Background The side-hop test by Gustavsson et al. (2006) helps with return to play decision making after lower extremity injury.

Objective Determine whether a reduced width of 30cm (compared with 40cm) may be more suitable for use in developmental athletes.

Design Cross-sectional study.

Setting School gymnasium.

Participants 18 developmental athletes, mean age 13.93 ± 1.37 years, (range 11.9 - 16.3 years), all members of an elite athletics project.

Assessment Each participant performed a 40cm and a 30cm side-hop test for each leg. Athletes were tested during a regular training session, sufficient resting time (≥ 4min) granted after each trial. All tests conducted within 10 days. Tests were supervised and evaluated by two coaches, one counting number of jumps during the trial with a customary hand-held mechanical clicker, the other one filming each attempt with a latest generation smartphone high-speed camera. Errors click-counted on home PC with videos running at half speed and error-rate calculated.

Main Outcome Measurements Comparison of total number of jumps and error-rates for both distances. Evaluation of correlations with time to/from peak height velocity (PHV), height, and chronological age.

Results Lower number of jumps for 40cm (59.33 ± 8.66) vs. 30cm (66.8 ± 9.91), but higher error-rate for 40cm (0.27 ± 0.13) compared with 30cm (0.22 ± 0.12). Paired t-tests show significant differences (p<0.02) for both. Based on cohen’s d, effect is large for comparing number of hops (d=0.81), but small when comparing error-rates (d=0.40). Pearsons’s correlations of error-rates with age, height, and PHV are stronger for 30cm (r_age=-0.664; r_height=-0.344; r_PHV=-0.351), than for 40cm (r_age=-0.537; r_height=-0.145; r_PHV=-0.155), however, respective effect sizes based on cohen’s q are small (q_age=0.201; q_height=0.213; q_PHV=0.210).

Conclusions While results remain statistically inconclusive when comparing 40 cm and 30 cm hop-tests in developmental athletes, 30 cm width may be favourable in a clinical context due to higher number of jumps and lower error rates. Further research is warranted to provide clarity.