target="_blank" rel="nofollow" href="#fb3_img_img_dbe73254-a9cb-598d-aaae-be877cad55c9.jpg" alt=""/> no difference in BMI, waist:hip ratio, lipid profile, and BP between those with LADA and those with type 1 diabetes
lower BMI, waist:hip ratio, lipids, and BP in patients with LADA and type 1 diabetes than in those with type 2 diabetes
similar baseline C-peptide levels in patients with LADA and type 1 diabetes but a more rapid decline with type 1 diabetes
similar prevalence of HLA haplotypes associated with high risk for diabetes in LADA and type 1 diabetes
LADA more commonly associated with presence of a single islet-specific antibody compared to type 1 diabetes
Study of 48 children with type 2 diabetes (13) found
8% ICA512 (fragment of ICA) positive; 30% GADA positive, 35% IAA positive
no correlation of antibody positivity with degree of obesity
thyroid autoimmunity in subjects with islet cell autoimmunity
Study of 37 African-American children and adolescents with type 2 diabetes (14) found
10.8% positive for GADA, IA-2, or both
no difference in treatment requirements (oral agent vs. insulin) between positive and negative patients
The accelerator hypothesis has been proposed to explain the development of diabetes-related autoimmunity in typical type 2 diabetes as the result of hyperglycemia secondary to insulin resistance inducing β-cell apoptosis (glucotoxicity) with the development of β-cell autoimmunity (15). Because of the high frequency of evidence of islet cell autoimmunity in otherwise typical type 2 diabetes, particularly in young people, ICA and GADA testing may be worthwhile in all pediatric patients considered to have type 2 diabetes.
Antibodies will indicate an earlier need for insulin.
Antibodies will indicate the need to check for thyroid autoimmunity and to consider other associated autoimmune disorders.
GADA may be the more important predictor of insulin therapy over the short term (3 years) (11).
The characteristic features of the main forms of diabetes seen in children and adolescents are summarized in Table 7.
Uncertainties of Classification
Unfortunately, the distinctions indicated by the features noted above are not as certain as we would like. The clinician is obliged to weigh the evidence in an individual patient for distinguishing between type 1 and type 2 diabetes. There are several reasons for this conundrum.
TABLE 7. Classification of the Types of Diabetes Seen in Children
With increasing obesity in childhood, as many as 20–25% of newly diagnosed type 1 diabetes patients may be obese (16). This will also be true with ADM.
Type 2 diabetes is common in the general population, with a random family history likelihood of ∼15% or greater in minority populations, reducing the specificity of this characteristic.
Positive family history for type 2 diabetes is increased for patients with type 1 diabetes as much as threefold over that of the nondiabetic population, and type 1 diabetes is more frequent in relatives of patients with type 2 diabetes (17).
Genetic interaction between type 1 diabetes and type 2 diabetes is further suggested by HLA haplotype interaction (18) and the finding of islet autoimmunity markers at onset in some children and adults with typical type 2 diabetes, as noted above.
There is considerable overlap in insulin or C-peptide measurements at onset of diabetes and over the first year or so in type 1 and type 2 diabetes because of the recovery phase of autoimmune diabetes (the honeymoon) and the degree of glucotoxicity/lipotoxicity impairing insulin secretion at the time of testing. Elevated C-peptide levels are indicative of type 2 diabetes, but normal concentrations, or suppressed levels in the presence of hyperglycemia, are not diagnostic.
One-third or more of pediatric patients with type 2 diabetes have ketonuria or ketoacidosis at diagnosis (9,19,20).
What Is the Magnitude of Possible Misclassification?
Among ∼700 patients 5–19 years old at three university centers in Florida newly diagnosed over a 5-year period from 1994 to 1999, 3% of those initially classified as having type 1 diabetes (17 of 605) were later diagnosed with type 2 diabetes, and 8% of those initially diagnosed with type 2 diabetes were reclassified as having type 1 diabetes (6 of 77) (21).
With increasing awareness of type 2 diabetes in children, the Florida experience probably reflects the true magnitude of the diagnostic problem.
DIAGNOSTIC STRATEGY
Features helpful in differentiating type 1 and type 2 diabetes in children and adolescents are summarized in Table 8.
Acute Onset
Nonobese individuals who are not African American are very likely to have type 1 diabetes and seldom require further testing.
TABLE 8. Differentiating Type 1 from Type 2 Diabetes in Children and Adolescents
Nonobese African American youths with a three-generation
