FormalPara Key Points

A systematic review of 61 studies indicated several methodological limitations (i.e. an inadequate assessment of motor competence, a lack of longitudinal observations and a failure to account for biological maturation) within the current literature that evaluates associations between motor competence, physical activity, physical fitness and psychosocial characteristics amongst adolescents.

Across several meta-analyses of 43 studies, motor competence was positively associated with physical activity, muscular endurance, muscular power, muscular strength, cardiovascular fitness, perceived motor competence and motivation, and inversely associated with weight status, speed and agility in adolescents.

Teachers, sports coaches, strength and conditioning coaches, and other stakeholders involved in health and performance interventions during adolescence should seek to synergistically develop motor competence, physical fitness and psychosocial characteristics for positive physical activity and health outcomes.

1 Introduction

The synergistic development of physical, psychosocial and motor skill domains throughout childhood and adolescence, across various environments, is important for the health and performance of all youth [1]. Such holistic development of “athleticism” (i.e. the composition of health-related fitness and psychosocial traits [1]) is crucial given the worldwide decline in youth health and fitness and therefore athleticism over past decades [2,3,4], confounded by reduced sports participation rates (e.g. [5, 6]), and fewer youth meeting the World Health Organisation’s ([7]) physical activity guidelines [8]. In turn, these trends may contribute to the increasing obesity pandemic amongst youth (e.g. UK [9], USA [10]).

Authors have postulated that motor competence underpins daily tasks, and engagement in health-enhancing activities (e.g. running, resistance training, recreational games, sport) across the lifespan [11]. Motor competence refers to an individual’s ability to perform a variety of motor skills, where outcomes are influenced by movement quality, control and coordination [12,13,14]. Furthermore, motor competence consists of simple, combined and complex movement capacities, which are inter-related. Motor competencies are often categorised into locomotor (e.g. running), object control (e.g. striking) and stability (e.g. balance) skills [15,16,17]; however, other domains (e.g. foundational movement skills, athletic motor skill competencies) have also been proposed [13, 18]. Research highlights that motor competence is crucial for physical and psychosocial development [19], as it enhances children’s and adolescents’ ability to meaningfully participate in games, sports and other physical activities [20]. Therefore, developing motor competence amongst youth should be a key focus of any physical activity, physical education or youth sport intervention, as it appears central to reversing the currently negative physical activity and obesity trends worldwide.

Previously, Stodden et al. [21] hypothesised that motor competence interacts with perceived motor competence (an individual’s identification and interpretation of their actual motor competence [14, 22]) and physical fitness during childhood to induce positive (e.g. increased physical activity engagement, healthy weight status) or negative (e.g. reduced physical activity engagement, unhealthy weight status) trajectories (Fig. 1). Accordingly, those expressing poor actual and perceived motor competence during childhood may present with reduced actual/perceived motor competence, physical fitness and physical activity engagement across the lifespan [23, 24]. Numerous studies have evaluated Stodden and colleagues [21] model, identifying positive associations between motor competence and physical activity engagement [25,26,27], musculoskeletal strength/endurance [12], cardiorespiratory fitness [12, 25] and inverse associations with weight status [12, 25]. Similarly, previous reviews (e.g. [14, 28, 29]) have shown that evidence levels differ for associations between different motor skills domains (e.g. locomotor, object control, stability/balance) and physical activity, physical fitness and/or psychosocial characteristics. However, most of the existing evidence involves children (e.g. [30,31,32,33]), or children and adolescents together (e.g. [34,35,36]).

Fig. 1
figure 1

Copyright © [2023] National Association for Kinesiology in Higher Education (NAKHE), reprinted by permission of Taylor & Francis Ltd, http://www.tandfonline.com on behalf of © [2023] National Association for Kinesiology in Higher Education (NAKHE)

Development model as proposed by Stodden et al. [21]. EC early childhood, LC late childhood, MC middle childhood.

Full size image

Childhood and adolescence are stages of youth development that require a divergent physical and psychosocial focus [37]. Adolescence represents a dynamic period of physical, psychosocial and highly individual development whereby the timing (i.e. the onset of change), magnitude (i.e. level of change) and tempo (i.e. rate) of biological maturation are asynchronous with chronological age [38, 39]. During biological maturation, growth rate increases rapidly, with peak height velocity (PHV; [38, 39]) typically occurring around 12 years for female individuals and 14 years for male individuals [40, 41]. This growth spurt can lead to temporary reductions in motor competence (i.e. adolescent awkwardness [42]). Furthermore, during adolescence, brain maturation is significant and ongoing. Psychosocial changes include an increased ability to process information [43], and improved executive function of the pre-frontal cortex [44], which underpins many self-regulatory mechanisms (e.g. behavioural/emotional/attentional regulation [45]). Thus, along with physiological changes, adolescents are developing their ability to self-evaluate and problem solve their own physical development. Factors such as age and maturity have been posed to contribute to the globally high percentage of adolescents who do not reach the World Health Organisation’s recommended physical activity guidelines, if appropriate interventions are not implemented [46]. Consequently, investigating motor competence within adolescent populations is an important consideration to enhance the health and athletic development of youths.

To the authors’ knowledge, no systematic review has examined the associations between motor competence and physical activity, physical fitness and psychosocial characteristics within adolescents alone. Further, because of the potential ramifications of the dynamic nature of growth and maturation, adolescence is a key period of the lifespan to focus upon such characteristics. While other reviews have investigated child and adolescent populations simultaneously (e.g. [12, 14, 47]), reporting findings simultaneously in studies may result in misinterpretation owing to a failure to distinguish between children and adolescent findings, leading to an unclear picture of adolescent research (e.g. [12]). Therefore, solely focusing on relevant research in adolescents is warranted to comprehensively review the types of research conducted, methods employed, measures used and the confounding effects these factors may have within this population. Additionally, it remains unclear which characteristics to target across adolescence to optimise health and performance outcomes [1]. Consequently, a systematic review and meta-analysis are required to highlight associations between motor competence and physical activity, physical fitness and psychosocial characteristics in adolescence. Such research will highlight potential focus points (e.g. population, characteristics of interest, methods of assessment) for future implementation and assessment of interventions, which is critical for understanding and potentially reversing the current negative physical activity and fitness trends among adolescents. Therefore, this study aimed to (1) analyse the scientific literature evaluating associations between motor competence and physical activity, physical fitness and/or psychosocial characteristics amongst adolescents; (2) evaluate the associations between motor competence and physical activity, physical fitness characteristics and/or psychosocial characteristics amongst adolescents; and (3) investigate the impact of moderator variables (i.e. age, sex, type of motor competence assessment) on the associations.

2 Methods

2.1 Study Design and Search Strategy

A systematic review and meta-analysis were conducted in accordance with the updated Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) statement [48]. Before commencing the review, the protocol was registered on the International Prospective Register of Systematic Reviews (PROSPERO) database (ref: CRD42021233441). A systematic search of eight databases (Academic Search Complete, CINAHL Complete, MEDLINE, SPORTDiscus and PsycINFO via EBSCOhost, PubMed, SCOPUS and SAGE Journals Online) was conducted to identify original research articles from the earliest record available up to and including 05/08/2022. Boolean search phrases were used to combine search terms relevant to adolescents (population), motor competence, physical activity and/or physical fitness, and/or psychosocial characteristics. Relevant keywords were identified for each search term through pilot searching (screening titles/abstracts, keywords, full texts and similar reviews previously published, e.g. [12, 14, 47].). Keywords were combined for each term using the “OR” operator, and the final search phrase was constructed using the “AND” and “NOT” operators as follows: (“Youth*” OR “Adolescen*” OR “Teen*” OR “Student*” OR “High school” OR “Secondary school” OR “Pube*”) AND (“Motor competenc*” OR “Movement competenc*” OR “Physical competenc*” OR “Motor development” OR “Motor skill*” OR “Motor abilit*” OR “Movement skill*” OR “Motor coordination” OR “Actual competenc*” OR “Object control” OR “Manipulative skill*” OR “Locomotor skill*” OR “Stability skill*” OR “Athletic competenc*” OR “Athletic skill*” OR “Motor proficiency” OR “Fundamental movement skill”) AND (“Physical activit*” OR “Activit*” OR “Sports” OR “Sports participation” OR “Body weight status” OR “Body composition” OR “Body fat” OR “BMI” OR “Physical fitness” OR “Fitness” OR “Cardiorespiratory fitness” OR “Cardiovascular endurance” OR “Muscular strength” OR “Muscular power” OR “Flexibility” OR “Mobility” OR “Endurance” OR “Muscular endurance” OR “psychological” OR “psycho-social” OR “Motivation” OR “Perceived motor competenc*” OR “Physical self-perceptions” OR “Self-confidence” OR “Self-efficacy” OR “Self-Competenc*” OR “physical self-concept”) AND (“correlate*” OR “determinant*” OR “predictor*” OR “relationship*” OR “association*”) NOT (“Adult*” OR “Child*” OR “Prepube*” OR “primary school” OR “Kid” OR “Kids” OR “Preschool” OR “Kindergart*” OR “preadolescen*” OR “Disease*” OR “Disab*” OR “Impair*” OR “Disorder*” OR “ill*”). Bibliographic screening and citation searching are powerful complementary tools to database searching alone [49, 50]. Therefore, bibliographic screening and forward citation searching (via Google Scholar) of previous reviews and included studies were conducted to identify articles that may have been missed by the search criteria.

2.2 Study Selection

Duplicate records were identified and removed before screening the remaining studies against the following pre-defined exclusion criteria: (1) studies not published in English; (2) previous reviews, conference abstracts, book (chapters), dissertations; (3) studies where the sample consists of only children (< 11 years old) or adults (> 18 years old) OR studies that included a combined sample of children/adults with adolescents; (4) participants with a physical or cognitive impairment; (5) studies that did not assess motor competence using a process (i.e. technique; e.g. Test of Gross Motor Development), product (i.e. outcome; e.g. Movement Assessment Battery for Children) or combined method (i.e. process and product; e.g. supine to stand test); (6) studies that did not report the association between motor competence and at least one measure of physical activity (e.g. pedometer, self-report questionnaire), physical fitness (e.g. assessments of body weight status, cardiorespiratory fitness, musculoskeletal strength) or psychosocial characteristics (e.g. perceived motor competence, motivation); and (7) full text not available. The screening process was conducted independently by two researchers (AB and FT) over two phases. Studies were initially excluded based on their title and abstract content, followed by a full-text review. There were no formal disagreements between reviewers regarding study selection; however, reviewers met virtually to discuss and clarify studies where there was more than one reason for exclusion. As there were no formal disagreements between reviewers, a third reviewer was not required.

2.3 Data Extraction

The lead author (AB) extracted the data using a specifically designed and standardised Microsoft Excel spreadsheet. Publication details (e.g. author, year), study type (e.g. cross-sectional, longitudinal, intervention), participant characteristics (i.e. sample size, age, sex, anthropometrics), motor competence assessment details and scores (i.e. measure used, type of measure), physical activity measure details and scores (i.e. measure used, type of measure), physical fitness measure details and scores (i.e. area of physical fitness assessed, measure used), psychosocial measure details and scores (i.e. measure used, psychosocial domain assessed), and the strength and orientation of associations between motor competence and physical activity, physical fitness and psychosocial characteristics were extracted. If any relevant data were missing, the paper’s corresponding author was contacted to provide the required information. Similarly, if the authors had performed a regression analysis on study variables, the authors were contacted to provide a correlation coefficient between the variables in question. Unlike similar reviews (e.g. [14].), reported regression coefficients were not converted to correlation coefficients using the Peterson and Brown [51] equation, as potentially large biases are associated with estimating mean population correlations in meta-analytic conditions [52]. Authors were contacted once in the first instance (followed by one further occasion if there was no response to the original query) for any missing details needed for the meta-analysis. Studies were excluded from the meta-analysis, but still utilised in the qualitative synthesis of the review, if authors did not respond or could not provide the requested information.

2.4 Risk of Bias Assessment

Consistent with previous research (e.g. [12, 14, 25, 47]), the criteria for assessing bias within included studies were adapted from the Strengthening the Reporting of Observation studies in Epidemiology (STROBE) [53] and Consolidated Standards of Reporting Trials (CONSORT) [54] statements. For this review, six criteria were determined to assess the risk of bias within included studies (Table 1). For each criterion, studies were scored with a tick (“✔”, low risk of bias), cross (“✖”, high risk of bias) or question mark (“?”, inadequate or unclear description). To create clear criteria and ensure high agreement between reviewers, the first (AB) and second reviewer (FT) individually screened the same five papers and subsequently discussed the scoring criteria via an online meeting. After refining the criteria, the first and second reviewer independently screened all the included studies and reconvened via an online meeting to compare final scores.

Table 1 Summary of risk of bias assessment criteria
Full size table

2.5 Data Analysis and Meta-analysis

This review’s qualitative synthesis and interpretation used descriptive data extracted from the articles. Where studies used a reverse scale measure (e.g. [55]), or where time (e.g. [56]) represented an outcome measure of motor competence, the effect size direction was reversed prior to analysis so that the association between variables represented the same orientation as other studies. This step accounted for studies where lower scores represented a greater outcome (e.g. faster time = greater motor competence). Within the meta-analysis, correlations of individual sexes were used where available. Additionally, associations of separate motor competence domains (i.e. overall, locomotor, object control, stability/balance, sports-specific competence) were analysed independently to avoid double counting. The fundamental movement skills concept was selected to define sub-group categories for this meta-analysis during a video call between co-authors (AB, IC, JCE, KT). This concept was clearly used by most studies to determine separate categories for correlations, thus allowing the maximum possible studies to be evaluated in the meta-analyses. Furthermore, the fundamental movement skill domains are widely acknowledged in the practical setting for prescribing and assessing motor skills (e.g. [17]). Studies were included more than once in the same meta-analysis where authors had correlated more than one measure of motor competence to the same variable (e.g. [57]), or had used the same measures on separate samples at different timepoints (e.g. [58]).

Random-effects meta-analyses were conducted using Comprehensive Meta-Analysis software (version 3.0; Biostat, Englewood, NJ, USA) to determine the magnitude, orientation and significance of the association between motor competence and physical activity, motor competence and physical fitness characteristics (e.g. strength, cardiovascular endurance), and motor competence and psychosocial characteristics (e.g. perceived motor competence, motivation). Several meta-analyses were conducted based on the relevant primary studies to explore the effect of hypothesised moderator variables (i.e. sex, age and type of motor competence measure [process, product or combined]) on the variation among study outcomes [59, 60].

The inputted data from each study included the sample size and the corresponding outcome measure (i.e. correlation coefficient). Each correlation coefficient (r) was converted to a Fisher’s z-score and standard error to obtain approximately normally distributed values. The Fisher’s z-score was then back transformed to a correlation coefficient and 95% confidence interval (CI) for interpreting the included studies’ summary statistic (i.e. pooled correlation coefficient). Pooled correlation coefficients were estimated for each comparison and moderator variable where possible. Pooled correlation coefficients were interpreted as: 0.00–0.10 (trivial), 0.10–0.30 (small), 0.30–0.50 (moderate), 0.50–0.70 (high), 0.70–0.90 (very high) and > 0.90 (nearly perfect) [61,62,63]. Statistical significance was interpreted for p < 0.05. Cochrane’s Q statistic and I2 statistic were used to determine heterogeneity, with I2 values of > 50%, and > 75% used to indicate moderate heterogeneity and high heterogeneity, respectively [64, 65]. The I2 statistic was supported by reporting the tau-squared statistic. A sensitivity analysis (one study removed function) was used for each comparison, which omitted study samples in turn to examine their influence on the magnitude, orientation or significance of pooled correlation coefficients.

2.6 Evaluation of Small Study Effects

Funnel plots were visually interpreted, along with Egger’s linear regression intercepts for each comparison, to evaluate potential small study effects and publication bias. An Egger statistic p-value < 0.05 indicated the presence of a small study effect.

3 Results

3.1 Overview of Studies

Following the removal of duplicates, a total of 4739 records were identified via the databases searched. Forty-nine additional records were identified from bibliographical screening and forward citation searching. From the title, abstract and full-text screening, 61 records were identified for the systematic review [36, 55,56,57,58, 66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121]. Of the studies identified for the systematic review, 14 [69, 71, 75, 77, 80, 83, 84, 87, 102,103,104, 115,116,117] were excluded from the meta-analysis because of missing data (e.g. unreported correlations, lack of sample size information for a reported correlation) required for conducting the meta-analyses (Fig. 2). A further four studies [78, 94, 109, 112] were also ineligible, as they had provided correlation coefficients for individual elements of a motor competence measure (e.g. overhead squat, frisbee competence), which did not correspond to the motor competence domains utilised for the meta-analysis (e.g. locomotor competence, sports-specific competence). Authors of the studies included in the review that were ineligible for the meta-analysis were contacted for the information required to be included in the meta-analysis. These authors either did not respond to our enquiries or could not be reached via their author contact details.

Fig. 2
figure 2

Flow diagram of the study selection process

Full size image

Extracted data from the included studies are presented in Table 2. Forty-five studies consisted of cross-sectional evaluations, ten studies [80, 103,104,105, 111,112,113,114,115,116] collected longitudinal evaluations, three studies [102, 117, 119] conducted a randomised controlled trial intervention, and three studies [66, 68, 110] involved validity and reliability methods. The included studies represented a total sample of 22,256 adolescents (mean = 371 ± 614 participants; range = 22–3638). Studies were conducted across 16 countries including Australia (n = 10 [66, 67, 71, 73, 74, 78, 83, 89, 110, 119]), Brazil (n = 7 [56, 79, 95, 96, 103, 106, 107]), Czech Republic (n = 1 [97]), England (n = 1 [82]), Finland (n = 9 [55, 58, 85, 87, 99, 111,112,113,114]), Germany (n = 1 [115]), Iceland (n = 2 [57, 108]), Ireland (n = 6 [70, 75, 76, 91, 102, 117]), New Zealand (n = 1 [72]), Norway (n = 1 [88]), Portugal (n = 1 [116]), Spain (n = 1 [98]), Switzerland (n = 1 [109]), the UK (n = 1 [118]), the USA (n = 4 [86, 93, 94, 121]) and Wales (n = 1 [100]). The remaining studies (n = 13 [36, 68, 69, 77, 80, 81, 84, 90, 92, 101, 104, 105, 120]) provided insufficient detail to determine where the data were collected.

Table 2 Overview of included studies
Full size table

Forty-nine studies [36, 55, 57, 58, 66,67,68, 70,71,72,73, 75,76,77, 79, 81, 83,84,85, 87,88,89,90,91, 93,94,95,96, 98,99,100, 102,103,104,105,106,107,108,109,110,111,112,113,114, 116, 117, 119,120,121] recruited their participants from schools (e.g. high school, middle school students/athletes), nine studies [69, 78, 80, 82, 86, 92, 97, 101, 118] consisted of sports-based samples (e.g. amateur male basketball, academy male youth soccer), and three studies [56, 74, 115] described their participants as “adolescents”. Forty-seven studies included both male and female participants, eight studies consisted of male individuals only [72, 78,79,80, 82, 97, 118, 119], while four studies consisted of female individuals only [89, 101, 106, 120]. Two studies [69, 92] failed to report the sex characteristics of their samples. Forty-four studies reported the mean age of their participants (overall mean age = 13.59 ± 1.4 years; range = 11.26–16.40 years), while seven studies [55, 68, 78, 87, 88, 106, 112] reported the age range, and ten studies [56, 58, 71, 73, 96, 99, 100, 105, 117, 118] reported the mean age by various sub-groups (e.g. normal weight, overweight/obese groups).

Eight studies measured the maturity status of their participants. Seven studies [56, 72, 80, 97, 100, 118, 120] used the Mirwald equation [122], whilst one study [82] used the Khamis and Roche method [123]. Authors used maturity status to: (1) compare their participants’ maturity status based on motor competence and/or physical fitness scores [80, 82, 120]; (2) identify associations between maturity status and motor competence, physical and/or psychosocial characteristics (e.g. the correlation between maturity status and motor competence) [72, 97, 100]; (3) highlight that different subgroups were of the same age and maturity status [56]; or (4) identify the influence of motor competence and maturity status on physical fitness outcomes [118]. When assessing correlations between motor competence and physical activity, physical fitness and psychosocial characteristics, 34 studies [56, 57, 66, 68,69,70,71,72,73, 76, 78, 80,81,82, 85,86,87, 89, 91, 92, 96,97,98, 101, 105,106,107,108,109, 111, 112, 118, 120] assessed motor competence against one characteristic, 24 studies [36, 55, 58, 67, 74, 75, 77, 79, 83, 84, 90, 93,94,95, 99, 100, 102, 103, 110, 113, 114, 116, 117, 119, 121] against two characteristics, and three studies [88, 104, 115] against all characteristics.

Of the 61 studies within this review, 25 studies [66, 67, 70,71,72,73, 75, 76, 78, 82, 83, 89, 91, 93, 94, 100,101,102, 104, 110, 117,118,119,120,121] used a process-orientated motor competence assessment, and 31 studies [36, 55, 57, 58, 68, 69, 79,80,81, 84, 85, 87, 88, 90, 95,96,97,98,99, 103, 105,106,107,108,109, 111,112,113,114,115,116] used a product-orientated assessment. Only one study [74] used a combined process and product assessment of motor competence, while four studies [56, 77, 86, 92] used a combined motor competence assessment but reported process and product scores separately. Across the included studies, the following 27 motor competence measures were used: the Körperkoordinationstest Für Kinder ([124], n = 11 [36, 68, 79, 95, 96, 98, 103, 105,106,107, 116]); a combination of individual measures (e.g. Figure 8 dribble test, the leaping test; n = 10 [55, 58, 85, 87, 99, 111,112,113,114,115]); the resistance training skills battery ([66], n = 4 [66, 72, 83, 119]); the Victorian FMS manual [125] (n = 3 [71, 73, 89]); a combination of measures from the test of gross motor development (TGMD [126]), TGMD-2 [127] and the Victorian FMS manual [125] (n = 4 [70, 75, 102, 117]); an adapted version of the Körperkoordinationstest Für Kinder [124] (n = 2 [80, 88]); the Bruininks-Oseretsky Test of Motor Proficiency-2 Short Form (BOT-2 Short; [128]; n = 2 [84, 97]); a combination of the functional movement screen™ [129, 130] and the Y-balance tests [131] (n = 2 [86, 92]); the PE Metrics Battery [132] (n = 2 [93, 94]); the Functional Movement screen™ (n = 3 [82, 101, 118]); an adapted version of the Get Skilled Get Active Battery [133] (n = 1 [67]); a combination of the TGMD, TGMD-2, Victorian FMS manual, and the Functional Movement Screen™ (n = 1 [76]); the McCarron Assessment of Neuromuscular Development [134] (n = 1 [74]); the TGMD (n = 1 [77]); an adapted version of the Athletic Ability Assessment [135] (n = 1 [78]); the Movement Assessment Battery for Children-2 (MABC-2; [136]; n = 2 [81, 108]); the supine to stand test [137] (n = 1 [56]), a combination of the MABC-2 and the test of motor competence [138] (n = 1 [57]); a combination of the TGMD, TGMD-2 and Get Skilled Get Active tests (n = 1 [91]); the Athletic Introductory Movement screen ([139]) and tuck jump assessment [140] (n = 1 [100]); the MABC (n = 1 [90]); the TGMD-3 [141] (n = 1 [121]); a combination of the TGMD-3 and the Victorian FMS Manual [125] (n = 1 [104]); the Motorische Basiskompetenzen (MOBAK) [142,143,144] (n = 1 [109]); the Life-Long Physical Activity Skills Battery [145] (n = 1 [110]); the back squat assessment [146] (n = 1 [120]); and an unreferenced measure of stability/balance (n = 1 [69]).

A total of 30 studies measured the association between motor competence and physical activity among adolescents [36, 55, 58, 67, 70, 71, 74, 77, 79, 84, 87, 88, 91, 94, 95, 103,104,105,106,107, 110,111,112,113,114,115,116,117, 119, 121]. Measures of physical activity engagement included non-referenced self-reporting questionnaires (n = 4 [55, 71, 87, 88]), accelerometery (n = 9 [75, 91, 94, 104, 110, 113, 116, 117, 119]), the Physical Activity Questionnaire for Older Children (PAQ-C; [147]; n = 7 [79, 95, 103, 105,106,107, 111]), the Adolescent Physical Activity Recall Questionnaire [148] (n = 1 [67]), the Flemish Physical Activity Questionnaire [149] (n = 1 [36]), pedometers (n = 2 [74, 121]), step activity monitors (n = 1 [84]), the Leisure Time Physical Activity Questionnaire [150] (n = 1 [58]), the Health Behaviour in School-Aged Children Survey [151] (n = 1 [112]); the International Physical Activity Questionnaire (Short Form) [152] (n = 1 [114]); the MoMo Physical Activity Questionnaire [153] (n = 1 [115]); and an unreferenced question about weekly engagement in sport, fitness or recreational activity (n = 1 [77]).

The association between motor competence and physical fitness was assessed across ten domains (Table 2). Motor competence was assessed against composite fitness scores (n = 9 [57, 66, 83, 90, 97, 104, 108, 114, 119]), weight status (n = 21 [56, 58, 69, 74, 77, 79, 83, 84, 91, 93,94,95,96, 100, 101, 103, 105,106,107, 110, 116]), muscular endurance (n = 10 [56, 69, 74, 77, 84, 93, 94, 99, 110, 113]), muscular power (n = 12 [69, 72, 78, 82, 86, 88, 92, 100, 110, 115, 118, 120]), speed (n = 5 [69, 72, 78, 82, 120]), agility (n = 5 [72, 74, 86, 92, 118]), muscular strength (n = 6 [56, 72, 74, 88, 115, 120]), cardiovascular endurance (n = 16; [56, 72,73,74, 77, 78, 80, 84, 88, 93, 94, 99, 110, 113, 115, 117]), flexibility (n = 6 [56, 74, 77, 88, 110, 115]) and functional mobility (i.e., timed up-and-down stairs test; n = 1 [84]).

A total of five psychosocial domains were assessed against motor competence among adolescents. The association between motor competence and motivation was evaluated by six studies [36, 55, 83, 85, 98, 100]. Studies measured motivation using the Behavioural Regulation in Exercise Questionnaire-2 (BREQ-2; [154]; n = 2 [83, 100]), the Sport Motivation Scale [155] (n = 2 [55, 85]), a Dutch version of the Behavioural Regulation in Physical Education Questionnaire [156] (n = 1 [36]) and a Spanish version of the Perceived Locus of Causality Scale (PLOC; [157]; n = 1 [98]).

Seventeen studies measured the association between motor competence and perceived motor competence [36, 55, 67, 68, 81, 83, 85, 88,89,90, 98,99,100, 104, 109, 115, 121]. Measures utilised to assess perceived motor competence included the Physical Self-Perception Profile (PSPP [158, 159]; n = 1; [67]), the PSPP Sports Competence Subscale (n = 3 [55, 85, 99]), the PSPP and the Pictorial Scale of Perceived Movement Skill Competence (PSPMSC [160]; n = 1 [89]), the PSPMSC (n = 1 [98]), the PSPMSC and the PSPMSC in Stability Skills [68] (n = 1 [68]), the Sport/Athletic Competence Subscale [161] of the Children and Youth Physical Self Perception Profile [162] (n = 1 [36]), the Self-Description Questionnaire-2 [163] (n = 1 [81]), the International Fitness Scale [164] (n = 1 [83]), the Norwegian version [165] of the Perceived Athletic Competence Subscale of the Self-Perception Profile for Adolescents [166] (n = 1 [88]), the Norwegian version [167] of the Self-Perception Profile for Children [161] (n = 1 [90]), the Perceived Physical Ability Scale for Children [168] (n = 1 [100]), the Self-Perception Profile for Adolescents [169] (n = 1 [104]), the Perceived Competence Scale for Children [170] (n = 1 [121]), the Selbstwahrnehmung der motorischen Kompetenz (SEMOK) [109] (n = 1 [109]), and the Physical Self-Description Questionnaire [171, 172] (n = 1 [115]). Pullen et al. [100] also analysed the association between motor competence and global self-esteem via the Rosenberg Self-Esteem Scale [173]. Fu and Burns [121] also measured the association between motor competence and physical activity enjoyment via the Sport Enjoyment Scale [174].

Five studies [70, 75, 76, 83, 102] measured the association between motor competence and self-efficacy/confidence. Smith et al. [83] used a self-efficacy scale related to resistance training [175], Fu and Burns [121] used a six-item self-efficacy scale [176] and the remaining studies [70, 75, 76, 102] used the Physical Self-Confidence Scale [177].

3.2 Risk of Bias Overview

The risk of bias overview of included studies is presented in Table 3. No studies met all six criteria, nine studies met five criteria [36, 57, 67, 73, 81, 97, 99, 102, 113], eight studies met four criteria [66, 68, 70, 76, 79, 90, 112, 117], 12 studies met three criteria [71, 74, 75, 82, 89, 92, 95, 100, 106, 108, 120, 121], 13 studies met two criteria [58, 72, 80, 83,84,85, 93, 94, 96, 98, 105, 114, 115] and 19 studies met one [55, 77, 78, 86, 87, 91, 101, 103, 104, 110, 111, 118, 119] or none [56, 69, 88, 107, 109, 116] of the criteria. Criteria one and four were the least met criteria (n = 15/61 and 17/61 respectively), followed by criterion five (n = 21/61), criterion six (n = 29/61) and criterion two (n = 30/61). Most studies met criterion three (40/61).

Table 3 Risk of bias assessment overview
Full size table

3.3 Meta-analysis

An overview of the associations between motor competence and physical activity, physical fitness and psychosocial characteristics in adolescence is presented in Fig. 3. Individual meta-analyses are presented in Figs. 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 and 16.

3.3.1 Pooled Correlation Coefficients for Motor Competence and Physical Activity

For motor competence and physical activity, correlation coefficients were analysed from 13 studies [36, 55, 58, 67, 74, 79, 88, 91, 110, 111, 113, 119, 121]. Figure 4 shows the pooled correlation coefficients (i.e. overall summary statistics) were significant, positive and small (r = 0.20–0.26) for each type of motor competence evaluated against physical activity.

Fig. 3
figure 3

Overview of the range of pooled correlation coefficients between motor competence and physical activity, physical fitness and psychosocial characteristics in adolescents

Full size image
Fig. 4
figure 4

Forest plots showing the pooled correlation coefficients between motor competence and physical activity (r ± 95% confidence interval [CI]). Bold font indicates the summary statistics for each type of motor competence represented. 03 2003 participants, 10 2010 participants, f female, loc locomotor competence, m male, obj object control competence, r correlation coefficient, stab stability/balance competence, ap < 0.05, bp < 0.01

Full size image
3.3.1.1 Pooled Correlation Coefficients for Motor Competence and Physical Fitness Characteristics

Composite Fitness Scores Seven studies analysed correlation coefficients for the association between motor competence and composite fitness [57, 66, 90, 97, 108, 114, 119]. Figure 5 shows that studies only reported correlation coefficients for overall competence, with the pooled correlation coefficient being significant, positive and moderate (r = 0.39).

Fig. 5
figure 5

Forest plots showing the pooled correlation coefficients between motor competence and composite fitness scores (r ± 95% confidence interval [CI]). Bold font indicates the summary statistics for each type of motor competence represented. f female, m male, MABC movement assessment battery for children, r correlation coefficient, TMC test of motor competence, ap < 0.05, bp < 0.01

Full size image

Weight Status The association between motor competence and weight status was analysed from 17 studies [56, 58, 74, 79, 88, 91, 93, 95, 96, 100, 101, 105,106,107, 110, 114, 119]. Pooled correlation coefficients ranged from trivial to moderate (r =  − 0.35 to − 0.10) and were all significant. The pooled correlation coefficients for locomotor and sports-specific competence consisted of fewer than three study samples (Fig. 6).

Fig. 6
figure 6

Forest plots showing the pooled correlation coefficients between motor competence and weight status (r ± 95% confidence interval [CI]). Bold font indicates the summary statistics for each type of motor competence represented. 03 2003 participants, 10 2010 participants, bmi body mass index, f female, fm fat mass, loc locomotor competence, m male, obj object control competence, proc process measure of motor competence, prod product measure of motor competence, r correlation coefficient, stab stability/balance competence, ap < 0.05, bp < 0.01

Full size image

Muscular Endurance Six studies [56, 74, 93, 99, 110, 113] examined the association between motor competence and muscular endurance (Fig. 7). The pooled correlation coefficient was significant, positive and moderate for overall competence (r = 0.34), locomotor competence (r = 0.44), object control competence (r = 0.31) and sports-specific competence (r = 0.36) to muscular endurance. Stability/balance competence had a significant, positive and high association with muscular endurance (r = 0.52). However, the pooled correlation coefficients for stability/balance competence and sports-specific competence to muscular endurance consisted of fewer than three study samples.

Fig. 7
figure 7

Forest plots showing the pooled correlation coefficients between motor competence and muscular endurance (r ± 95% confidence interval [CI]). Bold font indicates the summary statistics for each type of motor competence represented. loc locomotor competence, obj object control competence, proc process measure of motor competence, prod product measure of motor competence, r correlation coefficient, ap < 0.05, bp < 0.01

Full size image

Muscular Power The meta-analysis of the association between motor competence and muscular power evaluated the correlation coefficients from seven studies [72, 82, 86, 92, 100, 110, 118]. Figure 8 shows significant positive correlation coefficients for overall competence (r = 0.29, small) and stability/balance competence (r = 0.03, trivial) to muscular power.

Fig. 8
figure 8

Forest plots showing the pooled correlation coefficients between motor competence and muscular power (r ± 95% confidence interval [CI]). Bold font indicates the summary statistics for each type of motor competence represented. chest throw, f female, m male, r correlation coefficient, SBJ standing broad jump, stab stability/balance competence, VJ vertical jump, ap < 0.05, bp < 0.01

Full size image

Speed Two studies [72, 82] were analysed for the association between motor competence and speed. A pooled correlation coefficient was produced for overall competence only, which was significant, negative and moderate (r =  − 0.31; Fig. 9).

Fig. 9
figure 9

Forest plots showing the pooled correlation coefficients between motor competence and speed (r ± 95% confidence interval [CI]). Bold font indicates summary statistics for each type of motor competence represented, 10 10-m sprint time, 20 20-m sprint time, 30 30-m sprint time, r = correlation coefficient, ap < 0.05, bp < 0.01

Full size image

Agility The association between motor competence and agility was evaluated from three studies [86, 92, 118]. Figure 10 shows that pooled correlation coefficients for overall competence (r =  − 0.37, p = 0.01) and stability/balance (r =  − 0.21, p > 0.05) competence were negative, moderate and small, respectively.

Fig. 10
figure 10

Forest plots showing the pooled correlation coefficients between motor competence and agility (r ± 95% confidence interval [CI]). Bold font indicates the summary statistics for each type of motor competence represented. f female, m male, r correlation coefficient, stab stability/balance competence. ap < 0.05, bp < 0.01

Full size image

Muscular Strength A total of five studies [56, 72, 74, 88, 120] were evaluated for the association between motor competence and muscular strength. Pooled correlation coefficients were produced for overall competence (r = 0.36) and stability/balance competence (r = 0.41), which were significant, positive and moderate (Fig. 11).

Fig. 11
figure 11

Forest plots showing the pooled correlation coefficients between motor competence and muscular strength (r ± 95% confidence interval [CI]). Bold font indicates the summary statistics for each type of motor competence represented. abs absolute strength, f female, hgl hand grip test left hand, hgr hand grip test right hand, m male, proc process measure of motor competence, prod product measure of motor competence, PU push-up test, r = correlation coefficient, rel strength relative to body mass, SBJ standing broad jump test, ap < 0.05, bp < 0.01

Full size image

Cardiovascular Endurance The meta-analysis to evaluate the association between motor competence and cardiovascular endurance consisted of eight studies [56, 72,73,74, 88, 93, 110, 113]. Figure 12 shows the pooled correlation coefficients for each element of motor competence measured. The associations for all components with cardiovascular endurance were significant, positive and moderate (r = 0.37 to 0.48), except for locomotor (r = 0.60) and object control (r = 0.50) competence, which were significant, positive and high. However, the correlation coefficients for locomotor competence, object control competence and sports-specific competence consisted of fewer than three studies.

Fig. 12
figure 12

Forest plots showing the pooled correlation coefficients between motor competence and cardiovascular endurance (r ± 95% confidence interval [CI]). Bold font indicates the summary statistics for each type of motor competence represented. 13y 13 years old, 15y 15 years old, f female, loc locomotor competence, m male, obj object control competence, stab stability/balance competence, proc process measure of motor competence, prod product measure of motor competence, r correlation coefficient, VO2 max, VO2rel VO” max relative to body mass, ap < 0.05, bp < 0.01

Full size image

Flexibility A total of four studies [74, 88, 93, 110] were evaluated to identify the pooled correlation coefficients for motor competence and flexibility. Sports-specific competence had a non-significant negative trivial association with flexibility (r =  − 0.07), while significant positive small associations were identified for overall competence (r = 0.23) and stability/balance competence (r = 0.17) with flexibility. However, the meta-analyses for sports-specific competence and stability/balance competence consisted of fewer than three studies (Fig. 13).

Fig. 13
figure 13

Forest plots showing the pooled correlation coefficients between motor competence and flexibility (r ± 95% confidence interval [CI]). Bold font indicates the summary statistics for each type of motor competence represented. f female, m male, r correlation coefficient, ap < 0.05, bp < 0.01

Full size image
3.3.1.2 Pooled Correlation Coefficients for Motor Competence and Psychosocial Characteristics

Perceived Motor Competence For the association between motor competence and perceived motor competence, a total of 13 studies [36, 55, 67, 68, 81, 85, 88,89,90, 98,99,100, 121] were evaluated (Fig. 14). The associations between locomotor competence and stability/balance competence to perceived motor competence were significant, positive and small (r = 0.25 and 0.26, respectively). Further, significant positive moderate associations were identified for object control competence (r = 0.34) and overall competence (r = 0.34).

Fig. 14
figure 14

Forest plots showing the pooled correlation coefficients between motor competence and perceived motor competence (r ± 95% confidence interval [CI]). Bold font indicates the summary statistics for each type of motor competence represented. f female, loc locomotor competence, m male, obj object control competence, pmc fms perceived motor competence in fundamental movement skills, pmc loc perceived motor competence in locomotor skills, pmc obj perceived motor competence in object control skills, pmc self self-competence, pmc sport perceived motor competence in sports, pmc stab perceived motor competence in stability/balance skills, r correlation coefficient, stab stability/balance competence, ap < 0.05, bp < 0.01

Full size image

Self-Efficacy/Confidence Three studies [70, 76, 121] evaluated the association between motor competence and self-efficacy/confidence (Fig. 15). The association between overall competence and self-efficacy/confidence was small (r = 0.22); no further elements of motor competence were represented.

Fig. 15
figure 15

Forest plots showing the pooled correlation coefficients between motor competence and self-efficacy/confidence (r ± 95% confidence interval [CI]). Bold font indicates the summary statistics for each type of motor competence represented. 13y 13-year-olds, 14y 14-year-olds, 15y 15-year-olds, f female, FMS fundamental movement skill assessment, func functional movement screen assessment, m male, r = correlation coefficient, ap < 0.05, bp < 0.01

Full size image

Motivation A total of five studies [36, 55, 85, 98, 100] were analysed to identify the association between motor competence and motivation. The pooled correlations for all elements of motor competence were significant, except for object control competence, where the association was positive but trivial (r = 0.07). Associations for locomotor, overall and stability/balance competence were positive and small (r = 0.15 to 0.20). All elements of motor competence (except overall competence) were represented by fewer than three study samples (Fig. 16).

Fig. 16
figure 16

Forest plots showing the pooled correlation coefficients between motor competence and motivation (r ± 95% confidence interval [CI]). Bold font indicates the summary statistics for each type of motor competence represented. f female, loc locomotor competence, m male, obj object control competence, r correlation coefficient, stab stability/balance competence, ap < 0.05, bp < 0.01

Full size image
3.3.1.3 Heterogeneity

The degree of heterogeneity was moderate (> 50%) for locomotor competence to physical activity, stability/balance competence to weight status and object control competence to cardiovascular endurance. A high degree of heterogeneity (> 75%) was identified for overall, object control and stability/balance competence to physical activity, overall competence to weight status, locomotor competence to muscular endurance, overall competence to muscular power, stability/balance competence to muscular strength, overall and stability/balance competence to cardiovascular endurance, stability/balance competence to flexibility, and object control and stability/balance competence to perceived motor competence.

3.3.1.4 Sensitivity

The sensitivity analysis mainly showed minor changes. Independently eliminating three subgroup samples (male and female subgroup samples from the Huotari et al. 2010 cohort [58] and the male subgroup sample from O’Brien et al. [91]) altered the association between object control competence and weight status from small to trivial. The overall competence and muscular power association changed from small to moderate when individually removing each muscular power correlation from Pichardo et al. [72]; male and female vertical jump correlations from Kramer et al. [86]; and the female standing broad jump correlation from Kramer et al. [86]. The association between overall motor competence and speed increased from small to moderate when independently removing 20-m or 30-m sprint correlations from one study [72]. Removing Lloyd et al. [118] changed the association between overall competence and agility from moderate to small, while removing the female sample from Kramer et al. [86] changed the association between stability/balance competence and agility from non-significant and small to non-significant and trivial. The removal of the male sample from Haugen et al. [88] altered the association between stability/balance competence and cardiovascular endurance from moderate to high.

3.3.1.5 Evaluation of Small Study Effects

Inspection of the funnel plots and Egger’s regression intercepts revealed statistically significant Egger’s regression statistics for the association between overall competence and weight status (intercept =  − 4.21, 95% CI − 6.17, − 2.26, p < 0.01). The association between overall competence and weight status was not considered symmetrical, indicating the presence of a small study effect [178].

3.3.1.6 Moderator Variables

The subgroup analysis of the potential moderator variables (i.e., age, sex, type of motor competence assessment) is presented in supplementary Table 1. Pairwise comparisons showed three significant moderators; (1) the association between object control competence and physical activity was greater for male individuals (r = 0.33) compared with female individuals (r = 0.21, p = 0.04); (2) the association between overall competence and physical activity was greater in studies using product motor competence assessments (r = 0.31) versus process assessments (r = 0.18; p = 0.03); and (3) the association between overall competence and weight status was greater for studies with a mean age between 13 and 15 years (r =  − 0.37), compared with studies with a mean age between 11 and 12 years (r =  − 0.21; p = 0.03). There were no other significant differences in associations for motor competence and physical activity, physical fitness or psychosocial characteristics between any other potential moderators.

4 Discussion

4.1 Overview of the Main Findings

A key focus during adolescence is the synergistic development of motor competence, physical fitness and psychosocial characteristics [1]. The interaction between these characteristics is suggested to induce positive or negative physical activity and weight status trends amongst youth [21]; a hypothesis that potentially explains declining physical activity [8] and increasing obesity levels (e.g. UK [9], USA [10]) amongst these individuals. This systematic review with meta-analysis is the first to (1) analyse the scientific literature evaluating associations between motor competence and physical activity, physical fitness and/or psychosocial characteristics amongst adolescents; (2) evaluate the associations between motor competence and physical activity, physical fitness characteristics and/or psychosocial characteristics amongst adolescents; and (3) investigate the impact of moderator variables (i.e. age, sex, type of motor competence assessment) on these associations.

A total of 61 studies were reviewed [36, 55,56,57,58, 66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121], totalling 22,256 participants, providing a comprehensive systematic evidence base of the associations between motor competence and physical activity, physical fitness and psychosocial characteristics amongst adolescents. Findings from the qualitative review indicated that when examining the associations of motor competence during adolescence: (1) risk of bias is present across all studies; (2) longitudinal evaluations are limited, (3) few studies account for, or consider, maturity status, (4) few studies associate motor competence across multiple characteristics (i.e. physical activity, physical fitness, psychosocial) and (5) either process (i.e., technique) or product (i.e. outcome) measures are favoured when assessing motor competence compared to combined (process and product) methods.

Within the present study, physical activity, composite fitness, muscular endurance, muscular power, muscular strength, cardiovascular endurance, perceived motor competence, self-efficacy/confidence and motivation were positively associated with motor competence; weight status, speed and agility were inversely associated with motor competence. Flexibility showed positive and negative associations with motor competence depending upon the motor skills assessed. These findings align with previous evidence [12, 14, 21, 25, 28] across youth, suggesting that associations of motor competence continue throughout childhood and adolescence. Moderator comparisons (i.e. age, sex, type of motor competence assessment) presented three significant differences: (1) the association between object control competence and physical activity was greater for male individuals compared with female individuals; (2) the association between overall competence and physical activity was greater in studies using product motor competence assessments versus process assessments; and (3) the association between overall competence and weight status was greater for studies with a mean age between 13 and 15 years, compared with those between 11 and 12 years. These findings suggest that motor competence, physical activity engagement and physical fitness are complex during adolescence, when substantial physiological, biological and body composition changes are ongoing, meaning a greater understanding is required.

4.2 Summary of Study Methods

Risk of bias was present across all included studies (0/61 met all six criteria). Validity (criterion two) and reliability (criterion three) of motor competence assessments had the highest adherence. Thus, while numerous motor competence assessments are available, the most current assessments are valid and reliable for practitioners to utilise within their environments. Sampling characteristics (criterion one) and validity of physical activity/fitness/psychosocial measures (criterion four) presented the lowest adherence. The low adherence to criterion one is attributed to the limited detail regarding sampling methods (e.g. random/convenience) and participant demographics (e.g. age, sex, ethnicity). Criterion four’s low adherence highlights inconsistencies in reporting the validity of measures used. These inconsistencies could confound the results presented and indicate that future research should utilise valid measures of physical activity, physical fitness and psychosocial characteristics during adolescence. Such information is important to fully understand the confounding factors that may influence any associations evaluated. Thus, authors should provide detail regarding participant sampling characteristics (e.g. sampling method, sample size, age, sex, stage of maturity) and the validity and reliability of study measures (e.g. of motor competence, and physical activity/fitness/psychosocial measures) to enhance study quality.

Most studies (45 out of 61) included within the systematic review used cross-sectional study designs, with ten studies [80, 103,104,105, 111,112,113,114,115,116] collecting longitudinal evaluations. The remaining studies conducted randomised controlled trial interventions [102, 117, 119], or used validity and reliability methods [66, 68, 110]. This finding aligns with previous motor competence reviews [14, 29, 47], and supports the need for future longitudinal investigations. Whilst cross-sectional study designs allow researchers to highlight current trends at single timepoints, longitudinal designs may be more appropriate to understand the developmental trajectories of the associations between these characteristics, alongside the long-term influence of potential moderators (e.g. sex, age, maturity status). Furthermore, longitudinal research can confirm previous cross-sectional outcomes and highlight the most appropriate opportunities for interventions to enhance health, well-being and performance outcomes in adolescence [179].

When evaluating motor competence, physical activity, physical fitness and psychosocial characteristics across adolescence, maturity status should be considered. Maturity status is asynchronous with chronological age [38, 39] and can lead to temporary reductions in motor competence (i.e. adolescent awkwardness) during the adolescent growth spurt [42]. Eight studies within this review measured the maturity status of adolescents. For example, Ryan et al. [82] showed that Fundamental Movement Screen™ scores stagnated between pre-PHV and circa-PHV (d = 0.3; 95% CI − 0.6, 1.2), before increasing during post-PHV (circa- to post-PHV d = 1.4; 95% CI 0.5, 2.2), which supports the adolescent awkwardness hypothesis during peak growth. Furthermore, Kokstejn et al. [97] showed that during pre-PHV (estimated years from PHV =  − 2.88 ± 0.3 years), adolescents’ motor competence is negatively associated with maturity status (r =  − 0.29, p < 0.01), whilst Pichardo et al. [72] identified no association between maturity status and motor competence in circa-PHV male individuals (estimated years from PHV = 0.2 ± 0.9 years; r = 0.00, p > 0.05). These findings show that stages of maturity may influence health and performance characteristics differently. While measuring maturity status is a strength of these studies, no studies explored the effect of maturity status on associations between motor competence and physical activity, physical fitness and psychosocial characteristics. Future research should longitudinally track maturity status during adolescence and examine its influences on the association between motor competence and physical activity, physical fitness and psychosocial characteristics.

The hypothesised Stodden et al. [21] model has been responsible for most motor competence research, worldwide, over the last decade. However, while multiple motor competence associations were hypothesised, most studies (n = 34) within this review only compared motor competence to one characteristic (i.e. physical activity, physical fitness or psychosocial). Only three studies [88, 104, 115] evaluated associations across all characteristics. This finding supports that of Barnett et al. [29] who identified that few studies have investigated the entire model. One explanation for this finding is that multivariate approaches may be required to analyse associations between the variables within the Stodden et al. [21] model (e.g. physical fitness, psychosocial) because a univariate analysis can only determine relationships between two variables in a pairwise manner at any given time [180]. Nevertheless, based on currently available evidence, only inferences can be made on all aspects of the Stodden et al. [21] model in adolescents, and there is a need for more holistic longitudinal research to examine the model in its entirety.

When measuring motor competence, most studies used process (i.e. technique; n = 25) or product (i.e. outcome; n = 31) assessments. Only one study used a combined motor competence measure (i.e. process and product criterion; [74]), while four studies [56, 77, 86, 92] reported separate correlations for process and product elements. These findings support other reviews (e.g. [181]), which similarly show studies favouring process or product assessments of motor competence. Such methods limit the overview of an individual’s motor competence [182]. For example, evaluating an individual’s technique enables assessors to identify and correct inadequate movement patterns to inform training interventions [183], prevent injury [184] and increase perceived motor competence [28]. Conversely, product-based measures show long-term changes in movement outcomes [185]. Process evaluations are subjective and require experienced assessors [186], while product-based measures cannot identify individual differences in motor competence as they are outcome based [187]. Consequently, authors have developed valid approaches to assess combined motor competence (e.g. the Canadian Agility Movement Skills Assessment [188], and the Dragon’s Challenge [189]), which offer viable alternatives that practitioners should consider for assessing motor competence.

4.3 Summary of Meta-analyses

When assessing associations with physical activity, physical fitness and psychosocial characteristics, meta-analyses were conducted separately for different motor competence domains (i.e. overall competence, locomotor, object control, stability/balance, sports specific). This approach highlighted the scarcity of studies that provided correlations for the separate domains (see Figs. 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 and 16), meaning that for some characteristics, insufficient study samples were available to analyse their associations with motor competence. Therefore, care should be taken when reviewing some associations, owing to their limited evidence base. Where possible, future research should report associations with physical activity, physical fitness, and psychosocial characteristics as an overall score and separate motor competence domains.

4.3.1 Heterogeneity

The degree of heterogeneity varied depending on the characteristics measured. Higher heterogeneity occurred within meta-analyses consisting of greater study samples/sample sizes. Heterogeneity arises because of the grouping of studies that are methodologically diverse [64]. Thus, within the different meta-analyses, higher heterogeneity likely represents the diversity of the included studies’ population characteristics (e.g. sex, age, nationality) and the variety of motor competence assessments used across studies (27 different assessments identified). Thus, future research requires more consistent approaches for measuring associations between motor competence, physical activity, physical fitness and psychosocial characteristics among adolescents.

4.3.2 Association Between Motor Competence and Physical Activity

In the meta-analyses of 13 studies, a small association between motor competence and physical activity was seen among adolescents. The lowest association with physical activity was stability/balance competence, and the highest association with physical activity was overall competence, suggesting that a variety of motor skills such as throwing, catching, running, jumping and balancing are similarly important for physical activity engagement. A recent review indicated that supportive social environments are key to adolescent physical activity behaviours (e.g. active travel, sports participation) [190]. Perhaps, such environments may favour those with a broad range of motor skills that allow participation at the same levels as their peers (i.e. can engage successfully in a given social environment), particularly as displaying incompetence in front of others and exposure to embarrassment are perceived barriers to physical activity during adolescence [191]. Such experiences may be exaggerated in countries where there are strong links between school and sport (e.g. USA), although further research is required to test this hypothesis.

Current findings support previous reviews that identify positive associations between motor competence and physical activity in children and adolescents [25,26,27,28] but contradict the recent findings of Barnett et al., [29] who found no evidence supporting these associations. While Barnett et al. [29] explain their findings via a publication bias and a tendency within sports science research to only report significant associations, the present review shows no evidence of publication bias, with both non-significant and significant correlations extracted from the included studies. However, the present review’s sole focus on adolescent populations and the lack of longitudinal evidence presented may explain this contradiction.

Because of the variance in study methods (e.g. objective vs subjective physical activity assessments, participant characteristics, motor competence measures), comparing studies is challenging. Additionally, the tools used to assess physical activity and motor competence associations need acknowledging. For example, accelerometery does not capture the intensities of specific motor competencies (e.g. object control) [192], and consequently presents a lower association with these motor competencies [29]. This limitation is highlighted by O’Brien et al., [91], who measured physical activity in male individuals via accelerometery and reported a trivial association between physical activity and stability/balance competence. Measurement limitations should be considered during the research design process when assessing the associations between physical activity and motor competence in adolescents.

4.3.3 Association Between Motor Competence and Physical Fitness

Within these meta-analyses, various pathways of the Stodden et al. [21] model are represented. However, this study evaluated a broader range of physical fitness characteristics against motor competence compared with others (e.g. [12].), indicating the scarcity of evidence investigating individual physical fitness characteristics compared to other characteristics within the Stodden et al. [21] model. Thus, more research is required to strengthen the understanding of physical fitness and motor competence in adolescents.

Composite Fitness This review identified a moderate positive association between motor competence and composite fitness (r = 0.39). However, this association may represent similarities between product-based motor competence assessments and physical fitness measures (e.g. distance covered in standing long jump) that consist of similar neuromuscular actions [12]. For example, Vedul-Kjelsas et al. [90] measured physical fitness via a tennis ball throw, which bears similarities to components of the MABC-2 (e.g. ball skills). Within this meta-analysis, ten study samples utilised product-based assessments compared to two samples [66, 119] using process-based measurements. However, both samples identified a moderate positive association between process-orientated assessments and these characteristics, which suggests that associations may not be influenced by the type of motor competence assessment used. Future research should consider the similarities between product-based motor competence assessments and physical fitness measures in their methodologies. Adopting a combined (i.e. process and product) measure of motor competence is recommended when comparing to composite fitness scores, to account for measurement similarities and provide greater clarity on this particular association.

Weight status Weight status was negatively associated with motor competence in the meta-analyses (r =  − 0.36 to − 0.10). All motor competence domains were represented, although the meta-analysis for locomotor competence and sports-specific competence included insufficient study samples. These findings support similar evidence in youth [12, 25] and may be explained by the detrimental effect of increased body mass on motor competencies involving the projection of an individual’s body mass (e.g. jumping, running [12, 193]). However, body mass index was the most popular measure of weight status (n = 36/41 study samples). Measuring weight status via body mass index is a limitation of the current adolescent literature as lean/fat mass cannot be directly measured [84, 91, 194]. Within this review, Tadiotto et al. [56] identified that fat mass was negatively associated with motor competence, while lean mass was positively associated. This finding highlights the importance of differentiating between components of body composition when comparing associations with motor competence during adolescence, where lean mass gains occur, especially in male individuals [195]. Consequently, future research should focus on utilising more appropriate and practical measures of weight status that can differentiate between lean and fat mass (e.g. bioelectrical impedance) [28].

Of the meta-analyses undertaken, only the association for overall competence and weight status presented a small study effect, with the funnel plot indicating the presence of a significant publication bias. Explanations for the publication bias within this particular meta-analysis could include the use of a sedentary sample ([56] inclusion criteria = not physically active except for school time physical education and > 2 h of screen time per day), participants of a low socioeconomic status [96] and small sample sizes [95]. Therefore, care should be taken when interpreting the association between overall competence and weight status presented in this review, and future research should seek to limit publication bias.

Muscular Endurance, Power and Strength Compared with previous reviews (e.g. [12, 25].), this meta-analysis conducted a broader evaluation of motor competence associations with musculoskeletal fitness (e.g. muscular endurance, power, strength). The meta-analyses identified moderate positive associations between motor competence and musculoskeletal endurance, and muscular strength, as well as trivial-to-small positive associations with muscular power. Such findings suggest that musculoskeletal fitness and motor competence are mutually important for physical activity engagement [196]. For example, athletic tasks combine different skills that require both learnt levels of coordination and efficient force production/absorption capabilities (e.g. netball pass, jumping to catch a rebound in basketball) [12, 18]. Therefore, interventions should seek to synergistically develop musculoskeletal fitness and motor competence for positive health outcomes. Within this review, authors lacked consensus when classifying musculoskeletal fitness measures. For example, Kramer et al. [86] and Pichardo et al. [72] measured muscular power via a standing broad jump, whilst Haugen et al. [88] used this assessment to measure muscular strength. The limited consensus creates a cross-over in associations of motor competence and musculoskeletal fitness characteristics (i.e. muscular power scores contributing to muscular strength associations and vice versa), which could confound the associations presented. Therefore, future research requires more standardised measures to assess musculoskeletal fitness characteristics and facilitate between-study comparisons.

Speed and Agility Motor competence was negatively associated with speed and agility. No previous review has examined these associations because of focusing on health-related fitness (i.e. cardiovascular and musculoskeletal fitness; [12, 25, 29]). A broader focus on physical fitness components of athleticism [1] is a strength of the present study and allows the evaluation of additional characteristics required for physical activities/sports. These negative associations indicate that better speed and agility performance is synonymous with greater motor competence. However, readers should cautiously interpret the associations between motor competence, speed and agility because of the few studies (two for speed, three for agility) and study samples (four for speed, six for agility) evaluated. The need for caution is highlighted by a sensitivity analysis. Independently removing two study samples from the motor competence-speed meta-analysis changed this negative association from moderate to small, while the removal of one study from the agility meta-analysis changed this negative association from small to trivial. Nevertheless, low correlations between motor competence and speed/agility indicate the importance of other physical fitness characteristics for speed/agility. Previous research supports this hypothesis as relative strength is associated with longer step lengths (r = 0.79), and faster sprint speed (r = 0.42) [197]. Because of insufficient studies investigating the association between motor competence and speed, and agility, further research is required to understand these interactions fully.

Cardiovascular Endurance Overall, sports-specific and stability/balance competence were moderately associated with cardiovascular endurance (r = 0.38 to 0.60). However, a lack of study samples for locomotor competence (n = 2), object control competence (n = 2) and sports-specific competence (n = 1) means that these associations are inconclusive. Nevertheless, 12 study samples provide strong evidence for a moderate association between overall competence and cardiovascular endurance, which supports other findings across youth [12, 25, 28, 29]. Cattuzzo et al. [12] hypothesised that multiple physical fitness characteristics are both directly (i.e. via neuromuscular development) and indirectly (i.e. increased ability to participate in physical activities that promote cardiovascular fitness) linked with motor competence. For example, activities promoting cardiovascular endurance require repetitive, consecutive, concentric and eccentric contractions, which encompass contralateral limb coordination [12, 72]. These muscular actions may explain the high association between locomotor competence and cardiovascular endurance (r = 0.60) presented by two study samples within this meta-analysis. However, future study needs to explore this hypothesis owing to a lack of study samples for different motor competence domains.

Flexibility This meta-analysis shows that the association between motor competence and flexibility is inconclusive and concurrent with similar findings in youth [12, 25, 29]. The present results can be attributed to a lack of studies exploring this association. Nevertheless, both hyperflexibility and hypoflexibility can affect children’s movement capabilities [31]. Further, some adolescents experience temporary reductions in motor competence during circa-PHV [42], suggesting that maturation may affect flexibility. With limited consideration for maturity status throughout this review, further research is needed to clarify the association between motor competence and flexibility during adolescence.

4.3.4 Association Between Motor Competence and Psychosocial Characteristics

Perceived Motor Competence and Confidence The association between motor competence and perceived motor competence ranged from small to moderate, with all domains except sports-specific competence represented. The strongest evidence for this association was for overall competence (13 study samples included). Less evidence was available for locomotor, object control and stability/balance competence (four, five and five, respectively), suggesting that more in-depth evaluations of these associations are required. The present findings support those of De Meester et al. [14], who identified a small association between overall and perceived motor competence (r = 0.25). Previous understanding suggests that an individual’s accuracy of estimating motor competence increases with age [137]. However, because of insufficient study samples across different age groups (13 and 15 years [n = 7], followed by 11–12 years [n = 4] and 16 years and over [n = 1]), this meta-analysis was unable to evaluate any advances in self-evaluation ability and complexity of self-description that occur during adolescence. Additionally, maturity status likely influences self-perceptions and may moderate the associations with motor competence [198], although no studies within this review reported their findings in a way to examine this hypothesis. Therefore, future research should compare associations between motor competence and perceived motor competence by age group/stage of maturity.

The results of this meta-analysis may also reflect the alignment between actual and perceived motor competence measurements. For example, skills measured during actual motor competence assessments (e.g. Körperkoordinationstest Für Kinder—FMS) may not represent self-perceptions within existing broader measures (e.g. PSPP). Both Estevan and Barnett [22] and De Meester et al. [14] have recently advocated for better alignment between actual and perceived motor competence measurements. De Meester et al. [14] have called for authors to better articulate alignment and utilise different measures of perceived motor competence to assess the importance of alignment. Similarly, McGrane et al. [177] indicated the need for self-perception measures that capture differentiated perceptions of motor competence to a greater extent (e.g. PSPMSC—FMS). Thus, as our understanding of actual competence continues to develop (e.g. foundational movement skills [13], athleticism [18]), there is a need for commensurate development and alignment of perceived motor competence measurements with a particular research focus on process versus product motor competence measures and the variety of perceived motor competence measures available.

Self-Efficacy/Confidence Within this meta-analysis, only three studies reported associations between motor competence and self-efficacy/confidence. However, as per the definition of athleticism, youth engage with confidence as well as actual/perceived competence [1], suggesting that a greater understanding of this interaction is required. Again, as our understanding of actual competence develops, self-efficacy/confidence might be best understood related to specific motor competencies that are assessed. Consequently, more research exploring how perceived motor competence and self-efficacy/confidence are associated with actual motor competence is required.

Motivation Motor competence and motivation associations were trivial to small, with only overall competence represented by enough study samples (n = 4). Developing approaches to accurately determine motor competence can provide individuals with more realistic expectations of their competence, reduce the incidence of unsuccessful outcomes and reduce the incidence of lower motivation [36, 199]. Previous studies have found that significant amounts of autonomous motivation are explained by an adolescent’s perceived motor competence [200, 201]. However, most studies (3/5 studies) within this meta-analysis reported motivation via relative autonomy index scores. Thus, it is unclear how different components of motivation influence this association, although we hypothesise that greater motor competence is associated with greater autonomy for physical activity. Additional research is required to evaluate this association amongst adolescents and should account for the effect of perceived motor competence. Nevertheless, practitioners should promote success for all adolescents to encourage autonomous motivation and participation in physical activity, regardless of an individual’s actual motor competence [36, 200].

4.4 Moderator Variables

Overall, potential moderators (i.e. sex, age, type of motor competence assessment) produced a limited influence on the strength or orientation of motor competence associations during adolescence. However, a moderator analysis identified three significant findings. First, the association between object control competence and physical activity was greater for male individuals compared with female individuals. During motor competence assessments, male individuals often outperform female individuals in power and strength tasks, while female individuals perform better than male individuals during fine motor tasks, flexibility and balance [202]. Three out of five studies in this meta-analysis compared male and female associations between object control competence and physical activity using product measures (e.g. throwing). Such skills are complex multi-segmental motions that require energy transfer and timing [203, 204]. Inadvertently, this may explain the presented sex difference, as object control tasks require a prerequisite of strength and power to achieve desired outcomes.

Second, the overall competence and physical activity association was greater in studies using product motor competence assessments versus process assessments. This difference contradicts recent evidence in children, which suggests that process and product assessments are poor at explaining the variance in children’s physical activity [205]. Thus, this finding may suggest that as individuals develop into adolescence, successful physical activity engagement is synonymous with an individual’s ability to perform desired outcomes within the activities being explored, regardless of the technique behind it.

Last, the overall competence and weight status association was greater for studies with a mean age between 13 and 15 years, compared with studies with a mean age between 11 and 12 years. Indeed, excess weight can hinder the long-term development of motor competence [206]. This finding may therefore represent the negative trajectories of the developmental model [21] (i.e. poor weight status and motor competence leads to reduced physical activity, fewer opportunities to develop motor competence, and therefore, results in further weight gain as an individual develops). However, caution is needed when interpreting this finding, owing to the large difference in study samples within this moderator comparison (13 study samples for age 13–15 years; six study samples for age 11–12 years). Therefore, future research should explore the effects of age on motor competence and weight status associations.

Overall, the limited moderator findings are attributed to a lack of study samples per moderator, association and motor competence domain to draw meaningful conclusions. Thus, this section highlights the need for research exploring potential moderators (i.e. age, sex and type of motor competence assessment) for each motor competence domain to fully understand any moderator effects.

4.5 Strengths and Limitations

This systematic review with a meta-analysis is novel given the sole focus on adolescents owing to currently poor health-related trends during this stage of physical and psychosocial development. However, this review only included studies published in English, and important information may have been missed from non-English publications. Nevertheless, evidence was evaluated from 16 countries, which represents a broad overview of the associations between motor competence, physical activity, physical fitness and psychosocial characteristics in adolescence. Second, studies of physically/cognitively impaired adolescents were not included because of the already broad nature of this review. A previous review could not determine if the association between motor competence and perceived motor competence was stronger for typically developing or physically/cognitively impaired individuals because of a lack of study samples [14]. Consequently, it was beyond the scope of this review to further explore this comparison amongst adolescents. Nevertheless, this review highlighted the limited evidence regarding the influence of maturity status on the associations between motor competence and physical activity, physical fitness and psychosocial characteristics during adolescence. Furthermore, this review followed the updated PRISMA guidelines [48], included numerous database searches throughout the review process, followed clear and robust inclusion/exclusion criteria and utilised a second reviewer for screening purposes (title/abstract/full-text screening, study bias assessment).

5 Practical Applications

This review with meta-analysis provides several practical applications. First, researchers should (where feasible) include longitudinal assessments across adolescence, utilise combined motor competence tools (i.e. process and product), report overall and process/product scores, report scores for different motor competence domains (e.g. object control, stability/balance) and consider how maturity status influences such associations. Second, those seeking to design interventions to improve health-related outcomes in adolescence should focus on the synergistic development of motor competence, physical fitness and psychosocial characteristics (rather than focusing on sports alone) to increase physical activity opportunities. Adolescence is a complex and challenging period of the lifespan consisting of physical [38, 39, 42] and psychosocial changes [43,44,45], and organisations and practitioners need to recognise such complexities and collaborate to support continual development. For example, reflecting on current motor competence, physical fitness and psychosocial practices, and evaluating the importance of these characteristics can spark awareness of the developmental needs for different stages of maturity (e.g. [207].). Recent research (e.g. [2, 208,209,210].) also clarifies the goals and realities of an adolescent’s long-term developmental needs and provides suitable recommendations that practitioners and organisations can adopt to promote a fitter, healthier and more physically active adolescent population. However, it is apparent that such research needs translating into useful resources for coaches, teachers and organisations to utilise within their environments.

6 Conclusions

This paper aimed to (1) analyse the scientific literature evaluating associations between motor competence and physical activity, physical fitness and/or psychosocial characteristics amongst adolescents; (2) evaluate the associations between motor competence and physical activity, physical fitness characteristics and/or psychosocial characteristics amongst adolescents; and (3) investigate the impact of moderator variables (i.e. age, sex, type of motor competence assessment) on these associations. This study expands on previous reviews (e.g. [12, 25].), by focusing on adolescents, exploring broader physical fitness components of athleticism (e.g. muscular power, speed, agility) and discussing the potential influence of maturity status on associations. The risk of bias assessment highlighted suboptimal reporting of sampling methods, participant characteristics and the validity of physical activity/physical fitness/psychosocial measures. Furthermore, this review supports the need for longitudinal exploration of the Stodden et al. [21] developmental model during adolescence [12, 14, 25, 28, 29]. Present findings highlight several methodological differences when measuring motor competence, physical activity, physical fitness and psychosocial characteristics. Specifically, studies favoured either process or product motor competence evaluations, which when used independently, provide a limited overview of an individual’s motor competence [182]. Finally, the review showed that few studies considered the influence of maturity status on motor competence associations, even though adolescents can experience a transient decline in coordination during peak growth (i.e. circa-PHV; [12]).

The current meta-analyses support previous evidence [12, 14, 25,26,27,28, 47] exploring the hypothesised motor competence associations [21] and identified positive associations between motor competence and physical activity, composite fitness scores, muscular endurance, muscular power, muscular strength, cardiovascular endurance, perceived motor competence and motivation, in addition to inverse associations between motor competence and weight status, speed and agility. Interventions to enhance an adolescent’s health and well-being should synergistically target motor competence, physical and psychosocial development. However, improved evaluations of these characteristics are required to better inform such interventions during adolescence.