Thomas Astell-Burt and Xiaoqi Feng measured loneliness against hours spent in nature, in a nationally representative sample of 3,043 Australians surveyed in October 2020, with 2,421 answering again after four months and 2,123 after sixteen. Among those lonely at the start, 22.4 per cent were no longer lonely at four months and 29.4 per cent at sixteen. For emotional loneliness the figures were 34.2 and 41.2 per cent. One to two hours a week in nature, against less than one, went with relief from social loneliness at both follow-ups, and three to four hours went with relief from emotional loneliness.
Astell-Burt and Feng also report a null: time in nature was not associated with the incidence of loneliness at all. The relief figures apply only to people who were already lonely in October 2020.
Catharine Ward Thompson surveyed 406 adults in four of the most deprived communities in Scotland and measured the green space around each home objectively rather than by report. Seventy-five per cent of those with a garden or allotment fell into the low-stress group, against 52 per cent of those without, at p below 0.01. The share of green space in the neighbourhood contributed 20 per cent to the model, and the garden itself 5 per cent. But only above 60 per cent green coverage did the neighbourhood share shift people into the lower-stress group at all.
In three of the four communities, social isolation and place belonging predicted stress more strongly than any measure of green space did.
Yiyi Chen followed 11,718 Chinese adults aged 60 and over across three waves of a national panel between 2011 and 2015. A standard deviation more green cover across a city went with 0.052 units less loneliness in the unadjusted model, at a confidence interval of −0.073 to −0.031. After adjusting for demographics, health and city characteristics it held at between −0.023 and −0.020. Social activity accounted for about 2 per cent of the path, and less where air pollution was higher. Yiyi Chen calls the result modest associative evidence.
Tytti Pasanen tested the mechanism across 18 countries in 2017 and 2018, comparing 2,062 urban residents living alone with 6,218 living with a partner. Visiting green space went with better mental wellbeing, and the path ran through satisfaction with relationships and with the community rather than directly. It ran equally strongly for the people living alone.
Only Astell-Burt followed the same people over time with loneliness itself as the outcome. Chen’s is a panel with an exposure measured at city level, Ward Thompson’s a cross-section of 406 people, and Pasanen’s a cross-section of 8,280.