Clinical Summary: Treating Insulin Resistance With Metformin as a Strategy to Improve Clinical Outcomes in Treatment-Resistant Bipolar Depression (the TRIO-BD Study): A Randomized, Quadruple-Masked, Placebo-Controlled Clinical Trial
Patients with treatment-resistant bipolar depression often remain depressed despite multiple adequate medication trials, with high functional burden and few evidence-based next steps. This trial asks a clinically practical question: if insulin resistance is driving a subset of refractory bipolar depression, can reversing it with metformin improve depressive symptoms, anxiety, and functioning?
Key Findings
- At week 14, 10 metformin-treated patients (50%) no longer met IR criteria compared to 1 placebo-assigned patient (4%) (Fisher exact P = .0009).
- At week 14, significantly more converters were treatment responders, defined as reduction in baseline MADRS scores of ≥ 30%: 9/11 (81.8%) versus 11/28 (39.3%) among non-converters (Fisher exact P = .031).
- Depression improvement in converters was large and durable, with MADRS effect sizes of d = 1.17 at week 14 and d = 1.04 at week 26; improvement began at week 6 and was maintained until week 26.
- Functioning improved significantly in converters compared to non-converters, with GAF effect sizes of d = 1.47 at week 14 and d = 1.58 at week 26.
- Metformin-treated patients lost 1.67 kg by week 14 whereas placebo-assigned patients gained 1.35 kg (mean ± SD difference: 3.02 ± 1.31 kg; t122.99 = 2.29, P = .023), and YMRS mean scores remained low throughout the study.
In insulin-resistant treatment-resistant bipolar depression, the clinically meaningful signal was reversal of insulin resistance: patients who converted had substantially better depression, anxiety, and functioning outcomes, and metformin achieved conversion in 50% versus 4% with placebo at week 14. Screening for insulin resistance identifies a potentially treatable subtype of refractory bipolar depression.
Practice Implications
- Check for insulin resistance in patients with treatment-resistant bipolar depression using fasting plasma glucose and serum insulin to calculate HOMA-IR, as the trial enrolled patients with HOMA-IR ≥ 1.8 and excluded those with type 2 diabetes mellitus.
- When using metformin in this population, monitor both metabolic conversion and mood outcomes early; converters showed significantly greater improvement in MADRS and GAF beginning at week 6.
- Do not assume metformin assignment alone predicts antidepressant benefit; among non-converters, response rates were 4/10 (40%) with metformin and 7/24 (29.2%) with placebo, indicating that insulin resistance reversal was the key clinical marker.
- Discuss gastrointestinal adverse effects and short-term weight change with patients, but note that adherence was 97% for both treatment groups, no subjects withdrew because of side effects or noncompliance, and no serious adverse events occurred.
