So that's the hypothesis. Nothing to do with autophagy of healthy cells due to lack of butyrate. The "butyrate stops cancer cells from growing", that's done in vitro. The "helps prevent colorectal cancer [presumably in vivo]", that's purely hypothetical. And the "We can't be sure...", that's absolutely true. They can't. In fact, they don't know. They don't know if RS helps prevent colorectal cancer. But they'd very much like it if we believed it did, so we'd become their test subjects in self-reported experiments to find out. You know, a bit like what's been happening in the last 40 years with the saturated fat hypothesis.

In a comment on Richard Feinman's blog, I mentioned that ketones regulate their own metabolism, and he added that this is done with the TCA cycle. Butyrate is not a ketone, but if a ketone can do it, then it's possible that other substrates can do it as well. What does butyrate do to cancer cells to make them stop growing? Does it shut down anaerobic oxydation? This way, butyrate has priority, and makes the cancer cell behave like a normal cell instead. Just a hypothesis, of course, but I'm trying to understand what's going on so we don't have to rely on gut-produced butyrate, which we obviously can do since those cells obviously have access to the bloodstream.

What's the cause of colorectal cancer? It's not the lack of butyrate. Butyrate only stops cancer cells from growing, thereby implying that the prevention is done for existing cancer. If there's no cancer, butyrate loses its value. Butyrate's value is ex post facto.

Is colorectal cancer glucose-dependent? Then don't eat carbs. Is butyrate essential to stop cancer cells from growing, or are there other molecules that can do it too? Doesn't matter, butyrate can be obtained from the bloodstream, so any other cancer-fighting stuff can be obtained from there too. Is this all about just butyrate and RS? Doubtful. Dr Richard Feinman's work on insulin signaling says cancer is much more sensitive to insulin, therefore dietary carbs. And his work is done in vivo, no need to hypothesize about it. We know for a fact that a low-carb diet has a direct effect on cancer growth.

http://forum.lowcarber.org/showthread.php?t=447278
http://forum.lowcarber.org/showthread.php?t=448500

So, between a hypothesis and a fact, I'll take the fact every time.