Reference
Beyer KM, Kaltenbach A, Szabo A, Bogar S, Nieto FJ, Malecki KM. Exposure to neighborhood green space and mental health: evidence from the survey of the health of Wisconsin. Int J Environ Res Publ Health. 2014;11(3):3453-3472.
Design
Cross-sectional analysis comparing mental health status with neighborhood green space (NGS)
Participants
Residents of the state of Wisconsin who had previously completed the Survey of the Health of Wisconsin (SHOW) in 1 of the years between 2008 and 2011 (N=2,479). SHOW is an annual statewide probability population survey of the health status of Wisconsin state residents as determined by individual interview, physical exam, and biomarker collection.
Study Exposure
Exposure to NGS was determined by geomatching each participant’s residential address to his/her local US census block, which was assessed according to 3 different green space exposures:
- Neighborhood greenness, as determined by the 2009 Normalized Difference Vegetation Index (NDVI) dataset, a commonly used and validated database of neighborhood-level land-use type1;
- Neighborhood tree canopy coverage (NTCC), as determined by the 2001 National Land Cover Database; and
- an average of the above 2 measures (AVG).
Exposure conditions were dichotomized into NGS with areas of >10% or <10% tree canopy, as 10% tree canopy was the median value for the census blocks included in the study.
Outcome Measures
Mental health status was assessed by participant responses to the Depression, Anxiety and Stress Scales (DASS) instrument, a validated survey tool used by SHOW for measuring aspects of mental health status.2 DASS subscores are determined by Likert-type scales (0-3↑), each with 14 items, for a maximum score of 42. DASS subscores were each included as continuous variables in a multivariate linear regression model.
For every 25% increase in amount of neighborhood green space, a participant’s depression score was likely to be 1.0 points to 1.4 points lower after controlling for all the potential confounding factors.
To control for possible confounders, both individual-level (ie, age, gender, race/ethnicity, marital status, education level, annual household income, occupational status, residence housing type, type of health insurance) and neighborhood-level (ie, level of urbanicity, median household income, residential instability, percentage below poverty level, percentage of homeowners vs renters, percentage of unemployment, percentage of African American) factors were included in the regression analysis.
Key Findings
NGS significantly predicted DASS scores for all 3 measures of mental health. Depression scores were the most influenced by NGS, with NDVI, NTCC, and AVG all significantly negatively predicting depression (P<0.01). The regression demonstrated that for every 25% increase in amount of NGS, a participant’s depression score was likely to be 1.0 points to 1.4 points lower after controlling for all of the potential confounding factors.
Stress scores were also influenced by NGS by about half as much as depression scores. A 25% increase in NGS predicted a lower stress score by 0.5 points to 0.7 points (P<0.05), though this was only statistically significant for NTCC and AVG models.
Anxiety was also influenced by NGS, though only NDVI and AVG models were statistically significant (P<0.05). A 25% increase in NGS using these measures predicted 0.4 points to 0.5 points lower on the DASS anxiety subscale.
Analysis of the potential confounder variables revealed some statistically significant impacts on DASS scores, independent of NGS. In particular, low levels of income and higher rates of unemployment were significant predictors of all 3 DASS subscale measures (P<0.05), regardless of exposure to NGS.
Commentary
This paper is the latest in a surge of recent research publications looking at the relationship between green space and mental health.3 The awareness that natural environments have restorative and therapeutic purposes has been utilized since antiquity and is present in all known human civilizations.4 In the late 19th century, the founders of naturopathic medicine understood well that “the healing power of nature” was present as much in the world around us as it was in the food or herbs we consume or in the water we apply to our skin.5
It is not necessary to describe to the readers of this journal how society’s concept of health moved away from this holistic understanding to the more reductionistic, biomechanistic model that predominates today’s health landscape. Instead, what is needed is a reawakening of the realization that environmental exposures, such as the prevalence of NGS, are components in the greater milieu of total health determination, especially in this modern era.
Since Roger Ulrich first published his famous “room with a view” study in a 1984 issue of Science,6 we have had empirical evidence demonstrate that our surrounding environments directly influence both physical and mental health outcomes. Ulrich’s famous theory, now dubbed the psycho-evolutionary stress (PES) theory, describes a neuroaffective model for how environmental stimuli influence the psychophysiological cascade of events known as the “stress response.”7 Given that stress is a truly holistic process affecting all aspects of health and well-being,8 it is no surprise that the restorative influence of NGS can affect the severity of depression and anxiety reported by Wisconsin state residents.
It is worthwhile to note that the authors of this study specifically included measures of socioeconomic status (SES) such as income, education, and employment status in their analyses. SES has been a well-established social determinant of health for over a decade,9 and it is unsurprising that SHOW participants with low SES status had correspondingly higher DASS scores. Similarly, it is well-known that lower SES individuals and neighborhoods typically have less NGS; for the most part, richer neighborhoods are greener neighborhoods and poorer neighborhoods are less green neighborhoods.
While the authors did not analyze their data for an interaction between SES and NGS and the effects on mental health status, other similar studies have demonstrated such a relationship. A series of studies in Scotland have shown that people living in economically deprived communities have a lower burden of stress when NGS levels are increased, as measured by both self-report and salivary cortisol biomarkers.10,11 This evidence supports Ulrich’s PES theory and is potentially valuable for utilizing NGS to address the “wear and tear” effect that chronic poverty has on health via accumulated negative stress effects known as “allostatic load.”12 (As an aside, the PES/allostatic-load hypothesis was further supported by a now-classic geospatial analysis of British National Health Service data, which showed that in the whole of England [N=40,813,236], people with lower SES were more positively influenced by the restorative powers of NGS than were other higher SES groups.13)
All this information shows that the world around us influences our health in ways of which we may not be aware. In addition to the obvious implications for public health, urban planning, and environmental management, this research has applications for the clinical practitioner. First, it is important to include questions regarding a patient’s home environment in every first-time visit, in order to establish (among other things) the stressors and/or restorative opportunities to which patients are regularly exposed. Second, this information along with many other studies in this area14 suggest that frequent exposure to green space could be a preventive and even therapeutic intervention (in conjunction with other appropriate care) for patients with mental health concerns. Third, it reminds us as practitioners that if a patient’s home environment is detrimentally affecting his or her well-being, it may be prudent to apply the principle of tolle causum and suggest that the patient relocate to a healthier area if it’s in his or her economic means to do so.
