Angel Dzhambov1, Terry Hartig2, Iana Markevych3, Boris Tilov4, Donka Dimitrova5. 1. Department of Hygiene and Ecomedicine, Faculty of Public Health, Medical University of Plovdiv, Plovdiv, Bulgaria. Electronic address: angelleloti@gmail.com. 2. Institute for Housing and Urban Research and Department of Psychology, Uppsala University, Uppsala, Sweden. 3. Institute and Clinic for Occupational, Social and Environmental Medicine, University Hospital, LMU Munich, Munich, Germany; Institute of Epidemiology I, Helmholtz Zentrum München - German Research Center for Environmental Health, Neuherberg, Germany. 4. Medical College, Medical University of Plovdiv, Plovdiv, Bulgaria; Department of Management, Faculty of Economics and Management, University of Agribusiness and Rural Development, Plovdiv, Bulgaria. 5. Department of Health Management and Healthcare Economics, Faculty of Public Health, Medical University of Plovdiv, Plovdiv, Bulgaria.
Abstract
BACKGROUND: Urban greenspace can benefit mental health through multiple mechanisms. They may work together, but previous studies have treated them as independent. OBJECTIVES: We aimed to compare single and parallel mediation models, which estimate the independent contributions of different paths, to several models that posit serial mediation components in the pathway from greenspace to mental health. METHODS: We collected cross-sectional survey data from 399 participants (15-25 years of age) in the city of Plovdiv, Bulgaria. Objective "exposure" to urban residential greenspace was defined by the Normalized Difference Vegetation Index (NDVI), Soil Adjusted Vegetation Index, tree cover density within the 500-m buffer, and Euclidean distance to the nearest urban greenspace. Self-reported measures of availability, access, quality, and usage of greenspace were also used. Mental health was measured with the General Health Questionnaire. The following potential mediators were considered in single and parallel mediation models: restorative quality of the neighborhood, neighborhood social cohesion, commuting and leisure time physical activity, road traffic noise annoyance, and perceived air pollution. Four models were tested with the following serial mediation components: (1) restorative quality → social cohesion; (2) restorative quality → physical activity; (3) perceived traffic pollution → restorative quality; (4) and noise annoyance → physical activity. RESULTS: There was no direct association between objectively-measured greenspace and mental health. For the 500-m buffer, the tests of the single mediator models suggested that restorative quality mediated the relationship between NDVI and mental health. Tests of parallel mediation models did not find any significant indirect effects. In line with theory, tests of the serial mediation models showed that higher restorative quality was associated with more physical activity and more social cohesion, and in turn with better mental health. As for self-reported greenspace measures, single mediation through restorative quality was significant only for time in greenspace, and there was no mediation though restorative quality in the parallel mediation models; however, serial mediation through restorative quality and social cohesion/physical activity was indicated for all self-reported measures except for greenspace quality. CONCLUSIONS: Statistical models should adequately address the theoretically indicated interdependencies between mechanisms underlying association between greenspace and mental health. If such causal relationships hold, testing mediators alone or in parallel may lead to incorrect inferences about the relative contribution of specific paths, and thus to inappropriate intervention strategies.
BACKGROUND: Urban greenspace can benefit mental health through multiple mechanisms. They may work together, but previous studies have treated them as independent. OBJECTIVES: We aimed to compare single and parallel mediation models, which estimate the independent contributions of different paths, to several models that posit serial mediation components in the pathway from greenspace to mental health. METHODS: We collected cross-sectional survey data from 399 participants (15-25 years of age) in the city of Plovdiv, Bulgaria. Objective "exposure" to urban residential greenspace was defined by the Normalized Difference Vegetation Index (NDVI), Soil Adjusted Vegetation Index, tree cover density within the 500-m buffer, and Euclidean distance to the nearest urban greenspace. Self-reported measures of availability, access, quality, and usage of greenspace were also used. Mental health was measured with the General Health Questionnaire. The following potential mediators were considered in single and parallel mediation models: restorative quality of the neighborhood, neighborhood social cohesion, commuting and leisure time physical activity, road traffic noise annoyance, and perceived air pollution. Four models were tested with the following serial mediation components: (1) restorative quality → social cohesion; (2) restorative quality → physical activity; (3) perceived traffic pollution → restorative quality; (4) and noise annoyance → physical activity. RESULTS: There was no direct association between objectively-measured greenspace and mental health. For the 500-m buffer, the tests of the single mediator models suggested that restorative quality mediated the relationship between NDVI and mental health. Tests of parallel mediation models did not find any significant indirect effects. In line with theory, tests of the serial mediation models showed that higher restorative quality was associated with more physical activity and more social cohesion, and in turn with better mental health. As for self-reported greenspace measures, single mediation through restorative quality was significant only for time in greenspace, and there was no mediation though restorative quality in the parallel mediation models; however, serial mediation through restorative quality and social cohesion/physical activity was indicated for all self-reported measures except for greenspace quality. CONCLUSIONS: Statistical models should adequately address the theoretically indicated interdependencies between mechanisms underlying association between greenspace and mental health. If such causal relationships hold, testing mediators alone or in parallel may lead to incorrect inferences about the relative contribution of specific paths, and thus to inappropriate intervention strategies.
Authors: Alexander Wray; Gina Martin; Emma Ostermeier; Alina Medeiros; Malcolm Little; Kristen Reilly; Jason Gilliland Journal: Health Promot Chronic Dis Prev Can Date: 2020-04 Impact factor: 3.240
Authors: Wei-Lun Tsai; Raquel A Silva; Maliha S Nash; Ferdouz V Cochran; Steven E Prince; Daniel J Rosenbaum; Aimee A D'Aloisio; Laura E Jackson; Megan H Mehaffey; Anne C Neale; Dale P Sandler; Timothy J Buckley Journal: Environ Res Date: 2020-01-30 Impact factor: 6.498
Authors: I Gusti Ngurah Edi Putra; Thomas Astell-Burt; Dylan P Cliff; Stewart A Vella; Xiaoqi Feng Journal: Soc Psychiatry Psychiatr Epidemiol Date: 2022-01-17 Impact factor: 4.328
Authors: Eun-Young Lee; Louise de Lannoy; Lucy Li; Maria Isabel Amando de Barros; Peter Bentsen; Mariana Brussoni; Tove Anita Fiskum; Michelle Guerrero; Bjørg Oddrun Hallås; Susanna Ho; Catherine Jordan; Mark Leather; Greg Mannion; Sarah A Moore; Ellen Beate Hansen Sandseter; Nancy L I Spencer; Susan Waite; Po-Yu Wang; Mark S Tremblay Journal: Int J Behav Nutr Phys Act Date: 2022-06-15 Impact factor: 8.915