| Abstrakt: |
Parsing the relative contributions of fitness and fatness to various disease states has long interested investigators, despite their almost certainly intertwined causal relationships. In this issue of the Journal, Becofsky et al. (Am J Epidemiol. 2015;181(5):311–320) examined these constructs in relation to depressive symptoms assessed using the Center for Epidemiologic Studies Depression Scale, a widely used and well-validated instrument that ascertains several domains related to depression. The authors found stronger and more consistent associations of elevated depressive symptoms with objectively measured fitness than with any measure of fatness. These analyses highlight a serious concern in analyses of risk factors for complex clinical constructs—the likelihood of differential misclassification with respect to the risk factors of interest in scales with multiple, wide-ranging domains. In this interesting example, the multi-item Center for Epidemiologic Studies Depression Scale, which validly captures the multiple domains of depression, contains measures of psychomotor retardation that could easily reflect fitness itself. Because an instrument's validity is uncertain if selected items are removed and gold-standard measures of complex clinical constructs are rarely truly gold, options to remedy this problem in most epidemiologic studies are few. Nonetheless, their analysis illustrates the particular caution needed when examining exposures that plausibly relate to misclassification in multidimensional outcomes of interest. [ABSTRACT FROM PUBLISHER] |
