June 28 - 29, 2007 Advisory Committee Meeting
Dr. Haskell introduced Dr. Lee to provide an overview of issues surrounding
all-cause mortality. Dr. Lee outlined 5 key issues to be discussed; 1) Issues
related to study design; 2) Response issues; 3) What activities were measured in
any particular study; 4) Quantity versus quality of physical activity, and 5)
Physical activity as cause of mortality.
Dr. Lee reiterated the conflict between observational studies and randomized
clinical trials. For many chronic disease outcomes there is no randomized
clinical trial data. Also, observational studies are susceptible to bias and
without supporting data one cannot infer cause and effect from the
epidemiological data. For purposes of physical activity randomized trials are
not always feasible and every possible pattern of activity in daily life
cannot be tested. Also, data from observational studies may be much more open
to interpretation and judgment depending on who actually adheres to the arm.
If observational studies are to be utilized it will be important for the
Committee to agree on uniform criteria that all members refer to when
Responses to dose is not necessarily equal regarding input and output
especially when comparing different population sub-groups as well as when a
particular study was conducted. For example, the benefits to an 80-year old
sedentary woman will not be the same for a 20-year old marathon runner from
the same dose. Also, there appears to be different physical activity baseline
thresholds from studies conducted in the 1950's versus studies conducted now.
What Activities Where Measured?
When looking at a particular study it will be important to understand what
activities were measured. If one study records leisure time activities versus
commuting physical activity you may see a difference in how many people meet
specific amounts of recommended physical activity. This will not necessarily
mean one group is more active then the other.
Quantity vs. Quality
When looking at the issue of quantity versus quality in randomized
clinical trials versus observational studies it is clear there is clear
measurable data in randomized trials that will make it easier to interpret the
effects of quantity versus quality. In observational trials it may not be
clear. If one study says vigorous activity was better activity than moderate
activity, it really might not be that the intensity was important but the
total amount of activity that was important. In order to really look at this
issue one must do specific analysis that control for energy expenditure which
many chronic disease studies do not do.
Cause of Mortality
We now have data clearly showing that during the time that you exercise,
as well as a short period afterwards, vigorous physical activity can
precipitate sudden death, particularly in people who are not physically
active. Because only the most active exercise for a short portion of the day,
risk is increased during the short portion of the day, but decreased the
remainder of the day, when not exercising vigorously. Overall, we see a
protective effect of physical activity in all-cause mortality, but a clear
increase in risk over that period. While a small risk it should be explicitly