Article Text

Download PDFPDF

Do parents' and children's concerns about sports safety and injury risk relate to how much physical activity children do?
  1. Amanda Telford1,
  2. Caroline F Finch2,
  3. Lisa Barnett3,
  4. Gavin Abbott4,
  5. Jo Salmon4
  1. 1School of Medical Sciences, Discipline of Exercise Sciences, RMIT University, Bundoora, Victoria, Australia
  2. 2Australian Centre for Research into Injury in Sport and its Prevention (ACRISP), Monash Injury Research Institute, Monash University, Clayton, Victoria, Australia
  3. 3School of Health and Social Development, Deakin University, Burwood, Victoria, Australia
  4. 4Centre for Physical Activity & Nutrition Research, School of Exercise & Nutrition Sciences, Deakin University, Burwood, Victoria, Australia
  1. Correspondence to Associate Professor Amanda Telford, School of Medical Sciences, Discipline of Exercise Sciences, RMIT University, PO Box 71, Bundoora, Victoria 3083, Australia; amanda.telford{at}rmit.edu.au

Abstract

Objective To see whether concerns about injury risk relate to children's physical activity (PA).

Methods Two cohorts were recruited from 19 Australian schools and assessed in 2001 (T1), 2004 (T2) and 2006 (T3). The younger (n=162) was assessed at 6, 9 and 11years old, and the older (n=259) at 11, 14 and 16 years old. At T1 and T2, parents of the younger cohort reported on fear of child being injured, and whether child would be at risk of injury if they played organised sport; the older cohort self-reported injury fear. Accelerometers assessed PA at each time point. Linear regression models examined cross-sectional associations, and also associations between T1 injury fear and risk and T2 PA, and T2 injury fear and risk and T3 PA.

Results In the younger cohort at T2 (9 years), fear and risk were both negatively associated with moderate to vigorous PA (MVPA) (β=−0.17, 95% CI −0.30 to −0.03 and β=−0.26, 95% CI −0.41 to −0.10) and also vigorous PA (VPA). Fear was also associated with moderate PA (MPA). For the older cohort at T1, injury fear was negatively associated with MVPA (β=−0.21, 95% CI −0.35 to −0.07) and also MPA and VPA. Parental perception of risk at T1 (6 years) was negatively associated with children's MPA at T2 (9 years) (β=−0.17, 95% CI −0.32 to −0.02). Sex did not moderate any association.

Conclusions Younger children and their parents need to know which sports have low injury risks. Some children may need increased confidence to participate.

.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Children's physical activity (PA) has been associated with improved bone health,1 cardiovascular function2 and enhanced self-esteem.3 It is recommended that Australian children accumulate 60 min and up to several hours of PA everyday.4 Based on self-report, the 2007 Australian Children's Nutrition and Physical Activity Survey found just 40% of 9–13 year olds and 19% of 14–16 year olds meet PA recommendations.5 Although studies using accelerometry have reported that most children accumulate at least 60 min per day,6 ,7 declines throughout adolescence are consistently reported.6 ,8 The types and nature of PA that children engage in evolve from play to more structured participation in organised PA and sport; including contact or higher-risk sports.9

One of the few potentially adverse consequences of PA is injury;10 therefore PA needs to be undertaken safely so that it carries minimal risk of injury. Preventing injuries is also important because they have significant financial costs and other burdens,11 including future restriction of PA.11–14 ports injury hospital treatment rates are highest in children aged 5–14 years,15 and the immediate impact of sports injuries is also highest in this age group.12 Prevalence of PA injury is less common among girls, but increases significantly for both sexes with increased frequency of participation in PA.16

An analysis of an Australian state 2001 child health survey found more than one-quarter of the 5876 parents discouraged or prevented their active children aged 5–12 years from playing a particular sport/PA due to injury or safety concerns.17 Parental concerns can be a barrier to children's participation in PA as parents are often the key gate keepers for young people's PA.18 Children's perception of injury risk can also be a barrier to participation in PA and influence their decisions to drop out of sport.19 ,20 In contrast, some children may also perceive little risk of sporting injury. In a study of junior cricketers, players tended to see themselves as less likely to be injured than cricketers in general.21 Therefore, children's perception of sporting injury risk may either restrict or encourage PA and this can vary as they get older.19 ,20

To develop strategies for overcoming decreased participation in PA it is important that perceptions about injury risk in the context of PA as children grow older are understood. The aim of this study was to examine: (a) parental-reported and child-reported perceptions of sports injury risk; (b) cross-sectional and longitudinal relationships between parental and children's perceptions about sports injury risk and children's participation in PA; and (c) whether any associations were moderated by sex of the child.

Methods

Analyses were based on baseline data collected in 2001 for the Children's Leisure Activities Study (CLASS), and follow-up data collected in 2004 and 2006 for the Children Living in Active Neighbourhoods (CLAN) Study. The 2001 data collection point is termed as Time 1 (T1), and the 2004 and 2006 collections are called Time 2 (T2) and Time 3 (T3), respectively. Both studies were approved by the University Ethics Committee and the Department of Education and Early Childhood Development (Victoria), parental consent was also obtained. Details of the CLASS and CLAN study design, sample and methods are described in full elsewhere.22 ,23

Sample

In 2001 at T1, families of children aged 5–6 years and 10–12 years were recruited from 19 state primary schools in high (n=10) and low (n=9) socioeconomic areas in Melbourne. Schools were selected using stratified random sampling proportionate to school size. The original cohort consisted of 294 younger children and 926 older children. Participants were included in the current study if they had relevant data at all three time points resulting in a study sample of 162 younger children and 259 older children (referred hereafter as the younger and older cohort).

Group differences between the current sample and those not followed up were tested by χ2 tests for independence or independent sample t-tests, with observed differences reported as significant at p<0.05. Among the remaining younger cohort, responding parents were more likely to be female (91% vs 82%), have a tertiary degree (49% vs 36%) and be married/de facto (9%1 vs 77%). The younger cohort did not differ from non-participants on any of the PA measures, or on T1 and T2 parental perception of child fear about injury, or T2 parental perceived injury risk of child. However, parents of children without available data reported greater concerns (mean (M)=1.8, standard deviation (SD)=0.7) about injury risk during organised sport at T1 compared with parents with data at all three time points (M=1.6, SD=0.7). Parents of the older cohort were more likely to be female (88% vs 81%), have a tertiary degree (44% vs 31%), be married/de facto (88% vs 78%) and be employed part-time (41% vs 33%). The older cohort did not differ from the non-participants on most PA measures or on T2 child-reported injury fear; however, they performed less moderate vigorous PA (M=68.1, SD=27.2) and had less fear of injury (M=1.6, SD=0.8) at T1 (M=77.4, SD=78.3; M=1.7, SD=0.9, respectively).

Measures

Parents of the younger cohort completed a questionnaire at home at T1 and T2, which requested details of parents’ marital status, highest level of education (reported as some high school, high school or equivalent, tertiary) and employment status (full time, part time, other). Parents were asked to indicate level of agreement with the following statements: My child is scared that he/she will get hurt or injured; My child would be at risk of injury if her/she played organised sport. Both items used the scale 1=Strongly agree, 2=Agree, 3=Neither, 4=Disagree, 5=Strongly disagree and 6=Don't know and were reverse coded (1=Strongly disagree, 5=Strongly agree, etc) so that higher scores indicated greater agreement, and ‘Don't know’ responses were treated as missing. The older cohort completed a self-report survey in class at school at T1, and at home at T2, that asked (on a 6-point scale) how much they agreed with the following statement: I am scared that I might get hurt if I played sport (eg, Football, netball). These scores were reverse coded as with the parental reports for the younger cohort. The 2-week test–retest reliability of one item (My child is scared that he/she will get hurt or injured) was assessed with a different sample of 253 parents (mean age=38±5.6 years; 88% women). There were 156 parents of 10–12 year olds (49% boys) and 97 parents of 5–6 year olds (64% boys). The intraclass correlation (ICC) was 0.72 for parents of 10–12 year olds and 0.66 for parents of 5–6 year olds.

At all three data collection points, all children wore an Actigraph accelerometer (MTI Manufacturing and Technology, Inc, Fort Walton Beach, Florida, USA) for 8 days during all waking hours except during aquatic activities. Accelerometers have been shown to have acceptable validity among children.24 Accelerometer data were downloaded using 60 s epochs. Non-wear time was defined as sustained 20-min period of zero counts, and total duration of these periods represented the duration the total non-wear duration.25 For a valid day children were required to have produced counts for ≥600 min representing non-missing counts for at least 80% of a standard measurement day, defined as the length of time that at least 70% of the sample wore the monitor. The average accelerometer wear time per valid day was calculated. Only children with a minimum of four valid days (including 1 weekend day) were included in analyses, consistent with that recommended as a minimum amount of time to typify children's usual activity.24 Applying a published age-specific energy expenditure prediction equation,26 the time spent being moderately active (4.0–5.99 metabolic equivalent (MET)) and vigorously active (≥6 MET) per day was calculated. The minutes that children spent in moderate-intensity (MPA), vigorous-intensity (VPA) and moderate-intensity to vigorous-intensity physical activity per day (MVPA) were averaged across the days to provide a daily estimate (min/day).

Data analysis

The PA data were positively skewed and were square root transformed. All analyses were performed in STATA 11 using the transformed PA variables; however, for ease of interpretation, descriptive statistics are presented based on the raw untransformed PA data. Independent-samples t-tests and χ2 tests for independence were used to compare the younger and older child cohorts with regard to sociodemographic characteristics. Inspection of mean plots revealed that associations between injury concerns and activity levels did not violate assumptions of linearity. Paired-sample t-tests and repeated-measures analyses of variance were used to test for differences in parent and child injury concerns and activity levels, respectively, across all time points. A series of linear regression models were conducted with the transformed T1 and T2 PA variables as outcomes and the time-corresponding perceived injury risk and fear of injury entered (separately) as predictors to examine cross-sectional associations. Subsequently, longitudinal analyses were conducted where linear regressions were used to examine associations between the T1 injury concern and the T2 PA variables, and between the T2 injury concern and T3 PA variables. All models included maternal education and child sex as covariates, and adjusted for school clustering. The longitudinal analysis also adjusted for PA at T1 for the T2 model or adjusted for PA at T2 for the T3 model. Average wear time was included as a covariate as appropriate (eg, T1 average wear time was controlled for when the outcomes were T1 activity measures). In order to determine whether any associations differed by sex of child, all regression analyses were subsequently repeated including an interaction term which was the product of sex and the injury concern predictor for each model. If the interaction was non-significant it was removed from the model and separate analyses by sex deemed unnecessary. All analyses were performed separately by age group.

Results

Approximately half were boys and had a parent with a tertiary level education. Unsurprisingly, children in the younger cohort had significantly younger parents than the older cohort. Additionally, a greater proportion of parents of the older cohort reported working full-time compared to the younger cohort (see table 1).

Table 1

Sociodemographic characteristics of younger and older cohorts at T1 (2001)

Table 2 shows mean age for both cohorts at each time point and that, for both cohorts, the average amount of time spent in moderate, vigorous and MVPA decreased from T1 to T2, and then again from T2 to T3. In terms of injury concerns, for the younger cohort, parental injury fear was found to significantly decrease from T1 to T2 (M=1.7 SD=0.8 to M=1.5 SD=0.7, p= 0.005). No changes were observed from T1 to T2 for younger cohort parental perceptions of injury risk (M=1.6 SD=0.7 to M=1.6 SD=0.9, p= 0.999), or in the older cohort's self-reported fear of being injured (M=1.6 SD=0.8 to M=1.6 SD=0.9, p= 0.763).

Table 2

Mean (SD) minutes per day participating in moderate-intensity (MPA), vigorous-intensity (VPA) and moderate-intensity to vigorous-intensity (MVPA) among younger and older cohorts in 2001 (T1), 2004 (T2) and 2006 (T3)

Table 3 shows cross-sectional associations. For the younger cohort, no associations were found between T1 injury concerns and PA. However, parental perceptions of child fear and injury risk at T2 were associated with lower VPA and MVPA at T2. Additionally, perceived risk was associated with lower moderate activity at T2. For the older cohort, fear about injury at T1 was negatively associated with MPA, VPA and MVPA at T1. No cross-sectional associations were found between fear of injury and activity at T2 for the older children. No significant interactions were found between injury concerns and sex.

Table 3

Cross-sectional associations between injury concerns and moderate (MPA), vigorous (VPA) and moderate-intensity to vigorous-intensity physical activity (MVPA)†

Table 4 shows longitudinal associations. For the younger children, parental perceptions about children's injury risk at T1 were negatively associated with moderate activity at T2. No other significant longitudinal associations were found between perceived risk/fear and activity for either of the age groups. No significant interactions were found between injury concerns and sex.

Table 4

Longitudinal associations between injury concerns and moderate (MPA), vigorous (VPA), and moderate-intensity to vigorous-intensity physical activity (MVPA)†

Discussion

This study is the first to our knowledge that has examined associations between perceived injury risk and objectively assessed PA in children across time points. While the cross-sectional associations between parental and child concerns of being injured from organised sport were significant at age 9 and 11 years, there were no associations among children aged 6 or 14 years. However, parental concerns about their child being injured playing sport when they were 6 was prospectively inversely related to children's PA levels at age 9 years; this was in spite of an overall decrease in the level of parental concern between 6 and 9 years. These findings suggest that although higher levels of parental and child concerns about injury are associated with lower overall child PA levels, this varied by age (with children between 9 and 11 years being of particular concern) and sex did not moderate this relationship.

Parents may perceive their child is more afraid of injury when young as they are yet to participate in organised sports. However, parental concerns about their 6-year-olds’ injury risks were unrelated to the child's PA cross-sectionally. The fact the older children did not report change in their perceived fear of injury over time may mean that they have had enough experience in these activities by 11 years to have already formed a view of injury in sport. In Australia over half of the children by this age will have participated in organised sport beyond school.27 However, previous studies suggest a lack of experience with negative health consequences such as injury, particularly when a child first begins playing sport and has not been exposed to injury, which may desensitise children to potential health risks.28 An Australian state survey17 further supports this, as parents of 9–12 year olds were more likely to be concerned for the safety of their children in sport than parents of 5–8 year olds.

It is also possible that there were no associations between parental injury concerns and PA for the 6-year-old children, because these children were only just beginning to be involved in organised sports. Participation rates (ABS 2011) in organised sport for Australian children increase by around 10% between 5–8 and 12–14 years (64% and 74%, respectively).27 However in the present study, by the time these children were 3 years older there were associations with all PA intensities. This may indicate that parents, who perceive their child is afraid of injury, or who became more aware of injury risk once the child has started sport participation, are discouraging their child in PA as their child ages. This explanation is supported by the longitudinal data which showed parent perception of injury risk for their 6-year-old child was negatively associated with their child's moderate PA 3 years later. The delivery of most formal sports opportunities for junior participants introduces children to the skills and physical demands of their chosen sport in a very gradual way that recognises the changing developmental stages as well as chronological age.29 This has contributed to very low, or negligible, real injury risks in the youngest participants.30 ,31

For the 11-year-old children in the older cohort, fear of injury may have hampered sport involvement. Increased emphasis on competition, higher intensity and frequency of training and competition and the focus of specialising in a sport at an early age potentially leads to increased risk of injury.20 ,32 And yet when these same children reached 14 years old, their fear of injury was no longer associated with their PA. Also for these children, there was no longitudinal relationship between their fear of injury at 11 years old and subsequent PA. This may be because children's perceived PA competence or confidence increases with age so concerns about injury is therefore not a significant predictor of participation. Another potential reason for the lack of associations between 14 and 16 years is the decline in PA; therefore, concerns about injury from being active is less likely to be an issue (ie, reverse causality). The declines in PA levels among children in this study are consistent with previous research33 ,34 as are the age-related views on injury risk.19 ,20

The small number of longitudinal associations between perceived injury risk and PA may be due to the generic nature of the perceived risk of injury survey items. Not only have previous studies shown injury perceptions are specific to particular sports, they can also be context specific, for example, cricket players perceive that they will have a higher risk of injury when fielding closer to the batter.21 It is also possible that injury concerns among parents and children might not be expressed in terms of lower levels of PA, but instead in appropriate strategies such as change of sport to one perceived as less risky, or by providing/using appropriate safety gear. It is possible too, that our questions regarding injury fear and risk could be a ‘proxy’ for other issues. For example, a parent's own fear of injury (based on prior injury experiences) may have influenced parent responses. Or alternatively, parents may use injury fear as an excuse for not letting children take part in expensive/safety-demanding sports. However, this is probably unlikely in this cohort, as for the age groups discussed in this paper, most commonly played organised sports in Australia do not involve expensive safety equipment (eg, football, soccer, netball, basketball, tennis).

In this study, the relationship between parental fear and perception of injury and their child's PA did not differ according to sex of the child. A study examining parents’ judgements of their children's risk taking (in terms of safety gear) for common play situations, also found no sex differences in any of the analyses.35 In contrast, Boufous et al found that sex of the child was the most important predictor of a parent's decision to either prevent or discourage their 5–12 year old children from participating in PA.17 However, as the authors point out, this could be because boys are often attracted to higher risk sports such as football. This is reflected in data that show that the type of PA associated with injury differs by sex.36 Siesmaa et al20 found that girls were more likely than boys to report that their parents were concerned about sport injury on their behalf.

A limitation was the modest response rate that may have reduced the statistical power of the study to detect statistically significant relationships and potentially increased the potential for type 2 errors; nevertheless, the sample for both cohorts is still a reasonable size at all time points for a study of this nature. Also, the sample did include families from varying socioeconomic backgrounds. There was the possibility of selection bias given the reported differences in certain parent and child characteristics between those who remained in the study compared with those who dropped out. Test–retest reliability was not tested for two items that may be a potential source of measurement bias; however, test–retest reliability for the item regarding parental view of child's injury fear had adequate reliability. The use of an objective measure of PA is a study strength; however, the 60 s epoch length used was not ideal. These data were first collected in 2001; although recent research has recommended that a shorter epoch than this is necessary to minimise MVPA measurement error in children,37 for consistency over time we retained the 60 s epoch length.

Perception of injury risk as a correlate of children's PA has been relatively overlooked in the PA literature to date; instead the injury or safety focus has been on road safety and stranger danger.38 This is in contrast to injury prevention literature where it has been long understood that injury risk perceptions impact on safety behaviours.11 ,17 ,19 ,35 ,39–45 The implications of our findings are that parents and children need to be educated about the correct injury risks associated with children's participation in sport and also about the range of safety measures they can adopt to reduce this risk during PA. This may then impact on their child's activity while they are in primary school in a positive way so as to encourage them to support lifelong participation in safe sport by their child.

What this study adds

  • Negative associations between parental and child perception of injury fear and risk and PA appear to be more important when children are 9–11 years.

  • As the associations in this study differed according to age, but not by sex, an opportune time to educate parents about safe sporting opportunities may be when their child starts school and organised sports activities.

Acknowledgments

Data collection for this research at T1 was supported by the Financial Markets Foundation for Children and at T2 and T3 by the National Health and Medical Research Council (ID 274309). Lisa Barnett is supported by an NHMRC early career fellowship. Jo Salmon is supported by a National Heart Foundation of Australia and sanofi-aventis Career Development Award. Caroline Finch was supported by an NHMRC Principal Research Fellowship (ID: 565900).

References

View Abstract

Footnotes

  • Contributors Each of the five authors contributed equally to this manuscript (20% contribution by each person). Their specific contributions were as follows: Associate Professor Amanda Telford and Professor Jo Salmon were involved in all phases of the research and manuscript including conceptual development, methodology, data collection, data management, results and discussion. Professor Caroline Finch was involved in the conceptual development of the research and the background, results and discussion sections of the manuscript. Gavin Abbott was primarily responsible for the data analyses, preparation of the tables and results sections of the manuscript. Dr Lisa Barnett contributed to all sections of the paper including background, methodology, results and the discussion.

  • Funding Funding provided for health-based research.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Deakin University Human Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • ▸ References to this paper are available online at http://bjsm.bmjgroup.com