We inadvertently got some interesting data this week from my dd (and her dexcom). She chose to fast in Friday for religious reasons (disclaimers: she is 13, I kept a careful eye on her all day and this was her choice, nobody forced her to do it). During the day Friday her numbers were beautiful which I guess tells us our basal is set well. Side note: I find that giving low or no carb foods to basal test does not have the same effect. Her body must convert the protein to glucose pretty well so that it is not really an accurate basal test for us. But what really struck me was that though her numbers were great the first day she ran really, really low the next two nights. And actually much of the day in between too though that doesn't show up on the graphs as well. My conclusion: while fasting, she depleted her liver of its stored glucose and it took her a full 36 hours to replenish it. For that period it seemed like no matter how I poured carbs into her she would barely bump up and then come right back down as if her liver were just draining her blood of its glucose to replenish its own stores. Practical applications? Well, if your child ever chooses to fast for some reason, lower basals way down and underbolus for the next 36 hours. But I think this can also apply to unintentional fasts like when one has a stomach bug or when glucagon has been used for a low. Either one will rob the liver of its glucose and the evidence from our mini experiment is that it really did take 36 hours to replenish but that when it was done, it was done, just like that and her numbers were back to normal.