The first large-scale studies of the prediction of T1DM relied upon the detection of cytoplasmic islet cell autoantibodies (ICA) assays based on indirect immunofluorescence. High titer cytoplasmic ICA is most often associated with the presence of multiple islet autoantibodies and therefore a high risk of progression to diabetes . Screening for risk of T1DM now utilizes “biochemical” autoantibody assays for specific islet autoantigens. These include autoantibodies to insulin (IAA), glutamic acid decarboxylase (GAD), IA-2 (ICA512) and most recently ZnT8 . Individuals having a single positive autoantibody (insulin, GAD, IA-2, or ZnT8 autoantibodies) are at low risk for progression to T1DM. Single positive autoantibodies can be a non-reproducible false positive result (e.g. switched sample), transient, or represent the presence of an autoantibody reacting with the specific autoantigen that does not confer increased risk of diabetes (e.g. low affinity insulin autoantibodies) . Individuals expressing two or more positive autoantibodies, especially on multiple tests over time, are at very high risk of progressing to diabetes. This high risk may result from autoimmunity spreading to other autoantigens whose targeting increases destruction or from the high statistical specificity of expression of multiple autoantibodies. The Diabetes Autoantibody Standardization Program (DASP) workshop aims to improve and standardize measurement of autoantibodies associated with T1DM across laboratories .
Different islet autoantibodies have been associated with different risks of progression, with IA-2 autoantibodies most often associated with expression of other biochemical autoantibodies and high risk . Of note, insulin autoantibodies are extremely high at the onset of diabetes in young children while usually negative in individuals first presenting with diabetes after age 12. There is a log-linear inverse relationship between the levels of insulin autoantibodies and the age of onset of diabetes . In DAISY, 89% of children who progressed to diabetes expressed ≥2 autoantibodies with cumulative incidence of 74% by age 10 for those expressing three autoantibodies (Figure 2.3). In children expressing >2 autoantibodies, there is no significant difference in progression to diabetes between relatives and general population subjects.
Autoantibody screening among relatives
The initial large screening studies found approximately 3% of first-degree relatives to be positive for ICA.With analysis of biochemical autoantibodies, it has become evident that the presence of ICA in the absence of GAD65 or ICA512 autoantibodies is associated with a low risk of progression to diabetes [26,53]. The cumulative risk of developing diabetes within 15 years is only 2.8% for individuals with ICA but without GAD or ICA512 (IA-2) autoantibodies versus 66% for those with ICA and either or both GAD or ICA512 autoantibodies. As in many studies, in the DAISY study, the incidence of islet autoantibodies is much higher in first-degree relatives,
compared to the general population. The risk by age 10 is particularly high in the HLA-DR3/4,DQB1*0302 positive siblings (43%) and offspring (34%) and moderate-risk siblings (19%) . In the DAISY children without a first-degree relative with T1DM, the incidence of persistent islet autoantibodies by the age of 9 years is 10.6%, 5.5% and 3.4% in, respectively, high-,moderate- and average-HLA risk groups. The HLA genotype and having a diabetic relative, but not gender or ethnicity, predict development of islet autoantibodies.
Islet autoantibodies appear early in life. The BABYDIAB, DIPP, and DAISY studies have demonstrated that a significant proportion of first-degree relatives progressing to T1DM before age 15 develop islet autoantibodies before their 2nd birthday [31,55,56]. Although IAA usually appear first, a significant percentage of children followed from birth initially express GAD65 autoantibodies, while IA-2 and ZnT8 autoantibodies usually develop later.
Autoantibody screening in the general population
Islet autoantibody screening studies have included mainly first-degree relatives; however, 90% of T1DM cases occur in individuals with no family history of T1DM. Several ongoing studies have followed children from birth for the development of islet autoantibodies.The two studies with the longest follow-up are the DAISY study from Denver Colorado and the DIPP study in Finland . More recently, the multicenter international TEDDY study (The Environmental Determinants of Diabetes in the Young)  has enrolled over 8600 high-risk general population children identified by newborn screening into a prospective follow-up for islet autoantibody measurements from birth.
While the risk of developing islet autoantibodies is 3–4 times higher in relatives than in the general population with the same HLA-DR/DQ genotypes, the persistent presence of islet autoantibodies portends similar high risk of progression to diabetes for both relatives and the general population [52,59]. Higher titer and higher affinity of the autoantibodies as well as presence of autoantibodies to multiple autoantigens predict higher risk of diabetes .
Measurement of autoantibodies in adult-onset diabetes
After the age of about 15 years, measuring the incidence of T1DM is complicated by misclassification of some patients as type 2 diabetics. It has been estimated that at least 37% of T1DM is diagnosed after the age of 19 years and 15% after the age of 30 years . New cases of T1DM presenting in adult life tend to have a longer duration of symptoms before diagnosis and higher C-peptide levels remaining at diagnosis compared with those presenting in childhood, suggesting a slower rate of β-cell destruction. However, C-peptide levels in islet autoantibody positive adult diabetic patients are still significantly lower than in those with type 2 diabetes. Assays for GAD65 autoantibodies have been the most useful in identifying latent autoimmune diabetes in adults (LADA).
Between 5 and 10% of patients with a diagnosis of gestational diabetes have positive islet autoantibodies and the great majority progress to T1DM . Among adults diagnosed with type 2 diabetes and participating in the UKPDS, the proportion of patients with ICA and GAD65 decreased with increasing age at diagnosis: from 21% in patients aged 25–34 to 4% in those aged 55–65 for ICA and from 34% to 7% for GAD65  Most (94%) of patients with ICA and 84% of those with GAD65 required insulin therapy by 6 years, compared with 14% of those without the antibodies.
Measurement of beta-cell function
Direct measurement of the functional mass of islet cells is currently not possible. The first-phase insulin response (FPIR) to intravenous glucose can help predict progression to diabetes among individuals with positive islet autoantibodies [64,65]. The FPIR is usually calculated as the sum of the 1- and 3-min insulin levels measured after glucose is administered intravenously at 0.5 g per kg body weight. Low FPIR has been defined as below the 1st, 5th or 10th percentile of the distribution in nonobese healthy subjects. Subjects from the Joslin family study with FPIR below the first percentile on the first test had a 3-year diabetes-free survival rate of 13% compared with 78% for the group with higher FPIR . The FPIR is severely depressed at the time of detection of islet autoantibodies in many young children. In the DPT-1 study, loss of the FPIR was strongly associated with diabetes ; the average time from the fall of FPIR below the 1st percentile to the onset of diabetes was 1.8 years. A recent study from the Belgian registry reported similar results relative to first-phase insulin secretion with hyperglycemic clamps amongst individuals with IA-2 autoantibodies .
The oral glucose tolerance test (OGTT), performed in clinical trials of T1DM prevention largely to formalize the diagnosis of clinical diabetes, has long been known to have some value in predicting progression to T1DM among subjects with islet autoantibodies . DPT-1 reported that a risk score based on age, BMI, and OGTT indexes (glucose and C-peptide values), without use of IVGTTs or additional autoantibodies, accurately predicted T1DM in ICA-positive relatives .The likelihood of progression to diabetes increased with mild fasting or post oral glucose-load dysglycemia.
In 1993, Leech and coworkers reported that HbA1c measured by high-performance liquid chromatography may be slightly but significantly higher in nondiabetic ICA-positive teenagers compared to ICA negative controls . More recently, the DAISY study has demonstrated that Hb1Ac steadily increases within the normal range over a few years preceding diabetes onset and may therefore be useful in early detection of T1DM. Among children who have persistent islet autoimmunity, increase in HbA1c predicted increased risk of progression to T1DM, with a hazard ratio of 4.8 for each 0.4% (SD) increase in HbA1c, independent of random glucose and number of autoantibodies.
Recently, increased HbA1c level has been added by the joint American Diabetes Association (ADA), International Diabetes Federation, and European Association for the Study of Diabetes International Expert Committee (IEC) as a diagnostic tool for diabetes with a recommended threshold of 6.5% for diagnosis of diabetes on two tests . This threshold was established based on research conducted in adults with type 2 diabetes. However, studies in adolescents at high risk for T1DM suggest that HbA1c>6.5% is a specific but not sensitive early indicator for T1DM .
Through the Diabetes Complications and Control Trial (DCCT), it has been shown that patients who have sustained production of C-peptide have lower rates of severe hypoglycemia, microalbuminuria, and retinopathy [74,75]. In the Diabetes Prevention Trial-Type 1 (DPT-1), post challenge C-peptide levels begin to decrease appreciably in the 6 months before diagnosis and continue to decrease within 3 months after diagnosis . Recent data from the TrialNet study show a biphasic decline in C-peptide during the first 2 years post diagnosis, with a higher rate of fall during the first year .