- Design
- Open-label, randomised controlled pilot trial, hypothesis-generating (CTRI/2025/06/089606)
- Population
- 60 Indian adults with type 2 diabetes and subclinical hypothyroidism on background SGLT2 inhibitor therapy; 51 completed follow-up
- Primary outcome
- Change in urinary albumin-to-creatinine ratio and estimated GFR at six months
- Effect
- UACR -13.09 mg/g vs +31.84 mg/g; per-protocol difference -44.93 (95% CI -77.6 to -12.26, p=0.008), intention to treat -35.61 (-77.31 to 6.08, p=0.092); eGFR not significant
Subclinical hypothyroidism is common in type 2 diabetes and usually left alone. Whether treating it does anything for the kidney has not been tested properly. This open-label pilot, registered with the Clinical Trial Registry of India, randomised 60 adults with type 2 diabetes and subclinical hypothyroidism, all on background SGLT2 inhibitor therapy, to levothyroxine or no levothyroxine for six months.
Urinary albumin-to-creatinine ratio fell by 13.09 mg/g with levothyroxine and rose by 31.84 mg/g in controls. The between-group difference was significant per protocol (-44.93 mg/g, 95% CI -77.6 to -12.26, p=0.008) but not by intention to treat (-35.61 mg/g, -77.31 to 6.08, p=0.092). Estimated GFR changes were not significant either way. Thyroid-stimulating hormone fell as expected in the treated arm. No adverse events were reported.
The honest reading is in the gap between those two analyses. Nine of 30 treated participants and three of 30 controls did not complete follow-up, and per-protocol analysis discards exactly the people whose outcomes are most likely to differ - which is why intention to treat is the primary standard, and why the intention-to-treat result, crossing zero, is the one that describes what this trial showed. The authors call it hypothesis-generating, and that is right.
What makes it worth reading anyway is the design choice. Randomising on top of established SGLT2 inhibitor therapy asks the question that matters now: does treating subclinical hypothyroidism add anything to modern renoprotection? Nobody should start levothyroxine for albuminuria on this. But it is an Indian trial, in an Indian population, asking a question with real local relevance - subclinical hypothyroidism is highly prevalent here - and it deserves the larger trial it argues for.
- Do not start levothyroxine to treat albuminuria - the intention-to-treat result was not significant
- Continue to treat subclinical hypothyroidism on its existing indications, not renal ones
- Note the SGLT2 inhibitor background: this asks whether levothyroxine adds to it, not whether it works alone
- Read per-protocol results with suspicion when dropout differs between arms
- Recheck thyroid function in diabetic patients with rising albuminuria - as part of the workup, not as a treatment plan
The statistics, in plain English
The intention-to-treat interval, -77.31 to +6.08 mg/g, crosses zero: the data are consistent with a substantial reduction in albuminuria and with a small increase. Per-protocol analysis, which produced the significant result, includes only those who completed - and since more people dropped out of the treatment arm, that selection could produce the difference on its own. With 60 participants and a hypothesis-generating label, the trial was never sized to answer the question; it was sized to find out whether the larger study is feasible.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for nephrology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free