Health Benefits of Physical Activity Related to an Urban Riverside Regeneration

The promotion of physical activity through better urban design is one pathway by which health and well-being improvements can be achieved. This study aimed to quantify health and health-related economic impacts associated with physical activity in an urban riverside park regeneration project in Barcelona, Spain. We used data from Barcelona local authorities and meta-analysis assessing physical activity and health outcomes to develop and apply the “Blue Active Tool”. We estimated park user health impacts in terms of all-cause mortality, morbidity (ischemic heart disease; ischemic stroke; type 2 diabetes; cancers of the colon and breast; and dementia), disability-adjusted life years (DALYs) and health-related economic impacts. We estimated that 5753 adult users visited the riverside park daily and performed different types of physical activity (walking for leisure or to/from work, cycling, and running). Related to the physical activity conducted on the riverside park, we estimated an annual reduction of 7.3 deaths (95% CI: 5.4; 10.2), and 6.2 cases of diseases (95% CI: 2.0; 11.6). This corresponds to 11.9 DALYs (95% CI: 3.4; 20.5) and an annual health-economic impact of 23.4 million euros (95% CI: 17.2 million; 32.8 million). The urban regeneration intervention of this riverside park provides health and health-related economic benefits to the population using the infrastructure.


Index
 Table S1. Questions included in the survey conducted by Barcelona local authorities in 2014-2015  Table S2. Exposure-response function for each outcome  Table S3. List of assumptions considered for the assessment of health and health-related economic benefits of the urban riverside regeneration project  Table S4. Quartiles of basal levels of physical activity reported in Barcelona Health Survey  Table S5. Direct health-care cost for morbidity outcomes in Spain  Table S6. Descriptive analysis of the study population  Table S7. Sensitivity analysis using the minimum visit duration to the Park reported by walking commuter users  Table S8. Sensitivity analysis considering the minimum visit duration to the Park reported by cyclists >65 years old  Figure S1. Exclusion criteria used to define the sample of survey respondents for analysis in this study  Figure S2. Non-linear exposure-response function between physical activity in METhour/week and relative risks for all-cause mortality  Figure S3. Formulae for the calculation of the relative risk used to estimate the health benefits for our study population. These surveys were administered to Park users [1]. This is an adapted version of the original survey (in Catalan) and includes the key assumptions used for the model. report questions that have been used for this study. a,b "To be healthy" and "Others" are not physical activity categories. Thus, users who answered any of these two options were excluded of the study sample. c In order to use these values in the analysis, we assumed that: "more than 3 days/week"="5.5 days/week" [considering that the maximum expected days/week would be 7, and the minimum expected days/week would be 4. Thus, the mean of these values would be: 5.5 = (7+4)/2]; "1 day/week"="1.0 day/week"; "1 day/month"="0.25 days/week" (0.25 = 1 day/ 4 weeks); "occasionally"="0.0 days/week". We assumed these values being as conservative as possible. d In order to use these values in the analysis, we assumed: "less than 1 hour"="0.5 hours/day"; "1-2 hours"="1.0 hours/day"; "2-4 hours"="2.0 hours/day"; ">4.0 hours"="4 hours/day". We assumed these values being as conservative as possible. Exposure-response functions for ischemic stroke and dementia were available for subjects ≥65 years old. The study population was divided by age groups (18 to 64 years old, and ≥65 years old), with the aim of assigning appropriate age-specific incidence rates and exposure-response functions for each health outcome. Table 3. List of assumptions considered for the assessment of health and health-related economic benefits of the urban riverside regeneration project.

Assumption Justification
 The sample of survey respondents is representative of the study population. The survey was conducted by Barcelona local authorities and used in this study to estimate health and health-related economic benefits of the urban riverside regeneration project.
 Data and description of the procedure employed by local authorities to conduct the surveys was not available. However, this is official data which is being used by local authorities to assess the usability of the Besòs Riverside Park. To our knowledge, this is the only official data available at this moment.
 Scenario 1: 100% of the physical activity practised in the Besòs Riverside Park is new and related to the study intervention. Scenario 2: 50% of the physical activity practised in the Besòs riverside park is new and related to the study intervention (considered in the analysis), the other 50% was previously conducted somewhere else (e.g. on the beach, in a park, in the gym, etc) (not considered in the analysis).
 Data on the physical activity behaviour of the users of the Besòs Riverside Park before the intervention was not available. However, park infrastructure, including access to the riverbanks did not exist previously and the people could not use the area before the intervention completion.
 Base levels of physical activity of the study population are similar than those reported for the whole population of Barcelona [5,8]  Data on the base levels of physical activity of the specific study population was not available. Nevertheless, this data was available at city level and it was expected to be similar among the Barcelona population and the study population.

 Survey data on duration of the visits to the Park (Supplementary Material -
 Data from 2014-2015 surveys (number of users, duration, frequency and type of physical activity) is assumed to be constant though the time  Health benefits of physical activity do not emerge instantaneously and require regular practise. Data used for this study was collected 15 years after the completion of the riverside park, which means that the users who were using the infrastructure at this moment, may be users who have been using it for some time and that might continue using it in the future.       [1]. b Users who responded "to be healthy" or "others" as the "reason to come to the Besòs River" in the survey (see Supplementary Material -Table S1), were excluded of the sample of this study. This was because these activities could not be classified in a physical activity category according to the physical activity classification described by Ainsworth et al. 2011 [9]. This classification provides the energy cost of a wide variety of physical activities (e.g. dancing, walking, cycling, doing home activities like mopping or cleaning windows, etc.), which can be compared with other epidemiological studies providing data on self-reported physical activity. Figure S2. Non-linear exposure-response function between physical activity in METs hour/week and relative risks (RR) for all-cause mortality. Data obtained from a meta-analysis [4] including 22 studies. Shaded areas represent 95% confidence intervals.   Table S2. C = METs minutes/week basal value. Base levels of physical activity [5,8]. D = METs minutes/week basal + Scenario value. Base levels of physical activity [5,8] + estimated physical activity levels of the study population (Table 1). Y = power transformation of 0.25 [4].