The true burden of diabetes in sub-Saharan Africa (SSA) is unknown as most of the countries do not have good quality data. As such, the overall estimate of diabetes prevalence is largely based on modelled estimates, which may not be accurate.

Additionally, there is lack of clear guidance on which method and thresholds to use in the diagnosis of diabetes in African populations unlike in high income countries (HIC) where such guidance is clear. The limited data available shows that diabetes in Africa manifests differently for example occurring at younger age and in relatively lean individuals. Moreover, where the oral glucose tolerance test (OGTT) has been used to screen for diabetes, a significant proportion of individuals have isolated postprandial hyperglycaemia (IPH): The reasons for this differential manifestation are unclear and the diabetes progression of these unique phenotypes (for example in terms of risk of complications is unknown or response to treatment is unknown).

Therefore, the overall aim of this research is to undertake a large study to determine the true prevalence of diabetes and identify/characterize the different phenotypes; 2) establish a cohort patients with diabetes to understand the natural course of these different phenotypes, including how they respond to treatment (i.e. do the IPH or thin diabetics progress at the same rate as obese, and are the currently used intervention/therapeutic approaches equally effective in the different phenotypes?).

The collected data is likely to be directly relevant to an improved understanding of the cause and progression of diabetes, diagnostic test performance, and diabetes care in SSA, ultimately leading to better patient outcomes and well-being, as well as enhanced productivity.

The specific objectives will be to; (1) To determine the true prevalence of diabetes and prediabetes in a population-based cohort (2) To identify and characterize the different phenotypes of type 2 diabetes (3) To establish a base-line cohort of well-characterised people with diabetes to understand disease progression of the different phenotypes, including uptake and response to treatment. (4) To identify and recruit a cohort of non-diabetic volunteers (including prediabetes i.e., at high risk of developing diabetes) for longitudinal follow-up.

Study setting: This will be a population-based survey nested within the General Population Cohort (GPC). The GPC is a rural population-based cohort study of people living within the 25 villages of Kyamulibwa sub-county of Kalungu District in rural southwest Uganda. The primary study will be conducted in Uganda. The study shall recruit from 2 sites (rural and urban areas) across the greater Masaka region in southwest Uganda in order to explore the differences in lifestyle, and how these influence prevalence of diabetes and its outcomes; General Population Cohort (GPC) in Kyamulibwa will provide the rural population, and the town of Lukaya the urban population. Both rural and urban sites are established research areas. The entire adult population of the GPC will be eligible for this survey. Data from the most recent Census (2020-2021) in the GPC shows that there are 8,864 adults (aged 18+) of the total population of 20,751 in Kyamulibwa.

Eligibility: All adults aged 18 years and above will be invited to take part in the study. We will recruit all adults who are able to provide written informed consent and are residents at a household (i.e. not a temporary member with intent to leave) in either of the study areas. Individuals will be excluded from the study based on the following criteria: pregnant women (can participate six months after childbirth), live outside the geographical sampling frame for the relevant site or are unable to give informed consent, living outside the relevant site's catchment area.

Sampling method: This study is powered to estimate the prevalence of diabetes and other primary outcomes like prediabetes Impaired fasting glycaemia (IFG), isolated postprandial hyperglycaemia (IPH), isolated-IFG and isolated-Impaired Glucose Tolerance) with sufficient precision. We have conservatively based our assumptions on results from the 2014 Ugandan nation-wide cross-sectional survey suggesting that the prevalence of IFG and diabetes in Uganda is 2% (95% CI: 1.5-2.5%) and 1.4% (95% CI: 0.9 - 1.9%) respectively.3 That survey detected dysglycaemia using fasting blood glucose alone, rather than OGTT. Applying a rule of thumb that the margin of error should not exceed 0.25 of the prevalence and based on the following assumptions: A margin of error of 0.5% (a margin of error of 0.5% to 1% would be acceptable for low prevalent conditions) and Diabetes prevalence of 5% (assuming OGTT would detect more cases of diabetes thus doubling the prevalence) and estimated 5% level of significance (z-value = 1.96). We aim to screen a minimum of 11700 adults across all sites; a minimum of 5,850 from each site. All the adults in Kyamulibwa- the rural site (8, 864) will be invited to take part in our survey. For Lukaya-the urban site; initially, 3 of the 5 cells will be randomly selected and all the adults in the selected cells (approximate total adult population of 6000) invited to take part in the survey. Recruitment will be restricted to individuals who reside within defined areas (villages).

Enrollment: This study will consist of overlapping phases as described below:

Phase One: Main baseline survey. This phase comprises the selection and training of survey team, community sensitisation, obtaining the consent of the study participants, and conducting the baseline survey. All consented individuals will be interviewed using a structured pre-tested questionnaire (Data Collection Form) by trained study staff to collect relevant information including; demographic, socioeconomic and lifestyle in the language understood by the patient. Thereafter, participants will undergo biophysical measurements (weight, height, waist and hip circumferences, Mid-upper arm circumference (MUAC) ) and blood pressure after which they will be invited to come to the research hub in the next days for blood sample collection after an overnight fast. Participants are required to have fasted for at least 8 hours and those who have not will be asked to come back another day in a fasting state. Blood sample collection will be done using standardised procedures. Venous whole blood will be used for HbA1C, complete blood count, genetics and malaria. Serum will be used for biochemistry and HIV. NaF blood tube will be used for fasting plasma glucose. hCG urine testing will be carried out on the urine samples collected.

NB: Those with already clinically diagnosed diabetes and on treatment will come in fasted and baseline fasting samples taken off. They will not undergo OGTTT.

Phase Two: This phase involves assessment of the reproducibility of diagnostic approaches, and comparing their general performance. From Phase I of the study, all those with dysglycaemia (diabetes; prediabetes; i-IFG and i-IGT) and age and sex-matched non-diabetic subjects (1:1) will be invited back to the research hub for further assessment. In Phase 2, the aim will be 1) to assess the reproducibility of the diagnostic tests, notably the OGTT; 2) to assess the burden of glycaemia through continuous glucose monitoring; 3) to identify participants for Phase 3 studies.

Phase Three: This is aimed at assessing and describing any dysglycaemia clusters or phenotypes by comparing baseline characteristics and evaluating the distribution of various measures. Further tests may be done in the future to better understand the differences between the phenotypes. This phase involves utilisation of data generated from phases 1 and 2 to characterise individuals with dysglycaemia further: This is aimed at assessing and describing any dysglycaemia clusters or phenotypes by comparing baseline characteristics and evaluating the distribution of various measures inclusive of biochemical parameters (e.g., fasting and stimulated C-peptide), adiposity and anthropometric indices, continuous glucose measures in daily living. Further tests may be done at this level including vascular complications assessment, endogenous insulin secretion measurement, and exploration of the inflammatory and immune biomarker (including islet autoantibodies) profiles to better understand the differences between the phenotypes including Lean versus Non-lean, IPH vs IFG.

Phase Four: Aimed at the creation of well-characterised diabetic and non-diabetic long-term cohorts. The newly diagnosed diabetes participants will be referred to the local diabetes clinics where they will be managed according to the Uganda Ministry of health guidelines. These patients will be followed up in clinics (supported by the research team) to study natural course, including response to treatment, and disease progression. The non-diabetic participants will be followed in a longitudinal study nested in the GPC to determine the incidence of diabetes.

Data collected from the different clinical sites using the data collection form will be entered into a centralized data management tool (REDCap). All data will be anonymised before being stored in the data management system.

Ethical consideration: All participants will be required to provide written informed consent before participating in the study. Refusal to participate will not affect the quality of care of the participants at the clinical sites. The study has received ethical clearance from the Uganda Virus Research Institute (GC/127/21/09/858)

Inclusion Criteria:

• Age ≥ 18 years Signed informed consent
• Willing to provide written informed consent

Exclusion Criteria:

• Pregnant women - can participate six months after childbirth
• Living outside the geographical sampling frame for the relevant site
• Unable to give informed consent

Study participants are all Individuals aged 18 years or older, residing within the 25 villages of Kyamulibwa sub-county of Kalungu District and Lukaya town and willing to give informed consent.