Pupils' use of school outdoor play settings across seasons and its relation to sun exposure and physical activity

Long outdoor stay may cause hazardous exposure to ultraviolet radiation (UVR) from the sun even at high latitudes as in Sweden (Spring to Autumn). On the other hand, long outdoor stay is a strong predictor of primary school children´s free mobility involving moderate to vigorous physical activity (MVPA). UV‐protective outdoor environments enable long outdoor stay. We investigated the concurrent impact of different school outdoor play settings upon pupils´ sun exposure and levels of physical activity across different ages, genders, and seasons.


| INTRODUC TI ON
Children in Western countries increasingly sit indoors for hours on end during long, regulated schooldays. 1 Increasing time spent indoors may reduce children's opportunities for healthy levels of physical activity, defined as 60 minutes in moderate to vigorous physical activity (MVPA), 2 and occasional sun exposure, possibly enough for vitamin D production at mid-latitudes. 3 Long outdoor stay is a strong predictor of primary school chil-dren´s PA, at least in temperate climates. 4 On the other hand, long outdoor stay may cause hazardous exposure to ultraviolet radiation (UVR) from the sun even at high latitudes as in Sweden. 5,6 Further, overexposure to UVR at an early age increases the risk of skin cancer later in life. 7 It has been established that the occurrence of skin cancer increases rapidly among adults in Sweden as well as in other Western societies. 8 Reversely, sun in winter at high latitudes cannot maintain optimal vitamin D levels. 9,10 Low levels of PA combined with low levels of vitamin D may have implications for children's bone growth. 11 Studies on school children show the need of active intervention policies to increase PA, 12 and action to reduce sun exposure 13 whenever the risk of overexposure is high. Differences in children's UVR exposures could be explained in part by PA. For instance, passive outdoor pursuits have been associated with higher UVR exposures than outdoor PA and travel pursuits. 14 Results from the Colorado Kids Sun Care Program (a sun safety intervention trial) show that the promotion of sun safety, like avoidance of midday sun and sun protection during outdoor activities, is not likely to inhibit PA. 15 However, to the knowledge of the authors, UVR exposure and PA combined has never been studied in relation to play settings (PS).
In a systematic review examining children's and adolescent's PA during school recess, Ridgers et al found a positive correlation between the provision of loose equipment (eg, balls and skipping ropes) and MVPA. 16 Further, playground markings (for games and sports) and physical structures have been shown to increase school children's PA during recess. 12,17 In 11-to 12-year-old pupils, ball play areas may be related to even higher PA intensity levels, which are maintained with increasing age, especially in boys. 18 Woodland may serve as an opportunity for boys and girls to be physically active while playing together. 19 Studies of preschool environment show that greenery promotes physical activity in shelter from excessive UVR during free play. 20,21 For preschool outdoor environment, more specific configurations of factors inviting to behaviors that serve multiple health gains have been defined, 21-23 factors deemed relevant for school children only to a certain extent. 19 Impact of season, time outdoors, and amount of free sky upon UV exposure from the sun 5 and impact of area surface upon MVPA 24 have been previously investigated in this study population, but without relating the outcome to play settings. In this study, we investigated the concurrent impact of different school outdoor PSs upon Swedish pupils' sun exposure (UVR) and PA across different ages and seasons during one academic year.

| Play setting and weather observations
During fieldwork the observers were in charge of one class each (in mean 16 pupils/class), using a class specific protocol with the pupils' names to record arrivals and departures from the school premises, as well as in-and outdoor times (recesses and occasional outdoor lessons). Each child was categorized as being indoors when observed participating in activities inside the school building. Outdoor activity was categorized from the moment the child

| Assessment of individual UVR exposure
For individual assessment of erythemally effective UV radiation exposure, calibrated polysulphone film dosimeter badges 25,26 were applied at the shoulder in horizontal position. 27 The dosimeters were calibrated at the University of Manchester, School of Earth, Atmospheric and Environmental Sciences, UK, and returned and analyzed there after exposure with results listed as Standard Erythema Doses "SEDs" (1 SED = 100 J/m 2 ). 28 The absorbance of dosimeters at 330 nm is measured before and after exposure, and the difference entered into a polynomial function empirically formulated to give approximate results expressed as erythemal UV exposures weighted according to CIE's erythema reference action spectrum. 28 Polysulphone film dosimeters are proven to be reliable tools to measure personal UV exposures. 29  The pupils' mean daily UVR exposures (J/m 2 ) were calculated from measured accumulated week exposures of the individual pupils' UV dosimeters, however, without considering variations in global UVR due to different weather conditions from day to day as the dosimeters were not time-stamped. 5 The mean values were then compared with a three-step scale of exposures ranging from suberythemal and possibly insufficient to form vitamin D to suberythemal but potentially sufficient for vitamin D depending on clothing and up to erythemal. A "standard vitamin D dose" or "SDD" has been suggested 10 equivalent to the exposure of sun-sensitive persons (skin type I) to ¼ or less of a "Minimal Erythemal Dose" (1/4 MED) of ¼ of the skin area every other day, for example, equal to exposure every second day of hands, face, and arms 15-25 minutes at noon and at mid-latitude in March or equal to an oral dose of about 1000 IU. 10,30 In this study, UV fraction (%UV) is the ratio of accumulated exposure of each individual dosimeters to ambient accumulated available UVR during each individual child's stay outdoors at their respective sites. Available ambient UVR was calculated from the roof-top dosimeter readings modulated by the relative measurements of the UV-index meter-absolute calibration at the Davis factory was irrelevant in this study.

| Assessment of physical activity
For assessment of MVPA, hip-worn accelerometers (Actigraph GT3X+) were used and set to record movement in 10-second epochs with a sampling frequency of 60 Hz to get detailed PA data. 31,32 The sum of counts in each epoch was categorized into MVPA, using validated and recommended cut points (>383 counts/10 s) for MVPA 33,34 ). All 10-second epochs reaching the limit for MVPA were summarized into total daily minutes spent in MVPA. As a supplement to ocular observation, a built-in light sensor (Actilux) of the accelerometers was used to F I G U R E 1 Example of PSs at school no.1; 1-fixed play equipment. 2-paved surfaces. 3-sport fields. 4-green settings separate outdoor MVPA from indoor MVPA, as only outdoor MVPA was analyzed in this study. This procedure is clarified somewhere else. 24

| Evaluation and statistical analysis
The SPSS™ (Statistical Package for the Social Sciences for Windows, 22.0, 2014, SPSS Inc) and SAS for Windows software packages For confounder control, bivariate analysis was carried out from diary records (health status and PA outside school). None of them being significantly related to MVPA. 24

| Environmental conditions
Apart from temperature, the weather in terms of sky conditions did not differ significantly between seasons and schools. The sky was partly cloudy and overcast and clear in between with only one rainy day in September at one of the schools. The mean maximal outdoor temperature during the week of fieldwork was 16

| UVR exposure at different PS's and seasons
Weekly available UVR (CIE-weighted, erythemally effective 28

| MVPA at different PS's and seasons
Over the whole year, sport fields generated more minutes in MVPA than any other PS (P < .01) ( Table 2). Contrary to sport fields, paved areas yielded fewest minutes of MVPA, whereas green settings accumulated the next to most minutes of MVPA in combination with being protective against overexposure to UVR even in May. In March, the levels of MVPA were at the lowest compared with the other seasons. In September and May, boys were more active than girls in terms of MVPA, irrespective of grade (P < .05). In May, sport fields yielded more than half of the daily need of MVPA, and 18 pupils (12%) obtained ≥60 minutes of MVPA during school time ( Table 2).

| D ISCUSS I ON
Based on season and gender, the main finding was that school outdoor environments impact UVR exposures and minutes in MVPA in 7-to 11-year-old pupils differently depending on PS.
The higher levels of MVPA at sport fields correspond with the results from systematic reviews of PA showing that school facilities, playground markings, and unfixed equipment have a potential to increase PA levels during recess. 12,16 A previous study pointed out the important role of paved open spaces for PA as children tend to run a lot in many of the traditional schoolyard games including elements of hide and seek and chasing. 19 Our results, showing the lowest level of MVPA counts at paved open spaces, possibly depend on the fact that children with woodland at hand get more inclined to use these rather than the open paved areas. This is in line with Andersen et al 2015, showing that solid surfaces on the schoolyard such as paved areas generate less MVPA than other surfaces. 35 Low activity at sun-exposed paved spaces in May could also be explained by thermal discomfort due to high temperatures.
The gendered character of school play is well documented. 36 The present study shows that in September and March, boys had a significantly higher level of MVPA than girls, irrespective of grade. The gender differences showing that boys accumulate more MVPA than girls align with previous research. 37 Also, a decline of MVPA by age has been reported in several cross-sectional studies. 37,38 According to the result of this study, we would suggest as a routine to add trees and other greenery in proximity to sport fields to add sun protection as such fields are known to be well-used settings generating high levels of MVPA. Surface reflectance is another factor to consider. Graveled surfaces were common on soccer fields and under fixed play equipment and pupils who were active there were the ones most highly exposed to UVR. Graveled surfaces may reflect up to 25% of UVR whereas grass is almost non-reflectant. 39 This could explain some of the differences in UVR exposure between PSs. Grassy sport fields in green surroundings could lower the risk for hazardous UVR exposure in May, apart from taking opportunities to schedule outdoor activities in the mornings and late afternoons.
Further, the greening of sport fields as well as of paved areas and of areas with play equipment is motivated as the presence of trees and other greenery is known to increase the play value and thereby make outdoor settings more attractive for children.

F I G U R E 3
Mean UV exposures during school time by season and play setting. Total for 2nd and 5th graders and genders. The dotted black line represents the limit for hazardous UV exposure (1 MED), and the lower dotted gray line represents the lower limit for vitamin D production The results indicate that the chances to acquire vitamin D from solar exposure in March were low. Depending on clothing and thus sun -exposed skin area, all 2nd graders and more or less all 5th graders were potentially sufficiently exposed to UVR for own vitamin  19,40 when greenery contributes with many affordances. 41 Thus, these studies need to be supplemented with investigations outside the growing season when the surroundings are more plain.
The significance of a strategic design to reduce UVR exposures during play and to extend safe outdoor stays has been stressed in a recent study 42 and in numerous shade policies around the world. The relevance of shade planning is obvious from this study as well.
No interrater reliability testing of the observers was performed, but the observers were trained in studying pupils' movements and PA behavior in school environments. As the same population of pupils was studied during the three study periods, each observer became familiar with his/her assigned class of pupils, which simplified the task. Yet, as the children were not individually tracked, the results have to be interpreted with caution.

| CON CLUS ION
A set of defined common and well-used PSs in the outdoor environments of primary schools may promote children's health when better adapted to different seasons and children's grades and gender.
More outdoor activities in early spring are recommended. In May, greenery attractive for play could protect against overexposure to UVR and stimulate both girls and boys to vigorous play.