Introduction

Kicking is a fundamental movement skill and is highly prevalent and essential in many sports, including soccer, rugby, American Football and Australian Rules Football [15]. Kicking motion in all types of sports is a skill used  to score goals or points, which necessitates the generation of a high ball-releasing velocity [32]. Kicking performance is evaluated by ball release velocity [39] and accuracy [18]. Previous research has focused predominantly on the kinematics of the lower body [29, 35, 38], and the recommended instruction for coaching the upper-body motion is limited [15]. With the increasing number of recent kicking biomechanics studies, it is worth reviewing the necessity of adding an upper body rotation technique.

From a biomechanics perspective, generating high ball release velocities can be explained in terms of the kinetic link principle, where angular momentum is sequentially transferred from proximal to distal body segments. Although kicking performance includes accuracy [18] consistently, the kinetic link principle cannot account for this definition of kicking performance; instead, it explains  a coordinated sequencing of body segments to produce a high ball release velocity. From a sports coaching perspective, descriptions of the kicking skill [15] typically indicate that torque at the hip joint initiates the motion [35], followed by the peak extension of the ipsilateral knee joint [38], rather than from rotational motions of the upper body. These lower limb motions are explained by mathematical equations [2, 38] that indicate the transfer of angular velocities from the thigh to the shank (or lower leg) to generate a high foot velocity at ball impact. Foot velocity is further noted as the most significant predictor of ball release velocity [2, 9, 17]. However, the origin of the angular velocity of the hip action remains unclear. Thus, upper body segment motions should be considered (in addition to other body motions such as the action of supporting leg) to ascertain their contribution to prestrike foot velocity and, thus, ball release velocity in this review.

The kicking motion has already been defined as a ‘whip-like’ motion. The application of the kinetic link principle in biomechanics analysis is evident in studies examining other ‘whip-like’ motions, such as overarm throws [16, 19]. The angular momentum from upper torso rotation is sequentially transferred to the upper chest, upper arm, lower arm and hand to generate a maximum ball release velocity [19]. The possibility and its impact of an angular velocity transferred from the upper to the lower body segments in kicking warrants investigation.

The role of the upper body transferring angular momentum to the hip and knee joints during the kicking motion has been explored in previous studies [18, 32]. However, due to a variety of variables and differing definitions of the body segments involved, the contributions of the upper body segments remain unexplored in the current guideline [15, 21]. Thus, this systematic review aimed to synthesise findings in the extant literature from seven major sports science research databases to explore the role of upper body rotations in optimising kicking skills.

Methods

A predefined systematic review protocol was registered with the PROSPERO International Prospective Register of Systematic Reviews (registration ID: CRD42020176108). The review was conducted and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [31].

Literature Identification

An initial systematic search of existing literature was conducted using the combined keywords ‘All ball kicking sports AND Kick AND Upper body’ (Appendix 1). These keywords were generated in consultation with a specialist librarian to locate literature that explored the role of upper body rotations in kicking performance outcomes. Comprehensive searches were conducted on MEDLINE, CINAHL, SPORTDiscus, Web of Science, PUBMED, SCOPUS and EMBASE from the first record to 30th October 2023. Additional studies that met the inclusion criteria were identified from the reference lists of the included studies from the database search.

Inclusion Criteria

We included original research reporting on the relationship between upper body kinematics (segments above the hip) and ball kicking performance. We also included participants’ characteristics, including the type of sport, sex, age, health status, playing status of participant and languages. Studies reporting on the coordination patterns of the upper body during ball kicking, coaching and/or training applications, as well as adverse effects were included.

Exclusion Criteria

We excluded research reporting outcomes not related to the application of upper body rotation training on athletes such as robotic prototypes, muscle strength, or cross-sectional area, the lower extremity (segments including only the hip or below the hip). We also excluded editorials, conference papers, systematic reviews or unpublished studies without a peer-review process.

The title, abstract and full-text screening of the retrieved papers was independently conducted by three authors (AF, JL and JC) based on the inclusion and exclusion criteria. The final review of all included studies was then conducted by all authors. Any discrepancies were addressed through discussion with all authors.

Methodological Quality Assessment

The quality of each included study was assessed independently by three authors (AF, JL and JC) using an assessment checklist specifically designed for this review (Table 1 and Appendix 2). This checklist was modified based on the Standard Quality Assessment Criteria [26]. Items 1, 2, 4, 5, 6, 7, 9, 10, 12 and 13 were included from the existing list. Subsections were created within items 4, 5 and 6 and items 3, 8 and 11 were added to make this checklist specifically relevant towards biomechanical analysis in ball kicking motions. A score of 0 was given for each item when the corresponding criterion was not met, and a score of 2 when the criterion of the item was met fully. A score of 1 was given for each item when the information was reported partially and required interpretation or prior understanding. A total score of 80% or higher in a study indicated a high-quality study, as the methods used were considered reproducible to answer the intended research question. Any discrepancies in scoring were resolved through discussion by all authors.

Table 1 Methodological quality assessment checklist for included studies
Full size table

Data Extraction

Data extraction was undertaken by AF, JL and JC, with all authors checking for accuracy. Data extracted were the year of publication, the purpose of the study, mode of kicking, participant characteristics (age, sex, skill level), study design and methods, alongside the main finding (Table 2).

Table 2 Abridged summary of records screened in the systematic search. Listed by alphabetically by mode of kicking
Full size table

Data synthesis and Analysis

Due to the heterogeneity of included studies in terms of study methodology and analysis of joints and segments, a meta-analysis would be potentially misleading. Thus a narrative synthesis of the results was presented. The findings of the kicking motion in differing sports were discussed separately due to variances in ball size, mass and shape, as well as differences in kicking technique. Where reported, for each study, results were considered significant if the reported p-value was less than their stated critical p-value.

Results

Of the initial 1486 papers, 27 papers were deemed eligible for inclusion (Fig. 1) [3, 5, 7, 10, 12, 13, 20, 22,23,24,25, 27, 28, 32, 33, 36, 41,42,43,44,45,46,47,48,49, 51, 52] and a total of 457 participants included. The included papers involved the motion of the instep soccer kick (n = 21), the inside-of-the-foot soccer kick (n = 1), the rugby place kick (n = 4) with a stationary ball, and volley kick (n = 1). These papers analysed head (n = 2), upper torso (n = 3), thoracolumbar spine (n = 19), pelvic (n = 19), shoulder (n = 7) and elbow (n = 6) (Table 2). No eligible studies involving other sports including Australian Rules Football, and American Football were located. Participants in the included papers ranged from having no experience to professional athletes, allowing for a demonstration of a wide scope that distinguishes skill differences. None of the included studies studied the use of upper body rotation in coaching applications. A detailed summary of the results from each study is presented in Table 2.

Fig. 1
figure 1

Flowchart of the systematic literature search (n = 27)

Full size image

Methodological Quality Assessment

The quality assessment of the papers is reported in Table 1. All studies were included in the review regardless of methodological quality, acknowledging their research contributions. Seven studies, with high-quality scores, directly referred to the kinetic link principle using appropriate biomechanical definitions [3, 5, 7, 9, 10, 13, 32, 41]. All included studies except three used three-dimensional data capture and analysis techniques. In the four studies [13, 27, 36, 42], two digital cameras with orthogonal views were used to investigate the planar motions of participants from the front and side views.

Relationship Between Upper Body and Lower Body Rotations

The transfer of angular momentum from proximal to distal body segments is presented based on the kinetic link principle. Nine studies of the instep soccer kick [10, 13, 23, 25, 24, 32, 33, 41, 48] indicated that thoracolumbar rotation due to muscle moments about the longitudinal [13, 25, 28, 32, 33, 41] or transverse [10, 13, 23, 24, 28, 33, 48] axes resulted in a transfer of angular motion to the thigh and shank segments, with their respective segmental velocity magnitudes positively correlated. Langhout et al. [24] reported that during the leg-cocking phase, increasing the maximal range of thoracolumbar extension combined with non-kick side shoulder overhead extension motion produces a powerful tension arc which assists in the acceleration of hip flexion by the transference of angular momentum. This finding is supported by Langhout et al. [25], Smith et al. [48] and Carvalho et al. [13] that as the magnitude of the thoracolumbar rotation on the transverse axis increased, an increase in the range of motion of leg back swing accelerates hip flexion, which produces a maximum “whip-like effect”. This, in turn, resulted in an increased acceleration in knee extension [24, 32] in the leg-acceleration phase. Further, Augustus and colleagues [5, 7] concluded that the pelvis plays a pivotal role in generating transverse momentum to the kicking leg, and a strong relationship was demonstrated between change in pelvis transverse angular velocity and thigh-knee angular velocity upon ball contact. In contrast, the study conducted by Orloff et al. [36] revealed no significant impact of thoracolumbar angle on hip and knee joint angles. When the soccer ball was elevated for a volley kick, notable changes were observed for both the pelvis and the leg segments: there was an increase in pelvis clockwise rotation, a decrease in left hip linear velocity, a reduction in hip/knee flexion, and an increase in hip internal rotation [49].

For rugby place kicks, Bezodis et al. [12] reported that an increase in non-kick-side arm angular momentum about the transverse and longitudinal axes increased maximal kicking leg angular velocity later in the kicking cycle in support of the kinetic link principle. Green et al. [22] investigated rotations about the longitudinal axis, reporting that with an increase in the magnitude between the thorax and pelvic rotations (when viewed from above the head), thorax rotation and head rotation, the ball kicking distance increased. Although this finding implies that the kinetic link principle applied to the transference of angular momentum to the lower body, this was not directly examined. Atack et al. [3] found that kickers boasting a maximum placement kick distance beyond the international average (i.e., > 32 m) exhibited lower accuracy, attributed to a tension arc across their torso. This phenomenon resulted in heightened positive hip flexor joint exertion during the downward swing compared to their more accurate counterparts. Additionally, Zhang et al. [52] reported in the rugby place kick that lateral pelvic tilt towards the standing leg (1.2 m/s) and posterior rotation (0.3 m/s) played a significant role in the speed of foot movement during hip extension to hip flexion (1.6 m/s) and during knee flexion to knee extension (9.1 m/s). This observation highlights the transference of angular momentum from proximal to distal body segments.

Relationship Between Upper Body Rotations and Ball Release Velocity

Thoracolumbar rotations about the longitudinal axis were reported to have had a direct correlation with ball release velocity from an instep soccer kick following the kinetic link principle [20, 27, 41,42,43,44,45,46,47]. Fullenkamp et al. [20] found that individuals with a background in organised soccer experience, on average, have 53% more thoracolumbar ROM and 62% higher peak thoracolumbar rotation velocity than those with no soccer experience, which has a moderate, positive correlation with the ball velocity (r = 0.57, P < 0.01). On the other hand, although Lees et al. [28] reported that thoracolumbar rotations on the longitudinal and transverse axes followed classical proximal-to-distal temporal sequencing by transferring its angular velocity to the hip, knee and foot joints sequentially, there was no direct correlation with ball release velocity. Thus, there was no clear evidence that shoulder rotations or hip rotations on the longitudinal axis were related to ball release velocity or accuracy. Additionally, Scurr and Hall [42] reported that despite the magnitude of pelvic rotations about the longitudinal axis being greater in the 45° and the 60° approach angle conditions than in the self-selected approach angle (P < 0.05), there was no significant difference in ball release velocity (P = 0.59) between approach angles. When examining the transverse axis, thoracolumbar flexion and extension accounted for variances in ball release velocity [27, 41, 43,44,45]. A more recent study by Atack et al. [4] found that the width of the approach and a more forward body position can result in higher foot and ball velocities.

Participants performing the instep soccer kick with faster ball velocity exhibited an increase in the magnitude of thoracolumbar rotations about the longitudinal and transverse axes [20, 27, 43,44,45,46], leading to a greater ball release velocity when compared to novice participants. The selection of the preferred leg also plays a significant role. According to Shan and colleagues [45], greater ball release velocity was associated with the dominant foot, primarily due to an amplified magnitude of thoracolumbar rotations around both the longitudinal and transverse axes.

Discussion

This systematic review synthesised evidence on the role of upper body rotations in relation to the lower body and optimising kicking skills. Our findings revealed a clear relationship between upper body rotations and the lower body with kicking motion following the kinetic link principle. Most of the evidence was provided by studies of the instep soccer kick [5, 7, 10, 13, 20, 23,24,25, 28, 27, 32, 33, 36, 41,42,43,44,45,46,47,48,49], with additional supporting insights provided by studies of the rugby place kick [3, 12, 22, 52]. The thoracolumbar rotations along both the longitudinal and transverse axes amplify the magnitude of hip extension and enhance ball-release velocity.

Extant coaching literature indicates that the kicking motion is initiated at the hip joint [15], with angular momentum sequentially transferred to the thigh, shank and foot [38]. This review provides evidence that upper body motions should also be considered in the transfer of motion from proximal to distal body segments. Established definitions that indicate that hip extension initiates the kicking motion should now be expanded to indicate that the magnitude of hip extension is a consequence of the transfer of motion from thoracolumbar rotations about both the longitudinal [3, 12, 20, 22, 25, 27, 32, 33, 41,42,43,44,45,46,47, 52] and transverse axes [12, 13, 24, 27, 33, 43,44,45, 48]. This finding expands the application of the kinetic link principle applied to the kicking motion and thus has implications for coaching to improve kicking performance. However, an examination of segmental sequencing by evaluating the time of the maximum angular velocity or energy-flow analysis of all involved joints and body segments is required to fully establish a whole-body model of instep kicking performance [6, 13].

Lees and colleagues [29]underscored the significance of integrating temporal sequencing with fundamental biomechanical principles, encompassing factors like range of motion, the stretch–shortening cycle, end-point velocity, and the dynamics of action and reaction. A comprehensive understanding of these concepts not only allows for a mechanical insight into skill execution but also streamlines the process of conducting qualitative performance assessments. This concept can be universally applied across various kicking modes. Further, given the noteworthy evidence in our study, which links greater upper body rotations to increased ball release velocity via enhanced segmental sequencing across the body, we advocate for the inclusion of a whole-body analysis (encompassing both upper and lower body movements) to conduct a more thorough examination of a soccer kick. However, the upper body and pelvis rotation was not listed as the key skill criteria in the coaching manual for the development of kicking skills in children and adolescents [15]. Despite being updated in 2017 [21], it does not include any description related to the upper trunk and the pelvis, nor the sequencing of the rotations. Consequently, the valuable insights from research have not been effectively applied in youth coaching, resulting in limited improvement in practical training. This underscores the need for a holistic approach to skill analysis and for effectively translating experimental research into real-world coaching practices.

Performance of the instep soccer kick with the dominant foot produced highly coordinated, skilful movements as a result of long-term motor learning and practice in skilled participants [45, 51]. This in turn produced a greater ball release velocity than kicking with the non-dominant foot. This was attributed to a greater magnitude of upper body rotations during kicking with the dominant foot, further supporting the importance of upper body rotations in skilled performance. Although the better performance of the instep soccer kick with the dominant foot was attributed to a greater magnitude of upper body rotations than when kicking with the non-dominant foot, further research is required to understand the variations in coordination patterns and segmental sequencing during the performance with the non-dominant foot. Thus, the development of instructions that help improve the general kinematics during motor learning of the non-dominant foot would enhance coaching practices.

The association between upper body rotations and ball release velocity was best described, where thoracolumbar rotations increased resultant foot velocity to increase ball release velocity, including a discussion of the interactions of the hip, thigh and shank segments [44, 47]. The findings of these studies indicate that established equations representing the transfer of momentum [2, 38] can be expanded and re-established by incorporating the upper body. Atack et al. [2] analysed the sequencing and magnitude of a multi-segment model involving thorax, pelvis, hip, knee and foot rotations, which in turn increased ball release velocity in the rugby place kick. Results were discussed alongside a graphical representation to demonstrate the concurrence of the relative peak timing of each body segment. Thus, this approach must be applied to other modes of kicking to create a true understanding of kicking performance. Also, to educate coaches about the role of upper body rotations in increasing ball release velocity, future studies must identify the applications of the findings to coaching practice.

The examination of the kicking motion using a stationary ball was prevalent across all reviewed papers. However, this applies only in certain instances of sports involving the kick motion, such as free kicks, corner kicks or goal kicks in soccer and rugby place kicks. Enhancing the practicality and real-world applicability of kick motion analysis necessitates the inclusion of more dynamic scenarios, such as varying approaches to the ball and the choice of kicking foot. These factors have been shown to exert a significant influence on the overall kinematic patterns of the body [10]. Therefore, it is highly recommended that further research be conducted focusing on analysing the role of the upper limb in kicking motions while using a moving ball. A recent study [11] suggested that the difference between using a moving ball and a stationary ball may significantly impact whole-body kinematics and ball velocity during an instep kick in soccer. The study proposed that the ball approach angle influences various limb variables and upper body variables while kicking with the dominant foot results in higher linear and angular velocities of the swinging limb and a higher vertical position of the centre of mass. Furthermore, to improve our understanding of the relationship between upper body rotations and kicking accuracy, further research is necessary. Two studies [3, 22] investigated whether increased upper body rotations contribute to kicking accuracy. However, the sample size might be inadequate. It is recommended to provide formal priori sample size estimations for sports research. Otherwise it is still being determined what changes could be considered as a meaningful effect [1]. Hence, it is imperative for future research endeavours to address the sample size estimation to allow robust outcomes in soccer research.

Leg muscle mass is positively correlated with the ability to generate joint torques [40]. This, in turn, will influence prestrike foot velocity and hence ball release velocity, as prestrike foot velocity is the most significant predictor of ball release velocity [2, 8, 17]. Thus, when assessing movement patterns with skill level, it is essential to normalise the effects of muscle mass to counter the confounding effects of muscle mass. Limb length or height is another contributing factor. For example, Atack et al. [3] normalised the joint kinetics using height. It is suggested that ball release velocity may be normalised by dividing each participant’s stature to minimise the effect of longer limb length and associated mechanical advantage. However, none of the included studies checked the statistical assumption for using the scaling variables in normalisation [50]. The readers should interpret the results with caution. Further, anatomical and biological differences between sexes necessitate the need for separate analyses. By analysing male and female participants independently, comparisons between sex and skill levels can be made within and between studies [36, 46, 47]. It is recommended that future research include separate analyses of male and female participants, as suggested by Nimphius [34].

No studies investigated the feasibility of adding upper body rotation in kicking training for less-skilled athletes, particularly children or adolescent athletes. This is a gap in the literature as most athletes start developing skills acquisition during their younger years. Performance indicators and injury predictors also surface in the early stage of a career. The correlations between kinematic changes in the upper body and common soccer injuries, such as groin and hamstring injuries, need to be better understood. The current guidelines for improving kicking performance primarily focus on lower-body coordination training [15]. This raises the question of whether the movement pattern or the energy transfer from the upper body could complement these guidelines, reducing the risk of injuries or contributing to the occurrence of these injuries. Previous studies demonstrated that when the whip-effect is greatly increased, the chance of injury may also be increased [14, 37]. DeLang et al. indicated that he probability of injury of the inertial leg is 1.6 times that of the non-inertial leg in soccer. At the same time, while among youth athletes, the rate has increased by another 1.5 times [14]. Hamstring strain injuries have already been proven to be highly prevalent in sports [30]; risk of such injury may be increased by an excessive strain during eccentric contraction in the late kicking phase. Further studies could investigate the impact of energy flow between the trunk and the extremities and their contribution to injuries.

Currently, there is no standardised protocol to quantify the strength of evidence in biomechanics studies with different study designs. However, given that 10 of the 12 included studies demonstrated a positive relationship between upper body rotation and ball release speed. This review appears to provide at least moderate evidence that increased upper body rotations promote ball release velocity following the kinetic link principle. The impact of the upper body rotation is not highlighted in current coaching manuals, which should be updated to improve the kicking skill acquisition.

Perspective

This review provides moderate evidence that increased upper body rotations play a crucial role in promoting ball release velocity, aligning with the kinetic link principle. Specifically, developing thoracic and pelvic rotations has been found to have a positive impact on ball release velocity during the in-step kicking motion. However, an important gap emerged in the current coaching guidelines, as they do not emphasise the critical significance of upper body rotation, underscoring the urgent need for updates to enhance kicking skill acquisition. By gaining a deeper understanding of how the upper trunk and upper limb segments interact with the pelvis and lower limb segments, coaches can effectively prescribe techniques, and athletes can acquire the necessary skills for improved performance. To comprehensively optimise ball-kicking motion, thereby updating the coaching manual, further research is warranted to explore the effects of timing and the ranges of motion of all relevant upper and lower body segments on maximum ball release velocity.