Factors associated with thyroid dysfunction in children with newly diagnosed type 1 diabetes mellitus

Thyroid disease and type 1 but also type 2 diabetes mellitus (DM) are strongly associated, and this has important clinical implications for insulin sensitivity and treatment requirements. The pathophysiological basis of this association has only recently been better elucidated. It rests on a complex interaction of common signalling pathways and, in the case of type 1 diabetes and autoimmune thyroid disease, on a linked genetic susceptibility. The pathophysiological mechanisms underlying this linked regulation are increasingly being unravelled. They are exemplified in the regulation of 5' adenosine monophosphate-activated protein kinase (AMPK), a central target not only for the modulation of insulin sensitivity but also for the feedback of thyroid hormones on appetite and energy expenditure. The present review will discuss these concepts and their consequences for the clinical care of patients with DM and thyroid disorders. Moreover, it makes reference to the added effect of metformin in suppressing TSH.

The temporal development of autoantibodies was studied in 1,353 offspring of parents with type 1 diabetes. Islet cell antibodies (ICAs) and autoantibodies to insulin (IAAs), glutamic acid decarboxylase, and IA-2 were measured at birth, 9 months, 2 years, and 5 years of age. At birth, no offspring had islet autoimmunity other than maternally acquired antibodies, which were shown to influence antibody prevalence up to age 6 months. Antibodies detected thereafter were likely to represent a true de novo production, since prevalences were the same for offspring from mothers and fathers with diabetes, antibodies detected at 9 months were almost always confirmed in the 2-year sample and were associated with an increased likelihood of having or developing other antibodies. By 2 years of age, autoantibodies appeared in 11% of offspring, 3.5% having more than one autoantibody. IAAs were detected most frequently, and few had autoantibodies in the absence of IAAs. In 23 offspring with multiple islet autoantibodies, IAAs preceded other antibodies in 10 cases and were first detected concurrently with other antibodies in 12 and after detection of other antibodies in 1. Development of additional antibodies and changes in levels, including decline of IAAs at older age, was frequent. Nine children, all with IAAs and ICAs, developed diabetes. Overall cumulative risk for disease by 5 years of age was 1.8% (95% CI 0.2-3.4) and was 50% (95% CI 19-81) for offspring with more than one autoantibody in their 2-year sample. Autoimmunity associated with childhood diabetes is an early event and a dynamic process. Presence of IAAs is a consistent feature of this autoimmunity, and IAA detection can identify children at risk.