Keller pooled four cohorts from Denmark, France and the United Kingdom, 69,136 people surveyed in spring 2020, and put housing conditions against loneliness, anxiety and life satisfaction. Living alone carried an odds ratio of 3.08 for loneliness, at a confidence interval running from 1.87 to 5.07, and 1.27 for lower life satisfaction. Lacking outdoor space and living in a crowded household pointed the same way without reaching the same certainty.
Groot found the same shape in the young. Among 7,445 Danes of median age 20, surveyed at 18 and again three weeks into the first lockdown, living alone rather than with parents more than doubled the odds of loneliness starting, at 2.12. Having no access to outdoor space carried 1.72 for the onset of low mental wellbeing, and a denser household 1.26.
Haena Lee used eight years of the Health and Retirement Study, 1,893 older Americans, with trained observers recording the state of the neighbourhood and the home rather than asking residents about it. The odds of falling were 2.65 where stairways were poorly maintained and 2.10 where there was green space, both at p below 0.05. Both were significant only for the people living alone. Among people living with others, not one environmental risk factor reached significance across the eight years.
Bourion-Bédès surveyed 471 French children from 341 households during confinement. Living in an apartment gave an odds ratio of 1.8 for poor quality of life on the social support and peers dimension, never leaving home 2.6 and indoor noise 2.3. Conflict with the other occupants of the dwelling reached 2.9 for psychological wellbeing, 2.4 for the school environment and 2.2 for parent relations and autonomy.
Two studies describe the supply rather than the people. Phil Hubbard went through English planning appeals since the national space standard was introduced in 2015. Appeals by developers usually fail once a home is 10 per cent or more below it, but where a council has not adopted the standard, inspectors have set it aside: one 2019 decision approved an 18-square-metre flat, more than 50 per cent below the minimum for one person. Blanc and Scanlon report the counter-movement. Londoners aged 16 to 34 living alone fell 10 per cent between 2001 and 2011, from 225,978 to 202,525, while those sharing with other adults rose 28 per cent to 166,588.
Kianoush Vali measured what the dwelling then supplies. Sixteen older adults living alone wore spectral sensors through a winter and a summer. Median daytime melanopic illuminance was 31.2 lux, far below what circadian regulation needs, and the window-to-wall ratio of the building explained only a modest part of it.
Keller’s 69,136 and Groot’s 7,445 are the large samples, both surveyed during lockdown. Lee followed 1,893 people for eight years, the only long run here. Vali worked with 16 people, and Hubbard and Blanc are analyses of decisions and statistics rather than of people.