Familial Occurrence of Type 1 Diabetes Mellitus in Korean Children and Adolescents: A Multicenter Study

Article information

Diabetes Metab J. 2026;.dmj.2025.1149

Publication date (electronic) : 2026 March 5

doi : https://doi.org/10.4093/dmj.2025.1149

Hae Sang Lee^1,*, Hwa Young Kim^2,*, Mi Yang², Yun Jeong Kim², Hyun Wook Chae³, Kyungchul Song³, Aram Yang⁴, Hyo-Kyoung Nam⁵, Young-Jun Rhie⁶, Eungu Kang⁶, Mo Kyung Jung⁷, Yoonha Lee⁷, Sung Yoon Cho⁸, Insung Kim⁸, Minji Im⁸, Moon Bae Ahn⁹, Su Jin Park⁹, Soo Yeun Sim⁹, Yoo-Mi Kim¹⁰, Young-Lim Shin¹¹, Yong Hee Hong¹¹, Junghwan Suh¹², Sujin Kim¹², Seo Jung Kim¹², Min Hyung Cho¹, Yong Hyuk Kim¹³, Jieun Lee¹⁴, Su Jin Kim¹⁴, Jisun Park¹⁴, Eun Young Joo¹⁴, Myung Ji Yoo¹⁴, Minsun Kim¹⁵, Han Sol Kim¹⁵, Han Hyuk Lim¹⁶, Jung Eun Moon¹⁷, Kyungmi Jang¹⁷, Chan Jong Kim^,18, Jaehyun Kim^,2

¹Department of Pediatrics, Ajou University Hospital, Ajou University School of Medicine, Suwon, Korea

²Department of Pediatrics, Seoul National University Bundang Hospital, Seoul National University College of Medicine, Seongnam, Korea

³Department of Pediatrics, Gangnam Severance Hospital, Yonsei University College of Medicine, Seoul, Korea

⁴Department of Pediatrics, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul, Korea

⁵Department of Pediatrics, Korea University Guro Hospital, Korea University College of Medicine, Seoul, Korea

⁶Department of Pediatrics, Korea University Ansan Hospital, Korea University College of Medicine, Ansan, Korea

⁷Department of Pediatrics, CHA Bundang Medical Center, CHA University School of Medicine, Seongnam, Korea

⁸Department of Pediatrics, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea

⁹Department of Pediatrics, Seoul St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea

¹⁰Department of Pediatrics, Chungnam National University Sejong Hospital, Chungnam National University College of Medicine, Sejong, Korea

¹¹Department of Pediatrics, Soonchunhyang University Bucheon Hospital, Soonchunhyang University College of Medicine, Bucheon, Korea

¹²Department of Pediatrics, Severance Children’s Hospital, Yonsei University College of Medicine, Seoul, Korea

¹³Department of Pediatrics, Wonju Severance Christian Hospital, Yonsei University Wonju College of Medicine, Wonju, Korea

¹⁴Department of Pediatrics, Inha University Hospital, Inha University College of Medicine, Incheon, Korea

¹⁵Department of Pediatrics, Jeonbuk National University Hospital, Jeonbuk National University Medical School, Jeonju, Korea

¹⁶Department of Pediatrics, Chungnam National University Hospital, Chungnam National University College of Medicine, Daejeon, Korea

¹⁷Department of Pediatrics, Kyungpook National University Chilgok Hospital, School of Medicine, Kyungpook National University, Daegu, Korea

¹⁸Department of Pediatrics, Chonnam National University Children’s Hospital, Chonnam National University Medical School, Gwangju, Korea

Corresponding authors: Jaehyun Kim https://orcid.org/0000-0002-0203-7443 Department of Pediatrics, Seoul National University Bundang Hospital, Seoul National University College of Medicine, 82 Gumi-ro 173beon-gil, Bundang-gu, Seongnam 13620, Korea E-mail: jaehyun.kim@snu.ac.kr

Chan Jong Kim https://orcid.org/0000-0001-6519-2427 Department of Pediatrics, Chonnam National University Children’s Hospital, Chonnam National University Medical School, 42 Jebong-ro, Dong-gu, Gwangju 61469, Korea E-mail: cjkim@jnu.ac.kr

*The first two authors contributed equally to this work.

Received 2025 November 13; Accepted 2026 January 23.

Abstract

GRAPHICAL ABSTRACT

Highlights

INTRODUCTION

Type 1 diabetes mellitus (T1DM) is a chronic autoimmune disease characterized by pancreatic β-cell destruction, causing absolute insulin deficiency [1]. T1DM incidence has been steadily rising worldwide, particularly among children and adolescents, with the disease imposing a significant lifelong burden on affected individuals and their families [2]. Genetic predisposition is crucial for T1DM pathogenesis, with both environmental and immunologic factors contributing to disease onset [3,4].

A T1DM family history is a well-recognized risk factor, and the presence of an affected first-degree relative markedly increases the likelihood of developing T1DM compared with the general population [5-7]. Previous large-scale studies from Western countries have reported that approximately 5% to 15% of children with T1DM have a first-degree relative with the disease [6,8-12]. Moreover, familial T1DM cases often present with distinct clinical characteristics, including age of onset, glycemic control, and risk of diabetic ketoacidosis (DKA) differences, compared with sporadic cases [6,10,12,13]. Nonetheless, most of the available data are derived from Western cohorts; moreover, evidence from Asian populations, particularly from Korea, remains limited.

In Korea, although the incidence of T1DM remains relatively low, nationwide data indicate a steady rise among children under 15 years of age—from 1.36 per 100,000 during 1995–2000 to 3.19 during 2012–2014—with more recent analyses reporting 4.45 per 100,000 during 2007–2017 and a further increase to 5.28 per 100,000 in 2021 among those aged 0 to 14 years [14-18]. Considering potential ethnic and genetic differences, understanding the T1DM familial patterns in Korean children is essential for risk stratification and tailored clinical care. Therefore, we aimed to determine the frequency of first-degree relatives in Korean children with T1DM and compare the clinical presentation of familial and sporadic T1DM.

METHODS

Data collection

Demographic and clinical data at the time of T1DM diagnosis were collected by reviewing electronic medical records at each participating center. Collected variables included sex, age at diagnosis, random plasma glucose, glycosylated hemoglobin (HbA1c), presence of DKA, and autoantibody profiles at diagnosis. Autoantibody testing was performed as a part of routine clinical care at each participating center using different commercial assays and analytic platforms, and autoantibody positivity was determined based on the reference ranges and cutoff values applied at each institution. A detailed T1DM family history was ascertained for each proband, including the presence of T1DM in parents and siblings, along with the age at onset, DKA status, and metabolic parameters at diagnosis for affected relatives.

The index child was defined as the first sibling diagnosed with T1DM. Overall, 936 index cases were diagnosed with T1DM and were further categorized according to T1DM family history (Fig. 1). Those with siblings were classified into three subgroups: (1) having a parent with T1DM, (2) having one or more siblings with T1DM, or (3) sporadic cases with no T1DM in parents or siblings. Singletons were classified into two subgroups: (1) having a parent with T1DM or (2) sporadic cases.

Fig. 1.

Flow chart of grouping children based on the familial history of type 1 diabetes mellitus (T1DM).

RESULTS

Clinical and familial characteristics at diagnosis

Among 936 index children with T1DM, 605 and 331 were categorized as ‘children with siblings’ and singletons, respectively (Fig. 1). Overall, 32 (3.4%) patients had at least one affected first-degree relative, comprising 10 (1.1%) with an affected parent and 22 (2.4%) with one or more affected siblings. Among those with a parental history, paternal transmission was predominant (seven fathers, 70%), while maternal transmission was observed in two cases and both parents were affected in one case. Overall, 904 (96.6%) patients had no family history of T1DM, representing sporadic cases. Of the 936 index cases, 398 (42.5%) were male individuals; the median age at diagnosis was 10.2 years (IQR, 6.7 to 12.9). The median HbA1c was 12.6% (IQR, 11.0 to 14.0), and DKA was present in 453 (49.1%) patients (Table 1).

Table 1.

Clinical characteristics at diagnosis

Clinical characteristics of index and subsequent-affected children

Clinical characteristics at diagnosis of index and subsequent-affected children are summarized in Table 2. Subsequent-affected children demonstrated lower plasma glucose (median 300.0 mg/dL vs. 412.0 mg/dL, P=0.009) and HbA1c levels (median 10.4% vs. 12.6%, P<0.001) than did index children. DKA prevalence at presentation was significantly lower in the subsequent-affected children (13.8% vs. 49.7%, P<0.001), and DKA severity tended to be milder (P=0.001). Consistently, venous pH and serum bicarbonate levels were higher in subsequent-affected children (median 7.4 vs. 7.3, P=0.005; 22.0 mmol/L vs. 17.0 mmol/L, P=0.004, respectively), and the urine ketone levels were lower (P<0.001). No significant differences in the frequencies of diabetes-associated autoantibodies were found between the two groups. Among the 32 familial cases, 10 and 22 had an affected parent and at least one affected siblings, respectively. No significant differences in clinical characteristics were found between children with an affected parent and those with affected siblings (Supplementary Table 2).

Table 2.

Comparison of clinical characteristics between index and subsequent-affected children

Familial occurrence of T1DM

Among 605 index children with siblings, 23 subsequent-affected siblings with T1DM were identified among a total of 779 siblings, corresponding to a sibling prevalence of 3.0%. When parents were included in the analysis of first-degree relatives, 34 affected relatives with T1DM (23 subsequent-affected siblings plus 11 affected parents) were identified among 2,651 total first-degree relatives of 936 index children, yielding a prevalence of 1.3%. In regard to the analysis of seven twin pairs, wherein one twin in each pair was the index children, three cotwins were diagnosed with T1DM, resulting in a twin-pair prevalence of 42.9% (Table 3).

Table 3.

Prevalence of type 1 diabetes mellitus according to familial relationship to index children

DISCUSSION

In this large multicenter cohort study of 936 Korean children and adolescents with T1DM, 32 (3.4%) had a T1DM family history. Among familial cases, 1.3% and 3.0% had a first-degree relative and an affected sibling, respectively, while the concordance rate among twin pairs was 42.9%. Compared with index children, subsequent-affected children exhibited lower HbA1c and serum glucose levels at diagnosis, as well as a markedly lower DKA frequency. Moreover, among those with parental history, the majority represented paternal rather than maternal transmission.

Herein, the proportion of familial history was lower than that reported in Western cohorts, where 5% to 15% of children with T1DM had an affected first-degree relative [6,8-13,19-21]. In a Finnish nationwide cohort, 10.4% of newly diagnosed children had familial T1DM, with paternal transmission (5.1%) more common than maternal (2.8%) or sibling (1.9%) history [10]. Karges et al. [9] reported that 6.6% of pediatric patients with T1DM demonstrated a first-degree relative with the disease in a large population-based German registry. In Asia, reports on the prevalence of familial T1DM are lacking. However, a nationwide Japanese study found a prevalence of 5.5% (2.2% among parents and 3.3% among siblings), which is comparable to our findings [22]. Our results suggest that T1DM familial aggregation is less common in Korean children, consistent with the lower T1DM incidence in Asian populations compared with those in Western countries. Specifically, the sibling prevalence of 3.0% was higher than that for parental history of T1DM, which is consistent with findings from the Japanese study. These observations possibly reflect the multifactorial nature of T1DM, with population-specific differences in genetic susceptibility—particularly the distribution of high-risk human leukocyte antigen (HLA) class II haplotypes—as well as environmental and lifestyle modifiers contributing to variability in familial clustering across regions. High-risk HLA genotypes that strongly predispose to familial aggregation in Western populations are less prevalent in East Asian populations, including Koreans, which might partly explain the lower frequency of familial T1DM observed in our cohort [7,23,24]. In this context, our findings underscore the importance of population-specific epidemiologic data to guide genetic counseling, risk prediction, and early detection strategies in Asian children.

Notably, approximately 96.6% of patients in this cohort developed T1DM without any family history, highlighting that the vast majority of cases arise sporadically. However, given that the sibling prevalence of 3.0% corresponds to an estimated 11.7-fold higher risk compared with the general population prevalence of 0.26% among Korean children under 15 years in 2021, this magnitude of familial aggregation is broadly consistent with Western data reporting a 10–15-fold increased sibling risk [7,18,25]. These findings suggest that, although familial T1DM is uncommon in Korea, targeted screening or active surveillance among siblings of affected probands might be warranted for earlier detection and prevention of acute metabolic decompensation.

Importantly, subsequent-affected children exhibited milder metabolic decompensation at diagnosis compared with index children. Specifically, HbA1c and serum glucose levels were significantly lower, and the prevalence of DKA at diagnosis was reduced in the subsequent-affected children, thereby corroborating earlier reports demonstrating that children with a family history of T1DM tend to exhibit less severe metabolic decompensation at diagnosis. Prior studies have consistently attributed this pattern to heightened awareness within affected families, earlier recognition of hyperglycemic symptoms, and a lower threshold for seeking medical evaluation, which together facilitate diagnosis before substantial metabolic deterioration occurs [8-10]. This earlier recognition may account for the less severe metabolic profiles observed in our cohort. No significant differences were found in the frequencies of diabetes-associated autoantibodies between first- and subsequent-affected children, suggesting broadly similar underlying autoimmune processes at disease onset in familial and sporadic T1DM. However, inter-center variability in autoantibody assessment might have reduced the sensitivity for detecting subtle between-group differences.

When examining the affected first-degree relative type, children with a parent affected by T1DM tended to be diagnosed at a younger age compared with those with affected siblings. Nevertheless, the difference was not statistically significant. Turtinen et al. [10] demonstrated that index children with an affected parent tended to be diagnosed at a younger age compared with those with an affected sibling, thereby supporting the notion that parental transmission may predispose offspring to earlier disease onset. Contrastingly, Lebenthal et al. [6] found an opposite trend, reporting that children with a parental history of T1DM were diagnosed at an older age than those with an affected sibling. Further studies with larger cohorts and diverse ethnic backgrounds are warranted to clarify these discrepancies and better define the clinical distinctions between parent–offspring and sibling transmission.

Herein, among children with an affected parent, the majority had a father with T1DM, consistent with the findings of higher paternal than maternal transmission of the disease [26,27]. Several biological mechanisms have been proposed to explain the predominance of paternal transmission of T1DM. Genetic hypotheses include selective loss of fetuses carrying high-risk susceptibility genes in mothers with T1DM, preferential transmission of susceptibility alleles from fathers, and parent-of-origin effects influencing immune-related gene expression [28]. In addition, intrauterine environmental factors unique to maternal T1DM—such as exposure to maternal hyperglycemia, exogenous insulin, and transplacentally transferred islet autoantibodies—might promote immune tolerance in the offspring [28]. Other proposed mechanisms include maternal microchimerism, whereby transplacentally transferred maternal cells persist in the offspring and promote immune tolerance through regulatory T-cell induction, as well as the passive transfer of maternal antiviral antibodies that modulate early immune responses and reduce the risk of islet autoimmunity [28]. Although these hypotheses remain speculative, they might provide a plausible biological framework supporting the observed parent-of-origin differences in T1DM transmission.

This study has several strengths. It represents the first and largest multicenter analysis of familial T1DM conducted in Korea, with standardized ascertainment of autoantibody positivity at diagnosis and detailed phenotypic characterization. Furthermore, by differentiating between parental and sibling family history, we provide novel insights into the heterogeneity of familial T1DM presentation in this population.

Nevertheless, some limitations should be noted. First, owing to the retrospective multicenter design, the assessment of family history relied on abstraction from medical records at each participating center. Although family history is routinely obtained at the time of diagnosis, the depth and consistency of documentation might have varied across centers, and undiagnosed or unreported affected relatives cannot be excluded. In addition, autoantibody testing was performed as part of routine clinical care using different assays across centers, with positivity determined according to institution-specific reference ranges, which might have limited the comparability of autoantibody results. Second, despite the large overall cohort, the absolute number of familial cases was relatively small (32 of 936 index cases). Consequently, subgroup analyses—such as comparisons between parental and sibling transmission—should be interpreted with caution, as the statistical power to detect subtle differences was limited. Third, we did not have access to genetic or HLA data, which are central to understanding the pathogenesis and familial aggregation of T1DM. Given the retrospective, multicenter design of this study and the fact that HLA typing is not part of routine clinical practice in most centers, systematic genetic characterization was not feasible. This limitation is particularly relevant in Asian populations, where HLA risk alleles and their effect sizes differ from those reported in Western cohorts. Accordingly, our findings of familial clustering should be interpreted at the epidemiologic and clinical level, and future prospective studies incorporating standardized genetic and HLA analyses are warranted. Finally, our study did not include long-term outcomes such as glycemic trajectories, complications, or psychosocial burden, which warrant further prospective investigation.

In conclusion, this multicenter Korean study demonstrated that the prevalence of familial T1DM among children was lower than that reported in Western populations. Familial cases presented with milder metabolic derangements and earlier age of onset, particularly when a parent—most often the father—was affected. These findings underscore the influence of both genetic and environmental factors on disease onset and highlight the importance of tailored approaches to early detection and risk counseling in different ethnic populations.

SUPPLEMENTARY MATERIALS

Article information Continued