THC plus anti-inflammatory drug shows promise in Alzheimer's mouse models

Cannabis research has long sat on a frustrating contradiction: THC has genuine neuroprotective properties that could help the brain, but it's also known for impairing memory and learning — exactly what you'd want to avoid in an Alzheimer's treatment.

Now researchers at the University of Texas at San Antonio think they've found a way around that paradox. By pairing THC with celecoxib, a common anti-inflammatory drug already approved for human use, they've managed to preserve THC's protective effects while blocking its cognitive side effects.

The inflammation problem

The story starts with an enzyme called COX-2, which drives inflammation in the brain. In 2013, Chu Chen's lab made an unexpected discovery: when THC enters the brain, it actually increases COX-2 activity. That spike in inflammation is what causes the learning and memory problems associated with cannabis use — a major barrier to using THC as a therapy for neurological disease.

But here's where it gets interesting. The brain naturally produces its own cannabinoid molecules, called endocannabinoids, which activate the same receptors as THC but work differently. One of the most important, 2-arachidonoylglycerol (2-AG), actually suppresses COX-2 and reduces neuroinflammation. The body has its own way of using these molecules without causing cognitive damage.

So Chen's team asked: what if you could block the COX-2 spike that THC causes while keeping everything else intact?

Testing the combination

They added celecoxib — a selective COX-2 inhibitor already used to treat arthritis and pain — to low-dose THC and tested it in mouse models of Alzheimer's disease. The results showed the combination worked better than THC alone. The treated mice performed better on learning and memory tasks, had less beta-amyloid and tau buildup (the hallmark proteins of Alzheimer's), and showed reduced neuroinflammation markers.

Genetic analysis revealed something equally important: the treatment shifted genes involved in synaptic function and Alzheimer's risk back toward healthier patterns.

"What really mattered was behavior," Chen said. "If cognition is not improved, then the treatment doesn't matter. And that's where the combination clearly worked better than THC alone."

What makes this finding potentially significant is the practical advantage: both drugs are already approved for human use. That means researchers could move toward clinical trials without spending a decade developing and testing a new compound from scratch. "If you develop a new compound, it can take 10 to 20 years to reach patients," Chen noted. "In this case, both drugs are already approved. That gives us a real advantage."

What comes next

This study focused on preventing cognitive decline before symptoms appear. Chen's team is now investigating whether the combination can slow disease progression or even reverse damage once Alzheimer's has already taken hold — a much harder problem, but one that would have far greater clinical impact.

The research sits at that crucial moment where basic neuroscience is starting to point toward something that could actually reach patients.