The architecture of mental health: identifying the combination of apartment building design requirements for positive mental health outcomes

Background: Housing quality is a crucial determinant of mental health. While the construction of high-rise buildings is a popular policy strategy for accommodating population growth in cities, there is considerable debate about the health consequences of living in poorly designed apartments. Drawing on three Australian state government apartment design policies introduced to improve apartment design quality, this study aimed to identify the combination of design requirements that were optimally supportive of positive mental health. Methods: K-means cluster analyses identified groups of buildings (n = 172) that were homogenous in their implementation of a mix of n = 80 measured design requirements. Positive mental health was measured using the Warwick–Edinburgh Mental Well-being Scale (WEMWBS). Linear mixed-effects models controlling for demographic characteristics, self-selection factors and clustering of participants within buildings compared residents in the different clusters. Findings: Residents in the "high policy performance buildings", characterised by having a greater implementation of n = 29 design requirements across nine design elements, had significantly higher (+1.96 points) WEMWBS scores compared with residents in the "low policy performance buildings". Interpretation: This study is the first to empirically identify a mix of policy-specific architecture design requirements that are associated with positive mental health in apartment residents. These findings provide vital empirical evidence to inform national and international apartment and high-rise housing policies, and design instruments and practices to protect people's health in apartment dwellings. Funding: The High Life project is funded by a Healthway Research Intervention Project grant (#31986) and an Australian Research Council (ARC), Discovery Early Career Researcher Award (DECRA) (DE160100140). NE is supported by an Australian Research Council (ARC) Linkage Project (LP190100558). SF is supp