A diabetes drug with a plausible mechanism falls flat in the largest trial to date, except in one subgroup
STUDY: Dieperink E et al, Addiction 2026
STUDY TYPE: Randomized controlled trial
FUNDING: U.S. Department of Veterans Affairs
Background
The approved medications for alcohol use disorder — naltrexone, acamprosate, disulfiram (and, in France, baclofen) — work modestly and are rarely prescribed. Pioglitazone, a diabetes drug that activates a receptor involved in inflammation and neuroprotection (PPAR-gamma), looked promising in animal models and small observational studies. Here’s how it looks in the largest trial to date.
The Study
- 185 veterans with at least moderate alcohol use disorder, averaging 4.7 heavy drinking days per week at baseline
- Randomized to pioglitazone 45 mg/day or placebo for 14 weeks; all received a brief behavioral compliance intervention at each visit
- Primary outcome: heavy drinking days a week 14
Both groups drank less over time. Heavy drinking days dropped from 4.7 to about 2.4 in the pioglitazone group and from 4.7 to 2.0 in the placebo group — nearly identical. No differences appeared in total drinks per week, craving (Obsessive Compulsive Drinking Scale), depression (Beck Depression Inventory), anxiety, or PTSD symptoms.
One exploratory finding stood out: those with elevated C-reactive protein at baseline (greater than 5 mg/L) showed a greater reduction in heavy drinking days with pioglitazone compared to placebo. This subgroup represented about 25% of the sample.
Side Effects
Pioglitazone was well tolerated. Cough was the only side effect more common in the active treatment group (54% vs. 39%). No serious drug-related adverse events occurred. The FDA label describes a rare risk of bladder cancer.
Limitations
Almost entirely male veterans, limiting generalizability. The COVID-19 pandemic forced a shift from in-person to hybrid visits midway through the trial, and over half of participants reported that the pandemic increased their drinking.
Practice Implications
- Treat exploratory findings as observational. They don’t pass the laws of statistics, which manes pioglitazone is ready for further study but not ready for practice.
- It was also explored in bipolar depression, with mixed results there.
- Don’t despair, here are two new trials offering more hope for alcohol use disorder, using a naltrexone-baclofen combination or something as simple as carbonated water.
—Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report
