Am J Physiol. 1999 Aug;277(2 Pt 2):R441-54.
This paper started off with the first two pages as a review of different
techniques which had previously been used to mathematically describe baroreflex
curves measuring arterial pressure against nerve activity and heart rate. It
then went on to describe a new method, a 5-parameter equation, that was predicted to be able to account for asymmetry
sometimes seen in baroreflex curves. The argument in favor of this new method
was that it would be able to fit a curve better when the curve showed asymmetry
between upper and lower limits, or differences in slope on either side of the
midpoint. If no such asymmetry existed, the 5th parameter was averaged out and
the formula was able to approximate the curve the same way a traditional
4-parameter equation would.
The paper then went on to analyze pre-existing baroreflex data of MAP vs heart rate
and renal nerve sympathetic activity in rabbits and dogs via 4- and 5-
parameter curves - in both cases, they tested by forcing the curves to fit to
the resting points and tested again without forcing them. They found that when there was an
asymmetry in the curves of some subjects, the 5-parameter equation could
account for it, though for the most part the 5-parameter method did not report
values that were significantly different than the 4-parameter method. The biggest changes seemed to be caused by
forcing the curve through the resting point, and according to this paper, doing
that seems like a questionable technique. The paper did make a good point that even
though this technique may not result in significantly different results over
the 4-parameter technique in most cases, in cases where asymmetry is present,
it will yield a more accurate result - and if you need it for one parameter,
you should use the same technique for all parameters within a study. I think that's a pretty
fair argument. -DH
No comments:
Post a Comment