Accepted Manuscript

Temperament is Associated with Outdoor Free Play in Young Children: A TARGetKids! Study

Julia R. Sharp, Jonathon L. Maguire, Sarah Carsley, Kawsari Abdullah, Yang Chen,Eliana M. Perrin, Patricia C. Parkin, Catherine S. Birken

PII: S1876-2859(17)30467-9

DOI: 10.1016/j.acap.2017.08.006

Reference: ACAP 1077

To appear in: Academic Pediatrics

Received Date: 9 March 2017

Revised Date: 4 August 2017

Accepted Date: 12 August 2017

Please cite this article as: Sharp JR, Maguire JL, Carsley S, Abdullah K, Chen Y, Perrin EM, ParkinPC, Birken CS, on behalf of the TARGet Kids! Collaboration, Temperament is Associated with OutdoorFree Play in Young Children: A TARGet Kids! Study, Academic Pediatrics (2017), doi: 10.1016/j.acap.2017.08.006.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service toour customers we are providing this early version of the manuscript. The manuscript will undergocopyediting, typesetting, and review of the resulting proof before it is published in its final form. Pleasenote that during the production process errors may be discovered which could affect the content, and alllegal disclaimers that apply to the journal pertain.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

1

Temperament is Associated with Outdoor Free Play in Young Children: A TARGet Kids! Study

Julia R. Sharp,a Jonathon L. Maguire,b,c,d,e Sarah Carsley,b Kawsari Abdullah,b Yang Chen,c Eliana M. Perrin,f Patricia C. Parkinb,e and Catherine S. Birken,b,c,e,g on behalf of the TARGet

Kids! Collaboration†

Affiliations: aDepartment of Post-Graduate Medical Education, Sick Kids Hospital, University of Toronto, Canada; bPaediatric Outcomes Research Team, Sick Kids Hospital, Toronto, Canada; cThe Applied Health Research Centre of the Li Ka Shing Knowledge Institute, St. Michael's Hospital, Toronto, Canada; dDepartment of Paediatrics, St. Michael's Hospital, Toronto, Canada, eDepartment of Paediatrics, Sick Kids Hospital, Toronto, Canada, fDepartment of Pediatrics, Division of Primary Care, Duke University, Durham, North Carolina; gChild Health Evaluative Studies, Sick Kids Research Institute, Toronto, Canada

†TARGet Kids! Collaboration– Co-Leads: Catherine S. Birken, Jonathon L. Maguire; Advisory Committee: Eddy Lau, Andreas Laupacis, Patricia C. Parkin, Michael Salter, Peter Szatmari, Shannon Weir; Scientific Committee: Kawsari Abdullah, Mary Aglipay, Yamna Ali, Laura N. Anderson, Imaan Bayoumi, Catherine S. Birken, Cornelia M. Borkhoff, Sarah Carsley, Shiyi Chen, Yang Chen, David W.H. Dai, Denise Darmawikarta, Cindy-Lee Dennis, Karen Eny, Stephanie Erdle, Kayla Furlong, Kanthi Kavikondala, Christine Koroshegyi, Christine Kowal, Grace Jieun Lee, Jonathon L. Maguire, Dalah Mason, Jessica Omand, Patricia C. Parkin, Navindra Persaud, Lesley Plumptre, Meta van den Heuvel, Shelley Vanderhout, Peter Wong, Weeda Zabih; Site Investigators: Murtala Abdurrahman, Barbara Anderson, Kelly Anderson, Gordon Arbess, Jillian Baker, Tony Barozzino, Sylvie Bergeron, Dimple Bhagat, Nicholas Blanchette, Gary Bloch, Joey Bonifacio, Ashna Bowry, Anne Brown, Jennifer Bugera, Caroline Calpin, Douglas Campbell, Sohail Cheema, Elaine Cheng, Brian Chisamore, Evelyn Constantin, Erin Culbert, Karoon Danayan, Paul Das, Mary Beth Derocher, Anh Do, Michael Dorey, Kathleen Doukas, Anne Egger, Allison Farber, Amy Freedman, Sloane Freeman, Keewai Fung, Sharon Gazeley, Donna Goldenberg, Charlie Guiang, Dan Ha, Shuja Hafiz, Curtis Handford, Laura Hanson, Leah Harrington, Hailey Hatch, Teresa Hughes, Sheila Jacobson, Lukasz Jagiello, Gwen Jansz, Paul Kadar, Tara Kiran, Lauren Kitney, Holly Knowles, Bruce Kwok, Sheila Lakhoo, Margarita Lam-Antoniades, Eddy Lau, Fok-Han Leung, Alan Li, Patricia Li, Jennifer Loo, Joanne Louis, Sarah Mahmoud, Roy Male, Vashti Mascoll, Rosemary Moodie, Julia Morinis, Maya Nader, Sharon Naymark, Patricia Neelands, James Owen, Jane Parry, Michael Peer, Kifi Pena, Marty Perlmutar, Navindra Persaud, Andrew Pinto, Tracy Pitt, Michelle Porepa, Vikky Qi, Nasreen Ramji, Noor Ramji, Jesleen Rana, Alana Rosenthal, Katherine Rouleau, Janet Saunderson, Rahul Saxena, Vanna Schiralli, Michael Sgro, Susan Shepherd, Barbara Smiltnieks, Cinntha Srikanthan, Carolyn Taylor, Suzanne Turner, Fatima Uddin, Joanne Vaughan, Thea Weisdorf, Sheila Wijayasinghe, Peter Wong, Anne Wormsbecker, Ethel Ying, Elizabeth Young, Michael Zajdman; Research Team: Marivic Bustos, Charmaine Camacho, Dharma Dalwadi, Thivia Jegathesan, Tarandeep Malhi, Sharon Thadani, Julia Thompson, Laurie Thompson; Applied Health Research Centre: Christopher Allen, Bryan Boodhoo, Judith Hall, Peter Juni, Gerald Lebovic, Karen Pope, Jodi Shim, Kevin Thorpe; Mount Sinai Services Laboratory: Azar Azad.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

2

Address correspondence to: Julia R. Sharp c/o Catherine S. Birken Department of Paediatrics The Hospital for Sick Children, 686 Bay Street Toronto, ON M5G 0A4, Canada Phone:416-813-4930 E-mail: julia.sharp@mail.utoronto.ca Short Title: Preschoolers' Temperament and Outdoor Play Word Count: Main body: 2834 words Abstract: 235 words Funding Source: This work was supported by the Canadian Institute for Health Research, Sick Kids Foundation, St. Michael’s Hospital Foundation, and the Academic Pediatric Association through the Resident Investigator Award grant program. Funding agencies had no role in the design and conduct of the study, collection, management, analyses or interpretation of the results of this study or in the preparation, review, or approval of the manuscript Financial Disclosure: The authors have no financial relationships relevant to this article to disclose. Conflict of Interest: The authors have no conflicts of interest to disclose.

Abbreviations: CCHS: Canadian Community Health Survey. CBQ: Childhood Behavior Questionnaire. ECBQ: Early Childhood Behavior Questionnaire.PA: physical activity.

Key Words: Physical activity, Outdoor play, Active play, Free play, Preschoolers, Temperament, Negative affectivity

What's New: Temperament is associated with obesity risk in preschoolers, but its relationship with outdoor play is unknown. In this cross-sectional analysis of a cohort, temperament was associated with outdoor playtime, and thus may be an important consideration in preschoolers' health interventions.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

3

Contributors’ Statements:

Julia R. Sharp conceptualized the study, drafted the initial manuscript, and approved the final manuscript as submitted.

Jonathon L. Maguire designed the data collection instruments, reviewed the manuscript, and approved the final manuscript.

Sarah Carsley designed the data collection instruments and helped coordinate data collection and extraction, and approved the final manuscript as.

Kawsari Abdullah conceptualized the study, reviewed and approved the final manuscript.

Yang Chen designed and completed the statistical analyses, reviewed and approved the final manuscript.

David Dai (not an author but mentioned in Acknowledgements) completed the repeat statistical analyses for the paper revision, edited the methods section, and reviewed and approved the final Design and Methods section of the manuscript.

Eliana M. Perrin helped to design the study methods, reviewed the manuscript and approved the final manuscript .

Patricia C. Parkin designed the data collection instruments, reviewed the manuscript, and approved the final manuscript.

Catherine S. Birken acted as the principal investigator, conceptualized the study, completed extensive reviews of manuscript drafts, and approved the final manuscript as submitted.

All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the final manuscript.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

4

ABSTRACT

Background and Objectives. Outdoor free play is important for preschoolers' physical activity, health and development. Certain temperamental characteristics are associated with obesity, nutrition, and sedentary behaviors in preschoolers, but the relationship between temperament and outdoor play has not been examined. This study examined whether there is an association between temperament and outdoor play in young children.

Methods. Healthy children aged 1 to 5 years recruited to The Applied Research Group for Kids (TARGetKids!), a community-based primary care research network, from July 2008 to September 2013 were included. Parent-reported child temperament using the Childhood Behavior Questionnaire. Outdoor free play and other potential confounding variables were assessed through validated questionnaires. Multivariable linear regression was used to determine the association between temperament and outdoor play, adjusted for potential confounders.

Results. There were 3393 children with data on outdoor play. The association between negative affectivity and outdoor play was moderated by sex; in boys, for every 1-point increase in negative affectivity score, mean outdoor play decreased by 4.7 minutes/day. There was no significant association in girls. Surgency was associated with outdoor play; for every 1-point increase in surgency/extraversion, outdoor play increased by 4.6 minutes per day.

Conclusions. Young children's temperamental characteristics were associated with their participation in outdoor free play. Consideration of temperament could enhance interventions and strategies to increase outdoor play in young children. Longitudinal studies are needed to elucidate the relationship between children's early temperament and physical activity.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

5

INTRODUCTION Outdoor active free play is a major source of physical activity (PA) in young children and may

be associated with decreased BMI and obesity in preschoolers.1,2 Time outdoors may also be

associated with improved sleep, attention, imaginative play, social relationships, and reduced

symptoms of disease including asthma and chronic pain.3,4 Despite these benefits, many

preschoolers do not play outdoors on a daily basis.5 The American Academy of Pediatrics

recommends Pediatricians make it a priority to advocate for increased outdoor free play in

preschoolers, as an essential part of healthy development.6

Designing effective interventions to promote outdoor play in young children requires an

understanding of the determinants of outdoor play. Previous large cohort studies have shown that

increased outdoor play may be associated with access to green space,7,8 being white,2,5,8 being

male,5,8 having multiple playmates,5 and having a mother who does not work full-time.2,5

Research, however, has largely overlooked the role of individual child characteristics, such as

temperament, on outdoor play.

Child temperament is defined as constitutionally-based differences in behavioral style that are

evident from the earliest years.9 In their well-validated measure of temperament, the Children's

Behavior Questionnaire (CBQ),9 Rothbart et al define three temperament constructs: negative

affectivity, surgency/extraversion, and effortful control. Briefly, negative affectivity reflects the

tendency to experience and express negative emotional reactions. Surgency/extraversion reflects

level of motor activity, seeking closeness with others, and positive emotional reactions to high-

intensity stimuli. Effortful control reflects self-control of attention and behavior, and positive

reactions to low-intensity stimuli. Temperament has been shown to influence obesity and related

behaviors in young children, including nutrition and screen time,10,11 with greater levels of

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

6

negative affectivity inferring increased risk.10,12,13 Given the association of negative affectivity

with health behaviors, the primary objective of this study was to determine if negative affectivity

was associated with outdoor play. The secondary objectives were to determine if the other

temperament constructs, effortful control or surgency/extraversion were associated with outdoor

play. Based on evidence that age and sex may relate to expression of temperament,14 we assessed

whether child age and sex modified the relationship of temperament and outdoor play.

DESIGN AND METHODS

The Applied Research Group for Kids (TARGetKids!) is an ongoing open longitudinal cohort

study that has recruited healthy children from 9 primary care practices in Toronto, Canada.15

Children were invited to participate at each primary care visit, using parent-completed

questionnaires. Exclusion criteria included health conditions affecting growth (e.g. cystic

fibrosis), other chronic health conditions (excluding asthma), severe developmental delay, and

inability to complete English-language questionnaires. The full details of the cohort profile have

been published seperately.15 Data for the current study were collected between July 2008 and

September 2013. The exposure variable for this study, child temperament, was measured using a

suite of age appropriate validated questionnaires, the Very Short Forms (VSF) for the Early

Childhood Behavior Questionnaire (ECBQ)16 completed at the 18-month visit, and the Children's

Behavior Questionnaire (CBQ)9,17 completed at the 3-year visit. Using the VSF-ECBQ/CBQ,

parents were asked to consider their child’s reaction in the past 6 months to 36 situations (e.g.,

“gets angry when can’t find something s/he wants to play with”) on a 7-point scale from

“extremely untrue of your child” to “extremely true of your child.” For each child, the VSF-

ECBQ/CBQ produced independent scores for three broad temperament factors: negative

affectivity, effortful control and surgency/extraversion. Each factor was determined by 12 items

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

7

chosen from the original CBQ/ECBQ, determined to have high correlation to their associated

factor and low correlation to the other two.16,17 Children with high negative affectivity show a

predisposition to react with anger, frustration and fear, are negatively affected by situations with

high stimulation, and are difficult to soothe after distress or excitement. Children with high levels

of effortful control enjoy situations involving low stimulus intensity, have the capacity to

suppress inappropriate approach responses under novel or uncertain situations, and can maintain

focus on task-related activities. Children with high levels of surgency show a predisposition to

positive emotions; they are impulsive, highly active, enjoy situations involving high stimulation,

and do not show discomfort in social situations.9,14

The psychometric properties of these questionnaires have been evaluated in US, Chinese and

Japanese populations, with similar reliability (mean 0.78) and internal consistency (0.68 to

0.93).9 Parent-report measures of temperament have been shown to have low to moderate

convergence with laboratory and observational measures of temperament,18,19,20 and both

methods show evidence of validity. 9,20

The primary outcome variable in this study was parent-reported outdoor free play time, a

measure that has been shown to correlate moderately well with accelerometry-measured PA in

TARGetKids! data and in prior studies (spearman correlation r= 0.30 to 0.33)21,22 Specifically,

parents were asked “Aside from time in daycare and preschool, on a TYPICAL WEEKDAY, how

much time does your child spend outside in unstructured free play?” using wording from the

validated Canadian Community Health Survey (CCHS),23 and similar to wording in previous

validated measures of outdoor play.21

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

8

The covariates identified a priori as potential confounders in the relationship between

temperament and outdoor play were based on a review of the literature and included child’s age,

sex, zBMI, time of year when questionnaire was completed, maternal ethnicity, maternal

education level, neighborhood income, parental employment, daycare use, and presence of

siblings.2,5,7,8 Child's age was reported in months, and sex as male or female. To provide an age-

and sex-adjusted measure of weight status, zBMI was used. Time of year was dichotomized into

two categories; cooler months (November to April) and warmer months (May to October).

Median neighborhood household income was calculated using the six-digit postal codes for each

participant to obtain the median after-tax value (using the Statistics Canada Postal Code

Conversion File and data from the 2006 Canadian Census).24 Maternal ethnicity was measured

by the question "What were the ethnic or cultural origins of your child’s ancestors [...] on the

biological mother’s side?" using 20 categories of ethnicity from the CCHS. These categories

were then categorized to 7 major groups of European, East/Southeast Asian, South Asian, Arab,

African, Latin, and Other, as previously published.25

Descriptive analyses of the main exposure, outcome variable, and all covariates at baseline were

conducted. To assess the relationship between the primary exposure variable, negative

affectivity, and the outcome variable, outdoor play, we used a multiple linear regression model

adjusted for the covariates described above. We repeated this analysis with secondary exposure

variables surgency/extraversion and effortful control. Based on previous research, children’s age

and sex were thought to potentially modify the relationship between the temperament and time in

outdoor play, therefore they were included as interaction terms. To minimize model selection

and multiple testing, all interaction terms were tested simultaneously. The a priori plan was to

eliminate the interaction terms if no interaction terms were significant. If an interaction term was

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

9

felt to be significant, stratified models were then run to determine the size of the effect.

Regression coefficients, confidence intervals, and p-values were generated non-parametrically

using 5000 bootstrapped samples to avoid incorrect parametric inferences due to non-normal

model residuals.26 To assess the linearity of associations, model residuals were plotted against

the temperament variables, and showed no patterns of non-linearity. Because data were expected

to be missing at random, multiple imputation was used to reduce bias from performing complete

case analysis; none of the variables had more than 13% missing data. Fifty data sets were

computer generated and the results were pooled.27 All statistical analyses were conducted with

the use of SAS 9.3 for Windows and R version 3.3.1.

RESULTS

Baseline characteristics are shown in Table 1. There were 5161 children aged 1-5 years in the

TARGetKids! research cohort at the time of data extraction. Our sample included the 3393

children for whom data on outdoor play was available. Mean time outdoors on a typical weekday

was 61 minutes, with no significant differences by age group or sex. Mean scores for negative

affectivity were 3.6 (SD 0.9, range 1 to 7), surgency 4.8 (SD 0.8, range 1.7 to 7), and effortful

control 5.4 (SD 0.8, range 1.4 to 7).

In the multivariate linear regression analysis examining the relationship between negative

affectivity and outdoor play, the interaction with sex was approaching significance (95% CI -

0.61, 7.35, p = 0.06). The regression analysis was therefore run separately for girls and boys.

There was a statistically significant association between negative affectivity and outdoor play

among boys but not among girls (Table 2). In boys, for every 1-point increase in negative

affectivity score, mean outdoor playtime decreased by 4.7 (95% CI -7.92, -1.52, p = 0.002)

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

10

minutes per day (Table 3). There was no significant interaction with age (95% CI -0.08, 0.12, p =

0.792). Covariates associated with outdoor play included time of year, with 28.4 (95% CI 22.84,

33.91, p < 0.001) more minutes of daily outdoor play reported in the warmer months. There was

also a statistically significant effect for most categories of ethnicity, with all ethnicities having

decreased time in outdoor play compared with Europeans. Hours in daycare were also negatively

associated with outdoor play time; for every hour spent in daycare per week, mean daily outdoor

play time decreased by 0.5 (95% CI -0.67, -0.33, p <0.001) minutes per day. Of note, there was

no significant effect of zBMI on outdoor play. In girls, there was a similar significant association

of outdoor play with season and daycare attendance, but not with ethnicity.

Examining the relationship between surgency/extraversion and outdoor play, for every 1-point

increase in surgency/extraversion score, mean outdoor playtime increased by 4.6 minutes per day

(p = 0.002, 95% CI 2.73,7.70).There were no statistically significant interactions with age or sex

(Table 2). There was no statistically significant association found when examining the

relationship between effortful control and outdoor play (Table 2).

DISCUSSION

In this sample of healthy preschool children recruited from primary care practices, higher

negative affectivity was associated with less outdoor play in boys, and higher

surgency/extraversion was associated with more outdoor play in all children. Best practice

guidelines from the National Association of Sports and Physical Education28 recommend that

children 0 - 5 years old accumulate at least 60 minutes of unstructured physical activity daily,

including time outdoors. Similarly, the Canadian Society for Exercise Physiology29 recommends

children 0 - 4 years old accumulate at least 180 minutes daily of PA at any level, including play

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

11

outdoors, with the goal of progressing toward at least 60 minutes of energetic play by age 5. In

looking at our results, taking the example of two children with negative affectivity scores of 2

and 6, a 4.7 minute difference per unit change in temperament score could translate to a

difference of 19 minutes of outdoor play per day. Thus, these differences in outdoor play time

between children of varying temperaments could make a significant impact on meeting daily

goals for minimum active play in preschoolers.

A recent Canadian study of 216 preschoolers in childcare centres concluded that none of the

studied constructs of parent-reported temperament were predictive of accelerometry-measured

PA.30 This study is difficult to compare, however, as they used a different instrument with

different domains to measure temperament, and activity data were collected during childcare

hours only, a time that has been suggested to be quite sedentary for preschoolers.31

There has been increasing attention paid to how children’s temperament contributes to their

health; different temperamental characteristics may act as either protective or risk factors for

various childhood health and behavior outcomes,32,33 A recent systematic review which

summarized 18 studies on temperament and weight outcomes concluded that greater levels of

negative affectivity in early life were associated with increased risk of obesity in infants and

young children.10 Postulated mechanisms include increased intake of sweet food and drinks,12

decreased fruit and vegetable intake,12 and increased television exposure.7

In the present study, there was an association demonstrated between higher negative affectivity

and reduced outdoor play in boys. The following are hypothesized mechanisms that may support

these findings: discomfort with new situations, social gatherings, and high sensory stimuli may

lead boys with high negative affect to seek out more low-stimulus activities, a tendency toward

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

12

fear and difficult “soothability” may lead parents to keep these children indoors to avoid negative

reactions, time indoors could be used as punishment for negative behaviors, or reverse causality,

where boys who have less opportunities for outdoor play may be characterized by their parents

as having higher negative affectivity. Temperament has been shown to be a stable construct, so

should not be significantly affected by short-term exposures.14

Sex differences in the relationship of temperament and preschoolers' health behaviors have been

demonstrated in previous research.34 In the current study, while there was a significant

association of negative affectivity with outdoor play in males, there was no association in

females. This raises the possibility that caregivers' behaviors in response to a child's

temperament may be subject to sex differences. Preschool aged boys have been shown to

participate in more PA than girls, 35,36 but there is inconsistent evidence about whether similar

sex differences exist for participation in outdoor play,2,5,7,8 or overall PA. A sex difference in

overall outdoor play was also not identified in the current study.

There has been limited research to assess the association of surgency/extraversion with PA or

health-related outcomes in children. In a recent study, a temperament measure related to high

activity level was positively associated with accelerometry-measured PA, and negatively

associated with sedentary behavior.37 In our study, more highly surgent/extraverted children

spent more time in outdoor free play. Mechanisms that may explain this include: a tendency

toward enjoyment of high-stimulus activity leads these children to request outdoor activities

where they can run and jump, or their enjoyment of high-stimulus and gross motor activity leads

parents to send these children outdoors, or reverse causality, where increased time outdoors leads

to higher expression of surgency/extraversion. The association of outdoor play with season and

ethnicity are supported by previous studies in preschoolers.2,5,8,21 The association of daycare

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

13

attendance with decreased outdoor play outside of daycare demonstrated in our study was

statistically significant, though given the small magnitude of the effect the clinical significance is

questionable. Still, the direction of this effect is consistent with previous research.5,38

Hypothesized mechanisms that may explain this include: children in daycare are more likely to

have working parents who do not provide supervision for outdoor playtime, or parents’

perceptions that their child participated in outdoor play during daycare hours results in a lower

priority for time outdoors at home. Although these effects are interesting, the estimates are

exploratory in nature and this model was not designed to test the relationship of ethnicity,

season, or daycare attendance with outdoor play. This needs to be explored further in a separate

analysis.

Study strengths include a relatively large sample size and detailed clinical data, which allowed

inclusion of numerous potential confounding variables, including time of year and daycare

attendance, which have only been examined in a few previous studies.2,5,21

There are some limitations to our study that warrant careful consideration. Firstly, our data was

cross-sectional, and therefore does not allow us to infer causality. Our outcome measure of

outdoor play was assessed by parent-report, rather than direct measurement. Parent-reported

outdoor play has been shown to correlate moderately with accelerometry-measured PA in

preschoolers,,21 including our own data.22 In this study, we did not have data to assess parenting

strategies around outdoor play. There is some evidence to suggest that higher levels of parental

support for PA are associated with increased outdoor playtime in preschoolers.8 We did not have

data to assess the effect of built environment and access to green space on outdoor play, as

demonstrated in some studies..2,7,8 We also did not have data on parental concern for

neighborhood safety, though this has not been shown to correlate with young children's outdoor

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

14

play in previous studies.2,5 Although we hypothesized that parent perception of outdoor time at

daycare may be associated with time spent in outdoor play at home, we did not have data to

assess this in our sample. Finally, it is possible that other unmeasured confounding factors may

account for the associations.

It is important to consider how understanding the association of child temperament with outdoor

play might be applied to promote outdoor activity in young children. There is evidence for

parenting interventions that target difficult behaviors in children.39 Understanding how to adapt

and modify interventions for a "good fit" with the child’s temperamental characteristics,14,40

while considering other factors previously demonstrated to be associated with outdoor play time

in this age group, may result in more elegant models of health intervention that could be

developed and tested.

CONCLUSION

Our results highlight the association of preschoolers’ temperamental characteristics with outdoor

free play. Consideration of child temperament could help to enhance interventions and promote

parenting strategies aimed at increasing outdoor free play in young children. Child health

practitioners are uniquely positioned to help parents integrate temperament-based approaches for

health-related behaviors such as outdoor play. Further research is required to explore the causal

relationship between temperament and outdoor play using longitudinal data in young children.

ACKNOWLEDGEMENTS

The authors acknowledge David Dai for his substantial contributions to the final statistical

analyses and input for the Design and Methods section.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

15

The authors thank all participating families for their time and involvement in TARGetKids! and

are grateful to all practitioners who are currently involved in the TARGetKids! research network.

We would like to acknowledge the funding support provided by the Academic Pediatric

Association Resident Investigator Award 2014.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

16

REFERENCES

1) Ansari A, Pettit K, Gershoff E. Combating obesity in Head Start: Outdoor play and change

in children's body mass index. J Dev Behav Pediatr. 2015 Oct;36(8):605-12. 2) Kimbro RT, Brooks-Gunn J, McLanahan S. Links among neighborhood characteristics,

weight status, outdoor play, and television watching. Soc Sci Med. 2011 Mar;72(5):668-76.

3) McCurdy LE, Winterbottom KE, Mehta SS, et al. Using nature and outdoor activity to improve children's health. Curr Probl Pediatr Adolesc Health Care. 2010 May;40(5):102-17.

4) Xu H, Wen LM, Hardy LL et al. Associations of outdoor play and screen time with nocturnal sleep duration and pattern among young children. Acta Paediatr. 2016 Mar;105(3):297-303.

5) Tandon PS, Zhou C, Christakis DA. Frequency of parent-supervised outdoor play of US preschool-aged children. Arch Pediatr Adolesc Med. 2012 Aug;166(8):707-12.

6) Ginsberg KR. The importance of play in promoting healthy child development and maintaining strong parent-child bonds. Pediatr. 2007 Jan; 119(1): 182-91.

7) Marino AJ, Fletcher EN, Whitaker RC, et al. Amount and environmental predictors of outdoor playtime at home and school: a cross-sectional analysis of a national sample of preschool-aged children attending Head Start. Health Place. 2012 Nov;18(6):1224-30.

8) Grigsby-Toussaint DS, Chi SH, Fiese BH; STRONG Kids. Neighborhood greenness and outdoor physical activity among preschoolers. Int J Health Geogr. 2011 Dec 14;10:66.

9) Rothbart MK, Ahadi SA, Hershey KL, et al. Investigations of temperament at 3-7 years: The Children's Behavior Questionnaire. Child Dev. 2001 Sep-Oct;72(5):1394-408.

10) Anzman-Frasca S, Stifter CA, Birch LL. Temperament and childhood obesity risk: a review of the literature. J Dev Behav Pediatr. 2012;33:732–45.

11) Bergmeier H, Skouteris H, Horwood S, et al. Associations between child temperament, maternal feeding practices and child body mass index during the preschool years: a systematic review of the literature. Obes Rev. 2014 Jan;15(1):9-18.

12) Vollrath ME, Stene-Larsen K, Tonstad S, et al. Associations between temperament at age 1.5 years and obesogenic diet at ages 3 and 7 years. J Dev Behav Pediatr. 2012 Nov;33(9):721-7.

13) Thompson AL, Adair LS, Bentley ME. Maternal characteristics and perception of temperament associated with infant TV exposure. Pediatrics. 2013 Feb;131(2):e390-7.

14) Rothbart MK. Becoming Who We Are: Temperament and Personality in Development. New York, NY: Guilford; 2011.

15) Carsley S, Borkhoff CM, Maguire JL, et al. Cohort Profile: The Applied Research Group for Kids (TARGet Kids!).on behalf of the TARGet Kids! Collaboration. Int J Epidemiol.2015 Jun;44(3):776-88.

16) Putnam SP, Jacobs J, Gartstein MA, et al. Development and assessment of short and very short forms of the Early Childhood Behavior Questionnaire. Poster presented at International Conference on infant studies, Baltimore, MD, March 2010. <https://research.bodoin.edu/rothbart-temperament-questionnaires/files/2016/09/ecbq_article.pdf>

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

17

17) Very short form reference: Putnam SP, Rothbart MK. Development of short and very short

forms of the Children's Behavior Questionnaire.J Pers Assess. 2006 Aug;87(1):102-12. 18) Durbin CE, Hayden EP, Klein DN, et al. Stability of laboratory-assessed temperamental

emotionality traits from ages 3 to 7. Emotion. 2007; 7(2):388-399. 19) Majdandzić M, van den Boom DC. Multimethod longitudinal assessment of temperament

in early childhood. J Pers. 2007 Feb;75(1):121-68. 20) Gagne JR, Van Hulle CA, Aksan N, et al. Deriving childhood temperament measures from

emotion-eliciting behavioral episodes: scale construction and initial validation. Psychol Assess. 2011 Jun;23(2):337-53.

21) Burdette HL, Whitaker RC, Daniels SR. Parental report of outdoor playtime as a measure of physical activity in preschool-aged children. Arch Pediatr Adolesc Med. 2004 Apr;158(4):353-7.

22) Sarker H, Anderson LN, Borkhoff CM, et al; TARGet Kids Collaboration.Validation of parent-reported physical activity and sedentary time by accelerometry in young children. BMC Res Notes. 2015 Nov 30;8:735.

23) Statistics Canada. Canadian Community Health Survey. <http://www23.statcan.gc.ca/imdb-bmdi/instrument/3226_Q1_V2-eng.pdf.> Published January 2003.Updated 22 November 2012. Last accessed 30 July 2013.

24) Hanley GE, Morgan S. On the validity of area-based income measures to proxy household income.BMC Health Serv Res. 2008 Apr 10;8:79.

25) Ormand JA, Carsley S, Darling PB, et al. Evaluating the accuracy of a geographic closed-ended approach to ethnicity measurement, a practical alternative. Ann Epidemiol. 2014 Apr;24(4):246-53.

26) Efron B, TibshiraniRJ. An Introduction to the Bootstrap. New York: Chapman and Hall; 1993.

27) Rubin D. Multiple imputation after 18+ years. J Am Stat Assoc. 1996;91:473–89. 28) National Association for Sport and Physical Education. Active Start: A Statement of

Physical Activity Guidelines for Children from Birth to Age 5, 2nd Edition. Sewickley, PA: American Alliance for Health, Physical Education, Recreation, and Dance; 2009.

29) Canadian Society for Exercise Physiology. Canadian Physical Activity Guidelines for the Early Years Ages 0 - 4. < http://www.csep.ca/CMFiles/Guidelines/PAGuidelinesBackgrounder_EY2012_E.pdf>. Published 2012. Last accessed July 20, 2017.

30) Irwin JD et al. Temperament and objectively measured physical activity and sedentary time among Canadian preschoolers. Prev Med Rep. 2015 Jul 22;2:598-601.

31) Vanderloo LM, Tucker P, Johnson AM, et al. The influence of centre-based childcare on preschoolers' physical activity levels: a cross-sectional study. Int J Environ Res Public Health. 2014 Feb 5;11(2):1794-802

32) Hayes MJ, Parker KG, Sallinen B, et al. Bedsharing, temperament, and sleep disturbance in early childhood. Sleep. 2001 Sep 15;24(6):657-62.

33) Aminabadi NA, Ghoreishizadeh A, Ghoreishizadeh M, et al. Can child temperament be related to early childhood caries? Caries Res. 2014;48(1):3-12.

34) Faith MS, Hittner JB. Infant temperament and eating style predict change in standardized weight status and obesity risk at 6 years of age. Int J Obes (Lond). 2010 Oct;34(10):1515-23.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

18

35) Brasholt M, Chawes B, Kreiner-Møller E, et al. Objective assessment of levels and

patterns of physical activity in preschool children. Pediatr Res. 2013 Sep;74(3):333-8 36) Hinkley T, Crawford D, Salmon J, et al. Preschool children and physical activity: A review

of correlates. Am J Prev Med. 2008 May;34(5):435-441. 37) Schmutz EA, Leeger-Aschmann CS, Radtke T, et al. Correlates of preschool children's

objectively measured physical activity and sedentary behavior: a cross-sectional analysis of the SPLASHY study. Int J Behav Nutr Phys Act. 2017 Jan 5;14(1):1.

38) Carsley S, Liang LY, Chen Y, et al. The impact of daycare attendance on physical activity outside of daycare in young children. J Public Health. 2016 Feb 9 (epub).

39) Perrin EC, Sheldrick RC, McMenamy JM, et al.Improving parenting skills for families of young children in pediatric settings: a randomized clinical trial. JAMA Pediatr. 2014 Jan;168(1):16-24.

40) McClowry SG, Rodriguez ET, Koslowitz R. Temperament-Based Intervention: Re-examining Goodness of Fit. Eur J Dev Sci. 2008 Jun;2(1-2):120-135.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table 1. Baseline Characteristics

Characteristics Overall

n = 3373*

Male

n = 1762*

Female

n = 1631*

Age, months (mean, SD) 33.5 (17.9) 33.33 (18.14) 33.69 (17.72)

Typical weekday outdoor free play, minutes

(mean, 95% CI)

Age < 2 years

Age ≥2 years

59.4 (58.1, 64.0)

62.4 (60.0, 65.0)

59.80 (59.27)

62.64 (60.23)

59.00 (67.77)

62.30 (56.73)

Temperament score, of 7 (mean, SD)

Negative affectivity

Surgency/extraversion

Effortful control

3.6 (0.9, )

4.8 (0.8)

5.4 (0.8)

3.57 (0.94)

4.87 (0.79)

5.29 (0.75)

3.65 (0.92)

4.62 (0.82)

5.52 (0.73)

Maternal ethnicity (N, %)

African

Arab

East/SE Asian

European/Caucasian

Latin

South Asian

Other

143 (4.4)

61 (1.8)

341 (10.4)

2262 (68.7)

108 (3.3)

195 (5.9)

181 (5.5)

71 ( 4.1)

33 ( 1.9)

177 (10.3)

1172 (68.4)

61 ( 3.6)

106 ( 6.2)

93 ( 5.4)

72 ( 4.6)

28 ( 1.8)

164 (10.4)

1090 (69.1)

47 ( 3.0)

89 ( 5.6)

88 ( 5.6)

Maternal employment (N, %)

Full-time

Part-time

Unemployed

1571 (49.1)

555 (17.4)

1072 (33.5)

812 (48.8)

291 (17.5)

560 (33.7)

759 (49.4)

264 (17.2)

512 (33.4)

Maternal education (N, %)

High school and public school

College/University+

326 (9.9)

2963 (90.1)

180 (10.5)

1537 (89.5)

146 ( 9.3)

1426 (90.7)

Neighborhood income (mean $, SD) 59,690 (26,240) 58,640 (24,560) 60,830 (27,930)

Last week Hours of daycare (mean hours, SD) 10.8 (16.4) 10.77 (16.26) 10.73 (16.50)

Presence of siblings (N, % “yes”) 2198 (65.4) 1130 (64.6) 1068 (66.1)

Season (N, %)

Cooler months (October-April)

Warmer months (May-September)

1786 (52.6)

1607 (47.4)

924 (52.4)

838 (47.6)

862 (52.9)

769 (47.1)

zBMI (mean, SD) 0.13 (1.14) 0.25 (1.12) 0.15 (1.08) *Missing data: surgency n = 429 (12.6%), negative affectivity n = 430 (12.7%), effortful control n = 429 (12.6%), maternal ethnicity n = 102 (3.0%), maternal employment n = 195 (5.7%), maternal education n = 104 (3.1%), neighborhood income n = 242 (7.1%), hours of daycare n = 128 (3.8%), presence of siblings n = 29 (0.9%), season n = 0, zBMI n =136 (4%).

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table 2: Adjusted estimate of association between temperamental constructs and outdoor play in preschool children

Temperament Stratified By

Adjusted Estimate

(min/day)

95% CI*

Empirical p

Negative affectivity**

Female sex -0.23 -3.93, 3.01 0.966

Male sex -4.71 -7.92, -1.52 0.002

Surgency/extraversion† All 4.60 2.73, 7.70 0.003

Effortful control† All 1.83 -1.13, 5.02 0.244

*Bolded values are statistically significant p<0.05 **adjusted for age, maternal employment, maternal ethnicity, maternal education, neighbourhood income, hours of daycare per week, number of siblings, time of year, BMI †adjusted for age, sex, maternal employment, maternal ethnicity, maternal education, neighbourhood income, hours of daycare per week, number of siblings, time of year, BMI

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table 3. Multivariable Regression: Effect of Negative Affectivity on Outdoor Play in Boys Exposure variable (Reference) Adjusted Estimate

(min/day)

95% CI* Empirical p

Negative affectivity -4.71 -7.92, -1.52 0.002

Age in months (Mean) 0.16 0.01, 0.31 0.044

Maternal employment (Full Time)

Part-time

Unemployed

2.92

5.71

-4.30, 10.43

-0.48, 13.29

0.437

0.120

Maternal ethnicity (European/Caucasian)

African

Arab

East/Southeast Asian

Latin

South Asian

Other

-12.06

-30.04

-15.72

-22.25

-16.96

-10.08

-25.50, 5.94

-42.23, -16.95

-22.86, -7.77

-32.33, -12.62

-26.74, -4.47

-19.08, -0.21

0.130

<0.001

<0.001

<0.001

0.005

0.041

Maternal education (College/University)

Public school or high school

4.34

-4.89, 15.55

0.442

Neighborhood income (per $10,000) -0.06 -0.17, 0.04 0.260

Last week hours daycare (Mean) -0.50 -0.67, -0.33 <0.001

Number of siblings (Mean) -1.59 -4.83, 1.81 0.351

Warmer months 28.40 22.84, 33.91 <0.001

zBMI 0.35 -1.86, 3.12 0.976

*Bolded values are statistically significant p < 0.05

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Appendix Figure 1. Histogram of Score Distributions for Negative Affectivity

Negative Affectivity Score (1 - 7)

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 2. Histogram of Score Distributions for Surgency/Extraversion

Surgency/Extraversion Score (1 - 7)

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 3. Histogram of Score Distributions for Effortful Control

Effortful Control Score (1 - 7)