Considering each swimmer individually, a positive correlation was observed between the hip and CM values regarding velocity (ranging from 0.50 to 0.83), which is in accordance with Maglischo et al. (1987) in front crawl technique sellckchem (values between 0.86 and 0.96, with a mean coefficient of 0.87). These data, associated with the obtained high digitize-redigitize reliability values, evidence that, although there is an associated error that should be taken into account, the hip reflects satisfactorily the CM motion in front crawl when swimming at moderate intensity. The velocity to time curve obtained for one swimmer for both CM and hip showed similar patterns of positive and negative accelerations as described in the literature (Maglischo et al., 1987; Craig et al.

, 2006): both CM and hip decelerated during the downsweep phases (that are coincident with the recovery of the opposite arm) and in the transition from one propulsive phase to another, and both body points accelerated during the catch, insweep and upsweep phases. Thus, coaches should incorporate specific training drills aiming to perform faster transitions between propulsive phases, as well as to finish the stroke at maximal arm velocity. It was also evident that swimmers choose a catch-up inter-arm coordination mode that is typical of moderate paces due to a long gliding phase (Schnitzler et al., 2008; Seifert and Chollet, 2009; Seifert et al., 2010). In fact, the existence of a discontinuity between the end of the propulsion of one arm and the beginning of propulsion of the other arm is typical of front crawl swimming at moderate intensities (Seifert and Chollet, 2009; Seifert et al.

, 2010). Thus, coaches should not advise swimmers to adopt superposition arm synchronization when implementing aerobic pace training series. Furthermore, it was also evidenced that the hip presents higher and lower forward velocity peaks magnitude compared to CM, as shown by Maglischo et al. (1987) for higher swimming intensities. Notwithstanding that the forward velocity and displacement of the hip and CM are similar, and the evidence that the IVV determination using the hip is reliable, allows multiple cycles to be evaluated and enables the assessment of fatigue (Holm��r, 1979; Maglischo et al., 1987), differences between hip and CM were found for the IVV. Such differences corroborates the literature (Figueiredo et al.

, 2009), and might be explained by the inter-segmental actions during the front crawl swimming cycle that frequently changes the CM position (Barbosa et al., 2003). In addition, the CM vmax and vmin values seem to be over and underestimated (respectively) by the hip values, as previously proposed by Psycharakis and Sanders (2009). In fact, when the arms in front crawl accelerate the body Cilengitide mass, they simultaneously move backwards with respect to a body fix landmark refraining the acceleration of the CM.