A thorough literature search was carried out across multiple databases, namely Web of Science, PubMed (Pubmed Central), Scopus, Elsevier, and Springer. The search imposed no constraint on the publication date; however, it was confined to articles published in English. To pinpoint relevant studies, Boolean search logic was utilized, with specific search terms crafted around weightlifting training, plyometric training, traditional resistance training, and performance outcomes. The supplementary material holds a detailed account of the search strategy. Subsequently, two researchers independently went through the titles and abstracts of the identified studies to assess their relevance. Full - text versions of studies that appeared eligible were obtained and evaluated against the pre - set inclusion criteria. In cases where the two reviewers disagreed, discussions were held to iron out the differences.

Inclusion criteria for the literature were established based on the PICOS (Participants, Interventions, Comparators, Outcomes, and Study Design) guidelines. Studies involving healthy male or female participants were selected without restrictions on publication dates.

The participants included in this systematic review were individuals who maintained good health throughout the experimental process, with no restrictions on ethnicity, nationality, or sex. Age restrictions are limited to early adulthood (18–30 years old). Studies were excluded if they involved participants with chronic diseases or if they included animal subjects, which were identified and removed during the title and abstract screening phase.

Studies must have compared at least two of the following training modalities: weightlifting training (WT), traditional resistance training (TT), or plyometric training (PT). The intervention period was required to be a minimum of 6 weeks, with at least two training session per week, totaling at least twelve sessions.

The control group performed resistance training or the same regular interventions as the intervention group. For the network meta-analysis, we defined comparison groups based on the interventions studied. In this network, TT was designated as the reference treatment to allow for the quantification of relative treatment effects between PT, WT, and TT. Studies included in the network either compared one of these interventions against TT, or compared PT directly against WT.

Studies were required to report on at least one performance measure, including lower-limb maximum strength (1RM squat), jumping performance (squat jump, countermovement jump), or sprint performance (short-distance sprint).

Only randomized controlled trials (RCTs) published in English were included. For studies based on the same sample, only one article was included to avoid duplication.

Exclusion criteria included duplicate publications, insufficient data availability despite attempts to contact authors, and studies with significant methodological flaws or poor testing protocols. Studies failing to meet the minimum quality standard of 3 points on the PEDro scale.

Adhering to the PRISMA guidelines for literature selection and data extraction, the data collection form is based on the extraction form modified from the “Cochrane Handbook for Systematic Reviews of Interventions” (Higgins et al., 2019). Extracted data included: (1) Basic study information (first author, year of publication). (2) Participant demographics (gender, age, sample size, subject identity). (3) Intervention details (training methods, frequency, load, duration, number of sessions). (4) Outcome measures (1RM squat, squat jump, countermovement jump, sprint). During the process of data extraction, data from all intervention groups is obtained. In instances where direct acquisition of data is not feasible, methodologies such as GetData Graph Digitizer and Cochrane data conversion tools are employed to ensure uniform integration of the data. To maintain rigor, literature lacking crucial data that cannot be converted or that does not meet the established criteria is excluded directly. For the same intervention measures within the same literature, a combined treatment is applied.

The methodological quality of the included studies was assessed using the Physiotherapy Evidence Database (PEDro) scale, which evaluates external validity (Item 1) and internal validity (Items 2–9), as well as the adequacy of statistical information (Items 10–11). Each criterion met scores 1 point, with a maximum possible score of 10 (excluding Item 1). Studies scoring 6–10 were considered high quality, 4-5 moderate quality, and ≤3 low quality. Two authors independently evaluated the methodological quality of each study, with discrepancies resolved through discussion and consultation.

Publication bias was assessed using Egger’s regression test and funnel plot asymmetry adjustments (Egger et al., 1997).

The Bayesian network meta-analysis model was chosen for its ability to rank interventions based on posterior probabilities. This approach overcomes the limitations of frequentist methods, which estimate parameters by iteratively maximizing the likelihood function. Such methods can lead to instability and biased results. In contrast, the Bayesian model is more flexible and is currently the preferred method in evidence-based research.

The analysis was conducted using R (version 4.3.3 with the gemtc package) and Stata (MP 18). Given the different measurement methods for the outcome indicators, a random effects model was chosen to minimize heterogeneity and determine the overall effect of the training interventions on the outcomes, including 1RM, SJ, CMJ, and SP.

To objectively and accurately reflect the relative effects of the three intervention methods and to exclude the influence of the conventional training control group on effect sizes, direct comparisons were made between the three interventions. The pre- and post-training changes in the experimental and control groups were used to calculate the standardized mean difference (SMD) and its 95% confidence interval (CI) for pooling effect sizes. Heterogeneity was assessed I² statistic (low <25%, moderate 25%–50%, high >50%). Statistical significance was set at p < 0.05.

Ultimately, 17 studies were included in the analysis, comprising 15 two-arm trials and 2 three-arm trials. Among the 17 studies, there were a total of 394 participants distributed across 12 plyometric training groups (n = 134), 13 traditional resistance training groups (n = 142), and 11 Olympic weightlifting training groups (n = 118). The age range of participants was 18–30 years. The fundamental characteristics of the literature are presented in Table 1.

The methodological quality of the included studies was evaluated using the PEDro scale for physical therapy. Overall, the average PEDro score of 4.29 indicated that the included studies were of moderate quality as shown in Table 1.

Of the various methods used to assess publication bias, funnel plots are most frequently employed due to their straightforward nature; however, analyzing publication bias with funnel plots necessitates subjective interpretation to detect bias and cannot objectively describe symmetry within the graphical representation. Egger’s tests and adjustments for asymmetry in funnel plots were utilized to assess publication bias. Funnel plot results suggested a generally symmetrical distribution, while Egger’s regression test results indicated no significant publication bias with p > 0.05, thus indicating the absence of publication bias. The findings from both the funnel plot and Egger’s regression tests confirmed that publication bias was largely absent from the studies (Egger et al., 1997). See Figure 2.

The outcome measures selected for the meta-analysis related to lower extremity strength performance were maximum strength, jumping ability, and sprinting ability. Lower limb maximum strength is a key indicator of athletes’ performance in both jumping and sprinting movements. The maximum strength measures included the highest weight achieved in one repetition of squat, half-squat, or leg press exercises. The short-distance sprints were defined by the time taken to complete sprints ranging from 5 to 30 m. Three types of standing vertical jump were assessed: squat jump, countermovement jump.

The outcomes of the heterogeneity analysis indicated that the I ² values for direct pairwise comparisons of the interventions across all outcome indicators were generally below 75%, thus reflecting low heterogeneity. Moreover, the network comparisons of the interventions also had an I² value generally below 75% and was less than that of the direct comparisons. Please refer to Supplementary Material S4. for details on the heterogeneity analysis.

PT has been widely recognized for its ability to enhance lower extremity strength, power, and sprint performance through its unique physiological mechanisms. The core principle of PT lies in the stretch-shortening cycle (SSC), which involves a rapid eccentric (lengthening) phase followed by an immediate concentric (shortening) contraction (Watkins, 2009; Markovic and Mikulic, 2010; Turner and Jeffreys, 2010; Tomalka et al., 2020). This mechanism allows muscles and tendons to store elastic energy during the eccentric phase, which is then released during the concentric phase, resulting in a more powerful and efficient movement. The SSC is particularly effective in improving sprint performance due to its ability to enhance the rate of force development (RFD) and reduce ground contact time, both of which are critical for explosive movements such as sprinting and jumping.

Physiological adaptations induced by PT include increased tendon stiffness and muscle fiber hypertrophy, particularly in type IIa fibers, which are responsible for fast-twitch muscle contractions (Komi, 2000; Mero and Komi, 1986). Studies have shown that PT can lead to a 24.1% increase in tendon stiffness, which enhances the efficiency of force transmission from muscles to bones, thereby improving power output during explosive movements (Fouré et al., 2010). Additionally, PT has been shown to increase the cross-sectional area (CSA) of key lower extremity muscles, such as the quadriceps and hamstrings, by up to 10%, further contributing to improved strength and power (Vissing et al., 2008). These adaptations are particularly beneficial for sprint performance, as they allow athletes to generate greater force in a shorter amount of time, which is essential for rapid acceleration and high-speed running.

The short ground contact time (<250 ms) characteristic of PT exercises is another key factor that makes it particularly effective for improving sprint performance (Malisoux et al., 2006; Watkins, 2009; Flanagan and Comyns, 2008). During sprinting, the ability to quickly transition from the eccentric to the concentric phase of muscle contraction is crucial for maintaining high running speeds. PT training enhances this ability by improving neuromuscular coordination and increasing the stiffness of the lower extremity muscles and tendons, which reduces the time required to generate force during each stride (Bouguezzi et al., 2020). This is why PT is often more effective than other training modalities for improving sprint-related performance.

In terms of practical applications, PT should be tailored to the athlete’s experience level and specific performance goals. For beginners, it is recommended to start with low-intensity exercises involving 40–80 ground contacts per session, gradually progressing to higher intensities as the athlete becomes more experienced (Jensen and Ebben, 2005). Elite athletes, on the other hand, may benefit from high-intensity PT sessions involving 200–400 ground contacts, as these can further enhance power output and sprint performance. Controversially, many studies have found that high-intensity plyometric training is highly susceptible to injury and fatigue (Tofas et al., 2008; Drinkwater et al., 2009; Chatzinikolaou et al., 2010). For instance, one study found that high-volume plyometric training (212 ground contacts) already severely impairs a muscle’s ability to generate force and slow evoked contraction velocity, even if the athlete doesn’t feel fatigued (Drinkwater et al., 2009). It is crucial to balance training intensity with adequate recovery, as excessive PT can lead to neuromuscular fatigue and muscle damage, particularly in type II fibers, which are highly susceptible to overtraining (Macaluso et al., 2012). In addition, an anonymous survey study showed that many practitioners routinely use significantly lower amounts of sessions than previously recommended in the literature (p < 0.05) (Watkins et al., 2021). Thus, to align with the practical need for safe, effective PT programming (and address the nuance of volume prescription across athlete populations), training design must prioritize individualized volume adjustment (rather than a one-size-fits-all range) and integrated recovery strategies.

In summary, the physiological mechanisms underlying PT, particularly the SSC, make it highly effective for improving sprint performance by enhancing tendon stiffness, muscle fiber hypertrophy, and neuromuscular coordination. To optimize its benefits, PT should be implemented with careful consideration of the athlete’s experience level and recovery needs, ensuring that training intensity is appropriately matched to the desired performance outcomes.

The weightlifting training methods, including the clean and jerk, snatch, and deadlift, facilitate the achievement of coordinated triple extension in the primary lower extremity joints (hip, knee, and ankle) during exercise. Weightlifting dates back to ancient China, Egypt, and Greece, and it wasn’t until the 1950s that it found its application in foundational training for non-weightlifting athletes involved in team sports and athletics (Comfort et al., 2023; Todd, 2008; Shurley and Todd, 2012). Since that time, weightlifting-style training methods have been extensively applied in strength training. Fry’s study revealed that weightlifters exhibit a greater proportion of type IIA fibers and a comparatively higher relative content of myosin heavy chain IIA sub than normal adults. Furthermore, weightlifting performance shows strong correlation with the percentage of type IIA fibers (r = 0.94) and the area percentage of type IIA fibers (r = 0.83) (Fry et al., 2003). Numerous cross-sectional studies have provided evidence that high-intensity training in weightlifters results in hypertrophy of type IIA fibers (Fry et al., 2003; Tesch et al., 1984). Longitudinal HIRE studies on non-weightlifters also indicate the potential conversion from IIX to IIA fiber types (Fry, 2004). Olympic weightlifting essentially relies on movements like the clean, jerk, snatch, and deadlift, stressing the quick movement and release of the, thus representing a dynamic and rapid form of training that allows athletes to generate substantial force and power in a brief timeframe; this dynamic and rapid training method effectively improves an athlete’s explosive strength. Similarly, jumping involves quickly moving the body vertically and ‘releasing’ it into the air. This attribute of weightlifting training overlaps with the biomechanics of sprinting and jumping. Extensive longitudinal studies and systematic evaluations have reported the efficacy of weightlifting training in enhancing athletic strength performance. In weightlifting training, the muscle rate of force development and power output are maximized, with numerous researchers discovering that weightlifting significantly enhances power during jumps (Arabatzi and Kellis, 2012; Hawkins et al., 2009; Ostrowski et al., 1997). Garhammer’s study additionally identified similarities between the force-velocity loading profiles of weightlifting training and the movement patterns involved in jumping (Garhammer, 1993). Collectively, these findings suggest that weightlifting training can more effectively improve rapid muscle strength and power output to conventional training methods. Indeed, while weight training greatly enhances lower extremity performance, it also places significant demands on the flexibility of the hip, ankle, and upper limb joints due to its specific movement patterns. People with inadequate joint flexibility and insufficient technical proficiency may fail to experience training benefits, thereby raising their injury risk during exercise. Elite level weightlifters engage in high-intensity resistance exercises (HIRE, ≥80% of 1RM) twice or more each day, accumulating 6-7 sessions a week (Garhammer, 1993; Thrush, 1995). In elite weightlifters, partitioning the predefined training volume into two sessions on the same day can markedly enhance muscle strength, hypertrophy, and maximal neural activation of the trained muscle groups (Häkkinen and Kallinen, 1994; Hartman et al., 2007). Nevertheless, as most exercises executed by weightlifters revolve around competitive lifting and similar multi-joint activities, the same primary muscle groups experience extensive contractions during each training session. Existing evidence further suggests that repeated HIRE targeting the same muscle groups can result in sustained suppression of critical synthetic metabolic mediators, prolonged inflammatory response, and decreased muscle performance (Fry et al., 1994b; Fry et al., 1994a; Fry et al., 2006; Crewther and Christian, 2010; Crewther et al., 2011). Compared to conventional resistance training, weightlifting may yield lower strength development; nevertheless, it demonstrates higher outputs in speed-strength performances in young athletes (Channell and Barfield, 2008). Moreover, the rate of force development (RFD) during weightlifting is considerably greater than that observed during squats (Chimera et al., 2004; Gruber et al., 2007; Young and Bilby, 1993). While there is no direct evidence suggesting adverse effects of weightlifting training on children or adolescents, there remains significant debate regarding its use among younger populations (Storey and Smith, 2012). Consequently, there is still a lack of well-defined physiological guidelines and age-appropriate weightlifting training recommendations.

Indeed, numerous studies suggest that traditional heavy resistance training primarily centers on maximizing muscle strength and muscle hypertrophy, which in turn enhances lower extremity performance, thus accounting for the substantial increases in maximal strength. On the other hand, weightlifting and plyometric training methods tend to reduce muscle hypertrophy due to their distinct movement patterns while focusing on enhancing lower extremity explosiveness (Cormie et al., 2011a; Cormie et al., 2011b; Folland and Williams, 2007; Komi, 1984). In exploring the impact of plyometric training, weightlifting training, and TT on performance improvement, the specificity of training methods must be addressed. Concerning training specificity, considerable evidence demonstrates that employing training methods analogous to the targeted movement patterns can result in greater transfer of training effects and subsequently yield better outcomes. Stone indicated that the SSC associated with plyometric training and the triple extension in weightlifting are evidently better aligned with the biomechanics of vertical jump performance (Stone et al., 2022). Additionally, many studies have confirmed the effectiveness of both plyometric training and weightlifting training in increasing power output (Cormie et al., 2011a; Cormie et al., 2011b; Suchomel et al., 2015). In addition, studies have revealed that weightlifting exercises and ultra-length exercises share similarities with the dynamics of jumping (including maximum strength and power, the timing to reach maximum strength and power, and relative strength and power) (Canavan et al., 1996; Haff et al., 2005; Haff et al., 2015). Although TT executes actions that resemble vertical jumps with similar hip, knee, and ankle extension movements, the application of higher loads (above 80% of 1RM) inevitably result in some loss of movement velocity (Cormie et al., 2007; Hedrick and Wada, 2008). Even under lower load conditions (45% 1RM), which are closer to practical movement speeds, TT displays a prominent deceleration phase. This deceleration phase is related to slower concentric velocities, diminished concentric forces, and decreased levels of muscle recruitment; these factors together result in a reduction in maximum power output. Another perspective argues that specificity in strength training is not a requirement; rather, an increase in muscle contractile strength will naturally enhance an athlete’s jumping capability. Supporters of this theory believe that the effectiveness of plyometric training and weightlifting in improving vertical jump performance is due to their ability to enhance muscle contractile capacity (Toumi et al., 2004; Adams et al., 1992). To summarize, while weightlifting and plyometric training may show somewhat less effectiveness in enhancing maximum muscle strength relative to TT, they appear to be more beneficial in improving quick strength. It is advised to integrate weightlifting methods and plyometric training with TT in training regimens.