Monogenic diabetes (MD) is a rare form of diabetes caused by one or more defects in a single gene (1, 2). Up to date, 43 genes have been linked to MD pathogenesis. Despite the fact that MD accounts for up to 5% of all diabetes cases, it is often misdiagnosed either due to phenotypical overlap with other types of diabetes or absence of clinical suspicion (1–3).

MD is a heterogeneous group of disorders, the most common being maturity-onset diabetes of the young (MODY) and neonatal diabetes. MODY has well-established diagnostic guidelines in children and young adults, however, individuals over 25 are often overlooked despite the importance of genetic diagnosis for personalized treatment and family risk assessment (4). At least 14 genes have been confirmed to cause MODY, and eight of them (GCK, HNF1A, HNF4A, HNF1B, ABCC8, KCNJ11, 6q24, and INS) have clear clinical and therapeutic implications (2, 3).

The aim of this study was to highlight the burden of undiagnosed MD and analyze genetic sequencing results and their implications for treatment optimizations during 2017–2024.

A key strength of our study is real data from a national diabetes reference center and the inclusion of MD cases across all age groups, extending beyond the younger cohorts. We believe these findings may encourage greater attention to the genetic investigation of patients well into their 30s, 40s, and even 50s.

The frequency of GCK, HNF1A, HNF1B, and HNF4A variants were comparable in patients < 25 years and ≥25 years in our study, underscoring the importance of considering MD screening across all age groups. In pediatric cohorts, GCK-MODY typically predominates. For example, an Italian multicenter study of 172 families reported that ~80% of cases were due to GCK mutations, with relatively few HNF1A diabetes (5). Similarly, targeted sequencing in UK pediatric and young adult populations has confirmed that GCK and HNF1A together account for most MD cases (6). Studies of broader age spectrum cohorts also indicate that GCK, HNF1A, and HNF4A together account for at least 85% of MODY cases (7). Importantly, the penetrance of HNF1A is strongly age-dependent, with diabetes present in ~63% of carriers by age 25, ~94% by age 50, and ~99% by age 75 (8). Furthermore, population-based analyses suggest that HNF1A-MODY represents up to 70% of European cases, with the majority of affected individuals manifesting before the age of 25 (9). Taken together, these findings are consistent with our observation reinforcing the importance of offering MD genetic testing beyond pediatric and young adults' settings.

MD is frequently misdiagnosed, with many patients initially classified as having type 1 or type 2 diabetes due to overlapping clinical features and limitations of standard diagnostic algorithms, especially in patients over 25 years of age (1, 2). It is estimated that up to 80% of MODY cases are not correctly identified, resulting in significant diagnostic delays that often exceed a decade from the onset of hyperglycemia to the confirmation of a molecular diagnosis (10). Studies demonstrate that the time between the initial manifestation of hyperglycemia and genetic confirmation is prolonged, leading to extended periods of inappropriate management and missed opportunities for pathogenetically targeted therapy (1, 11–13). In our cohort, although the diagnostic interval was shorter than previously reported, with a mean duration of approximately up to 5 years from the initial diabetes diagnosis to genetic confirmation, substantial delays were still observed in some cases.

Delayed diagnosis in MD, particularly in HNF1A and HNF4A, compromises the efficacy of targeted treatments. Prolonged exposure to hyperglycemia accelerates β-cell failure through glucotoxicity, resulting in reduced endogenous insulin secretion (14, 15). Patients with HNF1A diabetes and longstanding diabetes exceeding a decade rarely achieve optimal glycemic targets [HbA1c < 7.5% (58 mmol/mol)] after initiating sulfonylurea monotherapy, in contrast to those diagnosed earlier who demonstrate success rates above 80% (12, 14). Diminishing β-cell function over time also necessitates adjunctive therapies despite genetically appropriate treatment. Importantly, early commencement of sulfonylureas has been associated with improved metabolic outcomes, greater preservation of insulin secretory capacity, and delayed onset of diabetes-related complications (2, 15). These findings emphasize that the therapeutic window for effective sulfonylurea monotherapy narrows considerably as diagnostic delays increase, highlighting the critical importance of early genetic identification in optimizing long-term management. A similar pattern was identified in our cohort, where individuals with a longer duration of diabetes demonstrated a reduced likelihood of achieving successful transition to sulfonylurea monotherapy, though we did not find statistical significance, likely influenced by the small sample sizes, limiting the study's statistical power.

The main limitation of our study is that comprehensive genetic testing, such as WES was not systematically performed in all patients. Instead, molecular diagnosis was largely based on targeted gene analysis, which may have led to an underestimation of the prevalence of rarer or atypical forms of MD. This approach also carries the risk of missing pathogenic variants in genes not included in the targeted panel, as well as non-coding or structural variants detectable only by broader sequencing methods. Consequently, the genetic spectrum described in our cohort may not fully capture the heterogeneity of MD. Future studies employing exome or genome sequencing across entire cohorts are warranted to provide a more complete picture of the underlying genetic architecture and to refine genotype–phenotype correlations.

The other limitaion is that this study identified several novel variants, however, no experimental functional studies (e.g., in vitro assays, splicing analysis, or protein function experiments) were performed to confirm their pathogenicity. This represents a limitation, as in silico prediction tools alone have restricted accuracy. However, segregation analysis was performed in families where additional samples were available, and segregation results contributed supporting evidence for variant interpretation. Final classification was based on a combination of ACMG/ACGS/ClinGen criteria, variant type, population frequency, predicted functional effect, and the consistency between the genotype and the patient's clinical phenotype. Nevertheless, we acknowledge that future studies including functional assays will be essential to strengthen the evidence for pathogenicity, particularly for the novel variants reported here.

Traditional screening criteria for MD—based on diagnosis before the age of 25 years, a non-obese phenotype, and a strong family history—have been widely applied increasingly recognized as overly restrictive (10, 16). Recent studies show that a substantial proportion of MD cases are diagnosed beyond the classical age threshold. In a Chinese cohort, 72.2% of patients with genetically confirmed MD were diagnosed after the age of 35, and only 38.9% had a family history of diabetes, highlighting the limitations of traditional screening criteria (17). Furthermore, a 2023 systematic review by Murphy et al. found that relying on the conventional age cutoff (< 25 years with family history) would miss about 63% of actual MD cases (16). These findings underscore the necessity to broaden current diagnostic criteria, extending genetic testing to adults diagnosed after the age of 25 who present with atypical diabetes features. Such an approach would facilitate earlier detection of MD and the timely implementation of pathogenetically appropriate therapies, thereby optimizing long-term clinical outcomes.

Finally, our data emphasize that accurate differentiation of MD from other more common types of diabetes is clinically essential, as it directly influences treatment decisions, long-term outcomes, and healthcare costs. In contrast to polygenic forms of diabetes, specific MD types have well established, gene-specific strategies. Misclassification of diabetes may lead to unnecessary or inappropriate treatment, including lifelong insulin therapy, increased risk of hypoglycemia, psychological burden, and avoidable healthcare costs. In this context, genetic testing—when applied to carefully selected patients—is cost-effective, as it enables precise diagnosis, optimization of therapy, and prevention of long-term complications related to overtreatment (2, 18, 19). Furthermore, it facilitates cascade screening of family members, allows early identification of affected relatives. Therefore, timely and accurate identification of MD is a key component of precision medicine in diabetes care, improving both individual patient management and broader family-based risk assessment.

Diagnostic delays in MD remain common, contributing to extended periods of non-optimized management. Our findings underscore the clinical value of genetic testing for MD across all age groups and support its role in guiding individualized, evidence-based treatment strategies.