Preschools’ Friendliness Toward Physical Activity Item Battery and Two Scores Developed by Mixed Methods Claudia S. Dörr, MD, Christina Bock, MSc, Dr.sc.hum., Joachim E. Fischer, MD, MSc, Freia De Bock, MD, MPH Background: Previous work suggests that many context-level characteristics of preschools are associated with increased physical activity (PA) in children. Purpose: To facilitate assessment of preschools’ friendliness toward PA, an item battery as well as comprehensive and short rating scores were developed based on mixed methods with validity and reliability tested. Methods: Organizational and contextual characteristics of 24 German preschools were assessed by direct observation and semi-structured interviews (September 2008 to April 2009; analyzed in 2012). Twenty-seven items across seven domains of friendliness toward PA were rated independently by two researchers as being limited (0); adequate (1); or extraordinarily (2) friendly toward PA. Values were summed to a comprehensive rating score value (RS-c). To increase feasibility, a short score (RS-s) was developed using an item subset identified by both itemitem correlations and positive association with moderate to vigorous PA. Validity of items and scores was tested by their association with preschoolers’ accelerometer-measured PA (n¼405) in covariate-adjusted multilevel models. Reliability was tested by inter-rater reliability coefficients. Results: Nine of 27 items were included in the RS-s. In the multilevel model, four single items and both the RS-c and RS-s were positively associated with children’s moderate to vigorous PA (RS-c: β¼0.5 minutes, p¼0.003, explained variance¼9%; RS-s: β¼1.3 minutes, po0.001, variance¼23%), but not with light activity or sedentary behavior. Inter-rater reliability was 0.80 (RS-c) and 0.88 (RS-s).
Conclusions: There is potential value for mixed methods in assessing preschools’ friendliness toward PA. (Am J Prev Med 2014;46(6):593–604) & 2014 American Journal of Preventive Medicine
Introduction
P
revious studies have shown that young children spend less than 10% of their daily time in moderate to vigorous physical activity (MVPA).1 Lack of physical activity (PA) is one of the main determinants of increasing overweight and obesity in children.2,3 As both a lack of PA and obesity track through childhood,4 PA promotion should start at a young age. From the Mannheim Institute of Public Health, Social and Preventive Medicine (Dörr, Bock, Fischer, De Bock), Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany; and the Children’s Hospital (De Bock), University Medicine Mannheim, Heidelberg University, Mannheim, Germany Address correspondence to: Freia De Bock, MD, MPH, Mannheim Institute of Public Health, Social and Preventive Medicine, Medical Faculty Mannheim, Heidelberg University, Ludolf-Krehl-Str. 7-11, D-68167 Mannheim Germany. E-mail:
[email protected]. 0749-3797/$36.00 http://dx.doi.org/10.1016/j.amepre.2014.01.015
& 2014 American Journal of Preventive Medicine
Preschools are important settings5,6 for early preventive efforts at the population level, as approximately 90% of children attend preschool for 57 hours daily on average.7,8 One way of promoting PA in preschools is to provide environments with organizational and contextual features that are associated with increased PA in children (e.g., teachers providing a PA role model9,10 or green grounds11). Previous valid and reliable instruments to measure preschools’ friendliness toward PA (e.g., Early Childhood Environment Rating Scale [ECERS]12,13 and Nutrition and Physical Activity Self-Assessment for Child-Care [NAP SACC]14,15) were based on quantitative surveys. However, qualitative methods might offer several benefits.16 Open-ended interview questions would give teachers the chance to provide data in their own words, avoid misunderstandings, and increase response rates.17 Furthermore, fewer restrictions or assumptions have to be
Published by Elsevier Inc.
Am J Prev Med 2014;46(6):593–604
593
594
Dörr et al / Am J Prev Med 2014;46(6):593–604
placed. Qualitative methods also include direct observation techniques, which minimize social desirability bias.18,19 These findings suggest that the use of qualitative methods for assessing preschool environments might meet the complexity of the environments20 and therefore deserve further in-depth investigation. However, no qualitative instrument exists to date, nor have qualitative data been used to complement quantitative methods.21 Thus, the aims of this study were (1) to develop a comprehensive item battery to assess preschools’ friendliness toward PA based on qualitative data and direct observation; (2) to develop comprehensive and short summary scores from the items; and (3) to test the items’ and scores’ reliability and validity in a mixed methods analysis (combining qualitative and quantitative methods) using data from a large field study.
Methods Study Setting The field study was conducted in Baden-Wuerttemberg, a federal state in Southwest Germany with 11 million residents and 47,600 preschools. Preschools were included via their participation in an existing state-sponsored PA program for which they had applied beforehand. Preschools in the program were representative for the German preschool system (mostly community- or church-run, with half-day attendance). The state-sponsored program offered regular gym lessons but did not interfere with contextual characteristics of the preschool environment. A superordinate clusterrandomized trial was designed to evaluate the program.22 Of 37 preschools forming the baseline, qualitative researchers visited 24 preschools with 568 children in a radius of 200 km maximally (September 2008 to April 2009). Valid questionnaire data were available for 496 of 568 (87.4%) children aged between 3 and 6 years, and 405 (71.3%) had accelerometry and heart rate data as well (analysis in 2012). The Ethical Committee of the Medical Faculty Mannheim, Heidelberg University, approved the study. The parents of all children participating in the study provided written informed consent.
Development and Evaluation of Rating Scores Item development. The development of the rating scores for assessing preschools’ friendliness toward PA encompassed several steps. First, the research team explored PA culture in preschools through informal talks with head teachers and direct environmental observation in a pilot study.23 Second, a literature search was performed (database, MEDLINE; keywords: preschool children, preschool environment, physical activity, preschool playground, playground characteristics, green grounds, preschool play equipment, gardening, outdoor, sports participation, and obesity; search date, July 2008). Referring to this literature search and by using the previously published quantitative instruments for preschool environment rating (e.g., ECERS and NAP SACC) as templates, seven domains
of friendliness toward PA on the preschool level were defined based on both the general systems theory22 and Bandura’s social learning theory9,10 as follows: (A) preschool space11,24–27; (B) preschool teacher role model regarding PA9,10; (C) regular PA sessions25,28; (D) regular outdoor time29–32; (E) available play objects25,26,33; (F) preschool engagement for PA9,10; and (G) season-related preschool PA activity.34 The third step encompassed the design of open-ended interview questions and the development of items for direct observation of the preschools’ environment. Fourth, iterative discussions within the research team and with other PA experts helped refine the question and observation items and led to a final set of 27 items (Table 1; original German questions available on request). Interviews and direct observation. Three qualitative researchers conducted semistructured face-to-face interviews with the person responsible for PA at each preschool (mostly head teachers) using the refined questions and documented the given answers in as much detail as possible directly after the interview. During the preschool visit, the preschool environment was also directly observed using a standardized document. All preschool characteristics were operationalized into quantitative items for a mixed methods analysis by rating the extent to which the preschool fostered children’s PA. Specifically, the interview notes and observations regarding the 27 predefined qualitative preschool characteristics were categorized as limited (0); adequate (1); or, for 18 items, extraordinarily (2) friendly toward PA by one of the authors and another researcher, which led to an intentional weighting of single items. These ratings were performed according to predefined coding schemes (Table 1). After inter-rater reliability testing, incongruent ratings were constituted by discussion. Constitution of rating scores. Responsiveness of all items was first analyzed by calculating the distribution of categories (Supplemental Table). Then, a comprehensive rating score (RS-c) value expressing the degree of friendliness toward PA was attained by summing the ratings of all 27 items (range, 045 points). To increase feasibility and efficiency, a subset of these initial 27 items was used to develop a shorter instrument based on both itemitem correlation and positive association with children’s MVPA. If correlation coefficients were 40.5, only one of the highly correlated items entered the shorter instrument. In addition, only items positively associated with children’s MVPA were included. Thereby, the original RS-c was reduced to a subset score of nine items called the short rating score (RS-s; range, 015 points). Assessment of reliability. Reliability of the scores was studied by inter-rater reliability measures. Specifically, two researchers’ independent ratings of all single 27 items per preschool were compared using contingency tables and the according Cohen’s κ and agreement of ratings per preschool.35 Assessment of validity. Importance of the RS-c was estimated based on a review of the final item battery by three head teachers not included in the field study (ordinal scale from 1¼very important to 6¼not important). Criterion validity was operationalized as the association between each preschool’s scores of friendliness toward PA and objectively measured PA of children enrolled at that preschool.
www.ajpmonline.org
595
Dörr et al / Am J Prev Med 2014;46(6):593–604
Table 1. Items used for assessing preschools’ friendliness toward physical activity
Domaina
Item
Categories
Coding scheme and according coding examples
Items for direct observation A
1. Overall, how would you rate the opportunities for being physically active at this preschool?
0¼limited 1¼adequate 2¼friendly toward PA
A
2. Is there an exercise room?
0¼no 1¼yes
A
3. Do children have free access to the exercise room?
0¼no 1¼yes
A
4. How large is the preschool’s playground?
0¼small 1¼medium-sized 2¼large
A
5. Do children have free access to the playground?
0¼no 1¼yes
A
6. Does the outdoor area have plenty of green space?
0¼no 1¼yes
A
7. Are there places for running or throwing and playing ball?
0¼no 1¼yes
A
8. Are there areas for cycling?
0¼no 1¼yes
A
9. Are there hills or differences in elevation?
0¼no 1¼yes
A
10. Is there a place for setting up a garden?
0¼no 1¼yes
A
11. Overall, how would you rate the amount of space for PA at this preschool?
0¼limited 1¼adequate 2¼friendly toward PA
If preschool is small and there is not much space for PA If there is sufficient space for PA If there is a large, open space for PA
E
12. To what extent is playground equipment available at the preschool?
0¼limited supply 1¼adequate supply 2¼extensive supply
If supply is not sufficient for number of children If supply is sufficient for number of children such that every child could have something to play with If supply is more than sufficient for number of children
0¼several suggestions 1¼few suggestions 2¼no suggestions
If there were 42 ideas or suggestions If there were 1 or 2 ideas or suggestions If there were no ideas or suggestions because of optimal conditions
r250 m2500 m2 600 m2900 m2 Z1,000 m2
If the outdoor area is made primarily of asphalt If outdoor area contains mainly grass and bushes or trees
Items for face-to-face interviews with head teachers A
1. What ideas or suggestions do you have for modifying the outdoor area and/or indoor space? (refers to optimal conditions for PA)
(continued on next page)
June 2014
596
Dörr et al / Am J Prev Med 2014;46(6):593–604
Table 1. Items used for assessing preschools’ friendliness toward physical activity (continued)
Domaina
Item
Categories
Coding scheme and according coding examples
B
2. To what extent do you consider yourself to be a PA role model for children at your preschool?
0¼no 1¼role model 2¼remarkable role model/special education or training/ special hobby
If no PA role model related terms are reported (e.g., too old for PA) If PA role model-related terms are reported (e.g., actively engaged in physical education lessons) If respondent is a member of a sports team or has a PA-related hobby such as horseback riding
B
3. To what extent do you feel that your team serves as a PA role model for children?
0¼no 1¼role model 2¼remarkable role model/special education or training/ special hobby
As above
C
4. How often during the week is PA scheduled at your preschool? (refers to physical education sessions)
0¼ oonce a week 1¼once a week 2¼ 4once a week
If physical education is offered oonce a week If physical education is offered once a week If physical education is offered 4once a week
D
5. How often during the week do you typically schedule PA outside? (refers to being outside/outdoors)
0¼oonce a day 1¼once a day 2¼ 4once a day
If outside PA is scheduled oonce a day If outside PA is scheduled once a day If outside PA is scheduled 4once a day
E
6. In thinking about the equipment available at your preschool, how do you motivate inactive children to be physically active? (refers to play equipment)
0¼other 1¼motivation 2¼high motivation
If respondent reports no inactive children at preschool If motivation through equipment with concrete tasks for children If motivation through special equipment (e.g., Montessori or Bobath equipment)
F
7. What kind of things do you do to ensure that all children engage in at least some fundamental form of PA? (e.g., swinging, playing on the sliding board)?
0¼answer shows no large commitment 1¼answer shows adequate commitment 2¼answer shows high commitment
If requirements of category 1 are not reached If specific and explicit observation of children’s PA, giving individual support (e.g., explicit observation of individual children during physical education, working with activity songs) If in addition to category 1: giving psychomotor support to vulnerable or weak children, individual documentation of children’s development, professional help by PA experts
F
8. What kind of habits exist in your preschool regarding PA? (e.g., how do you arrange special events? Refers to teachers’ level of engagement)
0¼answer shows no large commitment 1¼answer shows adequate commitment 2¼answer shows high commitment
If requirements of category 1 are not reached If respondent reports routine or regularly scheduled activities such as physical education sessions or days outside in the fields or in the forest If respondent reports specialb PA-related events such as field trips to the circus or a city rally
F
9. In thinking about standard procedures or rules at your preschool, what kind of things do you do to motivate inactive children to become physically active? (refers to teachers’ level of engagement)
0¼answer shows no large commitment 1¼answer shows adequate commitment 2¼answer shows high commitment
If no PA-related procedures are reported If respondent reports concrete PA-related procedures If respondent reports individual PA-related procedures
(continued on next page)
www.ajpmonline.org
Dörr et al / Am J Prev Med 2014;46(6):593–604
597
Table 1. (continued)
Domaina
Coding scheme and according coding examples
Item
Categories
F
10. Did you hold any special outdoor events for the children last month?
0¼no events 1¼1 event 2¼ 41 event or something very specialb
If no special outdoor events were held If 1 special outdoor events was held If 41 special outdoor event was held
F
11. In thinking about seasonal PA, what kind of seasonal activities do you organize? (e.g., apple harvesting, sledding)
0¼no large commitment 1¼adequate commitment 2¼high commitment
If maximally 2 seasonal activities are provided all over the year If one activity per season is provided If more than one activity per season or something specialb is provided
F
12. In which way do you approach decisions on what to do with the children outside?
0¼spontaneous 1¼planned
If respondent reports predominantly spontaneous decisions If respondent reports predominantly planned decisions
F
13. What kind of things do you look at to determine children’s interest in outdoor activities?
0¼answer shows no or little commitment and interest 1¼answer shows adequate commitment and interest 2¼answer shows high commitment and interest
If only observation of children’s interests If observation plus dialogue with children If concrete interviews of children about their interests were done
F
14. In thinking of ideas or suggestions for remodeling outdoor areas or indoor space, what do you do to implement them?
0¼answer shows no or little commitment and interest 1¼answer shows adequate commitment and interest 2¼answer shows high commitment and interest
If there were no ideas If there were ideas, but no plans for follow through If there were many ideas or extensive commitment (e.g., discussions with local community for permission)
G
15. In thinking about seasonal PA do you provide a distinct culture of seasonal activities?
0¼answer shows no distinct culture of seasonal activities 1¼answer shows adequate culture of seasonal activities 2¼answer shows distinct culture of seasonal activities
Report of no/few seasonal activities to show a distinct culture of seasonal activities If common seasonal activities are reported Report of many or specialb seasonal activities (e.g., a trip to the forest for a Christmas tree)
Note: Boldface indicates items included in the short rating score (RS-s). a Domains: (A) preschool space; (B) preschool teacher role model regarding PA; (C) regular PA sessions; (D) regular outdoor time; (E) available play objects; (F) preschool engagement for PA; and (G) season-related preschool PA activity. b Special was defined as whether events or activities in preschools reflected extraordinary high commitment and therefore differed from the majority of the other preschools. PA, physical activity
Physical Activity Measurement in Children Children’s accelerometry and heart rate data were recorded with the Actiheart monitors previously validated in adults36 and children23 (CamNtech, Cambridge UK).37 Recordings were performed during “core” preschool hours (9:00AM12:00PM) when children were mostly playing (not eating breakfast or lunch). Measurements used 15-second interval sampling to detect children’s rapid changes and short bursts of PA38 over at least one weekday (83% had Z3 days). Minutes spent in MVPA; June 2014
light activity (LA); or sedentary behavior (SB) were calculated based on cut-offs validated for preschoolers.23
Measurement of Covariates Covariates were included based on literature data1,39–45 and their association with children’s PA (Table 2).46 Covariates at the preschool level encompassed measurement season and preschool location (e.g., rural or urban),22 as both determine opportunities
598
Dörr et al / Am J Prev Med 2014;46(6):593–604 26,47,48
for PA. Covariates at the child level included age, gender, BMI, SES,44,49 immigrant background,50 and participation in organized sports. Sociodemographic information was collected through a parent-administered questionnaire (return rate¼87.4%).
Statistical Analysis
A mixed methods design21,51 was used to investigate the association of single items and rating scores with children’s MVPA, LA, and SB (first in bivariate, second in multivariate models). Intraclass correlation coefficients (ICCs) analyzed the degree of clustering of PA at the preschool level. ICCs 40.1252 for MVPA Table 2. Covariates at preschool and child levels in the multilevel model (ICC¼0.24); LA (0.16); and SB (0.15) required the use of multilevel linear regression Missing % (n) analyses.53 These yielded regression coefficients and COVARIATES ON PRESCHOOL LEVEL (n¼24) 95% CIs as well as variances of MVPA, LA, and SB 0 Location of preschoolsa explained by RS-c, RS-s, and Rural 70.8 single items adjusted for covariates at the preschool and Urban 29.2 individual level. All analyses b Season of measurement 0 were performed using SPSS for Windows, version 15.0 Spring 25.0 (SPSS Inc., Chicago IL), and Fall 54.2 SAS for Windows, version 9.2 (SAS Institute Inc., Cary NC). Winter 20.8 With PA data (n¼405) COVARIATES AT CHILD LEVEL Mean age (years) c
Mean BMI
M (SD)
Missing (n)
M (SD)
Missing (n)
4.9 (0.8)*
0
4.6 (0.8)*
1
15.2 (1.4)
1
15.3 (1.5)
90
% Boys Immigrant background
d
Without PA data (n¼163)
%
52.8
0
51.9
1
36.2
10
43.2
68
e
Maternal education level
12
69
Low
24.9
26.6
Medium
54.5
48.9
20.6
24.5
High Overweight
f
Regular participation in organized sports
g
5.0
1
4.1
90
64.8
16
61.5
72
Note: Percentages do not always add up to 100% because of rounding. a Preschool categorized as rural: located within or near forest, parks, and green spaces within the cutout and away from major highways or industrial areas (Google Earth, predefined altitude, accessed June 6, 2008). Others were categorized as urban. b Season of measurement: measurements were conducted during spring, fall, and winter, not in summer. c BMI, calculated using weight (kg, measured to the nearest 0.1 kg [Soehnle pharo, Nassau Germany]) divided by the height squared (m2, measured to the nearest 0.1 cm [Seca Deutschland, Hamburg Germany]) d Immigrant status of each child was defined as following: (1) child is a non-German national; (2) child’s parents are native speakers of a language other than German; or (3) child primarily speaks a language other than German at home.48 e Each child’s SES was estimated using his/her mother’s self-report of the highest level of educational attainment as a proxy.42 According to the International Standard Classification of Education (ISCED),47 we defined three categories of maternal education as follow: high, ISCED level of 56; middle, ISCED level of 4; and low, ISCED level of 13. f Overweight: BMI 490th percentile for children of the same age and gender g Maternal report of child’s regular participation in PA at a sports club or similar organization Z1 hour/week PA, physical activity n p¼0.001
Results Descriptive statistics of covariates at the preschool and child level are shown in Table 2 and preschool items are listed in Table 3. Children with and without PA data had comparable characteristics except for age (Table 2). Responsiveness (mean [SD], range, % missing) and bivariate significant associations with MVPA (nine items); LA (seven items); and SB (eight items) for preschool items are documented in the Supplemental Table. The score values ranged from 20.7 to 35.0 for RS-c (26.8 [4.0]) and from 7.0 to 15.0 for RS-s (9.0 [2.0]).
Psychometric Properties Cohen’s κ ranged between 0.46 and 1.00 for all single items and was 0.80 and 0.88 for RS-c and RS-s, www.ajpmonline.org
Dörr et al / Am J Prev Med 2014;46(6):593–604
Table 3. Description of sample characteristics at the preschool level (subset of items included in RS-s) Characteristics Availability of green spacea
n (%) Direct observation
Yes
21 (87.5)
No
1 (4.2)
Missing
2 (8.3)
Amount of outdoor space
b
Direct observation
2
Large (Z1,000 m )
9 (37.5)
Adequate (600 m 900 m )
9 (37.5)
small (r250 m 500 m )
5 (20.8)
Missing
1 (4.2)
2
2
2
2
Free access to the playground
Direct observation
Yes
12 (50.0)
No
11 (45.8)
Missing Exercise room available
1 (4.2) Direct observation
Yes
19 (79.2)
No
5 (20.8)
Missing
0 (0.0)
Overall opportunities for PA
Direct observation
To a significant extent
18 (75.0)
Adequate
4 (16.7)
Limited
1 (4.2)
Missing
1 (4.2)
Teacher as PA role model
Face-to-face interview
To a significant extent
7 (29.2)
Adequate
13 (54.2)
Limited
3 (12.5)
Missing
1 (4.2)
Regularly scheduled PA sessions
Face-to-face interview
41/week
4 (16.7)
1/week
16 (66.7)
o1/week
4 (16.7)
Missing
0 (0.0)
Teachers’ commitment to children’s outdoor interests
Face-to-face interview
High commitment
3 (12.5)
Adequate commitment
13 (54.2) (continued on next page)
June 2014
599
respectively. Mean percentage agreement between the two raters across all preschools was 90.7% (4.4%; range¼70.8%100.0%) for RS-c and 94.0% (7.3%) (range¼79.2% 100.0%) for RS-s. All items were rated as highly important by the teachers (range¼1.92.0 on a 6point scale), with no alternative or additional items suggested.
Association of Rating Scores and Single Items with Children’s PA Bivariate analyses. Compared to children from preschools in the lower third of RS-c, children from preschools in the upper third spent significantly more minutes in MVPA (median¼7.3 vs 6.5 minutes, p¼0.019) and LA (84.1 vs 69.1 minutes, po0.001) and fewer minutes in SB (87.6 vs 102.8 minutes, po0.001). Children in the upper third of RS-s spent more minutes in MVPA compared with children in the lower third (median¼7.4 vs 4.9 minutes, po0.001); however, the difference in SB or LA was not significant. Multilevel analyses. Four single items were positively associated with MVPA in the covariate-adjusted single-item multilevel model (large playground, þ1.9 minutes in MVPA, p¼0.04; free access to a playground, þ3.1 minutes, p¼0.02; high commitment in determining children’s
600
Dörr et al / Am J Prev Med 2014;46(6):593–604
Table 3. Description of sample characteristics at the preschool level (subset of items included in RS-s) (continued)
interventions to a respective outcome and surveying preschools’ friendliness toward different Characteristics n (%) intensities of PA. ExistNo or little commitment 5 (20.8) ing quantitative surveys (e.g., ECERS and NAP Missing 3 (12.5) SACC) hitherto did not Teachers’ ideas/suggestions for preschool modification Face-to-face interview differentiate between PA No suggestions 4 (16.7) intensities. Compared to these Few suggestions 8 (33.3) existing surveys, the valiSeveral suggestions 9 (37.5) dity and reliability of both Missing 3 (12.5) mixed methods scores appear to be adequate. Note: For corresponding questions, see Table 1 or Supplemental Table (available at www.ajpmonline.org). a Research team leader’s report of green space (grass and bushes or trees) around or near the preschool The current scores’ items b Estimated by the research team leader during the preschool visit achieved high importance PA, physical activity; RS-s, short rating score as rated by three head teachers. In comparison, 78% of ECERS items were rated interest in outdoor activity, þ5.5 minutes, p¼0.0012; as highly important by seven experts.13 ECERS conscheduled PA sessions Zone/week, maximally þ6.1 firmed predictive validity (as criterion validity) by minutes, po0.001). In addition, the multilevel regression positive association of the score with children’s academic models showed significant associations between RS-c and and cognitive outcomes.55 In the current study, criterion RS-s, respectively, and children’s MVPA (Table 4). An validity of both scores was operationalized as a positive increase of one point in RS-c and RS-s was correlated with association with children’s MVPA. Inter-rater reliability an increase of 0.5 and 1.3 minutes of MVPA, respectively (κ) for all single items ranged between 0.46 and 1.00 (p¼0.003 and po0.001). The final models for RS-c and RS(o0.6 only in five of 27 items, likely related to data s explained approximately 9% and 23% of the variance in quality or question wording). The κ range seems close to children’s MVPA. There were no significant associations that of NAP SACC (0.201.00).15 Thus, rating scores with children’s LA or SB. based on mixed methods data might be able to measure preschools’ friendliness toward PA in an adequate Discussion manner (as compared to quantitative scores). High ICCs (40.14) indicate significant clustering of The study aim was to develop an item battery based on children’s PA at the preschool level. The degree of mixed methods including data from semi-structured clustering increased with PA intensity. This finding interviews and direct observation to assess preschools’ underscores the general importance of the preschool friendliness toward PA. In the second step, a compreenvironment for PA. More specifically, it suggests that hensive rating score including all 27 items across seven for children to engage in MVPA, supportive preschool domains of preschool characteristics was derived. To environmental conditions seem crucial. This result is increase feasibility and efficiency, a shorter instrument further underlined by the fact that both scores are using a subset of items based on itemitem correlation correlated with children’s MVPA, but not with their and association with MVPA was developed. Inter-rater LA or SB. A major implication of these high ICCs is reliability of both scores was high. Validity of scores that future studies examining preschool determinants of and four single items was suggested by significant PA—in contrast to previous studies7,31—should account associations with objectively measured PA of 4400 preschoolers in a mixed methods multilevel regression for clustering of children’s PA by applying appropriate statistical methods. analysis. Some preschool characteristics from the item battery Strengths and Limitations were associated with MVPA (multilevel model: free access to54 or large playground,11,26 teacher commitTo our knowledge, both rating scores for assessing ment,10 and scheduled PA sessions25), whereas others preschools’ friendliness toward PA are the first to be based on mixed methods data. Strengths of this approach were associated with LA and SB (only bivariate analyses: include fewer assumptions,16 control of and insight into scheduled outdoor and seasonal activities, Supplemental Table). The item battery thereby facilitates tailoring the complexity of environments, and a higher response www.ajpmonline.org
601
Dörr et al / Am J Prev Med 2014;46(6):593–604
Table 4. Multilevel associations between rating scores and accelerometer-measured physical activity in preschool children (n¼405) Moderate to vigorous physical activity
Light activity
Sedentary behavior
RS-c: β (95% CI)
0.51** (0.18, 0.85)
0.10 (1.33, 1.52)
0.61 (2.00, 0.78)
RS-s: β (95% CI)
1.32*** (0.96, 1.68)
1.08 (3.50, 1.34)
0.25 (2.71, 2.21)
M
8
80
92
Median
6.4
80.1
93.5
0.24
0.16
0.15
Minutes spent in activity (9:00AM12:00PM)
ICC
Note: p-values r0.05 were considered significant. Each multilevel model was adjusted for covariates at the preschool level (measurement season and geographic location of preschools) and at the child level (age, gender, BMI, SES, immigrant background, and participation in organized sports). β, regression coefficient: increase of mean minutes spent in physical activity during 9:00AM12:00PM per rating score unit. nn pr0.01 nnn pr0.001 ICC, intraclass correlation coefficient, RS-c, comprehensive rating score, RS-s, short rating score
rate.17 Despite the necessity of repeatedly rating complex qualitative interview and observation data, the results show high inter-rater reliability. Thus, rigid coding schemes such as that presented seem to enable quantifying qualitative information in a reliable manner. Further strengths are the objective PA measurement37 for testing criterion validity,21 the consideration of clustered PA data using multilevel linear regression analyses,53 and a large sample size. Significant associations of both items and scores with MVPA suggest their validity at least in European preschool contexts, which might not be adequately represented in all other existing scores validated in the American context. A further strength of this study is to combine the “best of both worlds”: the development of a comprehensive item battery and score to rate preschools’ friendliness toward PA with increasing feasibility and efficiency by offering a short score based on only nine qualitative items. RS-s explains a large portion of the variance in MVPA. This might be due to the condition that items only entered the RS-s if they were associated with MVPA. However, it also means that researchers specifically interested in children’s MVPA could use the short score with only nine items without losing significant information compared to the comprehensive instrument. To measure LA and SB, RS-c might be better based on bivariate analyses. Despite these strengths, there are also some limitations. First, owing to cross-sectional data, reverse-causation bias precludes drawing causal conclusions. Second, children’s PA behavior might have been influenced by June 2014
children’s reactivity to wearing the monitor. Third, nonrepresentative weather conditions and activities might have affected the results. Fourth, as the scores were mainly developed for other researchers having the opportunity to visit preschools under study in person, they may be of limited utility to agencies, parents, or preschools themselves. Fifth, generalizability of the findings might be limited, as participating preschools were located in only one federal state, had applied for participation in a PA program, and therefore might be already more friendly toward PA. However, the latter would not alter the resulting associations. Although both rating scores are not yet cross-validated in an independent sample, there is no indication of a selection bias at the child level because the sociodemographic characteristics are similar to regional figures.56,57 At the preschool level, potential generalizability is underlined by its mixture of state- and church-sponsored preschools in rural and urban locations, which might be typical for western European preschools.
Clinical Relevance An increase of one score point leads to 0.5 and 1.3 more minutes spent in MVPA during preschool hours (9:00AM12:00PM) for RS-c and RS-s, respectively. Given that children spend an average of 22.5 hours per week at preschool,7 the findings suggest that children attending preschools with high friendliness toward PA could spend up to 79.2 more minutes (maximal RS-s value compared to minimal value) in MVPA per week. Children in preschools in the highest score value tertile compared
602
Dörr et al / Am J Prev Med 2014;46(6):593–604
to the lowest tertile still spend 37.4 more minutes in MVPA weekly. Considering current recommendations of at least 60 minutes of MVPA every day for children aged 512 years (specific recommendations for preschool age are lacking),58,59 children could thus accumulate up to a quarter of the recommended time if enrolled in preschools with high friendliness toward PA. This is consistent with current evidence reported by Bower et al.,7 who used an expansion of NAP SACC (Environment and Policy Assessment and Observation [EPAO] instrument)60 and found that children spend approximately 80 more minutes in MVPA per week when comparing centers with high and low PA environment scores.
Conclusion Both scores used to rate preschools’ friendliness toward PA demonstrate acceptable psychometric properties. These results indicate the utility of the mixed methods– based scores, but also of single items, in assessing preschools’ friendliness toward PA. As many young children do not meet PA recommendations59,61 and are at increased risk of obesity and resulting health problems,62–69 such assessment is important and might reinforce the role preschool environments play in promoting high-intensity PA at the population level. Both single items and scores could help to monitor preschools’ friendliness toward PA in Europe and to evaluate the success of PA-promoting interventions in preschools. Researchers interested in children’s MVPA should use the short rating score, whereas the comprehensive score might be more interesting for LA and SB. The comprehensive item battery based on mixed methods might be the starting point for the development of a comprehensive survey of preschools’ friendliness toward PA in the future. Studies could compare the newly developed scores with existing scores (e.g., ECERS and NAP SACC) hitherto primarily used in the U.S. The high level of clustering of children’s PA data at the preschool level suggests the need for appropriate statistical methods in future studies assessing determinants of PA behavior in preschools. We thank the parents, children, and preschools for their cooperation. We also thank our research assistants and associated researchers (Ann-Kristin Ehling and Jan Henkel) as well as Marion Becker for their help in conducting interviews, direct observations, and rating of interview and observation notes. We sincerely thank David Litaker, MD, PhD (Departments of Medicine, Epidemiology, and Biostatistics, Case Western Reserve University, Cleveland OH), for the quite detailed and thorough review of the manuscript and his help throughout the entire writing process. We also thank Marc
Jarczok (Mannheim Institute of Public Health, Heidelberg University, Germany) for his help with the data management.
Ethical Standards The parents of all children participating in the study provided their written informed consent. Ethical approval was granted by the ethics committee of the Medical Faculty Mannheim, Heidelberg University (ID 2008-275N-MA). No financial disclosures were reported by the authors of this paper.
References 1. Pearce MS, Basterfield L, Mann KD, Parkinson KN, Adamson AJ, Reilly JJ. Early predictors of objectively measured physical activity and sedentary behaviour in 8–10 year old children: the Gateshead Millennium Study. PLoS One 2012;7(6):e37975. 2. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the U.S., 1999–2004. JAMA 2006;295(13):1549–55. 3. Seth A, Sharma R. Childhood obesity. Indian J Pediatr 2013;80(4): 309–17. 4. Herman KM, Craig CL, Gauvin L, Katzmarzyk PT. Tracking of obesity and physical activity from childhood to adulthood: the Physical Activity Longitudinal Study. Int J Pediatr Obes 2009;4(4):281–8. 5. Pate RR, Pfeiffer KA, Trost SG, Ziegler P, Dowda M. Physical activity among children attending preschools. Pediatrics 2004;114(5):1258–63. 6. Finn K, Johannsen N, Specker B. Factors associated with physical activity in preschool children. J Pediatr 2002;140(1):81–5. 7. Bower JK, Hales DP, Tate DF, Rubin DA, Benjamin SE, Ward DS. The childcare environment and children's physical activity. Am J Prev Med 2008;34(1):23–9. 8. Pate RR, McIver K, Dowda M, Brown WH, Addy C. Directly observed physical activity levels in preschool children. J Sch Health 2008;78(8): 438–44. 9. Bandura A. Lernen am Modell. Ansätze zu einer sozial-kognitiven lerntheorie. Stuttgart Germany: Klett, 1976. . 10. Copeland KA, Kendeigh CA, Saelens BE, Kalkwarf HJ, Sherman SN. Physical activity in child-care centers: do teachers hold the key to the playground? Health Educ Res 2012;27(1):81–100. 11. Dyment JE, Bell AC. Grounds for movement: green school grounds as sites for promoting physical activity. Health Educ Res 2008;23(6): 952–62. 12. Harms T, Clifford RM, Cryer D. Early Childhood Environment Rating Scale—Revised (ECERS-R). New York: Teachers College Press, 1998. . 13. Clifford RM, Reszka SS, Rossbach H-G. Reliability and validity of the Early Childhood Environment Rating Scale. ers.fpg.unc.edu/d-reliabilityand-validity-early-childhood-environment-rating-scale. 14. Ammerman AS, Ward DS, Benjamin SE, et al. An intervention to promote healthy weight: Nutrition and Physical Activity SelfAssessment for Child Care (NAP SACC) theory and design. Prev Chronic Dis 2007;4(3):A67. 15. Benjamin SE, Neelon B, Ball SC, Bangdiwala SI, Ammerman AS, Ward DS. Reliability and validity of a nutrition and physical activity environmental self-assessment for child care. Int J Behav Nutr Phys Act 2007;4:29. 16. Britten N, Fisher B. Qualitative research and general practice. Br J Gen Pract 1993;43(372):270–1. 17. Doyle JK. Face-to-face surveys. In: Everitt BS, Howell D, eds. The encyclopedia of statistics in behavioral science. New York: Wiley, 2005. 18. Motl RW, McAuley E, DiStefano C. Is social desirability associated with self-reported physical activity? Prev Med 2005;40(6):735–9.
www.ajpmonline.org
Dörr et al / Am J Prev Med 2014;46(6):593–604 19. Jonas K, Stroebe W, Hewstone M. Social Psychology. An Introduction. 5th Edition. Heidelberg, Germany: Springer, 2007. 20. Sofaer S. Qualitative research methods. Int J Qual Health Care 2002;14 (4):329–36. 21. Creswell JW, Fetters MD, Ivankova NV. Designing a mixed methods study in primary care. Ann Fam Med 2004;2(1):7–12. 22. De Bock F, Fischer JE, Hoffmann K, Renz-Polster H. A participatory parent-focused intervention promoting physical activity in preschools: design of a cluster-randomized trial. BMC Public Health 2010;10:49. 23. De Bock F, Menze J, Becker S, Litaker D, Fischer J, Seidel I. Combining accelerometry and HR for assessing preschoolers’ physical activity. Med Sci Sports Exerc 2010;42(12):2237–43. 24. Cardon G, Van Cauwenberghe E, Labarque V, Haerens L, De Bourdeaudhuij I. The contribution of preschool playground factors in explaining children’s physical activity during recess. Int J Behav Nutr Phys Act 2008;5:11. 25. Taylor RW, Farmer VL, Cameron SL, Meredith-Jones K, Williams SM, Mann JI. School playgrounds and physical activity policies as predictors of school and home time activity. Int J Behav Nutr Phys Act 2011;8:38. 26. Dowda M, Brown WH, McIver KL, et al. Policies and characteristics of the preschool environment and physical activity of young children. Pediatrics 2009;123(2):e261–e266. 27. Ahmed AT, Oshiro CE, Loharuka S, Novotny R. Perceptions of middle school educators in Hawai'i about school-based gardening and child health. Hawaii Med J 2011;70(7S):S11–S15. 28. Antonogeorgos G, Papadimitriou A, Panagiotakos DB, Priftis KN, Nicolaidou P. Association of extracurricular sports participation with obesity in Greek children. J Sports Med Phys Fitness 2011;51(1):121–7. 29. 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;158(4):353–7. 30. Burdette HL, Whitaker RC. Resurrecting free play in young children: looking beyond fitness and fatness to attention, affiliation, and affect. Arch Pediatr Adolesc Med 2005;159(1):46–50. 31. Baranowski T, Thompson WO, DuRant RH, Baranowski J, Puhl J. Observations on physical activity in physical locations: age, gender, ethnicity, and month effects. Res Q Exerc Sport 1993;64(2):127–33. 32. Sallis JF, Nader PR, Broyles SL, et al. Correlates of physical activity at home in Mexican-American and Anglo-American preschool children. Health Psychol 1993;12(5):390–8. 33. Hannon JC, Brown BB. Increasing preschoolers’ physical activity intensities: an activity-friendly preschool playground intervention. Prev Med 2008;46(6):532–6. 34. Dowda M, Pate RR, Trost SG, Almeida MJ, Sirard JR. Influences of preschool policies and practices on children’s physical activity. J Community Health 2004;29(3):183–96. 35. Fleiss JL, Cohen J. The equivalence of weighted kappa and the intraclass correlation coefficient as measures of reliability. Educ Psychol Meas 1973;33:613–9. 36. Brage S, Brage N, Franks PW, Ekelund U, Wareham NJ. Reliability and validity of the combined heart rate and movement sensor Actiheart. Eur J Clin Nutr 2005;59(4):561–70. 37. Chen KY, Bassett DR Jr. The technology of accelerometry-based activity monitors: current and future. Med Sci Sports Exerc 2005;37 (11S):S490–S500. 38. Trost SG, McIver KL, Pate RR. Conducting accelerometer-based activity assessments in field-based research. Med Sci Sports Exerc 2005;37(11S):S531–S543. 39. Van Der Horst K, Paw MJ, Twisk JW, Van Mechelen W. A brief review on correlates of physical activity and sedentariness in youth. Med Sci Sports Exerc 2007;39(8):1241–50. 40. Rich C, Griffiths LJ, Dezateux C. Seasonal variation in accelerometerdetermined sedentary behaviour and physical activity in children: a review. Int J Behav Nutr Phys Act 2012;9:49.
June 2014
603
41. Trost SG, Pate RR, Sallis JF, et al. Age and gender differences in objectively measured physical activity in youth. Med Sci Sports Exerc 2002;34(2):350–5. 42. Grontved A, Pedersen GS, Andersen LB, Kristensen PL, Moller NC, Froberg K. Personal characteristics and demographic factors associated with objectively measured physical activity in children attending preschool. Pediatr Exerc Sci 2009;21(2):209–19. 43. Metcalf BS, Hosking J, Jeffery AN, Voss LD, Henley W, Wilkin TJ. Fatness leads to inactivity, but inactivity does not lead to fatness: a longitudinal study in children (EarlyBird 45). Arch Dis Child 2011;96 (10):942–7. 44. Gordon-Larsen P, McMurray RG, Popkin BM. Determinants of adolescent physical activity and inactivity patterns. Pediatrics 2000; 105(6):E83. 45. Vandendriessche JB, Vandorpe BF, Vaeyens R, et al. Variation in sport participation, fitness and motor coordination with socioeconomic status among Flemish children. Pediatr Exerc Sci 2012;24(1):113–28. 46. De Bock F, Genser B, Raat H, Fischer JE, Renz-Polster H. A participatory physical activity intervention in preschools: a cluster randomized controlled trial. Am J Prev Med 2013;45(1):64–74. 47. Tucker P, Gilliland J. The effect of season and weather on physical activity: a systematic review. Public Health 2007;121(12):909–22. 48. Beighle A, Alderman B, Morgan CF, Le Masurier G. Seasonality in children’s pedometer-measured physical activity levels. Res Q Exerc Sport 2008;79(2):256–60. 49. UN Educational Scientific and Cultural Organization. ISCED 1997— International Standard Classification of Education. New York: UNESCO, 2006. www.uis.unesco.org/Library/Documents/isced97-en.pdf. 50. Schenk L, Bau AM, Borde T, et al. A basic set of indicators for mapping migrant status. Recommendations for epidemiological practice [in German]. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2006;49(9):853–60. 51. Migiro SO, Magangi BA. Mixed methods: a review of literature and the future of new research paradigm. Afr J Bus Manage 2011;5(10):3757–64. 52. James L. Aggregation bias in estimates of perceptual agreement. J Appl Psychol 1982;67(2):219–29. 53. Ditton H. Mehrebenenanalyse. Grundlagen und Anwendungen des Hierarchisch Linearen Modells. Weinheim, Germany: Juventa, 1998. . 54. Dwyer GM, Higgs J, Hardy LL, Baur LA. What do parents and preschool staff tell us about young children’s physical activity: a qualitative study. Int J Behav Nutr Phys Act 2008;5:66. 55. Sammons P, Sylva K, Melhuish E, Siraj-Blatchford I, Taggart B, Elliot K. Measuring the impact of pre-school on children’s cognitive progress over the pre-school period. In: Sylva K, ed. The effective provision of preschool education (EPPE) project. London: Institute of Education, 2003. 56. Kurth BM, Schaffrath Rosario A. The prevalence of overweight and obese children and adolescents living in Germany. Results of the German Health Interview and Examination Survey for Children and Adolescents (KiGGS) [in German]. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2007;50(5–6):736–43. 57. Ministerium für Arbeit und Soziales B-W. Familien in BadenWürttemberg Bildung: integration von kindern und jugendlichen mit migrationshintergrund. Statistisches Landesamt Baden-Württemberg 2008. 58. Australian Government Department of Health and Ageing. Physical activity guidelines. health.gov.au/internet/main/publishing.nsf/con tent/health-pubhlth-strateg-phys-act-guidelines. 59. Tremblay MS, Warburton DE, Janssen I, et al. New Canadian physical activity guidelines. Appl Physiol Nutr Metab 2011;36(1):36–46; 47–58. 60. Ward D, Hales D, Haverly K, et al. An instrument to assess the obesogenic environment of child care centers. Am J Health Behav 2008;32(4):380–6. 61. Lampert T, Mensink GB, Romahn N, Woll A. Physical activity among children and adolescents in Germany. Results of the German Health Interview and Examination Survey for Children and Adolescents
604
62. 63.
64. 65.
66.
Dörr et al / Am J Prev Med 2014;46(6):593–604
(KiGGS) [in German]. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2007;50(5–6):634–42. Dietz WH. Health consequences of obesity in youth: childhood predictors of adult disease. Pediatrics 1998;101(3 Pt 2):518–25. Freedman DS, Dietz WH, Srinivasan SR, Berenson GS. The relation of overweight to cardiovascular risk factors among children and adolescents: the Bogalusa Heart Study. Pediatrics 1999;103(6 Pt 1):1175–82. Must A, Strauss RS. Risks and consequences of childhood and adolescent obesity. Int J Obes Relat Metab Disord 1999;23(2S):S2–S11. Wabitsch M. Overweight and obesity in European children: definition and diagnostic procedures, risk factors and consequences for later health outcome. Eur J Pediatr 2000;159(1S):S8–S13. Ebbeling CB, Pawlak DB, Ludwig DS. Childhood obesity: public-health crisis, common sense cure. Lancet 2002;360(9331):473–82.
67. Reilly JJ, Methven E, McDowell ZC, et al. Health consequences of obesity. Arch Dis Child 2003;88(9):748–52. 68. Lobstein T, Baur L, Uauy R. Obesity in children and young people: a crisis in public health. Obes Rev 2004;5(1S):S4–S104. 69. Botton J, Heude B, Kettaneh A, et al. Cardiovascular risk factor levels and their relationships with overweight and fat distribution in children: the Fleurbaix Laventie Ville Sante II study. Metabolism 2007;56(5):614–22.
Appendix Supplementary data Supplementary data associated with this article can be found in the online version, at http://dx.doi.org/10.1016/j.amepre.2014.01.015.
www.ajpmonline.org