We are missing a lot of context - much of which you can't supply here even if you wanted to. For example, it seems entirely plausible and not exactly rocket science that not eating the wrong foods (for a given person - what "wrong" means may vary widely and wildly depending on their context!) is better than eating it. For an example (of the example), if you eat a lot of high glycemic index foods it's not hard to guess why your blood sugar is doing better without it.

This is the problem with all those reports from people who say they are doing so much better during a fast.