Letters to the Editor
Department of Health Services Research; Palo Alto Medical Foundation Research Institute; Palo Alto, California; email@example.com (Knowles)
Department of Epidemiology; University of North Carolina; School of Public Health; Chapel Hill, North Carolina (Marshall, Millikan)
The authors respond:
We thank Sergentanis and Mariolis for their thoughtful comments1 and their suggestion to investigate age and years of playing experience as mutual effect modifiers. We have conducted an analysis evaluating the joint role of age and years of playing experience. The increase in the overall injury rate due to increasing age was slightly smaller for players with more playing experience (rate ratio = 1.14 [95% confidence interval = 0.89-1.47]) than for those with less playing experience (1.35 [0.86-2.13]) (Table 1). Any modification in the overall injury rate ratio is small, relative to the imprecision of the estimates. We assessed modification based on departure from pure additivity of effects.
We also stratified this analysis by a third factor, evaluating these effects separately for games and practices. An interesting picture emerges. For athletes with less experience, the effect of increasing age on injury occurred during practices (1.87 [0.82-4.26]) and not during games (0.97 [0.46-2.04]). For athletes with more experience, the converse is true: the effect of increasing age on injury was during games (1.40 [1.03-1.89]), rather than during practices (0.79 [0.48-1.29]).
As one would expect, the same result was observed when we examined age as a modifier of the effect of playing experience on injury risk, with older players showing an effect of increased experience in games (1.17 [1.08-1.25]) but not in practices (0.92 [0.83-1.02]). These results, though not strong, suggest that age is an effect modifier. One explanation for these results is that older athletes who are less experienced feel under greater pressure to secure a place on the starting roster for the team, and consequently play with greater intensity during practices (thereby exposing themselves to greater injury hazards). Inexperienced athletes who are younger may not feel the same pressure during practices. Furthermore, older athletes who have more years of experience with the sport are more likely to be “playmakers,” and probably receive more playing time during a game and are involved in a higher proportion of plays in each game.
Although there is close correlation between age and playing experience, these analyses suggest that these factors might combine synergistically or antagonistically, depending on whether it is a game or a practice session. Thus, there is potentially a 3-way interaction among game/practice, age, and playing experience.
Given the imprecise estimates and overlapping 95% confidence intervals, these conclusions should be seen as tentative. As sports injury research continues to evolve, the investigation of effect modification may provide additional insights for injury prevention.
Sarah B. Knowles
Department of Health Services Research
Palo Alto Medical Foundation Research Institute
Palo Alto, California
Stephen W. Marshall
Department of Epidemiology
University of North Carolina
School of Public Health
Chapel Hill, North Carolina
1. Sergantanis TN, Mariolis A. Interaction of playing experience and age. Epidemiology. 2009;20:622.
© 2009 Lippincott Williams & Wilkins, Inc.