I was doing a search for something else when I came accross an interesting study. 74 people without diabetes wore Dexcom sensors for a week. Lots of data was collected.
To me the most strange and interesting tidbit is that the youngest group had average sensor glucose values a full standard deviation above that of the oldest group, even though they had lower HbA1cs. The full data on this study would be interesting for all of us trying to figure out what blood sugars are healthy. Unfortunately, for many of the data points only the medians and interquartile ranges are given rather than the full ranges. That means the highest and lowest fourth of the people are ignored for many of these data points. Please remember that when looking at this data! Here is the study (click on results in the tabs at the top if that's not how it opens): http://clinicaltrials.gov/ct2/show/results/NCT00717977?term=dexcom&rank=15§=X6015#outcome1