Forests, health and inequalities in Scotland: a longitudinal approach

Abstract

Increasing international evidence shows that forests may enhance mental and physical health by providing opportunities for relaxation, physical activity, social interaction and through reducing air pollution. Studies also suggest that forests may have a role in reducing socioeconomic health inequalities by weakening the links between poverty, deprivation and poor health. Knowledge surrounding the relationship between forests, health and inequalities is limited as no national studies have been carried out, and findings to date are based on cross-sectional data. This thesis addresses these research gaps by examining associations between forests, health and inequalities for the whole of Scotland over a 20-year period. Firstly, changes in the socio-spatial distribution of forests in Scotland between 1991, 2001 and 2011 were assessed. Following this, relationships between different long-term patterns of individuals’ forest access and subsequent health outcomes were examined. The influence of cumulative forest access throughout life and levels of forest access at particular life stages on later mental health were also studied. Lastly, investigations into whether changes in forest access were associated with changes in general health were carried out. In order to understand whether forests might reduce socioeconomic health inequalities, each of the empirical analyses considered differences between sociodemographic groups. Measures of forest access in 1991, 2001 and 2011 were created in ArcGIS for all postcodes in Scotland and linked to a sample of individuals in the Scottish Longitudinal Study (SLS). The SLS contains linked census records collected in 1991, 2001 and 2011 for approximately 274,000 people (5.3% of the population). The study sample included those who had: complete data; were present in all three censuses; were aged 18+ in 1991; and lived in private residences on the Scottish mainland (n=97,658). Administrative health records from 2011 to 2016, including the prescribing of antidepressants and hospital admission data were linked to the sample members. A synthetic estimation of forest use based on SLS members’ characteristics and forest user information in the Scottish People and Nature Survey (SPANS) was also used to examine whether visiting forests explained the associations between forests and general health. Statistical techniques included Latent Class Growth Modelling (LCGM), hybrid effects models and tests for mediation. Over the study period, geographical access to forests improved throughout Scotland. However, there was evidence that individuals with low socioeconomic status in 1991 were more likely to have worse long-term patterns of forest access than those with higher socioeconomic status. There was evidence that these worse trajectories of forest access had implications for later health; individuals with better forest access trajectories had reduced risk of having worse health at the end of the study period. Women with a greater accumulation of forest access were less likely to attend a mental health outpatient clinic or be prescribed antidepressants during 2011-2016. For men and those without qualifications who had improved forest access between time points, the risk of having a long-term illness reduced, compared to those whose forest access did not change. Findings also suggested that better forest access across the life course and at particular stages in adulthood may be linked to reduced inequalities in mental health between men and women and between those with higher and lower socioeconomic status. Forest use partially explained the association between forest access and general health but there was also evidence of a direct effect of forest access on mental health. The key contribution of this thesis was the linkage of spatial environmental data to census and administrative health records for individuals and the application of a longitudinal approach. The thesis also contributes to the international literature by providing new insights into the causal mechanisms though which forests may influence health across the life course and how these may vary between social groups. The research has provided important evidence for policy makers such as Forestry Commission Scotland, about the social value of forestry in Scotland (and potentially elsewhere) and the opportunities that maintaining and enhancing forest access could have for improving population-level mental health and reducing health inequalities. In particular, those designing interventions to encourage forest use among disadvantaged groups should consider how interventions could be targeted at those with low individual-level socioeconomic status as well as deprived areas. Future research should use life course approaches to better specify the ways in which forests may support health for those with specific mental illnesses, and where possible consider the effect of forest access in childhood as well as adulthood on later life health outcomes.