Early Onset of Type 1 Diabetes in Kuwait: Distinct Clinical, Metabolic, and Immunological Characteristics

The incidence and prevalence of type 1 diabetes (T1D) have been rising globally [1]. Kuwait has one of the highest incidence rates of T1D in the world with an incidence rate of 40.9 per 100,000 per year [1]. Similar to international reports, Kuwait has also reported an increase in incidence of T1D in younger children with an annual increase of 3.2% for children 4 years and younger [1, 2]. The reasons behind such an increase of T1D globally in all age groups and specifically in younger children remain unclear [3].

T1D is a chronic autoimmune disorder characterized by insulin deficiency secondary to pancreatic β-cell destruction [4]. The etiology of T1D is quite complex and involves an interplay between inherent predisposition as suggested by twin studies as well as environmental factors [5]. It is still unclear what specific factors play a more important role in triggering T1D in younger age groups. However, it has been known that β-cell destruction is more aggressive in younger children with T1D [6].

Diabetes at such an early age poses numerous challenges for families as well as the care providers. Younger children at diagnosis with T1D are at a higher risk of developing diabetic ketoacidosis (DKA) due to the lack of classical symptoms such as polyuria, polydipsia, and weight loss [7]. It has been reported that younger children were more likely to present with DKA as well as severe DKA [8]. Therefore, exploring early-onset diabetes in terms of describing characteristics at the time of diagnosis is likely to aid in a better understanding of etiology and can also have implications on its management and prevention. The aim of this study was to investigate the prevalence of early-onset T1D in Kuwait as well as to describe the baseline clinical, biochemical, and immunological characteristics of children diagnosed at early onset in comparison to the older children with newly diagnosed T1D.

Medical records of children newly diagnosed with T1D and registered in the Childhood-Onset Diabetes electronic Registry (CODeR) in Kuwait [1] between the period of 2017 and 2022 were reviewed. CODeR is a comprehensive prospective population-based diabetes registry maintained by Dasman Diabetes Institute in collaboration with the Ministry of Health (MOH) of Kuwait. It was established in 2011 for surveillance of children (<14 years) diagnosed with diabetes. Children diagnosed with T1D were included in the study, and early-onset T1D was defined as age at diagnosis of less than 6 years.

Patient information included in the analysis were gender, age at diagnosis in years, body mass index (BMI) z scores, presentation with DKA and its severity, and admission to the pediatric intensive care unit (PICU). BMI z scores were calculated using the World Health Organization (WHO) child growth standards [9]. DKA and its severity were defined as per the International Society of Pediatric and Adolescent Diabetes (ISPAD) guidelines of 2018 [10]. Biochemical parameters at diagnosis included the following: hemoglobin A1C (HbA1C), C-peptide levels, vitamin D levels (25-OH-Vit D), thyroid-stimulating hormone levels, and free thyroxine levels.

Immunological parameters analyzed for the current study were glutamic acid decarboxylase antibodies, thyroid antibodies, and celiac antibodies. Positive thyroid autoimmunity was defined as positive thyroid peroxidase antibodies and/or positive thyroglobulin antibodies, whereas positive celiac autoimmunity was defined as positive tissue transglutaminase antibodies and/or positive endomysial antibodies.

Statistical analysis was performed using the STATA 17.0 SE-Standard Edition (Stata Corp LLC). Continuous variables were reported as mean ± SD when normally distributed and as median (interquartile range) otherwise. To test the equality in normally distributed continuous variables, t test (mean comparison test) was used; otherwise, the Wilcoxon rank-sum test was used. Categorical variables were reported as frequencies and percentages, and the testing of association in these variables was done using Pearson χ² test of independence. To model the association between binary outcomes (PICU admission and celiac autoimmunity) and possible confounders, namely, age in years, gender, and BMI z scores, multiple logistic regression modeling was implemented. Whereas linear regression modeling was used to model the association to continuous outcomes (HbA1C, C-peptide, and vitamin D levels) and possible confounders, namely, age in years, gender, and BMI z scores. A significant level was set at a p value of less than 5%. The study was approved by the Standing Committee for Coordination of Health and Medical Research at the MOH and the Ethical Committee at Dasman Diabetes Institute (RA 2011-006) and funded by the Dasman Diabetes Institute (RA 2022-1971).

A total of 2051 children aged <14 years were registered with new-onset T1D between 2017 and 2022, of which 657 (32.0%) were diagnosed at early onset. There has been a trend of a slight increase in the percentage of early-onset T1D after 2020 (15.2%) with a prevalence of 18.4% and 20.2% in 2021 and 2022, respectively (Fig. 1, p = 0.056).

Table 1 illustrates baseline characteristics of the study cohort comparing children at early-onset T1D with the children diagnosed at an older age (≥6 years). Children with early onset of T1D presented with lower BMI z scores compared to the older children (−0.14 vs. 0.37, p < 0.0001). Despite no significant difference in the presentation with DKA or its severity, children with early-onset T1D were more likely to be admitted to the PICU.

Biochemical and immunological parameters at diagnosis were significantly different between the two comparison groups (Table 1). Children at early onset had a lower HbA1C, C-peptide, thyroid-stimulating hormone, and free thyroxine, whereas their vitamin D concentration was higher at diagnosis compared to the older children. With regards to immunological parameters, younger children were less likely to have positive glutamic acid decarboxylase antibodies (71.7% vs. 78.4%, p = 0.011) and were slightly less likely to have positive celiac antibodies (7.2% vs. 10.1%, p = 0.066). There was no significant difference in the presentation with positive thyroid autoantibodies.

Table 2 illustrates the association between certain outcomes (PICU admission, celiac autoimmunity, HbA1C, C-peptide, and vitamin D levels) and the age at diagnosis of T1D and other possible confounders. Age at onset of T1D was inversely related to admission to the PICU (OR = 0.90, 95% CI: 0.85, 0.95, p < 0.0001) and was directly related to positive celiac autoimmunity (OR = 1.09, 95% CI: 1.01–1.18, p = 0.022). Furthermore, age at onset of T1D was directly related to higher HbA1C (p < 0.0001) and C-peptide levels (p < 0.0001). However, age at onset of T1D was inversely related to the higher vitamin D levels (p < 0.0001).

In the current study, we report on the baseline characteristics of a large cohort of children aged <14 years registered with new-onset T1D in Kuwait, a country with one of the highest incidences of T1D in the world [1]. We have observed a trend of an increase in the prevalence of new-onset T1D in children younger than 6 years of age after the year 2020. Furthermore, the study showed that children with early-onset T1D have distinct clinical, biochemical, and immunological characteristics at the time of diagnosis compared to the older children.

Our study had found a trend of a slight increase in the prevalence of new diagnosed T1D in children less than the age of 6 years. This finding was close to being statistically significant with p = 0.056. Furthermore, referring to Figure 1, the percentage of new diagnosis of T1D in children aged less than 6 years had increased (18.4% in 2021 and 20.2% in 2022) compared to the previous years. This might be related to a period corresponding to after the start of COVID-19 pandemic in the country. There has been a global shift in the age of onset of T1D to younger age groups even prior to the COVID-19 pandemic including Kuwait [1, 7]. Such an increase might be attributed to changing environmental conditions rather than genetic factors which might explain the further increase after the start of the COVID-19 pandemic caused by the SARS-CoV-2 virus. In children genetically susceptible, respiratory infections coupled with negative environment events have been suggested to trigger the development of T1D [11, 12]. In addition, the indirect effects which resulted from lockdown and social isolation measures must also be considered in the increased incident of T1D, especially in younger age groups [13]. Indeed, such measures might have accelerated the process of autoimmunity against in the β cells which had led to the earlier development of the condition at a younger age. However, it should be noted that the full pancreatic autoimmune panel for the study cohort was severely lacking during the study period, especially the pandemic period, which has limited further study of characteristics of the autoimmune process against the pancreas.

In this study, early-onset T1D was associated with distinct clinical and biochemical features at presentation. They presented with lower BMI, HbA1C, as well as lower C-peptide levels, all of which support the presentation with a more severe deficiency in β-cell residual function in this age group at diagnosis [14]. Studies of human pancreata suggest a more vigorous autoimmune response occurring in young children newly diagnosed with T1D [15]. For example, younger age of onset is associated with higher levels of CD20⁺ B cells, CD45⁺ cells, and CD8⁺ T cells in insulitis lesions, with fewer insulin-positive islet [16, 17]. It has been reported that lean children have lower β-cell residual function at the onset of T1D compared to obese and overweight children [18, 19]. Moreover, the process of decompensation in younger children is reported to take a shorter period resulting at a presentation with a lower Hb1C levels but higher glucose concentrations [14, 20, 21]. In addition to potentially being a manifestation of lower β-cell residual function, lower C-peptide levels had been associated with higher risk of development of hypoglycemia on follow-up which might be difficult to identify in this age group [22].

In our study, there was no difference in the presentation with DKA or severe DKA in children diagnosed with T1D at a younger age compared to older children. This is similar to findings from the CODeR registry for children newly diagnosed with T1D during the period 2011–2013, where the frequency of DKA was similar in all age groups, namely, 0–4 years, 5–9 years, and 10–14 years [14]. Furthermore, our report from the CODeR during the collective period of 2011–2017 also showed no difference in the prevalence of DKA or severe DKA in children younger than 6 years of age compared to older children [19]. Kuwait is a country with one of the highest incidences of T1D in the world with a recent incidence of 41.7 per 100,000 per year [1]. Children newly diagnosed with diabetes in the country are referred and managed at specialized diabetes centers at secondary governmental hospitals with readily available access to medical care across the country. Moreover, the country has a well-established, nation-wide diabetes registry that provides invaluable support to healthcare planning, national diabetes management programs, public health initiatives, and awareness programs. All of which might have contributed to similar DKA prevalence and severity among all age groups. Despite the lack of difference in the presentation with DKA and severe DKA between children younger than 6 years of age and older children, younger children were more likely to be admitted to the PICU at diagnosis with T1D. This might reflect the precautionary measures healthcare providers take in dealing with new diagnosis of diabetes in younger age groups as all secondary centers dealing with diabetes in a country with accessible intensive care facilities. A recent study from Italy reports that 4 out of 10 children newly diagnosed with T1D and admitted to the PICU were younger than 5 years of age [23]. This is a vulnerable group and is at higher risk of acute complications of diabetes at presentation like DKA and cerebral injury [24]. Cerebral injury is a devastating acute complication of diabetes with a mortality rate of 21–24% [24]. It is challenging to clinically diagnose cerebral injury, especially in younger age groups; hence, an age of less than 5 years is one of the minor criteria for the diagnosis of cerebral injury [24].

In the current study, early-onset T1D was associated with a lesser likelihood of development of positive celiac autoimmunity but not thyroid autoimmunity. Celiac disease (CD) is one of the most recognized associated autoimmune diseases with T1D [25, 26]. The Swedish National Diabetes Register had recently reported that CD was found to be significantly associated with the age at T1D onset, with nearly 20% of children aged 0–4 years being diagnosed [27]. This is different compared to our study; however, it should be noted that our study had investigated the development of celiac autoimmunity at diagnosis of T1D rather than confirmation of CD with results of intestinal biopsies which were not available for analysis at the time of the study. Both T1D and CD share genetic and immunological features [25] which might be different in children of young age groups in Kuwait as well as different for thyroid autoimmunity. Furthermore, environmental factors play a crucial role in development of CD in children with T1D as well as healthy children [25, 26]. Consumption of gluten along with viral infections have been identified as two of the most related environmental factors [26] which might be different in younger age groups compared to children presenting with T1D at an older age as well as children from other parts of the world. Our finding necessitates further study of the development of CD in children with T1D in the country as well as tailoring the screening practices for these specific age groups.

Early-onset T1D was associated with higher vitamin D concentration compared to older age in the present study cohort. Vitamin D has been reported to be a key regulator of autoimmunity and has been reported to protect against the development of T1D [28]. In infancy, vitamin D supplementation has been shown to be related to a lower risk of development of T1D [29]. Results from The Environmental Determinants of Diabetes in the Young (TEDDY) had showed that higher vitamin D concentrations throughout childhood and in early infancy were associated with lower pancreatic autoimmunity [30]. In our study cohort, the median vitamin D concentration for children younger than 6 years of age was at the insufficient range of 53.0 nmol/L in comparison to a median in the deficient range of 34.0 nmol/L in older children. The TEDDY study had suggested that categorization of vitamin D sufficiency is not informative with regards to their results that link vitamin D concentration to risk to pancreatic autoimmunity [30]. Although in our cohort the median of vitamin D concentration was below the sufficient range for both age comparative groups, it seems that the role of vitamin D in the development of T1D in younger children can be present at higher concentrations in comparison to older age groups. This again might be related to an accelerated autoimmune response in younger children that had led to the development of T1D.

Our study was based on data from a well-established diabetes registry in the country with a high ascertainment rate [1]. However, due to the cross-sectional nature of the study, future longitudinal follow-up studies are needed to decern the distinct differences between children presenting at a younger age with T1D and their counterparts. There has been missing information on some of the pancreatic antibodies, namely, ICA, anti-insulin antibodies, IA-IA2, and zinc transporter 8 antibodies, which have limited further study of the immunological characteristics at younger age groups. It should be noted that zinc transporter 8 antibody testing was not universally available in the country during the early study period. Information on intestinal biopsies to diagnose CD was also not available for analysis during the study. Furthermore, the registry did not capture full information on socioeconomic status and birth and neonatal history that might be linked to the development of diabetes.

Our study showed a trend of an increase in the prevalence of early-onset T1D after 2020 that might correspond to the COVID-19 pandemic period as well as distinct clinical, biochemical, and immunological characteristics compared to older children. These findings reinforce the need for increased attention to be given to study the development of T1D in children of younger age. This in turn will support special management and prevention measures targeted toward this vulnerable age group.

The authors would like to thank all the children and their families as well as all diabetes care providers in the country. Moreover, the authors thank Kuwait University, the Ministry of Health of Kuwait, and Dasman Diabetes Institute for their continued support of the study of Diabetes in Kuwait. The authors would also like to thank the founder of CODeR, Professor Azza Shaltout. The authors would like to thank Dr. Ahmed Albatineh for his efforts in preparing the datasets for analysis.

The study was approved by the Standing Committee for Coordination of Health and Medical Research at the Ministry of Health of Kuwait and the Research Ethics Committee at Dasman Diabetes Institute (RA 2011-006). The study was performed in accordance with the Declaration of Helsinki. The data used in this study were completely anonymized without any possibility to reidentify subjects. A waiver of consent was granted by the Standing Committee for Coordination of Health and Medical Research at the Ministry of Health and Dasman Diabetes Institute.

Authors report no conflicts of interest.

This study was funded by Dasman Diabetes Institute (RA 2022-1971).

Dalia Al-Abdulrazzaq was the principal investigator of the study and was responsible for the inception, planning, and execution of the project as well as managing the data analysis and interpretation of the results. Dalia Al-Abdulrazzaq wrote the manuscript. Mariam Qabazard, Fahad Al-Jasser, Ayed Al-Anizi, Iman Al-Basari, and Fawziya Mandani participated in the execution of the project as well as patients’ data management. Hessa Al-Kandari was responsible for the maintenance of CODeR. She also participated in the execution of the project as well as patients’ data management. The manuscript was approved.

The data that support the findings of this study are available from the Ministry of Health of Kuwait and Dasman Diabetes Institute, but restrictions apply to the availability of these data, which were used under license for the current study and therefore are not publicly available. Data are however available from the corresponding author upon reasonable request and with permission of the Ministry of Health of Kuwait.