Green spaces are recognized for improving mental health, but what particular kind of nature is required is yet not elucidated. This study explores the effect of specific types of recreational nature qualities on mental health. Longitudinal data (1999/2000 and 2005) from a public health survey was distributed to a stratified sample (n = 24,945) of a Swedish population. People from rural or suburban areas (n = 9230) who had moved between baseline and follow-up (n = 1419) were studied. Individual geographic residence codes were linked to five predefined nature qualities, classified in geographic information systems (GIS). Any change in the amount of or type of qualities within 300 m distance between baseline and follow-up was correlated to any change in mental health (as measured by the General Health Questionnaire) by logistic regression models. On average, the population had limited access to nature qualities both pre- and post-move. There was no significant correlation between change in the amount of qualities and change in mental health. However, the specific quality "serene" was a significant determinant with a significantly decreased risk for women of change to mental ill-health at follow-up. The objective definition of the potentially health-promoting quality may facilitate implication in landscape practice and healthy planning.
Keywords: GIS; longitudinal; mental health; nature type; public health survey; recreation; salutogenic; wellbeing.