The review protocol was prospectively registered in PROSPERO (Registration ID: CRD42024607126). This review is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement guidelines (18).

The search strategy was structured using the Observational Study PECO framework (Participants, Exposure, Comparison, Outcome, Study design) (19). Search strategy can be found in Table 1. One of the authors (anonymized for review) conducted a primary search in PubMed. Then, the search strings were translated for Web of Science, and EBSCO. Grey literature was excluded. The search covered the period from January 1, 2004, to 31st December, 2025. No language restrictions were applied. Detailed search string design and search results are available in Supplementary File S1.

All articles identified through the search were exported into NoteExpress reference management software [NoteExpress Reference Management Software (Version 3.9). Beijing Aegean Software Company, Beijing, China.], where automated duplicate detection and removal were performed. Following de-duplication, two independent reviewers systematically screened the titles, abstracts, and full texts of all remaining records to determine their eligibility for inclusion in the review. Any discrepancies between reviewers were resolved through discussion until consensus was reached, and a third reviewer was consulted when necessary.

The screening and selection process was conducted according to predefined eligibility criteria. Articles were included if they met all of the following conditions: (a) written in English; (b) observational cohort or cross-sectional study design; (c) reported at least one biomechanical outcome variable; (d) included both novice and non-novice (experienced) runners for comparison; and (e) the primary physical activity under investigation was running. Studies were excluded if they met any of the following criteria: (a) not written in English; (b) randomized controlled trials or intervention-based studies; (c) did not include analysis of biomechanical variables; (d) did not compare novice and non-novice runners; or (e) the participants were engaged in physical activities other than running.

Two investigators independently extracted data from the included studies. For each included study, first author, publication year, participant characteristics, experimental procedures and conditions, biomechanical variables, measurement instruments and parameters, and main findings were extracted.

For the definition of running experience, when multiple indicators were employed to categorize runners’ experience levels, the variable that was consistently applied across all groups was extracted. For instance, if both “running years” and “weekly running volume” were used to define experienced runners, but only “running years” was applied to define novice runners, then “running years” was regarded as the operational measure of running experience within that study.

Given the substantial heterogeneity in study designs, outcome measures, and definitions of running experience, a formal meta-analysis or statistical subgroup analysis was not feasible. Therefore, heterogeneity was explored using a structured narrative synthesis approach, in line with recommendations from the BMJ Synthesis without meta-analysis (SWiM) in systematic reviews: reporting guideline (20) and Cochrane Handbook for Systematic Reviews of Interventions (21). Findings were organised and compared according to hypothesised methodological modifiers, including the definition of running experience, running environment (treadmill vs. overground), and key experimental conditions.

Risk of bias (RoB) was independently assessed by two investigators using the National Institute of Health (NIH) observational cohort and cross-sectional Study Quality Assessment tool (22). This tool is widely accepted for assessing the potential bias in observational cohorts and cross-sectional studies.

The scale included 14 items. Item#10 [Was the exposure (s) assessed more than once over time?] and Item#13 (Was loss to follow-up after baseline 20% or less?) are removed because of their lack of applicability to specific types of studies. The original score for high-quality research is “≥11” (42). After removing 2 unrelated items, the score for high-quality research in this review should be “≥9.” Two investigators assessed studies independently, resolving disagreements by consensus or third-party adjudication.

The characteristics of the included studies are summarized in Table 2. In total, the reviewed studies involved 457 runners. Among these, eight studies examined kinematic variables, five assessed kinetic variables, two investigated spatiotemporal parameters, five focused on coordination, symmetry, and stability, and two analyzed muscle activity.

The definition of “novice runner” varied across studies. Seven studies distinguished novice and experienced runners based on years of running participation, five used weekly running volume, one applied a combination of both criteria, and one classified participant according to competition experience.

Regarding experimental settings, 12 studies (86%) conducted tests on a treadmill, whereas two studies (14%) employed overground running. Control for potential confounding variables also differed across studies (Table 3). Specifically, only two studies (13%) controlled for participants’ foot strike pattern, while five studies (33%) standardized footwear conditions, of which one required participant to run barefoot, and four instructed them to wear identical running shoes.

Across the included studies, participant age varied substantially, both between and within studies (Table 2). Mean ages ranged from early adulthood (approximately 21–24 years) to middle-aged runners (>40 years), with several studies including age-heterogeneous samples or reporting age differences between novice and experienced groups. Notably, studies defining running experience by years of running often included older experienced runners compared with novice runners [e.g., (16, 25)], whereas studies using weekly running volume tended to recruit younger cohorts with relatively small age differences between experience groups [e.g., (26, 30)].

Considerable heterogeneity was also observed in sex composition across studies (Table 2). Of the 14 included studies, five recruited male-only samples, three included female-only samples, and the remaining studies involved mixed-sex cohorts. Among studies with single-sex samples, biomechanical differences between novice and experienced runners were broadly consistent with findings from mixed-sex studies, particularly for joint ranges of motion, coordination, and stability-related outcomes.

The results of the risk of bias assessment are summarized in Table 4. Most studies (9, 64%) did not justify their sample size, and none (14, 100%) incorporated blinding procedures in their design or data collection. In total, six studies (43%) were rated as high quality.

The included studies defined runners’ experience based on running years, weekly running volume, or competition experience, and reported differences between novice and experienced runners in spatiotemporal, kinematic, kinetic, neuromuscular, and coordination-related variables (Table 2).

One study investigated the associations between running experience and dynamic stability. This study found that novice runners with lower weekly running volume displayed reduced stability at the ankle, knee, and hip joints compared with experienced runners (24).

Bilateral symmetry was evaluated in one study. This study reported that novice runners with fewer running years demonstrated poorer symmetry in GRF, VALR, and VILR (29).

To explore potential sources of heterogeneity in biomechanical findings, a structured narrative synthesis was conducted by stratifying studies according to hypothesised methodological modifiers, including the operational definition of running experience, running environment (treadmill vs. overground).

This systematic review synthesized evidence from 14 studies that investigated biomechanical differences between novice and experienced runners, with particular attention to how these differences vary according to diverse operational definitions of “running experience”. Across studies, novice runners consistently exhibited greater variability and poorer control in spatiotemporal parameters, larger joint ranges of motion (especially at the ankle and hip), and higher joint moments across multiple planes at the ankle. Novice runners also demonstrated reduced hip and trunk muscle strength, weaker inter-joint coordination, lower dynamic stability, and greater bilateral asymmetry during running.

The biomechanical characteristics of novice runners under different definitions are illustrated in Figure 2. In existing literature, running experience is operationalized as (i) years of running, (ii) weekly running volume (mileage), and (iii) competition or participation history. Across these definitions, novice runners exhibit broadly consistent biomechanical features. Specifically, runners with lower weekly mileage typically show increased joint ranges of motion in flexion–extension and rotation at the trunk and hip, reduced extension and external rotation moments, and diminished stability and symmetry. Those with fewer running years are characterized by shorter stride length, higher stride-time variability, weaker hip and trunk strength endurance, and inconsistent knee and ankle kinematics, whereas individuals with limited competition experience often display smaller joint excursions and lower propulsive ground reaction forces, indicating reduced propulsion efficiency and neuromuscular control.

Despite these consistent trends, discrepancies across studies are mainly concentrated in kinematic variables, particularly knee flexion/extension and ankle motion patterns, including variations in trunk rotation, hip internal–external rotation moments, knee flexion and extension angles, ankle dorsiflexion–plantarflexion and inversion–eversion ranges, stride length, stride rate, and coordination coupling. One possible explanation for this inconsistency is that “running years” primarily capture the long-term exposure and accumulated motor learning of running techniques, whereas “weekly running volume” reflects more recent training load, fatigue level, and current physical conditioning. These two metrics therefore represent different dimensions of running experience, one emphasizing chronic adaptation, the other acute training status.

The review found that novice runners experience exhibited greater spatiotemporal variability and lower baseline stride rates, particularly when required to adjust their step frequency, compared with experienced runners. This difference may stem from insufficient neuromuscular coordination, stability, strength, and endurance among novice runners (34). Novice runners may rely more heavily on conscious motor control and less on automatized gait patterns, leading to inconsistent temporal regulation and stride variability. Limited exposure to repetitive loading cycles may also restrict the optimization of central pattern generators responsible for rhythmical locomotion, resulting in less stable and less economical gait control (34). Moreover, inadequate muscle strength and endurance, particularly in the hip stabilizers, knee extensors, and ankle plantar flexors, can reduce the ability to maintain consistent limb trajectories and control impact forces. For example, studies have shown that novice runners often present poorer joint stability across the hip, knee, and ankle (24), which may impair shock absorption and alter timing between stance and swing phases. Deficient proprioceptive acuity and delayed muscle activation further compromise joint stiffness regulation, causing increased contact times and step-to-step variability (24).

Novice runners generally exhibit a greater ROM of ankle, particularly larger inversion–eversion excursions and increased inversion and internal-rotation moments than experienced runners (26). These differences likely stem from two interrelated factors. Physiologically, limited training exposure may lead to weaker peri-ankle musculature and reduced neuromuscular control. This can result in diminished joint stiffness and delayed muscle activation timing, reducing the ability to dynamically stabilize the ankle during stance and push-off (35). Consequently, the ankle may display exaggerated motion as muscles respond reactively rather than pre-emptively to ground contact. Methodologically, the 5 km pre-test run in that study may have induced fatigue among novices whose average weekly mileage was only 2–10 km, thereby amplifying joint excursions during measurement. Excessive ankle eversion during stance has been linked to elevated risks of Achilles tendinopathy, anterior cruciate ligament strain, and increased tibial stress (36, 37). Combined with the longer ground-contact times often observed in less experienced runners, these features may heighten tibial and soft-tissue loading. Moreover, the greater eversion and internal-rotation moments likely reflect compensatory stabilization efforts, demanding increased activation of the tibialis anterior and posterior tibialis muscles (38), which may contribute to the reduced running economy characteristic of novice runners.

The hip joint plays a critical role in lower-limb kinematics and stability, and its dysfunction has been identified as a key contributor to impaired running performance and injury risk (39). Across included studies, novice runners consistently exhibit greater hip RoM during the stance phase, suggesting reduced joint stability compared with experienced runners. This increased RoM may result from strength imbalances between the hip abductors and adductors, which compromise pelvic control and lead to excessive hip motion (30). Additionally, insufficient endurance of the hip and trunk musculature among runners with lower training volumes may exacerbate fatigue-induced instability. With prolonged running, novices appear less able to maintain neuromuscular control, resulting in greater lateral sway and difficulty modulating hip motion to absorb impact efficiently (27, 28).

When interpreted through the lens of age and sex, the biomechanical differences observed between novice and experienced runners appear to reflect partially distinct mechanisms. Age-related variability across studies was more closely associated with neuromuscular control, coordination, and stability-related outcomes. Studies involving older experienced runners tended to report more consistent differences in stride-to-stride variability, inter-joint coordination, and dynamic stability, suggesting that accumulated running exposure and age-related motor learning may jointly contribute to more stable and automatized gait patterns. In contrast, studies with younger cohorts and narrower age ranges more frequently reported differences in joint kinematics and kinetics, particularly at the ankle and hip, indicating that training status and recent mechanical loading may play a larger role when age-related adaptations are less pronounced.

In comparison, sex-related differences were more likely to influence the magnitude and direction of joint-specific kinematic and kinetic outcomes. Female-only studies more commonly reported greater frontal and transverse plane motions at the knee and hip, whereas male-only studies more frequently identified differences in joint moments, propulsion, and stability-related measures. These patterns align with known sex-based differences in pelvic morphology, lower-limb alignment, and muscle activation strategies during running.

This review highlights several methodological limitations in the existing literature that warrant attention in future research. First, substantial inconsistency exists in the operational definition of “running experience.” This definitional heterogeneity undermines cross-study comparability and constrains the extent to which biomechanical findings can be meaningfully synthesized. Future research would benefit from adopting standardized, multidimensional definitions of running experience that integrate indicators such as years of running, training frequency, weekly mileage, and performance level. Second, biomechanical assessments were conducted under heterogeneous experimental environments, with some studies employing treadmill running and others using overground protocols. Given well-documented biomechanical differences between these conditions (33), future studies should either standardize testing environments or explicitly compare treadmill and overground running to enhance ecological validity and interpretability. Third, potential confounding variables were insufficiently controlled across studies. Only a small proportion accounted for foot-strike pattern or standardized footwear, despite strong evidence that both substantially influence running biomechanics (13, 40). Improved methodological rigor in controlling or reporting these factors is essential to reduce uncertainty in biomechanical comparisons across experience levels. In addition, this review identified a notable gap in the investigation of trunk kinematics and their interactions with lower-limb biomechanics. Although trunk motion has been shown to influence lower-extremity loading and coordination (13), few studies have examined these relationships specifically in novice runners. Future research incorporating trunk-related biomechanical variables alongside ground reaction force measures may contribute to a more comprehensive understanding of whole-body adaptations associated with running experience. Finally, despite a growing body of research examining biomechanical differences between novice and experienced runners, considerable heterogeneity persists in participant characteristics and analytical approaches. This heterogeneity complicates the interpretation of findings and cautions against strong translation into practice. Future studies should incorporate age- and sex-stratified analyses or explicitly model age–sex–experience interactions to better distinguish experience-related biomechanical adaptations from demographic influences and to improve the generalizability of conclusions.