Figuring out the various studies that come out on the health effects of
this and that food is practically impossible. There are also many
contradicting studies. A lot of that has to do with the methodology of
comperative studies.

Generally, what they do is to solicit a group of people to participate
and then over a period of time they feed half of them with say cabbage
and the other half ("the control group") don't eat cabbage. Then after
a period of time they measure either the incidents of heart deasease or
some auxilliary factor (blood pressure, etc.) and using statistical
sampling theory conclude that there is a significant difference in the
two populations.

Obviously the conclusiveness of such studies will be all over the place
and you probably want to see more than one study before you stop eating
cabbage or load up on sauerkraut. Many factors can influence the
outcome of a study and not all of those can be controlled.

In addition, statistical sampling theory does come with its own
limitations on conclusiveness. Namely there will always be some
non-zero probability that the result was strictly random. Most studies
keep this probability under 5% but in calculating the significance of
the study, any researcher will have to use a number of assumptions that
may or may not be correct. There are well known examples of meaningless
statistically conclusive data such as the fact that the length of the
right and left foot are not statistically correlated.

Finally, many of these studies are sponsored by manufacturers of this
and that and even if the study was done by a third party research unit,
you still have to wonder.