"Yo-Yo Dieting"

First and foremost, upon reading the news article, I was somewhat
mislead by the headline "Panel finds no major risk from 'yo-yo'
dieting."  Only in the 4th paragraph are readers informed that the
study was a meta-analysis of dozens of previously-published research
articles, and not an original study.  In and of itself, there is no
harm done, but it was not elaborated upon.  By saying "in reviewing
nearly 30 years of medical data," the news article seems to be
implying that a massive amount of data was researched and
categorized--and that it was all conducted properly with adequate
numbers of subjects.  However (as we soon discovered), there were only
43 studies researched, and many had few subjects.

The journal article then cast more doubts on the veracity of its
claims.  Table 1 appears to be an attempt to summarize the previous 30
years' studies, but the data is not presented in any comprehensible
way and do not take into account the vastly differing sample sizes,
from 4 to 10,529 (!)  The definition of "weight cycling" also needs
some clarification.  Although the researchers attempted to standardize
on a set of criteria, they admit that "such basic information as the
normal degree of weight fluctuation during short periods in
nonclinical populations is currently unknown."  Conceding such an
essential criteria and an also admitting that "not all confounding
variables could be adequately controlled" leaves me with some doubt as
to the credibility of such an analysis.  In addition, social
definitions of weight, obesity and weight cycling have changed over
the past 30 years, making direct comparison of data less viable.