Latent Autoimmune Diabetes in Adults (LADA) is a slowly progressive autoimmune form of diabetes presenting in adults. There is still debate regarding its classification and whether it should be considered a slowly progressing form of type 1 (T1) diabetes mellitus (DM) or a distinct type of DM (1). A recent consensus statement from the American Diabetes Association and the European Association for the Study of Diabetes (EASD) defines age of onset (>30 years), presence of diabetes associated autoantibodies and absence of insulin requirements for at least 6 months after diagnosis as key diagnostic criteria for LADA (2). Also, they recommend an individualized approach that differs from the classical management of T1DM.

It has been shown that individuals with LADA have a distinct metabolic profile compared to patients with T1DM and type 2 (T2) DM, with higher C-peptide concentrations than subjects with T1DM and a lower frequency of metabolic syndrome than people with T2DM (3). In addition, there is some evidence that LADA is associated with a genetic background that is different from T1DM and T2DM, with an attenuated influence of key T1DM-associated HLA haplotypes compared to T1DM and an increased presence of PTPN22, INS and SH2B3 polymorphisms, compared to T2DM (4).

The PTPN22 gene encodes a protein tyrosine phosphatase, a negative regulator for T cell activation, and a down-regulator of the immune response. The first study on the association between PTPN22 gene variants with T1DM was reported by Bottini et al (5) in 2004. Among the single nucleotide polymorphisms (SNPs) in the gene, rs2476601 has been linked to T1DM (4, 6). The INS gene encodes the (pro) insulin protein. Variants in the gene or its promotor region might affect its thymic expression, leading to decreased tolerance and a higher likelihood of developing an autoimmune response (7).

In the present study, we evaluated differences in the HLA region, as well as in PTPN22 and INS, in patients diagnosed with LADA compared to patients diagnosed with T1DM and T2DM.

A total of 578 participants were included: 248 with T1DM (178 diagnosed below 30 years of age and 70 above that age), 256 with T2DM and 74 with LADA. Their main features are displayed in Table 2.

Our results show genetic differences between T1DM and LADA in all the analyzed genes (HLA, PTPN22 and INS), but only in patients with T1DM diagnosed before 30 years of age. However, no significant differences were found between LADA and the group with T1DM diagnosed after the age of 30. When compared with T2DM, LADA patients had similar or higher frequencies of T1DM risk alleles.

The role of high risk alleles in the age of diagnosis of autoimmune diabetes has been previously studied by other authors, suggesting an earlier onset in those patients with the highest risk alleles (13), thus establishing a continuum from the pediatric age to adulthood. Still, environmental factors can alter this “continuum”, favoring earlier onset in those patients with lower genetic risk exposed to environmental triggers (12) such as viral infections, nutritional practices or increased BMI (14). In this context, LADA seems to be one of the lowest risk forms of autoimmune diabetes, presenting at a later age. Still, it is worth mentioning that LADA also shows a “spectrum” of presentation, with age of onset having been associated with the genetic risk (15) or GAD antibody titers (16). As Desai et al. phrased it, “supporting the hypothesis that autoimmune diabetes occurring in adults is an age-related extension of the pathophysiological process presenting as childhood-onset T1DM” (15).

To our knowledge, this is one of the first genetic studies to directly compare LADA and T1DM diagnosed after the age of 30. Indeed, previous genetic studies comparing LADA and T1DM are scarce overall. Other strengths of this study include the relatively large population and the adjustment for potential confounders (age and BMI).

As genetic analysis has become more accessible to medical and research centers around the world, the number of studies evaluating the genetic differences between the subtypes of DM has increased in recent years, providing novel insights into the pathogenic routes of the disease. In 2018, Cousminer et al (4) published the first Genome Wide Association Study (GWAS) in subjects with LADA (see Table 5), whom they compared not only to healthy controls, but also to patients diagnosed with T1DM and T2DM. They showed mixed genetic influence, sharing risk loci with both T1DM and T2DM. Compared to T1DM, the only significant differences were found in the HLA region, with LADA showing lower frequencies of high T1DM risk alleles. Compared to T2DM, LADA patients showed an increased frequency of T1DM risk alleles in HLA, INS (rs689), PTPN22 (rs2476601) and SHB3 (rs3184504). Similar results were published by Buzzetti et al. in 2007 (23), reporting higher frequencies of moderate and high risk class II HLA haplotypes in subjects diagnosed with LADA compared with patients diagnosed with T2DM.

In the present study, the patients diagnosed with T1DM overall also showed a higher frequency of the HLA risk haplotype DRB1*04-DQB1*03:02-DQA1*03:01 compared to LADA. Furthermore, the absence of risk alleles was more frequent in patients diagnosed with LADA than in patients diagnosed with T1DM, in agreement with Cousminer et al. (4). Indeed, the authors conclude that the leading genome-wide signals support the concept of LADA being a late-onset T1DM. Kisand et al. (20) also reported differences in HLA risk alleles between T1DM (mean age 22 years) and LADA in an Estonian population, i.e. an increased prevalence of T1DM HLA protective haplotypes in patients with LADA. This finding is in accordance with earlier studies suggesting that the increased frequency of protective HLA genotypes is a specific feature of patients with slowly-progressing T1DM. We did not find significant differences in genotypes when comparing LADA and T1DM with onset after 30 years of age, although given the low number of patients in the >30 years age group compared to the overall T1DM population we cannot rule out the existence of differences. Indeed, other authors have reported differences between these two groups (18, 24).

The presence of the T allele in rs2476601 in PTPN22 has been reported to increase the risk of T1DM (20, 26) and LADA (17, 25) by some authors, though it is not a generalized finding in the latter (20). Studies focused on evaluating the differences between its frequency in patients with T1DM and patients with LADA are scarce. We found the risk variant more frequently in patients diagnosed with T1DM before de age of 30 years than in those with LADA, even though the result was statistically borderline (p=0.045). No differences were found on those diagnosed after the age of 30 (see Table 4). Authors have reported an increased presence of the risk (T) allele in both patients with T1DM and LADA when compared to healthy controls (17). Dong et al. reported a similar effect of the PTPN22 gene (rs2476601) in their 2014 meta-analysis, when comparing patients diagnosed with LADA with healthy controls. They found an increased risk for LADA in carriers of the risk allele (21). Similarly, this T allele has also been shown to confer an increased risk of T1DM (5, 25–27).

On the other hand, Kisand et al. (20) reported no association between PTPN22 genotypes and LADA susceptibility, but did find an increased risk for T1DM associated to rs2476601, when comparing both types of DM with controls. Interestingly, there was a 30-year difference in the mean age of diagnosis between the groups. In their 2019 meta-analysis, Ramu et al. (22) also compared subjects with LADA and healthy controls, and reported a significant association for the risk allele in rs2476601 with LADA in the codominant, dominant and recessive models.

Regarding the INS rs689 SNP, we also found a different genotypic distribution among our patients, with a reduced frequency of the (T) minor allele in patients diagnosed with T1DM when compared with LADA or T2DM, though no significant difference was seen between LADA and T1DM patients with an onset of >30 years (see Table 4). Kisand et al. (19) reported similar findings after comparing their patients with T1DM, T2DM and LADA. The T allele was associated with a reduced risk of T1DM, but was not associated with the risk of LADA. In their meta-analysis, Ramu et al. (22) also showed an inverse association between LADA and the presence of the T allele in the codominant, dominant and recessive models.

We acknowledge this study has some limitations. Even though our study is one of the largest published to date comparing genetic differences between T1DM and LADA, the relatively small sample size could limit our ability to detect minor differences, especially in the group with T1DM diagnosed after the age of 30 years. For unknown reasons, the number of male participants exceeds that of females. This may have influenced some of the results. Also, we limited our study to three of the main genes related to T1DM risk. More extensive studies assessing polygenic risk scores, would be interesting to perform.

To summarize, we present one of the few studies to directly compare differences in the genetic profile of patients with T1DM and LADA. We believe this differentiation is important because the underlying autoimmune process against the β-cell can have consequences for the prognosis, complications and treatment of patients (24). We report a similar T1DM risk allele distribution in patients with LADA compared to individuals with T1DM diagnosed after the age of 30, but less risk alleles than in those diagnosed before the age of 30 years. Our results suggest that, from a genetic point of view, LADA holds genetic similarities to late-onset forms of T1DM. Our study supports the importance of genetic studies in the characterization of LADA, but also of comparing the study group with appropriate controls, although the confidence in the results is limited by sample size.

The original contributions presented in the study are included in the article/supplementary files. Further inquiries can be directed to the corresponding authors.

The studies involving human participants were reviewed and approved by Ethics Committees of Hospital de la Santa Creu i Sant Pau, Barcelona and Hospital Universitari Arnau de Vilanova, Lleida. The patients/participants provided their written informed consent to participate in this study.

MH and DM were responsible for study conception and design, RF, EP, AE, EC and MH participated in data acquisition. RF and EP performed data analysis. AM and YN-M drafted the first version of the manuscript. MH, DM, AM and YN-M made subsequent revisions to the manuscript. All authors revised and approved the final version of the paper.

This project was supported by a grant from Instituto de Salud Carlos III, Spain (Project FIS 061104).

The authors thank Associació Catalana de Diabetis (ACD) for providing the samples and clinical information of the patients included in the register. We acknowledge Joan Verdaguer, from the University of Lleida (UdL), for logistic support with the genetic studies.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.