play · education · health
playground gains that don't show at a year
In a 2025 meta-analysis, playground enrichment raised children's activity significantly only in studies measured before 12 months, and two cluster trials lasting a school year or more, one adding risk and loose parts and one adding zoned markings with weekly-changing games, found no gain in measured activity
Source: Oppici et al., “How can physical enrichment of school playgrounds improve…” Open ↗source: primaryLast updated: 2026-11-12
Problem statement
Primary-school playgrounds are one of the few places where nearly every child gets daily, unstructured play, and schools try to make them more playable: painted line markings and zones, climbing structures, loose parts (tyres, crates, fabric), fewer rules, more room for risk. A 2025 systematic review and meta-analysis of 28 studies (19,753 participants; 23 in primary schools, 3 in preschools, 3 in secondary schools) found that these "physical enrichment" changes raise moderate-to-vigorous physical activity (MVPA) by a very small amount on average (Hedges' g = 0.17), but the effect was significant only in studies measured before 12 months (g = 0.29), not in studies measured at 12 months or later (12 months was the median assessment point; this is a comparison between studies, not a measured decline within them). Two year-long cluster trials in ordinary schools show what that looks like on the ground. In New Zealand (PLAY, a two-year trial in 16 schools, 840 children), schools added risk and challenge, cut rules and brought in loose parts; their play environments measurably improved and staff believed children were more active, but accelerometers showed no gain in MVPA at break time or lunch time at one or two years (lunch time at two years: 0.4 minutes, 95% CI −1.1 to 2.0; across the whole school day, intervention children did 3.3 minutes less than control children at two years) and no difference in weight. In the Netherlands (21 schools, 1,154 children, mean age 9), a programme that divides the playground into five colour-coded zones and changes the games in each zone every week, run on one and later two days a week, produced no effect on recess physical activity or social–emotional functioning over a school year, and the control children's motor skills improved more. The unsolved problem is a playground change whose effect on children's active play lasts a whole school year, in recess periods as short as 10 minutes, without adding adult organising time.
Why this matters
The review's authors note that over 770 million children are enrolled in primary and secondary education worldwide, which is why playgrounds are treated as a reach-everyone setting. Schools are spending on this now: interventions in the review cost from almost nothing (recycled materials, markings painted by researchers or school staff) to about €10,000–20,000 per school for markings and equipment, and up to €300,000 per school for full renovations. One loose-parts and outdoor-play programme for schools, OPAL, reports having run in over 2,000 UK schools. If the gains are mostly a novelty effect, much of that spending buys a few months of extra activity. The evidence base is weak in a way that matters for decisions: only 3 of the 28 studies in the review had low risk of bias, the MVPA prediction interval runs from a moderate effect in favour of control to a large effect in favour of enrichment (−0.37 to 0.70), and the review rates the level of evidence for all the outcomes it considered as low.
What’s been tried and why it hasn’t worked
Risk, challenge and loose parts (PLAY, New Zealand). Eight intervention schools redesigned their play environments; at one year they scored 4.50 points higher than control schools on a playworker's play-evaluation score (baseline mean 19.0). Schools liked the change and reported more activity, but objectively measured activity did not increase. The authors' reading is that "children may have been involved in different, rather than additional activities." A companion paper from the same trial found intervention children more likely to be happy at school (OR 1.64 at 2 years) and to play with more children (OR 1.66 at 1 year), and also more likely to report being pushed or shoved (OR 1.33) but less likely to tell a teacher (OR 0.69). A 2017 systematic review of loose-parts play found mostly null results for social outcomes (peer acceptance, social skills, peer group size) in controlled studies and concluded the evidence was insufficient either way (its controlled studies include the PLAY bullying paper and the Sydney Playground Project below, so it is not independent of them). Zoned markings with weekly novelty (Beweeg Wijs, Netherlands). This is close to the obvious fix for fading novelty: an online platform supplies a new weekly plan of activities for each of five zones, a supervisor introduces them, and older pupils act as junior coaches; intervention days start at one a week and rise to two halfway through. Over one school year it had no effect on MVPA, sedentary time, or social–emotional outcomes, and implementation fidelity did not change that. The authors' explanations include short recesses (10–45 minutes) eaten up by organising — choosing activities and repeating instructions as new children join; a programme that may be too adult-directed for a time children see as theirs; and, for earlier positive studies, a possible novelty effect. Most schools approached declined to take part because of workload. What the pooled evidence says about design. In the 2025 meta-regression, line markings (alone or with structures) were associated with larger MVPA effects, and involving end-users in the design was associated with smaller ones. Only 4 of the 28 studies built in novelty and only one built in scaffolding, the two principles the review's authors single out: novelty, which they suggest could be the key to maintaining the short-term gain, and scaffolding, which they expect to raise children's engagement; the Dutch trial, published after the review, suggests that novelty supplied by adults one or two days a week does not do it on its own. A separate 2025 review limited to playground markings (11 studies) pooled two RCTs, found no significant difference, and rated the certainty of evidence very low, which tempers the meta-regression's line-marking result. One within-study look at decay among the included studies, a non-randomised UK markings-and-structures study in 26 schools, found a positive but not statistically significant effect that was stronger at 6 months than at 12. Child-reconfigurable loose parts have already been tried. The Sydney Playground Project (12 schools, 226 children aged 5–7) put recycled materials with no obvious play purpose into playgrounds, with no defined adult role, plus a risk-reframing workshop for parents and teachers; MVPA rose by 1.8 minutes against control at the 13-week post-test, taken in the final week of the intervention (the trial's 2013 paper, Engelen et al., says children in one intervention school were retested after two years and "maintained the gains"; the PLAY paper counts 16 children in that retest and calls its result not significant, and the 2013 paper's abstract mentions no control school in the retest). An Australian loose-parts study in one intervention school and one matched control school (LEAP, not randomised, pedometers and observation) reported effects still present at 8 months. PLAY also added loose parts and found no gain at one or two years, though its playworker's loose-parts score did not differ significantly between intervention and control schools at one year. No trial found here has measured a loose-parts-only change at 12 months or later in more than those 16 children.
What would unlock progress
A way to keep the playground new to children without an adult running it, over a full school year. Loose parts are the existing child-reconfigurable option, and they have shown short-term gains (13 weeks), but past 12 months there are only two measurements: the Sydney project's two-year retest of one intervention school (16 children by the PLAY paper's count), where the trial's authors report the children "maintained the gains", and PLAY, which included loose parts and found no gain. What is left unsolved is the rotation and renewal side: how the set of materials, markings or layouts changes over a year (who swaps what in, how often, from where it is stored) without adult organising time eating a 10-minute recess. The Dutch authors call for "practical, easy-to-implement strategies" fitted to each school; the review's authors propose borrowing the environmental-enrichment principles (complexity, variety, novelty, targeting needs, scaffolding) used in animal research. Digital layers are being tried: a 2025 laboratory study of an augmented-reality scavenger hunt on an indoor playground (17 pairs of 5- to 8-year-olds) found children rated the AR hunts as more enjoyable than the high-structured non-AR hunt and stayed engaged longer in the low-structured AR version, but spent more time in low-intensity postures than without AR, and the authors list the effect of novelty over time as untested. The second unlock is measurement: the review notes most trials had few clusters and says none reported a power calculation (PLAY's own paper does report one), and assessment timing varied widely, which is why the short- versus long-term question is still open.
How we’d know it’s being solved
Recess MVPA (accelerometry, minutes per recess) at the end of a full school year, not at 4–12 weeks, compared with control schools — the measurement that separated short-term from long-term effects in the review. Alongside it: adult minutes needed per recess to run the change (the barrier the Dutch trial points to), the share of recess spent in set-up or waiting, and a play-diversity measure such as an observation tool, so that "different rather than additional activity" can be detected. Beware the proxy that PLAY exposed: staff impressions of more activity were not borne out by accelerometers.
Where to start
Facility-free door: using the open-access review (its study-characteristics table and forest-plot text give each included study's enrichment strategy and assessment time point; its supplementary materials list intervention costs), build a "novelty decay" model — effect size against months since installation and against which enrichment principles each study used — and write it up as a design guide stating which kinds of changes have any evidence of lasting past 12 months. Design door: starting from the Sydney Playground Project and LEAP loose-parts designs, prototype what neither tested, a year-long rotation system (the Sydney project added new objects periodically during its 13 weeks, and LEAP added up to two new types of material a week during its intervention period): stencils or modular floor markings, a loose-parts set with storage and a year-long swap schedule children can run, aimed at near-zero adult set-up, and measure set-up and reset time with adult volunteers acting the role of a lunchtime supervisor; any test with children needs parental consent and ethics review. Partner door (design the trial, don't run it): with a school district, write the protocol for a cluster trial comparing a child-reconfigured playground with a fixed enrichment, with accelerometer-measured recess MVPA at 12 months as the primary outcome, adult minutes per recess as a secondary outcome, and a pre-registered analysis plan. Relevant skills: landscape and product design, human factors, kinesiology, evaluation methods.
Genes
Related problems
Stuck the same way, in other fields.
field: water & energyinvisible water, no treaty468 aquifers cross international borders and countries report only 13 aquifer-specific cooperation arrangements — five of which do not actually operatebuilt without the people who'd live with itsource: primaryfield: making & materialsless copper per tonne, more energy to get itCopper ore grade decline creates compounding energy-emissions processing spiralbuilt without the people who'd live with itsource: secondaryfield: health & medicinethe anchor that doesn't go inAustralia put off a stability rule for furniture that tips onto young children and mandated warning labels instead, leaving prevention to wall anchors that many households cannot fit: they need a stud or masonry tools, sometimes a tradesperson, and in some States a landlord's permission; nine in ten businesses checked in 2025 had at least one product breaking the label rulebuilt without the people who'd live with itsource: primaryfield: water & energywhere the sludge goes, nobody counts43% of the world uses pits and septic tanks, but data on whether the sludge is ever treated covers only 1% of the global population — the faecal-sludge accounting gapbuilt without the people who'd live with itsource: primarySources
- Oppici L, Aadland KN, Aadland E, Li M, Rudd JR. "How can physical enrichment of school playgrounds improve movement behaviours and developmental outcomes in children and adolescents? A systematic review with meta-analysis." The International Journal of Behavioral Nutrition and Physical Activity 22:161, 2025, doi:10.1186/s12966-025-01856-y, accessed 2026-10-22. Link ↗
- Brocken JEA, Meeter M, de Bruijn AGM. "Recess as a Setting for Child Development: A Randomized Controlled Trial on Physical Activity, Motor Skills and Social-Emotional Functioning." The Journal of School Health 96(8):e70198, 2026, doi:10.1111/josh.70198, accessed 2026-10-22. Link ↗
- Farmer VL, Williams SM, Mann JI, Schofield G, McPhee JC, Taylor RW. "The effect of increasing risk and challenge in the school playground on physical activity and weight in children: a cluster randomised controlled trial (PLAY)." International Journal of Obesity 41(5):793-800, 2017, doi:10.1038/ijo.2017.41 (full text read 2026-11-12 from the publisher PDF), accessed 2026-10-22. Link ↗
- Farmer VL, Williams SM, Mann JI, Schofield G, McPhee JC, Taylor RW. "Change of School Playground Environment on Bullying: A Randomized Controlled Trial." Pediatrics 139(5):e20163072, 2017, doi:10.1542/peds.2016-3072 (abstract read via Europe PMC record), accessed 2026-10-22. Link ↗
- Gibson JL, Cornell M, Gill T. "A Systematic Review of Research into the Impact of Loose Parts Play on Children's Cognitive, Social and Emotional Development." School Mental Health 9(4):295-309, 2017, doi:10.1007/s12310-017-9220-9, accessed 2026-10-22. Link ↗
- Stearne SM, Beynon A, Lund Rasmussen C, et al. "Augmented Reality Playgrounds to Promote Physical Activity in Young Children: Feasibility Study Using a Repeated Measures Laboratory Design." JMIR Pediatrics and Parenting 8:e75302, 2025, doi:10.2196/75302, accessed 2026-10-22. Link ↗
- Russell W, Barclay M, Tawil B. "Exploring Play Sufficiency: Reflections and learning from the Exploring Play Sufficiency webinar series," Ludicology and University of Gloucestershire, December 2025, accessed 2026-10-22. Link ↗
- Bundy A, Engelen L, Wyver S, Tranter P, Ragen J, Bauman A, Baur L, Schiller W, Simpson JM, Niehues AN, Perry G, Jessup G, Naughton G. "Sydney Playground Project: A Cluster-Randomized Trial to Increase Physical Activity, Play, and Social Skills." The Journal of School Health 87(10):751-759, 2017, doi:10.1111/josh.12550 (full text read 2026-11-12 from the accepted manuscript), accessed 2026-10-22. Link ↗
- Hyndman BP, Benson AC, Ullah S, Telford A. "Evaluating the effects of the Lunchtime Enjoyment Activity and Play (LEAP) school playground intervention on children's quality of life, enjoyment and participation in physical activity." BMC Public Health 14:164, 2014, doi:10.1186/1471-2458-14-164 (full text read 2026-11-12 from the PubMed Central full text), accessed 2026-10-22. Link ↗
- Ridgers ND, Fairclough SJ, Stratton G. "Twelve-month effects of a playground intervention on children's morning and lunchtime recess physical activity levels." Journal of Physical Activity & Health 7:167-175, 2010, doi:10.1123/jpah.7.2.167 (abstract read via Europe PMC record), accessed 2026-10-22. Link ↗
- Cruciani F, Mitrova Z, Brigoni P, Kalemi T, Masini A, Saulle R. "[Playground marking for contrasting sedentary lifestyle: systematic literature review]." Epidemiologia e Prevenzione 2025, doi:10.19191/ep25.2-3.a838.028 (abstract read via Europe PMC record), accessed 2026-10-22. Link ↗
- Engelen L, Bundy AC, Naughton G, Simpson JM, Bauman A, Ragen J, Baur L, Wyver S, Tranter P, Niehues A, Schiller W, Perry G, Jessup G, van der Ploeg HP. "Increasing physical activity in young primary school children--it's child's play: a cluster randomised controlled trial." Preventive medicine 56(5):319-325, 2013, doi:10.1016/j.ypmed.2013.02.007 (abstract read via Europe PMC record), accessed 2026-11-12. Link ↗
Revision notes
This brief gets checked against its sources and contains the following revisions. Future revisions get added here.
Worth knowing
- the Sydney trial's own authors report that children in one school retested after two years kept their gains; another trial's paper counts that retest at 16 children and calls it not significant
Full check
Sources read
- Oppici et al., meta-analysis of playground enrichment (2025): full text
- Brocken et al., Dutch zoned-playground cluster trial (2026): full text
- Gibson et al., review of loose-parts play (2017): full text
- Stearne et al., augmented-reality playground feasibility study (2025): full text
- Russell, Barclay & Tawil, play sufficiency report (University of Gloucestershire, 2025): full text (the section on OPAL)
- Farmer et al., PLAY cluster trial (International Journal of Obesity, 2017): full text
- Bundy et al., Sydney Playground Project (2017): full text
- Hyndman et al., LEAP loose-parts study (2014): full text
- Engelen et al., Sydney Playground Project trial (Preventive Medicine, 2013): abstract only
- Farmer et al., PLAY trial and bullying (Pediatrics, 2017): abstract only
- Ridgers et al., playground markings study (Journal of Physical Activity and Health, 2010): abstract only
- Cruciani et al., review of playground markings (Epidemiologia e Prevenzione, 2025): abstract only
- Dodd et al., OPAL programme in two English primary schools (preprint, 2026): abstract only
What changed
- Stated more strongly than the source: PLAY was used as evidence about loose parts; its loose-parts score did not differ significantly between intervention and control schools at one year, and the page now says so; the UK markings study's effect is now described as positive but not statistically significant, stronger at 6 months than at 12.
- Contradicted by the source, corrected: the Sydney project's 13-week measurement was called its only follow-up; it was a post-test in the intervention's final week, and a two-year retest of one school also exists, so the claims of no loose-parts evidence past 12 months were narrowed to match; the design starting point said the Sydney and LEAP studies lacked a rotation system; both added new materials during the intervention, so it now says neither tested a year-long rotation.
- Numbers: PLAY's 0.4-minute figure is lunch-time activity at two years; the page now says so and adds that across the school day, intervention children did 3.3 minutes less activity than control children at two years; LEAP was called a single-school study; it compared one intervention school with one matched control school.
Left open
- the PLAY bullying results (abstract only)
- the UK markings study's findings (abstract only)
- the 2025 markings review's findings (abstract only)
- the size, comparison and test of the Sydney two-year retest (abstract only)
Separate check of specific claims: Display title changed from "The Playground Gains That Fade" to "Playground Gains That Don't Show at a Year", because the sources show no gain at a year, not a measured decline.
Separate check of specific claims: The Sydney trial's own paper was added, with its report that one school retested after two years kept its gains, and the review's statement on power calculations is now marked as the review's.
Earlier full check: Four statements were brought in line with the sources: what the review says of novelty and scaffolding, who painted the markings, when the Sydney project measured, and how one non-randomised study looked at decay.
Separate check of specific claims: The full and display titles were toned down to a between-study result, the Dutch trial's one-to-two-days-a-week dose was added, and the write-up now starts from earlier loose-parts trials it had missed.