Chasing the top quartile of cross-sectional data: Is it possible with resistance training?

Resistance exercise is believed to be important for improving several biomarkers of health and contributing to decreased all-cause mortality. However, a careful examination of the literature reveals a growing body of cross-sectional studies finding stronger associations between strength independent of physical activity on these various outcomes and biomarkers (i.e., C - reactive protein, homocysteine, total cholesterol) of health. Considering that engagement in resistance type exercise is associated with an increase in strength, such cross-sectional data is often used to promote engagement in resistance training. When one examines the cross-sectional studies together with the experimental work, however, this relationship becomes less clear. For example, being in a higher tertile or quartile for grip strength may be associated with more favorable biomarkers or health outcomes, yet, experimental work shows that grip strength does not change with traditional resistance exercise. This seems to suggest that our interpretation and application of this literature may be incorrect. It is our opinion that much of the cross-sectional work may more accurately illustrate inherent differences between individuals who are "weak" and individuals who are "strong". In addition, there is conflicting experimental evidence and a lack of longitudinal support that resistance exercise may have a meaningful effect on long-term health outcomes. Using this evidence, we present our "human baseline hypothesis", which proposes that the baseline value of strength that a person possesses prior to training may be a more appropriate indicator regarding long-term health outcomes than the act of training itself.