REF-1793 · Zeng et al., 2026 · Scandinavian Journal of Medicine & Science in Sports, · Paper · PubMed

In 40 highly trained male sprinters randomised to two kinds of maximal-intent lifting at 50-55% 1RM for 6 weeks, both two-way ballistic and jump-squat training improved jump height, 30-m sprint and 6-s cycle sprint, with no between-group difference reported for these; ballistic training gained more early isometric pulling impulse and jump squats more reactive strength (no non-lifting control group)

Main caveat: Two active arms and no non-training control: the shared improvements cannot be separated from regular training, the season or test learning.
moderate evidenceMeasured on highly trained athletesRead from the abstract only

Zeng and colleagues asked whether two ways of lifting at maximal intended velocity steer adaptation differently. Forty highly trained male sprinters were randomised to two-way ballistic training or to jump squats, each done as 5 sets of 8 at 50-55% of 1RM, 12 sessions over 6 weeks.

Before and after, the team measured countermovement jump (CMJ) and repeated CMJ height, 30-m sprint, isometric mid-thigh pull (IMTP), a 6-s cycle sprint, vastus lateralis architecture and the activation of vastus lateralis and biceps femoris during repeated jumps.

Both groups improved CMJ, repeated-jump height, 30-m sprint and 6-s cycle performance, with similar changes in muscle architecture. Ballistic training improved IMTP impulse over the first 100 ms more, jump squats improved the modified reactive strength index more, and vastus lateralis activation rose more after ballistic training at take-off and landing.

The authors read this as two routes to explosive performance and advise choosing the movement by the quality to be trained.

Participants

40 highly trained male sprinters, labelled Tier 3 by the authors; 20 per arm.
Age, event, personal bests, training history and baseline strength: not in the abstract.

Protocol

Both arms: 12 sessions over 6 weeks, 5 × 8 repetitions at 50-55% 1RM, maximal intended velocity.
TWB: "two-way ballistic" movement; the exact exercise and how it differs from a jump squat are not described in the abstract. JS: jump squats.
Tests before and after: CMJ, consecutive CMJ (CCMJ) height, 30-m sprint, IMTP (including 100-ms impulse), 6-s cycle sprint, VL muscle architecture, VL and BF surface EMG during CCMJ.
Not in the abstract: what other training the sprinters did during the 6 weeks, the season phase, blinding of testers, dropouts, and the sizes of the within-group changes.

Key Results

OutcomeResult
CMJ, CCMJ height, 30-m sprint, 6-s cycle sprintimproved in both groups (p < 0.05); no between-group difference reported in the abstract; magnitudes not given
VL muscle architecturesimilar increases in both groups
IMTP 100-ms impulsegreater improvement after TWB (ηp² = 0.25)
Modified reactive strength index (RSImod)greater improvement after JS (ηp² = 0.11)
VL activation, take-offincreased more after TWB (p < 0.001, ηp² = 0.30)
VL activation, landingincreased more after TWB (p < 0.001, ηp² = 0.58)
No percentage changes, sprint times or jump heights are given in the abstract.

Practical Application

For a sprinter, ballistic lifting and jump squats at 50-55% 1RM both went with better 30-m sprint and jump results over 6 weeks; the choice between them shifted which quality moved (early force for ballistic work, reactive strength for jump squats), not whether speed improved. One trial in trained male sprinters, without a group that did no extra lifting.
Within-group gains are not proof the lifting caused them. With no control arm, normal sprint training, the season and learning the tests could explain part of the improvement.
Inference, not tested here: if a sprinter's weakness is early force (a slow start), ballistic work fits; if it is short ground contacts, jump squats fit. The differences were in force-plate and jump measures and EMG, not in sprint time.
This sits on top of the base's sprint plan, not in place of it: SYNTH-59 keeps sprinting as the core and adds heavy strength and jumps; this trial tested only light, fast lifting and does not compare it with heavy strength work.

Limitations

Two active arms and no non-training control: the shared improvements cannot be separated from regular training, the season or test learning.
Men only, one 6-week block, 12 sessions; long-term or competition effects not measured.
EMG and architecture are markers; the between-group differences in performance were in IMTP impulse and RSImod, not in sprint time.
The ballistic movement, co-training, season phase, blinding and dropouts are not described in the abstract; no magnitudes or intervals are given for the main performance outcomes.
Written from the abstract and the PubMed record; full text not read.
How this entry was checked and changed
Why this level

a parallel RCT with 20 per arm read from the abstract: risk of bias not read (abstract only); the within-group gains have no non-training control, and the EMG and architecture results are markers

In 40 highly trained male sprinters randomised to two kinds of maximal-intent lifting at 50-55% 1RM for 6 weeks, both two-way ballistic and jump-squat training improved jump height, 30-m sprint and 6-s cycle sprint, with no between-group difference reported for these; ballistic training gained more early isometric pulling impulse and jump squats more reactive strength (no non-lifting control group) — BLOCKS Science · Blocks