We respectfully disagree with both issues raised by Shaibi, Cruz, and Goran (1). First, we will address the comment that “associations between fitness and activity are weak at best.” The review article referenced by Shaibi, Cruz, and Goran was based on observational studies published before 1995, a time at which physical activity was measured almost exclusively by questionnaire. Since 1995, several studies have shown that questionnaire measures of physical activity in youngsters only modestly correlate with direct measures obtained by accelerometry (2,3) and, more importantly, that accelerometry-based measures of physical activity relate to cardiorespiratory fitness (CRF) in the pediatric population (4,5). Furthermore, intervention studies have reported significant improvements in CRF in response to physical activity programs in children and youth (6,7). Thus, while we agree that CRF is in part accounted for by genetics, current evidence also substantiates our claim that CRF is largely influenced by physical activity.
We would also like to comment on the contention that CRF would not have been an independent predictor of the metabolic syndrome had we controlled for adiposity. Recall that one of the five components of the metabolic syndrome as defined in our study was abdominal obesity (e.g., high waist circumference) (8). Because adiposity itself was a component of the outcome measure, from a statistical standpoint it would have been inappropriate to control for part of the outcome measure (adiposity) when examining the relation between the exposure (CRF) and outcome (metabolic syndrome). Thus, while we agree that adiposity, particularly abdominal fat, is a central feature of the metabolic syndrome, we disagree with the flawed rationale of Shaibi, Cruz, and Goran that CRF should not be considered an independent predictor of metabolic syndrome. In fact, we suggest that the effect of physical activity and CRF on adiposity is one of the primary pathways by which these exposures influence the metabolic syndrome.