REF-1795 · Xie et al., 2026 · International Journal of Sports Medicine · Paper · PubMed

Complex training (heavy lifts paired with jumps) beat traditional strength training on countermovement jump (ES 0.58, 62 studies), squat jump (0.77), agility (−1.05, faster) and sprint (−0.65, 50 studies) in a three-level meta-analysis of 75 randomised trials in trained and developmental athletes, with moderate certainty by GRADE; the dose optima come from between-study meta-regression

Main caveat: Heterogeneity moderate to substantial for three of four outcomes (I² 56-69%); confidence intervals for the main effects are not in the abstract.
fairly strong evidenceMeasured on trained athletesRead from the abstract only

Xie and colleagues pooled randomised trials that compared complex training, in which heavy resistance exercise is paired with explosive or plyometric work, with traditional strength training in athletes who train three or more times a week and compete at local level.

Searching five databases to June 2025, they included 75 studies and used three-level meta-analysis, moderator analyses, dose-response meta-regression and GRADE.

Complex training improved countermovement jump (effect size 0.58, 62 studies), squat jump (0.77, 23 studies), agility (−1.05, 29 studies; negative means faster) and sprint (−0.65, 50 studies) more than traditional strength training. Effects were larger in technical or composite-sport athletes and in younger athletes.

Dose relationships were non-linear: the jump effect peaked at "1 rep minimum" and "14 repetitions maximum" (the abstract's terms, not explained), and agility and sprint peaked around six sets and 12 repetitions. The authors rate the certainty as moderate and recommend complex training for higher-level athletes, with higher volumes for technical-sport and younger athletes.

Participants

75 randomised trials; trained or developmental athletes training three or more times a week and competing at local level.
Total participants, sexes, ages and sports: not in the abstract (technical/composite-sport and younger athletes appear as subgroups).

Protocol

Search: five databases, updated June 2025.
Inclusion: RCTs of complex training lasting at least 4 weeks against traditional strength training, with jump, agility or sprint outcomes.
Analysis: three-level meta-analysis; moderators (athlete type, age among them); dose-response meta-regression on weeks, sets and repetitions; GRADE certainty.
Not in the abstract: the risk-of-bias tool and its results, the exact definition of complex training used, confidence intervals for the main pooled effects, and whether the two programmes were matched for volume.

Key Results

Outcome (complex vs traditional strength training)StudiesEffect sizepI²
Countermovement jump620.58< 0.0133.99%
Squat jump230.77< 0.0168.59%
Agility (negative = faster)29−1.05< 0.0163.97%
Sprint (negative = faster)50−0.65< 0.0556.04%

Dose-response peaks (standardised mean difference, 95% CI):

OutcomeDoseSMD (95% CI)
Countermovement jump"1 rep minimum"1.69 (0.56 to 2.81)
Countermovement jump"14 repetitions maximum"1.36 (0.34 to 2.38)
Agilitysix sets−1.60 (−2.37 to −0.83)
Sprintsix sets−1.28 (−2.01 to −0.57)
Sprint12 reps ("min.rep")−1.33 (−2.26 to −0.40)
Larger effects in technical/composite-sport athletes and in younger athletes; sizes not in the abstract.
Certainty: moderate (GRADE); the abstract gives one overall rating, not one per outcome.

Practical Application

For athletes who need jump, sprint or change-of-direction speed, pairing heavy lifts with jumps gave moderate-to-large extra gains over lifting alone (CMJ 0.58, squat jump 0.77, sprint −0.65, agility −1.05) across 75 trials, with moderate certainty. Measured in trained and developmental athletes of sports the abstract does not list; no endurance outcome was pooled.
This agrees with the base's sprint guide: SYNTH-59 feeds sprint speed with heavy strength and jumps, and this pool says combining them in one session beat strength alone for jumps and sprints in athletes; it gives no sprinter-only estimate.
Treat the dose numbers as a starting range, not a prescription. Around six sets and 12 repetitions went with the largest sprint and agility gains, but these are comparisons between studies, and the repetition terms for the jump peaks are not explained in the abstract.
Inference, not tested here: for a runner or cyclist the jump and sprint gains do not show faster endurance racing; the endurance dose of strength and jumps stays with SYNTH-44 and SYNTH-45.

Limitations

Heterogeneity moderate to substantial for three of four outcomes (I² 56-69%); confidence intervals for the main effects are not in the abstract.
The dose optima and the athlete-type and age moderators are between-study associations, not tested doses; the abstract's repetition terms for the jump peaks are unclear, and its sentence on moderators is garbled.
The risk-of-bias tool, the number of participants, sexes, sports and whether programmes were volume-matched are not in the abstract.
The recommendation for "higher-level athletes" goes beyond the inclusion criterion of local-level competition.
Written from the abstract and the PubMed record; full text not read.
How this entry was checked and changed
Why this level

a meta-analysis of RCTs whose abstract reports its GRADE certainty and the result (moderate), so the abstract-only cap does not apply; held at A- for moderate-to-substantial heterogeneity (I² 34-69%) and dose findings that are between-study meta-regressions