Environment International 134 (2020) 105132 Contents lists available at ScienceDirect Environment International journal homepage: www.elsevier.com/locate/envint Changing the urban design of cities for health: The superblock model T a,b,c a,b,c,d a,b,c,e,f a,b,c Natalie Mueller , David Rojas-Rueda , Haneen Khreis , Marta Cirach , David Andrésg, Joan Ballestera,b,c, Xavier Bartollh,i, Carolyn Dahera,b,c, Anna Delucaa,b,c, Cynthia Echaveg, Carles Milàa,b,c, Sandra Márqueza,b,c, Joan Paloug, Katherine Pérezh,i, Cathryn Tonnea,b,c, Mark Stevensonj, Salvador Ruedag, Mark Nieuwenhuijsena,b,c, ⁎ a ISGlobal, Barcelona, Spain b Universitat Pompeu Fabra (UPF), Barcelona, Spain c CIBER Epidemiología y Salud Pública (CIBERESP), Madrid, Spain d Department of Environmental and Radiological Health Sciences, Colorado State University, Fort Collins, CO, USA e Texas A&M Transportation Institute (TTI), College Station, TX, USA f Center for Advancing Research in Transportation Emissions, Energy, and Health (CARTEEH), College Station, TX, USA g Agència d'Ecologia Urbana de Barcelona (BCNEcologia), Barcelona, Spain h Agència de Salut Pública de Barcelona (ASPB), Barcelona, Spain i Institut d'Investigació Biomédica (IIB Sant Pau), Barcelona, Spain j Melbourne School of Design/Melbourne School of Population and Global Health, University of Melbourne, Australia ARTICLE INFO ABSTRACT Handling Editor: Zorana Jovanovic Background: Car-dependent city planning has resulted in high levels of environmental pollution, sedentary AndersenKeywords: lifestyles and increased vulnerability to the effects of climate change. The Barcelona Superblock model is an Active transport innovative urban and transport planning strategy that aims to reclaim public space for people, reduce motorized Health impact assessment transport, promote sustainable mobility and active lifestyles, provide urban greening and mitigate effects of Public space climate change. We estimated the health impacts of implementing this urban model across Barcelona. Sustainability Transport planning Methods: We carried out a quantitative health impact assessment (HIA) study for Barcelona residents ≥20 years Urban planning (N = 1,301,827) on the projected Superblock area level (N = 503), following the comparative risk assessment methodology. We 1) estimated expected changes in (a) transport-related physical activity (PA), (b) air pollution (NO2), (c) road traffic noise, (d) green space, and (e) reduction of the urban heat island (UHI) effect through heat reductions; 2) scaled available risk estimates; and 3) calculated attributable health impact fractions. Estimated endpoints were preventable premature mortality, changes in life expectancy and economic impacts. Results: We estimated that 667 premature deaths (95% CI: 235–1,098) could be prevented annually through implementing the 503 Superblocks. The greatest number of preventable deaths could be attributed to reductions in NO2 (291, 95% PI: 0–838), followed by noise (163, 95% CI: 83–246), heat (117, 95% CI: 101–137), and green space development (60, 95% CI: 0–119). Increased PA for an estimated 65,000 persons shifting car/motorcycle trips to public and active transport resulted in 36 preventable deaths (95% CI: 26–50). The Superblocks were estimated to result in an average increase in life expectancy for the Barcelona adult population of almost 200 days (95% CI: 99–297), and result in an annual economic impact of 1.7 billion EUR (95% CI: 0.6–2.8). Discussion: The Barcelona Superblocks were estimated to help reduce harmful environmental exposures (i.e. air pollution, noise, and heat) while simultaneously increase PA levels and access to green space, and thereby provide substantial health benefits. For an equitable distribution of health benefits, the Superblocks should be Abbreviations: BCNEcologia, Barcelona Urban Ecology Agency; CadnA, Computer Aided Noise Abatement Software; CI, Confidence interval; ERF, Exposure response function; HIA, Health impact assessment; IPAQ, International Physical Activity Questionnaire; Lden, EU day-evening-night noise indicator with 5 dB and 10 dB weights for the evening and night time, respectively; LE, Life expectancy; METs, Metabolic equivalents of task; NO2, Nitrogen dioxide; NDVI, Normalized Difference Vegetation Index; PA, Physical activity; PAF, Population attributable fraction; PI, Prediction interval; PM2.5, Particular matter with diameter ≤ 2.5 μm; RR, Relative risk; SDGs, Sustainable Development Goals; TRAP, Traffic-related air pollutants; UHI, Urban heat island; VSL, Value of a statistical life; WHO, World Health Organization; %GS, Percentage green space ⁎ Corresponding author at: ISGlobal, Dr. Aiguader 88, 08003 Barcelona, Spain. E-mail address: mark.nieuwenhuijsen@isglobal.org (M. Nieuwenhuijsen). https://doi.org/10.1016/j.envint.2019.105132 Received 6 May 2019; Received in revised form 20 August 2019; Accepted 26 August 2019 Available online 09 September 2019 0160-4120/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/). N. Mueller, et al. Environment International 134 (2020) 105132 implemented consistently across the entire city. Similar health benefits are expected for other cities that face similar challenges of environmental pollution, climate change vulnerability and low PA levels, by adopting the Barcelona Superblock model. 1. Introduction accommodating pedestrians, horse-drawn carriages and public tram lines. Nevertheless, Barcelona's streets became filled with concrete Worldwide, car-dependent city planning has resulted in low levels structures, cars and traffic (Barcelona City Council, 2018). Conse- of physical activity (PA) and high levels of environmental pollution (i.e. quently, air and noise pollution levels are high, persistently exceeding air pollution, noise, and anthropogenic heat) (Nieuwenhuijsen and WHO limits, and are estimated to cause a large health burden (Mueller Khreis, 2016). Contemporary cities devote up to 70% of public space to et al., 2017a, 2017b). The city's densely-constructed urban design, ac- accommodate motor vehicles (Crawford, 2002; Manville and Shoup, commodating > 1.6 million people on 100 km2, leaves little available 2005), whereas no more than 25% is suggested for a sustainable design space for green and public open space, and amplifies the generation of (Barcelona Urban Ecology Agency, 2018; Dávalos et al., 2016). Com- anthropogenic heat. Temperature in the city center can be up to 8 °C paratively little space in cities is assigned to public open and green higher in comparison with less urbanized surrounding areas because of space. Reclaiming these spaces, however, for recreational (e.g. PA the urban heat island (UHI) effect (Moreno-Garcia, 1994). performance) and community activities, would add aesthetical appeal, To recuperate Cerdà's progressive design and remedy the negative and could provide urban resilience and climate change adaptation effects of the current situation, the Barcelona Superblock model has through the provision of eco-system services (i.e. passive air and noise been proposed: an innovative land use intervention that aims to reclaim pollution control, cooling through shading and evapo-transpiration of space for people, reduce motorized transport, promote sustainable water) (Nieuwenhuijsen et al., 2017; Wolf and Robbins, 2015). mobility and active lifestyles, provide urban greening and mitigate the The original plan for the ‘extension’ of the city of Barcelona (i.e. effects of climate change (Rueda, 2018). A total of 503 Superblocks, Eixample), by 19th century progressive Catalan urban planner Ildelfons stretching over the city of Barcelona, have been developed by the Urban Cerdà, considered the human needs for natural lighting, ventilation, Ecology Agency (BCNEcologia), a public consortium integrated into the open space and greenery, and a transport network fairly Barcelona City Council. Superblocks are constructed cells transforming Fig. 1. Barcelona's 503 proposed Superblocks. 2 N. Mueller, et al. Environment International 134 (2020) 105132 the city into sustainable (and healthy), compact and connected neigh- less congested, because of avoided turns into the Superblocks borhoods with a mixed land use, and high potential for social capital. (Barcelona City Council, 2014; Rueda, 2018). In addition to re- The Superblock model also works towards achieving the ambitious configuration of transport, liberation and re-allocation of public space targets set by the Sustainable Development Goals (SDGs) that define are planned: the Superblock model foresees the development of public sustainable city and community development in SDG 11 as a pressing open and green space throughout the city, consisting of plazas, parks, issue and leverage point to overcome global challenges related to green corridors, green patches and general greening in and outside the poverty, inequality, climate, environmental degradation, prosperity, Superblocks (Fig. 3 and Fig. 4). peace, and justice (United Nations, 2015). While by mid 2019 the Barcelona City Council has implemented 2.2. Health impact assessment three Superblocks (Poblenou, Sant Antoni, Horta neighborhoods) (Fig. 1), and is committed to six more (http://ajuntament.barcelona. A quantitative health impact assessment (HIA) was carried out for cat/superilles/es/), the present study aimed at estimating the health the Barcelona population ≥20 years (N = 1,301,827; 2017) (Table 1) impacts related to the potential implementation of all 503 proposed (Barcelona City Council, 2017a), at the projected Superblock area level Superblocks. We estimated the impact on preventable premature mor- (N = 503) (Fig. 1). Anticipated changes in the urban form and transport tality, changes in life expectancy (LE) and economic impacts based on system are expected to result in changes of (a) transport-related PA, (b) expected changes in urban and transport planning related exposures of air pollution, (c) road traffic noise, (d) green space, and (e) mitigation (a) transport-related PA, (b) air pollution, (c) road traffic noise, (d) of the UHI effect through heat reductions. urban green space, and (e) mitigation of the urban heat island (UHI) Preventable premature mortality, changes in LE and economic im- effect through heat reductions. pacts related to the Superblock model were estimated. We used a comparative risk assessment framework (Murray et al., 2004), and 2. Materials and methods followed standard HIA methodologies (Mueller et al., 2017a, 2017b), comparing the baseline situation with the counterfactual scenario (i.e. 2.1. The Barcelona Superblock model Superblocks). We obtained exposure-response functions (ERFs) from the literature to quantify the strength of association between the ex- In the Eixample neighborhood, which covers most of the city center posures and mortality (Table 2). We obtained the annual natural-cause and is laid out in an orthogonal grid pattern, a Superblock will cover mortality rate for Barcelona (1,144 deaths/100,000 persons; 2015) approximately 400 m × 400 m. In other parts of the city, the (Table 1) (Barcelona Public Health Agency, 2017). We combined the Superblock design can deviate (Fig. 1). Within the Superblocks, pacified exposure data with the ERFs, population data and mortality statistic to interior roads will provide a local road network that is accessible pri- calculate population attributable fractions (PAFs) to quantify the Su- marily to active transport (i.e. walking and cycling) and secondarily to perblock scenario attributable preventable mortality burden: residential traffic with a maximum speed of 20 km/h (Fig. 2) (Rueda, n 2018). The Superblocks will be framed by the basic road network that p RRi i= 1 i 1 PAF = connects the city and accommodates through traffic at a maximum n p RRi i= 1 i speed of 50 km/h (Rueda, 2018). Besides accommodating cars/motor- cycles, the basic road network will contain segregated cycling and pe- where pi is the proportion of population at exposure level i, RRi is the destrian infrastructures and segregated bus lanes for rapid transit. For relative risk that quantifies the strength of association between the level optimal access, bus stops will be placed every 400 m at the main in- of exposure i and mortality, i is the level of exposure (i.e. the difference tersections of the Superblocks (in the non-grid-like neighborhoods this in exposure between baseline situation and Superblock scenario), and n distance can vary) and buses will circulate at a high frequency, making the number of exposure categories (i.e. 503 Superblocks) (Zapata- public transport an attractive alternative. With the implementation of Diomedi et al., 2018). the 503 Superblocks, private motorized traffic is expected to decrease We assumed immediate build-up of health benefits and the popu- considerably, and traffic flow on the basic road network is expected be lation size, age structure and mortality rate to stay constant. Children Fig. 2. Road hierarchy and traffic circulation aimed at with the Superblock model. 3 N. Mueller, et al. Environment International 134 (2020) 105132 Fig. 3. Green space development for the Eixample neighborhood. and adolescents were excluded from the analyses as their exposure- 2.3. Exposure data health relationships; their transport behaviors and the natural-cause mortality rate are different from the adult population. Analyses were 2.3.1. Transport-related physical activity conducted in R (v 3.5.2), Microsoft Excel (v 12.0) and QGIS (v 3.4.2). With the implementation of the 503 Superblocks, the private mo- torized transport share is estimated to be reduced by 19.2% (Table 3). This scenario is based on transport projections made for the Barcelona 4 N. Mueller, et al. Environment International 134 (2020) 105132 Fig. 4. Sant Antoni Superblock (2019). Table 1 have the distance of an average baseline bicycle trip (i.e. 3.84 km), and Barcelona population characteristics. a new walking trip would have the distance of an average baseline Population ≥20 years Natural-cause mortality Expected natural-cause walking trip (i.e. 1.2 km) (Table 3). Accounting for longer trips, we (2017) rate (2015) mortality assumed a new public transport trip would replace an average baseline car/motorcycle trip (i.e. 8.45 km) and would include a total 10 min 1,301,827 1,144/100,000 persons 14,893 walk to/from public transport to final destinations (Rojas-Rueda et al., 2012). We assigned the new bicycle trip 6.8 METs (Ainsworth et al., 2011), and the new walking trip and 10 min walk to/from public Urban Mobility Plan 2013–2018, that included infrastructure, quality transport 3.5 METs (Ainsworth et al., 2011). The association between and safety measures to promote public and active transport and reduce PA and mortality was quantified using a curvilinear ERF, applying a private motor vehicle use (Barcelona City Council, 2014). Hence, of the 0.25 power transformation (Table 2) (Woodcock et al., 2011). We cal- current almost 1,190,000 car/motorcycle trips/weekday (Barcelona culated the relative risk (RR) and the population attributable fraction City Council, 2017b), almost 228,000 trips would be shifted to public (PAF) for both baseline PA and gained PA. The estimated number of and active transport. We estimated health impacts associated with ex- preventable deaths for baseline PA was subtracted from the estimated pected changes in transport-related PA levels resulting of shifting car/ number of preventable deaths for the gained PA. motorcycle trips to public transport, cycling or walking for 5 days/ week. Supported by transport data, we assumed that Barcelona re- sidents carry out on average 3.5 trips/weekday (Metropolitan Area of 2.3.2. Air pollution Barcelona, 2013) (Table 3). Metabolic equivalents of task (METs) were Baseline nitrogen dioxide (NO2) concentrations (2012) were calcu- used as a measure of energy expenditure during PA. We calculated the lated and averaged at the Superblock area level using the Street 5.2 air gain in marginal METs for persons substituting car/motorcycle trips quality model (KTT Umweltberatung und Software Dr. Kunz GmbH, (i.e. 65,092 persons/weekday) with public transport, cycling or 2005). Changes in NO2 concentrations were modeled and reflect the walking, considering baseline PA levels. Baseline PA levels were de- expected 19.2% car/motorcycle share reduction, including changes in rived from the Barcelona Health Survey 2016/2017. The Barcelona vehicle fleet and increased average speeds due to avoidance of con- Health Survey draws on a population-based random sample of 4,000 gestion (Barcelona City Council, 2014) (Fig. 5). The association be- residents studying their health status, life-styles and use of health ser- tween NO2 and mortality was quantified using a linear ERF (Table 2) vices (Bartoll et al., 2018). PA levels were assessed in the Health Survey (Atkinson et al., 2018). The RR and PAF corresponding to the exposure with the International Physical Activity Questionnaire (IPAQ) short level difference between baseline and the Superblock scenario were form (IPAQ, 2005). We included Health Survey participants ≥20 years calculated at the Superblock area level. The expected changes of other with complete PA data (i.e. 3,217 adults) to determine baseline PA traffic-related air pollutants (TRAPs), such as particulate matter (e.g. levels as a combination of walking, moderate and vigorous-intensity PA PM2.5), were not modeled for the Superblock scenario, and therefore, (Table A.1). Accounting for shorter trips and peoples' willingness to health impacts could not be estimated. cycle or walk for transport, we assumed that a new bicycle trip would Table 2 Risk estimates for mortality by exposure domain. Exposure Risk estimate Exposure Health effect Study design Reference Physical activity RR = 0.81 11 versus 0 MET-h/week All-cause mortality Meta-analysis Woodcock et al., 2011 (95% CI 0.76–0.85) 3 NO2 RR = 1.02 per 10 μg/m All-cause mortality Meta-analysis Atkinson et al., 2018 (PI 0.99–1.06) Noise HR = 1.04 per 10 dB Lden (Road) CVD mortality Cohort study Héritier et al., 2017 (95% CI 1.02–1.06) Green space RR = 0.99 per 10% increase in greenness All-cause mortality Meta-analysis Gascon et al. 2015 (95% CI 0.98–1.01) Heat RR = 1.19 99th versus 74th temperature percentile All-cause mortality Time-series study Guo et al., 2014 (95% CI 1.16–1.23) CVD = cardiovascular disease; HR = hazard rate; Lden = EU day-evening-night noise indicator with 5 dB and 10 dB weights for the evening and night time, re- spectively; MET = metabolic equivalent of task (1 MET = 1 kcal/kg/h); PI = prediction interval; RR = relative risk; 95% CI = 95% confidence interval. 5 N. Mueller, et al. Environment International 134 (2020) 105132 Table 3 Barcelona transport data. Mode Baseline Superblocks Difference Transport data Mode Trips/weekday Persons/ Mode Trips/weekday Persons/ Persons/weekday Mean Mean Mean speed share (%)a weekdayb share (%)c weekdayb distance duration (h) (km/h) (km) Car/motorcycle 26.10 1,189,219 339,777 21.10 961,399 274,685 − 65,092 8.45d 0.41f 20.60g Public transport 39.50 1,799,776 514,222 41.30 1,881,791 537,655 + 23,433 6.22d 0.51f 12.20g Bicycle 2.10 95,684 27,338 2.50 113,910 32,546 + 5,208 3.84e 0.28f 13.71h Walking 32.30 1,471,716 420,490 35.10 1,599,295 456,941 + 36,451 1.20e 0.25f 4.80h Total 100.00 4556395b 1,301,827 100.00 4556395b 1,301,827 a Barcelona City Council, 2017b. b Calculated based on 3.5 trips/person/day (2003−2012) (Metropolitan Area of Barcelona, 2013). c Barcelona City Council, 2014. d Calculated as duration (h) * speed (km/h). e Raser et al., 2018. f Encuesta de Movilidad en Día Laborable (EMEF) 2015 (Metropolitan Transport Authority, 2015). g Barcelona City Council, 2016. h Calculated as distance (km)/duration (h). 2.3.3. Road traffic noise (Table 2). Daily mean ambient air temperatures (2016) were simulated Baseline annual mean 24-h noise indicator Lden (in dB) for road with the UrbClim Model (v 1.1) (resolution 1:100) provided by the traffic (2016) was calculated in 5 dB intervals by using the Computer Flemish Institute for Technological Research (VITO) within the Eur- Aided Noise Abatement (CadnaA) software (v 3.7) (DataKustik GmbH, opean climate-fit.city project (Ridder et al., 2015), and were averaged 2007) and was averaged at the Superblock area level. Changes in Lden at the Superblock area level (Fig. 5). On 132 days during 2016, at least levels were modeled reflecting the 19.2% car/motorcycle share reduc- one Superblock exceeded the MMT of 21.5 °C (Table 4). The exposure tion (Fig. 5). Missing noise values were imputed by using the mean. level difference between the MMT and the simulated daily mean tem- Currently, only evidence for Lden and cardiovascular mortality exists perature was calculated at the Superblock area level. To account for (van Kempen et al., 2018). It is believed though that the greatest con- seasonal variability in mortality, we based the daily mortality rate on tribution of noise to mortality is through cardiovascular effects (van averaged daily mortality count data for Barcelona for the meteor- Kempen et al., 2018), which was recently reinforced by a mega cohort ological summer season (June through August) provided by the Spanish study (Héritier et al., 2017). The association between Lden and cardio- Institute for National Statistics (i.e. 3 deaths/100,000 persons/day). vascular mortality was quantified using a linear ERF (Table 2) (Héritier The corresponding RR and PAF were calculated. et al., 2017). The hazard ratio and PAF corresponding to the exposure Thermal simulations of heat transfer from finite elements (i.e. sur- level difference between baseline and the Superblock scenario were face) to ambient air temperatures for one centric Superblock in the calculated at the Superblock area level. densely-constructed city center (Fig. 1 and Fig. B.1), were conducted using RadTherm Heat Transfer Software (v 9.0.1) (Thermo Analytics 2.3.4. Green space Inc., 2016). Considering statistical summer day weather conditions, the Baseline and Superblock green space was estimated and developed use of semi-permeable materials and additional green space, heat by BCNecologia exclusively for the Eixample neighborhood (Fig. 5), transfer simulations simulated in a daily average 2 °C heat transfer re- using cartographic data (2010) (resolution 1:1,000) from the Barcelona duction (Fig. B.2 and Fig. B.3). To be more conservative and to allow for City Council (Barcelona City Council, 2019). Using PostgreSQL/PostGIS varying land uses, densities, traffic volumes, surface elements and the (v 2.3), the current percentage green space (%GS) and projected in- presence of green and blue spaces across the city, we assumed a 1 °C crease in %GS under the Superblock scenario was calculated for the 202 ambient air temperature reduction as a more realistic scenario for the Superblock areas compromising the Eixample neighborhood (popula- implementation of all 503 Superblocks. Supporting this assumption, tion ≥20 years N = 464,216) (Fig. 3). The exposure difference between compared to rural background, the city-wide UHI effect for summer the current %GS and the Superblock %GS was calculated. A linear ERF 2016 was estimated to be around 2 °C (Fig. 5). Given that rural back- was used to quantify the association between green space and mortality ground conditions are unlikely to be achieved in the city, reducing the (Gascon et al., 2016b). The RR and the corresponding PAF were cal- UHI effect by half (i.e. 1 °C) seemed a more plausible scenario for the culated for each of the 202 Eixample Superblocks. The remaining 301 Superblocks. Nevertheless, we also present results of the 2 °C Superb- Superblock areas framing the Eixample neighborhood (i.e. population lock heat reduction scenario as a sensitivity analysis in the appendix ≥20 years N = 837,611) and either representing the historic city center (Table B.1). (i.e. Ciutat Vella) or neighborhoods located at Barcelona's periphery in Under the Superblock scenario, we reduced 2016 daily mean air the north, west and south, were excluded from the analyses as no temperatures by 1 °C at the Superblock area level. The exposure dif- quantifiable green space vision exists yet for these neighborhoods. ference was calculated for each Superblock for each day still exceeding 21.5 °C (i.e. 116 days) (Table 4). The corresponding RR and PAF were calculated. The number of deaths attributable to ambient air tempera- 2.3.5. Heat tures theoretically reduced by 1 °C was subtracted from the number of Daily mean temperatures (2012–2017) were available through a deaths attributable to the simulated ambient air temperatures for 2016. weather station located in the south-west of the city (Fig. 1) (Klein Tank et al., 2002), and were averaged to obtain typical temperatures for one calendar year. Following an empirical model (Guo et al., 2014), the 2.4. Life table analysis and economic impact 74th daily mean temperature percentile, which defines the minimum mortality temperature (MMT) for Spain, was determined to be 21.5 °C We estimated average changes in LE anticipated with the for Barcelona (2012–2017). Between the 74th and 99th temperature Superblock model, using life tables for Barcelona (2016) (Table C.1) percentiles, a linear mortality ERF was assumed (Guo et al., 2014) (IDESCAT, 2018), applying life table methods, as first described by 6 N. Mueller, et al. Environment International 134 (2020) 105132 NO2 (µg/ m3) A B Road noise (Lden dB) A B Fig. 5. Baseline and Superblocks environmental exposure levels. 7 N. Mueller, et al. Environment International 134 (2020) 105132 Green space (%) A B Temperature (ºC) A C A=Baseline exposure level B=Superblock exposure level C=UHI effect summer 2016 Fig. 5. (continued) 8 N. Mueller, et al. Environment International 134 (2020) 105132 Table 4 Baseline and Superblocks exposure levels and annual preventable premature deaths. Exposure (year) Baseline Superblocks Preventable deaths (95% CI) Transport physical activity (2017) Mean 29.39 MET-h/week Public transport +16.78 MET-h/week 36 (SD 35.93 MET-h/week) Bicycle +23.99 MET-h/week (26–50) Walking +5.99 MET-h/week NO2 (2012) 47.18 μg/m3 35.72 μg/m3 291 (Range 31.64–90.47 μg/m3) (Range 27.31–56.83 μg/m3) (0–838) Noise (2016) 54.18 dB Lden 51.26 dB Lden 163 (Range 39.15–74.60 dB Lden) (Range 37.52–71.01 dB Lden) (83–246) Green space (Eixample) (2010) 6.54%GS 19.57%GS 60 (Range 0.00–66.78%GS) (Range 0.20–97.95%GS) (0–119) Heat (2016) 132 days at least one Superblock > 21.50 °C 116 days at least one 117 Mean exceedance 3.55 °C Superblock > 21.50 °C (101–137) (SD 2.11 °C) Mean exceedance 3.15 °C (SD 1.77 °C) Total 667 (235–1,098) dB = decibel; Lden = EU day-evening-night noise indicator with 5 dB and 10 dB weights for the evening and night time, respectively. MET = metabolic equivalent of task (1 MET = 1 kcal/kg/h); SD = standard deviation; %GS = percentage green space; 95% CI = 95% confidence interval. Brunekreef (1997) for air pollution. We estimated average gains in LE were estimated to be reduced to 35.7 μg/m3 (−24.3%), and baseline for the Barcelona population ≥20 years by changing the probabilities city-wide annual mean road traffic noise levels of 54.2 dB Lden were of dying because of estimated city-wide mean improvements in en- estimated to be reduced to 51.3 dB Lden (−5.4%) (Table 4). In the vironmental exposure levels (i.e. NO2, noise, heat and green space) and Eixample neighborhood, baseline mean percentage green space of 6.5% increases in transport-related PA for persons expected to habitually was estimated to be increased to 19.6%. In 2016, on 132 days at least replace car/motorcycle trips with public and active transport. We also one Superblock exceeded the MMT of 21.5 °C daily mean temperature, estimated economic impacts of the changed premature mortality risk whereas with an assumed reduction of ambient air temperatures by based on the value of a statistical life (VSL) (2,510,000 EUR for Spain, 1 °C, on 116 days at least one Superblock would have exceeded 21.5 °C. 2015) (WHO, 2017). With the implementation of the 503 Barcelona Superblocks, we estimated that 667 premature deaths (95% CI: 235–1,098) could be 3. Results prevented annually (Fig. 6). The greatest proportion of preventable premature deaths could be attributed to reductions in NO2 levels (291, With the implementation of the 503 Superblocks, the car mode 95% PI: 0–838), followed by road traffic noise (163, 95% CI: 83–246), share was projected to be reduced by 19.2% (Barcelona City Council, heat (117, 95% CI: 101–137), and green space development in the 2014), that translated into almost 230,000 car/motorcycle trips/ Eixample neighborhood (60, 95% CI: 0–119). The estimated increased weekday being shifted to public transport, cycling or walking (Table 3). PA from the habitual shift from cars/motorcycles to public and active Accordingly, baseline city-wide annual mean NO2 levels of 47.2 μg/m3 transport for approximately 65,000 persons (Table 3) resulted in 36 Fig. 6. Annual preventable premature deaths estimated for the Barcelona Superblock model. 9 N. Mueller, et al. Environment International 134 (2020) 105132 preventable premature deaths annually (95% CI: 26–50) among this highlighted: quality spaces with less disturbing factors (e.g. cars and population. motorcycles) help create place attachment and provide security that In terms of changes in LE, we estimated that Barcelona residents positively affects the formation of identity, sense of community and ≥20 years would gain on average 198 days (95% CI: 99–297) because emotional and social well-being (Rollero, 2013). Public space helps of reductions in NO2, noise and heat. Additionally, Eixample neigh- facilitate social interaction and therefore contribute to social cohesion borhood residents ≥20 years were estimated to gain on average (Holland et al., 2007). In addition, Superblocks can help facilitate safe 37 days (95% CI: 0–76) because of green space development. Adults and independent child play. Research on Play Streets (i.e. street closure habitually replacing car/motorcycle trips with public transport, cycling from traffic to foster child play) showed that children's PA levels sig- or walking were estimated to gain on average 97 days (95% CI: nificantly increased, enhanced their sociability and reduced parental 79–159), 125 days (95% CI: 102–192), and 44 days (95% CI: 36–95), safety concerns (Cortinez-O'Ryan et al., 2017; D'Haese et al., 2015). respectively, because of increases in PA (Table C.1). Also, leisure time PA levels, although not considered in the present The reduction in 667 premature deaths resulted in economic im- analysis, are expected to increase with compact neighborhoods, mixed pacts of 1.7 billion EUR annually (95% CI: 0.6–2.8). land-use and quality spaces, and result in even greater health benefits. 4. Discussion 4.1. Limitations and strengths The Superblocks are a new innovative model in urban and transport As with all HIA studies, causal inferences may be an issue and re- planning that reframe the current mobility paradigm and places people sults have to be understood as estimations based on best available and well-being at the center. We estimated that with the implementa- epidemiological evidence. All changes in environmental exposures were tion of the 503 planned Superblocks, 667 premature deaths could be modeled and depend on assumptions made in the models and data prevented annually in Barcelona (Fig. 6), translating into a substantial availability. Further work is needed to validate these models with ac- economic impact of 1.7 billion EUR and considerable average gains in tual post-intervention measurements. Despite the longstanding epide- LE of almost 200 days due to reductions in harmful environmental ex- miological study of air pollution, the evidence base for a causal re- posures. Further gains in LE are expected with green space development lationship between long-term NO2 exposure and mortality remains and uptake of public and active transport. uncertain partly due to concerns relating to potential confounding ef- The present study is the first study to comprehensively estimate the fects by co-pollutants, especially fine particulate matter (i.e. PM2.5), and health impacts of this proposed real-life innovative intervention, holi- the heterogeneity of individual study results (Atkinson et al., 2018). stically considering the expected changes in the multiple urban and The evidence base for PM2.5 for inference of a causal effect on mortality transport planning related exposures. Previous HIA studies for the city is stronger (U.S. Environmental Protection Agency, 2009), but exposure of Barcelona estimated that 20% of premature mortality and 13% of the level changes of PM2.5 under the Superblock scenario were not mod- total burden of disease could be prevented under the assumption that eled, and PM2.5 can have many other sources than road traffic (i.e. international exposure guidelines for PA, exposure to air pollution, industry, the port and ships, construction, domestic, natural, etc.). noise, heat and access to green space were complied with (Mueller Particularly in the city of Barcelona, which has the highest-traffic et al., 2017a, 2017b). However, these scenarios were not associated density in Europe (Barcelona City Council, 2018) and a large share of with an actual urban and transport planning intervention. diesel-powered vehicles, NO2 is believed to be a good marker for road In contrast to other HIA studies of transport scenarios, that con- traffic related air pollution. A 2014 meta-analysis found a statistically sidered the impacts of PA, besides other risk factors, and commonly significant pooled effect of 1.04 (95% CI: 1.02–1.06) for NO2 on mor- estimated PA to be the most important health determinant (de Hartog tality (Faustini et al., 2014). If the true effect of NO2 on mortality was et al., 2010; Mueller et al., 2015; Rojas-Rueda et al., 2011; Woodcock indeed that reported by Faustini and colleagues, then Superblock-re- et al., 2009), our estimated transport-related PA benefits were smaller lated NO2 reductions would result in more certain and stronger health compared to those attributable to air pollution, noise and heat reduc- benefits of 570 preventable premature deaths (95% CI: 291–838). tions. This result can be explained by the different baseline populations While the largest contribution of noise to mortality is believed to be for which health impacts were estimated for. While motorized transport through cardiovascular effects (van Kempen et al., 2018), under- emission reductions (i.e. NO2 and noise) and mitigation of the UHI ef- estimation of the effect of noise on multiple causes of mortality cannot fect would benefit the general Barcelona adult population be ruled out: a recent and growing body of evidence has associated (N = 1,301,827), green space benefits were estimated for Eixample noise exposure with a wide range of physiological and psychological neighborhood residents only (N = 464,216) and benefits linked to in- health reactions (e.g. metabolic effects, respiratory symptoms, effects creases in transport-related PA relate exclusively to the population that on cognition, annoyance and sleep disturbance) (WHO, 2018). Despite is expected to habitually switch from cars/motorcycles to public and the Swiss population, from which the noise risk estimate was obtained active transport (N = 65,029). (Héritier et al., 2017), probably not being representative of the Barce- By far the strongest mortality impacts for the Superblocks resulted lona population, confounding by socio-economic position is unlikely, from air pollution reductions (291 deaths) - under the assumption that because the Swiss analyses were adjusted for socio-economic covari- the association between NO2 and mortality is causal. This finding is ates. linked to the assumption of reduction of traffic volume and decline in Although our heat analyses were supported by thermal simulations, congestion, that reduces NO2 concentrations substantially (Zhang and Superblock heat mitigation analyses were limited by the fact that the Batterman, 2014) (Table 4). Estimated road traffic noise and heat mi- conversion from surface to ambient air temperature is largely uncertain tigation effects were of considerable magnitude (163 and 117 deaths, and hardly generalizable because many different factors related to respectively), emphasizing the importance of these understudied en- urban structures and macro- and microclimate play a role. Additional vironmental exposures in urban environments. Further, green space uncertainties might have been introduced with modeling the associa- development in the Eixample neighborhood alone resulted in a con- tion between heat and mortality linearly and omission of the UHI effect siderable number of preventable deaths (60 deaths), underlining the possibly being protective during the winter months preventing cold- epidemiological significance of vegetation in cities in times of in- related deaths, which we did not account for. creasing population densities, general space scarcity and climate Using cartographic data of green spaces and the corresponding ex- change. posure unit %GS does not differentiate between intensity and quality of In addition to the health benefits presented in this study, other green areas. The Normalized Difference Vegetation Index (NDVI), potential co-benefits that we were not able to quantify should be which detects live green vegetation canopy using multispectral remote 10 N. Mueller, et al. Environment International 134 (2020) 105132 sensing data, has been suggested to be a better indicator (Gascon et al., Superblock model calls for a simultaneous improvement of the sub- 2016a), and has been linked to stronger health benefits (Crouse et al., urban commuter network that provides people with true alternatives. 2017; James et al., 2016), but is more difficult to project in the future as Improvement of the less-well developed public transport system for the it is vegetation type-specific. Moreover, until now, a quantifiable green wider Metropolitan Region is therefore a prerequisite for the Superb- space vision exists only for the Eixample neighborhood. However, the locks to function well. Furthermore, gentrification is a potential risk, other 301 Superblocks are also expected to receive green space inter- which can occur when depressed areas are occasionally improved (with ventions, which would translate into larger health benefits than esti- green and other popular infrastructure) and become attractive for upper mated here. classes, resulting in rent rises and subsequently forced migration (Cole For the transport-related PA analysis, although supported by re- et al., 2017). Also, the possibility of undesired relocation of car/mo- cords, we assumed that Barcelona residents conduct 3.5 trips/day on torcycle traffic (to potentially already deprived areas) outside the Su- average (Metropolitan Area of Barcelona, 2013). This assumption perblocks needs to be considered and avoided and therefore supple- strongly determines the traveling population of Barcelona and thus the mentary interventions that further discourage the use of private population shifting from private to public and active modes of transport motorized transport in the city are necessary (e.g. reduction of on-road and consequently the magnitude of health benefits. parking, congestion charging, low emission zones, etc.). In order to Furthermore, there might have been exposure and outcome mis- mitigate the identified risks, consistent and equitable implementation classification. Exposure, transport, health and demographic data were of the model across the entire city is recommended. only available for different years. Additionally, it is uncertain whether people spend most of their time in their Superblock of residence and 5. Conclusions therefore if exposures were assigned correctly. Besides the benefit of allowing the comparison of severities between The Barcelona Superblock model is a promising urban model as well the exposures, when considering multiple exposures, there may be the as public health strategy to reclaim public space for people and help the risk of ‘double-counting’ of premature deaths, especially if the ex- city become cleaner, greener, more physically active and climate posures are correlated. Independence of health effects has so far only change resilient through the reconfiguration of urban and transport been demonstrated for noise and air pollution (Tétreault et al., 2013). structures. We showed that the Superblocks have the potential to re- Health effects of air pollution could possibly be modified by tempera- duce the premature mortality burden and increase LE considerably ture (Li et al., 2017), and health benefits of green spaces may in fact through reductions in air pollution, noise and heat and increased access result from increases in PA or mitigation of air pollution, noise and heat to green space and transport-related PA performance. The magnitude of (Gascon et al., 2016b). In fact, exposure to multiple risk factors asso- estimated health impacts should make a case for the rapid im- ciated with the same health outcome (e.g. mortality) leads to the health plementation in Barcelona and the scale-up to other cities, where si- outcome being preventable in more than one way, from which could milar health benefits can be expected. For full and equitable distribu- potentially follow that preventable cases are being counted more than tion of health benefits, the Superblock model should consistently be once (Rowe et al., 2004). Therefore, the presented sum of preventable implemented across the entire city. deaths should be interpreted with caution. Finally, this HIA study assumed immediate and simultaneous im- Funding sources plementation of all 503 Superblocks; this is rather unlikely and also time-lags in the change of transport behaviors and exposure levels This research did not receive any specific grant from funding would delay health benefits. Unaccounted political, social, and cultural agencies in the public, commercial, or not-for-profit sectors. factors will influence the Superblock implementation process and therefore health impacts, something important to consider in times of Declaration of competing interest increasing urban populations, demographic as well as climate change and transformations in transportation and vehicle fleets. We declare no competing interests. Despite these limitations, there are many strengths of this study: This HIA should be understood as a robust overall estimation, based on Acknowledgements best epidemiological evidence according to the current research, of what health impacts of the real-life Superblock intervention may look We would like to thank Maria Foraster for helpful advice on the like. The multiple urban and transport planning related exposures were noise analyses and Glòria Carrasco for taking the pictures of the Sant considered holistically and where uncertainty on causal inferences ex- Antoni Superblock. JB gratefully acknowledges funding from the isted, assumptions were defined with caution and impacts were esti- European Union's Horizon 2020 research and innovation programme mated conservatively. Further, the presented premature mortality im- under grant agreements No 727852 (project Blue-Action), 730004 pact has to be understood as the ‘tip of the iceberg’ because mortality is (project PUCS) and 737480 (Marie Sklodowska-Curie fellowship an extreme event. Non-fatal impacts such as the expected reduction in ACCLIM). The climate-fit. city is developed as part of the PUCS project, chronic disease, improvements in quality of life, social cohesion and which has received funding from the European Union's H2020 Research mental health have not been quantified in the present analysis. and Innovation Programme under Grant Agreement No. 73004. Consequently, the total health and well-being impact of the Superblocks can be expected to be considerably larger than the numbers presented Appendices. Supplementary data here, which would make an even stronger case for the rapid im- plementation of this urban and transport planning – as well as public Supplementary data to this article can be found online at https:// health – intervention. doi.org/10.1016/j.envint.2019.105132. 4.2. Other implications References Certain concerns and possible negative consequences of the Ainsworth, B.E., Haskell, W.L., Herrmann, S.D., Meckes, N., Bassett, D.R., Tudor-Locke, Superblock model need to be acknowledged, especially in the context of C., Greer, J.L., Vezina, J., Whitt-Glover, M.C., Leon, A.S., 2011. 2011 compendium of physical activities: a second update of codes and MET values. Med. Sci. Sports Exerc. health equity and environmental justice. Since almost half of the two 43, 1575–1581. https://doi.org/10.1249/MSS.0b013e31821ece12. million daily suburban commuter trips from the Metropolitan Region Atkinson, R.W., Butland, B.K., Anderson, H.R., Maynard, R.L., 2018. Long-term con- are made in car/motorcycle (Barcelona City Council, 2016), the centrations of nitrogen dioxide and mortality. Epidemiology 29, 460–472. https:// 11 N. Mueller, et al. Environment International 134 (2020) 105132 doi.org/10.1097/EDE.0000000000000847. date: 10 October 2016. Barcelona City Council, 2014. Urban mobility plan of Barcelona PMU 2013-2018. [WWW James, P., Hart, J.E., Banay, R.F., Laden, F., 2016. Exposure to greenness and mortality in Document]. URL. http://www.bcnecologia.net/sites/default/files/proyectos/pmu_ a nationwide prospective cohort study of women. Environ. Health Perspect. 124, angles.pdf, Accessed date: 11 April 2019. 1344–1352. https://doi.org/10.1289/ehp.1510363. Barcelona City Council, 2016. Dades bàsiques de mobilitat 2015. [WWW Document]. Klein Tank, A., A.M.G., et al., 2002. Daily dataset of 20th-century surfaceair temperature URL. http://mobilitat.ajuntament.barcelona.cat/sites/default/files/DB_2015.pdf, and precipitation series for the European climate assessment. Int. J. Climatol. 22, Accessed date: 25 May 2018. 1441–1453 EUROPEAN CLIMATE ASSESSMENT & DATASET (ECA&D). Barcelona City Council, 2017a. Population Characteristics. Official Figures of Population. KTT Umweltberatung und Software Dr. Kunz GmbH, 2005. Street 5.2. 2017. [WWW Document]. URL. http://www.bcn.cat/estadistica/angles/dades/ Li, J., Woodward, A., Hou, X.-Y., Zhu, T., Zhang, J., Brown, H., Yang, J., Qin, R., Gao, J., tpob/pad/ine/a2017/edat/edatq11.htm, Accessed date: 17 July 2018. Gu, S., Li, J., Xu, L., Liu, X., Liu, Q., 2017. Modification of the effects of air pollutants Barcelona City Council, 2017b. Dades bàsiques de mobilitat 2016. [WWW Document]. on mortality by temperature: a systematic review and meta-analysis. Sci. Total URL. http://mobilitat.ajuntament.barcelona.cat/sites/default/files/DADES- Environ. 575, 1556–1570. https://doi.org/10.1016/j.scitotenv.2016.10.070. BASIQUES-MOBILITAT16_13_1.pdf, Accessed date: 25 May 2018. Manville, M., Shoup, D., 2005. Parking, people, and cities. J. Urban Plan. Dev 131, Barcelona City Council, 2018. Qualitat de l'Aire. Sabies que…? [WWW Document]. URL. 233–245. https://doi.org/10.1061/(ASCE)0733-9488(2005)131:4(233). https://ajuntament.barcelona.cat/qualitataire/ca/sabies-que, Accessed date: 17 Metropolitan Area of Barcelona, 2013. Mitjana desplaçaments per persona i dia dels re- April 2018. sidents a l'EMT i a l'AMB, 2003-2013 [WWW Document]. URL. http://www.amb. Barcelona City Council, 2019. CartoBCN. [WWW Document]. URL. http://w20.bcn.cat/ cat/web/area-metropolitana/dades-estadistiques/mobilitat-i-transport/mobilitat- cartobcn/, Accessed date: 15 December 2018. quotidiana, Accessed date: 25 May 2018. Barcelona Public Health Agency, 2017. Llibre mortalitat anual 2015 [WWW Document]. Metropolitan Transport Authority, 2015. Encuesta de Movilidad en Día Laborable Barcelona 2015, ambdós sexes. Mortalitat segons Grup d'edat i causes defunció. URL. (EMEF). [WWW Document]. URL. https://www.atm.cat/web/es/observatori/ http://www.aspb.cat/docs/sisalut/SISalutLlibresIndicadors/LlibreMortalitat_2015. encuestas-de-movilidad.php, Accessed date: 21 January 2019. html, Accessed date: 23 May 2018. Moreno-Garcia, M.C., 1994. Intensity and form of the urban heat island in Barcelona. Int. Barcelona Urban Ecology Agency, 2018. Charter for Designing New Urban Developments J. Climatol. 14, 705–710. and Regenerating Existing Ones. (Barcelona). Mueller, N., Rojas-Rueda, D., Cole-Hunter, T., de Nazelle, A., Dons, E., Gerike, R., Götschi, Bartoll, X., Pérez, C., Pasarín, M., Rodríguez, M., Borrell, C., 2018. Enquesta de salut de T., Int Panis, L., Kahlmeier, S., Nieuwenhuijsen, M., 2015. Health impact assessment Barcelona 2016/17. (Barcelona). of active transportation: a systematic review. Prev. Med. (Baltim). 76, 103–114. Brunekreef, B., 1997. Air pollution and life expectancy: is there a relation? Occup. https://doi.org/10.1016/j.ypmed.2015.04.010. Environ. Med. 54, 781–784. Mueller, N., Rojas-Rueda, D., Basagaña, X., Cirach, M., Cole-Hunter, T., Dadvand, P., Cole, H.V.S., Lamarca, M.G., Connolly, J.J.T., Anguelovski, I., 2017. Are green cities Donaire-Gonzalez, D., Foraster, M., Gascon, M., Martinez, D., Tonne, C., Triguero- healthy and equitable? Unpacking the relationship between health, green space and Mas, M., Valentín, A., Nieuwenhuijsen, M., 2017a. Urban and transport planning gentrification. J. Epidemiol. Community Health 71, 1118–1121. https://doi.org/10. related exposures and mortality: a health impact assessment for cities. Environ. 1136/jech-2017-209201. Health Perspect. 125, 89–96. https://doi.org/10.1289/EHP220. Cortinez-O’Ryan, A., Albagli, A., Sadarangani, K.P., Aguilar-Farias, N., 2017. Reclaiming Mueller, N., Rojas-Rueda, D., Basagaña, X., Cirach, M., Cole-Hunter, T., Dadvand, P., streets for outdoor play: a process and impact evaluation of “Juega en tu barrio”(play Donaire-Gonzalez, D., Foraster, M., Gascon, M., Martinez, D., Tonne, C., Triguero- in your neighborhood), an intervention to increase physical activity and opportu- Mas, M., Valentín, A., Nieuwenhuijsen, M., 2017b. Health impacts related to urban nities for play. PLoS One 12, e0180172. https://doi.org/10.6084/m9.figshare. and transport planning: a burden of disease assessment. Environ. Int. 107, 243–257. 3437033. https://doi.org/10.1016/j.envint.2017.07.020. Crawford, J., 2002. Reclaiming cities for citizens. [WWW Document]. URL. https:// Murray, C., Ezzati, M., Lopez, A., Rodgers, A., Vander Hoorn, S., 2004. Comparative www.opendemocracy.net/ecology-climate_change_debate/article_480.jsp, Accessed Quantification of Health Risks: Conceptual Framework and Methodological Issues, date: 16 August 2018. Comparative Quantification of Health Risks. Global and Regional Burden of Disease Crouse, D.L., Pinault, L., Balram, A., Hystad, P., Peters, P., Chen, H., van Donkelaar, A., Attributable to Selected Major Risk Factors. World Health Organziation, Geneva. Martin, R.V., Ménard, R., Robichaud, A., Villeneuve, P.J., 2017. Urban greenness and https://doi.org/10.1007/s12263-011-0248-4. mortality in Canada’s largest cities: a national cohort study. Lancet Planet. Heal. 1, Nieuwenhuijsen, M.J., Khreis, H., 2016. Car free cities: pathway to healthy urban living. e289–e297. https://doi.org/10.1016/S2542-5196(17)30118-3. Environ. Int. 94, 251–262. https://doi.org/10.1016/j.envint.2016.05.032. DataKustik GmbH, 2007. Computer Aided Noise Abatement (CadnaA) Software. Nieuwenhuijsen, M.J., Khreis, H., Triguero-Mas, M., Gascon, M., Dadvand, P., 2017. Fifty Dávalos, D., Maldonado, D., Polit, D.J., 2016. The hidden potential behind the city shades of green: pathway to healthy urban living. Epidemiology 28, 63–71. https:// planned for cars. Procedia Eng 145, 924–931. https://doi.org/10.1016/j.proeng. doi.org/10.1097/EDE.0000000000000549. 2016.04.120. Raser, E., Gaupp-Berghausen, M., Dons, E., Anaya-Boig, E., Avila-Palencia, I., Brand, C., De Hartog, J., Boogaard, H., Nijland, H., Hoek, G., 2010. Do the health benefits of cycling Castro, A., Clark, A., Eriksson, U., Götschi, T., Int Panis, L., Kahlmeier, S., Laeremans, outweigh the risks? Environ. Health Perspect. 118, 1109–1116. https://doi.org/10. M., Mueller, N., Nieuwenhuijsen, M., Orjuela, J.P., Rojas-Rueda, D., Standaert, A., 1289/ehp.0901747. Stigell, E., Gerike, R., 2018. European cyclists’ travel behavior: differences and si- D’Haese, S., Van Dyck, D., De Bourdeaudhuij, I., Deforche, B., Cardon, G., 2015. milarities between seven European (PASTA) cities. J. Transp. Heal. https://doi.org/ Organizing “play streets” during school vacations can increase physical activity and 10.1016/j.jth.2018.02.006. decrease sedentary time in children. Int. J. Behav. Nutr. Phys. Act. 12, 1–9. https:// Ridder, K.D., Lauwaet, D., Maiheu, B., 2015. Urban climate UrbClim – a fast urban doi.org/10.1186/s12966-015-0171-y. boundary layer climate model. Urban Clim. 12, 21–48. https://doi.org/10.1016/j. Faustini, A., Rapp, R., Forastiere, F., 2014. Nitrogen dioxide and mortality: review and uclim.2015.01.001. meta-analysis of long-term studies. Eur. Respir. J. 44, 744–753. https://doi.org/10. Rojas-Rueda, D., de Nazelle, A., Tainio, M., Nieuwenhuijsen, M.J., 2011. The health risks 1183/09031936.00114713. and benefits of cycling in urban environments compared with car use: health impact Gascon, M., Cirach, M., Martínez, D., Dadvand, P., Valentín, A., Plasència, A., assessment study. BMJ 343, 1–8. Nieuwenhuijsen, M.J., 2016a. Normalized difference vegetation index (NDVI) as a Rojas-Rueda, D., de Nazelle, A., Teixidó, O., Nieuwenhuijsen, M.J., 2012. Replacing car marker of surrounding greenness in epidemiological studies: the case of Barcelona trips by increasing bike and public transport in the greater Barcelona metropolitan city. Urban For. Urban Green. 19, 88–94. https://doi.org/10.1016/j.ufug.2016.07. area: a health impact assessment study. Environ. Int. 49, 100–109. https://doi.org/ 001. 10.1016/j.envint.2012.08.009. Gascon, M., Triguero-Mas, M., Martínez, D., Dadvand, P., Rojas-Rueda, D., Plasència, A., Rollero, C., 2013. The town in my mind: how place attachment and identification are Nieuwenhuijsen, M., 2016b. Residential green spaces and mortality: a systematic linked to place perception. Estud. Psicol. 34, 309–314. https://doi.org/10.1174/ review. Environ. Int. 86, 60–67. 021093913808349343. Guo, Y., Gasparrini, A., Armstrong, B., Li, S., Tawatsupa, B., Tobias, A., Lavigne, E., de Rowe, A.K., Powell, K.E., Flanders, W.D., 2004. Why population attributable fractions can Sousa Zanotti Stagliorio Coelho, M., Leone, M., Pan, X., Tong, S., Tian, L., Kim, H., sum to more than one. Am. J. Prev. Med. 26, 243–249. https://doi.org/10.1016/j. Hashizume, M., Honda, Y., Guo, Y.-L.L., Wu, C.-F., Punnasiri, K., Yi, S.-M., amepre.2003.12.007. Michelozzi, P., Saldiva, P.H.N., Williams, G., 2014. Global variation in the effects of Rueda, S., 2018. Superblocks for the design of new cities and renovation of existing ones. ambient temperature on mortality: a systematic evaluation. Epidemiology 25, 781–9. Barcelona’s case. In: Nieuwenhuijsen, M., Khreis, H. (Eds.), Integrating Human doi:https://doi.org/10.1097/EDE.0000000000000165. Health into Urban and Transport Planning. Springer International Publishing, pp. Héritier, H., Vienneau, D., Foraster, M., Eze, I.C., Schaffner, E., Thiesse, L., Rudzik, F., 135–154. Habermacher, M., Köpfli, M., Pieren, R., Brink, M., Cajochen, C., Wunderli, J.M., Tétreault, L., Perron, S., Smargiassi, A., 2013. Cardiovascular health, traffic-related air Probst-Hensch, N., Röösli, M., 2017. Transportation noise exposure and cardiovas- pollution and noise: are associations mutually confounded? A systematic review. Int. cular mortality: a nationwide cohort study from Switzerland. Eur. J. Epidemiol. 32, J. Public Health 58, 649–666. https://doi.org/10.1007/s00038-013-0489-7. 307–315. https://doi.org/10.1007/s10654-017-0234-2. Thermo Analytics Inc, 2016. RadTherm Heat Transfer Software. Holland, C., Clark, A., Katz, J., Peace, S., 2007. Social Interactions in Urban Public Places. U.S. Environmental Protection Agency, 2009. Integrated Science Assessment for The Policy Press, Bristol. https://doi.org/10.1186/1745-6215-12-264. Particulate Matter, EPA/600/R-08/139F. Final Report. (Washington, DC). IDESCAT, 2018. Taula de vida abreujada. Tots dos sexes Barcelona. 2012-2016. [WWW United Nations, 2015. Transforming our World: The 2030 Agenda for Sustainable Document]. Life Expect. Heal. Popul. Indic. URL. http://www.idescat.cat/pub/?id= Development. New York. https://doi.org/10.1007/s13398-014-0173-7.2. iev&n=8641&geo=prov%3A08&lang=en&t=201400, Accessed date: 20 July Van Kempen, E., Casas, M., Pershagen, G., Foraster, M., 2018. WHO environmental noise 2018. guidelines for the European region: a systematic review on environmental noise and IPAQ, 2005. Guidelines for Data Processing and Analysis of the International Physical cardiovascular and metabolic effects: a summary. Int. J. Environ. Res. Public Health Activity Questionnaire (IPAQ) – Short and Long Forms. [WWW Document]. URL. 15, 379. https://doi.org/10.3390/ijerph15020379. http://www.institutferran.org/documentos/scoring_short_ipaq_april04.pdf, Accessed WHO, 2017. Health Economic Assessment Tool (HEAT) for Walking and for Cycling. 12 N. Mueller, et al. Environment International 134 (2020) 105132 Methods and User Guide on Physical Activity, Air Pollution, Injuries and Carbon health benefits of strategies to reduce greenhouse-gas emissions: urban land trans- Impact Assessments. [WWW Document]. URL. http://www.euro.who.int/__data/ port. Lancet 374, 1930–1943. https://doi.org/10.1016/S0140-6736(09)61714-1. assets/pdf_file/0010/352963/Heat.pdf?ua=1, Accessed date: 11 April 2019. Woodcock, J., Franco, O.H., Orsini, N., Roberts, I., 2011. Non-vigorous physical activity WHO, 2018. Environmental Noise Guidelines for the European Region. https://doi.org/ and all-cause mortality: systematic review and meta-analysis of cohort studies. Int. J. 10.3390/ijerph14080873. Epidemiol. 40, 121–138. https://doi.org/10.1093/ije/dyq104. Wolf, K.L., Robbins, A.S.T., 2015. Metro nature, environmental health, and economic Zapata-Diomedi, B., Barendregt, J.J., Veerman, J.L., 2018. Population attributable frac- value. Environ. Health Perspect. 123, 390–398. tion: names, types and issues with incorrect interpretation of relative risks. Br. J. Woodcock, J., Edwards, P., Tonne, C., Armstrong, B.G., Ashiru, O., Banister, D., Beevers, Sports Med. 52, 212–213. https://doi.org/10.1136/bjsports-2015-095531. S., Chalabi, Z., Chowdhury, Z., Cohen, A., Franco, O.H., Haines, A., Hickman, R., Zhang, K., Batterman, S., 2014. Air pollution and health risks due to vehicle traffic. Sci. Lindsay, G., Mittal, I., Mohan, D., Tiwari, G., Woodward, A., Roberts, I., 2009. Public Total Environ. 307–316. https://doi.org/10.1016/j.scitotenv.2013.01.074.Air. 13