Letter Regarding Article by Christou et al, “Fatness Is a Better Predictor of Cardiovascular Disease Risk Factor Profile Than Aerobic Fitness in Healthy Men”
To the Editor:
Christou et al1 postulated that fatness is a better predictor of cardiovascular disease (CVD) risk factors than aerobic fitness. The interpretations of the data may be misleading. Theoretically, 2 risk factors can be (1) independently associated with a disease or (2) may be different steps along the same causal pathway. In a chain of causation in which one factor affects another factor, which eventually leads to disease, we must interpret the analyses with caution. An example could be fatness leading to atherosclerosis, which could lead to CVD. If we adjust for the intermediate factor (atherosclerosis), then the effect of fatness disappears completely if the effect of fatness is mediated entirely through this factor. If the effect of fatness was mediated through mechanisms other than atherosclerosis as well, only part of the association between fatness and CVD would disappear.2 The fact that the 6 figures in the article showing the association between fatness and the risk factors with and without adjustment did not change when adjusted for fitness strongly supports the hypothesis that fatness is an intermediate factor closer to the end point in a causal chain from fitness to the risk factor; however, the authors came to the opposite conclusion.
Another concern with the study is a mean fitness level of 42 mL · min−1 · kg−1, with few subjects below 30 mL · min−1 · kg−1. The association between fitness and CVD risk factors is not linear, but risk increases steeply when fitness decreases below 30 to 35 mL · min−1 · kg−1.3 Because individuals with elevated risk factor levels were excluded, they also excluded individuals for whom the effect of physical fitness was apparent.
Accordingly, the 3 implications the authors mentioned should be altered. (1) “First, men who are overweight or obese should be encouraged to reduce body fatness, regardless of their aerobic fitness.” Low fitness may be the principal cause of fatness, mediated through insulin resistance. (2) “Second, weight management and prevention of excess adiposity should be a primary cardiovascular health goal for men.” We would add that this could be done by staying active and insulin sensitive. (3) “Third, in lean trained men, the cardioprotective influence of habitual physical activity may be mediated in part through the maintenance of optimal body weight and fatness.” Studies analyzing changes in physical activity or fitness in relation to CVD risk or mortality have shown a lower CVD rate among subjects who increase levels of activity or fitness that is independent of weight change.4,5
In conclusion, low fitness and fatness may both be important predictors of CVD risk factors.
Andersen and colleagues make several statements regarding our article published in Circulation,1 and we would like to address each point separately.
First, the authors state that we “postulated” that fatness is a better predictor of cardiovascular disease (CVD) risk factors than aerobic fitness. This is an inaccurate précis of our article. We stated no such hypothesis. Rather, our stated purpose was to determine which of these putative influences better predicts traditional CVD risk factors in healthy men. We took no a priori position on this question.
Second, the authors state that the interpretation of our data may be misleading because 2 risk factors can either be independently associated with a disease or be different steps of the same causal pathway. Our analysis was designed to examine the independent association of fitness (or fatness) with CVD risk factors, while partialling out the effects of fatness (or fitness) and age. Our study design did not allow determination of the sequence of events in a causal pathway, as suggested by Andersen and colleagues.
Third, the authors state that the mean V̇o2max of 42 mL · min−1 · kg−1 is of concern. Our subject cohort included men with a broad fitness range, including highly trained competitive athletes. As such, a mean V̇o2max of 42 mL · min−1 · kg−1 is not remarkable. The authors also state that few subjects had a V̇o2max below 30 mL · min−1 · kg−1 and that risk increases steeply below 30 to 35 mL · min−1 · kg−1. In fact, 15% of our participants had a V̇o2max below 30 mL · min−1 · kg−1, and almost 40% of the subjects led sedentary lifestyles. The statement by Andersen and colleagues that risk increases steeply below a V̇o2max of 30 to 35 mL · min−1 · kg−1 also is misleading. This conclusion is drawn from a study2 conducted on adolescent boys and girls in which body fatness was not measured, and both V̇o2max (from submaximal exercise heart rates) and risk factors (from questionnaires or field tests) were estimated.
Fourth, the authors state that “[b]ecause individuals with elevated risk factor levels were excluded, they also excluded individuals for whom the effect of physical fitness was apparent,” inferring that our analysis may have been biased by subject exclusion criteria. Again, our analysis included subjects with a broad range of fitness levels and a large proportion of sedentary men. In addition, we clearly stated the limitation that our findings should be restricted to healthy men (top of page 1913).
Finally, Andersen and colleagues advise that we change our conclusion on the basis of their interpretation of our data. In particular, they emphasize the need to include information about the insulin-sensitizing effect of physical activity. However, on page 1911 of our article under the subheading “Aerobic Fitness and CVD Risk Factor Profile,” we discuss this issue and state, “These observations support the need for regular aerobic exercise as an important strategy to promote optimal insulin sensitivity.”
Overall, it appears that a central underlying concern of Andersen and colleagues is that the conclusions of our study underestimate and/or underemphasize the importance of physical activity/aerobic fitness in controlling CVD risk factors. This is not the case. The primary take-home message from our article is not that habitual physical activity and fitness are unimportant, but rather that body fatness should not be ignored as a key determinant of the CVD risk factor profile.