Rheumatoid arthritis and kidney stones

I just wanted to tidy up some loose ends I have in my head about rheumatoid arthritis, diet, kidney infections and kidney stones before I can get on to other ideas. Squiggs is over chickenpox and multiple on-call shifts are over for a while, so I just might get some blogging done...

My understanding of RA is based around Prof Ebringer's work, summarised here, linking urinary proteus infection to antibody production against its urease enzyme. The antibody against this bacterial protein cross reacts with the collagen in small joints and the end result is RA.

Why is RA so intractable? Why doesn't a short course of antibiotics clear up both the urinary proteus and the RA? There are several reasons.

As you well know, RA patients have a high incidence of kidney stones.

This is probably important and here's why:

Ammonia is pretty toxic to mammals, sufficiently toxic that we expend energy in joining two molecules of ammonia to one of carbon dioxide to give a relatively non toxic compound, urea. We can then excrete this with ease. Of course to a bacterium, with no concerns about the toxicity of ammonia to mammals, urea is just a food source. Splitting urea with a urease releases both the energy invested by the mammal and, unfortunately, the rather unpleasant ammonia.

What happens to the ammonmia? OK, it gets urinated out. But it also does two things on the way. First it renders the urine alkaline and second it provides NH4+ ions. So what? Both things are bad, but can be made worse.

There is also an association between RA and metabolic syndrome. Two of the hallmarks of metabolic syndrome are hyperglycaemia and hyperinsulinaemia. Importantly it is also associated with magnesium deficiency.

Because modern diets tend to be grossly magnesium deficient, you would expect the body to hang on to its magnesium as tightly as possible. But that doesn't seem to happen. Both hyperinsulinaemia and hyperglycaemia cause urinary magnesium loss.

So magnesuria in the face of magnesium deficiency is a feature of metabolic syndrome. What about a link between phosphate loss and insulin resistance? Metabolic syndrome is associated with phosphate loss in the urine too.

BTW: Coupled with the urinary calcium loss which also occurs, this is how you urinate your bones down the loo to get osteoporosis if you follow mainstream nutritional advice.


What do you get when you mix ammonium ions, magnesium ions and phosphate ions at an alkaline pH? Answer: Struvite urinary stones. Magnesium ammonium phosphate.

Struvite is a common form of urinary tract stone. It forms with ease when a person with metabolic syndrome (magnesium and phosphate in the urine) gets a low grade urinary infection with a urease producing bacterium (providing ammonium ions and an alkaline pH in the urine). Struvite is porous and harbours those very same urea splitting bacteria in a location where it is remarkably difficult to get antibiotics to penetrate... These stones can become enormous. They don't go away unless you sort out the metabolic syndrome as well as the urinary infection, which is often sub clinical and the person carrying it doesn't even know they have it. In fact the first warning sign many people with a kidney stone get is the sudden agony when the stone enters and stretches a ureter...

Proteus is one of the best struvite generating bacteria available. It's a normal commensal in the gut where it does well in the anaerobic conditions of the colon. It is highly motile and gets from the gut to the urinary tract with some ease. Ascending infection would be expected to be commoner in females than in males, owing to anatomical considerations. So RA should be commoner in females than males. It is.

So to summarise: A combination of metabolic syndrome with Proteus mirabilis infection is a generator of struvite stones. The bacterial enzyme used to extract energy from urea while generating ammonium ions has a peptide sequence remarkably similar to the collagen in small joints and is the best candidate to trigger rheumatoid arthritis. The process of urea splitting to release energy is an anaerobic reaction, the bacterium is just extracting the energy previously used to make the ammonia safe... Urea splitting is obviously well suited to the urinary tract, which is as anaerobic as the colon.

So antibiotic therapy tends to be limited in effect as it is difficult to clear the urinary infection in the presence of struvite stone(s) and difficult to dissolve the stones in the presence of metabolic syndrome. It is also essentially impossible to eliminate all proteus from the gut using antibiotics, provided the gut environment is convivial to the microbe. How does all of this fit in with diet?

I want to flick through the diet trials which have been used and those which haven't but perhaps should be. First thing to note is that, as detailed by Kjeldsen-Kragh, these trials are appallingly difficult to conduct.

Skoldstam's group started it all off back in the 1970s and looked at both fasting and vegetarianism. The fasting had some effect but there was always that niggly problem of only being able to use it for short periods and the vegetarian diet immediately re established the problem. They seem to have gone the vegetarian route as there was huge popular confidence in this approach at the time. It didn't work.

Another swedish group did that vegan diet diet study which was initially gluten free and found some effect, but only in the subgroup who adhered well to the diet (lapsing from the diet shows as the production of anti gliadin antibodies, only those who stayed anti gliadin antibody negative improved). You can't say from this study if it was the successful gluten avoidance which worked or the vegan aspect. But I can guess! They didn't look at proteus.

Kjeldsen-Kragh's group in Norway used a similar vegan gluten free diet followed by a vegetarian diet with some success. They were associated with Ebringer's proteus group from King's College. Here it was the subgroup which lowered both faecal proteus count and blood anti proteus IgG count which improved. They didn't look at anti gliadin antibodies but the initial diet was gluten free.

Skoldstam's group has more recently tried applying the Mediterranean Diet, at least as Mediterranian as envisaged in the Lyon heart study diet, based around canola oil gloop. Again they got some improvement, possibly due to improved lipid composition of the diet (more omega threes) rather than changing the disease process itself. I doesn't look like remission to me.


So how do you integrate all of this in to one concept? My feeling is that people will have to have a genetic predisposition. This will relate to those human leucocyte antigens known to be associated with RA (there are several). No one can change these and they simply set limits on what you can "get away with" in terms of your internal environment. These leucocyte antigens determine what you see as self or non-self. I don't see them as the "cause" per se. In general I feel genetics is phenomenally important in determining how we "break" under adverse conditions. Your genetics don't do the breaking. Mostly your diet does the breaking. Another post there to clarify that.

They have to have metabolic syndrome. This is side step-able.

They have to have intestinal dysbiosis with a predominance of proteus in their colonic bacteria. The best way to get this is probably to eat gluten but clearly gluten is not the be all and end all of the problem.

They have to have recurrent or continuous low grade urinary tract proteus infection, probably without signs and probably derived from their colonic bacterial population.

There may well be background maintenance of the immune response to proteus due to its ability to cross in to the blood stream directly from the gut and so be seen by the immune system, without necessarily invading the urinary tract. I think it is unlikely to be around and expressing urease for very long in the systemic circulation.

So that sums up the diets and the problems to me. It looks like you need to side step the metabolic syndrome while damping down proteus levels.

So which diet is out there in the mainstream and just crying out to be tried in RA?

Let's go back to Ray Audette and Neanderthin (bit of a collector's item by the current used prices!). This little book was not in my early reading but I've had a copy for some time. It's an interesting read, especially the early sections. Here Audette describes the miseries of his life with RA, which were later compounded by the additional miseries of type two diabetes. He did the paleo thing for his diabetes and virtually immediately ate his way to normoglycaemia. The RA went the same way as the diabetes.

Anyone reading Stephan's blog on diabetes trials will remember Lindeberg's paleo diet trial for human type two diabetes. Essentially, Lindeberg replicated Audette's approach in a bigger group than n=1. It was certainly quite effective for type two diabetes, despite not being a particularly LC approach. I suspect that reducing fruit consumption would improve Lindeberg's diet a lot, but it's pretty good as it is.

So what I would love to see is Lindeberg's diet applied to RA.

Okay, I'd really rather see the Optimal Diet applied to RA, but I'm not waiting around for that one...

Peter