Is training age predictive of physiological performance changes in developmental rugby league players? A prospective longitudinal study

Abstract

The aim of this study was to determine if training age – rugby league training age, resistance training age and cardiovascular training age – was predictive of changes in physiological performance in developmental rugby league players. Developmental rugby league players from one elite club ( n = 147, mean age = 15.8 ± 1.1 years) playing at Under 16 and Under 18 representative levels were tracked longitudinally over two consecutive seasons. Estimated aerobic capacity, jump height, upper body strength endurance and change of direction were assessed. Mixed method random effect analysis determined relationships between training age domains and physiological change. Positive relationships were observed between rugby league training age and change in estimated V̇O2max ( β = 0.13, p = 0.018), and change in strength endurance ( β = 0.51, p = 0.033), and resistance training age and change in jump height ( β = 0.11, p = 0.007). A negative relationship was observed between rugby league training age with change in change of direction time ( β = –0.06, p = 0.026). Moderate pre-season training effect sizes were identified for change of direction (effect size = –0.42, confidence interval (CI) = –0.8 to 0.1) and jump height (effect size = 0.41, CI = 0.1–0.7), while small pre-season training effect size was apparent for strength endurance (effect size = 0.33, CI = 0.0–0.7) and estimated V̇O2max (effect size = 0.16). Small in-season performance decrements occurred in strength endurance (effect size = –0.23, CI = –0.6 to 0.1) and change of direction (effect size = 0.30, CI = –0.1 to 0.7). Higher pre-study rugby league training age was associated with greater improvements in estimated V̇O2max, strength endurance and change of direction. Additionally, higher resistance training age was associated with greater improvements in vertical jump. Overall, the magnitude of resistance training age and cardiovascular training age may have been too inadequate to meaningfully impact training adaptations in the population. Decrease in effect size for physiological performance changes during the in-season phase suggests that performance enhancements were not maintained across the competitive season.

Statistical analysis considerations within longitudinal studies of physical qualities in youth athletes: A qualitative systematic methodological review

“…Six articles also reported a range of participant observations of three or more [ 34 , 36 , 47 , 51 , 65 , 67 ]. The remaining articles either confirmed that all participants were observed at all time points [ 13 , 28 , 30 , 33 , 35 , 46 , 56 , 58 , 59 , 64 , 66 ], or the number of participant observations was not stated [ 29 , 31 , 32 , 37 , 42 , 48 , 49 , 57 ].…”

“…Other reasons included within and between season differences, comparison of sub-groups over time [ 39 ], non-normal data [ 30 ] and longitudinal design [ 28 ]. Methods for supporting the justification included referencing a journal article ( n = 14; 35%) [ 29 , 31 – 34 , 42 , 46 , 47 , 51 , 52 , 61 – 63 , 65 ], the MLwiN software ( n = 5; 13%) [ 36 , 38 , 44 , 45 , 63 ], books ( n = 2) [ 28 , 54 ] and Hopkins website ( n = 1) [ 65 ].…”

“…This is a consequence of the specific research questions, rather than limitations of the analysis methods used. For example, Booth et al [ 31 ] used multilevel models to identify the influence of different training ages by including each type of training age individually within their own model. As such Booth et al [ 31 ] were able to identify the magnitude of change over time (regression slope) for each individual training age type but may have missed potential inter-relation effects between them.…”

“…For example, Carvalho et al [ 32 ] found both quadratic and cubic terms for age significant (p ≤ 0.01) in height and Yo-Yo intermittent recovery test level 1 models suggesting a non-linear relationship between time and physical qualities was present. Non-linear terms were not included within all articles making use of multilevel models [ 31 , 33 , 37 , 54 ], while others included the assessment of non-linear terms in the methods although they failed to reach significance (e.g. Valente Dos Santos et al [ 61 , 63 ] and skeletal age).…”

“…Consequently, the development of physical qualities was only assessed as a result of time (age grade or timepoint). However it is also apparent that a single independent variable was used where a multivariable analysis could be applied, with multilevel [ 28 , 31 ] and segmented linear modelling [ 42 ]. This is a consequence of the specific research questions, rather than limitations of the analysis methods used.…”