I'm looking for references to PROSPECTIVE scientific studies regarding
what forms of exercise in healthy non-obese men have the greatest
health benefits.

A prospective study is one in which the subjects are divided into
groups, and, for example, each group is assigned to (a) swim, (b) jog,
or (c) work out on exercise machines.  Alternatively, the study might
prospectively compare swimming 20 minutes a day to swimming 30 minutes
a day.

I am not interested in retrospective studies (i.e.,  examining the
difference in life expectancy between people who choose to be
swimmers, and people who choose to jog).

I am also not interested in articles that are largely based on
hypothesis (e.g., swimming is the best exercise because it is not
traumatic to the joints), or simple recommendations by the American
Heart Association that say you should eat right and exercise 3-4 times
a week.  IMHO, most specific recommendations for exercise are based on
hypothesis rather than evidence of decreased mortality or morbidity.

Ideally, the end point of the study should be a real-life measure,
such as increased life expectancy, or decrease in heart attack rates. 
Less ideal, but also good, would be changes in resting heart rate,
blood lipid levels, or blood pressure.  I'm not interested in exercise
as a form of weight loss.

Thanks!
dan89