I think that Atkins does not do a good job will full disclosure on the effects of coffee. The evidence that caffeine causes insulin spikes is inconclusive. The only study that gets cited is rather out of date. And I think that the wide sample of folks on this site who drink coffee (black or with heavy cream) speaks to the fact that any spiking is probably minor, if at all. You just need to experiment yourself.