The GABA Paradox: When Calming Supplements Don't Work

What if “too much glutamate” and “not enough GABA” isn’t always the real issue in autism? What about when calming supplements and medications aren’t working, or are even making things worse?

It’s pretty popular to talk about a GABA/glutamate imbalance. GABA is an inhibitory, calming neurotransmitter, while glutamate is an excitatory neurotransmitter. At least, that is how it’s supposed to work.

But here’s something that almost nobody is talking about.

In autism, almost half (40%!) of people have a paradoxical response to GABA, that is, GABA has become excitatory just like glutamate.

The fact that this can happen was noted in scientific literature as far back as 1987. (PMID: 2826308)

A recent study from February 2026 revisits bumetanide, a loop diuretic, as a potential treatment in autism that appears to significantly improve symptoms in a subset of patients: 40%. (PMID: 41633963)

There were several other studies for bumetanide in autism going back to 2010. There are also studies with similar diuretics such as torasemide, azosemide, and furosemide. Bromide, too, works on the same pathophysiology but by a different mechanism (by displacing the chloride).

The idea behind these diuretics is that they block the chloride channel NKCC1, which brings chloride ions into neurons. If chloride is too high, then when GABA binds to its receptor, it lets chloride OUT instead of IN. This depolarizes the membrane and makes the neuron more likely to fire (excitatory). It’s the opposite of what is supposed to happen: GABA binds, lets chloride in, and this hyperpolarizes the neuron, making it less likely to fire, an inhibitory effect.

The counterpart to NKCC1 is KCC2, which is an ion channel that exports chloride out of the cell. If for any reason KCC2 becomes downregulated, or less active, this would cause chloride to build up in the cell, thus creating the issue mentioned previously.

Here is the important part: inflammation, particularly cytokines like IL-1β can cause downregulation of KCC2, allowing chloride to build up and create a paradoxical effect for GABA.

Therefore: if calming supplements or medications such as benzodiazepines (sometimes given in emergency rooms or before surgeries to calm a stressed kid down) do not work for your child or only make things worse, it is very likely due to too much chronic inflammation.

Even worse, it can become a vicious cycle: GABA receptors also exist in immune cells, and a paradoxical response here can contribute to prolonged inflammatory states, which perpetuates downregulated KCC2.

Still, you hear the advice all the time in biomedical circles: “give calming supplements.” The truth is, this doesn’t work in a very significant minority. Not without lowering inflammation first.

And lowering inflammation does work! I have tested and confirmed it myself, with my son.

Years ago, we were trying all sorts of biomedical interventions, but every gain was only temporary, as he would regress with the next illness. This turned out to be PANS caused by mold. PANS is a form of autoimmune encephalitis, which of course involves chronic neuroinflammation. There was one exception to “nothing had a lasting effect”: potassium bromide. I now realize that this is because bromide bypassed inflammation and corrected a pathophysiology that was caused by it. Bromide made him able to use a fork and spoon, and drink from a regular cup on his own. These are movements that require the cerebellum and Purkinje cells—the same cells that were preserved by torasemide in a rat autism study. (PMID: 36879996) When I once removed bromide to see what would happen, he lost this ability again.

With a combination of modulating the immune system with gut balancing and KPV, and reducing mold exposure, we were able to lower inflammation enough to put PANS into remission. I then thought it would be a good time to try removing bromide again, to test my hypothesis about inflammation being the real source of the issue. This time, there was no regression. I tested this further by adding supplements that promote GABA, such as L-theanine and Serene Calm from Researched Elements. No paradoxical reaction!

Why is nobody talking about this? I suppose perhaps because many interventions do lower inflammation and end up correcting the issue accidentally. But I wouldn’t necessarily bet on it, and I do think it is worth discussing, especially since “give calming supports” is often seen as a one-size-fits-all recommendation.

Despite the fact we were trying other interventions for years, following advice that is very common, we weren’t seeing lasting progress. Going to a functional doctor, changing the diet, adding mitochondrial support, trying folinic acid / leucovorin, doing antimicrobial-heavy gut protocols, zeolite detox sprays—weren’t working for us. Or if it did, whatever was gained would be lost with the next illness.

On the other hand, gut balancing brought my son out of a months-long flare, even before we started addressing mold. Gut protocols that focus on killing off bad bacteria can be MORE inflammatory by releasing endotoxin / LPS. But gut balancing which places more focus on building up the good bacteria, such as the ones that produce anti-inflammatory butyrate, did show benefit.

Mineral balancing should also reduce inflammation by lowering toxic metal burden. This is another intervention we added before taking away bromide. Though we tried detoxing with zeolite sprays before, it did not work as well because it can only pick up what is already floating around, not what is deep in tissues replacing essential minerals at enzyme sites.

If you’re struggling with repeated regressions and/or the lack of success with calming supplements, it could be that inflammation is chronically too high, downregulating KCC2 and causing GABA to be excitatory. Excessive excitation may not be a glutamate issue after all, and increasing GABA won’t fix it.