That has been puzzling me too. In fact if the error is as you suggest,
it is off by a factor of a hundred, not ten. It has not yet been
published on paper. Its status is claimed to be that of corrected
electronic draft released on the web prior to paper
publication. That's an odd error to escape peer review and correction,
and there's still time for further correction before it reaches hard
publication status.

It could be the author meant the size of the range rather than its end
points, a common ambiguity. Even so, there's still an interesting size
discrepancy between that and other sources. IIRC ranges of size
between 5mg/dl and 15 mg/dl are often quoted. The large range you
quote is often suspected of being an average contaminated by
undiagnosed diabetes and impaired glucose response rather than a
normal figure, i.e. the range of undamaged blood glucose homeostasis.

The full text requires subscription. I'll find out tomorrow if my
university account has access privileges.