Long COVID Treatments: What Medications Are Being Tested and What We Don’t Know Yet

There’s no FDA-approved treatment for Long COVID yet. That’s not because doctors aren’t trying - it’s because the condition is still a mystery. Millions of people worldwide are living with symptoms like brain fog, crushing fatigue, heart palpitations, and dizziness months after their initial COVID infection. And while there’s no cure, doctors are testing a handful of existing medications, hoping one might work. But here’s the catch: we don’t fully understand how these drugs behave in Long COVID patients. What’s safe for someone with rheumatoid arthritis might not be safe for a 30-year-old who had a mild case of COVID. The stakes are high, and the unknowns are still bigger.

Baricitinib: A Promising Drug With Serious Risks

Baricitinib is one of the most talked-about drugs in Long COVID research. It’s already approved for rheumatoid arthritis and alopecia areata, and it even helped reduce deaths during the acute phase of COVID-19. Now, it’s being tested in the REVERSE-LC trial - a major NIH-funded study involving over 40 researchers across six U.S. sites. Early signs suggest it might help with fatigue and brain fog by calming an overactive immune system.

But baricitinib isn’t harmless. The FDA label warns of increased risk of serious infections, blood clots, and even lymphoma. In rheumatoid arthritis trials, about 1 in 5 patients got a serious infection. For someone with Long COVID, whose immune system is already acting strangely, this could be dangerous. Some doctors are cautious. Dr. Priscilla Hsue from UCSF points out that Long COVID patients may already have immune dysregulation - adding a drug that suppresses immunity could backfire. Results from the REVERSE-LC trial won’t be in until late 2026. Until then, it’s still experimental.

Metformin: The Unexpected Hero

Metformin, a cheap, widely used diabetes drug, surprised researchers. A 2023 study from the University of Minnesota found that people who took metformin within five days of a COVID diagnosis were 41% less likely to develop Long COVID. That’s a big deal. It’s not just a guess - it’s a statistically significant drop, with a 95% confidence interval showing real effect.

But it comes with a cost. Over a third of participants in that trial had stomach problems - nausea, diarrhea, cramps. For someone already dealing with fatigue and brain fog, that’s a tough trade-off. Still, many patients are self-prescribing it. According to the Body Politic Long COVID Support Group, 32% of respondents have tried metformin off-label. Some say it helped with energy levels. Others quit because the side effects were worse than the symptoms. There’s no standard dose for Long COVID yet. Most are taking 500-1000 mg daily, but without clinical guidance, it’s a gamble.

Low-Dose Naltrexone (LDN): The Off-Label Favorite

LDN is another off-label option gaining traction. Usually, naltrexone is used at 50 mg to treat opioid addiction. But at 1-5 mg, it appears to have anti-inflammatory effects. A 2024 study from Nova Southeastern University tracked 120 Long COVID patients on LDN. Over 60% reported less fatigue. That’s promising. But 28% had trouble sleeping, and 19% got headaches. One patient in Sydney told me she started LDN after months of being bedridden. She felt better after three weeks - but couldn’t sleep through the night. She stopped.

There’s no large, controlled trial for LDN in Long COVID. The data comes from small, patient-reported surveys. That means we don’t know the long-term risks. Is it safe to take for years? Could it interfere with other medications? No one knows. But because it’s cheap and widely available, many patients are trying it anyway.

Paxlovid: Does It Work for Long COVID?

Paxlovid - the antiviral combo of nirmatrelvir and ritonavir - was a game-changer for early COVID. But does it help with Long COVID? Two conflicting studies say no and yes.

A small UC San Francisco study in 2024 found that 38% of patients on a 15-day Paxlovid course felt better, compared to 22% on placebo. But a much larger NIH trial published in JAMA Internal Medicine in February 2025 found no difference at all. The numbers were almost identical: 34.1% vs. 32.8%. So what’s going on? Maybe it only helps certain subtypes of Long COVID. Or maybe the timing matters - was it too late for most patients in the big trial?

Then there’s the bitter taste. Nearly 8 out of 10 people on Paxlovid said it tasted awful. And ritonavir - the booster in Paxlovid - interferes with over 40% of common medications. If you’re on blood thinners, cholesterol drugs, or even some antidepressants, Paxlovid could be dangerous. It’s not worth the risk unless you’re in a trial.

Other Drugs in the Pipeline - And the Ones That Failed

Not every trial has gone well. The BC007 trial, designed to neutralize harmful autoantibodies in Long COVID, was stopped in March 2025. It didn’t work better than placebo. Worse, three patients had serious infusion reactions. That’s a red flag. It shows how easily things can go wrong.

Meanwhile, other drugs are still being tested. Sipavibart, a monoclonal antibody used for COVID prevention in Europe and Japan, is now being tested in a 100-person trial. Early safety data from its prevention use shows mild reactions at the injection site - but we don’t know how it affects Long COVID patients yet. AER002, a long-acting immunoglobulin, is in phase 2 trials. So far, only mild infusion reactions have been reported.

And then there’s GLP-1 agonists like tirzepatide (Mounjaro). These drugs, approved for weight loss and diabetes, are being looked at because many Long COVID patients have metabolic issues - insulin resistance, blood sugar swings. But they cause nausea, vomiting, and diarrhea. If you’re already struggling with gut symptoms, this might make things worse.

The Biggest Unknowns

We don’t have a blood test for Long COVID. No biomarker. No scan. No clear definition of who has it and who doesn’t. That makes trials messy. One person’s fatigue might be from mitochondrial damage. Another’s might be from lingering virus particles. A third’s might be from autoantibodies attacking nerves. The NIH now recognizes at least four different types of Long COVID. One drug won’t fix them all.

Even more troubling: most safety data comes from people with different illnesses. Baricitinib was tested on older adults with arthritis. Metformin on diabetics. LDN on chronic pain patients. But Long COVID patients are often young, previously healthy, and without other conditions. Their bodies react differently. We’re guessing.

And what about long-term use? If you take baricitinib for six months, what happens to your liver? Your immune system? Your cancer risk? No one has studied that. The same goes for LDN. What if you take it for two years? We simply don’t know.

What Patients Are Really Doing

A survey of 15,000+ members of the Body Politic support group found that 68% have tried at least one medication off-label. Metformin is top, then LDN, then Paxlovid. But 57% said it didn’t help enough. And 41% said the side effects were worse than their Long COVID symptoms. That’s a huge red flag. We’re treating symptoms with drugs that weren’t designed for this.

And then there’s sleep. The RECOVER-SLEEP trial at UC Health is testing sedatives for Long COVID-related hypersomnia. Early feedback? 18% of participants started worrying about dependence. Sleep is already fragile in Long COVID. Adding a pill that makes you drowsy might help tonight - but hurt next month.

What Comes Next

The NIH’s RECOVER initiative has poured $1.15 billion into this research. By 2027, we might have our first FDA-approved treatment. But it won’t be a magic bullet. It’ll likely be one drug for one subtype of Long COVID. And even then, we’ll need years of follow-up to know if it’s truly safe.

For now, the safest thing might be patience - and careful monitoring. Don’t self-prescribe. Talk to a doctor who understands Long COVID. If you’re in a trial, you’re helping science. If you’re not, wait for the data. The next two years will tell us more than the last five.