Increased Functional Mobility in Healthy Elderly Individuals After Six Months of Adapted Taekwondo Practice

Abstract

Age-related muscle mass and force reduction may ultimately pave the way for loss of independence, reduced quality of life (QoL) and increased falls, which represents one of the primary causes of hospitalization and death among the elderly. Physical exercise (PE) is widely proposed to reduce the risk of falls (RoF) in the elderly, and Taekwondo (TKD) seems a particularly suitable activity for this purpose. Therefore, this single-arm longitudinal observational study aims to evaluate (a) the effects of a six-month adapted TKD course on RoF by means of an instrumented version of the timed up and go test (TUG) and (b) differences after six months of TKD on TUG temporal and kinematic variables. Twenty elderly participants of both sexes (age: 64.6 ± 4.2 years; mass: 76.0 ± 15.0 kg; stature: 1.63 ± 0.10 m) were assessed before (T0) and after (T6) six months of adapted TKD by means of the i-TUG test. TUG, walking phases and time to sit duration were shorter than at T0. Furthermore, during walking phases, antero-posterior linear acceleration increased significantly. The adapted TKD course was suitable to improve functional mobility in the elderly, and the use of the instrumented TUG allowed us to further our understanding of the mechanisms underlying these improvements.

1. Introduction

Health and socio-economic progress, combined with the increased prevention of chronic diseases and reduced mortality, mean mankind is faced with increased longevity [1]. A longer senescence involves physiological and para-physiological modifications necessary to manage and maintain elderly health. Age-related modifications are heterogeneous and widespread to all body systems; among them, muscle–skeletal systems may be affected by pathological conditions like osteoporosis and sarcopenia [2]. Physiologically, the elderly are subjected to a decrease in muscle mass caused by a loss of between 10 and 40% of type 2 fibre in terms of number and size [2,3]. This loss is exacerbated by physical inactivity: ten days of immobilization is associated with a loss of one kg of muscular mass [3]. Furthermore, muscle fibre undergoes a reduction of 75% in the transversal section, which is directly associated with a reduction in strength, particularly in the lower limbs [4,5]. These modifications, influenced by behavioural factors [2], may heavily impact the quality of life of the elderly [6] and their ability to perform activities of daily living (ADLs), drastically increasing their dependence on caregivers [7]. Moreover, the independence and QoL of the elderly are threatened by falls, which correlate with weakness [8] and low levels of physical performance [9]. Falls constitute a relevant issue in elderly care because they are responsible for increased rates of morbidity, comorbidity and death in this population: individuals that have fallen have a higher risk of falling again within a year, and most of them do not recover their pre-accident functional levels [9]. Furthermore, falls are the seventh most common cause of death in the elderly: one-third of patients who suffer hospitalization after a fall, especially for hip fractures, do not survive more than a year [9,10]. Therefore, enhancing autonomy and QoL and reducing the risk of falls in the elderly are important goals to achieve for the care of older age groups [11] and to reduce medical costs associated with hospitalization. The risk of falls can be clinically assessed using different types of tests. One of the most employed is the “timed up and go” test (TUG) [12]. TUG is a widely used screening test in geriatrics that is designed to assess static and dynamic balance. The ability and speed of ambulation, mobility and functional strength of an elderly individual is essential in performing ADLs. The test consists of different phases, namely, rise from a chair, walk forward for three meters, turn around 180°, walk back towards the chair, turn around 180° and sit on the chair. Although TUG is widely used for the functional evaluation of the elderly, its effectiveness as a tool for the prediction of the risk of falls in the elderly remains a matter of debate [13,14]. Recently, TUG has been implemented using an inertial measurement unit (i-TUG) [15,16]. This implementation has been shown to improve predictive test accuracy by up to 80%, enabling the more accurate preventive detection of elderly individuals at risk of fall. Physical activity (PA) and physical exercise (PE) are widely used in elderly care [9,17] due to their impact on the age-related modifications of improving strength, resistance and body mechanics [10,18]. Indeed, it is well established that PE has the potential to reduce the risk of falls in the elderly by up to 25% [19]. Recently, martial arts have been broadly considered as an effective approach to prevent falls in this population [20,21,22]; as reported by Miller and colleagues [23], martial arts practice improves limb strength, flexibility, gait speed and balance (both static and dynamic), irrespective of gender. In this framework, Bubela and colleagues [24] and Kim and colleagues [22] evaluated the effect of Tai Chi [24] and Taekkyon [22] on functional mobility, as assessed through the TUG test. Both studies reported a significant reduction in TUG duration. Noteworthily, Bubela and colleagues [24] recruited participants who had previous experience in Tai Chi, whereas Kim and colleagues [22] recruited only women.

On the other hand, Taekwondo (TKD), a dynamic discipline that encompasses movements such as kicks and strikes and requires strength, flexibility and agility [25], may represent a suitable yet innovative exercise for the elderly. As reported by recent studies on this topic [25,26], elderly individuals practicing TKD show increased gait speed, balance, lean body mass [25,26] and handgrip strength [25]. In particular, a study by Li and colleagues [25] evaluated the effect of an adapted eleven-week TKD protocol on the functional mobility of fifteen older individuals: after the training, a significant improvement in functional mobility was observed, as shown by a reduction in the TUG test duration. Similar outcomes were observed by Cromwell and colleagues [27] in twenty participants after eleven weeks of a TKD protocol specifically designed for older adults. The effectiveness of TKD in improving functional abilities in the elderly may possibly be accounted for by the functional demands placed on the lower limbs to maintain balance during this activity [27,28]. Furthermore, the effectiveness of TKD may be accounted for by its dynamic nature, as opposed to other martial arts (i.e., Tai Chi), which are predominantly characterized by movements performed slowly and gradually [27]. The specific characteristics of TKD may result in the enhanced walking ability that is usually compromised in old individuals.

Moreover, even if TKD has been already proposed as an effective intervention to ameliorate functional mobility assessed by means of the TUG test, to the best of our knowledge, an objective evaluation of the effect of an adapted TKD training of the TUG subtask has not been performed so far. In this respect, the i-TUG test has not been used to evaluate the TKD effects.

Therefore, in light of the above considerations, the aim of this study is twofold: first, to evaluate the effect of a six-month adapted TKD exercise on risk of falls as assessed by the i-TUG; and second, to evaluate differences, if any, on gait kinematics and temporal i-TUG parameters after six months of training. We hypothesize that six months of adapted TKD training may result in an improved functional mobility in a group of elderly individuals. Specifically, since balance and power may be effectively increased by TKD practice, we expect that after the adapted TKD, the elderly individuals will be able to perform the turning phases of TUG more effectively.

2. Materials and Methods

2.1. Study Design

This study was conceived as a single-arm longitudinal observational study. To investigate the effectiveness of an adapted TKD course on functional mobility in older adults, participants were evaluated before (T0) and after (T6) six months of TKD. In each testing session, the participants were asked to perform two trials of the timed up and go test (TUG) at their normal walking pace.

2.2. Participants

Thirty-eight participants were initially recruited for this study. Thirteen of them dropped out during the training protocol and five were not able to attend the post-intervention session. Therefore, twenty older adults (females n = 9; males n = 11) were included in the final analysis. The anthropometric characteristics of the participants at T0 are reported in

Table 1. Anthropometric characteristics of the participants at T0. Acronyms: F = females; M = males; n = number of participants. Data are mean and standard deviation.

	Participant Characteristics
	F (n = 9)	M (n = 11)
Age (years)	67 ± 3.3	62.5 ± 3.9
Stature (m)	1.53 ± 0.02	1.71 ± 0.06
Body Mass (kg)	66.3 ± 7.4	84 ± 15.1

. An a priori sample size calculation was performed using the software (G-Power 3.1.9.2) with the results of the TUG duration reported in the study of Park and colleagues [26]. This analysis showed that seventeen participants were sufficient to achieve a statistical power of β = 0.95 with an effect size of 0.48 and significance level of α = 0.05. A conservative drop-out rate of 50% was taken into account for the sample-size calculation. Participants included in the study were self-sufficient, physically active and reported no physical impairment preventing them from walking and no contraindication to the practice of physical activity. They were asked to provide medical approvement for the practice of light-intensity physical activity. Participants were excluded from the study if they reported any contraindication to physical activity; neurological, cardiovascular or respiratory disorders; had a history of osteo-articular injuries in the six months preceding the study; failed to provide the medical approvement or if they had previous experience in TKD. Before the first testing session, the study aims, procedures, risks and benefits were carefully explained; moreover, participants were asked to maintain their usual activity levels for protocol duration. Participants were asked to sign a written informed consent approved by the Ethical Committee of Calabria region (n° 86/16 March 2023). All the procedures of this study comply with the Declaration of Helsinki on studies with human participants.

2.3. Experimental Protocol

In both assessments (T0; T6), participants’ age and anthropometric characteristics were recorded. Both assessments were held in the morning. Body weight was measured by using an InBody scale (InBodyR20, Seoul, Korea), whereas height was recorded using a measuring tape to the nearest 0.01 m. In each testing session, participants were asked to perform two trials (T1; T2) of the timed up and go test (TUG). For this test, they had to stand up from a chair (without using their arms), walk at their normal speed for three meters (marked on the floor using a cone), circle the cone, walk back to the chair, turn around (180° turn) and sit down [12,29]. The test was performed completely barefoot on a habitual Taekwondo mat. During the TUG test, participants were equipped with a tri-axial inertial measurement unit (IMU; G-SENSOR 2, BTS S.p.A, Padova, Italy; sampling frequency: 200 Hz; full scale: ±2 g; 2000°/s for the accelerometer and gyroscope, respectively). The IMU was firmly attached to each participant’s lower back, approximately over the L4 vertebrae, using elastic and adhesive straps. After the six-month intervention, the participants were retested. The adapted TKD training consisted of two weekly supervised sessions, each lasting 1 to 1.5 h, and was structured from a professional TKD trainer of the Italian Taekwondo Federation (FITA). The training course had a progressively increasing difficulty and with a low–moderate intensity. Each training session included coordination exercises, which were basic punching and kicking techniques of TKD interspersed with a large recovery time to maintain a low–moderate intensity. In this way, the training did not have a strong cardiovascular impact, keeping the focus of the training on learning basic punching and kicking techniques. In particular, the adapted TKD practice aimed at allowing participants to learn the first two forms (poomsae).

2.4. Data Analysis

The extraction of the features of the i-TUG was carried out by analysing the data collected by the IMU. All the analyses detailed in the following paragraphs were carried out using MATLAB software (The Mathworks, version 2023b). To obtain a stable and repeatable reference system across participants, trials and assessments, the orientation of the sensor was aligned with the direction of gravity, when participants were in a static position [30]. This procedure enables the axes of the sensor to approximate the cranio-caudal, medio-lateral and antero-posterior anatomical axes. Subsequently, to remove possible high-frequency noise, angular velocity and linear acceleration about the three axes were filtered using a 4th-order low-pass Butterworth filter with a cut-off frequency of 25 Hz [30,31,32]. The TUG was segmented into six subtasks, as reported by Weiss et al. [33]. Namely, the subtasks were sit to stand; forward walk; mid turn; return walk; turn to sit and stand to sit. This segmentation has been claimed as the most appropriate to segment the TUG test by a previous systematic review [15].

The beginning of the sit-to-stand (SITSTA_START) and stand-to-sit (STASIT_START) phases was detected on the medio-lateral component of the angular velocity signal as the first point deviating from zero. Similarly, the end of the above-mentioned phases (SITSTA_END; STASIT_END) was identified on the same angular velocity component as the first time point returning to zero (please see Figure 1 for a detailed description). Sit-to-stand (SITSTA_DUR) and stand-to-sit (STASIT_DUR) duration were computed as the time that elapsed between SITSTA_START and SITSTA_END and STASIT_START STASIT_END. The mid-turn and turn-to-sit phases were detected on the cranio-caudal component of the angular velocity signal. On this latter signal, two peaks were detected representing the peak angular velocity about the cranio-caudal axis of the mid-turn (^MIDTURω_CC-PEAK) and turn-to-sit (^TURSITω_CC-PEAK) phases. Before and after each peak, the time points in which the cranio-caudal angular velocity was equal to zero were identified as the beginning and the end of the mid-turn (MIDTUR_START; MIDTUR_END) and turn-to-sit (TURSIT_START; TURSIT_END) phases. The durations of these phases (MIDTUR_DUR; TURSIT_DUR) were computed as the time interval between MIDTUR_START and MIDTUR_END for the mid-turn phase and TURSIT_START and TURSIT_END for the turn-to-sit phase. Forward walk duration (FORWAL_DUR) was computed as the time interval occurring between SITSTA_END and MIDTUR_START, whereas the return walk duration (RETWAL_DUR) was computed as the time interval that elapsed between MIDTUR_END and STASIT_END. The total duration of TUG was computed as the time that elapsed between SITSTA_START and TURSIT_END. The root mean square (RMS) of the three-dimensional linear acceleration signal was computed during the forward walk (^APRMS_FW; ^CCRMS_FW; ^MLRMS_FW) and return walk (^APRMS_RW; ^CCRMS_RW; ^MLRMS_RW). Moreover, to provide a comprehensive description of the TUG test, the peak trunk flexion angles during sit to stand (^SITSTAα_FLEX) and stand to sit (^STASITα_FLEX) were computed by integrating the filtered medio-lateral angular velocity signal. Last, the peak angular velocity during the trunk flexion of the sit-to-stand and stand-to-sit (^SITSTAω_FLEX; ^STASITω_FLEX) phases were identified as the positive peaks on the medio-lateral component of the angular velocity. Conversely, the peak angular velocities during trunk extension of sit to stand and stand to sit (^SITSTAω_EXT; ^STASITω_EXT) were computed as the negative peaks occurring on the medio-lateral component of the angular velocity. All the variables defined in this paragraph are described in

Table 2. Acronyms and respective definitions of variables extracted from i-TUG analysis.

Acronyms	Variable Definitions	Unit
TUGDUR	Overall TUG duration = Δt (Stasitend − Sitstastart)	s
SITSTADUR	Sit-to-stand phase duration = Δt (Sitstaend − Sitstastart)	s
FORWALDUR	Stand-to-sit phase duration = Δt (Stasitend − Stasitstart)	s
MIDTURDUR	Mid-turn phase duration = Δt (Midturend − Midturstart)	s
RETWALDUR	Turn-to-sit phase duration= Δt (Tursitend − Tursitstart)	s
TURSITDUR	Forward walk phase duration= Δt (Midturstart − Sitstaend)	s
STASITDUR	Return walk phase duration =Δt (Stasitend − Midturend)	s
MIDTURωCC-PEAK	Peak angular velocity about the cranio-caudal axis of the mid-turn phase	°/s
TURSITωCC-PEAK	Peak angular velocity about the cranio-caudal axis of the turn-to-sit phase	°/s
APRMSFW	Root mean square of linear acceleration signal on antero-posterior axis during forward walk	m/s2
CCRMSFW	Root mean square of linear acceleration signal on cranio-caudal axis during forward walk	m/s2
MLRMSFW	Root mean square of linear acceleration signal on medio-lateral axis during forward walk	m/s2
APRMSRW	Root mean square of linear acceleration signal on antero-posterior axis during return walk	m/s2
CCRMSRW	Root mean square of linear acceleration signal on cranio-caudal axis during return walk	m/s2
MLRMSRW	Root mean square of linear acceleration signal on medio-lateral axis during return walk	m/s2
SITSTAαFLEX	Peak trunk flexion angle during sit to stand	°
STASITαFLEX	Peak trunk flexion angle during stand to sit	°
SITSTAωFLEX	Peak angular velocity during trunk flexion of sit to stand	°/s
STASITωFLEX	Peak angular velocity during trunk flexion of stand to sit	°/s
SITSTAωEXT	Peak angular velocity during trunk extension of sit to stand	°/s
STASITωEXT	Peak angular velocity during trunk extension of stand to sit	°/s

.

2.5. Statistical Analysis

All statistical procedures detailed in this manuscript were carried out using the Statistical Package for Social Sciences (SPSS version: 20 IBM, Milan, Italy). All variables were tested for normal distribution using the Shapiro–Wilk test. Subsequently, to investigate the effect of the adapted TKD intervention, the variables of interest were submitted to separate repeated measure analysis of variance (RM-ANOVA) with time of assessment (T0 vs. T6) and trials (T1 vs. T2) as repeated measures. Where significant interaction effects were observed, pairwise comparisons were carried out. Bonferroni correction for multiple comparisons was applied. For all the statistical procedures detailed in the present paragraph, a significance level of 0.05 was adopted. Effect size and achieved power are reported by means of μ² and 1-β, respectively. Data are reported as mean and SD and 95% confidence intervals (CIs).

3. Results

After six months of TKD, a significant reduction in BMI (p = 0.047, μ² = 0.19; 1-β = 0.52) was observed (T0 = 28.5 ± 3.5; CI: [26.86, 30.14] vs. T6 = 28.0 ± 3.1; CI: [26.50, 29.46]). For clarity, temporal and kinematic variables of the i-TUG will be described in two separate paragraphs.

Temporal variables:

The RM-ANOVA showed a significant main effect of time on TUG_DUR (p = 0.002, μ² = 0.40; 1-β = 0.92), FORWAL_DUR (p = 0.002, μ² = 0.41; 1-β = 0.93), RETWAL_DUR (p = 0.001, μ² = 0.47; 1-β = 0.97) and STASIT_DUR (p = 0.043, μ² = 0.20; 1-β = 0.54). Specifically, at T6, TUG_DUR (T0 = 9.40 ± 1.06; CI: [8.90, 9.90] vs. T6 = 8.60 ± 0.96; CI [8.15, 9.05]), FORWAL_DUR (T0 = 1.87 ± 0.33; CI: [1.71, 2.02] vs. T6 = 1.56 ± 0.38; CI: [1.38, 1.73]), RETWAL_DUR (T0 = 1.67 ± 0.33; CI: [1.52, 1.83] vs. T6 = 1.39 ± 0.29; CI: [1.25, 1.52]) and STASIT_DUR (T0 = 2.11 ± 0.28; CI = [1.98, 2.25] vs. T6 = 1.87 ± 0.40; CI: [1.69, 2.06]) were all shorter than at T0. Moreover, a significant effect of trial was observed for TUG_DUR (p = 0.0016), with T2 being shorter than T1, and for TURSIT_DUR (p = 0.007), with T1 being shorter than T2. Furthermore, a significant time by trial interaction (p = 0.023) was observed for MIDTUR_DUR. Following Bonferroni correction, none of the pairwise comparisons were significant. No significant effect of time, trial or time by trial interaction was observed for the other variables.

Kinematics variables

Considering kinematic variables, the RM-ANOVA revealed a significant effect of time on ^APRMS_FW (p = 0.031; μ² = 0.22; 1-β = 0.60) and ^APRMS_RW (p = 0.034; μ² = 0.03; 1-β = 0.60), with both being larger at T6 than at T0 (^APRMS_FW: T0 = 2.95 ± 0.65; CI: [2.64, 3.25], T6 = 3.36 ± 0.82; CI: [2.80, 3.75]; ^APRMS_RW: T0 = 2.38 ± 0.85; CI [1.99, 2.78] vs. T6 = 2.95 ± 1.15; CI: [2.41, 3.49]), and a significant effect of trial on ^CCRMS_FW (p = 0.024) and on ^CCRMS_RW (p = 0.026), both being lower at T6 compared to T0. Moreover, a significant time by trial interaction effect was observed for ^MIDTURω_CC-PEAK (p = 0.023). After Bonferroni correction, pairwise comparisons were not significant. No significant effect of time, trial or time by trial interaction was observed for the other variables.

Table 3. Temporal and kinematic variables of the i-TUG before (T0) and after (T6) the TKD course. Data are expressed as mean (M) and standard deviation (SD). * Denotes significant effect of time; † denotes significant effect of trial, ‡ denotes significant interaction time by trial.

Variables	T0		T6		p-Value
	T1	T2	T1	T2	Time	Trial	Time × Trial
TUGDUR (s)	9.45 ± 1.10	9.36 ± 1.10	8.77 ± 1.00	8.44 ± 0.96	* 0.002	† 0.016	0.107
SITSTADUR (s)	1.45 ± 0.23	1.48 ± 0.26	1.46 ± 0.32	1.49 ± 0.28	0.930	0.455	0.995
FORWALDUR (s)	1.92 ± 0.36	1.80 ± 0.41	1.55 ± 0.38	1.57 ± 0.50	* 0.002	0.486	0.369
MIDTURDUR (s)	1.82 ± 0.27	1.90 ± 0.31	2.00 ± 0.37	1.85 ± 0.30	0.289	0.396	‡ 0.023
RETWALDUR (s)	1.68 ± 0.37	1.67 ± 0.37	1.41 ± 0.35	1.36 ± 0.35	* 0.001	0.421	0.842
TURSITDUR (s)	1.36 ± 0.24	1.44 ± 0.27	1.34 ± 0.23	1.40 ± 0.23	0.532	† 0.007	0.781
STASITDUR (s)	2.13 ± 0.44	2.12 ± 0.34	1.93 ± 0.39	1.80 ± 0.51	* 0.043	0.376	0.457
MIDTURωCC-PEAK (°/s)	162.2 ± 29.1	154.9 ± 32.2	149.6 ± 25.2	164.1 ± 35.4	0.807	0.427	‡ 0.001
TURSITωCC-PEAK (°/s)	197.5 ± 40.0	197.5 ± 40.1	197.3 ± 34.5	198.9 ± 35.2	0.895	0.867	0.864
APRMSFW (m/s2)	2.97 ± 0.74	2.93 ± 0.63	3.36 ± 0.83	3.37 ± 0.85	* 0.031	0.823	0.683
CCRMSFW (m/s2)	9.65 ± 0.20	9.70 ± 0.22	9.52 ± 0.36	9.57 ± 0.35	0.105	† 0.024	0.883
MLRMSFW (m/s2)	1.40 ± 0.27	1.45 ± 0.28	1.54 ± 0.32	1.58 ± 0.34	0.090	0.096	0.838
APRMSRW (m/s2)	2.44 ± 0.94	2.32 ± 0.82	2.95 ± 1.17	2.94 ± 1.16	* 0.034	0.333	0.416
CCRMSRW (m/s2)	9.89 ± 0.26	9.93 ± 0.29	9.73 ± 0.61	9.80 ± 0.46	0.145	† 0.026	0.364
MLRMSRW (m/s2)	1.54 ± 0.35	1.69 ± 0.50	1.67 ±0.48	1.54 ± 0.38	0.748	0.685	0.176
SITSTAαFLEX (°)	0.40 ± 0.80	0.37 ± 0.84	0.10 ± 0.28	0.24 ± 0.42	0.329	1.00	0.835
STASITαFLEX (°)	6.95 ± 9.82	12.0 ± 15.6	8.97 ± 10.7	10.5 ± 11.0	0.932	0.147	0.380
SITSTAωFLEX (°/s)	99.4 ± 20.5	94.9 ± 21.5	100.4 ± 21.9	102.7 ± 23.2	0.328	0.342	0.149
STASITωFLEX (°/s)	54.5 ± 13.6	53.3 ± 13.3	59.4 ± 13.7	51.7 ± 32.9	0.721	0.268	0.349
SITSTAωEXT (°/s)	−57.8 ± 17.0	−55.3 ± 11.8	−52.9 ± 15.1	−54.5 ± 18.0	0.387	0.743	0.295
STASITωEXT (°/s)	−80.0 ± 19.0	−85.5 ± 16.7	−85.6 ± 20.6	−85.0 ± 20.8	0.530	0.123	0.089

shows the temporal and kinematic variables for each trial before and after the TKD course.

4. Discussion

This single-arm longitudinal study aimed to evaluate the efficacy of a six-month TKD protocol to reduce the risk of falls in a population of old individuals using i-TUG and, secondly, to evaluate differences in kinematics and temporal i-TUG’s parameters after the participation in the TKD course. We showed that after the TKD, anthropometric variables were significantly enhanced. Moreover, after the TKD, temporal and kinematic variables of the i-TUG were modified. In particular, most of the temporal variables showed a reduction in their duration, whereas antero-posterior linear acceleration during the forward and backward walk increased after the TKD course.

In agreement with the study of Park et al. [26], the results of our study confirmed that the practice of TKD positively impacts functional mobility, as reflected by the reduced TUG duration (TUG_DUR) from T0 to T6. Before the adapted TKD course, our sample was slightly slower than the reference values for the TUG test provided in two meta-analyses [34,35]. In these meta-analyses, the confidence intervals for the TUG performed in individuals of the same age range as our study (60–69 years old) was 7.1–9.0 s and 6.62–9.20 s for the studies of Bohannon and colleagues [34] and Long and colleagues [35], respectively. Importantly, after the TKD, the performance of the TUG test by our participants fell within the interval of confidence provided by both studies, thus resulting as a promising approach to increase functional mobility and reduce the risk of fall among the elderly. A reduced TUG duration was observed in an elderly population using other types of training: two meta-analyses showed significant reductions in TUG duration after a resistance training protocol [36] and after Tai Chi practice [37]. Lopez and colleagues [36] showed that resistance training was effective in reducing TUG duration only when combined with other type of training such as balance or gait retraining. Interestingly, TKD has been shown to simultaneously improve balance, lean body mass and strength [25,26]. Similarly, the systematic review of Chen and colleagues [37] reported a significant reduction in TUG duration by approximately one second in older individuals practicing Tai Chi. This reduction in TUG duration is comparable to that observed in our study. This may be accounted by some similarities between the two disciplines. Indeed, both Tai Chi and TKD pose a strong emphasis on balance, both in static and in dynamic conditions. Moreover, as TKD is performed on mats, the unstable surface may further stress the balance system. In light of the above considerations, adapted TKD can be considered a suitable and alternative activity to be proposed to old individuals.

The reduction in TUG duration was associated with a reduction in most of the temporal variables computed using the inertial measurement unit and especially with a significant reduction in the duration of both FORWAL_DUR and RETWAL_DUR. Since in the TUG the distance to be walked is fixed (3 m forward walk; 3 m return walk), a reduction in the duration of the walking phases implies an increase in walking speed, which has been previously observed to be a significant predictor of risk of fall [38] disability and death [33]. The observed reduction in FORWAL_DUR and RETWAL_DUR after the TKD intervention suggests an increased ability of our participant to accelerate the centre of mass of their body. The observed increases in ^APRMS_FW and ^APRMS_RW confirm the increased ability to accelerate the centre of mass. Reasonably, this enhanced ability was ascribed to improved neuromuscular characteristics of lower limb muscles, involved during walk subtasks. In fact, in a longitudinal study, Park et al. [26] showed an increase in the thigh’s cross-sectional area after TKD training. Moreover, in a similar study design, Lee and colleagues [39] showed improvements in maximum leg strength after three months of TKD training. In addition, TKD more than other sport disciplines seems to induce changes in muscle morphology [40]. As it is well established that variations in muscular morphology affects muscle mechanics and function [41], the practice of TKD may positively impact activities of daily living, such as walking.

Similarly to what was observed for the walking subtask of the i-TUG test, STASIT_DUR showed a significant reduction from T0 to T6. According to Weiss and colleagues [29], STASIT_DUR relates to the risk of falls of an individual. As STASIT represents the transition from a standing to seated position, its decrease may be linked to a general improvement in the general motor control of the body accounted for by enhanced lower limb strength and overall balance. Improved neuromuscular control, especially of the rectus femoris and of the biceps femoris, has already been posited by a previous study focusing on muscle activation during stand to sit in older adults as a discriminating factor between old and young adults [42]. Relevant to the present findings, improved neuromuscular control of lower limb muscles following martial art training has been observed in previous studies focusing on muscle coordination during kicking actions in Karate, a martial art sharing some kicking techniques with Taekwondo [43,44,45,46]. Similarly, following twelve weeks of TKD, an increase in maximal leg strength was observed by Lee and colleagues [39]. A study by Cho and Roh [47] further supports our speculation of improved neuromuscular control following the TKD program. Indeed, Cho and Roh [47], showing an improved chair stand test performance in elderly individuals after 16 weeks of TKD training, posited that this enhancement may be connected to specific movements of the discipline involving the lower limb, such as kicking actions. Interestingly, no effect of TKD emerged in the apparently opposite movement: the SITSTA_DUR. Although similar, recent evidence showed that sit to stand and stand to sit exhibit marked differences in terms of the centre of mass displacement and that stand to sit is more sensitive than sit to stand in detecting individuals more at risk of fall [48] due to its higher demand in terms of neuromuscular control. With this in mind, the reduced STASIT_DUR observed after the TKD program seems even more important. It has to be acknowledged that the interpretation of the possible mechanisms of action of the adapted TKD is speculative, and further studies are needed to establish the exact mechanisms accountable for the observed increase in functional mobility.

Some considerations regarding the safety of the adapted TKD and the possibility of adverse effects need to be addressed. According to an epidemiological study [49], the most recurrent TKD injuries are contusions and sprains, located in the lower limbs. Moreover, these injuries had higher frequency in professional athletes than amateur practitioners. Considering that our TKD protocol has been tailored to an elderly population and that it mainly focuses on learning the poomsae with constant supervision by a qualified TKD instructor of the Italian Taekwondo Federation, the possibility of adverse effects in our study, although not eliminated, was reduced to a minimum. Therefore, adapted, low-to-moderate intensity and supervised TKD can be considered a safe yet efficient activity to be practiced in the elderly.

This study is not devoid of limitations. The first limit of our study is the absence of a control group. This limitation does not enable us to ascertain if the observed effect can be attributed solely to the practice of TKD or to any other activity. The main issues related to the absence of a control group rely on the fact that the observed results may be ascribed to a learning effect and/or to the fact that the observed improvements are not related to the intervention itself. Concerning the first issue (learning effect), it must be noted that the TUG test has been shown to have excellent reliability in community-dwelling older adults (ICC: 0.97; [5]). Its excellent reliability results in a limited learning effect, which indeed might have hampered the strength of our results. Concerning the second point, an improvement in functional mobility in the elderly is rather unlikely to occur without any external intervention: we would like to emphasize that, despite the absence of a control group, any improvement, at this age, is extremely difficult to achieve, thus underlying the effectiveness of the proposed activity. However, this study may represent a promising starting point for a detailed evaluation of the effects of adapted TKD on TUG performance, especially focusing on its subtasks. Second, although we showed significant effects of TKD practice on a number of temporal and kinematic variables, studies using a larger sample size are warranted to thoroughly investigate possible predictors of the risk of falls using the i-TUG subtasks. Moreover, a larger samples size would allow us to stratify the effect of the adapted TKD practice according to gender and age. Third, an accurate evaluation of body composition and especially of the appendicular muscle mass may have allowed us to establish whether the improvements in the TUG performance may be associated with an increase in muscle mass and consequently determine possible adaptations in muscle morphology underlying strength gains in the lower limbs.

5. Conclusions

In this study, we showed that a six-month adapted TKD program was associated with an enhancement in the functional mobility in old individuals by means of the i-TUG test. Moreover, the observed reduction in the i-TUG duration and changes occurring in the centre of mass kinematics allowed us to speculate on the underlying mechanisms accountable for the increased functional mobility. Future studies with larger sample size, a comparison group and a randomized design, should evaluate the effects of an adapted TKD program on the risk of falls in older individuals combining centre of mass kinematics with muscle activation strategies to unravel the underlying mechanisms of the adapted TKD program. In addition, the effect of the adapted TKD program should be investigated bearing in mind possible differences between gender and age groups.