I wasn't able to find anything in PubMed on the causal connection as you describe, though I did find papers like http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3684604/ on "Archaea and Fungi of the Human Gut Microbiome: Correlations with Diet and Bacterial Residents" which describes the correlation.

> Methanobrevibacter and Candida were positively associated with diets high in carbohydrates, but negatively with diets high in amino acids, protein, and fatty acids. A previous study emphasized that bacterial population structure was associated primarily with long-term diet, but high Candida abundance was most strongly associated with the recent consumption of carbohydrates. Methobrevibacter abundance was associated with both long term and recent consumption of carbohydrates.

How do you know it's Candida and not Methanobrevibacter?

It's a flawed experiment, though. Literally the smallest possible sample size, extremely susceptible to the placebo effect, no controls.

I'd be interested in a large, rigorously controlled study on this, where the control group is fed convincing imitations of sweet (via artificial sweeteners) and processed (via presentation) foods. A single person's biased self-analysis isn't particularly valuable.

Yes, there has been a lot written. And there is no evidence.

>but I would ask you to try staying away from processed foods and sugar for a few weeks and see if you feel better. It is an easy experiment to do.

95%+ of the food I eat comes from my property. There's nothing processed to cut out. I feel no different than I did when I was eating little ceasars pizza and coke every day for a year. This "experiment" is not an actual experiment, and would not support the claim in question even if it were.