Postural control in top-level Para Karate athletes with visual and intellectual impairment: an observational study Authors: Henrieta Horníková, Radovan Hadža, Michal Hrubý, Dušana Augustovičová Published: 2025 ABSTRACT AbstractBackgroundAlthough the World Karate Federation (WKF) lacks a standardized test battery, Para Karate classification follows International Paralympic Committee (IPC) principles, with balance recognized as a key physical ability affecting kata performance. This study aimed to compare postural stability across categories and classes of top-level Para Karate athletes and to provide evidence supporting classification refinement.MethodsFifty-nine athletes competing at the 2023 World Para Karate Championship were classified into three groups: visually impaired (VI, n = 24, age 23.8 [20.6] years), intellectual impairment (II1, n = 17, 26.5 [9.7] years), and intellectual impairment with additional impairment (II2, n = 18, 28.0 [11.3] years). A 30-second bipedal stance test with eyes open was used to assess total path and area of the centre of pressure (CoP), including anterior-posterior (AP) and medio-lateral (ML) directions. Non-parametric Kruskal-Wallis H test and Mann-Whitney U test were used for analysis.ResultsSignificant group differences were observed for total CoP path (H(2) = 13.205, p < 0.001) and area (H(2) = 12.929, p = 0.002). II1 athletes demonstrated significantly lower CoP path than VI (padj=0.001), with better AP stability compared to both groups. II2 class had lower total CoP area than VI (padj<0.001). They showed the lowest area in AP direction compared to both VI and II1 (padj=0.002 and padj=0.040, respectively).ConclusionsThe 30-second balance test proved to be sensitive to between-group differences in postural control among impairment types. VI athletes exhibited greater CoP path than II1 athletes and greater CoP area than II2 athletes, highlighting the need for targeted balance training. In the AP direction, II1 athletes showed better control in path length compared to both groups, while II2 athletes outperformed them in sway area. However, the area parameter demonstrated limited reliability, and its results should therefore be interpreted with caution. Based on these findings, detected inter- and intra-group variability may contribute to building a robust evidence base for future classification development of Para Karate athletes.Keywords: Athlete classification, Balance assessment, Centre of pressure, Kata performance, Posturography Introduction Since 2006, karate has been developed and promoted among athletes with disabilities. Para Karate is a modality of karate focused on developing the ancient discipline of “kata” which involves executing choreographed sequence of martial arts movement. This sport is for wheelchair athletes, thus athletes with physical impairment (PI), athletes with visual (VI) and intellectual impairment (II) and is a fully integrated part of the operations of the World Karate Federation (WKF). The intellectually impaired category is divided into two sport classes: for individuals with an Intelligence Quotient (IQ) of less than 75 (II1) and significant limitations in adaptive behaviour; and II2, for athletes who meet the II1 criteria and also present with a significant additional impairment (e.g. Down Syndrome or mild cerebral palsy). Their impairment must be diagnosed in the developmental stage of their life. To better understand functional differences within these intellectual impairment classes, standardized assessments of gross motor skills are commonly used. Gross motor skills, such as the Test of Gross Motor Development-2 (TGMD-2) are frequently employed to evaluate differences among athletes with intellectual impairments. Another tests, the Bruininks-Oseretsky Test of Motor Proficiency-Second Edition (BOTMP-2) allows the assessment of an array of motor skills of students with normal development as well as the assessment of those with up to moderate motor skill deficits [1]. Studies using the TGMD-2 and BOTMP-2 consistently demonstrate that children and adolescents with intellectual disabilities perform significantly worse than typically developing peers, with those of greater severity showing the most pronounced deficits across coordination, strength, and motor proficiency [2–4]. Interestingly evidence shows that motor deficits are more strongly associated with the severity of intellectual disability, and thus with IQ, than with the specific type of disability [5]. To date, a provisional test battery has been proposed by the WKF, incorporating elements adapted from other Para sports (e.g., swimming, table tennis). While promising, the battery remains in a pilot phase and lacks published data on validity and reliability within the Para Karate population. The development of such tools is consistent with the IPC’s mandate for federations to improve classification systems through multidisciplinary research initiatives [6]. Each Para Karate athlete must undergo a classification before participation in the competition. This classification aims to determine who is eligible to compete in WKF-labelled events, while ensuring that it is not the degree of impairment, but sporting excellence, that ultimately determines which athlete is victorious. A compensation score, in the form of extra points, is given according to the degree of impairment assessed by expert WKF classifiers during the classification session for VI and PI athletes [7]. Objective, evidence-based functional assessment are essential for para-sport classification [8]. In this study, we focus on one key physical ability – balance – as a determinant of kata performance. Balance is fundamental for kata, requiring precise transitions between static stances and dynamic motions, rapid weight shifts, changes of direction, and reactive stability [9, 10]. The central nervous system integrates somatosensory, vestibular, and visual inputs to maintain balance [11, 12]. VI athletes often exhibit increased postural sway and reduced stability due to limited visual input [13, 14], while those with intellectual disabilities may show deficits in motor planning and sensorimotor integration [15, 16]. For instance, individuals with intellectual impairments display decreased postural control, stability, balance skill as well as manual dexterity due to inefficient integration of proprioceptive, vestibular, and visual inputs. These deficits manifest as difficulties in coordinating sensory input with motor output, stemming from altered processing in brain regions associated with motor control and multisensory integration [17, 18]. Furthermore, athletes classified as II2 may present compounded balance limitations resulting from secondary neurological or musculoskeletal conditions [19]. These known deficits provide a rationale for comparing balance performance between these groups in Para Karate. To evaluate balance capacity, the 30-second balance test is frequently used in athletic populations. Most often, this involves maintaining a quiet stance either on one leg or both feet for 30 s, with eyes open or closed. Parameters such as centre of pressure (CoP) displacement, velocity, sway area, and path length, which reflect both static and sensorimotor control strategies, are commonly analysed to quantify postural stability [20, 21]. A posturographic signal can be described in terms of two postural factors: postural accuracy and steadiness [9]. Postural accuracy (quality of movement of CoP) is quantified from the measurement of the sway area of CoP (mm2). A reduced sway area value corresponds to greater stance precision. Steadiness, i.e. the ability to stand as motionless as possible (quantity of movement of CoP), is measured using path of CoP (mm). In karate, several studies have shown that such tests are reliable and sensitive in detecting performance differences among healthy athletes, especially when vision or surface conditions are manipulated [22]. The test is non-invasive, simple to administer, and yields quantifiable metrics. In Para Karate athletes, balance assessment must account for the nature of the impairment. For instance, hearing-impaired Para Karate athletes demonstrated comparable or superior static balance, particularly in single-leg stance, compared to visually impaired athletes [23]. However, little is known about balance performance across different Para Karate impairment categories at the elite level. Given that classification in Para Karate is currently based primarily on medical diagnosis rather than functional assessment, objective tools like CoP analysis could provide additional insight into how impairments translate into sport-specific functional limitations [24]. The current classification system in Para Karate separates athletes into distinct sport classes (VI, II1, II2) based on medical diagnosis and predefined eligibility criteria. However, understanding the functional consequences of these impairments particularly in relation to balance, a key determinant of kata performance, is essential for the future refinement of evidence-based classification. Athletes within the same sport class may exhibit considerable variability in postural control, especially within the VI group, where the degree of visual loss (e.g., B1–B3) can influence reliance on proprioceptive and vestibular feedback. Similarly, II1 athletes with emerging or secondary physical impairments may seek reclassification into the II2 category; thus, objective data on balance performance may aid classifiers in determining the appropriateness of such transitions. As Para Karate seeks alignment with the IPC’s centralized classification system, which emphasizes the impact of impairment on sport performance rather than diagnosis alone, it becomes crucial to quantify functional outcomes such as postural stability. Therefore, this study aims to: (1) compare the balance profiles of Para Karate athletes classified as VI, II1, and II2, and (2) analyse intra- and inter-class variability in postural control using objective measures of static balance. By doing so, we seek to contribute empirical evidence to support the refinement of functionally-based classification criteria aligned with IPC standards [8, 25]. A detailed analysis of performance during balance assessments can elucidate inter-class differences among Para Karate athletes, offering critical insights into specific performance deficits. Targeted interventions based on these findings may contribute to improved training strategies and, consequently, enhanced athletic performance [26]. The results also may aid in the development of an evidence-based classification system. Materials and methods Design This observational, cross-sectional study was conducted in a laboratory setting during the 2023 World Para Karate Championship during official classification session. Postural control was assessed in athletes from different impairment categories without any intervention, using standardized balance measurements. Participants Total of 59 Top-Level Para Karate Athletes divided into three groups according to kata classification of WKF - athletes with visual impairment (VI: n = 24 [17 men and 7 women] aged 23.8 [20.6]), intellectual impairment (II1: n = 17 [8 men and 9 women], aged 26.5 [9.7]) and athletes with II and additional impairment (II2: n = 18 [10 men and 8 women], aged 28.0 [11.3]). The sample consisted of athletes who participated in the World Karate Championship in 2023. Athletes were considered ‘Top-Level’ based on their qualification for the 2023 World Para Karate Championship, an elite-level international competition sanctioned by the WKF. Participation in the study was voluntary, and all eligible athletes and their coaches were informed in advance via official communication from the World Karate Federation (WKF). Prior to data collection, the purpose of the study supporting the development of evidence-based classification for Para Karate and its relevance for sport development and potential Paralympic inclusion was fully explained during the informed consent process. All participants expressed willingness to take part. The participant´s data on body height, body mass, years of experience, or training load were not available. The inclusion criteria were as follows: (1) athletes with visual impairment who provided confirmation of their visual impairment from an ophthalmologist, due to data privacy limitations, detailed International Blind Sports Federation (IBSA) classifications (B1–B3) were not available. In accordance with WKF classification rules, athletes were eligible if they met the minimum impairment criteria consistent with recognized medical standards, demonstrating a visual acuity of LogMAR 1.0 (6/60) or worse and/or a visual field restricted to less than 40 degrees in diameter; (2) athletes with intellectual impairment falling under the criteria determined by VIRTUS eligibility criteria. For the II1 sport class, these criteria include impairment of intellectual functioning (IQ ≤ 75), limitations in adaptive behaviour, and evidence of the impairment during the developmental period; for the II2 sport class, additional impairments physical and/or sensory, such as cerebral palsy, visual impairments, or Down syndrome are also associated [27]. To conduct this study, we received approval from Institutional Review Board of Ethical Committee of the Faculty of Physical Education and Sport of Comenius University in Bratislava, Slovakia, under reference number (UK5/2023). All registered participants at the World Karate Championship or their parent(s)/legal guardian(s) were sent information about the upcoming evidence-based research for classification development, including a description of all the procedures one week prior to the testing session. They were also sent the informed consent with clear explanation of the research’s objectives, activities, data collection methods. All participants or their legal guardians confirmed that the participant’s decision to participate is voluntary. Procedures The testing took place during the classification session indoors in a quiet room to ensure that the athletes were not distracted. Each athlete was accompanied by their assistant or coach, who provided instructions according to the examiner’s guidelines and translated them into the athlete’s native language if necessary. The athletes did not perform any warm-up before the test. After familiarisation, which included a verbal explanation of the test procedure as well as a demonstration of its practical execution, the 30-second bipedal balance test was performed. The test procedure was conducted barefoot. Participants were asked to step onto pre-marked tags with fixed position (small feet marks in the middle of the plates) (e.g. distance apart) on the force plates calibrated according to the manufacturer’s guidelines. They performed the test in a parallel stance, with eyes open, hands on hips, no dual-task condition, meaning no additional tasks or external stimulus. All athletes were also instructed maintain eye contact with a high-contrast picture with simple black symbol of eyes on white paper of A4 format (landscape) placed approximately 1.5 m in front of them on the brown wooden wall, while the B1 athletes (very low visual acuity and/or no light perception) of VI group [28] were instructed maintain upright position of the whole body and head. Each athlete completed two attempts, meaning maintaining the starting position independently throughout exactly 30 s of the test without deviation from this position or required manual assistance preventing falling. They had a 60-second rest interval between trials. For each participant, the best attempt was used for analysis, as this approach minimizes the potential confounding effects of familiarization and fatigue on performance measures. In addition to the verbal explanation, they were also shown a demonstration of the correct position to adopt. If the task was not understood, the entire familiarisation procedure and demonstration were repeated until two successful trials were completed. Signal processing Kistler force plates (9287 C) (Kistler Instruments Ltd., Winterthur, Switzerland) were used to collect the data. Data were sampled at 1 kHz and captured using BioWare (5.3.0.7, Kistler). The raw data recorded were filtered by a second-order Butterworth low-pass filter and a cutoff frequency of 12 Hz following measurement. The parameters obtained included the total path (total displacement) [mm] and area [mm2] of the centre of pressure (CoP) calculated as 95% confidence ellipse. Both path and area of CoP were measured also in anterior-posterior (AP) and medio-lateral (ML) directions [29]. Statistical analysis All statistical analyses were conducted using SPSS for Windows (version 23.0, SPSS, Inc., Chicago, IL, USA). The reliability parameters such as intraclass correlation coefficient (ICC(3,1)), standard error of measurement (SEM) and minimal detectable change (MDC) were calculated for each CoP metric (Table 1). Table 1Medians [IQR] and reliability parameters of the CoP metrics during the 30-second balance testCoP parameterMedian [IQR]ICC (95% CI)SEM (%)MDC (%)Trial 1Trial 2Total path322.7 [93.8]333.80 [97.4]0.728 (0.582–0.829)43.4 (12.4)120.1 (34.3)Path AP231.1 [69.4]227.80 [79.4]0.719 (0.569–0.823)35.1 (14.3)97.4 (39.5)Path ML184.2 [53.3]182.6 [54.7]0.709 (0.552–0.814)28.6 (14.6)79.2 (40.4)Total area322.4 [353.2]383.9 [246.1]0.397 (0.157–0.593)213.8 (51.7)592.7 (143.3)Area AP82.7 [58.1]94.2 [67.7]0.482 (0.259–0.656)35.3 (36.1)97.7 (99.9)Area ML28.4 [26.9]30.3 [15.0]0.574 (0.376–0.722)13.1 (37.9)36.4 (105.2) Medians [IQR] and reliability parameters of the CoP metrics during the 30-second balance test The non-parametric methods were used because the criteria for normality, as determined by the Shapiro-Wilk test, or the homogeneity of variance, as determined by Levene´s test, were not met (see details in Table 2). Table 2Normality and homogeneity of variance tests for all variables across groupsVariableGroupShapiro–Wilk W p Levene’s F p Total pathVI0.9200.0584.0000.024II10.9110.105II20.9310.199Path APVI0.9180.0525.0760.009II10.8230.004II20.9790.933Path MLVI0.8830.0010.5000.640II10.8960.058II20.9210.136Total areaVI0.9270.0834.6380.014II10.9560.556II20.9640.686Area APVI0.9130.0413.2430.047II10.9810.964II20.9540.491Area MLVI0.796< 0.0010.5540.578II10.773< 0.001II20.8860.033 Normality and homogeneity of variance tests for all variables across groups Therefore, non-parametric descriptive statistics as the median and inter-quartile range [IQR] were used. Using the interquartile range (IQR) rule: lower bound = Q1 − 1.5 × IQR; upper bound = Q3 + 1.5 × IQR, any data point outside this range was considered an outlier. The differences among groups were calculated by the Kruskal-Wallis H test. When a Kruskal–Wallis H test was significant, post-hoc pairwise comparisons were conducted using the Mann–Whitney U test. P-values were adjusted for multiple comparisons using the Holm correction. The effect size “r” [r = |Z|/sqrt(N)] (where N was the sum of participants in the two groups being compared) was used to determine the effect size with the following interpretation [30]: 0.1 ≤ r < 0.3 (small effect), 0.3 ≤ r < 0.5 (medium effect), r ≥ 0.5 (large effect). The level of significance was set at p < 0.05. Results The reliability analysis showed that the path parameter reached ICC values slightly above 0.70, with SEM between 12 and 14%. For the area parameter, however, reliability was markedly lower: ICC = 0.397 for the total value, slightly higher for the AP direction (ICC = 0.482) and the ML direction (ICC = 0.574), with corresponding SEM values of ~ 52% for total and 36–37% for AP and ML directions. The mean difference between the first and second trial was 121.6 mm². The majority of participants (53 out of 59) showed differences smaller than 200 mm², only six exceeded this threshold (two > 200 mm², one > 300 mm², and three > 500 mm², including one with a difference of approximately 1000 mm²). Given that the mean performance across both trials was ~ 412 mm², most athletes demonstrated relatively small trial-to-trial variation, whereas a few extreme cases markedly increased the estimates of variability (SEM, MDC). Descriptive data expressed as medians [IQR] for each group (VI, II1 and II2) and parameter of the 30-second balance test measured (path and area – total, AP and ML direction) are comprised in the Table 3. Table 3Descriptive data for each group and parameter measuredGroup (n)VI (24)II1 (17)II2 (18)Parametermedian [IQR]median [IQR]median [IQR]Path of CoP [mm]total352.7 [145.7]277.3 [58.0]316.6 [54.0]AP254.6 [118.8]184.9 [45.1]227.8 [56.3]ML196.2 [77.2]167.8 [46.2]168.1 [58.1]Area of CoP [mm2; mm]total400.5 [344.7]305.1 [253.2]259.8 [122.1]AP96.9 [49.2]81.3 [48.7]62.8 [28.8]ML28.4 [18.9]21.2 [14.1]24.1 [12.7] Descriptive data for each group and parameter measured The results showed significant differences in the length of the pathway of the centre of pressure (CoP) during the 30 s balance test among the three groups of Para Karate athletes (H(2) = 13.205, p = 0.001). Specifically, post-hoc pairwise comparisons using the Mann–Whitney U test with Holm correction indicated that VI and II1 differed significantly (padj = 0.001) with large effect size (r = 0.55). The comparison between II1 and II2 showed a trend towards significance (raw p = 0.035), but this difference did not remain significant after Holm adjustment (padj = 0.070). However, medium effect size was detected (r = 0.36). The groups significantly differed also in path in anterior-posterior (AP) (H(2) = 15.507, p < 0.001) and medio-lateral (ML) planes (H(2) = 6.669, p = 0.036). Athletes of II1 class achieved significantly lower median of path´s length in AP direction compared to VI and II2 groups (padj = 0.001 and padj = 0.001, respectively) with medium to large effect sizes (r = 0.57 and r = 0.48). In the ML direction, a medium effect size was observed in favour of group II1 compared to VI (r = 0.37), although the difference did not reach statistical significance after Holm correction (padj = 0.051). The comparisons of groups in the total path and path in AP and ML directions are shown in Figs. 1, 2 and 3. Comparison of the total path of CoP in the 30 s bipedal stance test among Para Karate athletes with different impairments Comparison of the anterior-posterior path of CoP in the 30 s bipedal stance test among Para Karate athletes with different impairments Comparison of the medio-lateral path of CoP in the 30 s bipedal stance test among Para Karate athletes with different impairments Significant differences in the sway area of the CoP among Para Karate groups were found (H(2) = 12.929, p = 0.002). More specifically, karatekas of II2 have significantly lower median [IQR] of area in the 30-second balance test compared to those with VI (padj < 0.001) with large effect size (r = 0.55). No differences between VI and II1 (padj = 0.094, r = 0.31 - medium effect) and between II1 and II2 groups were found (padj = 0.184; r = 0.23 – small effect). Para Karate groups significantly differed also in AP direction (H(2) = 22.465, p < 0.001), but not in ML direction (H(2) = 3.259, p = 0.196). Karatekas of II2 group recorded significantly lower median [IQR] of area in AP direction compared to both VI and II1 groups (padj = 0.002, and padj = 0.040, respectively) with medium to large effect sizes (r = 0.73 and r = 0.39, respectively). Additionally, II1 class showed significantly lower median [IQR] than VI group, with a medium effect size (padj = 0.040, r = 0.35). When examining performance ranges within subgroups using the better trial of the two, we found very wide differences: in the VI group, total area ranged from ~ 142.7 mm² (best) to ~ 929.3 mm² (worst); in II1 from ~ 84.8 mm² to ~ 631.3 mm²; in II2 from ~ 141.6 mm² to ~ 498.6 mm². The comparisons of groups in the total area and path in AP and ML directions are shown in Figs. 4, 5 and 6. Comparison of the total area of CoP in the 30 s bipedal stance test among Para Karate athletes with different impairments Comparison of the anterior-posterior area of CoP in the 30 s bipedal stance test among Para Karate athletes with different impairments Comparison of the medio-lateral area of CoP in the 30 s bipedal stance test among Para Karate athletes with different impairments Discussion As a first step, we evaluated the trial-to-trial reliability of the 30-second bipedal stance test. While the Path parameter demonstrated acceptable reliability, the area parameter showed limited and potentially problematic reliability, largely due to a few extreme trial-to-trial discrepancies that strongly inflated estimates of variability. In para sport populations, it is common to observe greater trial-to-trial variability than in able-bodied athletes, which complicates reliability estimation (e.g. larger time-differences in Para vs. able-bodied skiers) [31]. Despite this, most participants exhibited relatively small differences between trials. The observed pattern suggests that the limited reliability of the area parameter reflects the heterogeneity of the study population rather than systematic measurement error. However, due to the poor reliability of the area parameter, any statistical findings (e.g., p-values or effect sizes) involving this metric must be interpreted with a considerable degree of caution. This limitation represents a fundamental weakness that must be kept in mind throughout the interpretation of area-related findings. Extreme performances can disproportionately influence overall reliability coefficients, while the majority of athletes demonstrate relatively stable trial-to-trial performance. It should also be noted that even in able-bodied populations, the 30-second bipedal stance test has shown only modest reliability for comparable sway distance parameters with eyes open conditions (ICC = 0.37–0.57) [32] suggesting that certain CoP measures may inherently exhibit lower stability regardless of population characteristics. The findings of this study demonstrate that the 30-second bipedal balance test is sensitive to impairment-related differences in postural control differences among Para Karate athletes with various types of impairments. Significant differences were observed in the total path length of CoP across the three monitored Para Karate groups, where athletes with II1 exhibited significantly lower CoP path values compared to VI group. Interestingly, the raw Mann–Whitney U test suggested a potential difference between II1 and II2, but this effect did not survive correction for multiple comparisons. This indicates that while there may be a tendency towards a difference, the evidence is not strong enough to conclude statistical significance once the increased risk of Type I error from multiple testing is controlled for. However, from a practical perspective, the medium effect size still suggests a potentially meaningful difference between the groups that warrants further investigation. The reduced sway observed in II1 athletes may reflect a conservative postural strategy, characterized by reduced mobility but not necessarily superior functional control in dynamic sport-specific contexts. Additionally, no significant differences were found between the VI and II2 groups, indicating possible similarities in postural strategies under eyes-open, bipedal stance conditions. Significant differences were found only in AP direction, more specifically, II1 outperformed both VI and II2 groups. Further significant differences were found in the sway area of the CoP during the 30-second balance test. Although the sway area parameter demonstrated limited trial-to-trial reliability, the analysis indicated that athletes with VI exhibited a significantly greater sway area, suggesting reduced postural control in static standing conditions. A medium effect size also points to a potentially meaningful difference between the VI and II1 groups, even though this was not statistically significant after correction. When analysing directional sway, significant differences emerged in the AP direction, but not in the ML plane, among groups. Specifically, athletes with additional impairments (II2) demonstrated the smallest AP sway area, followed by athletes of II1, both significantly outperforming their VI counterparts. These findings indicate a potential trend in directional sway control; however, they should be interpreted with caution given the lower reliability of the area parameter and the relatively high intra-group variability observed in this population. Since there is a notable lack of research focused exclusively on Para Karate, direct comparisons with existing literature are limited. Previous research on para-sports in general highlights the challenge of developing classification systems that fairly distinguish functional capacities without relying solely on medical diagnosis [8, 24]. Thus, most comparisons here are drawn from studies on able-bodied karate athletes and athletes with disabilities in other sports. While these sources provide valuable context, they may not fully capture the sport-specific demands of Para Karate athletes. Moreover, the current discussion is more oriented toward neuromuscular and sensorimotor aspects of postural control (e.g., reliance on visual, proprioceptive, and vestibular inputs; compensatory mechanisms; and biomechanical stability) rather than sport-specific performance outcomes. The significantly lower CoP path length observed in athletes with II1, compared to those with VI, align with previous findings suggesting that visual input plays a crucial role in maintaining postural stability. As vision is a dominant sensory modality for balance, athletes with VI typically rely more heavily on proprioceptive and vestibular inputs, which may not fully compensate in static conditions [33–35]. These findings are consistent with previous research indicating that visually impaired individuals exhibit greater postural sway, particularly under eyes-open conditions, compared to sighted individuals [36, 37]. The VI group likely included athletes with varying degrees of impairment, from partial to complete vision loss. Interestingly, athletes with II2 showed postural control outcomes that were intermediate between the II1 and VI groups in total path andwith significantly lower total area compared to VI group. This is consistent with previous studies indicating that individuals with additional impairment, especially Down Syndrome, often present with hypotonia, ligamentous laxity, decreased muscle strength and impaired proprioception, all of which contribute to motor coordination deficits and balance challenges [38, 39]. However, their preserved visual input may offer an advantage over athletes with VI in static, visually guided tasks. The significantly greater total and AP path length observed in the VI group compared to both II1 and II2 groups suggests that visual impairment imposes considerable challenges to postural control in static conditions. This is consistent with findings on sensorimotor integration, showing that when sensory inputs are restricted (e.g., vision, proprioception), the postural control system often compensates through altered feedback mechanisms. For example, Peterka [33] demonstrated that postural regulation changes when visual or support-surface information is degraded, requiring adjustments in stiffness and damping. Similarly, expertise in activities requiring fine postural control (such as gymnastics) is associated with reduced attentional demand and more stable sway under challenging stance conditions [35]. Interestingly, athletes in the II1 group demonstrated the lowest total and AP path length, suggesting a markedly reduced sway distance. One plausible explanation is the adoption of a more conservative or rigid postural strategy, often seen in individuals with intellectual disabilities, who may prioritize stability over adaptability to reduce perceived risk of imbalance [40]. However, such a reduction in sway distance does not necessarily reflect superior dynamic postural control or sport-specific readiness, particularly in activities that require responsive balance adjustments. Interestingly, athletes in the II2 group, who have intellectual impairments combined with additional motor or sensory limitations exhibited the smallest sway area, particularly in the AP direction. This might reflect not better balance, but rather reduced exploratory postural behaviour or motor output, possibly due to greater neuromotor constraints [41, 42]. Such a reduction in sway could stem from increased stiffness, passive stabilization strategies, or limited motor variability, which are not indicative of superior postural control but rather a compensatory reduction in movement degrees of freedom. Moreover, the greater sway in the AP plane across groups, compared to the ML direction, is consistent with previous biomechanical studies that highlight greater instability in the sagittal plane during quiet standing [43]. The most marked group differences were found in the AP direction that may reflect functional differences in how each impairment affects anterior-posterior postural control. This is particularly relevant in karate stances involving weight shifting and stepping forward/backward. Interestingly, the Para Karate athletes demonstrated comparable best performances in the 30-second bipedal stance balance test. This suggests that their top-level balance ability was similar across classes. However, noticeable differences emerged in the maximum values (i.e., the poorest performances within each group), indicating greater variability and the presence of individuals with lower balance ability in some groups. Such intra-class variability reinforces the need for functional classification systems that account for individual ranges of performance, rather than relying solely on diagnostic criteria. Performance profiling or composite classification models could provide greater fairness and alignment with IPC standards [44]. Collectively, these results support the hypothesis that the 30-second bipedal balance test is sufficiently sensitive to detect between-group differences in postural control related to specific impairments. They also highlight the importance of considering the sensory and motor characteristics of each disability when interpreting balance performance, rather than assuming uniform deficits across Para Karate athletes. Given that kata frequently involves sagittal plane weight shifts and forward/backward stepping, the AP sway measures identified in this study may be particularly relevant for future functional classification tests, as well as for targeted training interventions to enhance kata performance. These findings are hypothesis-generating and point to the need for future studies that directly link CoP-derived balance metrics, such as AP sway, with classifier decisions and kata performance scores. Such prospective investigations align with IPC’s 2025 Classification Code, which emphasizes the development of evidence-based classification systems through valid, reliable, and sport-specific functional assessments. Limitations of the study Despite the strengths of this study, several limitations should be acknowledged. First, the classification of visually impaired (VI) athletes lacked granularity, as data on IBSA sub-classes (e.g., B1–B3) were not available. This may have introduced intra-group variability in postural control, as different levels of residual vision can differentially affect balance performance. Second, the use of a static bipedal stance test, while practical and reliable, may not fully reflect the dynamic balance demands of kata, which involves rapid directional changes, weight transfers, and complex motor sequences. Moreover, best-of-two can bias results toward stability and reduce variability. An additional limitation concerns the reliability of the area parameter, which showed lower consistency across trials and was strongly affected by intra-group heterogeneity and a few extreme performances. These factors inflated overall variability indices (e.g., ICC, SEM, CV) and should be considered when interpreting group differences in sway area. Future studies should incorporate dynamic or task-specific assessments to better approximate sport-specific functional performance. Also, the lack of a standardized warm-up protocol represents a potential limitation, as it may have influenced neuromuscular activation levels and subsequently affected performance measures. More, the study did not include a control group of able-bodied karate athletes, which limits the interpretability of impairment-related deficits relative to normative performance. Finally, the sample consisted exclusively of elite-level athletes selected for World Championship participation, which, while ensuring a high-performance cohort, may limit the generalizability of findings to lower-tier or developmental Para Karate populations. Implications for future research Absence of IBSA sub-classes (B1–B3) leads to recommendation of future stratified analyses. Future research should also employ dynamic or sport-specific balance assessments to more accurately capture functional differences and training effects. In addition, the use of wearable sensors and AI-assisted analysis during kata performance would allow ecologically valid evaluation beyond static tests, aligning with IPC priorities to integrate technological innovation into classification systems and supported by emerging evidence on AI and wearable technologies in sports biomechanics [45]. Including an able-bodied control group would aid in contextualizing Para Karate athletes’ balance performance. Additionally, exploring the relationship between balance measures and kata performance, as well as test responsiveness to training interventions across impairment types, will provide further insights. Conclusions The 30-second bipedal balance test with eyes open sensitively distinguished postural control among Para Karate classes, particularly in the AP direction. The test distinguished between groups, with VI athletes demonstrating significantly greater CoP path than II1 athletes and greater CoP area than II2 athletes. In the anterior-posterior direction, II1 athletes outperformed both groups in the path parameter, while II2 athletes showed a superior sway area compared to both groups. This study demonstrates potential for integrating balance testing into evidence-based classification for Para Karate, showing that objective CoP measures can capture impairment impact on sport performance beyond medical diagnosis. However, the area parameter showed limited reliability due to high trial-to-trial variability in some participants, and its results should therefore be interpreted with caution. Consequently, findings based on CoP area should be considered exploratory until further validated by studies with improved measurement consistency. Despite this limitation, the generally acceptable reliability of the path parameter and moderate intra-group variability suggest that postural stability metrics can provide functionally meaningful information for classification. Identifying athletes with stable versus inconsistent balance control could help refine evidence-based class profiles and improve the objectivity of classification decisions. The methods used here require further validation and cannot yet be applied within official classification protocols. These results provide preliminary functional indicators rather than finalized criteria. For VI athletes, reduced postural control supports the need for proprioceptive training; for II1 athletes, relatively stable AP control suggests functional capacity under static conditions; and for II2 athletes, postural instability reflects compounded impairment effects. These findings can inform classifier training and guide further evidence-based classification development in line with IPC’s classification research priorities. Abbreviations Artificial intelligence Anterior-posterior Bruininks-Oseretsky Test of Motor Proficiency- Second Edition Confidence interval Centre of pressure International Blind Sports Federation Coefficient of variation Intraclass correlation coefficient International Paralympic Committee Intelligence Quotient Interquartile range Intellectual impairment Intellectual impairment - individuals with an IQ of less than 75 and significant limitations in adaptive behaviour Intellectual impairment – individual who meet the II1 criteria and also present with significant additional impairment Minimal detectable change Medio-lateral Physical impairment Standard deviation Standard error of measurement Test of Gross Motor Development-2 Visual impairment World Karate Federation Authors’ contributions HH conceived and designed the study. RH, MH and DA collected the data. HH and RH performed the statistical analyses. HH drafted the manuscript. RH, MH and DA contributed to the interpretation of data and critical revision of the manuscript. All authors read and approved the final manuscript. Funding This research was supported by the Scientific Grant Agency of the Ministry of Education, Science, Research and Sport of the Slovak Republic and the Slovak Academy of Sciences (No. 1/0611/23 and 1/0725/23). Data availability The data that support the findings of this study are not openly available due to reasons of sensitivity and privacy of the individuals involved as they are part of the protected documents available mainly to the classifiers of the classification process under the auspices of the WKF. The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate The study was conducted in accordance with the ethical standards of the responsible institution and the Declaration of Helsinki. Approval was obtained from the Institutional Review Board of Ethical Committee of the Faculty of Physical Education and Sport of Comenius University in Bratislava, Slovakia, under reference number (UK5/2023). Written informed consent was obtained from all participants involved in the study. Informed consent was obtained from all participants, with written consent from the parent(s)/legal guardian(s) for those under 18. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The data that support the findings of this study are not openly available due to reasons of sensitivity and privacy of the individuals involved as they are part of the protected documents available mainly to the classifiers of the classification process under the auspices of the WKF. The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.