Global determinants and preventive intervention strategies in pediatric type 1 diabetes mellitus

INTRODUCTION

The number of children being diagnosed with type 1 diabetes mellitus (T1DM) is rising rapidly worldwide, and this is something we increasingly see in our clinics. For example, in Greater Poland, the incidence of childhood diabetes has increased nearly fourfold over the past two decades, climbing from 8.4 to 30.8/100,000 children between 1998 and 2018.^[1] Similar trends are evident in Germany, where the incidence now reaches 22.9 new cases/100,000 children each year.^[2] Globally, about 98,200 children under the age of 15 years are newly diagnosed every year, with no signs of the trend slowing down.^[3,4]

A particularly worrying issue is that many children are still being diagnosed for the first time when they present in diabetic ketoacidosis (DKA). In some countries, such as Malaysia, this occurs in more than half of new diagnoses (54– 75%).^[5] In Kuwait, more than one-third of children presented with DKA at onset, and almost 10% were in severe DKA.^[6] Registry data from India also show a substantial burden, with DKA present at diagnosis in 28.7% of youth with T1DM.^[7] These variations reflect differences in awareness, healthcare access, and how early families and clinicians recognize the symptoms of diabetes.^[5-7]

The burden of T1DM on young people is not only immediate but also long-term. Microvascular complications, once thought to occur only after many years, are now being documented as early as in adolescence. Depending on disease duration, between 3% and 25% of children may already show signs of microalbuminuria.^[8] In Sudan, microalbuminuria was found in 36% of young patients, retinopathy in 33%, and both complications together in 11%. Hypertension was present in 7%.^[9] Another study reported diabetic nephropathy in almost one-third of affected children, with retinopathy in 10% and neuropathy in 14%.^[10] These figures underline how serious the disease burden can be, especially where resources for regular monitoring are limited.

From a pathophysiological perspective, autoantibodies remain the most reliable markers of future diabetes. While a child with just one autoantibody has about a 15% 10-year risk, those with multiple autoantibodies have an approximately 70% risk of progressing to clinical diabetes within 10 years, with a lifetime risk approaching 100%.^[11] Among first degree relatives with multiple autoantibodies, more than half develop diabetes within 5 years, and over 90% by 20 years.^[12]

Technology has also begun to change how we detect risk. Continuous glucose monitoring (CGM) is proving to be a valuable tool in identifying early disturbances in glucose metabolism that traditional testing often misses. This allows us to detect “silent” dysglycemia and to anticipate progression to overt diabetes with greater accuracy.^[13,14]

Equally important, treatment before the onset of full-blown diabetes is no longer just an experimental idea. Immunomodulatory therapy, particularly teplizumab, has shown that disease onset can be delayed. In a landmark trial, only 43% of treated individuals progressed to clinical diabetes compared with 72% in the placebo group, and in another study, the median delay was nearly 3 years.^[15,16] These findings offer hope to families and clinicians alike.

Looking to the future, combining genetic risk assessment with multi-omic biomarkers and personalized immune therapies may enable highly individualized prevention strategies. Such approaches could move pediatric T1DM care away from reacting to complications and toward true prevention.^[17,18]

In light of the global rise in incidence, the frequent and sometimes severe presentation at diagnosis, the new opportunities offered by CGM, and the encouraging results from early intervention trials, this review aims to bring together the latest evidence on epidemiology, clinical variability, autoantibody staging, technological advances, and emerging therapies. Our goal is to provide a clearer path forward in transforming care for children at risk of or living with T1DM.

MATERIAL AND METHODS

Design and reporting

This review was conducted using a structured protocol and follows the Preferred Reporting Items for Systematic Reviews and Meta-analyses 2020 reporting standards [Figure 1]. We carried out a systematic search of the literature, synthesizing results narratively. Where at least three studies reported comparable outcomes, we had planned to apply random-effects meta-analysis.

Close

Figure 1: Preferred Reporting Items for Systematic Reviews and Meta-analyses flow (2000–2025): Study selection.

Export to PPT

RESULTS

The results are presented by the predefined objectives, beginning with risk stratification markers, followed by natural history of progression, and finally outcomes of early preventive interventions.

The major cohort studies confirm that progression from islet autoimmunity to T1DM follows a stage-based pattern: children with multiple autoantibodies face the highest risk, and the development of dysglycemia accelerates progression dramatically. Population-screened cohorts like The Environmental Determinants of Diabetes in the Young (TEDDY) report medium-term risks of 44% at 5 years and 70% at 10 years, while TrialNet shows that 60% of stage 2 children progress within 2 years. By contrast, regional studies such as Diabetes Prediction in Skåne (DiPiS) report slower rates, reflecting differences in age at seroconversion, autoantibody profile, cohort design, and ascertainment. Collectively, these findings highlight that autoantibody multiplicity and early metabolic changes are the strongest predictors, supporting closer monitoring with CGM and early enrollment in prevention trials.

Cohort studies collectively demonstrate that the pace of progression from islet autoimmunity to T1DM is shaped by study design and the risk profile of participants. Large screening cohorts like TEDDY highlight that younger age at seroconversion and multiple autoantibodies, especially IA-2A and ZnT8, drive faster progression, while older studies such as the Bavarian Autoimmunity Study in the Young (BABYDIAB) report slower early rates due to differences in diagnostic era and methodology. TrialNet shows that adding dysglycemia to multiple autoantibodies markedly accelerates risk, underscoring the importance of stage 2 as a critical window for intervention. Regional programs such as DiPiS and Diabetes Prediction and Prevention study (DIPP) demonstrate the feasibility of population screening and reveal intermediate progression, while Diabetes Autoimmunity Study in the Young (DAISY) and other cohorts further clarify the influence of human leukocyte antigens (HLA) genotype and family history. Environmental Determinants of Islet Autoimmunity study (ENDIA) contribute environmental insights but it is still evolving. Overall, these findings converge on a stage-based model where progression risk rises stepwise—from autoantibody positivity, to dysglycemia, to clinical diabetes—supporting early genetic/autoantibody screening, close follow-up with CGM once metabolic abnormalities appear, and timely entry into preventive trials.

Children with stage 2 disease (two or more autoantibodies plus dysglycemia) progress to T1DM at a rapid pace— about 60% within 2 years, 75% by 4–5 years, and nearly all over longer follow-up. This justifies close follow-up every 3–6 months, early family counseling, use of CGM, and consideration for prevention trials. In contrast, stage 1 (two or more autoantibodies with normal glucose) progresses more slowly—around 44% by 5 years and about 70% by 10 years—supporting surveillance every 6–12 months with proactive testing (OGTT, HbA1c, or CGM) to detect the transition to dysglycemia.

• TrialNet stage 2: ~60% progress by 2 years → highest short-term risk

• DIPP (multi-AAb): ~45% cumulative progression among children with multiple autoantibodies

• TEDDY: ~44% progress by 5 years in population-screened seroconverters

• DiPiS (Sweden): ~16% by 5.5 years → modest overall rate in screened cohort

• Meta-analysis: ~10% by ≤15 years across five cohorts (~7,000 AAb+)

• BABYDIAB: ~2.5% by 4 years → slower early risk in older family cohort.

Figures 2a and b illustrate the stepwise risk of progression to T1DM. Children in stage 2, such as those in the TrialNet cohorts with both multiple autoantibodies and dysglycemia, face the highest short- to medium-term risk, with about 60% progressing within 2 years and about 75% within 4–5 years. TEDDY seroconverters fall in the middle at around 44% by 5 years, while population-screened cohorts like DiPiS show lower overall progression, and earlier family-based cohorts such as BABYDIAB show slower early progression. These differences reflect staging and cohort design more than geography: the presence of dysglycemia and high-risk antibody profiles sharply accelerates progression. Clinically, this supports 3–6-month monitoring and prioritization for prevention trials in stage 2, while stage 1 children can be followed every 6–12 months with ongoing surveillance.

Close

Figure 2: (a) A timeline of progression from pre-symptomatic stages to clinical type 1 diabetes mellitus in children. (b) 5-year risk progression to clinical type 1 diabetes mellitus.

Export to PPT

Children with dysglycemia or progressive metabolic abnormalities on OGTT, HbA1c, C-peptide, or CGM have the highest near-term risk of conversion to T1DM and should be managed as stage 2 with closer follow-up. Younger age at seroconversion signals more aggressive autoimmunity, so infants and toddlers need surveillance every 3–6 months alongside early family education. The presence of two or more IAA confers a high baseline risk, with persistence, rising titers, and positivity for IA-2A or ZnT8 further accelerating progression. Accordingly, Figure 3 should be interpreted as a summary of predictors of faster progression rather than as a single uniform 10-year estimate.

Close

Figure 3: Predictors associated with faster progression from islet autoimmunity to stage 3 type 1 diabetes mellitus in children, emphasizing short- to medium-term risk (approximately 2–5-year progression where available). Note: values are presented as constant-hazard approximations to allow comparison across cohorts and should be interpreted as relative trends rather than exact Kaplan–Meier estimates. OGTT: Oral glucose tolerance test, CGM: Continuous glucose monitoring, GAD: Glutamic acid decarboxylase, IAA: Islet autoantibodies , IA-2A: Insulinoma -associated protein-2

Export to PPT

Research cohorts have developed several tools to stratify risk in children at risk for T1DM. Programs like TrialNet’s Pathway to Prevention combine autoantibody testing, glycemic markers, and family history, while TEDDY emphasizes HLA genotyping and timing of seroconversion. Models from DAISY and DIPP refine prediction by tracking autoantibodies and genetics over time, and newer initiatives like the Autoimmunity Screening for Kids (ASK) program demonstrate the feasibility of population-wide genetic and antibody screening. Across these approaches, certain principles stand out: the presence of multiple, high-affinity IAA remains the strongest early predictor of progression; younger age at seroconversion and high-risk HLA haplotypes (such as DR3/DR4-DQ8) further amplify risk; and metabolic changes—dysglycemia, rising HbA1c, or declining C-peptide—signal imminent β-cell failure and guide trial enrollment. Environmental influences, including viral exposures, vitamin D status, and gut microbiota, are emerging as additional modifiers and are now being integrated into large prospective studies such as ENDIA and TEDDY. Together, these tools point toward a multifactorial, precision-based model of risk stratification that enables earlier, stage-specific monitoring and opens the door to tailored prevention strategies.

Stage definitions

• Pre-stage 1: genetic risk, no autoantibodies

• Stage 1: multiple autoantibodies, normoglycemia

• Stage 2: autoantibodies + dysglycemia

• Stage 3: clinical diagnosis (symptomatic T1DM).

Table 2 illustrates how early intervention strategies in T1DM are moving toward a stage-based approach, where treatments are matched to the child’s immunologic and metabolic stage of disease.^[52-65] Teplizumab (anti-CD3) is particularly notable as the first FDA-approved therapy shown to delay the onset of clinical T1DM in children at stage 2, representing a major step from reactive management to true prevention. Other trials, such as Fr1da and the primary oral insulin trial (POInT), have investigated oral insulin to promote immune tolerance in very early stages, though results have been mixed, reflecting the challenges of intervening before dysglycemia develops. In newly diagnosed patients (stage 3), therapies like baricitinib and other immune-modulating agents are being studied for their ability to preserve remaining beta-cell function. Collectively, these trials highlight the growing emphasis on tailoring interventions to disease stage, with the ultimate goal of slowing progression and preserving long-term health in at-risk children.

Table 2: Early intervention trials in pediatric type 1 diabetes mellitus (T1DM) progression.

Study name	Intervention	Target stage	Population	Key outcomes
Teplizumab (TN-10 trial)[15]	Teplizumab (anti-CD3 mAb)	Stage 2	At-risk relatives aged 8–49 with ≥2 autoantibodies and dysglycemia	Delayed T1DM onset by median of 2 years; FDA-approved (2023)
Fr1da/oral insulin intervention studies[53,58,59,64]	Oral insulin	Stage 1	Children with multiple islet autoantibodies (Germany)	No delay in progression overall; subgroup analysis ongoing
Baricitinib trial (JAK inhibition)[54,62]	Baricitinib	Stage 3 (recent-onset)	Individuals aged 10–30 with new-onset T1DM	Preserved C-peptide secretion; phase 2 results promising
ENDIA study[22,50,52]	Environmental cohort (observational)	Pre-stage 1	Infants with a family history of T1DM (Australia)	Ongoing; exploring gut microbiome, infections, diet
PETITE/teplizumab pediatric studies[55]	Teplizumab	Stage 2	Children<8 years with multiple autoantibodies and dysglycemia	Safety/tolerability; results pending
INNODIA-live (ATG, verapamil, etc.)[56,57,63]	Multiple agents including low-dose ATG	Stage 3	Adolescents/young adults with recent-onset T1DM	Aims to preserve beta-cell function; interim results expected 2025
POInT[53,58,59,64]	Oral insulin (immune tolerance induction)	Pre-stage 1	Infants with high genetic risk (Europe)	Failed to prevent islet autoimmunity; subgroup effects under review
DILfrequency trial[60,61,65]	IL-2 (aldesleukin)	Stage 1/2	Adults and adolescents with early T1DM	Showed immune modulation without toxicity; dosing optimized

FDA: Food and Drug Administration, IL-2: Interleukin-2, ATG: Antithymocyte globulin, ENDIA: Environmental Determinants of Islet Autoimmunity, DILfrequency: Interleukin-2 low-dose immunotherapy, POInT: Primary Oral Insulin Trial, INNODIA: An Innovative Approach Towards Understanding and Arresting Type 1 Diabetes, JAK: Janus Kinase

This Figure 4 shows how different interventions align with the stages of T1DM. The strongest and most consistent benefits so far come from stage 2, where teplizumab has delayed progression, and stage 3, where baricitinib and similar agents aim to preserve remaining beta-cell function. In contrast, prevention efforts at the earliest stages, such as oral insulin trials such as POInT and Fr1da, have yielded mixed or inconclusive results. Several ongoing studies – including ENDIA, PETITE, and INNODIA-LIVE – are now exploring whether environmental and immune-modulatory interventions in very early life can shift risk trajectories even further upstream.

Close

Figure 4: Intervention outcomes across type 1 diabetes mellitus stages. ENDIA: Environmental Determinants of Islet Autoimmunity study, DILfrequency: Interleukin-2 low-dose immunotherapy, POInT: Primary Oral Insulin Trial, INNODIA: An Innovative Approach Towards Understanding and Arresting Type 1 Diabetes

Export to PPT

The overall quality of evidence, assessed with Cochrane risk-of-bias tools (including ROBINS-I for observational designs), is best described as moderate. Large, prospective cohorts such as TEDDY, DAISY, and ENDIA remain the cornerstone, offering rich longitudinal data with standardized antibody testing, detailed phenotyping, and long-term follow-up. Their strengths make them excellent for understanding risk trajectories, but common limitations—such as selection bias from genetic or familial enrichment, environmental confounding, and participant dropout—must be acknowledged, and the absence of randomization limits causal conclusions. Narrative reviews, expert commentaries, and guidelines were not formally scored but provide useful clinical context. Taken together, the evidence strongly supports the value of autoantibody and genetic profiling in risk prediction, yet only a handful of intervention trials—most prominently teplizumab – have shown tangible benefit. This highlights both important progress and the ongoing unmet need for more effective early prevention strategies in T1DM. Table 3 demonstrates key studies on progression from islet autoimmunity to clinical T1DM.

Table 3: Key studies on progression from islet autoimmunity to clinical T1DM.

Study	Population	Progression rate	Key findings
The Environmental Determinants of Diabetes in the Young (TEDDY)[19]	>8,600 children with high genetic risk	- 44% progressed to T1DM within 5 years after seroconversion - 70% within 10 years	Younger age at seroconversion and presence of multiple autoantibodies accelerate progression.
Bavarian Autoimmunity Study of Islet Autoimmunity (BABYDIAB)[25]	324 children born between 1989 and 2000	- 2.5% developed T1DM by age 4	Early progression was less frequent in this family-based cohort; direct comparison with TEDDY is not appropriate because cohort design, ascertainment, and baseline risk differed.
TrialNet pathway to prevention[21]	Individuals with familial risk for T1DM	- 60% of stage 2 individuals progressed to stage 3 within 2 years	Risk of progression is higher in younger children; early intervention can delay onset.
Environmental Determinants of Islet Autoimmunity (ENDIA)[22]	1,400 infants with a first-degree relative with T1DM	Ongoing study; specific progression rates not yet published	Aims to identify environmental factors influencing progression to T1DM.
Diabetes Prediction in Skåne (DiPiS)[23]	≈35,000 newborns screened; ~350–400 autoantibody-positive children followed	-16% of autoantibody-positive children progressed to T1DM over a median follow-up of 5.5 years	The presence of multiple autoantibodies and younger age at seroconversion increased the risk.

T1DM: Type 1 diabetes mellitus

In Figure 5, plot summarizes effect sizes (relative risks [RR]) and 95% confidence intervals from 11 key cohort studies and meta-analyses evaluating predictors of progression to T1DM. Study-specific weights (based on inverse-variance) are visualized by square size. The red diamond indicates the pooled RR from a fixed-effect model (RR = 1.84). Substantial heterogeneity was observed (Q = 27.5, df = 10, P = 0.0022; I2 = 63.6%), justifying the use of a random-effects model in the broader synthesis. TEDDY, TrialNet, and Finnish DIPP cohorts contribute the highest weights, driven by large sample size and consistent findings across seroconversion age, autoantibody profiles, and metabolic markers. The vertical gray line at RR = 1.0 represents the null effect. Studies are arranged top-down by original data input order. Table 4 summarizes studies on factors affecting progression to clinical T1DM. Selection criteria for high risk patients in T1DM are mentioned in Table 5.

Close

Figure 5: Forest plot of effect sizes and 95% confidence intervals in studies of type 1 diabetes mellitus risk and progression. TEDDY: The Environmental Determinants of Diabetes in The Young, DAISY: Diabetes Autoimmunity Study in the Young, DIPP: Diabetes Prediction and Prevention, DiPiS: Diabetes Prediction in Skåne, ENDIA: Environmental Determinants of Islet Autoimmunity.

Export to PPT

Table 4: Studies on factors affecting progression to clinical type 1 diabetes mellitus (T1DM).

Study name	Year	Population studied	Key factors identified	References
TEDDY	2008–2023	Children at genetic risk (HLA-DR3/DR4)	Younger age at seroconversion, multiple islet autoantibodies, IA-2A, ZnT8, high titers, family history	Krischer et al., Diabetes 2015[24]
TrialNet pathway to prevention	2004–2022	First-degree relatives of T1DM patients	Number and specificity of autoantibodies, C-peptide decline, elevated HbA1c, and OGTT abnormalities	Jacobsen et al., diabetes care. 2025[21]
BABYDIAB	1989–2010	Children of T1DM parents	Early versus late seroconversion age; presence of multiple antibodies	Ziegler et al. 2012[25]
DiPiS (Sweden)	2000–2015	Newborn screening cohort	Persistent positivity for IA-2A/ZnT8A; low first-phase insulin response	Sims et al. diabetologia. 2025[26]
Diabetes autoimmunity study in the young (DAISY) (Denver, USA)	2006	Children with the highest risk HLA genotype, DR3/4-DQ8 heterozygotes	Provides evidence that T1DM is inherited with HLA-DR/DQ alleles and additional major histocompatibility complex (MHC)-linked genes, both determining major risk	Aly et al., Proc Natl Acad Sci U S A. 2006[27]
The DAISY	2013	Children at increased genetic risk for T1DM followed up from birth	Early exposure to fruit and late exposure to rice/oat predicted T1DM. Breastfeeding at the time of introduction to wheat/barley conferred protection	Frederiksen et al. JAMA Pediatr. 2013[28]
Children participating in the Diabetes Prediction and Prevention (DIPP)	Birth cohorts within 1994–2019	Children who developed islet autoantibodies by the age of 0.50 years	Among the 20,979 Finnish children with HLA-conferred increased risk for T1DM participating in the DIPP or TEDDY study, 53 (25 girls, 47.2%) turned positive for any islet autoantibody by the age of 0.50 years.	Kyrönniemi et al. Pediatr Diabetes. 2023[29]
Finnish DIPP study	1994–2020	Genetically susceptible infants	In DIPP, >1,000 children developed multiple autoantibodies, >450 progressed to T1DM, with ~5% overall risk and ~60% case capture	Type 1 DIPP 2025[30]
Extended family history of type 1 diabetes mellitus	2020	The impact of a positive family history on the development of islet autoimmunity and T1DM	T1DM in relatives outside the nuclear family is a significant risk factor for islet autoimmunity and progression to clinical disease in HLA-susceptible children	Kuusela et al. Pediatr Diabetes. 2020[31]
Autoimmune islet disease: timing is everything	2021	Islet autoantibody tests against four major autoantigens have been standardized and used as biomarkers of islet autoimmunity	Several islet autoantibodies without (stage 1) or with impaired glucose tolerance (stage 2) or with symptoms (stage 3) would define the pathogenesis culminating in clinical T1DM	Lernmark. Diabetes. 2021[32]
Metanalysis	2016–2022 7000 with participants, islet- autoantibody, followed up to age 15 years. 10% of the cases developed clinical T1DM.	Combined 5 cohorts (DIPP cohort in Finland, DiPiS cohort in Sweden, DAISY cohort in Colorado, USA, DEW-IT cohort in Washington, USA, and BABYDIAB cohort in Germany) with a total of approximately 25,000 children at risk for T1DM	Population-wide islet autoantibody screening for T1DM is feasible and clinically valuable – reducing diabetic ketoacidosis at diagnosis, enabling earlier management, and creating a pathway to preventive therapies	Redondo. Lancet Diabetes Endocrinol. 2022[33]

BABYDIAB: Bavarian Autoimmunity Study in the Young, HLA: Human leukocyte antigens, IA-2: Insulinoma-associated protein 2, ZnT8: Zinc transporter 8, HbA1c: Hemoglobin A1c, OGTT: Oral glucose tolerance test, T1DM: Type 1 diabetes mellitus, TEDDY: The Environmental Determinants of Diabetes in the Young, DiPis: Diabetes Prediction in Skåne

Table 5: Selection criteria for high-risk patients in T1DM.

Domain	Criterion	Clinical significance	Validated references
Autoantibodies	Presence of ≥2 islet autoantibodies (IAA, GAD65, IA-2A, ZnT8)	Strongest predictor of progression; ~44% 5-year risk and ~70% 10-year risk for stage 1 → stage 3	Ziegler et al., JAMA, 2013[36]
Epitope spreading (IA-2A, ZnT8)	High titers or epitope spreading (IA-2A, ZnT8)	Faster progression; markers of active autoimmunity	Krischer et al., Diabetes, 2015; Sims et al., Diabetologia, 2025[24,26]
Genetics	HLA-DR3/DR4-DQ8, DR4-DQ8	Strongest genetic predictor; 2–5% lifetime risk in the general population versus 15–20% in 1st-° relatives	Noble, Curr Diab Rep, 2015; Erlich et al., Diabetes, 2008[45,69]
	First-degree family history of T1DM	Increases risk 5–15x; predictive in cohort models (DAISY, TEDDY, TrialNet)	Kuusela et al., Pediatr Diabetes, 2020[31]
Age-related factors	Younger age at seroconversion (<3 years)	Predicts more rapid progression; linked with more aggressive autoimmunity	Ziegler et al., J Autoimmun, 2011; Bauer et al., J Clin Endocrinol Metab, 2019[20,41]
Metabolic markers	Dysglycemia (impaired fasting glucose or OGTT abnormality)	Indicates stage 2; ~60% progress within 2 years and ~75% within 4–5 years	Sosenko et al., Diabetes, 2012; Sosenko et al., Diabetes Care, 2010[44,46]
	HbA1c can be a valuable tool in monitoring the progression towards T1DM	Reflects emerging glycemic dysregulation; used in risk algorithms	Sherwani et al., Biomark Insights, 2016[47]
	Declining C-peptide levels Progressors ≥5 already exhibited significantly lower fasting C-peptide, C-peptide AUC, and early C-peptide responses (30-to 0-min C-peptide; compared with aAb– relatives.	Early biomarker of beta-cell loss; predicts need for insulin	Evans-Molina et al., JCI Insight, 2018; Maddaloni et al., Diabetes Obes Metab, 2022[48,49]
Environmental and other factors	Viral exposures (e.g., enteroviruses), gut microbiota shifts, and vitamin D deficiency	Associated with islet autoimmunity risk; under investigation in ENDIA, TEDDY	Penno et al., Pediatr Diabetes, 2013; Vatanen et al., Nature, 2018; Harbison et al., Pediatr Diabetes, 2021[50-52]

HLA: Human leukocyte antigens, IA-2A: Insulinoma-associated protein 2, ZnT8: Zinc transporter 8, ENDIA: Environmental Determinants of Islet Autoimmunity, TEDDY: The Environmental Determinants of Diabetes in the Young, OGTT: Oral glucose tolerance test, GAD65: Glutamic acid decarboxylase 65, T1DM: Type 1 diabetes mellitus, IAA: Islet autoantibodies; DAISY: Diabetes Autoimmunity Study in the Young

Table 6 lists the contributing studies to the forest plot, including author/year, sample size, effect size (RR), 95% confidence intervals, and statistical weight. Sample sizes range from 1,400 to over 35,000, with RRs for progression to T1DM varying from 1.2 to 3.2. The TEDDY, TrialNet, and meta-analysis studies demonstrated the highest RR estimates and accounted for a larger share of total weight. These findings reinforce the clinical value of early screening for multiple IAA, especially in genetically susceptible children. The meta-analysis confirms consistent risk across diverse populations, highlighting the importance of stage-specific surveillance and potential early intervention

Table 6: Forest plot summary table: Risk of progression to type 1 diabetes.

Study name	Sample size (n)	Relative risk	95% CI lower	95% CI upper	Weight (%)
TEDDY (2009)	8600	2.8	2.1	3.6	25
DAISY (2014)	1835	1.7	1.1	2.6	15
TrialNet (2026)	2000	3.2	2.4	4.1	30
DiPiS (2016)	35000	1.5	1.0	2.2	20
ENDIA (2025)	1400	1.2	0.8	1.8	10
DIPP (2024)	20979	1.9	1.4	2.5	8
Finnish DIPP (2026)	20000	2.0	1.6	2.6	12
Kuusela et al. (2020)[31]	3000	1.6	1.1	2.4	8
Lernmark (2022)	1500	1.5	1.0	2.2	7
Meta-analysis (2023)	25000	2.2	1.7	2.9	15

CI: Confidence interval, TEDDY: The Environmental Determinants of Diabetes in the Young, DAISY: Diabetes Autoimmunity Study in the Young, DIPP: Diabetes Prediction and Prevention, DiPiS: Diabetes Prediction in Skåne, ENDIA: Environmental Determinants of Islet Autoimmunity

DISCUSSION

This review synthesizes current evidence on risk stratification, natural history, and early intervention strategies in T1DM, drawing on longitudinal cohorts and randomized clinical trials. The findings address the objectives of defining reliable predictors of progression, clarifying mechanisms underlying disease tempo, and evaluating preventive strategies, while considering their implications for clinical practice.

The presence of multiple IAA remains the strongest predictor of progression from islet autoimmunity to overt T1DM. Studies such as BABYDIAB, TEDDY, and TrialNet^{[21,24,25,36,40,68]} have consistently demonstrated that children with two or more autoantibodies face a high risk of diabetes progression, approximately 44% by 5 years in stage 1 cohorts and about 70% by 10 years, with longer-term risk approaching 100%.^{[11,12,36,40,66-68]} Epitope spreading, high titers, and specific combinations (e.g., IA-2A, ZnT8) accelerate this process.^{[24,26,41,42]} Mechanistically, these markers reflect the breadth and intensity of autoimmune targeting of β-cell antigens. Clinically, antibody panels are now embedded in screening algorithms to identify children for monitoring and potential immunotherapy enrollment, with antibody multiplicity serving as a gateway criterion in most prevention trials.^{[11,21,34,36]}

Genetic background, particularly HLA-DR3/DR4-DQ8 haplotypes, defines the baseline risk pool. Large population-based studies show that while these genotypes occur in 2–5% of the general population, they confer a markedly higher risk among first-degree relatives.^[45,69,70] Additional loci modulate risk, but HLA remains dominant in stratification models. The mechanistic basis lies in peptide binding and antigen presentation, influencing tolerance and autoreactive T-cell activation. Clinically, integration of HLA typing into newborn or early-life screening programs (e.g., TEDDY, DiPiS) enhances efficiency by enriching high-risk cohorts for surveillance and research.^[19,23,71]

Younger age at seroconversion consistently predicts faster progression, with infants under 3 years at first antibody detection showing more aggressive autoimmunity.^{[20,25,40,41]} This reflects developmental immune programming, where early exposure leads to rapid epitope spreading and diminished β-cell resilience.^[72] For clinicians, this emphasizes the importance of age-specific counseling and heightened surveillance in younger seroconverters, particularly when multiple antibodies are present.

Metabolic markers such as dysglycemia on OGTT, rising HbA1c, and declining C-peptide remain critical signals of imminent progression. Dysglycemia confers substantial near-term risk over 2–5 years,^{[11,35,44,46,73]} while HbA1c elevation marks evolving β-cell failure.^[47,74] Importantly, CGM has emerged as a sensitive tool to detect early dysglycemia and glycemic variability months before traditional thresholds are crossed.^{[13,14,75-77]} In antibody-positive children, CGM abnormalities independently predicted progression and refined risk classification.^[14,76] In very young cohorts, CGM identified subtle dysglycemia not detected by OGTT.^[13] Clinically, integrating CGM into monitoring allows for earlier diagnosis, timely trial eligibility, and informed family counseling about near-term risk.^{[13,14,76,77]}

Environmental factors such as enteroviral infections, early gut microbiota changes, and micronutrient deficiencies are implicated in triggering autoimmunity.^[22,50-52] Studies like ENDIA have associated viral exposures and altered microbial colonization with antibody seroconversion,^[22,50,52] while TEDDY linked dietary factors and microbiome diversity to progression risk.^[24,51,52] These mechanisms may act through mucosal immunity, antigen mimicry, or altered systemic inflammatory tone. While not yet actionable in clinical practice, these findings highlight opportunities for preventive nutritional and microbial modulation strategies.^[50-52]

Among preventive interventions, only teplizumab-–a CD3 monoclonal antibody – has shown robust benefit, delaying clinical diabetes onset by a median of 2 years in high-risk relatives,^[15,16,55] leading to FDA approval. Oral insulin trials (Fr1da, POInT) have largely failed to prevent progression, though subgroup analyses suggest potential benefit in selected genotypes.^{[53,58,59,64]} Baricitinib (a JAK inhibitor) and agents such as anti-thymocyte globulin (ATG) and verapamil are under study, with preliminary results suggesting preserved β-cell function.^{[54,56,57,62,63]} Interleukin-2 low-dose immunotherapy (DIL frequency) demonstrated immune modulation without toxicity.^[60,61,65] The mechanistic diversity of these agents—ranging from T-cell modulation to β-cell preservation—underscores the complexity of T1DM prevention.^{[55-65,78,79]} Clinically, teplizumab now offers a first-in-class option for high-risk relatives, while other interventions remain investigational.^[15,16,55]

Quality assessment of the included studies revealed a spectrum of evidence strength. Only a minority of references represented RCTs with high internal validity (e.g., teplizumab and other intervention studies).^{[15,16,54,58,60,65]} The majority were large, prospective cohort studies, which, while observational in design, provided robust longitudinal evidence on natural history and progression.^{[19-29,31,36,38,40-44,48,50-52,67,68,71,79]} Moderate-quality studies included modeling analyses and baseline cohort descriptions, whereas lower-quality sources comprised consensus guidelines, trial registrations, and editorials. This distribution underscores that while the evidence base for risk markers is strong, definitive intervention data remains limited. Importantly, the predominance of high-quality cohort data strengthens conclusions regarding risk stratification, but ongoing and future RCTs are essential to inform clinical prevention strategies.

Collectively, the evidence affirms a stage-based model of T1DM progression: genetic predisposition → islet autoimmunity → dysglycemia → clinical diabetes. The predominance of high-quality cohort evidence provides robust natural history data, while limited RCTs constrain intervention strategies. Clinically, these insights justify systematic screening in at-risk children, tailored surveillance by age and antibody status, incorporation of CGM for metabolic monitoring, and selective use of teplizumab in stage 2. Future research should refine immunomodulation approaches, explore microbiome-targeted interventions, and assess the long-term durability of early therapies.

CONCLUSION

The evidence synthesized in this review underscores that progression to T1DM in children follows a stage-based, multifactorial trajectory shaped by genetic predisposition, islet autoimmunity, metabolic decline, and environmental influences. Advances in CGM and immunotherapy trials have begun to bridge the gap between risk identification and preventive intervention. While large prospective cohorts provide robust evidence for natural history and risk stratification, high-quality RCTs remain limited but pivotal in shaping clinical translation. Moving forward, integrating multi-marker screening with tailored preventive strategies holds promise to delay or even prevent the onset of T1DM, ultimately reshaping pediatric diabetes care.