Philippa Howden-Chapman ran the only randomised controlled trial in this collection. Across seven low-income communities in New Zealand, 1,350 households containing 4,407 people were randomised by cluster, single-blinded, to receive a standard retrofit insulation package or not. The physical effect was modest: bedroom temperatures rose 0.5 °C in winter, relative humidity fell 2.3 per cent, and bedrooms spent 1.7 fewer hours a day below 10 °C, while insulated houses used 81 per cent of the energy the uninsulated ones did. The health effects were larger. Self-rated health was an adjusted odds ratio of 0.50, wheezing in the past three months at 0.57, children taking a day off school at 0.49, adults taking a day off work at 0.62 and visits to a general practitioner at 0.73. Hospital admissions for respiratory conditions fell to 0.53 and that reduction was not statistically significant, at p = 0.16, which the paper reports as a trend and not a result.
Jan Sundell and thirteen colleagues from different disciplines reviewed the literature to 2005 and judged 27 peer-reviewed papers to carry enough information on both ventilation rates and health effects to be usable. Consistency held across different investigations, different epidemiological designs and different populations, and several health endpoints showed the same relationship with ventilation rate. In offices, higher ventilation up to about 25 litres per second per person is associated with a lower prevalence of sick building syndrome symptoms. The review is explicit about what it cannot say: the literature gives no clear evidence on which agent in the air produces the effect.
Jenni Ervasti linked the recorded sick leaves of 1,678 teachers to what pupils in 92 Finnish secondary schools said about the air, in 2001 and again in 2004. Where the pupils called the air good at both points, teachers had an odds ratio of 0.6 for short sick leaves of one to three days, against schools where it was consistently poor. An improvement in the air gave the same 0.6. Worsening pupil satisfaction gave 1.8 and more bullying 1.5.
Amanjeet Singh followed two workforces through a move into certified green buildings and asked them before and after. Work hours a month affected by asthma and respiratory allergies fell from 16.28 to 6.32, and hours affected by depression and stress from 20.21 to 14.06. The numbers answering fell from 46 to 33 across the same period.
Juan Palacios had a natural experiment when 70 per cent of a Dutch municipality’s workforce moved into a building designed for health and sustainability. Perceived conditions improved and sick building syndrome symptoms fell, and a mediation analysis traces a path from health and wellbeing through job satisfaction to a 2 per cent lower prevalence of sick leave. The effect was still there two years later and was largest among employees over 50.
Howden-Chapman’s 4,407 participants are the only randomised trial in this collection and the only one with registered health outcomes rather than reported ones. Ervasti’s 1,678 teachers were also matched to records rather than asked. Sundell pools 27 papers. Forty-six people answered Singh’s first survey and 33 the second. Palacios gives the share of a workforce that moved and not a number of people.