol Consumption and the Risk of Type 2 Diabetes: A Systematic Review and Dose-Response Meta-Analysis of More Than 1.9 Million Individuals From 38 Observational Studies’, Diabetes Care 38, no. 9 (2015): 1804–12, https://doi.org/10.2337/dc15-0710.

[45] Knott et al., ‘Alcohol Consumption and the Risk of Type 2 Diabetes’.

[46] Institute for Health Metrics and Evaluation (IHME), Global Burden of Disease 2021: Findings from the GBD 2021 Study (Seattle, WA, 2024).