Prevalence and Incidence of Thyroid Dysfunction in Type 1 Diabetes, Type 2 Diabetes and Latent Autoimmune Diabetes of Adults

The Fremantle Diabetes Study Phase II

Kirsten E. Peters; Stephen A. Paul Chubb; David G. Bruce; Wendy A. Davis; Timothy M. E. Davis


Clin Endocrinol. 2020;92(4):373-382. 

In This Article


Baseline Patient Characteristics

The 1617 participants were significantly older than the 51 who were excluded (mean age 63.8 vs 55.5 years, P = .001), reflecting the younger age of the FDS2 subjects with secondary and monogenic diabetes who comprised almost half of those excluded (Figure 1), but diabetes duration, sex, race/ethnicity and thyroxine use were similar (all P ≥ .11). The baseline characteristics of the included participants are summarized by diabetes type in Table 1. Patients with type 1 diabetes were younger and had longer diabetes duration than those with LADA or type 2 diabetes (P < .001). The proportions of males, Anglo-Celts and those taking T4 were similar by diabetes type (P ≥ .072).

Figure 1.

Details of the participants with type 1 diabetes (T1D), latent autoimmune diabetes of adults (LADA) or type 2 diabetes (T2D) recruited to the Fremantle Diabetes Study Phase II (FDS2) with thyroid function test (TFT) data available for the present analyses. Known thyroid disease was defined as hospitalization for/with thyroid disease, and/or taking medication for thyroid disease. Undiagnosed thyroid disease was defined by biochemical analysis of thyroid function

Prevalence of Thyroid Disease

The prevalence of known thyroid disease (based on prior hospitalization and/or use of thyroid medication) was 11.7% (189/1617), with more females affected than males (20.4% vs 3.8%, P < .001). There was a nonsignificant trend to a higher prevalence in type 1 diabetes (16.9%) vs LADA (11.4%) and type 2 diabetes (11.2%) (P = .17). After adjustment for age and sex, the differences in prevalence between diabetes types were significantly different (P = .003) but further adjustment for anti-TPO rendered the association nonsignificant.

For the remaining 1428 participants without documented thyroid disease, the distributions of TSH and FT4 are shown in Table 2. There were 93 individuals (6.5%) with an abnormal TSH, nine with type 1 diabetes, five with LADA and 79 with type 2 diabetes (P = .62). There were no significant differences in the number of participants with an abnormal TSH by gender by diabetes type (all P ≥ .33). Across all diabetes types, 5.1% had subclinical hypothyroidism, 1.1% overt hypothyroidism, 0.1% subclinical hyperthyroidism, and 0.2% overt hyperthyroidism. For type 1 participants, none had overt hypothyroidism, one had overt hyperthyroidism (0.9%), 6.5% had subclinical hypothyroidism and 0.9% had subclinical hyperthyroidism. For participants with LADA, the prevalence of subclinical hypothyroidism was 7.1%, with no overt hypothyroidism or any hyperthyroidism detected. In type 2 participants, the prevalences of overt and subclinical hypothyroidism were 1.2% and 4.9%, respectively, and for overt and subclinical hyperthyroidism, 0.2% and 0.1%, respectively.

The overall prevalence of any thyroid disease (known/previously undiagnosed) in the cohort at baseline was 17.4% (282/1617). By diabetes type, 23.8% of type 1, 17.7% of LADA, and 16.8% of type 2 patients had prevalent thyroid disease (P = .13). After adjustment for age and sex, the differences in prevalence were significantly different (P = .001) but further adjustment for anti-TPO rendered the associations with diabetes type nonsignificant. The total prevalence of known and overt disease was 12.8% (207/1617), and 8.7% (18/207) of these cases did not report being treated nor had they been coded as having thyroid disease in the WADLS.

There were 183 anti-TPO positive individuals, with a higher proportion of type 1 patients compared to LADA and type 2 (28.5% vs 19.0% vs 9.3%, P < .001). Significantly more females were anti-TPO positive compared to males (17.3% vs 5.9%, P < .001). After adjustment for age and sex the prevalence of anti-TPO positivity by diabetes type remained significantly different (P < .001). There was a significant association between anti-TPO positivity and glutamic acid decarboxylase (GAD) antibody positivity in the type 1 patients (P = .006). All LADA patients were GAD antibody positive and all T2 patients GAD antibody negative due to the definitions of LADA and type 2 diabetes in this study.

Longitudinal Changes in Participants With Baseline Thyroid Dysfunction

Of the 93 participants with an abnormal baseline TSH, 57 (61.3%) had follow-up samples available, taken a mean of 4.3 ± 0.4 years after study entry (Table 3). There was no significant difference in baseline age, sex, BMI, Anglo-Celt ancestry and diabetes duration between individuals with and without follow-up (P > .53). Patients without follow-up had either died (16.1%) or had withdrawn from active participation (22.6%).

Among participants with type 2 diabetes and baseline subclinical hypothyroidism, 40 returned for follow-up at 4 years, four of whom had been started on thyroid replacement therapy by their usual care clinicians during follow-up (Table 3). Of the remainder, subclinical hypothyroidism was transient in 15/36 (41.7%), had persisted in 20/36 (55.6%) and had progressed to overt hypothyroidism in one (2.8%). Among eight individuals with undiagnosed overt hypothyroidism at baseline, 2 started thyroid replacement during the follow-up period. Of the remainder, none had TSH >8.2 mU/L or FT4 <10.4 pmol/L at baseline; two remained overtly hypothyroid, one became subclinically hypothyroid and three reverted to normal thyroid function. There was one patient with subclinical hyperthyroidism at baseline whose status remained unchanged at follow-up. All three type 1 participants with subclinical hypothyroidism at baseline reverted to euthyroidism with two starting thyroid replacement therapy during follow-up. One participant had subclinical hyperthyroidism at baseline and was overtly hyperthyroid at follow-up despite treatment with carbimazole. Of the three LADA participants with baseline subclinical hypothyroidism, one reverted to euthyroidism and two remained subclinically hypothyroid at follow-up (Table 3). Thus, 14.8% (8/54) of the participants with raised TSH at baseline started thyroxine replacement during follow-up.

Incidence of New-onset Thyroid Disease

Serum thyrotropin concentrations were available at entry and Year 4 for 844 participants with no history of thyroid disease and normal baseline TSH. During 5694 patient-years (mean ± SD: 6.7 ± 1.0 years) of follow-up to end-June 2016, 25 (3.0%) had a first hospitalization for/with thyroid disease or started thyroid medication, with more females affected than males (4.5% vs 2.0%, P = .04). There were similar rates of known incident thyroid disease in patients with type 1 and type 2 diabetes (2.8% and 3.1%, respectively) and no new cases in patients with LADA. There were too few cases to allow adjustment for age and sex.

Follow-up thyroid function testing of the remaining 819 participants without evidence of thyroid disease at baseline is summarized in Table 4. The median serum TSH increased during follow-up (from 2.1 [1.6–2.9] to 2.3 [1.6–3.1] mIU/L; P < .001); 34 (4.2%) progressed to biochemical thyroid disease, including 28 with subclinical hypothyroidism, two with overt hypothyroidism and four with subclinical hyperthyroidism. A greater proportion of patients with type 1 versus type 2 diabetes progressed (8.3% vs 4.6%), but the difference was not significant (P = .14), and there were too few cases to permit adjustment for age and sex.

The incidence of any new thyroid disease, including those with an abnormal follow-up TSH and those with known incident thyroid disease, was 59/844 (7.0%) during 4.3 ± 0.5 years of follow-up, with a trend for more females affected than males (8.9% vs 5.7%, P = .075). After adjustment for age and sex, the incidence of any new thyroid disease was significantly higher in type 1 compared to type 2 diabetes (P = .026). The total incidence of new overt disease was 3.2% (27/844), with 7.4% (2/27) of these patients unaware of the condition. All of the patients with LADA remained euthyroid during follow-up.