Background
Just under half of children aged from 2 to 6 years achieve 60 min per day in physical activity [
1]. Given that this is a minimum recommendation in this age range, this is a serious concern. Focusing on physical activity promotion in this age range is important as physical activity behaviors established during early childhood may track into later childhood [
2] and adolescence [
3].
Motor competence in children has gained credence in the last decade as an important correlate of physical activity and other health related behaviours and outcomes [
4], including higher cardiorespiratory fitness and healthier weight status [
5]. Motor competence is a broad term that encompasses fundamental movement skill ability, including locomotor, object control and stability factors [
6], and motor coordination [
4].
Previously it has been hypothesized that children with better motor competence participate in higher levels of physical activity and that this in turn helps to further develop higher actual and perceived motor competence [
7]. Countless sporting activities and games need competence in fundamental motor skills (e.g., running, jumping, catching, throwing) to enable physical activity participation [
4]. Now, close to a decade later, there is convincing evidence that physical activity and actual motor skill competence are associated in children and youth [
4,
5,
8]. In addition, high perceptions of physical competence contribute to increased physical activity in children and youth [
9]. Yet as most evidence is cross-sectional, the relationship between physical activity and actual and perceived motor competence is not fully understood.
The inherent hypothesis posed by Stodden and colleagues [
7] in terms of the relationship between motor competence and physical activity being important in both directions has had limited testing [
4,
10]. More is known about the skill to physical activity relationship than the physical activity to skill relationship. These relationships are hypothesized to differ at different stages, i.e., engagement in physical activity is important for the development of motor competence but then as children develop, motor competence is of more importance for physical activity participation. In a path analysis study in adolescents, Barnett and colleagues found a reciprocal relationship between object control skill competence and physical activity, but a one way relationship from physical activity to locomotor skills. Perceived competence acted as a mediator in both proposed path directions, although the strength differed [
10].
There are few longitudinal studies to draw from. The northern Finland Birth Cohort found that age at walking supported and age at standing unaided predicted sports participation at 14 years [
11]. Studies in older children found childhood motor skill competence (age 10 and age six respectively) was a predictor of subsequent physical activity [
12,
13] and physical activity in Grade 7 was a predictor of motor competence in Grade 8, but only for boys [
14]. No longitudinal study to our knowledge has examined physical activity as a determinant of actual or perceived motor competence in very young children. A systematic review in preschool children determined physical activity was a key cross-sectional correlate of motor competence [
15] indicating that it is worthwhile to investigate such associations at this young age. Understanding more about the role of physical activity in the development of motor competence in young children will potentially help inform recommendations and advice to parents, carers, health professionals and educators of children in this age group. This study therefore sought to investigate whether physical activity in the toddler and/or preschool years influences subsequent motor skill competence at school starting age (5 years). The hypothesis was that physical activity behaviour would predict actual and perceived motor competence.
Results
There were no differences in actual movement skill at age 5 between children who had valid physical activity data and children who did not at any time point. Those who remained in the sample had a trend to be less active at 19 months old than those who dropped out (28 min per day compared to 25 min, p = 0.050).
Descriptive data for the parent sample (based on the time point with the largest number of parents) is presented in Table
1. Parent respondents were largely mothers born in Australia who spoke English at home. Around 60 % of respondent parents were working part- or full-time and a similar proportion had a university education.
Table 1
Descriptive data for the parent sample that were included when their child was 19 months (n = 193)
Residential Socioeconomic position | High | 37 | 19.2 |
Medium | 123 | 63.7 |
Low | 33 | 17.1 |
Respondent Parent | Mother | 184 | 95.3 |
Country of birth | Australia | 158 | 81.9 |
Main language spoken at home | English | 186 | 96.4 |
Employment status | Working full or part time | 113 | 58.5 |
Education level | University degree or higher | 116 | 60.1 |
Most children (
n = 164, 84.0 %) were not attending school at 5 years old. Children’s MVPA at each time point and actual and perceived skills at age 5 are described in Table
2. Children were least active at 19 months, and most active at 5 years old. Average perceived skill was 41.4 out of a possible 48 (range of 29 to 48), showing children tended to respond on average between ‘pretty good’ and ‘really good’ for each skill. Based on standardized TGMD scores nearly all children were average (object control
n = 81, 63.8 %, locomotor
n = 87, 68.5 %), or below average (object control
n = 35, 27.6 %, locomotor
n = 31, 24.4 %) in their skill level at age 5.
Table 2
Child sample, MVPA at each time point and actual and perceived skills at age 5
19.0 mths (2.2) | 193 | 98 (50.8) | 95 (49.2) | 39.2 | 25.1 (9.4) | – | – | – | – | – | – |
3.5 years (0.2) | 118 | 53 (44.9) | 65 (55.1) | 24.0 | 42.5 (16.1) | – | – | – | – | – | – |
5.0 years (0.1) | 127 | 59 (46.5) | 68 (53.5) | 25.8 | 52.8 (17.9) | 49.7 (9.7) | 26.0 (5.5) | 23.3 (6.1) | 41.4 (4.7) | 21.1 (2.3) | 20.3 (2.9) |
MVPA as a predictor of total skill competence at age 5 years
MVPA at 19 months was not a predictor of total skill (actual or perceived) at age 5. MVPA at 3.5 years was predictive of perceived total skill at age 5 years (B = 0.059,
p = 0.044) and approached significance for actual skill at age 5 years (B = 0.109,
p = 0.059). MVPA at 5 years was not associated with total skill (actual or perceived). See Table
3.
Table 3
MVPA at each age (19 months, 3.5 years, 5 years) as a predictor of actual total skill and perceived total skill at age 5 years
Actual Total Skill Age 5 |
−.116 | .073 | .117 | −.260,.029 | .109 | .057 | .059 | −.004,.222 | .078 | .050 | .121 | −.021,.178 |
Sex (boy)*, Age*. | No sig. adjustment variables | No sig. adjustment variables |
Perceived Total Skill Age 5 |
−.028 | .037 | .446 | −.102,.045 | .059 | .029 | .044* | .002,.116 | .016 | .025 | .514 | −.033,.066 |
No sig. adjustment variables | No sig. adjustment variables | No sig. adjustment variables |
MVPA as a predictor of object control skill at age 5 years
MVPA at any age was not associated with object control skill (actual or perceived) at age 5 years. See Table
4.
Table 4
MVPA at each age (19 months, 3.5 years, 5 years) as a predictor of actual and perceived object control skill at age 5 years
Actual Object Control Skill Age 5 |
−.061 | .046 | .185 | −.153, .030 | .032 | .034 | .348 | −.035,.098 | .029 | .031 | .356 | −.033,.090 |
Age*, Sex (Boy)** | Age*, Sex (Boy)* | Sex (Boy)* |
Perceived Object Control Skill Age 5 |
−.032 | .022 | .151 | −.077,.012 | .033 | .017 | .061 | −.002,.067 | .011 | .015 | .483 | −.020, .041 |
Sex (Boy)* | Treatment group (Control)* | No sig. adjustment variables |
MVPA as a predictor of locomotor skill at age 5 years
MVPA at 19 months was not a predictor of locomotor skill (actual or perceived) at age 5. MVPA at 3.5 years was predictive of actual locomotor skill at age 5 years (B = 0.073,
p = 0.033), but not perceived locomotor skill. At 5 years, MVPA was not cross-sectionally associated with locomotor skill (actual or perceived). See Table
5.
Table 5
MVPA at each age (19 months, 3.5 years, 5 years) as a predictor of actual and perceived locomotor skill at age 5 years
Actual Locomotor Skill Age 5 |
−.048 | .042 | .251 | −.131, .035 | .073 | .034 | .033* | .006,.139 | .043 | .029 | .134 | −.014,.100 |
No sig. adjustment variables | Sex (Girl)* | No sig. adjustment variables |
Perceived Locomotor Skill Age 5 |
.004 | .019 | .842 | −.034,.042 | .026 | .015 | .085 | −.004,.056 | .006 | .012 | .651 | −.019, .030 |
No sig. adjustment variables | No sig. adjustment variables | No sig. adjustment variables |
Discussion
This study explored whether early physical activity behaviour impacts on subsequent actual or perceived motor competence. We found that children had positive perceptions overall. MVPA at age 3.5 years was predictive of perceived total skill. MVPA in preschool years predicted locomotor but not object control skills. Stodden et al. [
7] postulated that, in young children, engagement in physical activity is important for the development of motor competence but then as children develop, motor competence is of more importance for physical activity participation. Our findings support the first part of this hypothesis, in that MVPA in the preschool years did contribute to subsequent motor competence in the current sample. A recent review concluded that whilst there is strong evidence of an association between motor competence and physical activity (e.g., [
5,
8,
27]), there were still questions in terms of antecedent/consequent mechanisms [
4]. Based on our sample, a preschool child who spent 15 min a day more in MVPA would demonstrate approximately one unit higher for locomotor competence. This is similar to performing one additional component correctly in one of the six locomotor skills (e.g., “both feet come off the floor together and land together” in the jump). This may be meaningful, as mastering particular components may enhance play and game opportunities. For instance being able to do a ‘step hop’ sequence means being able to play games involving the skip. As such, this finding warrants recommending to parents and preschool staff that MVPA at this age does make a difference to skill outcomes.
Other research has demonstrated ‘free play’ does not contribute much to motor skill competence [
28], so it is suggested that in very young children the type and quality of physical activity relate more to motor skill development than simply movement quantity and intensity. This may explain why the relationship found in this study was with subsequent locomotor competence rather than object control competence. This would be expected as the activities that preschool-aged children are likely to engage in are more locomotive in nature (e.g., running and playing in the garden). High participation in ball games would be more likely to be associated with better object control competence. One study investigated different types of organized physical activity participation in preschool aged children (dance, ‘kindy gym’, swimming) and found different associations in terms of object control and locomotor competence [
29]. This provides further evidence that young children need specific opportunities for skills to be taught and practised. Whilst objective measurement of physical activity has many advantages, it does not enable isolation of the types and quality of physical activity that children are engaging in. A recent systematic review of motor skill correlates in children and youth found that whilst physical activity was a correlate of some aspects of motor competence (i.e., total skills and motor coordination), evidence was indeterminate for physical activity being a correlate of object control or locomotor skill competence [
30]. This illustrates further that when examining relationships between physical activity and motor competence, that it is important to examine relationships according to the way motor competence is operationalized.
Stodden and colleague’s model [
7] also hypothesized that relationships between motor competence and physical activity increases in strength as children age and develop. A recent narrative review supports this premise, with the conclusion that overall, whilst data strongly supports a positive relationship between motor competence and physical activity, associations are low to moderate in early to middle childhood [
4]. At first glance our null findings at age 5 years appears counter intuitive, and contrary to other investigations (e.g., [
29,
31]) but it could also be viewed that previous time in MVPA is more important to the current level of skill than any pattern of current MVPA. Motor competence needs time to develop and if we also consider that MVPA only tracks to a moderate level in early childhood [
32], it is reasonable that previous time in MVPA may be more important in some cases to skill development than current time in MVPA.
It is likely that the null finding in terms of toddler activity being a predictor of subsequent motor competence is heavily influenced by the developmental stage of children at this early age. Even though we adjusted for age at first walking, many children would still be using crawling as a common form of locomotion and that behaviour is not picked up well by accelerometers. Since children’s activity levels are so influenced by their developmental stage at this age they therefore may not be a true reflection of characteristic physical activity levels of each individual child. This may help to explain why we found no relationships between toddler MVPA and subsequent actual and perceived skill.
Few studies have investigated perceived physical competence and physical activity in preschool age children [
33]. Those studies which have specifically examined perceived motor competence (rather than general perceptions of physical competence), have focused on an older group of children (first three years of school) [
34,
35]. Cross-sectional findings have suggested that whilst school children’s perceptions of their motor competence are associated with their actual skills [
34,
35], perceptions did not relate to their current physical activity levels [
34,
36]. Yet, in the present sample, MVPA among preschool-aged children was associated with subsequent overall perceived motor competence at age 5 years. Our finding can be interpreted as a child who spends 15 more minutes in MVPA per day would be approximately one unit higher in perceived competence; similar to shifting from a self-perception of ‘sort of good’ to ‘pretty good’ in one of the 12 skills. This confirms what was found with actual skill, in that previous physical activity behaviour appears (in this sample) to have more influence on children’s perceptions than current physical activity behaviour. Considering that developmentally children tend to rate themselves favourably, this indicates that the amount of time in MVPA still has the ability to discriminate between children with different self-perceptions. This finding is meaningful, as a child who believes they are ‘pretty good’ at catching as opposed to ‘sort of good’, may potentially be more motivated to participate in a whole range of ball sports that involve catching.
The strengths of the current study include objective measurement of physical activity at three time points in a good sized sample which reflects well on generalizability of findings. A further strength is the use of an instrument to assess perceived motor competence aligned with a measure of actual motor competence. Limitations are that actual and perceived motor competence were only measured at age 5 years. This meant a mixed model which investigated change over time could not be used. Whilst there are instruments which could have assessed motor competence broadly at each time point, it was not possible for fundamental movement skill competence to be assessed at each time point as the TGMD-2 is only suitable to a minimum age 3 years, and motor milestones are relevant at age 19 months. The TGMD-2 is the only instrument to our knowledge which comprehensively assesses a broad range of fundamental movement skills in young children. It was not possible to measure perceived motor competence prior to age 5 as younger children could not be expected to complete this assessment due to their cognitive developmental stage. The number of children with valid physical activity assessments at each time point (43–71 % of age 5 sample) is a limitation, however there were no differences in skill between those with valid assessments and those without, showing this did not likely influence internal validity of results. Although those who stayed in the study were less physically active at 19 months old.
Acknowledgements
Thank you to the parents and children who were involved. The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation.