Genevive Meredith searched fourteen databases, screened 10,917 titles, read 155 in full and kept 14. Ten minutes of sitting or walking in a natural setting was enough to move psychological and physiological markers of wellbeing significantly. Thirteen of the 14 studies used college students, ten were run in Japan, and 22 different measures appear across them, so the effects being pooled are not all the same effect.
Estelle Michinov put 47 employees through a short guided relaxation, 25 of them in a virtual reality headset and 22 watching the same thing as a 2D video. Negative emotions fell at a partial eta squared of 0.70, stress at 0.67 and perceived restoration rose at 0.44, all before and after the break. What the headset added was nothing: there was no effect of the condition and no interaction with it on any of the affective measures. Acceptability and reported user experience were higher in virtual reality, and some participants named trouble with their breathing rhythm and discomfort from the headset. For a company deciding what to buy, the 2D video did the same work.
Robert Meade sat 40 adults aged 64 to 79 through a nine-hour simulated heat wave at a heat index of 37 °C; 20 of them spent hours five and six resting in a room at about 23 °C. At the end of the break their core temperature was 0.8 °C below the control group’s, and an hour after returning to the heat it was still 0.3 °C below, both at p below 0.001. By hours eight and nine the two groups were statistically equivalent, within 0.3 °C. The demand on the heart dropped during the break in the same way and had also evened out by the end. Meade supports the guidance that heat-vulnerable people should go somewhere air-conditioned, and adds a limit the guidance leaves out: the cooling is temporary, at 0.8 °C when the break ends, 0.3 °C an hour later and gone by hour eight.
Glen Kenny had 30 adults aged 18 to 30 and 30 aged 50 to 69, half of each group women, walk on a treadmill in coveralls at a wet-bulb globe temperature of 26 °C until their core temperature reached 38.0 °C, then switched to cycles of 15 minutes rest and 45 minutes work for up to four hours. The time to 38 °C was shorter in the afternoon than in the morning and back to normal the next morning. Women reached it faster than men, and the older women fastest of all. Once the cycles started, mean rectal temperature stayed at or below 38.3 °C in more than 90 per cent of work bouts, with no difference by sex or age. How hot people said they felt did not reliably track what their bodies were doing.
Bin Jiang put 160 people through a standard laboratory stress test and then showed each of them one of ten six-minute videos of a neighbourhood street, chosen at random, with tree canopy ranging from 1.7 to 62 per cent, while salivary cortisol and skin conductance were recorded. For men the curve was an inverted U: recovery improved as canopy rose to 24 per cent, was flat between 24 and 34, and got slower above 34. For women there was no relationship at any density. The quadratic model explains 22 per cent of the variance in the summary stress index.
Meredith pooled other people’s work and rates the base narrow. Jiang ran the largest single experiment here with 160 people, Kenny worked with 60 across a day and a half, Michinov with 47 employees and Meade with 40 older adults. Every one of these breaks was measured in a laboratory or a controlled chamber. None was observed during a working day.