Physical literacy defined as the motivation, confidence, physical competence, knowledge, and understanding to value and take responsibility for engagement in physical activities for life.1 It is theorized that physical literacy and physical activity have a bidirectional relationship such that increased physical literacy leads to increased physical activity participation, which in turn increases physical literacy over time, thus positioning physical literacy as an important determinant of health across the lifespan.2 We previously reported that physical literacy is associated with favorable health indicators in school-age children, and that moderate to vigorous physical activity (MVPA) mediated the relationship between physical literacy and aerobic fitness.3 Among a sample of youth from northern Canada, physical literacy was moderately associated (r = .20–.44) with self-reported physical activity.4 In another study, physical literacy explained 30% of the variance in pedometer-measured physical activity in a sample of 110 boys and girls.5 Using a latent profile approach with over 2000 youth, 5 profiles of physical literacy were identified in children and youth across 2 timepoints. The children in the highest profile for physical literacy also demonstrated the highest physical activity over time,6 providing evidence for the theory that increased physical literacy over time leads to increased physical activity. To date, the proposed opposite relationship of increased physical activity over time leading to increased physical literacy has not yet been well described or explored in the literature and warrants further investigation.
There is growing understanding about longitudinal physical activity patterns across childhood. For example, MVPA has been shown to increase in the early years7 and decrease in the school-age years and into adolescence. A systematic review reported that MVPA declined 3.4 minutes per day per year across childhood, and the decline was steeper in girls than boys8; however, not all children follow the same declining trajectories. Group-based trajectory modeling is a novel, empirical method to identify clusters of individuals following similar patterns or developmental trajectories over time, such as changes in physical activity participation. In contrast to traditional methods that may determine group membership a priori or based on assumptions about the patterns over time (ie, dividing participants into those that increase vs decrease physical activity), group-based trajectory modeling determines group membership based on the observed patterns within the specific dataset being analyzed.9 Using this method to analyze physical activity data from ages 10 to 16, Pate et al10 reported that less than 5% of the sample remained highly active over time, 55% were active at baseline and declined, and 40% had consistently low physical activity that declined over time. The high physical activity group included predominantly males.10 In another study, 4 MVPA trajectories from 5 to 19 years old were reported: consistently active (18%), consistently inactive (15%), declining physical activity (53%), and substantially declining physical activity (14%).11 While these studies have established trajectories for physical activity in school-age children, what is missing is a description of physical activity trajectories across the preschool to school-age years. We previously reported that physical activity patterns vary greatly from year-to-year to early childhood, suggesting patterns are not yet set.12 It is believed that young children’s physical activity patterns are impacted by the transition from childcare or home settings to primary school,13,14 yet these studies have reported on changes within an entire sample rather than investigating if children follow different trajectories of physical activity. Given the proposed longitudinal relationships between physical activity and physical literacy,2 the aforementioned positive, cross-sectional associations between physical activity and physical literacy,3–5 and the positive associations between increased physical literacy over time and physical activity,6 it is now necessary to determine if, and how, physical activity trajectories across childhood are associated with physical literacy.
The Health Outcomes and Physical activity in Preschoolers (HOPP) Study was a prospective, observational study of physical activity and health outcomes that followed a cohort of children annually for 3 years.15 The School-age Kids health from early Investment in Physical activity (SKIP) Study was a follow-up to the HOPP Study that used the same measures to follow the same cohort for an additional 3 years, with physical literacy assessments added to the final timepoint. Given the limited research available on the longitudinal physical activity patterns from preschool to school-age, the primary aim of the current study was to examine group-based trajectories of device-assessed measured physical activity from the preschool to school-age years. To address the gaps in literature about the proposed associations between longitudinal physical activity patterns and physical literacy, the secondary aim was to determine if trajectory group membership was associated with school-age physical literacy. The maturity status and proportion of boys and girls in each trajectory group was also examined. Trajectories of both total volume of physical activity and MVPA, assessed as accelerometer counts per minute (CPM) and minutes per day (min/day), respectively, were examined so the results may be relevant to the entire age spectrum of participants, to public health guidelines and to physical activity promotion.16,17 We hypothesized that children would follow several unique physical activity trajectories over time and that children who engaged in consistently high physical activity would demonstrate the highest school-age physical literacy scores.
Methods
Study Design and Participants
Participants were part of the HOPP Study, the methodology of which has been reported previously,15,18 and its follow-up study, the SKIP Study, both conducted at McMaster University. All assessments were completed during annual study visits in a laboratory located at McMaster Children’s Hospital. Physical literacy assessments were added to the final timepoint of the SKIP Study. The Hamilton Integrated Research Ethics Board provided ethical approval for both studies. Written informed consent from the child’s legally authorized representative (eg, parent) was obtained, and participants aged 7 years and older provided assent to participate in the study.
Four hundred and eighteen 3-, 4-, and 5-year-old children enrolled in the HOPP Study (data collected from 2010 to 2014), and 279 of these participants subsequently enrolled in the SKIP Study (data collected from 2015 to 2019). The physical literacy assessment at timepoint 6 was completed by 222 participants (see Supplementary Figure S1 [available online] for a flow chart of participants through the HOPP and SKIP Studies). This study includes data from both the HOPP and SKIP studies; however, only n = 279 participants who consented to the SKIP study are included. For this subset of participants (n = 279), the average total follow-up period from timepoint 1 to timepoint 6 was 6.3 (0.6) years (range: 5.00–8.37 y) and average time between timepoints was 1.3 (0.1) years. Reasons for not participating in the physical literacy assessments included: the visit was scheduled before ethics approval was granted for these additional assessments (n = 4), participant did not complete a timepoint 6 laboratory visit (n = 7), participant and/or parent denied participation in the extra assessment (n = 19), or a trained assessor was unavailable (n = 5).
Physical Activity
Physical activity was assessed using an accelerometer (ActiGraph GT3X line) worn for all waking hours over the right hip for 7 days. Parents and/or participants were asked to keep a written record of when the device was removed, including for water activities. Data were downloaded and analyzed in 3-second epochs and analyzed with Actilife software (version 6.13.4, ActiGraph, Pensacola, FL). Nonwear time was identified as periods of at least 60 minutes of continuous zero counts and device removal as indicated in the participant’s logbook.19 At least 3 days with a minimum of 10 hours of wear per day were required for inclusion in the analyses.20 Physical activity was expressed as CPM and MVPA. CPM are a measure of overall volume of total physical activity that combines all movement recorded in the vertical/y-axis while the device is worn (total counts divided by total wear time). MVPA was expressed as minutes per day based on cut points (>574 counts/ 15 s) developed and validated by Evenson et al.21
Physical Literacy
The Physical Literacy Assessment for Youth Tools were developed by Sport for Life and represent a series of assessment tools to assess the multiple domains of physical literacy. The Physical Literacy Assessment for Youth Tools were designed for children 7 years and older. In combination, the PLAYfun, PLAYparent, and PLAYself tools provide a multiperspective assessment of a participant’s physical literacy.22 Participants in the SKIP Study completed PLAYfun and PLAYself, and participant’s parent or guardian completed PLAYparent. The administration of PLAYfun, PLAYparent, and PLAYself has been previously described.3
Briefly, the PLAYfun assessment includes 18 movement skills within 5 domains: running, locomotor, object control (upper body and lower body), balance, stability, and body control.22 All PLAYfun assessments were administered and scored by one of 2 assessors (HATC or NAD). PLAYfun has very good to excellent interrater reliability and acceptable internal consistency.23,24 PLAYfun scores also increase with increasing age as developmentally expected.23,24 The PLAYself questionnaire is a 22-item self-evaluation of a child’s own physical literacy that includes 4 subsections: environment, physical literacy self-description, relative rankings of literacies (literacy, numeracy, and physical literacy), and fitness. The PLAYparent questionnaire is an evaluation of a parent’s perception of their child’s physical literacy, including questions about the child’s ability, confidence, and participation. PLAYparent provides researchers with an additional perspective of the children’s physical literacy and identifies positive and negative factors that affect the child’s ability to lead a healthy lifestyle. The Physical Literacy Assessment for Youth parent questionnaire is divided into 5 subsections: physical literacy visual analogue scale, cognitive domain, environment, motor competence (locomotor and object control), and fitness. Both PLAYself and PLAYparent have demonstrated acceptable internal consistency.23
Physical literacy z scores were calculated for PLAYfun, PLAYself, and PLAYparent as the individual values minus the group mean, divided by the SD to achieve variables that had a mean of 0 and SD of 1. The physical literacy composite score was calculated as the sum of the PLAYfun, PLAYparent, and PLAYself z scores, with higher values suggesting higher physical literacy. This measure has been used previously and is associated with MVPA and health indicators in school-age children.3
Anthropometry and Maturity
At each study visit, the participant’s height, weight, and sitting height were measured. Height and weight were used to calculate body mass index (kg/m2) and age-specific percentiles were calculated at each timepoint.25 Sitting height was used to calculate years from peak height velocity (YPHV) at timepoint 6 as an indicator of maturity using predictive equations.26
Statistical Analyses
All statistical analyses were conducted in STATA (version 14.2). Means, SD, minimum and maximum values for all variables were calculated to describe the distribution of each variable. Sex-dependent variation in all measures were examined with independent t tests. To identify patterns of trajectories of physical activity, group-based trajectory modeling with the TRAJ package in STATA27,28 was conducted using the CNORM distribution for continuous data. Each participant’s total volume of physical activity and MVPA (at 6 timepoints) were grouped within a pattern of conditional probabilities based on structural equation modeling theory that assumes individuals differ qualitatively as members of latent subgroups.29 Individual-specific probabilities of belonging to each subgroup allowed assignment to a subgroup based on the highest probability. The relationships between total volume of physical activity and age, and MVPA and age, were fitted up to a quadratic polynomial model for 2, 3, and 4 groups. The final number of groups and best fitting polynomial model were determined by the Bayesian Information Criterion (BIC), the proportion of participants in each group, and the change in BIC between models (−2ΔBIC).30 A 10-fold difference in BIC is considered a meaningful difference.10 In addition, to confirm the number of groups chosen, posterior probabilities and odds of correct classification (OCC) were calculated. Posterior probabilities >0.7 suggest the trajectory includes participants with similar patterns or change, and an OCC > 5 is recommended for all groups.31 Once the number of groups was determined, the models were rerun to eliminate nonsignificant quadratic terms. Under the assumption that data were missing at random, participants with incomplete data were included. Trajectories and the 95% confidence intervals surrounding each trajectory were plotted. The proportions of girls and boys in each trajectory group were examined.
After identifying latent total physical activity and MVPA subgroups that followed similar profiles, physical literacy (PLAYfun, PLAYparent, PLAYself, and composite score) were added to the models as distal outcomes. The differences in physical literacy outcomes between groups were then examined using postestimation analysis with Wald tests. A P value of .05 was used to indicate statistical significance. To examine the role of maturity status, YPHV was added as a covariate to the models to determine if YPHV was associated with trajectory group membership.
Results
Participant characteristics are included in Table 1. Age (P = .524), YPHV (P = .108), and sex (P = .097) were similar between participants who did and did not provide consent and assent to complete the physical literacy assessments at timepoint 6. One participant who provided consent and assent became distressed in the visit and withdrew participation from PLAYfun and PLAYself, and the parents of 2 participants did not complete PLAYparent. Timepoints 1 and 2 were separated by 1.0 (0.05) years, 2 and 3 by 1.0 (0.1) years, 3 and 4 by 2.3 (0.1) years, 4 and 5 by 1.0 (0.1) years, and 5 and 6 by 1.0 (0.1) years.
Descriptive Characteristics of Participants Across the 6 Timepoints
| All | Girls | Boys | P | |||
|---|---|---|---|---|---|---|
| Mean (SD) | Minimum | Maximum | Mean (SD) | Mean (SD) | ||
| Timepoint 1 (n = 279) | ||||||
| Age, y | 4.5 (0.9) | 3.0 | 6.0 | 4.5 (0.9) | 4.4 (0.9) | .759 |
| Height, cm | 106.5 (7.9) | 89.2 | 132.7 | 105.9 (7.9) | 107.2 (7.7) | .139 |
| Weight, kg | 17.8 (3.1) | 12.1 | 34.2 | 17.8 (3.1) | 18.1 (3.1) | .110 |
| BMI %ile | 50.3 (27.8) | 0.0 | 99.6 | 52.2 (27.8) | 48.6 (27.9) | .286 |
| Total physical activity (CPM) | 669.0 (149.4) | 384.7 | 1186.4 | 614.9 (133.0) | 715.5 (147.6) | <.001* |
| MVPA, min/d | 70.1 (18.2) | 35.3 | 130.7 | 62.2 (15.3) | 77.0 (17.9) | <.001* |
| Timepoint 2 (n = 279) | ||||||
| Age, y | 5.5 (0.9) | 4.0 | 7.0 | 5.5 (0.9) | 5.5 (0.9) | .717 |
| Height, cm | 113.5 (7.9) | 95.1 | 143.1 | 112.8 (7.8) | 114.1 (7.8) | .144 |
| Weight, kg | 20.1 (3.8) | 12.9 | 43.0 | 19.9 (3.8) | 20.4 (3.7) | .309 |
| BMI %ile | 49.7 (27.0) | 0.2 | 99.2 | 51.9 (27.7) | 47.6 (26.3) | .186 |
| Total physical activity (CPM) | 701.0 (155.2) | 312.7 | 1419.1 | 666.1 (168.4) | 733.0 (134.8) | <.001* |
| MVPA, min/d | 73.9 (18.0) | 30.8 | 123.2 | 67.6 (17.5) | 79.8 (16.5) | <.001* |
| Timepoint 3 (n = 276) | ||||||
| Age, y | 6.5 (0.9) | 5.0 | 8.0 | 6.5 (0.9) | 6.5 (0.9) | .759 |
| Height, cm | 120.0 (7.9) | 101.2 | 151.5 | 119.2 (7.9) | 120.7 (7.9) | .104 |
| Weight, kg | 22.6 (4.5) | 15.3 | 52.9 | 22.5 (4.8) | 22.8 (4.3) | .636 |
| BMI %ile | 48.3 (27.8) | 0.8 | 99.3 | 51.5 (28.5) | 45.5 (27.0) | .087 |
| Total physical activity (CPM) | 713.2 (183.6) | 314.0 | 1564.1 | 671.3 (163.1) | 751.2 (193.3) | <.001* |
| MVPA, min/d | 76.0 (21.6) | 34.4 | 171.1 | 69.2 (19.0) | 82.2 (22.0) | <.001* |
| Timepoint 4 (n = 279) | ||||||
| Age, y | 8.7 (1.1) | 6.2 | 11.2 | 8.8 (1.0) | 8.7 (1.1) | .252 |
| Height, cm | 133.6 (9.2) | 111.6 | 172.8 | 133.1 (8.8) | 134.1 (9.5) | .997 |
| Weight, kg | 29.8 (7.2) | 17.6 | 65.2 | 29.8 (7.3) | 29.8 (7.2) | .328 |
| BMI %ile | 48.6 (28.8) | 0.6 | 99.8 | 49.7 (28.7) | 47.6 (28.9) | .550 |
| Total physical activity (CPM) | 646.6 (189.1) | 184.6 | 1422.7 | 612.0 (178.1) | 678.8 (194.0) | .004* |
| MVPA, min/d | 69.83 (21.2) | 23.8 | 147.3 | 63.5 (17.9) | 75.7 (22.4) | <.001* |
| Timepoint 5 (n = 259) | ||||||
| Age, y | 9.8 (1.1) | 7.2 | 12.5 | 9.8 (1.0) | 9.7 (1.1) | .247 |
| Height, cm | 139.9 (9.6) | 117.4 | 178.1 | 139.4 (9.4) | 140.3 (9.7) | .829 |
| Weight, kg | 33.8 (8.7) | 19.3 | 77.6 | 33.9 (8.8) | 33.7 (8.6) | .323 |
| BMI %ile | 48.7 (29.8) | 0.6 | 98.8 | 49.6 (29.7) | 47.9 (30.0) | .643 |
| Total physical activity (CPM) | 606.8 (176.3) | 222.2 | 1373.2 | 557.1 (150.3) | 650.8 (186.1) | <.001* |
| MVPA, min/d | 67.9 (21.1) | 26.1 | 136.8 | 60.9 (18.4) | 74.4 (21.4) | <.001* |
| Timepoint 6 (n = 249) | ||||||
| Age, y | 10.7 (1.1) | 8.1 | 13.7 | 10.8 (1.0) | 10.7 (1.1) | .265 |
| YPHV, y | −1.8 (1.3) | −4.3 | 1.6 | −1.0 (1.0) | −2.6 (0.9 | <.001* |
| Height, cm | 145.7 (9.) | 124.5 | 186.1 | 145.6 (9.9) | 145.8 (10.0) | .279 |
| Weight, kg | 37.9 (10.5) | 22.3 | 91.8 | 38.3 (10.5) | 37.5 (10.5) | .124 |
| BMI %ile | 47.5 (30.6) | 0.4 | 99.0 | 48.5 (31.0) | 46.6 (30.8) | .613 |
| Total physical activity (CPM) | 557.7 (175.8) | 227.5 | 1367.8 | 510.1 (160.0) | 601.3 (179.0) | <.001* |
| MVPA, min/d | 62.7 (20.8) | 18.4 | 124.9 | 56.9 (19.5) | 68.1 (20.5) | <.001* |
Abbreviations: BMI, body mass index; CPM, accelerometer counts per minute, a measure of the total volume of physical activity; MVPA, moderate to vigorous physical activity; YPHV, years from peak height velocity. Note: P value represents the results of independent t tests to determine differences between boys and girls.
*P < .05.
Physical Activity Trajectories
Valid accelerometer data were available for 251 participants (90% of sample; 46.2% girls) at timepoint 1, 263 participants (94%; 47.9% girls) at timepoint 2, 265 participants (96%; 47.9% girls) at timepoint 3, 264 participants (95%; 48.2% girls) at timepoint 4, 243 participants (95%; 47.0% girls) at timepoint 5, and 237 participants (95%; 48.1% girls) at timepoint 6. All participants had at least one valid accelerometer timepoint (68.8% had 6 timepoints, 17.6% had 5 timepoints, 8.6% had 4 timepoints, 2.9% had 3 timepoints, 1.1% has 2 timepoints, and 1.1% had 1 timepoint). Of the participants with valid wear time, the average number of days worn across all timepoints was 5.7 (1.3) to 6.3 (1.1) days (range: 3–7 d), and daily wear time was 723.4 (40.9) to 768.7 (44.4) minutes per day. There were no observed differences in valid days or wear time between males and females (P = .364–.864) at any timepoints. Summary data on total volume of physical activity (CPM) and MVPA, at each timepoint, are shown in Table 1.
Three physical activity trajectories were identified for total physical activity and for MVPA using BIC, 2ΔBIC and consideration of group sizes. The group trajectories were supported by posterior probabilities >0.7 and OCC values >5.31 For total volume of physical activity, posterior probability values were >0.9 (OCC 9.1). For MVPA groups, posterior probability values were >0.9 (OCC > 11.4).
The longitudinal trajectories for total volume of physical activity are shown in Figure 1. Three distinct trajectories were identified. Total volume of physical activity and MVPA were lowest in Group 1. TPA Group 1 was comprised of 62.2% girls, and increased from timepoints 1 to 2, and then declined from timepoints 2 to 6. TPA group 2 comprised of 43.5% girls, increased from timepoints 1 to 3, and declined from timepoints 3 to 6. TPA Group 3, the most active group at all timepoints, included 27.3% girls and increased from timepoints 1 to 3 and then declined from timepoints 3 to 6. Figure 2 displays the 3 longitudinal trajectories of MVPA. The MVPA groups 1 and 2 displayed an overall decline in MVPA over the course of the study. Participants in MVPA group 3 (7.8% of participants) increased MVPA from timepoints 1 to 4 and declined from timepoint 5 to 6. Participation in MVPA remained higher in MVPA group 3 than MVPA groups 1 and 2 across all timepoints. MVPA group 1 included 63.1% girls, MVPA group 2 was 34.9% girls, and MVPA group 3 was 9.5% girls.
—Model-predicted trajectories of total volume of physical activity (CPM) across childhood based on latent group membership, represented as mean minutes and 95% confidence intervals by trajectory groups. Percentages represent the proportion of participants in each trajectory group. CPM indicates counts per minute.
Citation: Journal of Physical Activity and Health 19, 4; 10.1123/jpah.2021-0635
—Model-predicted trajectories of MVPA (min/d) across childhood based on latent group membership, represented as mean minutes and 95% confidence intervals by trajectory groups. Percentages represent the proportion of participants in each trajectory group. MVPA indicates moderate to vigorous physical activity.
Citation: Journal of Physical Activity and Health 19, 4; 10.1123/jpah.2021-0635
Physical Literacy
The average PLAYfun score was 49.1 (out of 100), with scores across from 39.4 to 97.5. The average PLAYself score was 73.5 (out of 100) and PLAYparent was 128.0 (out of 150). Boys displayed higher PLAYfun scores than girls (Table 2; P = .017), but there were no differences in PLAYself and PLAYparent between boys and girls. The physical literacy composite score ranged from −8.8 to 5.6, with no difference between boys and girls.
PLAYfun, PLAYself, PLAYparent, and Physical Literacy Composite Scores
| Whole sample (N = 222) | Girls (n = 113) | Boys (n = 109) | P | |||
|---|---|---|---|---|---|---|
| Mean (SD) | Minimum | Maximum | Mean (SD) | Mean (SD) | ||
| PLAYfun | 49.1 (7.6) | 21.2 | 67.7 | 47.9 (7.11) | 50.3 (7.8) | .017* |
| PLAYself | 73.5 (10.5) | 39.4 | 97.5 | 73.7 (10.8) | 73.3 (10.4) | .820 |
| PLAYparent | 128.0 (16.0) | 76.3 | 149.9 | 127.1 (15.9) | 128.9 (16.1) | .423 |
| Physical literacy composite score | −0.01 (2.2) | −8.8 | 5.6 | −0.22 (2.1) | 0.22 (2.3) | .151 |
Abbreviation: PLAY, Physical Literacy Assessment for Youth. Note: P value represents the results of independent t tests to determine differences between boys and girls.
*P < .05.
Physical Activity Trajectories and Physical Literacy
Physical literacy scores for total physical activity and MVPA groups are included in Table 3. Total physical activity group 1 (least active) had lower PLAYfun, PLAYparent, PLAYself, and physical literacy composite scores than groups 2 and 3 (P < .05). Total physical activity group 3 (most active) had higher PLAYfun, PLAYparent, and physical literacy composite scores than group 2 (P < .05), but similar PLAYself scores to group 2 (P = .591). Similarly, the least active MVPA group had lower PLAYfun, PLAYparent, PLAYself, and physical literacy composite scores than groups 2 and 3. MVPA group 3 (most active) had higher PLAYfun and physical literacy composite scores than group 2 (P = .003–.005), and similar PLAYself and PLAYparent scores to Group 2.
Physical Literacy at Timepoint 6 Across Total Volume of Physical Activity and MVPA Trajectory Group Membership
| Group mean (95% CI) | Pairwise comparisons (P values) | |||||
|---|---|---|---|---|---|---|
| Total volume of physical activity (N = 279) | Group 1 (n = 110) | Group 2 (n = 123) | Group 3 (n = 46) | Group 1 vs 2 | Group 1 vs 3 | Group 2 vs 3 |
| PL composite score | −1.2 (−1.7 to −0.7) | 0.3 (−0.2 to 0.7) | 1.6 (0.9 to 2.3) | <.001* | <.001* | .002* |
| PLAYfun | 46.3 (44.7 to 48.0) | 49.5 (47.9 to 51.0) | 54.6 (52.0 to 57.1) | .011* | <.001* | .001* |
| PLAYself | 69.5 (67.0 to 72.0) | 75.5 (73.2 to 77.9) | 76.8 (73.2 to 80.4) | <.001* | .001* | .591 |
| PLAYparent | 121.3 (117.6 to 125.0) | 129.8 (126.5 to 133.0) | 137.5 (132.4 to 142.6) | .001* | <.001* | .016* |
| Group mean (95% CI) | Pairwise comparisons (P values) | |||||
|---|---|---|---|---|---|---|
| MVPA (N = 279) | Group 1 (n = 148) | Group 2 (n = 109) | Group 3 (n = 22) | Group 1 vs 2 | Group 1 vs 3 | Group 2 vs 3 |
| PL composite score | −0.8 (−1.2 to −0.4) | 0.6 (0.1 to 1.1) | 2.2 (1.2 to 3.2) | <.001* | <.001* | .005* |
| PLAYfun | 46.84 (45.5 to 48.2) | 50.6 (49.0 to 52.1) | 56.6 (53.0 to 60.2) | .002* | <.001* | .003* |
| PLAYself | 71.1 (69.0 to 73.1) | 75.5 (73.2 to 77.8) | 79.4 (74.2 to 84.6) | .004* | .003* | .193 |
| PLAYparent | 123.1 (120.1 to 126.0) | 132.1 (128.8 to 135.5) | 138.5 (130.9 to 146.1) | <.001* | <.002* | .127 |
Abbreviations: APHV, age at peak height velocity; CI, confidence interval; MVPA, moderate to vigorous physical activity; PLAY, physical literacy assessment for youth; PL, physical literacy.
*Statistical significance P < .05.
Physical Activity Trajectories and Maturity Status
The YPHV at timepoint 6 was considered as a covariate in the trajectory analysis and, being less mature increased the likelihood of being in group 2 (log-odds estimates: −0.38 to −0.52; P = .001–.006), or in group 3 (log-odds estimates: −0.51 to −0.61; P = .003–.029). These results suggest being less mature was associated with higher physical activity participation over the previous data collection period.
Discussion
The objectives of this paper were to examine the group-based trajectories of device-assessed physical activity from the preschool to school-age years and to determine if trajectory group membership was associated with school-age physical literacy. We determined that children followed varied physical activity trajectories from preschool to school-age, and trajectory group membership was associated with school-age physical literacy. For MVPA trajectories, over 90% of participants displayed a decline in MVPA, while 8% of participants increased MVPA from preschool to school-age. For total volume of physical activity, all 3 trajectory groups gradually declined over time. The groups with the highest physical activity were comprised of predominantly boys, while the groups with the lowest physical activity were predominantly girls. Results indicated that low and declining physical activity were associated with lower physical literacy.
Using group-based trajectory analysis, our study observed overall group declines in MVPA in groups 1 (−1.5 min/d per timepoint) and 2 (−0.9 min/d per timepoint), which included over 90% of participants. We identified a small group of participants (group 3; 7.8%), which experienced consistently high MVPA across timepoints with a slight increase (+0.7 min/d per timepoint). At all timepoints, average MVPA for participants in MVPA groups 2 and 3 remained above the recommended physical activity guideline of 60 minutes of daily MVPA,16 while average MVPA of participants in MVPA 1 declined over time and fell below the physical activity guidelines at timepoints 4, 5, and 6. The changes in MVPA that we observed were less dramatic than the average decline (−3.4 min MVPA/d per year) reported in Farooq et al’s8 systematic review on physical activity across childhood (2020). In the Iowa Bone Development Study, 4 distinct MVPA trajectories were identified in a cohort of over 500 children from ages 5 to 19 years old, with 1 group (18% of sample) displaying consistently high physical activity over time while the remaining participants were consistently inactive (15%), decreased physical activity (53%), or substantially decreased physical activity (14%). In the same study, the substantially declining group saw the steepest decline starting at around 13–14 years old, a trend that may be seen if our cohort continues to be followed over time.11 The initial increases in MVPA we found across the preschool years align with previously reported increases in MVPA across the early childhood years.7
We observed distinct trajectory groups for total volume of physical activity; however, physical activity in all 3 groups increased in the preschool years, and then declined to lower than baseline at timepoint 6. Likewise, other studies have also noted a decline in physical activity across the school-age years. In a sample of 650 students followed from fifth to 11th grade (ages 10–16 y), 3 total physical activity trajectory groups were observed: 1 group remained high and increased from ages 14 to 16 years (4%), the second declined and remained low as children aged (40%), and the third had the lowest physical activity that declined over time (56%).10 In summary, we observed several trajectory groups of physical activity across childhood and consistent with previous studies, the most active trajectory groups included the smallest proportion of participants.
Previous research supports our finding that girls were overrepresented in the lowest, and underrepresented in the highest physical activity trajectory groups. In the Iowa Bone Development Study, the inactive MVPA trajectory group was predominantly females, and the consistently active group was predominantly males.11 Farooq and colleagues32 examined trajectories of physical activity from age 7 to 15 years separately for boys and girls in a cohort of 545 children and youth. Using similar analysis methods to ours, they observed steeper declining trajectories of total physical activity in girls compared to boys for TPA, and all declining MVPA trajectories for girls with 3 declining and 1 increasing MVPA trajectory in boys.32 Maturity was also associated with physical activity trajectory group membership. We determined that being less mature increased the likelihood of being in the moderate and high trajectory groups, compared to the low physical activity groups, but this may be attributed to a higher percentage of boys in the high physical activity groups. In a study by Pate et al,10 the high physical activity group was predominantly male and the least mature, while the least active group was predominantly female and the most mature. The participants in the least active groups in the Iowa Bone Development Study were also the most mature at the last timepoint.33 Future studies should continue to explore the associations between physical activity, gender, and maturity status to better understand longitudinal physical activity patterns as children grow and mature. Future research may also wish to examine additional differences between participants in the different trajectory groups, such as socioeconomic status or sports participation.
It is theorized that increased physical activity is associated with physical literacy in a positive feedback cycle, and therefore, physical literacy is a determinant of health given the established associations between increased physical activity and improved health.34,35 Within the current study, low and declining total volume of physical activity and MVPA were associated with lower physical literacy, whether expressed as individual assessments (PLAYfun, PLAYself, and PLAYparent), or as an overall composite score. If low physical activity levels are sustained over childhood, children may not develop their physical literacy and in turn, they may not be prepared to engage in physical activity later in life. There is limited research on the associations between physical activity and physical literacy, though the field is rapidly growing. In a cross-sectional analysis in the same cohort of participants as the current study, we found that physical literacy composite score explained 24% of the variance in daily minutes of MVPA in school-age children.3 Similarly, in a sample of over 2000 grade 5 boys and girls, 5 unique physical literacy profiles were identified which were associated with physical activity, such that the participants in the high physical literacy profile had the highest physical activity participation. The physical activity differences observed between profiles remained stable over 3 years and the authors concluded that physical literacy plays an important role in physical activity participation over time in children6; however, the current study is the first to assess the opposite relationship that physical activity patterns from preschool to school-age are associated with school-age physical literacy. Despite lower physical activity participation across all timepoints and lower PLAYfun scores, girls displayed similar PLAYself and PLAYparent scores to boys. PLAYself and PLAYparent scores reflect the motivation, confidence, knowledge, understanding, and valuing physical activity components of physical literacy.22 As such, the lower physical activity levels in girls may not be associated with lower confidence and motivation, but may reflect the physical activity opportunities provided to them. For example, previous work identified that circus arts-based physical education versus traditional physical education may help close the gender gap in physical literacy of school-age children, as the circus arts curriculum was exciting and enjoyable to students of all genders.36 Further work should continue to examine the similarities and differences in physical literacy between boys and girls, and particularly how girls and boys respond to physical activity programming or opportunities that are designed to meet the preferences of children of all genders and target all domains of physical literacy (physical competence, confidence and motivation, and knowledge and understanding). Overall, we observed that longitudinal physical activity trajectory groups from preschool to school-age were associated with school-age physical literacy, and physical activity patterns over time may play an important role in the development of physical literacy in both boys and girls. However, without assessments of physical literacy over time, we cannot confirm that the opposite (physical literacy impacts physical activity) is true as well.
Several limitations in this study should be considered. Some data were missing from participants who missed study visits or did not meet our accelerometry wear-time criteria; however, overall missing data were minimal (<10%). We employed Evenson accelerometer cut points across our entire sample to assess MVPA, despite these cut points being validated in 5-to 8-year-olds.21 Nevertheless, it was more appropriate to use the same cut points across all timepoints than to apply different cut points to the preschool and school-age data. It is recommended that the Evenson cut points be used to estimate physical activity in 5-to 15-year-olds, an age range which encompasses most of our sample.37 Our sample size limited our ability to generate separate trajectories for boys and girls. The high physical activity groups were relatively small and future research with a larger sample is needed. The physical literacy composite score was included to reflect the multiple elements of physical literacy,1 but psychometric properties of this measure have not been investigated. Group-based trajectory modeling is only specific to the study sample and not all group members perfectly follow a group’s trajectory. It cannot be assumed that all children will follow one of the distinct trajectory groups we identified based on our sample.31 Finally, we were not able to analyze if longitudinal changes in physical literacy are associated with physical activity because physical literacy was only assessed at the final timepoint. Further investigation of these possibly reciprocal relationships over time is warranted. Despite these limitations, a major strength of this study was the longitudinal study design and device assessments of physical activity across 6 timepoints. Physical activity was expressed as total volume of physical activity and MVPA to increase the generalizability of these findings to other studies. Finally, group-based trajectory analysis, an emerging method to assess data longitudinally, was used to identify the subgroups of participants who followed distinct physical activity trajectories.
In conclusion, this study identified distinct patterns of physical activity and trajectory group membership was associated with school-age physical literacy, such that children with the lowest physical activity over time had the lowest school-age physical literacy. Findings suggest that school-age physical literacy may be fostered with higher physical activity participation in early childhood. Future work is needed to determine how physical literacy changes in response to changes in physical activity over time, and how other measures of physical activity (such as light physical activity) change over time and their relationship with physical literacy.
Acknowledgments
This study would not have been possible without the ongoing support and contributions of the members of the Child Health & Exercise Medicine Program. Finally, the authors would like to thank the many participants and their families who contributed to the HOPP and SKIP studies. This research was funded by Canadian Institutes for Health Research (CIHR Award#: MOP 102560 and 137026) and the North American Society for Pediatric Exercise Medicine Marco Cabrera Student Research Award Program. B.W.T. is supported by a Tier II Canada Research Chair in Child Health and Exercise Medicine. H.A.T.C. was supported by an Ontario Graduate Scholarship.
References
- 1.↑
International Physical Literacy Association. Physical Literacy. Published 2014. Accessed April 19, 2018. https://www.physical-literacy.org.uk/
- 2.↑
Cairney J, Dudley D, Kwan M, Bulten R, Kriellaars D. Physical literacy, physical activity and health: toward an evidence-informed conceptual model. Sports Med. 2019;49(3):371–383. PubMed ID: 30747375 doi:10.1007/s40279-019-01063-3
- 3.↑
Caldwell HAT, Di Cristofaro NA, Cairney J, Bray SR, Macdonald MJ, Timmons BW. Physical literacy, physical activity, and health indicators in school-age children. Int J Environ Res Public Health. 2020;17(15):5367. doi:10.3390/ijerph17155367
- 4.↑
Stearns JA, Wohlers B, McHugh TLF, Kuzik N, Spence JC. Reliability and validity of the PLAYfun tool with children and youth in Northern Canada. Meas Phys Educ Exerc Sci. 2018;23(1):47–57. doi:10.1080/1091367X.2018.1500368
- 5.↑
Bremer E, Graham JD, Bedard C, Rodriguez C, Kriellaars D, Cairney J. The association between PLAYfun and physical activity: a convergent validation study. Res Q Exerc Sport. 2019;91(2):179–187. PubMed ID: 31617795 doi:10.1080/02701367.2019.1652723
- 6.↑
Brown DMY, Dudley DA, Cairney J. Physical literacy profiles are associated with differences in children’s physical activity participation: a latent profile analysis approach. J Sci Med Sport. 2020;23(11):1062–1067. PubMed ID: 32475780 doi:10.1016/j.jsams.2020.05.007
- 7.↑
Hnatiuk JA, Lamb KE, Ridgers ND, Salmon J, Hesketh KD. Changes in volume and bouts of physical activity and sedentary time across early childhood: a longitudinal study. Int J Behav Nutr Phys Act. 2019;16(1):1–9. doi:10.1186/s12966-019-0805-6
- 8.↑
Farooq MA, Martin A, Janssen X, et al. Longitudinal changes in moderate-to-vigorous-intensity physical activity in children and adolescents: a systematic review and meta-analysis. Obes Rev. 2020;21(1):1–15. doi:10.1111/obr.12953
- 9.↑
Nagin DS, Odgers CL. Group-based trajectory modeling in clinical research. Annu Rev Clin Psychol. 2010;6(1):109–138. doi:10.1146/annurev.clinpsy.121208.131413
- 10.↑
Pate RR, Schenkelberg MA, Dowda M, McIver KL. Group-based physical activity trajectories in children transitioning from elementary to high school. BMC Public Health. 2019;19(1):1–7. doi:10.1186/s12889-019-6630-7
- 11.↑
Kwon S, Janz KF, Letuchy EM, Burns TL, Levy SM. Developmental trajectories of physical activity, sports, and television viewing during childhood to young adulthood: Iowa bone development study. JAMA Pediatr. 2015;169(7):666–672. PubMed ID: 25984811 doi:10.1001/jamapediatrics.2015.0327
- 12.↑
Caldwell HAT, Proudfoot NA, King-Dowling S, Di Cristofaro NA, Cairney J, Timmons BW. Tracking of physical activity and fitness during the early years. Appl Physiol Nutr Metab. 2016;41(5):504–510. PubMed ID: 27045869 doi:10.1139/apnm-2015-0338
- 13.↑
Cooper AR, Goodman A, Page AS, et al. Objectively measured physical activity and sedentary time in youth: the International children’s accelerometry database (ICAD). Int J Behav Nutr Phys Act. 2015;12(1):1–10. doi:10.1186/s12966-015-0274-5
- 14.↑
Taylor RW, Williams SM, Farmer VL, Taylor BJ. Changes in physical activity over time in young children: a longitudinal study using accelerometers. PLoS One. 2013;8(11):e81567. PubMed ID: 24282607 doi:10.1371/journal.pone.0081567
- 15.↑
Timmons BW, Proudfoot NA, MacDonald MJ, Bray SR, Cairney J. The health outcomes and physical activity in preschoolers (HOPP) study: rationale and design. BMC Public Health. 2012;12(1):284. doi:10.1186/1471-2458-12-284
- 16.↑
Tremblay MS, Carson V, Chaput J-P, et al. Canadian 24-hour movement guidelines for children and youth: an integration of physical activity, sedentary behaviour, and sleep. Appl Physiol Nutr Metab. 2016;41(6 suppl 2):S311–S327. doi:10.1139/apnm-2016-0151
- 17.↑
Tremblay MS, Chaput J-P, Adamo KB, et al. Canadian 24-hour movement guidelines for the early years (0–4 years): an integration of physical activity, sedentary behaviour, and sleep. BMC Public Health. 2017;17(suppl 5):874. doi:10.1186/s12889-017-4859-6
- 18.↑
Proudfoot NA, King-Dowling S, Cairney J, Bray SR, MacDonald MJ, Timmons BW. Physical activity and trajectories of cardiovascular health indicators during early childhood. Pediatrics. 2019;144(1):e20182242. doi:10.1542/peds.2018-2242
- 19.↑
Pfeiffer KA, Dowda M, McIver KL, Pate RR. Factors related to objectively measured physical activity in preschool children. Pediatr Exerc Sci. 2009;21(2):196. PubMed ID: 19556625 doi:10.1123/pes.21.2.196
- 20.↑
Rich C, Geraci M, Griffiths L, Sera F, Dezateux C, Cortina-Borja M. Quality control methods in accelerometer data processing: defining minimum wear time. PLoS One. 2013;8(6):e67206. PubMed ID: 23826236 doi:10.1371/journal.pone.0067206
- 21.↑
Evenson KR, Catellier DJ, Gill K, Ondrak KS, McMurray RG. Calibration of two objective measures of physical activity for children. J Sports Sci. 2008;26(14):1557–1565. PubMed ID: 18949660 doi:10.1080/02640410802334196
- 22.↑
Sport for Life. Physical Literacy Assessment for Youth. Published 2013. Accessed February 22, 2017. http://play.physicalliteracy.ca/what-play
- 23.↑
Caldwell HAT, Di Cristofaro NA, Cairney J, Bray SR, Timmons BW. Measurement properties of the Physical Literacy Assessment for Youth (PLAY) Tools. Appl Physiol Nutr Metab. 2020;46(6):571–578. doi:10.1139/apnm-2020-0648
- 24.↑
Cairney J, Veldhuizen S, Graham JD, et al. A construct validation study of PLAYfun. Med Sci Sport Exerc. 2018;50(4):855–862. doi:10.1249/MSS.0000000000001494
- 25.↑
Kuczmarski R, Ogden C. 2000 CDC growth charts for the United States: Methods and development. In: Vital Health Statistics. Vol. 246. 2002:1–190. Accessed December 3, 2014. http://europepmc.org/abstract/MED/12043359
- 26.↑
Mirwald RL, Baxter-Jones ADG, Bailey DA, Beunen GP. An assessment of maturity from anthropometric measurements. Med Sci Sports Exerc. 2002;34(4):689–694. Accessed June 29, 2016. http://www.ncbi.nlm.nih.gov/pubmed/11932580
- 27.↑
Jones BL, Nagin DS. A stata plugin for estimating group-based trajectory models. Carnegie Mellon Univ. 2012:1–29. doi:10.1184/r1/6470963.v1
- 28.↑
Nagin DS, Jones BL, Passos VL, Tremblay RE. Group-based multi-trajectory modeling. Stat Methods Med Res. 2018;27(7):2015–2023. PubMed ID: 29846144 doi:10.1177/0962280216673085
- 29.↑
Jones BL, Nagin DS. Advances in group-based trajectory modeling and an SAS procedure for estimating them. Sociol Methods Res. 2007;35(4):542–571. doi:10.1177/0049124106292364
- 30.↑
Jones BL, Nagin DS, Roeder K. A SAS procedure based on mixture models for estimating developmental trajectories. Sociol Methods Res. 2001;29(3):374–393. doi:10.1177/0049124101029003005
- 31.↑
Nagin DS. Group-Based Modeling of Development. Cambridge, MA: Harvard University Press; 2005. doi:10.4159/9780674041318
- 32.↑
Farooq MA, Parkinson KN, Adamson AJ, et al. Timing of the decline in physical activity in childhood and adolescence: gates head Millennium Cohort Study. Br J Sports Med. 2018;52(15):1002–1006. PubMed ID: 28288966 doi:10.1136/bjsports-2016-096933
- 33.↑
Janz KF, Letuchy EM, Burns TL, Francis SL, Levy SM. Objectively measured physical activity trajectories predict adolescent bone strength: Iowa Bone Development Study. Med Sci Sports Exerc. 2015;47(10):2201–2206. PubMed ID: 25751769 doi:10.1249/MSS.0000000000000648
- 34.↑
Poitras VJ, Gray CE, Borghese MM, et al. Systematic review of the relationships between objectively measured physical activity and health indicators in school-aged children and youth. Appl Physiol Nutr Metab. 2016;41(6 Suppl. 3):S197–S239. doi:10.1139/apnm-2015-0663
- 35.↑
Warburton DER, Bredin SSD. Health benefits of physical activity. Curr Opin Cardiol. 2017;32(5):541–556. PubMed ID: 28708630 doi:10.1097/HCO.0000000000000437
- 36.↑
Kriellaars DJ, Cairney J, Bortoleto MAC, Kiez TKM, Dudley D, Aubertin P. The impact of circus arts instruction in physical education on the physical literacy of children in grades 4 and 5. J Teach Phys Educ. 2019;38(2):162–170. doi:10.1123/jtpe.2018-0269
- 37.↑
Trost SG, Loprinzi PD, Moore R, Pfeiffer KA. Comparison of accelerometer cut points for predicting activity intensity in youth. Med Sci Sport Exerc. 2011;43(7):1360–1368. doi:10.1249/MSS.0b013e318206476e
