Funded projects

Poor mental health is a major public health problem. For example, it is estimated that 27% of the EU adult population experienced poor mental health in the past year; that's 83 million people. These problems are expensive too. In Scotland alone, where this project is based, they are thought to cost £10.7billion. Improving people's mental health and wellbeing is therefore a priority. The problem is we are not sure how to do this when so many people are affected. The environments we live in might be able to help. Seeing and visiting natural environments, such as woodlands and parklands, can be good for health and wellbeing. Not only do these environments make you feel better, they can also benefit your body and mind in measurable ways. In particular, they can reduce stress. We know this from experiments which take people into woodlands or gardens. The results show that they feel better and that those effects last when they return to their normal lives. This is useful, but it's not clear how much they match what happens in people's everyday experience because the results are from scientific experiments, not real life. To find out whether better physical environments can help with this public health problem, we need to explore whether changing the environments in which people live has an impact on their health. In particular, we would like to find out if it can help poorer communities, who often face particularly high levels of stress and mental health problems.

Although there are many woods in and around deprived communities, they are often not easy for local people to enter and use, they lack facilities or information, and ways to enjoy the woods are not well promoted. The Forestry Commission Scotland (FCS) has a programme to work with local people to open these sites up, add better facilities and offer activities like walking groups to encourage use. Our study will evaluate the impacts on health of this programme. We are looking for answers to some specific questions that will tell us how well projects under the programme have worked: do local people become more aware of their nearby woodlands and start to use them more? Do people's stress levels go down? Does this happen equally across the community, or do men benefit more than women, for example? We will use a standard way of measuring how stressed people are; it's been used in other studies before and is good at picking up the effects of different environments on stress. We will also study what has most effect: is it physical change in the woodland environment itself or is it local people becoming involved in activities like organised walks afterwards? Finally, we will also work out whether any effects that the scheme has are good value for money.

The study will take place in central Scotland. The study will work with three deprived communities where local woods are changed, and three where they are not. This comparison between places that experience change in their woodlands and places that don't will allow us to be more sure that any changes we see in the communities' mental health is really due to the forestry scheme. Local residents will be surveyed three times as the FCS work progresses. All data collected are anonymous and will be held securely by the research team; no-one has to take part if they don't want to. We will also carefully monitor what actually changes in the woodlands, and talk with local people to hear their views on the scheme and how is has affected them (or not).

Our team has considerable experience in this kind of work and has already piloted the research in a smaller study with similar communities. The FCS is paying for all changes to the woods and their promotion. We need funding to carry out the surveys, monitor what changes in the woods and the community, and share the results with those responsible for public health, land planning and management, as well as with the communities themselves. Our findings will give organisations and policy makers better information on how to plan future forestry and green space schemes for the health of local communities.

Design, Setting & Target Population:
Following guidance from the MRC Population Health Sciences Research Network, we will treat the WIAT intervention as a natural experiment. In partnership with FCS, we will select 6 woodland sites within the Scottish Lowlands Forest District with associated communities that meet current WIAT inclusion criteria and are in the worst 30% of socio-economic deprivation in Scotland. Woodland sites will not have received investment or direct promotion within the last 2 years. Three intervention sites will receive the WIAT programme between 2012 and early 2015, three control sites matched on woodland and community characteristics will not. The intervention is in two stages: firstly, it makes changes to the physical woodland environment designed to facilitate greater use; secondly, it undertakes community engagement activities to advertise and promote woodland use. We expect the intervention to increase engagement and contact with the woodland, leading to improvements in mental wellbeing, reflected in lower PSS scores. The design draws on i) data from a population-level study in Denmark, which measured associations between PSS and use of green space, ii) our pilot evaluation of an earlier phase of WIAT, focused on change in woodland visits.

The design combines a repeat, cross-sectional survey of individuals resident in intervention and control communities, with three waves of data collection to assess health impacts, and a longitudinal mixed-method study to track the environmental changes in woodlands, and promotional activities which take place. For our quantitative study, we will recruit adults from each of the study communities using a random sample of addresses within 1km of the woodland sites in each community. Our sampling frame will be the postcode address file.

Intervention being evaluated:
The intervention is guided by a woodland development plan, created in partnership with the community. In stage 1, physical changes will be made to improve access to and within the woods (e.g. clearing shrubs, creating paths and adding signage). In stage 2, activities designed to increase awareness and use of the woods by the local community (e.g. led-walk programmes, leafleting and event days) will take place. The same programme will be applied across the 3 intervention sites, though precise detail will be site-specific. WIAT projects involve an initial capital expenditure on environmental improvements of £10k - £100k and, thereafter, approximately £30k per year on management by wardens, rangers and foresters as well as associated maintenance costs.

Measurement of outcomes:
The primary outcome will be a measure of mental wellbeing, assessed using the Perceived Stress Scale (PSS) (35). It has has been used in studies relating natural environments and stress, and is sensitive to change (34). Secondary outcomes as detailed in Section C3 will be measured using well-tested measures of woodland use and experience (36,38 43-45), the Connectedness to Nature Scale (46), the NICE approved, single-item physical activity assessment (47), General Health and Quality of Life (48) and social capital and cohesion (49).

Changes in the nature and quality of the woodland sites will be monitored both independently (by experts surveyors using GIS) and with community-led audits. Community engagement interventions will be assessed through focus groups.

Sample size:
The literature suggests there are likely to be gender differences in the observed effects (25). Based on data from Stigsdotter et al (26), to detect a difference between intervention and control groups, and identify gender differences in effect, we need a total of n=1680 (420 men and women in each group). This would allow us to detect a difference in mean PSS scores of 1.2 with a common standard deviation of 6.2 based on a two-sided, two-sample test with a 5% level of significance, 80% power. We have allowed for a 25% additional sample to take into account any clustering effect. Thus our total sample size at each wave of survey will be 2,100 (1,050 per intervention or control group).

We have not powered the study for further sub-group analysis. We will, however, consider other demographic and personal variables in analysis of the data, which will also take account of the clustering of respondents within sites, the sequential nature of the intervention, and confounders such as life events. Sub group analyses will explore effects by age group, baseline behavioural characteristics and stress levels.

Planned analyses:
As data will be collected via anonymous cross-sectional surveys, we will not be able to link pre and post intervention results from the same individuals. To address the primary research question, we will compare the outcome variables in the baseline with the stage 1 and stage 2 post-intervention periods, for intervention and control sites. By undertaking three separate regressions, one at each wave, we will examine the effect of residence in intervention or control sites on PSS, adjusting for differences on key confounding variables (sex, age, SES, ethnicity, education level, employment status, financial strain, limiting illness and life events). To address secondary research questions, we will adjust the models to control for level of community engagement with the woodlands; run new regressions with different outcomes of interest forming the dependent variable of an appropriate form of regression; examine the difference in coefficients between survey waves 1 and 2, and then 2 and 3; and run models stratified by gender and by distance from home postcode to site. Health economics analysis will initially take a cost-consequence approach, costing each stage of the intervention and comparing those costs to the differences achieved in the intervention group over the control group on primary & secondary outcomes. In the second part, the overall cost-consequence analysis from the first exercise will be extended to a more formal economic appraisal, estimating the likely quality adjusted life year (QALY) benefits that might be expected from the programme.

A Grounded Theory approach will be used in the qualitative work, to explore how people experience and respond to the WIAT activities. Any unexpected positive or negative outcomes will be recorded both from the perspective of the communities and from the FCS staff and partners.

Project timetables:
Stage 1 of the intervention will occur in 2012/13, stage 2 in 2014/early 2015. The first wave survey will establish a baseline in 2012, with the two subsequent waves (in 2014 and 2015) undertaken at the same time of year to minimise the effects of seasonality, and to fall at least 3 months after each intervention stage. Qualitative work will be ongoing at regular intervals throughout the project.

£948,024