Should we test routinely for autoimmunity among children with newly diagnosed type 1 diabetes mellitus?

All tests have costs, not just financial, but also attitudinal, in shaping future behaviors of patients, their family members, and healthcare personnel. There could be changes in actions taken, including getting more tests or implementing measures for risk reduction, and emotional reactions, including higher anxiety levels, or complacency and reduced adherence to future necessary actions. Therefore, in an ideal world where wastage is abhorred, whenever a test is ordered, the health care team should be aware of these factors, be sure it is money well spent, and be clear about what it intends to do with the test results. For example, there is a strong case for universal screening of thyroid function in newborns, with reasonable clarity about what to do with the reports. About 98–99% newborns would be considered normal; the very few with abnormal reports would be advised confirmatory retesting; and the one in 1,000–3,000 affected newborns would be started on thyroxine replacement so that brain development is not hampered, and intellectual disability is prevented. It is another matter that this screening still eludes over 70% newborns worldwide.^[1]

The situation in type 1 diabetes (T1D) is less straightforward. T1D is recognized to be an autoimmune disorder, with a higher propensity to develop other autoimmune conditions, including hypothyroidism and Graves’ disease, celiac disease, Addison disease, and vitiligo; hence, it does seem logical to document thyroid and adrenal autoantibody status.^[2] How exactly these tests help, however, is not always thought through. A paper in this issue of the journal reports a retrospective case review of thyroid (thyroid peroxidase [TPO]) and adrenal antibody status and function of children presenting to the Royal Hospital for Children, Glasgow, with newly diagnosed T1D over 5 years.^[3] In clinical medicine, analysis of experience always provides valuable insights, and this paper, too, has interesting nuggets of data. However, it also throws up an opportunity to raise several questions in the way we approach the child with T1D.

Of 504 children diagnosed with T1D, data were available for 384—TPO antibodies were tested at diagnosis in 95% of them. Of the 69 (18%) who had raised TPO antibodies, 10 (14%) developed hypothyroidism, with 8 diagnosed at the time of initial TPO testing. Over the median 4 years of follow-up, only two more TPO-positive children went on to develop hypothyroidism—2.3 and 2.4 years later. The paper made no mention of how many TPO-negative children developed hypothyroidism during the period of follow-up. We need to keep the following facts in mind: (a) the need to test TSH every 1–2 years or if symptoms development continues, regardless of thyroid antibody status, (b) hypothyroidism can develop in thyroid antibody negative individuals, (c) the cost of testing TSH is a fraction of the cost of testing thyroid antibodies, (c) thyroxine replacement is started only when TSH rises, regardless of thyroid antibody status, and (e) at present, medical science cannot prevent hypothyroidism. The International Society for Pediatric and Adolescent Diabetes (ISPAD) guidelines say that “screening children for thyroid antibodies can help stratify which youth with diabetes to follow most closely for development of hypothyroidism.”^[2]

Apart from that somewhat tenuous suggestion, thyroid antibody testing seems to fulfill very little purpose, while adding to costs. Moreover, an antibody-positive report may further increase the caregivers’ anxiety at a time when they are coping with the devastating diagnosis of T1D, while a negative report may increase complacency and delay routine TSH testing, which is important because symptoms of hypothyroidism can be quite subtle and easy to miss.

Detectable adrenal autoantibodies are reported in up to 2% individuals with T1D, and only 0.2% develop Addison disease; the ISPAD guidelines mention, but do not outright recommend, routine screening.^[2] Yet, adrenal antibody status was obtained in 99% of the children at T1D diagnosis, with two testing positive at diagnosis (one was positive before diagnosis), and no adrenal dysfunction was noted. Once again, the cost–benefit ratio is debatable.

ISPAD guidelines on classification of diabetes advise that “if the diagnosis is unclear, include diabetes-associated autoantibodies: glutamic acid decarboxylase 65 autoantibodies (GAD); tyrosine phosphatase-like insulinoma antigen 2 (IA2); insulin autoantibodies (IAA); and β-cell specific zinc transporter 8 autoantibodies (ZnT8). The presence of one or more of these antibodies confirms the diagnosis of T1D in children. The possibility of other types of diabetes should be considered in the child who has negative diabetes-associated autoantibodies.”^[4] Such advice is based on the high rates of diabetes-associated antibody positivity among Caucasians. However, Asian or African populations have much lower positivity rates,^[5,6] so being antibody negative does not rule out a diagnosis of T1D. Unfortunately, it is not uncommon for general physicians unfamiliar with nuances to misunderstand this, conclude that children who are antibody negative do not have T1D, and try prescribing oral drugs and/or alternative therapies. This leads to high rates of ketosis, even DKA, to debilitating chronic complications, and to considerable distress as individuals and their family members have to unlearn a “simple” diagnosis and management and learn the more arduous proper management of T1D with a basal-bolus insulin regimen based on multiple daily home blood glucose monitoring. Therefore, in the wrong hands, the tests can cause actual harm, with higher morbidity and even mortality.

Possible misinterpretation or misuse cannot be the basis for avoiding essential tests. However, there is a case to be made for discouraging the routine conduct of tests with limited utility. The importance of rising costs is in proportion to increasing resource limitations; thus, money spent on tests with limited utility would impact the ability to spend on more critical management needs. There would be little, if any, doubt in diagnosing T1D in a 4–8-year-old child presenting with a history of polyuria and weight loss, an HbA1c of 10–15%, especially if ketosis is also present. Here, money spent on expensive diabetes-associated, adrenal, or thyroid antibody testing would be better utilized to initiate continuous glucose monitoring, with blood tests for TSH and anti-tissue transglutaminase (TTG). However, with the current worldwide epidemic of obesity and consequent type 2 diabetes (T2D) and hypertension, the lines between T1D and T2D are blurring in a rising proportion of individuals. Thus, an obese adolescent with diabetes, presenting with a history of body mass index dropping from 38 to 36 kg/m², acanthosis, and blood pressure of 130/90 mmHg, poses a genuine diagnostic challenge, especially initially. Here, testing for autoimmunity would be useful, since diabetes-associated antibody-, TPO- or TTG-positivity would have a crucial impact on management.

A further point: physicians and policymakers often feel that expenses are somehow more distressing when out-of-pocket. When parents have to pay an enormous hospital bill at the time of discharge after diagnosis (DKA), with all screening tests done, it definitely causes immediate concern. However, even when the expense is borne by insurance or by the government, it is still indirectly being paid by them, though insurance premiums or taxes. We forget there simply is no free lunch. Therefore, minimizing costs by sensibly minimizing tests of limited utility is beneficial in the long run, in all settings, not only those with limited resources.

As experts familiar with nuances, experienced pediatric endocrinologists may be able to read between the lines of the guidelines and judiciously individualize which tests and treatments to prioritize. However, guidelines are used by healthcare professionals with varying experience and expertise, and may also serve as legal documents. It might be useful if guidelines on pediatric and adolescent diabetes would consider recommendations for nuanced rather than routine autoimmunity testing.