Subclinical thyroid dysfunction: evaluating risks, underlying mechanisms and treatment efficacy using pooled individual data in an international consortium
Description: Background: Subclinical thyroid dysfunction (SCTD) affects 10-15% of adults =65 years, but who should be treated remains unclear. With past SNF grants, we showed that SCTD is associated with cardiovascular (CV) events and fractures across pooled prospective cohorts. In our EU-funded randomized controlled trial (RCT) (TRUST) of levothyroxine therapy (LT4) in 737 older adults with subclinical hypothyroidism (SHypo), LT4 did not reduce symptoms. Despite being the largest RCT to date, it was not powered for CV events, similar to all previous RCTs and our recent IEMO RCT that reported an HR of 0.61 (95%CI 0.24-1.50) for CV events and less deaths. As larger RCTs with CV endpoints are unlikely to be undertaken, it remains unclear whether LT4 can reduce CV risk and potential mechanisms, as LT4 only slightly improves lipid profile. In subclinical hyperthyroidism (SHyper), there is still no large RCT of clinical outcomes and the groups at risk remain unclear. Emerging evidence suggests that SCTD might be associated with risk of specific cancers and venous thromboembolism (VTE).Aims: To address these knowledge gaps, we aim to assess potential causal relationships and mechanisms with SCTD and risk of CV and other clinical outcomes. Our specific aims are:Aim 1. To better understand the link of SCTD and CV events:1a. To determine which of peripheral hormones (free T3, T4) vs. central hormone (TSH) in SHyper is a better predictor of heart failure (HF) events1b. To evaluate whether the associations of SCTD with CV events and HF are partially mediated through epicardial adipose tissue (EAT)1c. To explore the potential causal relationship of thyroid function with HF through Mendelian Randomization (MR)1d. To assess whether LT4 leads to different lipidomic profile in RCTsAim 2. To estimate the CV benefits of LT4 in SHypo, combining evidence from RCTs with multiple cohorts, using emulation methodsAim 3. To perform new exploratory analyses to assess whether SCTD is associated with VTE, thyroid and selected common cancers, then perform MRMethods: For Aim 1a/b+3, we will analyze individual participant data (IPD) from our international Thyroid Studies Collaboration (TSC), with now 28 prospective cohorts (n=156,718), with high power to identify clinical risks. We will collect thyroid function tests, therapies, and updated outcome data (CHD, HF, VTE, cancer) over long follow-up. We will assess if these risks differ by subgroups (e.g. age, TSH, persistent SCTD). For Aim 1c, we will use data from ThyroidOmics Consortium (n=271,000) and HERMES consortium (47,309 HF cases/930,014 controls), complemented by additional analyses, as well as for cancer/VTE (if associations are confirmed). For aim 1d, we will use biobanked blood from TRUST and IEMO RCTs (n=842) to assess novel insights on LT4 effects (vs. placebo and overt hypothyroidism) using lipidomics. For Aim 2, we will emulate one trial per TSC cohort to estimate the effects of starting LT4 on CV events over long follow-up, then use cross-design evidence-synthesis methods to combine data from emulated trials with RCT data.Expected value: Thyroid function affects CV risk and might impact other conditions relevant for older adults, such as VTE and cancer. This proposal uses existing databases and expertise uniquely available to our consortium to address important unresolved controversies. Strengths of this proposal include: 1) unique international collaborations within TSC, yielding sufficient power to clarify these risks, 2) advanced methods including lipidomics, mediation analysis and MR to better understand potential mechanisms, 3) advanced causal inference methods to pool RCT and cohort data, 4) efficient use of resources funded by SNSF and EU with analyses nested within TSC and large RCTs. This proposal will generate important scientific knowledge to help clinicians accurately identify patients at increased risk and to individualize SCTD therapy.
Amount: CHF 999,800
Funding type: research
Funder scheme: Project funding
Start date: 2025-07-01
End date: 2029-06-30
Investigator: Rodondi, Nicolas
Institution: University of Berne – BE
Subfield: Endocrinology, Diabetes and Metabolism
Field: Medicine
Domain: Health Sciences
Funded outputs count: 0
Funder award ID: 10002064
OpenAlex ID: G5731167772