There goes that hypothesis

What a lot of non-scientists don't understand is the role of the hypothesis, and the role of the theory.

I just got my blood tests back from the challenge diet.  My plasma ammonia levels went from a baseline of 42 (normal) to.... 30.  Statistically the same.

What does this mean?  It means that this hypothesis, that my chronic pain was caused by hyperammonemia, has been disproven.  Assuming the lab didn't screw up, that's not what's going on.  When I eat protein, I don't end up with a buildup of ammonia in my blood.

So I'll toss that hypothesis, look at the data again, and form a new one.  But these hypotheses that I'm testing are meant to answer the specific question, "what is the mechanism that connects my diet with my chronic pain?"  They don't address the larger question, "Does my diet affect my chronic pain?", nor the slightly smaller question, "Is protein intake directly related to my pain level?"  Both of those have been answered to my satisfaction.  The evidence I've collected, such as the fact that arginine helps tremendously all on its own, and that my feet are better when I eat less protein *and* elevate them for a period, all fits together.  I can put the pieces together to make a big picture that makes sense.

That big picture is my theory: I have some form of protein-metabolism disorder.  Did the ammonia test disprove my theory?  No.  It just showed me that the picture is more complex than I had first assumed.  All of the other evidence still supports my theory.

I have learned that I was asking the wrong question.  Now I need to find the right one to ask next. 

The broadest swath of protein-metabolism disorders are UCDs, so that's the label I've been using to find information.  The next logical question to ask would be, "Do all UCDs result in elevated plasma ammonia levels?"  If the answer is "No", I may have just narrowed the possibilities dramatically.  If the answer is "Yes", I need to broaden my search terms to other metabolic disorders.

This is what scientists mean when we say that failure is just a different kind of information.  My results weren't what I expected, so now I use that to figure out what to do next.  It's all part of the process.

(But I don't mind telling you, my first reaction was "Damn, we're getting into really deep Zebra territory here.")