Improving Precision Diagnosis and Treatment of Diabetes in Low-Resource Settings in Africa Through Structured Clinical Phenotyping: A Narrative Review Authors: Davis Kibirige, Anxious Niwaha, Jean Claude Katte, William Lumu Published: 2026 ABSTRACT Precision diabetes care is an evolving evidence-based, personalised approach that uses an individual’s clinical, metabolic, and genetic data to accurately classify them into distinct subgroups and tailor diabetes prevention, treatment, and monitoring strategies. The use of genomics data, advanced biomarker testing, and artificial intelligence-driven diagnostics has improved precision diabetes care in high-income countries. This has enabled accurate identification of diabetes subtypes and targeted treatment. Despite the rising burden of diabetes, high rates of undiagnosed cases, and emerging evidence of atypical manifestations of diabetes in Africa, such advancements in precision diabetes care remain unavailable in many low-resource settings, directly impacting treatment outcomes. To improve precision diabetes care in low-resource settings in Africa, we suggest using a structured clinical phenotyping approach based on routinely collected clinical characteristics. In this review, we explore how using a combination of readily available clinical characteristics, such as age at diagnosis, body mass index, clinical presentation, family history of diabetes, and coexisting medical conditions, can help healthcare professionals in Africa to diagnose and treat different diabetes subtypes accurately. By adopting a pragmatic, structured clinical phenotyping approach, African healthcare systems can deliver region-specific, cost-effective precision diabetes care without overreliance on advanced laboratory tests and digital technologies.Supplementary InformationThe online version contains supplementary material available at 10.1007/s13300-026-01881-5. Key Summary Points There is a growing burden of diabetes in Africa, in addition to emerging evidence of differential manifestation of diabetes and an increasing frequency of atypical diabetes subtypes.Because of these observations, precision diagnosis and treatment of diabetes are important to improve diabetes outcomes.The use of highly specific tests, such as C-peptide, islet-cell autoantibody, and genetic tests, has aided the progress of precision diabetes care in high-income countries.These advanced tests remain unavailable in most clinical settings in Africa.In this narrative review, we discuss how, in the absence of such expensive laboratory tests, structured clinical phenotyping based on routinely collected clinical characteristics can be used as a pragmatic strategy to advance precision diabetes care in low-resource settings in Africa. There is a growing burden of diabetes in Africa, in addition to emerging evidence of differential manifestation of diabetes and an increasing frequency of atypical diabetes subtypes.Because of these observations, precision diagnosis and treatment of diabetes are important to improve diabetes outcomes.The use of highly specific tests, such as C-peptide, islet-cell autoantibody, and genetic tests, has aided the progress of precision diabetes care in high-income countries.These advanced tests remain unavailable in most clinical settings in Africa.In this narrative review, we discuss how, in the absence of such expensive laboratory tests, structured clinical phenotyping based on routinely collected clinical characteristics can be used as a pragmatic strategy to advance precision diabetes care in low-resource settings in Africa. Introduction Burden of Diabetes in Africa According to the 11th edition of the International Diabetes Federation (IDF) Diabetes Atlas, an estimated 25 million adults are living with diabetes in Africa, representing an age-standardised prevalence of 5% (the lowest among the IDF regions) [1]. This proportion of people living with diabetes in Africa is projected to increase by 142% to 60 million by 2050 [1]. Africa has the largest percentage of undiagnosed diabetes (17.8 million adults, representing a prevalence of 72.6%) [1]. This increases the likelihood of diabetes complications and premature death. Manifestation of Diabetes in Africa Emerging data and clinical observations show marked heterogeneity in the presentation of diabetes in Africa. Generally, compared with Black populations living with diabetes in high-income countries (HICs), Black Africans develop diabetes at an early age (about 10 years earlier) and at a lower body mass index (BMI) [2]. Furthermore, at the time of diabetes diagnosis, they present with higher glycated haemoglobin (HbA1c) concentrations and pronounced pancreatic beta-cell dysfunction, as evidenced by lower markers of pancreatic beta-cell function, such as fasting or stimulated C-peptide levels and the insulinogenic index [2]. Several atypical diabetes subtypes, such as ketosis-prone diabetes (KPD), fibrocalculous pancreatic diabetes (FCPD), malnutrition-related diabetes (MRD), and type 2 diabetes in individuals without overweight or obesity (NOW-T2D), have been widely described in Black African populations [3–5]. The IDF recently recognised MRD as type 5 diabetes following a 2-day international consensus meeting of global diabetes experts in Vellore, India, in January 2025, which discussed atypical diabetes in low-BMI individuals in depth [6]. These four diabetes subtypes exhibit distinct phenotypic characteristics that do not fit the traditional classification of type 1 diabetes (T1D) and type 2 diabetes (T2D). Recently, a novel non-autoimmune, insulin-deficient diabetes subtype has been described in children and young adults of Black African ancestry with previously presumed T1D [7]. This novel diabetes subtype is phenotypically distinct from classical autoimmune T1D. It is characterised by a negative islet-autoantibody status, a lower median T1D genetic risk score (GRS), severe insulin deficiency (though higher random C-peptide concentrations than in T1D, but lower than in classical T2D), equal male and female sex distribution, and a high frequency of parental history of diabetes [7]. Monogenic diabetes, an atypical form of diabetes caused by a single-gene mutation, can be mistaken for both T1D and T2D [8, 9]. Although robust observational studies on monogenic diabetes in Black Africans are scarce, data from HICs reveal that the condition represents about 2.5–6.5% of all diabetes in both paediatric and adult populations [9, 10]. A diagnosis of monogenic diabetes requires molecular testing using Sanger or next-generation sequencing techniques, which are largely absent in most clinical settings in Africa [11]. There is limited literature on the burden of different monogenic diabetes forms in young and adult Africans with a previous diagnosis of T1D, with most evidence obtained from case reports and one cross-sectional study [12–14]. This heterogeneity in diabetes manifestation among Black African populations highlights the clinical challenge that healthcare professionals in Africa face in accurately diagnosing and managing the various diabetes subtypes. Despite the diverse diabetes subtypes observed in Africa, anecdotal evidence suggests that healthcare professionals in most clinical settings continue to use the traditional one-size-fits-all diabetes care model to treat diabetes. This generalised treatment approach does not consider differences in an individual’s genetic makeup, sex, or external environmental factors, resulting in the selection of inappropriate diabetes treatment and increasing the risk of adverse events and diabetes complications. Precision Diabetes Care: Current Perspectives Precision medicine is a contemporary, evidence-based medicine that considers individual variability in genetic makeup, the external environment, and lifestyle to provide personalised diagnosis, treatment, prediction, monitoring, and prevention of disease [15, 16]. This approach helps reduce the chances of error and greatly improves treatment outcomes. To efficiently address the heterogeneity of diabetes, precision diabetes care accounts for inter-individual differences in genetics, lifestyle, and environment. It advocates an individualised, holistic approach to managing diabetes rather than the traditional, step-by-step, one-size-fits-all approach [15, 16]. The components of precision diabetes care include precision diagnosis, precision treatment, precision therapeutics, precision monitoring, and precision prognosis [15, 16]. In HICs, the precise diagnosis of diabetes is primarily supported by genomics data (T1D and T2D genetic risk scores), C-peptide and islet autoantibody testing, proteomics and metabolomics, advanced radiological imaging modalities, data-driven clustering approaches, and artificial intelligence-driven diagnostics [15, 16]. These facilitate accurate identification and subclassification of diverse diabetes subtypes, as well as targeted treatment. Rationale for Improving Precision Diabetes Care in Africa Africa is a continent with 54 universally recognised independent countries. It is also home to an estimated 3000 distinct ethnic groups, speaking over 2000 languages [17]. As the cradle of mankind, Africa has the greatest genetic heterogeneity globally [18, 19]. This high genetic diversity in the most underrepresented population in genetic studies globally has the potential to enable the discovery of novel genetic variants and improve phenotyping of atypical diabetes subtypes. In addition to Africa’s vast genetic heterogeneity, external environmental factors, such as civil strife, spells of famine and drought, childhood and adulthood undernutrition, tropical infections, such as malaria, and infectious diseases, such as HIV, tuberculosis, and human herpes virus type 8, play an essential role in the differential manifestation of diabetes in the diverse African populations [20–23]. This close interplay between genetics and environmental factors, and their influence on epigenetics, may partly explain the increasing prevalence of atypical diabetes subtypes that do not fit the traditional T1D and T2D classifications [24]. As a result, accurately diagnosing and treating these diabetes subtypes is critical for achieving optimal treatment responses and reducing the likelihood of adverse events and diabetes complications. This can be achieved by implementing a simple, inexpensive, and pragmatic approach to advance precision diabetes diagnosis and treatment within the healthcare system. Anecdotal evidence suggests that most clinical settings in Africa lack advanced diagnostic tests for diabetes, such as genetic risk scores, C-peptide and islet autoantibody tests, advanced radiological imaging modalities, and digital health technologies to improve diagnostic precision and treatment. This adversely affects diabetes treatment outcomes. In this narrative review, we discuss how structured clinical phenotyping, based on simple, routinely collected, and readily available clinical characteristics, such as age at diagnosis, body mass index, and clinical presentation, can be used in low-resource clinical settings in Africa as a foundation for advancing practical precision diabetes diagnosis and treatment. This approach may help address the clinical challenge posed by a lack of advanced diagnostic tests required for subclassifying diabetes subtypes, resulting in more accurate treatment of these diabetes subtypes. Methods Literature Search Strategy We searched PubMed and Google Scholar for literature on precision diagnosis and treatment of diabetes, regardless of subtype, published between 2000 and January 2026. The following search terms and keywords were used: “precision medicine” OR “personalised medicine” OR “precision diabetes care” OR “Individualised diabetes care” OR “precision diabetes management” OR “precision diabetes treatment” OR “precision diabetes diagnosis”. Study Selection Criteria The preliminary screening of titles and abstracts to identify potentially eligible articles was done independently by one reviewer (DK). The inclusion criteria of studies were cross-sectional studies, retrospective studies, prospective cohort studies, randomised clinical trials (RCTs), case reports, case series, review articles, editorials, systematic reviews and meta-analyses, published in the English language, and reporting any information on how any of these simple and routinely collected clinical characteristics can be used to guide precision diagnosis and treatment of different diabetes subtypes: age at diagnosis, body mass index (BMI), clinical presentation and signs, coexisting medical conditions, and clinical history of diabetes in the family. In addition to the search of the databases above, we also hand-checked the references of the articles whose full texts were retrieved for any additional studies. We excluded studies published in languages other than English and studies whose full texts could not be retrieved. Ethics Statement This article is based on previously conducted studies and does not include any new studies involving human participants or animals conducted by any of the authors. Results The literature search returned a total of 441 articles. From these, 105 duplicates were removed. We reviewed the titles and abstracts of the remaining 336 articles, and 74 articles were identified for full-text retrieval. A total of 59 articles were excluded, leaving 15 articles. After reviewing the references to the articles whose full texts had been retrieved, an additional 23 articles were identified, bringing the total number of articles included in the narrative review to 38. Of these, 19 articles were review articles [4, 5, 9–11, 25–38], 8 articles were prospective cohort studies [39–46], 6 articles were cross-sectional studies [47–52], 3 articles were RCTs [53–55], one article was an editorial [56], and one article was a systematic review and meta-analysis [57] (Supplementary Material – PRISMA flow diagram). Discussion To improve the precision diagnosis and treatment of diabetes in Africa, we propose a structured clinical phenotyping, simplified as the ABC approach. This approach encompasses five simple and routinely collected clinical characteristics: A—Age at diagnosis; B—Body mass index (BMI); C—Clinical presentation and signs; Coexisting medical conditions; and Clinical history of diabetes in the family. We discuss in detail in the next section how these clinical characteristics can guide healthcare professionals in low-resource settings in Africa to accurately diagnose and treat commonly encountered diabetes subtypes in clinical practice (summarised in Fig. 1, Tables 1 and 2).Fig. 1The ABC approach to guide the precision diagnosis of diabetesTable 1The ABC approach to guide the precision diagnosis of diabetesClinical characteristicsDiabetes subtypesT1DT2DMonogenic diabetesLADAKPDFCPDMRDA—Age at diagnosis6 months to < 30 years≥ 30 years< 6 months (NDM)6 months to < 30 years (MODY)≥ 30 years≥ 30 years< 30 years< 30 yearsB—BMI at diagnosis< 25 kg/m2Can develop at any BMI level18.5 to < 25 kg/m2≥ 25 kg/m2≥ 25 kg/m2 (in majority)< 18 kg/m2 or18.5 to < 25 kg/m2 (in a minority)< 18 kg/m2C—Clinical presentation Clinical presentation in DKACommonCommonly seen with T2D-NOWUncommon, except with NDMCommonCommonUncommonUncommon Clinical sign of acanthosis nigricansAbsentPresent with T2D-ORAbsentMay be presentPresentAbsentAbsent Clinical presentation with signs of malnutrition or malabsorptionAbsentAbsentAbsentAbsentAbsentPresentPresent Coexisting autoimmune conditions or family history of autoimmune conditionsPresentAbsentAbsentPresentAbsentAbsentAbsent Coexisting ASCVD and/or OSAAbsentPresentAbsentMay be presentPresentAbsentAbsent Coexisting congenital renal, urinary tract, & uterine abnormalitiesAbsentAbsentPresent in cases of MODY-5AbsentAbsentAbsentAbsent Clinical history of diabetes in first-degree relativesUncommonCommonCommonUncommonCommonUncommonUncommonASCVD atherosclerotic cardiovascular diseases, BMI body mass index, DKA diabetic ketoacidosis, FCPD fibrocalculous pancreatic diabetes, KPD ketosis-prone diabetes, LADA latent autoimmune diabetes of adults, MODY maturity onset diabetes of the young, MRD malnutrition-related diabetes, NDM neonatal diabetes mellitus, OSA obstructive sleep apnea, T1D type 1 diabetes, T2D type 2 diabetes, T2D-NOW type 2 diabetes in individuals without overweight or obesity, T2D-OR obesity or overweight-related type 2 diabetesTable 2Age at diagnosis and body mass index as guides to the precision treatment of diabetesClinical characteristicsDiabetes subtype and accurate treatmentA—Age at diagnosis < 6 monthsConfirmed neonatal diabetes mellitus Initial insulin and fluid therapy to correct DKA episode Low-dose SU therapy for long-term glycaemic management 6 months to < 30 yearsConfirmed MODY Low-dose SU therapy for long-term glycaemic management (for the commonly observed MODY subtypes)Confirmed T1D Basal-bolus insulin therapy (three pre-meal short- or rapid-acting insulin doses and one bedtime long-acting insulin dose)B—Body mass index in individuals with T2D BMI < 25 kg/m2(non-obese phenotype)Good response to DPP-IV inhibitor and SU therapyPoor response to GLP-1 receptor agonist therapyAn indicator of insulin deficiency, hence an indication for early insulin initiation in some individuals (especially those with low BMI < 18 kg/m2) BMI ≥ 30 kg/m2(obese phenotype)Good response to TZD therapyPoor response to DPP-IV inhibitors in the presence of coexisting hypertriglyceridaemiaA compelling indication for initiation of SGLT2i and GLP-1 receptor agonists as first-line glucose-lowering agents in combination with metforminBMI body mass index, DKA diabetic ketoacidosis, DPP-IV dipeptidyl peptidase IV, GLP-1 glucagon-like peptide 1, MODY maturity onset diabetes of the young, SGLT2i sodium-glucose co-transporter 2 inhibitors, SU sulfonylureas, T1D type 1 diabetes, T2D type 2 diabetes, TZD thiazolidinediones The ABC approach to guide the precision diagnosis of diabetes Table 1The ABC approach to guide the precision diagnosis of diabetesClinical characteristicsDiabetes subtypesT1DT2DMonogenic diabetesLADAKPDFCPDMRDA—Age at diagnosis6 months to < 30 years≥ 30 years< 6 months (NDM)6 months to < 30 years (MODY)≥ 30 years≥ 30 years< 30 years< 30 yearsB—BMI at diagnosis< 25 kg/m2Can develop at any BMI level18.5 to < 25 kg/m2≥ 25 kg/m2≥ 25 kg/m2 (in majority)< 18 kg/m2 or18.5 to < 25 kg/m2 (in a minority)< 18 kg/m2C—Clinical presentation Clinical presentation in DKACommonCommonly seen with T2D-NOWUncommon, except with NDMCommonCommonUncommonUncommon Clinical sign of acanthosis nigricansAbsentPresent with T2D-ORAbsentMay be presentPresentAbsentAbsent Clinical presentation with signs of malnutrition or malabsorptionAbsentAbsentAbsentAbsentAbsentPresentPresent Coexisting autoimmune conditions or family history of autoimmune conditionsPresentAbsentAbsentPresentAbsentAbsentAbsent Coexisting ASCVD and/or OSAAbsentPresentAbsentMay be presentPresentAbsentAbsent Coexisting congenital renal, urinary tract, & uterine abnormalitiesAbsentAbsentPresent in cases of MODY-5AbsentAbsentAbsentAbsent Clinical history of diabetes in first-degree relativesUncommonCommonCommonUncommonCommonUncommonUncommonASCVD atherosclerotic cardiovascular diseases, BMI body mass index, DKA diabetic ketoacidosis, FCPD fibrocalculous pancreatic diabetes, KPD ketosis-prone diabetes, LADA latent autoimmune diabetes of adults, MODY maturity onset diabetes of the young, MRD malnutrition-related diabetes, NDM neonatal diabetes mellitus, OSA obstructive sleep apnea, T1D type 1 diabetes, T2D type 2 diabetes, T2D-NOW type 2 diabetes in individuals without overweight or obesity, T2D-OR obesity or overweight-related type 2 diabetes The ABC approach to guide the precision diagnosis of diabetes < 6 months (NDM) 6 months to < 30 years (MODY) < 18 kg/m2 or 18.5 to < 25 kg/m2 (in a minority) ASCVD atherosclerotic cardiovascular diseases, BMI body mass index, DKA diabetic ketoacidosis, FCPD fibrocalculous pancreatic diabetes, KPD ketosis-prone diabetes, LADA latent autoimmune diabetes of adults, MODY maturity onset diabetes of the young, MRD malnutrition-related diabetes, NDM neonatal diabetes mellitus, OSA obstructive sleep apnea, T1D type 1 diabetes, T2D type 2 diabetes, T2D-NOW type 2 diabetes in individuals without overweight or obesity, T2D-OR obesity or overweight-related type 2 diabetes Table 2Age at diagnosis and body mass index as guides to the precision treatment of diabetesClinical characteristicsDiabetes subtype and accurate treatmentA—Age at diagnosis < 6 monthsConfirmed neonatal diabetes mellitus Initial insulin and fluid therapy to correct DKA episode Low-dose SU therapy for long-term glycaemic management 6 months to < 30 yearsConfirmed MODY Low-dose SU therapy for long-term glycaemic management (for the commonly observed MODY subtypes)Confirmed T1D Basal-bolus insulin therapy (three pre-meal short- or rapid-acting insulin doses and one bedtime long-acting insulin dose)B—Body mass index in individuals with T2D BMI < 25 kg/m2(non-obese phenotype)Good response to DPP-IV inhibitor and SU therapyPoor response to GLP-1 receptor agonist therapyAn indicator of insulin deficiency, hence an indication for early insulin initiation in some individuals (especially those with low BMI < 18 kg/m2) BMI ≥ 30 kg/m2(obese phenotype)Good response to TZD therapyPoor response to DPP-IV inhibitors in the presence of coexisting hypertriglyceridaemiaA compelling indication for initiation of SGLT2i and GLP-1 receptor agonists as first-line glucose-lowering agents in combination with metforminBMI body mass index, DKA diabetic ketoacidosis, DPP-IV dipeptidyl peptidase IV, GLP-1 glucagon-like peptide 1, MODY maturity onset diabetes of the young, SGLT2i sodium-glucose co-transporter 2 inhibitors, SU sulfonylureas, T1D type 1 diabetes, T2D type 2 diabetes, TZD thiazolidinediones Age at diagnosis and body mass index as guides to the precision treatment of diabetes Confirmed neonatal diabetes mellitus Initial insulin and fluid therapy to correct DKA episode Low-dose SU therapy for long-term glycaemic management Confirmed MODY Low-dose SU therapy for long-term glycaemic management (for the commonly observed MODY subtypes) Confirmed T1D Basal-bolus insulin therapy (three pre-meal short- or rapid-acting insulin doses and one bedtime long-acting insulin dose) BMI < 25 kg/m2 (non-obese phenotype) Good response to DPP-IV inhibitor and SU therapy Poor response to GLP-1 receptor agonist therapy An indicator of insulin deficiency, hence an indication for early insulin initiation in some individuals (especially those with low BMI < 18 kg/m2) BMI ≥ 30 kg/m2 (obese phenotype) Good response to TZD therapy Poor response to DPP-IV inhibitors in the presence of coexisting hypertriglyceridaemia A compelling indication for initiation of SGLT2i and GLP-1 receptor agonists as first-line glucose-lowering agents in combination with metformin BMI body mass index, DKA diabetic ketoacidosis, DPP-IV dipeptidyl peptidase IV, GLP-1 glucagon-like peptide 1, MODY maturity onset diabetes of the young, SGLT2i sodium-glucose co-transporter 2 inhibitors, SU sulfonylureas, T1D type 1 diabetes, T2D type 2 diabetes, TZD thiazolidinediones For this narrative article, the diabetes subtypes of interest are T1D, T2D, monogenic diabetes (maturity onset diabetes of the young [MODY] and neonatal diabetes mellitus [NDM]), KPD, latent autoimmune diabetes of adults (LADA), FCPD, and MRD. Using the ABC Approach to Guide Precision Diagnosis of Diabetes Age at diagnosis as a guide Acute hyperglycaemia and diabetic ketoacidosis (DKA) in infants under 6 months are diagnosed as NDM [11]. It is a form of monogenic diabetes that follows a variable course and severity, and is subclassified as permanent NDM or transient NDM. Both subtypes of NDM arise from activating heterozygous mutations in either gene encoding the subunits of the pancreatic beta-cell KATP channel (KCNJ11 or ABCC8) [11, 27]. New-onset diabetes in individuals aged > 6 months to < 30 years is caused by either classical autoimmune T1D [7, 58], MODY [11, 27, 28], FCPD [4, 29, 30], or MRD [4, 31, 32]. These four diabetes subtypes are associated with early onset of diabetes. In adults aged ≥ 30 years, new-onset diabetes may arise from T2D (early- and late-onset) [2], LADA (also called slowly evolving autoimmune diabetes of adults) [33], or KPD [4, 40]. Compared to white populations in HICs, Black Africans aged ≥ 30 years have a low prevalence of classical autoimmune T1D [47, 49]. Body Mass Index as a Guide BMI, as an indirect and simple marker of adiposity, is a key clinical characteristic that can guide the precision diagnosis of diabetes. In the absence of BMI cutoffs specific to Africans, the traditional World Health Organization (WHO)-recommended cutoffs of < 18.5 kg/m2, ≥ 18.5–24.9 kg/m2, and ≥ 25 kg/m2 are used to classify individuals as underweight, normal weight, and overweight or obese, respectively [59]. New-onset diabetes in underweight individuals may arise due to T1D [5, 8], MRD [4, 11, 12], FCPD [4, 9, 10], or T2D (with severe weight loss preceding the diagnosis) [60]. In individuals with new-onset diabetes and a normal BMI of ≥ 18.5–24.9 kg/m2, consider T1D [7, 58], MODY [11, 27], FCPD (in a smaller proportion of patients) [30], and T2D (without coexisting insulin resistance) [61] as the differential diagnoses. In individuals with new-onset diabetes and overweight or obesity, T2D (with prevalent insulin resistance) [2], LADA [33], and KPD [4, 34, 40, 56] are the possible differential diagnoses. Clinical Presentation and Signs as a Guide In young and adult individuals aged > 6 months presenting with DKA (diagnosed in low-resource settings based on the presence of severe hyperglycaemia (> 15 mmol/l), the presence of ketonuria, and classic symptoms and signs, such as vomiting, abdominal pain, and Kussmaul’s breathing), T1D [5, 8], T2D-NOW [60], and KPD [4, 34, 40, 56] are the probable differential diagnoses. Absence of ketoacidosis in the setting of severe hyperglycaemia and low BMI of < 18.5 kg/m2 is indicative of either FCPD [4, 29, 30] or MRD [31, 32]. The presence of acanthosis nigricans on physical examination (skin hyperpigmentation often at the back of the neck and armpits) suggests increased insulin resistance and is common in T2D associated with overweight or obesity [50, 51] and KPD [34, 41, 56]. In individuals presenting with new-onset diabetes, BMI < 18 kg/m2, and clinical signs of malnutrition and/or malabsorption syndrome, such as skin and hair changes, parotid gland enlargement, palmar hyperpigmentation, atrophic glossitis, angular cheilitis, pallor, and sarcopenia, consider either FCPD [4, 29, 30] or MRD [4, 11, 12] as the likely clinical diagnosis. Coexisting Medical Conditions The presence of coexisting autoimmune conditions, such as Graves’ disease, autoimmune thyroiditis, pernicious anaemia, and Addison’s disease, in an individual with new-onset diabetes, suggests classical autoimmune T1D or LADA as the underlying differential diagnosis [35]. These autoimmune conditions form part of the autoimmune polyglandular syndromes [35]. Furthermore, a family history of autoimmune conditions may raise the possibility that the individual with new-onset diabetes has T1D [62]. The presence of comorbidities, such as obstructive sleep apnea and atherosclerotic cardiovascular diseases (ASCVD), such as coronary artery disease, stroke, and peripheral arterial disease, is more prevalent in individuals with overweight or obesity-related T2D [36, 42] and KPD [52]. They are infrequent in individuals with recently diagnosed FCPD [4, 29, 30], MRD [4, 11, 12], and MODY [11, 27]. In the presence of congenital renal, urinary tract, and uterine abnormalities in young individuals with newly diagnosed diabetes, such as renal cysts, renal hypoplasia, bicornuate or unicornuate uterus, in a young individual (< 30 years) presenting with new-onset diabetes, consider hepatocyte nuclear factor-1-beta (HNF1B) MODY or MODY-5 as the definitive diagnosis [37]. This rare form of monogenic diabetes is caused by mutations or deletions of the HNF1B gene [37]. Clinical History of Diabetes in the Family as a Guide A strong family history of diabetes, especially among first-degree relatives, in an individual with new-onset diabetes, suggests either T2D, KPD [34, 38, 56], or MODY [11, 27] as the probable diagnosis. Specifically, MODY is characterised by a history of early-onset diabetes in multiple generations when a family pedigree chart is drawn, given its autosomal dominant pattern of inheritance [11, 27]. The T1D, FCPD, and MRD subtypes are not significantly associated with a family history of diabetes. Additional Clinical Characteristics to Complement the ABC Approach in the Precision Diagnosis of Specific Diabetes Subtypes Because T1D and MODY occur in the same age group, the severity of hyperglycaemia at diagnosis is important to distinguish the two diabetes subtypes. Compared with T1D, individuals with MODY often present with mild-to-moderate hyperglycaemia (HbA1c < 7.5% or 58 mmol/mol) at diagnosis [28]. Individuals with FCPD and MRD share many clinical characteristics, including a history and clinical signs of undernutrition, socioeconomic underprivileged status, residence in rural areas, and severe hyperglycaemia at diagnosis [30, 31]. Radiological imaging of the pancreas, either with abdominal X-ray or ultrasound (both readily available in most low-resource settings in Africa), is critical for distinguishing between these two diabetes subtypes [30, 31]. Individuals with FCPD have intra-pancreatic calcifications and large ductal dilatation (primarily involving the main pancreatic duct) [30]. In contrast, those with MRD have no evidence of intra-pancreatic calcifications, ductal dilatation, or hypoplasia [31]. Using the ABC Approach in the Precision Treatment of Diabetes Age at Diagnosis as a Guide Age at diagnosis, when combined with BMI and specific clinical characteristics, is an important clinical characteristic to guide the precision treatment of diabetes, especially in cases of suspected monogenic diabetes, T1D, and T2D. In infants aged less than 6 months, transient or permanent NDM is the definitive diagnosis for new-onset hyperglycaemia [11, 27, 28]. Babies diagnosed with these forms of monogenic diabetes initially require insulin and fluid replacement therapy to correct the DKA [11, 27, 28]. When stable, they can be switched to long-term oral sulfonylurea treatment. This treatment option has been shown to maintain long-term (> 10 years) optimal glycaemic control in children diagnosed with NDM [11, 27, 28]. For children and young adults diagnosed with diabetes between 6 months and 30 years with a BMI < 25 kg/m2, T1D [7, 58] or MODY [11, 27, 28] is the most common diagnosis. Because classical autoimmune T1D is associated with absolute insulin deficiency, individuals diagnosed with T1D require early insulin initiation [63]. A basal-bolus approach to insulin administration (three pre-meal short- or rapid-acting insulin doses and one bedtime long-acting insulin dose) is preferred to mimic the normal physiological pattern of insulin secretion [63]. For children or young adults diagnosed with MODY, accurate treatment depends on the MODY type. Most individuals with a genetic diagnosis of MODY (> 80%) will have either MODY arising due to glucose kinase mutation (GCK-MODY or MODY-2), hepatocyte nuclear factor-1-alpha mutation (HNF1A-MODY or MODY-3), or hepatocyte nuclear factor-4-alpha mutation (HNF4A-MODY or MODY-1) [11, 27, 28]. GCK-MODY is associated with non-progressive, mild, and asymptomatic hyperglycaemia and does not require treatment, except possibly during pregnancy [11, 27, 28]. Both HNF1A-MODY and HNF4A-MODY require low-dose sulfonylurea treatment to achieve sustained optimal glycaemic control [11, 27, 28]. The optimal treatment of HNF1B-MODY or MODY-5 (a rare type of MODY) is not well established. However, insulin therapy can be used in individuals diagnosed with this rare type of MODY [11, 27, 28]. Body Mass Index as a Guide BMI plays an important role mainly in guiding the precision treatment of T2D. It is an important and widely available anthropometric measure for categorising people with type 2 diabetes into two clinically distinct subgroups: non-obese and overweight or obese. These two subgroups differ in their underlying pathophysiologic mechanisms, cardiometabolic risk profiles, and therapeutic responses to glucose-lowering therapies [55, 60]. The non-obese T2D subgroup is characterised by a BMI < 25 kg/m2, lower markers of metabolic syndrome, and pancreatic beta-cell dysfunction is the predominant underlying pathophysiologic defect [60, 61, 64]. Conversely, the obese T2D subgroup is marked by a high prevalence of metabolic syndrome, and insulin resistance is predominant [60]. Emerging evidence from studies conducted in Asian and white populations suggests that non-obese and obese T2D subgroups respond differently to several glucose-lowering therapies [43–46, 53–55, 57]. Studies have shown that individuals with T2D and obesity have a greater glycaemic reduction on thiazolidinediones (TZD) therapy, such as pioglitazone [55]. However, this obese subgroup (BMI ≥ 30 kg/m2) with high circulating triglyceride concentrations shows a poor therapeutic response to dipeptidyl peptidase IV inhibitors (DPP-IVi), such as sitagliptin [43, 55]. This greater response to DPP-IVi has also been observed in Asian populations with T2D [53, 57]. Asians, like most Black African populations, develop T2D at a young age, have a lower BMI, and less insulin resistance [39, 65]. Additionally, in the A Diabetes Outcome Progression (ADOPT) trial, men without obesity showed a greater long-term reduction in HbA1c with sulfonylurea therapy than with TZD therapy [54]. However, women with obesity showed a better long-term response when treated with TZD compared with sulfonylureas [54]. In another study by Jones et al., lower markers of pancreatic beta-cell function, such as lower fasting C-peptide concentrations and severe insulin deficiency (fasting C-peptide ≤ 0.25 nmol/l), were associated with a reduced response to glucagon-like peptide 1 receptor agonists [44]. A low or normal BMI of < 25 kg/m2, which is common in Black African populations, is an important clinical predictor of insulin deficiency at the time of diagnosis of T2D [48]. In the absence of C-peptide testing, which is common in most low-income settings in Africa, a low or normal BMI can be used to guide early initiation of insulin therapy in individuals with new-onset diabetes. Despite being primarily used as a potent weight loss drug in people with T2D and obesity, some studies in Asian and white populations have found that metformin is an equally effective glucose-lowering drug in both people with or without obesity and T2D [45, 46]. Studies investigating the effectiveness and safety of glucose-lowering therapies in Black African populations remain limited. This raises the clinical question of whether the findings from the studies conducted in white and Asian populations discussed above can be directly applied to Black African populations living with T2D. Robustly designed RCTs are urgently needed in this population to address this evidence gap and inform clinical practice in Africa. In clinical practice, BMI helps inform broader cardiometabolic risk assessment. Because BMI is strongly associated with several cardiovascular risk factors, such as hypertension, atherogenic dyslipidaemia, and ASCVD, such as coronary heart disease and peripheral arterial disease, the presence of an overweight or obese phenotype is a compelling indication for the use of SGLT2 inhibitors and glucagon-like peptide 1 receptor agonists as first-line glucose-lowering therapies in combination with metformin [25, 26]. This is due to their proven cardiorenal protective effects and weight-loss benefits [66]. Alongside guiding accurate glucose-lowering therapy choices in individuals with T2D, BMI can also be used to predict specific adverse events associated with certain glucose-lowering treatments. For instance, euglycaemic DKA, an adverse event related to SGLT2 inhibitor therapy, occurs more frequently in individuals without obesity or overweight [50]. This allows healthcare professionals to advise patients who are not obese on these medications to be vigilant for signs and symptoms of DKA, such as sudden unexplained abdominal pain, vomiting, and diarrhoea despite normal blood glucose levels. Improving Precision Diabetes Care in Africa: Proposing Region-Specific Strategies In the absence of precision tests for diabetes diagnosis and treatment, such as C-peptide and islet autoantibody tests, genomics, and proteomics, structured clinical phenotyping based on basic clinical characteristics provides a solid foundation for improving precision diabetes care in low-resource African settings. As a result, electronic medical record systems in clinical settings must be improved to capture all of the necessary basic clinical data. Integrating structured clinical phenotyping into diabetes care across all levels of healthcare is crucial for early and precise diagnosis and treatment of Africa’s diverse diabetes subtypes. This approach aims to enhance treatment outcomes and lower the risk of diabetic complications. Following implementation, well-conducted large-scale studies are required to assess the effectiveness, acceptability, and translatability of this structured clinical phenotyping approach (the ABC framework) in improving the precision diagnosis and treatment of diabetes in our local clinical settings. There is a need to enhance the technical capacity of local laboratories in Africa to perform essential diagnostic tests for diabetes, such as C-peptide and islet autoantibody tests, to accurately diagnose the different diabetes subtypes and reduce the chances of using inappropriate and costly diabetes treatments. Over the past years, these tests have become more affordable and available in most parts of the world. To further the role of genomics in precision diabetes care across the continent, more geneticists need to be trained, and African centres of excellence in genetic research must be established to conduct comprehensive genetic research and testing to support clinical practice. Blood samples from different African countries can be shipped to these genetic centres of excellence for specific genetic tests to be performed and ensure accurate diagnosis of non-syndromic and syndromic atypical diabetes subtypes. Because most diabetes precision treatment studies have been conducted in Asian and white populations, robust clinical trials investigating the efficacy and safety of glucose-lowering agents in diverse African populations are urgently needed to provide practical solutions specific to the African setting. Furthermore, because atypical diabetes subtypes are highly prevalent in Black African populations, more robustly conducted studies are needed to comprehensively characterise the underlying causes and natural history of these atypical diabetes subtypes. This presents a unique opportunity to discover novel underlying pathophysiological mechanisms, diagnostic tests, and accurate treatment approaches. Conclusion Precision diabetes care is a growing paradigm in diabetes care that focuses on personalised diagnosis, treatment, prevention, and monitoring of diabetes based on an individual’s genetic makeup, lifestyle, and external environment. In the absence of specialised precision diagnostic tests for diabetes, structured clinical phenotyping based on basic, routinely collected clinical characteristics offers a promising, cost-effective, and easily scalable approach to improve precision diagnosis and treatment in low-income settings in Africa. It lays a solid foundation for advanced future precision diabetes care strategies in Africa. Supplementary Information Below is the link to the electronic supplementary material.Supplementary file1 (PDF 230 KB) Supplementary file1 (PDF 230 KB)