Apparently the studies found no link between prostate cancer and calcium
intake, as some previous studies had found.  But one of the studies
found a difference depending on how the calcium was taken.  In one case
the comparison was skim milk vs. whole milk and the other it was skim
milk vs. total calcium from other sources, I think.

One general comment.  If a study is done with one purpose and then the
group is further subdivided into separate groups looking for possible
differences, it is possible to find spurious relations.  Many
statisticians reject outright any such subsequent subgroup analysis
unless the study was carefully designed with the subgroups in mind.

I don't know to what extent this objections applies here.  I would say
that it is premature to change one's dietary habits on the basis of
these results.  We would have to await confirmation with more studies.

The basic problem is that unless you have a biological mechanism in
mind, studies which show a statistical association can be misleading
unless the results are so strong that there is little doubt.  For
example, years before a biological mechanism for how smoking causes lung
cancer was established, the fact that the lung cancer rates among
smokers was so much higher than that for nonsmokers clearly established
the risk.   On the other hand a modest association allows for other
plausible conclusions.

Perhaps I should add that I've been using nonfat milk products for
years.  I suppose that is it possible it increased my risk of prostate
cancer, but personally I doubt it.