Screening Strategy for Islet Autoantibodies in Diabetes Patients of Different Ages

Abstract

Background:

The detection of islet autoantibodies is essential for the accurate classification and diagnosis of diabetes mellitus (DM). The islet autoantibody distribution varies by age. However, screening strategies for DM patients with different onset ages remain lacking.

Materials and Methods:

This cross-sectional study included 17,536 DM patients from 46 medical centers across China. The seroprevalence of glutamic acid decarboxylase autoantibody (GADA), insulinoma-associated-2 autoantibody (IA-2A), zinc transporter 8 autoantibody (ZnT8A), and insulin autoantibody (IAA) was determined in younger and older patients with type 1 DM (T1DM) (n = 287 and 285, respectively), younger and older patients with latent autoimmune diabetes (LAD) (n = 140 and 121, respectively), and younger and older patients with type 2 DM (n = 200 in each group).

Results:

The cutoff age between younger and older patients was 35 years using restricted cubic spline method (n = 17,536, adjusted R ² = 0.97, residual standard error = 1.32; P < 0.001). The seroprevalence rates of four islet autoantibodies were higher in patients aged 15–35 years than in those ≥35 years (GADA: 17% vs. 5.6%, IA-2A: 8.5% vs. 1.3%, ZnT8A: 6.3% vs. 2.3%, IAA: 2.2% vs. 1.0%). The prevalence of ZnT8A was higher in LAD patients than in T1DM patients, especially in older LAD patients. The results indicated that ZnT8A detection can increase the detection rate of older LAD patients from 70.2% (based on GADA detection alone) to 91.7%.

Conclusions:

In patients stratified according to the cutoff age of 35 years, the optimal detection sequence should be GADA, IA-2A, and ZnT8A in younger patients and GADA, ZnT8A, and IA-2A in older patients, so as to reduce the screening cost while improving the detection rate. Particularly, the ZnT8A test is recommended in older patients to avoid a missed LAD diagnosis.

Introduction

Diabetes mellitus (DM) is a multifactorial disease, and accurate classification is essential to protect beta cell function and delay disease progression. “The Consensus Statement on Latent Autoimmune Diabetes in Adults (LADA) management,” drafted by an international expert panel in 2020, emphasized the importance of islet autoantibody detection in the classification and diagnosis of DM.¹ The panel also highlighted several main diabetes-related autoantibodies, including glutamic acid decarboxylase autoantibody (GADA), insulinoma-associated-2 autoantibody (IA-2A), zinc transporter 8 autoantibody (ZnT8A), and insulin autoantibody (IAA), of which GADA is dominant in both Western countries and China.

The GADA positivity rate varies substantially according to ethnicity and region, especially among those first diagnosed with type 2 DM (T2DM).² For instance, among Chinese LADA patients, the GADA positivity rate was observed to be only 67%,³ which contrasts starkly with the >90% rate observed in Caucasian individuals,⁴ underscoring the need for multiple autoantibody screening in China. But multiple autoantibody detection will increase the cost of testing. The reported seropositivity rates of other islet autoantibodies also vary in different populations and may be related to ethnicity as well.⁵

In addition, antibody positivity rates is influenced by age,⁶ and DM subtypes are classified by age at onset.⁷ Therefore, clarifying the antibody detection priorities for DM patients of different ages is crucial to reduce the rate of missed diagnosis and screening cost, especially in China. However, the screening strategies for younger and older DM patients remain to be elucidated.

Type 1 DM (T1DM) is mainly grouped into two subtypes—rapid and slow—according to the progression rate of beta cell destruction. Slowly progressive insulin-dependent DM is also called latent autoimmune diabetes (LAD), “hybrid diabetes,” “type 1.5 diabetes (T1.5D),” and “double diabetes (DD).”^8,9 Regarding onset age, LAD is divided into LAD in older individuals (referred to as LADA) and LAD in younger individuals (referred to as LADY).^8,10

However, the cutoff age between younger and older LAD patients remains an ongoing international debate. Previous studies used different cutoffs, including 18, 25, 30, and 35 years old, and the most widely used is 30 years old, which are recommended by the Immunology of Diabetes Society in 2005.¹¹ However, the use of these cutoffs was always arbitrary, determining the cutoff age based on the seroprevalence of islet autoantibodies would allow formulate accurate autoantibody detection strategies for DM patients of different onset ages.

Here, we attempt to define, for the first time, a cutoff age according to the distribution of islet autoantibodies and clarify the different islet autoantibody detection priorities for DM patients of different ages to optimize the screening cost while increasing the detection rate of autoimmune DM (ADM).

Materials and Methods

Patient selection

This cross-sectional study included 17,536 DM patients aged 15–79 years newly diagnosed with DM from April 2015 to October 2017. All patients from 46 tertiary hospitals in 25 major cities across China were continuously enrolled in the “Diagnosis and Treatment Optimization of Autoimmune Diabetes in Chinese Adults” project based on the National Key R&D Program of China (2013BAI09B12). Research nurses at each participating hospital were trained in standard procedures and data collection methods.¹² The inclusion criteria were (1) onset age ranging from 15 to 79 years, (2) disease duration of <1 year, and (3) treatment at outpatient clinics of the Departments of Endocrinology of participating hospitals.

The exclusion criteria were (1) pregnancy at the time of diagnosis or a diagnosis of a special DM type such as gestational diabetes mellitus, fulminant T1DM, and maturity onset diabetes of the young; (2) acute infectious and traumatic, or other stressful conditions, such as acute myocardial infarction; and (3) tumors or severe diseases. All patients were screened for GADA.

Our study determined the cutoff age to distinguish between younger and older patients according to the GADA test results of 17,536 patients. All younger patients (n = 1851) from the total cohort were included in the study. Sample patients were selected using SPSS version 25.0. A total of 1800 patients matched for age, sex, GADA seropositivity rate, and DM duration were selected from the total cohort (n = 15,685) (Fig. 1) to explore antibody screening strategies for subjects of different ages and DM subtypes. Serum samples were stored at −80°C until analysis. All samples were centrally assayed for GADA, IA-2A, and ZnT8A. IAA assays were performed only in patients who had not used insulin because of the difficulty in distinguishing IAA from insulin antibody upregulated by insulin therapy.

FIG. 1.

Flowchart of patient selection and screening. GADA, glutamic acid decarboxylase autoantibody; IA-2A, insulinoma-associated-2 autoantibody; IAA, insulin autoantibody; LAD, latent autoimmune diabetes; T1DM, type 1 diabetes mellitus; T2DM, type 2 diabetes mellitus; ZnT8A, zinc transporter 8 autoantibody.

According to the islet autoantibody detection results, islet function, clinical characteristics, and age, patients were classified into six groups: younger T1DM (n = 287), younger LAD (n = 140), younger T2DM (n = 200, randomly selected from 1424 patients), older T1DM (n = 285), older LAD (n = 121), and older T2DM (n = 200, randomly selected from 1651 patients) patients. Since only 28 out of 1800 patients aged ≥35 years were diagnosed with T1DM, all T1DM patients with onset age ≥35 years were included in the analysis to ensure that the number of older T1DM patients was comparable with the sizes of the other DM groups.

The diagnostic criteria for T1DM were (1) insulin dependence from the time of diagnosis and (2) meeting at least one of the following conditions according to two independent endocrinologists: fasting C-peptide (FCP) <200 pmol/L, seropositivity for at least one islet autoantibody (GADA, IA-2A, ZnT8A, or IAA), obvious diabetes-related metabolic disorder symptoms, and diabetic ketosis or ketoacidosis. The diagnostic criteria for LAD were (1) being positive for at least one islet autoantibody and (2) not requiring insulin at least 6 months after diagnosis. T2DM patients met the 1999 World Health Organization (WHO) DM diagnostic criteria. A diagnosis of T2DM was excluded if patients were seropositive for islet autoantibodies.

This study was approved by the Research Ethics Committee of The Second Xiangya Hospital of Central South University (Approval No. [2014] Lun Zhen [Division] No. 32), and all participants or their guardians gave written informed consent.

Autoantibody assays

GADA, IA-2A, and ZnT8A were detected in serum samples by radioligand assays (RLAs),¹³ with a respective sensitivity and specificity of 82% and 97.8% for GADA, 76% and 100% for IA-2A, and 72% and 100% for ZnT8A according to the 2016 islet autoantibody standardization program (IASP 2016). The presence of IAA was determined by electrochemiluminescence (ECL) method,¹⁴ with a sensitivity of 52% and specificity of 100%, according to the IASP 2020. Based on the 99th percentile observed in 405 healthy participants, cutoff values were 18 U/mL for GADA and 3.3 U/mL for IA-2A (where U is a WHO unit), and a GADA titer ≥180 U/mL was considered high.² The threshold antibody indices for ZnT8A and IAA was 0.011 and 0.005, respectively.

Biochemical assays

The serum levels of fasting blood sugar (FBS), triglycerides, total cholesterol, high-density lipoprotein cholesterol, and low-density lipoprotein cholesterol were measured on the day of treatment using an automatic chemistry analyzer. Serum FCP and postprandial C-peptide (PCP) levels were measured using a chemiluminescence system (ADVIA Centaur, Siemens, Germany). The intra-assay and interassay variation coefficients of the C-peptide test were 1.0%–3.3% and 3.7%–4.1%, respectively. HbA1c was measured by automated liquid chromatography (Bio-Rad VARIANT-II Hemoglobin Testing System).

Statistical analysis

All statistical analyses were performed using IBM SPSS version 25.0. The normality of the distribution of continuous variables was assessed using Q–Q plots. Normally distributed variables are presented as means ± standard deviations, and non-normally distributed variables are shown as medians (interquartile ranges). Alternatively, some continuous variables were binarized and represented as the number of positive cases, constituent ratio, or ratio. Categorical variables were expressed as percentages (number, n). Continuous variables were compared by analysis of variance (ANOVA) or nonparametric tests, as appropriate. Frequencies were compared using χ² tests or Fisher's exact tests, as appropriate. A two-tailed P-value of <0.05 was considered significant.

Given the nonlinear relationship between age and the GADA positivity rate, a restricted cubic spline (RCS) method was used to visualize this relationship and determine the cutoff age. RCS analysis is currently one of the most common methods for analyzing nonlinear relationships.^15,16 The procedure connects piecewise cubic polynomials across different intervals of a continuous variable.¹⁷ We chose 3 knots with 61 age points to obtain an adequate fit of the model.¹⁸ A set of nodes with the best fit was selected by setting the nodes continuously and comparing the fit of the curve at different nodes according to the residual standard error and adjusted R ². After carefully inspecting the curve, the node with the largest change in slope was used as the cutoff age.¹⁹ The same method was applied to verify the cutoff age for the sample patients. The RCS was plotted and analyzed with R version 4.0.3 for Windows.

Results

Determination of the cutoff age to distinguish between younger and older patients

The correlation between the GADA positivity rate and age (patients aged 75–79 years were grouped together due to the small sample size and large variability, n = 361) was assessed by RCS analysis. The GADA positivity rate initially decreased significantly with age and then tended to stabilize (Fig. 2A). After comparing the fit of the curves at different nodes, the nodes at ages 28, 35, and 55 years were selected. Curve fitting showed that the goodness of fit was highest for this set of nodes (adjusted R ² = 0.97 [95% confidence interval (CI), 0.96–0.98], residual standard error = 1.32, P < 0.001). The cutoff point at which the GADA positivity rate started to stabilize was 35 years (Fig. 2A).

FIG. 2.

RCS analysis for the determination of cutoff age. (A) Determination of cutoff age. (B) Verification of cutoff age. Each dot represents an age point, and the last dot represents the age group 75–79 years. The lines are fitted curves, and the shaded area represents the 95% confidence interval for the association between age and seropositivity for islet autoantibodies. RCS, restricted cubic spline.

In the sample patients, the rates of positivity for any islet autoantibody (patients aged 75–79 were pooled because of the small sample size and large variability, n = 38) were used as dependent variables to draw the RCS. The results showed a similar changing tendency and the same cutoff age (adjusted R ² = 0.88 [95% CI = 0.81–0.94], residual standard error = 3.11, P < 0.001) (Fig. 2B).

Prevalence of islet autoantibodies in patients of different ages

The seropositivity rates for islet autoantibodies were significantly higher in the age group 15–35 years than in the age group ≥35 years (Table 1). The autoantibody with the highest expression level in both age groups was GADA (17% and 5.6% in the younger and older group, respectively). The prevalence of IA-2A was higher than that of ZnT8A in patients aged 15–35 years (8.5% vs. 6.3%, P = 0.008). In contrast, the prevalence of ZnT8A was higher than that of IA-2A in patients aged ≥35 years (2.3% vs. 1.3%, P = 0.033); In addition, the positivity for ZnT8A alone was higher than that for IA-2A in patients aged ≥35 years (1.3% vs. 0.2%, P < 0.001). The seropositivity profiles are shown in Supplementary Figure S1.

Table 1.

The Seropositivity Rates of Islet Autoantibodies in Patients of Different Ages

	<35 years (n = 1851)	≥35 years (n = 1800)	P
GADA	17% (315/1851)	5.6% (101/1800)	<0.001
IA-2A	8.5% (158/1851)	1.3% (24/1800)	<0.001
ZnT8A	6.3% (116/1851)	2.3% (41/1800)	<0.001
IAA	2.2% (17/786)	1.0% (10/1011)	0.042

Data are shown as percentages (positive cases/total cases). IAA was screened only in patients who had not used insulin.

GADA, glutamic acid decarboxylase autoantibody; IA-2A, insulinoma-associated-2 autoantibody; IAA, insulin autoantibody; ZnT8A, zinc transporter 8 autoantibody.

Prevalence of islet autoantibodies in patients of different DM subtype

Among the four ADM subtypes, the seropositivity rate of GADA was dominant, with the highest rate observed in younger LAD patients (83.6%). For IA-2A, the seroprevalence rate was higher than that of ZnT8A in younger T1DM patients (40.4% vs. 25.1%, P < 0.001) and higher in these patients than in patients with other DM subtypes (P < 0.05). ZnT8A was positive in a larger proportion of LAD patients than of T1DM patients (31.4% and 32.2% in younger and older LAD patients, respectively), and ZnT8A positivity was higher than that of IA-2A (14%) in older LAD patients (P = 0.001) (Supplementary Table S1).

Furthermore, the seropositivity of ZnT8A alone (the other islet autoantibodies were negative) was not only higher than that of IA-2A alone in LAD patients (younger LAD: 8.6% vs. 0.2%, P = 0.017; older LAD: 19% vs. 1.7%, P < 0.001) but also higher than that of ZnT8A alone in T1DM patients (Table 2). In older subjects with LAD, when patients with concurrent GADA and ZnT8A positivity were excluded, the positivity of ZnT8A was 21.5% (26/121), indicating that the additional detection of ZnT8A increased the detection rate of older LAD from 70.2% (when only GADA was tested, Supplementary Table S1) to 91.7%. The positivity for two or more autoantibodies was higher in younger individuals (36.0% and 37.9% in younger T1DM and LAD patients, respectively) than in older subjects (21.1% and 20.7% in older T1DM and LAD patients, respectively).

Table 2.

Single Seroprevalence Rates for Islet Autoantibodies in Four Autoimmune Diabetes Subtypes

	Younger T1DM (n = 287)	Younger LAD (n = 140)	Older T1DM (n = 285)	Older LAD (n = 121)
GADA	28.9% (84/287)	47.1% (66/140)^***	42.1% (120/285)^***	52.1% (63/121)^***
IA-2A	2.4% (7/287)	0.2% (3/140)	1.4% (4/285)	1.7% (2/121)
ZnT8A	0.7% (2/287)	8.6% (12/140)^***	1.1% (3/285)^†††	19% (23/121)^*** ^{††‡‡‡}
IAA	3.7% (1/27)	9.5% (6/63)	0% (0/38)	13.6% (8/59)

Data are percentages (single positive cases/total cases). IAA was screened only in patients who had not used insulin.

***

P < 0.001 versus younger T1DM patients; ^†† P < 0.01, ^††† P < 0.001 versus younger LAD patients; ^‡‡‡ P < 0.001 versus older T1DM patients.

LAD, latent autoimmune diabetes; T1DM, type 1 diabetes mellitus.

Variations in clinical characteristics among DM subtypes

Age, body mass index (BMI), and the levels of FBS, FCP, PCP, HbA1c, and serum lipids were compared among six DM subtypes to assess the rationality of the cutoff age (Supplementary Table S2). The average age was lower in younger T1DM than in the younger LAD group (P < 0.05). The FCP levels were 106.6 pmol/L (95% CI: 42.7–193.2), 287.4 pmol/L (95% CI: 170.0–489.9), and 485 pmol/L (95% CI: 295.4–740.5) in younger T1DM, younger LAD, and older LAD patients, respectively (P < 0.01 for all pairwise comparisons), and PCP exhibited a similar trend. HbA1c levels had an opposite trend. FCP and PCP levels and BMI were lower in LAD patients than in the T2DM group regardless of age.

Discussion

A total of 17,536 patients were selected from 46 tertiary care hospitals in 25 cities across China; thus, these findings are strongly representative of subtypes and clinical characteristics of DM patients nationwide. Four islet autoantibodies were detected, greatly reducing the rate of missed diagnosis and misdiagnosis. Although multiple autoantibody screening is of great significance in decreasing the missed diagnosis rate of ADM, it will also raise the screening cost. This study examined differences in islet antibody profiles among age groups and DM subtypes to perform a more economical screening and differential diagnosis in the clinic. Although patients younger than 15 years of age were not included in the analysis, the GADA positivity rate seems to remain stable until the age of 10–15 years.^20,21

Higher islet autoantibody seropositivity rates are strongly associated with younger age.²² Our results are consistent with previous findings, and the prevalence rates of the four islet autoantibodies were higher in patients aged 15–35 years than in those aged ≥35 years. Similarly, the positivity rates for multiple islet autoantibodies were higher in younger patients; the difference in the age-related distribution between younger and older patients was similar to that in a small European study,²³ indicating that autoimmunity was more common in younger individuals than in older individuals. Moreover, the hierarchy of autoantibody seroconversions has been demonstrated to be age related.²⁴

In our study, we compared all types of patients and found that the rank order of seropositivity frequency differed between younger and older patients. In younger patients, GADA had the highest seropositivity, followed by IA-2A and ZnT8A, whereas in older patients, GADA had the highest seropositivity, followed by ZnT8A and IA-2A. (Due to the low number of patients tested for IAA, which may have increased the inaccuracy of the results, IAA is not included in the discussion.)

This finding suggests that it would be helpful to adapt the islet autoantibody screening strategy according to onset age to reduce screening costs while increasing detection rate of ADM. Moreover, LAD patients exhibited an absolute advantage in terms of the positivity rate for ZnT8A compared with that of IA-2A, especially in older patients. Compared with the detection of GADA alone, the presence of ZnT8A can increase the detection rate of older LAD by more than 1/4, demonstrating that GADA testing should be followed by ZnT8A screening in older patients to distinguish more LAD patients from T2DM patients. Similarly, ZnT8A was found to have considerable diagnostic value, second only to GADA, in older Argentinian LAD patients.²⁵

We performed a comparison of the autoantibody distributions between T1DM and LAD patients. Our results showed that younger and older T1DM patients shared a similar autoantibody pattern, which is consistent with the findings in a European study.²³ In contrast to T1DM patients, the autoantibody pattern differed between younger and older LAD patients, with GADA > IA-2A = ZnT8A in younger LAD patients and GADA > ZnT8A > IA-2A in older LAD patients. These results suggest that younger LAD patients may represent a DM type different from T1DM and older LAD patients.

The seroprevalence rates of ZnT8A alone were significantly higher in LAD patients (especially in older patients) than those in T1DM patients. The results demonstrated that ZnT8A is related to the slow progression of DM and older age; in other words, older LAD patients have a higher frequency of positivity for ZnT8A.²⁵ In T1DM or LAD patients, the positivity for two or more islet autoantibodies was higher in the younger groups. Moreover, the rates were lower in LAD than in T1DM irrespective of age. These findings suggest a greater expansion of antibody epitopes among related molecules in T1DM patients, particularly in younger patients.²⁶ Moreover, the number of positive islet antibodies in LAD patients reflects the intensity of the autoimmune response to islet cell damage.²⁷

We also compared clinical characteristics among DM subtypes grouped according to the cutoff age. Age, BMI, islet function, and plasma glucose control in younger LAD patients were all intermediate between younger T1DM and older LAD patients. Moreover, younger LAD patients in this study had lower weight, worse islet function, and better blood lipid status than younger T2DM patients. Therefore, at the cutoff age of 35 years, not only did the islet autoantibody profile differ, but the clinical features among the subgroups of DM patients also varied.

In addition, previous studies have suggested that the use of truncated GADA with radiobinding assays (RBAs) and ECL are more disease-specific for ADM^28,29; however, RLAs were used in the present study. Thus, we will verify our results with the detection method of RBAs and ECL and explore the disease specificity of positivity for ZnT8A alone in LAD patients to more effectively distinguish LAD from T2DM.

In conclusion, the results of this study were based on a nationally representative sample and found that the optimal cutoff to screen Chinese DM patients for islet autoantibodies was 35 years old. To increase the detection rate of ADM, and taking into account the testing costs, it appears that different screening strategies for patients of different ages would be valuable: for patients 15–35 years old, the detection sequence should be GADA, IA-2A, and ZnT8A, while for patients ≥35 years old, the detection sequence should be GADA, ZnT8A, and IA-2A. Notably, ZnT8A screening is recommended in older patients to avoid missing the diagnosis of LAD.

The major limitation of this study is the cross sectional design. Therefore, the findings are tentative, and the proposed screening strategy should be tested in different populations before being recommended for routine practice. Moreover, follow-up studies on beta cell function and treatment in autoantibody-positive subjects and controls are needed to further support our islet autoantibody screening strategies and determine that the strategy will lead to different clinical management.

Footnotes

Authors' Contributions

X.N. designed the study, assayed autoantibodies, and analyzed data, software, and drafted the article; X.L. and Y.X. designed the study and contributed to the data analysis and discussion and edited the article; X.Y. and H.Z. collected the data and serum samples and assayed autoantibodies and contributed to project administration; X.T., J.C., X.N., J.L., Q.J., and L.J. collected data and serum samples and contributed to project administration; G.H. and Z.Z. designed the study, contributed to funding acquisition, project administration, data analysis, discussion, and edited and reviewed the article. All authors have read and approved the final article. G.H. is the guarantor of this work and had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Acknowledgments

The authors thank all the patients, nurses, doctors, investigators, and technicians involved at the 46 participating centers of National Clinical Research Center for Metabolic Diseases for their efforts in data and sample collection (see the for more information). We also thank Prof. Guang Ning for his kind help in providing the population census data of the general population in China in 2010. We thank Prof. Xilin Yang for recommending us the method of RCS to determine the cutoff age. We thank Prof. Bill Hagopian for providing the recombinant hGAD65 and IA-2 plasmid and thank Prof. John C. Hutton for providing the recombinant ZnT8 plasmid. We also thank Prof. Yu Liping for technical support of IAA test based on ECL.

Author Disclosure Statement

No competing financial interests exist.

Data and Resource Availability

The datasets generated during the current study are available from the corresponding author upon reasonable request.

Funding Information

This study was supported by the Chinese National Key Research and Development Project (2018YFC1315600), the National Research and Development Program of China (2018YFC2001005), the National Key R&D Program of China (2018YFC1315603, 2013BAI09B12, 2016YFC1305001), the National Natural Science Foundation of China (81820108007), Science and Technology Major Project of Hunan Province (2017SK1020), and National Science and Technology Major Project (2020ZX09201-028). The funders had no role in study design, data collection, data analysis, interpretation, writing of the report, and decision to submit the article for publication.

Supplementary Material

Supplementary Figure S1

Supplementary Table S1

Supplementary Table S2

References

Buzzetti

, Tuomi

, Mauricio

, et al.: Management of latent autoimmune diabetes in adults: a consensus statement from an international expert panel. Diabetes, 2020; 69:2037–2047.

Zhou

, Xiang

, Ji

, et al.: Frequency, immunogenetics, and clinical characteristics of latent autoimmune diabetes in China (LADA China study): a nationwide, multicenter, clinic-based cross-sectional study. Diabetes, 2013; 62:543–550.

Xiang

, Huang

, Shan

, et al.: Glutamic acid decarboxylase autoantibodies are dominant but insufficient to identify most Chinese with adult-onset non-insulin requiring autoimmune diabetes: LADA China study 5. Acta Diabetol, 2015; 52:1121–1127.

Hawa

, Kolb

, Schloot

, et al.: Adult-onset autoimmune diabetes in Europe is prevalent with a broad clinical phenotype: action LADA 7. Diabetes Care, 2013; 36:908–913.

Yang

, Luo

, Huang

, et al.: The diagnostic value of zinc transporter 8 autoantibody (ZnT8A) for type 1 diabetes in Chinese. Diabetes Metab Res Rev, 2010; 26:579–584.

Williams

, Long

: What has zinc transporter 8 autoimmunity taught us about type 1 diabetes?. Diabetologia, 2019; 62:1969–1976.

Tuomi

, Santoro

, Caprio

, et al.: The many faces of diabetes: a disease with increasing heterogeneity. Lancet, 2014; 383:1084–1094.

Pozzilli

, Guglielmi

, Caprio

, Buzzetti

: Obesity, autoimmunity, and double diabetes in youth. Diabetes Care, 2011; 34(Suppl 2):S166–S170.

Leslie

, Palmer

, Schloot

, Lernmark

: Diabetes at the crossroads: relevance of disease classification to pathophysiology and treatment. Diabetologia, 2016; 59:13–20.

10.

Zachariah

, Sharfi

, Nussey

, Bano

: Latent autoimmune diabetes in the young. Clin Med (London, England), 2008; 8:552–553.

11.

Fourlanos

, Dotta

, Greenbaum

, et al.: Latent autoimmune diabetes in adults (LADA) should be less latent. Diabetologia, 2005; 48:2206–2212.

12.

Tang

, Yan

, Zhou

, et al.: Prevalence and identification of type 1 diabetes in Chinese adults with newly diagnosed diabetes. Diabetes Metab Syndr Obes, 2019; 12:1527–1541.

13.

Huang

, Xiang

, Pan

, et al. Zinc transporter 8 autoantibody (ZnT8A) could help differentiate latent autoimmune diabetes in adults (LADA) from phenotypic type 2 diabetes mellitus. Diabetes Metab Res Rev, 2013; 29:363–368.

14.

Wang

, Li

, Yan

, et al.: [A novel electrochemiluminescence method for detecting insulin autoantibody]. Zhonghua yi Xue za Zhi, 2020; 100:424–429.

15.

, He

, Li

, et al.: Predictive value of fasting glucose, postload glucose, and hemoglobin A1c on risk of diabetes and complications in Chinese adults. Diabetes Care, 2019; 42:1539–1548.

16.

Bhaskaran

, dos-Santos-Silva

, Leon

, et al.: Association of BMI with overall and cause-specific mortality: a population-based cohort study of 3.6 million adults in the UK. Lancet Diabetes Endocrinol, 2018; 6:944–953.

17.

Leng

, Zhang

, Wang

, et al.: Plasma levels of alanine aminotransferase in the first trimester identify high risk Chinese women for gestational diabetes. Sci Rep, 2016; 6:27291.

18.

Harrell

: Regression Modelling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis. New York, Spinger-Verlag, 2001.

19.

, Huo

, Cao

, et al.: Bile acid metabolites in early pregnancy and risk of gestational diabetes in Chinese women: a nested case-control study. EBioMedicine, 2018; 35:317–324.

20.

Pollanen

, Ryhanen

, Toppari

, et al.: Dynamics of islet autoantibodies during prospective follow-up from birth to age 15 years. J Clin Endocrinol Metab, 2020; 105: e4638–e4651.

21.

Bell

, Mayer-Davis

, Beyer

, et al.: Diabetes in non-Hispanic white youth: prevalence, incidence, and clinical characteristics: the SEARCH for Diabetes in Youth Study. Diabetes Care, 2009; 32(Suppl 2):S102–S111.

22.

Steck

, Johnson

, Barriga

, et al.: Age of islet autoantibody appearance and mean levels of insulin, but not GAD or IA-2 autoantibodies, predict age of diagnosis of type 1 diabetes: diabetes autoimmunity study in the young. Diabetes Care, 2011; 34:1397–1399.

23.

Rogowicz-Frontczak

, Pilacinski

, Wyka

, et al.: Zinc transporter 8 autoantibodies (ZnT8-ab) are associated with higher prevalence of multiple diabetes-related autoantibodies in adults with type 1 diabetes. Diabetes Res Clin Pract, 2018; 146:313–320.

24.

Vehik

, Bonifacio

, Lernmark

, et al.: Hierarchical order of distinct autoantibody spreading and progression to type 1 diabetes in the TEDDY study. Diabetes Care, 2020; 43:2066–2073.

25.

Yohena

, Penas-Steinhardt

, Muller

, et al.: Immunological and clinical characteristics of latent autoimmune diabetes in the elderly. Diabetes Metab Res Rev, 2019; 35:e3137.

26.

Claessens

, Wesselius

, van Lummel

, et al.: Clinical and genetic correlates of islet-autoimmune signatures in juvenile-onset type 1 diabetes. Diabetologia, 2020; 63:351–361.

27.

Nishimura

, Matsumura

, Kikuno

, et al.: Slowly progressive type 1 diabetes mellitus: current knowledge and future perspectives. Diabetes Metab Syndr Obes, 2019; 12:2461–2477.

28.

Achenbach

, Hawa

, Krause

, et al.: Autoantibodies to N-terminally truncated GAD improve clinical phenotyping of individuals with adult-onset diabetes: action LADA 12. Diabetologia, 2018; 61:1644–1649.

29.

Zhu

, Qian

, Liu

, et al.: Glutamic acid decarboxylase autoantibody detection by electrochemiluminescence assay identifies latent autoimmune diabetes in adults with poor islet function. Diabetes Metab J, 2020; 44:260–266.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.18 MB

0.02 MB