Effects of neuromuscular training on pain, physical function, balance, muscle strength, and lower limb kinematics in individuals with patellofemoral pain: a systematic review and meta-analysis of randomized controlled trials Authors: Ali Asghar Maleki, Evangelos Pappas, Sara R. Piva, Seyed Hamed Mousavi Published: 2026 ABSTRACT BackgroundPatellofemoral pain (PFP) is a common musculoskeletal condition that impairs daily function and quality of life. Neuromuscular training (NMT) has been proposed to address pain, functional limitations, and biomechanical deficits in PFP, but its overall effectiveness across multiple outcomes remains unclear.MethodsWe conducted a systematic review and meta-analysis of randomized controlled trials (RCTs) evaluating the effects of NMT in individuals with PFP. PubMed, Web of Science, Scopus, and Embase were searched from inception to May 2025. Eligible studies compared NMT with control conditions or active interventions such as strength training (ST) and reported at least one of the following outcomes: pain, physical function, balance, muscle strength, or lower-limb kinematics. Data were pooled using random-effects models to calculate mean differences (MD) with 95% confidence intervals (CI). The certainty of evidence was assessed using the GRADE approach.ResultsSeventeen RCTs (n = 871; 66% female; mean age 23 years) met the inclusion criteria. Compared with control groups, NMT reduced pain (MD = − 0.60; 95% CI: − 0.96 to − 0.23; moderate certainty) and improved physical function (MD = 10.83; 95% CI: 9.71 to 11.94; moderate certainty). Compared with strength training, NMT produced greater pain reduction (MD = − 0.55; 95% CI: − 0.77 to − 0.34; moderate certainty) but did not improve physical function beyond ST alone (MD = − 2.32; 95% CI: − 10.31 to 5.66; low certainty). Although NMT appears effective for reducing pain and improving physical function, its effects on lower-limb biomechanics remain inconclusive, with current evidence showing inconsistent findings across studies.ConclusionsNMT is an effective intervention for reducing pain and improving physical function in individuals with PFP. Although not superior to ST for functional outcomes, NMT may provide added value when incorporated into multimodal rehabilitation programs targeting neuromuscular control, proprioception, and dynamic stability. Further high-quality trials with standardized protocols are needed to clarify its effects on balance, strength, and kinematics.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13102-026-01730-9. Introduction Patellofemoral pain (PFP) is one of the most common musculoskeletal disorders of the knee and is characterized by diffuse pain around or behind the patella, typically aggravated by activities such as squatting, stair negotiation, running, and prolonged sitting [1, 2]. It affects a wide range of populations and represents a substantial clinical burden, with prevalence estimates of approximately 20–25% among adolescents and physically active adults. PFP also accounts for a large proportion of knee‑related consultations in sports medicine and orthopedic settings [3, 4]. The condition is associated with persistent symptoms, reduced physical function, and diminished quality of life, and symptoms may persist for years if inadequately managed [1, 5]. Epidemiological studies consistently report a higher prevalence in women, with a female‑to‑male ratio of roughly 2:1 [6, 7]. PFP is also common among adolescents, athletes, and individuals exposed to repetitive knee loading [7, 8]. Despite its high prevalence, the etiology of PFP is complex and multifactorial [9, 10]. Current models suggest that symptoms arise from the interaction of biomechanical, structural, psychological, and overuse‑related factors rather than a single underlying cause [11]. Biomechanical impairments are considered particularly important contributors. Individuals with PFP frequently demonstrate weakness or impaired activation of key muscle groups, including the gluteus medius and gluteus minimus (hip abductors), gluteus maximus (hip external rotator), quadratus lumborum (lateral trunk stabilizer), and the quadriceps femoris—particularly the vastus medialis oblique, a primary stabilizer of the patella [12, 13]. These deficits may compromise dynamic lower‑limb stability and contribute to altered movement patterns such as excessive femoral internal rotation, hip adduction, and increased knee valgus during functional tasks. Such alterations increase patellofemoral joint stress and are believed to play a central role in symptom development and persistence [13, 14]. Conservative management is the cornerstone of PFP treatment, with exercise therapy widely recommended as a first‑line intervention. In recent years, neuromuscular training (NMT) has gained attention as a comprehensive rehabilitation strategy [15–18]. NMT typically includes balance and proprioceptive exercises, dynamic stability drills, movement retraining, agility tasks, and plyometric exercises, all designed to improve neuromuscular control, strength, coordination, and movement quality [19, 20]. Beyond rehabilitation, NMT has been shown to reduce injury risk and enhance performance in athletic populations, underscoring its potential relevance for individuals with PFP [19, 21]. Preliminary randomized controlled trials suggest that NMT may effectively address several impairments associated with PFP, including pain, physical function, muscle strength, balance, and lower‑limb kinematics [22–26]. However, the existing evidence remains fragmented, with considerable variability in intervention protocols, outcome measures, and methodological quality across studies [27, 28]. Moreover, previous reviews have often focused on isolated outcomes—such as pain or function—without comprehensively evaluating the broader neuromuscular and biomechanical impairments relevant to PFP. These gaps highlight the need for an updated and comprehensive synthesis of the literature examining the effects of NMT on pain, physical function, muscle strength, balance, and lower‑limb kinematics in individuals with PFP. Therefore, this systematic review and meta‑analysis was conducted to evaluate the effectiveness of NMT on these outcomes and to provide an integrated understanding of its therapeutic potential in PFP. Methods This systematic review and meta-analysis were conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [29]. The study protocol was prospectively registered in the International Prospective Register of Systematic Reviews (PROSPERO; registration number: CRD42025645819; registration date: 17 March 2025). Eligibility criteria Using the PICOS (Population, Intervention, Comparator, Outcomes, and Study Design) framework [30], we established the following inclusion and exclusion criteria: Population: Studies enrolling individuals diagnosed with PFP.Interventions: NMT delivered as the primary intervention.Comparator: Studies included comparison of NMT against control groups or against other active exercise modalities, such as strength training (ST), or evaluations of outcomes pre versus post NMT.Outcomes: Studies were required to report at least one relevant outcome, including pain, physical function, muscle strength, balance, or lower limb kinematics.Study Design: Only RCTs published in peer-reviewed journals and written in English were included. Population: Studies enrolling individuals diagnosed with PFP. Population: Studies enrolling individuals diagnosed with PFP. Interventions: NMT delivered as the primary intervention. Interventions: NMT delivered as the primary intervention. Comparator: Studies included comparison of NMT against control groups or against other active exercise modalities, such as strength training (ST), or evaluations of outcomes pre versus post NMT. Comparator: Studies included comparison of NMT against control groups or against other active exercise modalities, such as strength training (ST), or evaluations of outcomes pre versus post NMT. Outcomes: Studies were required to report at least one relevant outcome, including pain, physical function, muscle strength, balance, or lower limb kinematics. Outcomes: Studies were required to report at least one relevant outcome, including pain, physical function, muscle strength, balance, or lower limb kinematics. Study Design: Only RCTs published in peer-reviewed journals and written in English were included. Study Design: Only RCTs published in peer-reviewed journals and written in English were included. Search strategy A comprehensive search was conducted in PubMed, Web of Science, Scopus, and Embase from database inception to May 10, 2025. The search strategy combined keywords and controlled vocabulary (MeSH and Emtree terms) related to “patellofemoral pain,” “anterior knee pain,” and “neuromuscular training.” Boolean operators and database‑specific indexing terms were adapted for each database. ClinicalTrials.gov was also searched for ongoing or unpublished trials, and reference lists of eligible studies and relevant reviews were screened manually. Full search strategies are provided in Supplementary Table S1. Study selection All retrieved articles were imported into EndNote X7 (Thomson Reuters, New York, NY, USA) for duplicate removal and screening. Titles and abstracts were independently screened by two reviewers (AAM and SHM) based on predefined criteria. Full texts of potentially eligible studies were then assessed. Disagreements were resolved through discussion or consultation with a third reviewer (EP). Inter‑rater agreement for both screening stages was calculated using Cohen’s kappa statistic. Data collection process Two authors (AAM and SHM) independently extracted all pertinent details from the selected studies, and these data were subsequently verified for accuracy by a third author (EP). The extraction process systematically recorded information on the publication year, study design, sample size, participant demographics (age and sex), specific intervention characteristics (including type and duration), and the key findings. Quality assessment Methodological quality was assessed independently by two reviewers using the Physiotherapy Evidence Database (PEDro) scale [31]. The PEDro scale was selected due to its widespread use and validation in physiotherapy and exercise-based randomized controlled trials, allowing for standardized appraisal of internal validity and reporting quality. Studies were categorized as high quality (≥ 6 points), moderate quality (4–5 points), or low quality (≤ 3 points) [31, 32]. Disagreements in scoring were resolved by discussion. If a study did not provide sufficient information regarding any PEDro scale elements, a zero score was assigned for that criterion. This approach ensured a consistent and objective evaluation across all included studies. Synthesis of results All statistical analyses were performed in RevMan 5.4 using two-sided random-effects models to account for anticipated variability in intervention protocols, participant characteristics, and timing of outcome assessments. Unstandardized mean differences (MD) with corresponding 95% confidence intervals (CI) were pooled across eligible studies. Statistical heterogeneity was quantified using the I² statistic and its chi-square p value, with thresholds interpreted as minimal (< 25%), low (25–50%), moderate (50–75%), or high (> 75%). For outcomes reported in ten or more studies, funnel-plot symmetry was visually inspected; when asymmetry suggested potential small-study effects, the trim-and-fill method was applied to generate bias-adjusted MD estimates. Not all studies identified in the qualitative synthesis were included in the quantitative pooling. Trials were excluded from meta-analysis if they did not report the necessary statistical data (e.g., mean change scores with standard deviations), used outcome measures that were not comparable with the majority of included trials, or assessed pain at follow-up time points that differed substantially from the predefined analysis windows. Each pooled effect was subsequently appraised using the GRADE framework across five domains: risk of bias (based on PEDro scores), inconsistency (magnitude of I² and CI overlap), indirectness (alignment with the PICO question), imprecision (width of CI and total sample size), and publication bias. The certainty of evidence for each outcome was then classified as high, moderate, low, or very low [33] (S4 Table). Results Study selection The initial literature search identified 2253 studies from PubMed, Web of Science, Scopus, and Embase. After removing duplicates, 1379 studies remained. Following the application of the inclusion criteria, 1324 studies were excluded. After title and abstract screening, 55 papers were selected for full text review, and of those, 17 studies met the inclusion criteria. Figure 1 shows the flow diagram of the selection process and the number of excluded studies at each stage. Fig. 1PRISMA flow diagram of the study selection process PRISMA flow diagram of the study selection process Study characteristics The overview of the characteristics of the studies included in the analysis. Research has examined the effects of NMT on individuals with PFP between 2015 and 2025. A large portion of this research (73%) occurred between 2020 and 2025. The studies included a total of 871 participants, with females accounting for 66% of the sample. On average, the participants were 23 years old and had undergone 6 weeks of training. The outcomes of the studies varied: 14 investigated the impact on pain, 11 on physical function, 7 on muscle strength, 7 on balance, and 5 on lower limb kinematic outcomes (S2 Table). Quality assessment The results of the quality assessment using the PEDro scale. The quality of 5 studies [22, 34–37] was moderate (ranging from 6 to 8), whereas 12 studies [16, 23, 28, 38–46] were of high quality (between 9 and 10). The average quality assessment score for the studies was 8.82 (S3 Table). Outcomes Effects of NMT on pain Fourteen studies [16, 22, 23, 28, 34–41, 45, 46] examined the effects of NMT on pain in individuals with PFP. Nine studies compared NMT with control groups, and most reported substantial pain reduction following NMT [23, 28, 34–36, 38–40, 45]. As illustrated in Fig. 2, there is moderate‑certainty evidence for a reduction in pain, with a pooled mean difference (MD) of − 0.60 (95% CI: −0.96 to − 0.23, P < 0.001, I² = 0%). On a 0–10 VAS scale, this corresponds to a small but clinically relevant improvement. Additionally, four studies compared NMT to ST [16, 22, 38, 46]. These findings provide moderate‑certainty evidence that NMT results in a greater reduction in pain than ST, with a pooled MD of − 0.55 (95% CI: −0.77 to − 0.34, P = 0.005, I² = 43%) (Fig. 3). Considering commonly cited MCID thresholds for PFP pain (approximately 1.0–2.0 points on a 0–10 scale), this effect size may be statistically significant but may fall below the threshold for clear clinical meaningfulness. Fig. 2Results of the meta-analysis of the NMT vs. the control groups for pain reduction Results of the meta-analysis of the NMT vs. the control groups for pain reduction Fig. 3Results of the meta-analysis of the NMT vs. the ST for pain reduction Results of the meta-analysis of the NMT vs. the ST for pain reduction Effects of NMT on physical function A review of eleven studies [16, 23, 28, 35, 36, 40, 41, 43–46] evaluated the impact of NMT on physical function. Seven studies comparing NMT with control groups demonstrated improvements in physical function [23, 28, 35, 36, 40, 44, 45]. A meta‑analysis of four studies showed moderate‑certainty evidence that NMT improved physical function compared with controls, with a pooled MD of 10.83 (95% CI: 9.71 to 11.94, P < 0.001, I² = 9%) (Fig. 4). Four studies comparing NMT with ST found no significant advantage of NMT (MD: −2.32, 95% CI: −10.31 to 5.66, P > 0.05, I² = 94%) (Fig. 5) [16, 41, 44, 46]. The wide confidence intervals and high heterogeneity suggest that the non‑significant findings may reflect variability in protocols and potentially insufficient statistical power. Fig. 4Results of the meta-analysis of NMT vs. control groups for physical function Results of the meta-analysis of NMT vs. control groups for physical function Fig. 5Results of the meta-analysis of NMT vs. ST for physical function Results of the meta-analysis of NMT vs. ST for physical function Effects of NMT on balance The findings of three studies showed that, in comparison to the control groups, NMT resulted in enhancements in balance for individuals with PFP [22, 28, 35]. The results of one study, which compared balance training and NMT, revealed that both exercises improved balance and that there was no difference between the exercises [37]. One study investigating the comparison between ST and NMT revealed that both methods effectively improved balance. However, NMT showed a more pronounced impact on balance enhancement [42]. Meta‑analysis was not performed due to substantial heterogeneity in balance assessment tools and outcome metrics. Effects of NMT on muscle strength The results of five studies revealed that, compared with the control groups, NMT increased the strength of the knee extensor muscles [23, 34, 36, 38, 39]. The results of a study comparing NMT with ST revealed that ST produces greater strength increases [16]. Meta‑analysis was not conducted due to inconsistent strength‑assessment methods and non‑uniform reporting. Effects of NMT on lower limb kinematics Three studies indicated that NMT improved knee valgus and other kinematic parameters in individuals with PFP [22, 23, 46]. However, two studies reported no significant kinematic changes following NMT [16, 41]. Differences in measurement tools—such as 3D motion capture versus 2D video analysis—may contribute to variability in findings. The meta‑analysis (Fig. 6) showed no significant improvements in kinematic parameters following NMT, including knee valgus (MD: 5.38, 95% CI: −1.34 to 12.11, P > 0.05, I² = 97%), knee flexion (MD: −0.46, 95% CI: −2.45 to 1.54, P > 0.05, I² = 0%), hip flexion (MD: −4.35, 95% CI: −9.75 to 1.03, P > 0.05, I² = 82%), hip adduction (MD: −0.27, 95% CI: −2.23 to 1.70, P > 0.05, I² = 0%), and ankle flexion (MD: 0.90, 95% CI: −0.66 to 2.45, P > 0.05, I² = 0%). Fig. 6Results of the meta-analysis of the effects of NMT on lower limb kinematic parameters Results of the meta-analysis of the effects of NMT on lower limb kinematic parameters Discussion This systematic review and meta-analysis, synthesizing evidence from 17 randomized controlled trials involving 871 participants, demonstrates that NMT is effective in reducing pain and improving physical function in individuals with PFP. The pooled findings indicate a moderate reduction in pain following NMT compared with both control conditions and ST. Although improvements in physical function were consistently observed relative to control groups, these gains did not exceed those achieved through conventional ST. This suggests that although NMT offers meaningful benefits compared with non‑exercise controls, it does not appear to provide substantial advantages over strength‑focused interventions. Clinically, NMT may still be particularly beneficial for individuals with notable deficits in balance, proprioception, or neuromuscular control. The mechanisms underlying the benefits of NMT likely relate to its emphasis on enhancing neuromuscular control, proprioceptive feedback, and dynamic stability [47–49]. Neurophysiological evidence—including improvements in joint position sense and normalization of electromyographic activation patterns—supports the idea that NMT promotes more efficient sensorimotor integration [50, 51]. By targeting aberrant movement patterns such as altered patellar tracking or impaired lower‑limb coordination, NMT may facilitate more balanced force distribution during functional tasks. These adaptations not only contribute to pain reduction but may also enhance performance during activities such as stair climbing, squatting, and running [52, 53]. In this way, NMT addresses both symptomatic and biomechanical contributors to PFP. Many NMT programs also incorporate elements of biomechanical re‑education, which aim to correct faulty movement patterns and optimize lower‑limb alignment during functional tasks [54]. This typically involves movement retraining, feedback‑based cues (visual, verbal, or tactile), and task‑specific drills designed to reduce excessive hip adduction, femoral internal rotation, and dynamic knee valgus—mechanical factors strongly implicated in PFP [55]. By promoting more efficient and stable movement strategies, biomechanical re‑education complements the neuromuscular improvements induced by NMT and may contribute to reductions in patellofemoral joint stress [54, 55]. Although NMT and ST differ in their primary emphasis, ST can also improve neuromuscular control, proprioceptive feedback, and dynamic stability through well‑established physiological mechanisms [56]. Progressive resistance training induces muscle hypertrophy and increases maximal force‑generating capacity, but these structural changes are accompanied by substantial neural adaptations, particularly in the early phases of training [57]. These adaptations include improved motor‑unit recruitment, increased firing frequency, enhanced synchronization, and reduced antagonist co‑activation, all of which enhance the efficiency and precision of muscle activation patterns [57]. ST also contributes to improved joint stability by increasing the capacity of key muscle groups—such as the quadriceps, gluteus medius, and gluteus maximus—to control lower‑limb alignment during dynamic tasks [58]. Stronger muscles can more effectively resist excessive femoral internal rotation, hip adduction, and knee valgus, thereby reducing patellofemoral joint stress [58]. Stronger muscles can more effectively resist excessive femoral internal rotation, hip adduction, and knee valgus, thereby reducing patellofemoral joint stress. These biomechanical improvements overlap with the goals of NMT, even though they arise through different training pathways. Additionally, resistance training has been shown to enhance proprioceptive acuity through increased muscle stiffness and improved tendon properties, which enhance the sensitivity of muscle spindles and mechanoreceptors [59]. Because both NMT and ST influence the neuromuscular system—albeit through different mechanisms—their effects on functional outcomes may converge [60]. Functional tasks such as stair climbing, squatting, or running rely on coordinated force production, joint stability, and sensorimotor integration. Improvements in maximal strength and neuromuscular efficiency can therefore translate into functional gains similar to those achieved through NMT, which may explain why the two interventions produce comparable improvements in physical function [17, 60–62]. This highlights the importance of identifying patient subgroups who may benefit more from one intervention than another. Individuals with pronounced neuromuscular deficits, balance impairments, or poor movement coordination may respond more favorably to NMT, whereas those with significant muscle weakness may benefit more from targeted strengthening [60, 63]. Future research should incorporate stratification based on biomechanical alignment, muscle strength profiles, pain chronicity, or symptom severity to better tailor interventions. Investigating whether sequential or combined NMT and ST protocols yield synergistic effects is also warranted. Despite the positive findings, substantial heterogeneity was observed in several outcomes, particularly those related to lower‑limb kinematics. Intervention protocols varied widely, with session frequencies ranging from 2 to 5 sessions per week, durations from 30 to 60 min, and total intervention periods spanning 4 to 12 weeks. Additionally, studies employed diverse biomechanical assessment tools—including 2D video analysis and 3D motion capture—which likely contributed to inconsistent kinematic results. Standardizing intervention parameters and adopting uniform, validated assessment methods would enhance comparability across future trials. Several limitations should be considered when interpreting these findings. Most included studies predominantly involved female participants, which may limit generalizability to male populations. Given known sex‑related differences in lower‑limb biomechanics and neuromuscular control, this imbalance may influence the applicability of results. Methodological variability, including inconsistent blinding procedures and differences in adherence monitoring, may also introduce bias. Furthermore, restricting the search to English‑language publications may have introduced language bias, potentially excluding relevant studies from non‑English‑speaking regions. Taken together, the findings suggest that an integrated rehabilitation approach combining NMT with targeted ST and biomechanical re‑education may offer the most comprehensive strategy for PFP management. A phased framework—beginning with NMT to restore neuromuscular control and progressing toward higher‑load strengthening—may be particularly effective. Future research should aim to determine the optimal dosage, intensity, and duration of NMT protocols and incorporate advanced biomechanical assessment tools, such as wearable sensor technologies, to enhance precision and clinical applicability. Conclusion This systematic review and meta-analysis provides moderate‑certainty evidence that NMT is an effective intervention for individuals with PFP, producing meaningful reductions in pain and improvements in physical function compared with non‑exercise controls. Although NMT does not outperform conventional ST for functional outcomes, it offers distinct therapeutic value through its emphasis on neuromuscular control, proprioception, and dynamic stability—key components in addressing the multifactorial nature of PFP. Limited but promising evidence also suggests that NMT may improve balance and increase knee extensor strength, although the small number of studies and methodological variability limit the strength of these conclusions. In contrast, the effects of NMT on lower‑limb kinematics remain inconclusive, largely due to inconsistent measurement approaches and substantial heterogeneity across studies. Overall, the findings support the integration of NMT into multimodal rehabilitation programs for PFP, particularly for individuals presenting with neuromuscular or proprioceptive deficits. Future high‑quality trials with standardized intervention protocols, consistent outcome measures, and advanced biomechanical assessment methods are needed to clarify the full scope of NMT’s benefits and to identify patient subgroups most likely to respond favorably. Supplementary Information Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2.