Bipolar disorder is a challenging illness with various clinical presentations. In "type one" people struggle with alternating symptoms of full blown mania, and most also have depressive episodes as well. In "type two," depression is the primary state, with the occasional rare bit of hypomania. By mania, I mean increased energy, increased sexuality, racing thoughts, insomnia, feeling grandiose or very irritable, sometimes to the point where you detach from reality and become psychotic. Medication and specific kinds of therapies focused on monitoring symptoms, adjusting lifestyle and regulating sleep-wake cycles have been proven to be helpful in decreasing the number of manic and depressive episodes. Typically, the medications are also anti-seizure medicines, such as valproate, lamotrigine, and carbamazepine.
Ketogenic diets, which are low in carbohydrate and protein while high in fat have been used to treat epilepsy for a hundred years. Since anti-seizure medicines are clearly useful for bipolar disorder (notwithstanding many side effects), would a ketogenic diet that can control seizures be useful in bipolar disorder (1)?
In the literature for epilepsy, patients were encouraged to fast for 12-36 hours to promote ketosis, and then to follow a dietary plan with less than 20 g carbohydrate daily (or even lower, in most research ketogenic diets). In doing so their brains would be flooded with ketones, and promote having some excess protons floating around in the space between the cells. Critically, it seems that in order for a ketogenic diet to help control seizures, there must be reduced sodium molecules floating outside the cells as well. There are several seizure medicines (such as gabapentin) that don't work well for bipolar disorder. When scientists look closely, they find that only the seizure medicines that promote a reduced extracellular sodium concentration are helpful in bipolar disorder. Ketosis does exactly that.
To really understand what is going on, we need to take a close look at how nerves work. It's pretty cool, really, but involves a little science. Now a picture, courtesy the US Government and Wikipedia:
Nerve impulses and signals travel along the nerve fibers via electricity. How that happens is that the extracellular levels of ions and the intracellular levels of ions are maintained at a very different level. Inside neurons, the sodium concentration is about 10mM, but outside, it is 130mM or more - rather like there are a bunch of balls stored in a container on top of a hill. Open a little door on the side of the container, and the balls come pouring out and down the hill. Potassium is the opposite - levels are very high inside the cells, and quite low outside (3).
Neurons have plasma membranes like other cells in our bodies. Those membranes are somewhat like a tarp that has been oiled on both sides. Charged ions such as sodium and potassium can't get through unless they go through special ion pumps that are located in the cell membranes. The sodium pump, in fact, may use up to 50% of the energy in the brain (4)!
The result of all these shenanigans is that our nerve cell membranes are left somewhat negatively charged (-75 mV, in fact). The neurotransmitters (such as serotonin, norepinephrine, dopamine, acetylcholine, glutamate, GABA, etc.) work by changing these membrane potentials in various ways. Neurotransmitters can open ion channels, allowing sodium to enter the cell and causing a wave of electrical impulse that travels along the neuron. Neat!
When the electrical impulse (or "action potential") reaches the end of the neuron, the "presynaptic terminal," neurotransmitters are released into the space between the nerve cells, called the synapse. At this part of the neuron, calcium is the important ion (though sodium plays a role too). The electrical impulse (originally mediated by sodium at the dendrite) that traveled down the nerve causes extracellular calcium to pour into the cell, which then leads to the release of the neurotransmitters into the synapse, which then can affect communication with the next neuron. Voila! Your neurons have now sent messages to one another. Yee haw. The sodium and calcium membrane potential craziness can be set back to baseline by the transport of potassium, so everything is all set for a new signal to be sent.
It's Friday (at least when I am writing this article!). I know. But it's important to understand the above to some extent to figure out why a ketogenic diet might change the ionic environment in our brains.
Before we get to a ketogenic diet, let's look at lithium, carbamazepine, and valproate, all medications that have anti-seizure and mood stabilizing properties. Lithium is especially interesting, because it looks a lot like sodium, so much so that our kidneys can become confused between the two. Seems our brains can be confused as well. In rats treated with lithium, the lithium displaces the intracellular sodium in the neurons, and overall sodium is decreased. The changed sodium gradient may be central to the mood-stabilizing effects of lithium. (Any doctors out there will be squinting at me right now - hey, lithium isn't an anti-seizure med! Well, actually, in the old days it was used as one. Lithium can be horribly toxic at levels high enough needed to control seizures, so it is never used for seizures now. Carbamazepine is a little mysterious, but one of its effects is to definitely inhibit the voltage-sensitive sodium channels. (Lamotrigine, another anti-seizure and mood stabilizing drug, does something very similar). Valproate has a whole load of actions, and can increase GABA, making it a pretty good anti-anxiety med. Another thing it does is to decrease the rapid-fire ability of the spazziest neurons, probably by inhibiting the sodium channels.
Get the picture here? All these meds can be life-saving if you have bad bipolar disorder or seizures, but they can all be pretty toxic and have a host of side effects. But all of them work (effectively) as insulators in the brain, decreasing the ability of the neurons to send out out-of whack sodium messages leading to neurotoxic calcium overload. This calcium overload is speculated to be the cause not only of seizures, but also migraines and bipolar symptoms, which is why anti-seizure meds can be used to treat all three conditions.
Enter the ketogenic diet. Ketogenic diets (severely carbohydrate restricted diets) result in ketone bodies (made from fat) being used by the brain as fuel in lieu of glucose. The ketone bodies, acetoacetate and beta-hydroxybutyrate, are acidic. That simply means that there will be extra H+ protons hanging out, compared to a non-ketotic brain. Well, protons can be pumped into neurons in exchange for sodium, acting a little bit like lithium. And a few extra protons outside the cell do all sorts of interesting things, such as reduce the excitability of the neurons and reduce the activity of the excitatory neurotransmitters. Protons seem to block the calcium channels at the NMDA receptors, for example (5). GABA (the inhibitory neurotransmitter and anti-seizure also) is increased in the brain in ketogenic diets, along with many other neurotransmitter changes (6).
Sounds good! Well, some intrepid doctors in Israel had a bipolar patient who didn't respond that well to medication, and after discussion with the patient and family, it was decided to try a ketogenic diet (7). The patient fasted for 48 hours and began what is described as a "classic" ketogenic diet for two weeks. Oddly, she didn't have any ketones in her urine, which is a reliable sign of being in ketosis, especially early on. After two weeks, the doctors added medium chain triglyceride oil, which can induce a more reliable ketogenic state. The patient was pretty gung-ho on the diet, and the doctors made note that her compliance was good for the month the diet was tried. She showed no clinical improvement, no loss of weight, no urinary ketosis, and no changes in liver function. Seems odd that she wouldn't get ketosis or weight loss while fasting or on a strict ketogenic diet, but perhaps that's why it didn't work. I'll discuss a second case below.
On the internet, one can find a number of anecdotes about people improving bipolar symptoms with ketogenic or low carb diets. But it is very important to understand that there are absolutely no systematic scientific studies, not even a small pilot trial. The last thing you would want to do is, all on your own, ditch your meds and try a home-made ketogenic diet without anyone's help. Ketogenic diets can have pretty bad side effects - constipation, menstrual irregularities, elevated serum cholesterol (if you care) and triglycerides (not good), hemolytic anemia, elevated liver enzymes, kidney stones, and gallstones. Up to 15% of kids on a ketogenic diet will get changes in the heart conduction which puts them at higher risk for death (this is thought to be mediated via selenium deficiency). Valproate + ketogenic diet seemed to worsen the side effects (8).
Now some of these ketogenic diet studies were done at the height of the low-fat era, and likely designed by some pretty fat-hating nutritionists. And, like Atkins TM, many of the original ketogenic diets will have no regard for the omega6:omega3 ratio. Here's a case where Atkins made manic psychosis a lot worse - I wonder about that, as arachadonic acid (omega 6 metabolite) administration has also been shown to worsen psychosis. In that case, the patient was on valproic acid also, and that may have been part of the problem. A natural sort of ketogenic diet (think Inuit in the winter) would probably have a lot fewer of these complications and side effects.
But it makes you think, doesn't it? Many of our ancestors probably spent many a winter in ketosis, and other times lack of food or long fasts would have brought on brief periods of ketone body use in the brain too. Maybe our brains work better if we spend time in ketosis. Speculation, of course, but not an unimportant question to research further.