Introduction

Hyperglycaemia is a common metabolic abnormality among acutely ill hospitalised patients and may occur in individuals with previously known diabetes mellitus, previously undiagnosed diabetes, or transient stress hyperglycaemia during acute illness.1,2 In inpatient settings, a random blood glucose concentration of at least 7.8 mmol/L is commonly used to identify clinically significant hyperglycaemia, although this threshold is intended for screening and risk stratification rather than for definitive diagnosis of diabetes mellitus.1 In this context, stress hyperglycaemia refers to elevated blood glucose during acute illness with no evidence of chronic hyperglycaemia, whereas previously undiagnosed diabetes refers to hyperglycaemia occurring in a patient without known diabetes but with glycated haemoglobin levels suggestive of chronic abnormal glucose regulation.2

The definition of inpatient hyperglycaemia varies across studies, with reported cut offs ranging from 5.6 mmol/L to 11.1 mmol/L depending on the study population, sample type, and timing of glucose measurement.3,4 Acute illness can provoke hyperglycaemia through increased secretion of counter‑regulatory hormones such as cortisol, catecholamines, glucagon, and growth hormone, which increase hepatic glucose production and impair peripheral glucose utilisation.5 Stress hyperglycaemia has been associated with adverse outcomes, including increased mortality and longer hospital stay, in both critical care and non-critical care settings.6,7

Although the epidemiology of diabetes is well documented globally, less is known about the burden of inpatient hyperglycaemia among acutely ill adults in many low resource settings. The global burden of diabetes and undiagnosed diabetes continues to rise, particularly in low- and middle-income countries.8,9 In sub-Saharan Africa, studies have reported substantial variation in hyperglycaemia prevalence across different hospital populations, while Ugandan evidence remains limited and largely restricted to selected patient groups such as severe traumatic brain injury or emergency unit attendees with established hyperglycaemia.4,10–12

Understanding hyperglycaemia in newly admitted medical patients is important because it may reflect underlying diabetes, acute physiological stress, or both. In addition, prior diabetes status may directly influence baseline glycaemic control, while illness severity, as reflected by the Modified Early Warning Score, may increase stress‑related hyperglycaemia during acute illness. Education level may also be relevant as a marker of social and behavioural factors that influence long term metabolic risk. However, few Ugandan studies have examined the burden of hyperglycaemia and its associated factors among general adult medical emergency patients, particularly outside the national referral setting. We therefore conducted a cross-sectional study to determine the prevalence of hyperglycaemia and the factors associated with it among newly admitted adult medical patients at the emergency ward of Mbale Regional Referral Hospital in eastern Uganda.

MethodsStudy Design

This was an analytical cross-sectional study conducted in the emergency ward of MRRH in Eastern Uganda from December 2024 to March 2025.

Study Area

This study was conducted at the emergency ward of MRRH, which is in Eastern Uganda, a regional referral hospital located in Mbale City in eastern Uganda. The hospital serves an estimated catchment population of about 4.6 million people across sixteen districts and provides emergency, inpatient, laboratory, and specialist services, including diabetic care and follow‑up. MRRH has 12 yards and a capacity of 450 beds. The Emergency Ward receives is the main entry point for acutely ill adult medical patients requiring urgent assessment and initial stabilisation before transfer to inpatient wards or other appropriate units. During the study period, the ward received approximately 105 adult medical patients per week, and the average length of stay was about 30 hours. This setting was considered appropriate because it enabled assessment of glycaemic status at the point of emergency admission among a broad range of acutely ill adult medical patients.

Study Population

The source population comprised all adult medical patients admitted to the MRRH emergency ward with hyperglycaemia during the study period. The study population included newly admitted adult medical patients aged 18 years or older who were screened at presentation and consented to participate after initial stabilisation. All eligible newly admitted adult medical patients were screened, and hyperglycaemia status was then determined at enrolment. Patients were excluded if they were younger than 18 years, declined participation, or had insufficient history or key data required for classification of study variables.

Sampling Procedure and Enrolment

Consecutive sampling was used. All newly admitted adult medical patients who met the eligibility criteria during the study period were approached consecutively until the required sample size was attained and exceeded. Screening was done at admission to the Emergency Ward. Eligible patients, or surrogates where applicable in line with ward procedures after stabilisation, were given information about the study in English or Lumasaba, and written informed consent was obtained before enrolment.

Data Collection Procedures

Data collection was conducted using a structured interviewer‑administered case record form rather than a semi‑structured questionnaire. The form captured age, sex, education level, smoking history, alcohol use, family history of diabetes, selected medication exposure, vital signs, weight, height, waist circumference, blood pressure, random capillary blood glucose, glycated haemoglobin for participants meeting the glucose threshold, and complete blood count. Data were collected at the point of admission by the study team after patient stabilisation. Bedside clinical assessment, random capillary blood glucose measurement, and blood sampling for glycated haemoglobin and complete blood count were then undertaken at enrolment. The average enrolment rate was approximately 11 participants per day.

Clinical Assessment and MEWS

The Modified Early Warning Scores (MEWS) were used as a pragmatic indicator of illness severity at presentation. MEWS is a composite score derived from routinely measured physiological parameters used in acutely ill patients to support recognition of clinical deterioration. In this study, the score was generated from the vital signs recorded at enrolment and was categorised as less than 5 versus 5 or more, in line with the threshold analysed in the manuscript. To address reviewer concerns, MEWS is now defined explicitly in the methods, and the text clarifies that it was derived from the recorded admission vital signs rather than introduced later only at the analysis stage. MEWS was included because greater physiological derangement during acute illness can increase stress‑related hyperglycaemia through counter‑regulatory hormonal activation.

Anthropometric and Blood Pressure Measurements

Anthropometric measurements included weight, height, body mass index, BMI, and waist circumference. BMI was calculated as weight in kilograms divided by the square of height in metres and categorised as underweight, normal weight, or overweight based on the study coding used in analysis. Waist circumference was classified using the thresholds applied in the original study, namely greater than 94 cm in males or greater than 80 cm in females for elevated central adiposity. Blood pressure was measured at enrolment, and elevated blood pressure was defined as systolic blood pressure greater than 140 mmHg, diastolic blood pressure greater than 90 mmHg, or current use of antihypertensive medication.

Glucose and HbA1c Measurements

Random capillary blood glucose was measured at enrolment for all participants, and hyperglycaemia was defined as a random capillary blood glucose concentration of 7.8 mmol/L or more. This threshold was used as a clinical cut-off for inpatient hyperglycaemia screening and risk stratification in acutely ill patients; it was not used as a standalone diagnostic threshold for diabetes mellitus. The threshold was selected because it is recognised in inpatient diabetes care guidance and has been used in prior hospital-based studies, despite variation across the literature regarding alternative cut‑offs.

Participants with capillary blood glucose of 7.8 mmol/L or more had blood drawn for glycated haemoglobin, HbA1c, and complete blood count. Participants were classified as having stress hyperglycaemia if the random capillary blood glucose was above 7.8 mmol/L and the HbA1c was 6.4% or lower. Participants were classified as having previously undiagnosed diabetes mellitus if the random capillary blood glucose was 7.8 mmol/L or more and the HbA1c was 6.5% or higher, in the absence of a prior diagnosis of diabetes. Participants with a previous history of diabetes mellitus and admission hyperglycaemia were classified under known diabetes with hyperglycaemia.

Study Variables

The dependent variable was hyperglycaemia at admission, coded as a binary outcome based on the random capillary blood glucose threshold of 7.8 mmol/L. Independent variables were grouped into social and demographic variables, clinical variables, anthropometric and metabolic variables, and laboratory variables. Age and sex were included as standard descriptive and potential confounding variables. Prior diabetes status was included because preexisting hyperglycaemia directly influences the likelihood of presenting with hyperglycaemia. MEWS was included as a marker of illness severity because acute physiological stress may precipitate stress hyperglycaemia. Education level was retained as a social determinant proxy that may reflect longer‑term socioeconomic and behavioural influences on metabolic risk. Smoking, alcohol use, BMI, and waist circumference were included because they are established or plausible correlates of metabolic risk and chronic hyperglycaemia. Blood pressure and selected medication exposure were included because cardiovascular and treatment‑related factors may influence glucose metabolism. This rationale has been added to make the conceptual basis for variable selection clearer to readers.

Sample Size Determination

Sample size was calculated using the Kish and Leslie formula for estimation of a single proportion, because the primary objective was to estimate the prevalence of hyperglycaemia among newly admitted adult medical patients. A prevalence of 49.6% was used, taken from a cross-sectional study from Benin that reported a comparable hyperglycaemia related prevalence estimate in an African setting at the time of protocol development.11 The standard assumptions applied were a 95% confidence level, corresponding to a Z value of 1.96, and a margin of error of 5%. This yielded a minimum required sample size of 384 participants. A total of 449 participants were enrolled, which exceeded the minimum required sample size to increase the power of the study.

Data Management and Quality Assurance

Data were collected electronically using KoboToolbox and exported to Microsoft Excel and STATA SE version 18 for cleaning and analysis. The data was checked for completeness, consistency checks between related variables, range checks for continuous measurements, verification of implausible entries against the original electronic forms where possible, and coding of categorical variables before analysis. Participants with missing key data required for exposure or outcome classification were excluded from the final analytic dataset, as reflected in the study flow chart and enrolment summary.

Statistical Analysis

Data analysis was performed using STATA SE version 18. Descriptive analysis summarised continuous variables using means and standard deviations where appropriate, and categorical variables using frequencies and percentages. The prevalence of hyperglycaemia was calculated as the proportion of enrolled participants with random capillary blood glucose of 7.8 mmol/L or more. The prevalence of previously undiagnosed diabetes mellitus was calculated as the proportion of enrolled participants with hyperglycaemia and HbA1c of 6.5% or higher, without a prior diagnosis of diabetes mellitus.

At bivariate level, cross‑tabulations were used to compare hyperglycaemia status across predictor variables, and Pearson’s chi‑square test was used for categorical associations. Variables were considered for inclusion in the multivariable model based on clinical relevance, prior evidence of association with hyperglycaemia, biological plausibility, and statistical evidence from bivariate analysis using a p‑value threshold of <0.20. For multivariable analysis, a modified Poisson regression approach was used to estimate prevalence ratios and 95% confidence intervals. The variables were considered for multivariable modelling based on clinical relevance, prior evidence, and statistical signal at bivariate analysis. The final model retained variables that met these criteria and were central to the study question, including prior diabetes status, MEWS category, education level, sex, tobacco smoking, and waist circumference. Statistical significance was set at p less than 0.05.

Results

Between December 2024 and March 2025, 491 adult medical patients presenting to the emergency ward of Mbale Regional Referral Hospital were screened for eligibility. Of these, 449 were enrolled in the study, while 42 were excluded; 16 were younger than 18 years, 15 did not provide consent, and 11 had incomplete data.

Socio Demographic and Clinical Characteristics of the Study Patients

As shown in Table 1, slightly more than half of the participants were female, 230 of 449, 51.2%. Most participants, 326 of 449, 72.6%, were aged 40 years or older, and the mean age was 53.2 ± 19.8 years. Nearly half, 221 of 449, 49.2%, had attained primary education, while 294 of 449, 65.5%, had normal body mass index. The mean body mass index was 21.7±6.2 Kg/m2.

Table 1 Social Demographic and Clinical Characteristics of the Study Patients

Prevalence and Glycaemic Categories Among Newly Admitted Medical Patients at the Emergency Ward in MRRH

Among the 449 participants included in the study, 192 (42.8%) had random capillary blood glucose levels of at least 7.8 mmol/L at admission and were therefore classified as having hyperglycaemia. Among these 192 participants, 53 (27.6%) had a prior diagnosis of diabetes mellitus, 74 (38.5%), had hyperglycaemia with HbA1c levels of 6.4% or less and were classified as stress hyperglycaemia, and 65 (33.9%), had hyperglycaemia with HbA1c levels of at least 6.5% and were classified as previously undiagnosed diabetes in this study. Overall, the proportion classified as previously undiagnosed diabetes in the full study population was 14.5% (65 of 449). The distribution of these glycaemic categories is shown in Figure 1. No post‑discharge follow‑up was conducted in this cross-sectional study to confirm subsequent glycaemic status beyond the admission assessment.

Figure 1 Study flow diagram.

Factors Associated with Hyperglycaemia Among Newly Admitted Medical Patients

In multivariable analysis, tertiary education level, prior diagnosis of diabetes mellitus, and Modified Early Warning Score of at least 5 remained significantly associated with hyperglycaemia, as shown in Table 2. The bivariate analysis is shown in Supplementary Table 1. Compared with participants with no formal education, those with tertiary education were more likely to have hyperglycaemia (adjusted prevalence ratio [aPR 1.6, 95% CI: 1.1 to 2.3, p = 0.019]). Participants with a prior diagnosis of diabetes mellitus were also more likely to have hyperglycemia (aPR 1.9, 95% CI: 1.6 to 2.4, p < 0.001). Similarly, participants with a Modified Early Warning Score of at least 5 were more likely to have hyperglycaemia than those with a score below 5 (aPR 1.5, 95% CI: 1.2 to 1.9, p = 0.002).

Table 2 Unadjusted and Adjusted Prevalence Ratios of the Factors Associated with Hyperglycaemia

Discussion

This study found that 42.8% of newly admitted adult medical patients at the emergency ward had hyperglycaemia at admission, and that 14.5% of the total study population were classified as having previously undiagnosed diabetes based on admission glucose and HbA1c measurements. Hyperglycaemia was independently associated with tertiary education, prior diagnosis of diabetes mellitus, and Modified Early Warning Score of at least 5. These findings address the study objectives by quantifying the burden of admission hyperglycaemia and identifying clinically relevant factors that may help target early detection in emergency settings.

In this study, close to half of the newly admitted adult medical patients had hyperglycaemia at admission. This prevalence is comparable to reports from Germany and the United States, where admission hyperglycaemia among acutely ill patients was reported at 47% and 42.5%, respectively, but lower than estimates reported in India and Egypt of 55.4% and 64.7%, respectively.10,13–15 These comparisons should, however, be interpreted cautiously because the studies were conducted in different clinical settings, used different glycaemic thresholds, and involved different patient populations and admission diagnoses. Thus, the prevalence estimates reported across these studies should not be interpreted as directly equivalent, because variation in glycaemic definitions, glucose measurement methods, admission timing, and the clinical composition of study populations can substantially influence the proportion classified as having hyperglycaemia.

The random capillary blood glucose threshold of ≥7.8 mmol/L used in this study should be interpreted as a screening and risk stratification cut‑off rather than a diagnostic criterion for diabetes mellitus. This threshold is useful for identifying clinically relevant admission hyperglycaemia among acutely ill hospitalised patients and for prompting further assessment, including HbA1c testing where appropriate, but it does not by itself confirm diabetes mellitus.1 Therefore, the use of this relatively sensitive threshold may have increased the estimated prevalence of admission hyperglycaemia compared with studies that applied higher thresholds such as ≥11.1 mmol/L, fasting glucose criteria, or venous plasma glucose-based definitions. This should be considered when interpreting the 42.8% prevalence observed in this study and when comparing it with estimates from studies that used different glycaemic definitions.

The prevalence observed in our study was also higher than that reported in earlier Ugandan studies. Matovu et al reported a prevalence of 16.2% among patients with severe traumatic brain injury, while the Mulago emergency unit study reported a lower prevalence using a higher glycaemic threshold of 11.1 mmol/L.4,12 The higher prevalence in our study may therefore reflect differences in case mix, admission diagnoses, and glycaemic definitions, as our population comprised general adult medical emergency patients rather than selected neurological or more narrowly defined emergency cohorts.

In our study, 14.5% of the full study population were classified as having previously undiagnosed diabetes based on admission hyperglycaemia together with raised HbA1c. This finding is clinically important because it suggests that a substantial proportion of adult medical emergency patients may have previously unrecognised chronic hyperglycaemia. It is also broadly consistent with International Diabetes Federation estimates showing a high burden of undiagnosed diabetes in sub–Saharan Africa.8 However, because our classification was based on admission measurements without post‑discharge follow up, this subgroup should be interpreted as probable previously undiagnosed diabetes within the context of this study design.

A Modified Early Warning Score of at least 5 was independently associated with hyperglycaemia, suggesting that patients with greater physiological instability were more likely to have elevated glucose levels at admission. This is biologically plausible because severe acute illness activates counter‑regulatory hormones and inflammatory pathways that increase hepatic glucose output and reduce peripheral glucose utilisation, resulting in stress‑related hyperglycaemia.5,6 Our finding is also consistent with evidence that admission hyperglycaemia is associated with greater illness severity and poorer outcomes in acutely ill populations.3,7,16

Prior diagnosis of diabetes mellitus was associated with hyperglycaemia. This could be due to poor glycaemic control among individuals with established diabetes mellitus, as observed in other studies in Ethiopia,17 Uganda18 and Tanzania.19 In Ethiopia, poor knowledge about diabetes control among patients, poor adherence to interventions and therapeutic inertia among health care workers were associated with poor glycaemic control.20

Tertiary education was independently associated with hyperglycaemia in our study. This finding should be interpreted cautiously, as education may be acting as a proxy for broader socioeconomic and behavioural pathways rather than exerting a direct biological effect. In some low- and middle-income settings, higher education is associated with occupations involving less physical activity, greater dietary transition, and increased exposure to sedentary urban lifestyles, all of which may contribute to metabolic risk.17,18,21 At the same time, education may also reflect differences in health‑seeking behaviour, prior diagnosis, or access to care, which could influence the likelihood of identifying hyperglycaemia at presentation. Our finding is therefore consistent with the broader epidemiologic transition described across low- and middle-income countries, but residual socioeconomic confounding cannot be excluded.18,19,22

In contrast, body mass index, waist circumference, smoking history, and alcohol consumption were not significantly associated with hyperglycaemia in the adjusted analysis. This may reflect the acute nature of the study setting, where admission hyperglycaemia is influenced not only by chronic metabolic risk factors but also by immediate illness‑related physiological stress. It is also possible that the sample size within some exposure categories limited statistical power to detect weaker associations. These findings suggest that, in emergency care settings, acute illness severity and prior diabetes status may be more informative predictors of admission hyperglycaemia than conventional behavioural or anthropometric risk factors alone.

From a clinical perspective, these findings support the value of early glucose assessment among newly admitted adult medical patients in emergency settings, particularly among those with known diabetes or greater physiological instability. The finding that approximately one in seven participants were classified as having previously undiagnosed diabetes also suggests that admission screening may identify patients who would otherwise remain undetected. In practice, this may improve emergency triage, prompt inpatient monitoring, and strengthen referral for follow‑up metabolic assessment after discharge. From a public health perspective, the results highlight the need for simple screening pathways that can facilitate earlier recognition of hyperglycaemia in resource‑limited hospital settings.

Strengths and Limitations

The study had several limitations. First, glycaemic subgroup classification was based on admission random capillary blood glucose and HbA1c measurements, and no post‑discharge follow‑up was conducted to confirm subsequent glycaemic status. Accordingly, the categories of stress hyperglycaemia and previously undiagnosed diabetes should be interpreted within the context of this cross-sectional design. Second, other causes of hyperglycaemia, including endocrine disorders such as thyroid or adrenal disease, were not investigated, and the specific type of diabetes could not be determined. Third, pregnancy was ruled out on clinical grounds alone, which may have missed early first‑trimester gestations. Finally, the emergency ward population was clinically heterogeneous, and admitting diagnoses were not formally stratified in the analysis; this may limit comparability with disease‑specific cohorts from other settings and introduces the possibility of residual clinical confounding.

A key strength of this study was the use of HbA1c in addition to admission glucose measurement to improve glycaemic classification beyond single glucose testing alone. This is particularly relevant in a setting where transient stress hyperglycaemia and chronic hyperglycaemia may coexist in acutely ill patients. To the best of our knowledge, this is among the first studies to describe admission hyperglycaemia and the burden of probable previously undiagnosed diabetes among newly admitted adult medical patients in eastern Uganda.

Because glycaemic subgroup classification was based on admission random capillary blood glucose and HbA1c measurements without post discharge follow up, the categories of stress hyperglycaemia and previously undiagnosed diabetes should be interpreted within the context of the study design.

Conclusions

We found that 42.8% of newly admitted adult medical patients at the Mbale Regional Referral Hospital emergency ward had hyperglycaemia at admission, and that 14.5% of the total study population were classified as having previously undiagnosed diabetes based on admission glucose and HbA1c measurements.

Hyperglycaemia was independently associated with prior diagnosis of diabetes mellitus, tertiary education, and Modified Early Warning Score of at least 5. These findings suggest that admission hyperglycaemia is common among adult medical emergency patients and may be particularly relevant in patients with known diabetes and greater illness severity. In practice, these findings may help inform clinical vigilance for hyperglycaemia at admission and support referral for follow up metabolic assessment after discharge.

Although the findings support consideration of early glucose assessment in emergency settings, they should be interpreted considering the single site design, short study period, and clinical heterogeneity of the study population. Future prospective studies are needed to validate these findings in other settings, assess patient outcomes longitudinally, and determine how admission glucose assessment can best inform emergency triage, inpatient monitoring, and follow up pathways.

Data Sharing Statement

The datasets used during our study are available from the corresponding author upon formal request.

Ethics Approval and Consent to Participate

Ethical approval was obtained from the Busitema University Faculty of Health Sciences Research and Ethics Committee (REC), No. BUFHS-2024-228. Administrative clearance was also obtained from Mbale Regional Referral Hospital. Patient care was not interrupted for the purposes of the study; the PI and research assistants approached only stabilised patients for participation. Written informed consent was taken from patients to participate in the study. All data and study documents were stored securely, according to Good Clinical Practice 11 and the principles of the Declaration of Helsinki on secure storage of research materials and confidentiality.

Acknowledgments

We would like to acknowledge the management of Mbale Regional Hospital for permitting us to conduct this study and for their support throughout the data collection period. We are specifically thankful to Brian Makoko who assisted in data analysis and the Emergency ward staff who helped us with the data collection.

Author Contributions

Oreb Nankunda: Writing – review & editing, Writing – original draft, Validation, Project administration, Methodology, Formal analysis, Data curation, Conceptualization. Denis Bwayo: Supervision, Methodology, Formal analysis, Data curation, Writing – original draft. Julius Imalingat: Writing – review & editing, Writing – original draft, Validation. Jasper Kebesu: Writing – review & editing, Writing – original draft, Validation, Methodology, Formal analysis, Data curation, Conceptualization. Erias Kirabo: Data Curation, Validation, Writing – original draft. Peter Masaba: Methodology, Data curation, Conceptualization, Writing – review & editing. Esther Apio: Writing – review & editing, Writing – original draft, Validation, Methodology. Richard Katuramu: Supervision, Methodology, Formal analysis, Data curation, Conceptualization, Writing – original draft. All authors gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agreed to be accountable for all aspects of the work.

Funding

The study was funded by the principal investigator. No external funding was received.

Disclosure

The authors declare no conflict of interest.

References

1. American Diabetes Association Professional Practice Committee. 16. diabetes care in the hospital: standards of medical care in diabetes-2022. Diabetes Care. 2022;45(Suppl 1):S244–11. doi:10.2337/dc22-S016

2. Malcolm J, Halperin I, Miller DB, et al. In-hospital management of diabetes. Can J Diabetes. 2018;42:S115–S123. doi:10.1016/j.jcjd.2017.10.014

3. Evans NR, Dhatariya KK. Assessing the relationship between admission glucose levels, subsequent length of hospital stays, readmission and mortality. Clin Med. 2012;12(2):137–139. doi:10.7861/clinmedicine.12-2-137

4. Matovu P, Kirya M, Galukande M, et al. Hyperglycemia in severe traumatic brain injury patients and its association with thirty-day mortality: a prospective observational cohort study in Uganda. PeerJ. 2021;9:e10589. doi:10.7717/peerj.10589

5. Vedantam D, Poman DS, Motwani L, Asif N, Patel A, Anne KK. Stress-induced hyperglycemia: consequences and management. Cureus. 2022;14(7):e26714. doi:10.7759/cureus.26714

6. Sukkari MA, Fang S, Aiyedipe S, et al. 1149-P: stress hyperglycemia is an independent marker of both acute illness and diabetes risk. Diabetes. 2022;71(Supplement_1). doi:10.2337/db22-1149-P

7. Zhang C, Shen HC, Liang WR, et al. Relationship between stress hyperglycemia ratio and allcause mortality in critically ill patients: results from the MIMIC-IV database. Front Endocrinol. 2023;14:1111026. doi:10.3389/fendo.2023.1111026

8. International Diabetes Federation. IDF Diabetes Atlas. 11th ed. Brussels: International Diabetes Federation; 2025.

9. Sun H, Saeedi P, Karuranga S, et al. IDF diabetes atlas: global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabet Res Clin Pract. 2022;183:109119. doi:10.1016/j.diabres.2021.109119

10. El Toony LF, El Zohri MH, Elghait AAA. Admission hyperglycemia and its implications on outcome in patients attending medical intensive care units at Assiut University Hospital. J Diabetol. 2021;12(2):134–139. doi:10.4103/jod.jod_28_18

11. Amegan NH, Amidou AS, Houehanou CY, et al. Prevalence and factors associated with hyperglycemia in a rural population of Tanvè and Dékanmey in Benin in 2019. PLOS Glob Public Health. 2022;2(5):e0000471. doi:10.1371/journal.pgph.0000471

12. Tumwebaze A, Kiboneka E, Mugalu J, Kikabi EM, Tumwine JK. Prevalence and outcome of stress hyperglycaemia among severely malnourished children admitted to Mulago referral and teaching hospital in Kampala, Uganda. BMC Nutr. 2018;4:49. doi:10.1186/s40795-018-0258-3

13. Liu L, Qian J, Yan W, Liu X, Zhao Y, Che L. Relationship between hyperglycaemia at admission and prognosis in patients with acute myocardial infarction: a retrospective cohort study. Postgrad Med J. 2022. doi:10.1136/pmj-2021-141454

14. Mandal A, Iyyadurai R, Ninan KF, Gunasekaran K. Prevalence of hyperglycemia on admission and its effect on short-term cardiovascular outcomes of non-ST elevation myocardial infarction: a retrospective study from a tertiary care center in South India. Arch Med Health Sci. 2023;11(1):70–74. doi:10.4103/amhs.amhs_28_23

15. Zelihic E, Poneleit B, Siegmund T, Haller B, Sayk F, Dodt C. Hyperglycemia in emergency patients--prevalence and consequences: results of the GLUCEMERGE analysis. Eur J Emerg Med. 2015;22(3):181–187. doi:10.1097/mej.0000000000000199

16. Fadini GP, Morieri ML, Boscari F, et al. Newly diagnosed diabetes and admission hyperglycemia predict COVID-19 severity by aggravating respiratory deterioration. Diabet Res Clin Pract. 2020;168:108374. doi:10.1016/j.diabres.2020.108374

17. Bauman A, Ma G, Cuevas F, et al. Cross-national comparisons of socioeconomic differences in the prevalence of leisure-time and occupational physical activity, and active commuting in six Asia-Pacific countries. J Epidemiol Community Health. 2011;65(1):35–43. doi:10.1136/jech.2008.086710

18. Seiglie JA, Marcus ME, Ebert C, et al. Diabetes prevalence and its relationship with education, wealth, and BMI in 29 low- and middle-income countries. Diabetes Care. 2020;43(4):767–775. doi:10.2337/dc19-1782

19. Williams ED, Tapp RJ, Magliano DJ, Shaw JE, Zimmet PZ, Oldenburg BF. Health behaviours, socioeconomic status and diabetes incidence: the Australian Diabetes Obesity and Lifestyle Study (AusDiab). Diabetologia. 2010;53(12):2538–2545. doi:10.1007/s00125-010-1888-4

20. Dubale M, Gizaw K, Dessalegn D. Magnitude and predictors of poor glycemic control in patients with diabetes at Jimma Medical Center, Ethiopia. Sci Rep. 2023;13(1):15952. doi:10.1038/s41598-023-42774-y

21. Mishra SR, Ghimire S, Shrestha N, Shrestha A, Virani SS. Socio-economic inequalities in hypertension burden and cascade of services: nationwide cross-sectional study in Nepal. J Hum Hypertens. 2019;33(8):613–625. doi:10.1038/s41371-019-0165-3

22. Zhang Y, Tong G, Ma N, et al. Associations between education and ideal cardiovascular health metrics across 36 low- and middle-income countries. BMC Med. 2025;23(1):204. doi:10.1186/s12916-025-04032-y