key: cord-0715591-9r754nxp
authors: Wali, Behram; Frank, Lawrence D.
title: Hospitalizations and Mortality Relationships with Built Environment, Active and Sedentary Travel
date: 2021-08-21
journal: Health Place
DOI: 10.1016/j.healthplace.2021.102659
sha: 3a5e8540982bd04817d10bb26d2449174c32dfb6
doc_id: 715591
cord_uid: 9r754nxp

Most of the existing literature concerning the links between built environment and COVID-19 outcomes is based on aggregate spatial data averaged across entire cities or counties. We present neighborhood level results linking census tract-level built environment and active/sedentary travel measures with COVID-19 hospitalization and mortality rates in King County Washington. Substantial variations in COVID-19 outcomes and built environment features existed across neighborhoods. Using rigorous simulation-assisted discrete outcome random parameter models, the results shed new lights on the direct and indirect connections between built environment, travel behavior, positivity, hospitalization, and mortality rates. More mixed land use and greater pedestrian-oriented street connectivity is correlated with lower COVID-19 hospitalization/fatality rates. Greater participation in sedentary travel correlates with higher COVID-19 hospitalization and mortality whereas the reverse is true for greater participation in active travel. COVID-19 hospitalizations strongly mediate the relationships between built environment, active travel, and COVID-19 survival. Ignoring unobserved heterogeneity even when higher resolution smaller area spatial data are harnessed leads to inaccurate conclusions.

Built environment design significantly impacts public health. The importance of built environment design 25 was well realized in the latter 1800s and early 1900s when infectious diseases (such as Influenza) were the 26 primary public health threat faced by humanity (Frumkin et al., 2004) . Over decades, compact built 27 environment design also proved to be effective in combatting noncommunicable chronic diseases, including Carlson et al., 2015) . Causal evidence also suggests 38 that the gains in physical activity and active travel due to being exposed to compact and more walkable 39 environments over time are effective in protecting and prescriptively reducing mortality and morbidity from 40 the chronic disease epidemic (Schmid et al., 2015) . For example, more walkable neighborhoods are 41 associated with lower incidence of diabetes (Creatore et al., 2016) , obesity and body mass index (Author et 

Previous studies made important contributions by shedding new light on the role of density as the 79 pandemic evolved. However, several major gaps remain. First, population density has been largely used as 80 a proxy measure to capture the overall level of walkability, yet it is well understood that density alone does 81 not make a place walkable and less car dependent (Cervero and Kockelman, 1997b) . There are several other 82 aspects of community design that constitutes a walkable environment including heterogeneity of land uses, 83 and connected street networks along with a supportive pedestrian environment 1 (Saelens and Handy, 2008) 84 1 With regards to the built environment measures, the use of population density as a key surrogate measure of the built environment in understanding COVID-19 outcomes is problematic. While healthier and more walkable neighborhoods are denser, dense environments are not necessarily walkable (especially in developing countries). Consider the city of Karachi which is among the most populous cities in the world, with a population density of around 25,229 individuals / km 2over three times that of Vancouver (5,400 individuals / km 2 ). However, the two cities are on completely opposite ends of the livability spectrumwith Karachi rated as one of the least livable cities globally with poor levels of pedestrian-oriented infrastructure and connectivity, residential allotment, public services, and transportation. As discussed elsewhere (Author and Author, 2021), greater density brings people closer and is a completely opposed spatial concept to distancing. It is not logically possible that greater population density would be unrelated or negatively correlated with COVID-19 infection rate. In addition, given the strong positive correlations between density and measures of walkable environments (especially in developed countries), it is not possible to disentangle the effect of density from the effects of walkable built environment features. When density is used as a key built environment J o u r n a l P r e -p r o o f (Sallis et al., 2018; Khattak and Rodriguez, 2005) . Second, reliable estimates of COVID-19 incidence and 85 severity were arguably impossible to gauge early on in the pandemic when several studies focused on this 86 outcome were published. Testing regimes varied considerably and likely also systematically across urban 87 form further confounding the ability to assess how built environments correlate with COVID-19. Denser 88 urban areas may have been more rigorous in their testing procedures early on relative to sprawling lower 89 density suburban areas. Likewise, earlier data on COVID-19 mortality rates were extremely dynamic and 90 changed rapidly by location making it a difficult outcome to pin down. In addition to the focus of previous 91 studies on density as a key built environment measure discussed earlier, this may explain the inconsistent 92 and contradictory statistical links reported between density and infection/mortality rates early on the 93 pandemic (discussed earlier). In fact, statistically significant negative correlations between broader 94 measures of the built environment and COVID-19 mortality rates at a US county level were observed only 95 after November 2020 when the data became relatively more stable 

Keeping in view the above gaps, the present study contributes by conducting a neighborhood-level 117 analysis of the associations between built environment, COVID-19 hospitalization and mortality rate using 118 fine-grained census-tract level data from King County, Washington. Findings can help to inform how policy 119 makers respond to the threat of not only COVID-19 but future infectious disease threats. By including 120 neighborhood-level data on COVID-19 hospitalizations (in addition to mortality data), we examine the role 121 of hospitalization rates in mediating the relationships between built environment measures and COVID-19 122 mortality. In addition to the built environment, the present study sheds light on the effects of neighborhood-123 level active (sedentary) travel on COVID-19 hospitalizations and mortality. While the focus is on 124 hospitalization rates as a key mediator, we also test and include COVID-19 positivity rate as an additional 125 mediator linking built environment and active/sedentary travel with COVID-19 mortality outcomes. 126 predictor, the effects of other measures of walkable environments are manifested through density. In addition to the methodological issue of unobserved heterogeneity (discussed next), this mechanism could be a key factor behind previous findings concluding that population density is unrelated or negatively related with COVID-19 spread. 2 As one example, thanks to advancements in objective-built environment assessment methods, measures of walkability/built environment provide influential data about key elements of the physical environment known to support active travel. However, these data still do not approach the level of detail needed to fully capture the entire fabric of built environment. It is impractical to expect that information on all the relevant environmental factors can be collected and/or observed in the data typically available for analysis. ignore this important methodological concern which can lead to erroneous findings when significant 180 heterogeneity exists in the underlying data which is then masked through aggregation.

This results in the inability to accurately attribute a given coefficient to a large proportion of the 182 sample. Likewise, due to inherent differences (Salon, 2015) , it is unrealistic to posit that each individual 

To achieve the study objectives, multiple data streams are spatially joined at the census tract level. auto-oriented intersections typically are barriers to pedestrian and bicyclist mobility. Thus, it was given a 228 zero weight. Likewise, four-way intersections were given more weight since it promotes street connectivity 229 more effectively than three-way intersections. To capture regional auto-oriented accessibility, a regional 230 auto centrality index ranging between 0 and 100 is used (D5CEI). This measure reflects the proportional 231 accessibility to regional destinations by automobile and capture working age population accessibility (via (e.g., between destination accessibility and land-use/diversity). Finally, regarding diversity and land use 242 configuration, a five-tier employment and residential entropy measure is used. Developed originally by 243 Cervero (1989) (Cervero, 1989) , this measure was first applied to predict active travel by Author variables. As is evident, the coefficients on exposure must be fixed at 1 to be able to model rates per capita 292 (instead of counts). For details, see (Osgood, 2017) .

The fixed parameter Poisson models (Equations 1 through 6) unrealistically assumes that the 294 coefficients associated with each of the explanatory factors are fixed/constant across all the census-tracts.

Given the methodological concern of unobserved heterogeneity (explained in section 2.1), random 296 parameters can be incorporated as (subscript " " for variables dropped for brevity): (Table 1) 

Likewise, random parameter Poisson model led to a substantial improvement in predictive fit (see Figure   437 3). From an inferential standpoint, only two variables were statistically significant in the fixed parameter 438 model (compared to seven (7) statistically significant variables in the random parameter counterpart).

Finally, the random parameter Poisson model for positivity rate substantially outperformed the fixed 440 parameter counterpartwith a reduction of 6146.9 points in the AIC and a reduction in MAPE from 50.95 441 to 3.49 for random parameter model (Table 2) . Overall, these findings provide compelling evidence related 442 to the importance of accounting for unobserved heterogeneity. Importantly, the results suggest that the 443 associations between built environment and COVID-19 outcomes even vary significantly within a county 444 J o u r n a l P r e -p r o o f 13 (as opposed to inter-county heterogeneity reported elsewhere ). This finding is 445 intuitive given the personalized nature of behavioral interactions between users and the environment 4 . 446 Table 3 shows the estimation results for COVID-19 positivity, hospitalization and fatality rate 447 outcomes. The direct and indirect effects of exogenous variables are shown in Table 4 

This discussion of key findings is based on the results of the best-fit random parameter models. The key 474 findings with regards to the direct and indirect effects (see Figure 1 ) are presented. Higher levels of street 475 connectivity and more mixed-use correlate with less sedentary and more active travel. The opposite is true 476 for the regional auto centrality index. Results suggest that neighborhoods with more pedestrian-oriented 477 street connectivity, more mixed use (in terms of residential and employment mix), and lower auto 478 accessibility will on-average have lower COVID-19 hospitalizations and fatalities. We postulate this 479 relationship with COVID-19 is largely the result of increase in physical activity and lower levels of obesity 480 and chronic disease associated with these built environment characteristics.

After controlling for sociodemographic and other unobserved factors, pedestrian-oriented street 484 density (indicating higher connectivity) and 5-tier employment and household entropy (proxy for land-use 485 mix) are negatively correlated with COVID-19 hospitalizations. A ten-unit increase in pedestrian-oriented 486 street intersection density correlates with a 1.039% reduction in COVID-19 hospitalizations per capita.

Likewise, a one percent increase in employment and household mix is associated with a 0.241% reduction 488 in COVID-19 hospitalizations per capita. As expected, a percent increase in auto centrality index associates 489 with a 0.986% increase in hospitalizations (Table 3 and 4). However, the associations between employment 490 and household mix and hospitalizations exhibit substantial heterogeneity in varying degrees. For example, 491 with a mean structural parameter estimate of -0.002 and standard deviation of 0.008 (Table 3) Table 3 ). The application of more advanced random parameter methods found that 498 designing more walkable neighborhoods could be an effective strategy to combat COVID-19 severity while 499 sedentary auto-oriented development may be further discouraged. 

Hospitalizations and Fatalities 581 Table 4 (Table 4 ). For instance, a one 597 percent increase in walking for commute was indirectly associated with a 9.259% reduction in COVID-19 598 fatalities per capita through its associations with hospitalizations ( 

Despite the higher-resolution spatial data and rigorous methodological framework, the study is correlational 631 in nature and causal insights cannot be made. This study is based on neighborhood-level data from King 632 County, WA. Thus, caution must be exercised in generalizing the findings. However, the neighborhood-633 level built environment features harnessed in this study appear to be similar to what is found nationally. In 634 terms of housing and household size, the owner-occupied housing rate and household occupancy (persons 635 per household) is 56.9% in King County versus 64% nationally and 2.45 in King County versus 2.62 636 nationally. The proportion of high school graduates (93.1%) in King County is a bit greater than the national 637 average (88%). Regarding accessibility, the mean travel time to work is 29.6 minutes (compared to 26.9 638 minutes nationwide). Poverty in King County (7.7%) is lower than the national average (10.5%). These 639 statistics may allow readers to infer broader similarity patterns between King County and counties in other 640 states. The findings from this study highlight several promising avenues for future research. As discussed, 641 chronic disease was not considered due to the unavailability of reliable neighborhood-level data. Due to 642 data unavailability, the active/sedentary travel measures considered in this study only relate to work- 

Integrated 681 census-tract level data on COVID-19 mortality, hospitalization, and positive cases, built environment, 682 travel behavior, and sociodemographic factors were harnessed. Informed by rigorous simulation-assisted 683 discrete outcome random parameter models, the results shed new lights on the direct and indirect 684 connections between built environment, active/sedentary travel behavior, hospitalization, and mortality 685 rates

Auto-oriented built environment design (greater auto accessibility) is positively correlated with 687

COVID-19 fatality rate. Conversely, more mixed land use and greater pedestrian-oriented street 688 connectivity is associated with lower fatality and/or hospitalization rates

Sedentary (auto) travel is associated with greater COVID-19 fatality rate. Conversely, active travel 690 (biking and walking) is correlated with lower COVID-19 hospitalization/fatality rates

COVID-19 hospitalizations strongly mediate the relationships between built environment, active 692 travel, and COVID-19 fatality rate. Given the strong mediation role of hospitalizations, the indirect 693 effects of built environment and active travel on COVID-19 fatalities are relatively more profound

Black and elder populations are more vulnerable to develop serious illness (hospitalizations) and 695 die from COVID-19

Accounting for unobserved heterogeneity not only led to remarkable improvement 698 in predictive fit but provided more accurate insights as well. Importantly, ignoring unobserved 699 heterogeneity leads to incorrect conclusions -such as concluding that built environment is not

Healthier (active) 701 transportation infrastructure can be effective in combating the severity of COVID-19 and perhaps other 702 highly contagious infectious diseases. Population-level health and well-being of residents can be improved 703 by continuing to advocate for active-travel supportive transport infrastructure with more mixed land-use 704 and greater pedestrian-oriented street connectivity. Independent of mixed land use, the negative 705 associations between street connectivity and COVID-19 hospitalizations assert the independent role of 706 pedestrian-oriented street design in promoting walkability and combating COVID-19

Neighborhoods with more mixed land-use, greater street connectivity, and lower auto-oriented accessibility 709 promote active (walk, bike) and discourage sedentary travel. Related to population health, the importance 710 of walkable built environment in combatting COVID-19 is further highlighted by our findings related to 711 active and sedentary travel behaviors -suggesting that greater participation in active travel and lower 712 participation in automobile travel significantly reduces COVID-19 hospitalization and/or mortality rates

In assessing the relationships between built environment and COVID-19 outcomes, broader built 714 environment measures should be considered to paint a more accurate and informed picture of the underlying 715 mechanisms and pathways connecting transportation infrastructure and built environment with infectious

Hierarchical modeling and analysis for spatial 719 data

Multivariate random parameters collision count 721 data models with spatial heterogeneity

Walking, cycling, and obesity rates in Europe

Built environment and health

A copula-based approach to accommodate residential self-727 selection effects in travel behavior modeling

Accidents on Belgium's motorways: a network autocorrelation 730 analysis

Coronavirus disease 2019 mortality: a multivariate 732 ecological analysis in relation to ethnicity, population density, obesity, deprivation and 733 pollution

Measuring the built environment for physical 735 activity: state of the science

Modeling unobserved heterogeneity using 738 finite mixture random parameters for spatially correlated discrete count data

Multimodel inference: understanding AIC and BIC in 741 model selection

Obesity and COVID-19: an Italian snapshot

Obesity and COVID-19 severity in a designated hospital in 745

Association between neighborhood walkability and 747 GPS-measured walking, bicycling and vehicle time in adolescents

Urban density and COVID-19

Obesity, Race/Ethnicity, and COVID-19. Centers for Disease Control and 752 Prevention

Urban environments and objectively-assessed 758 physical activity and sedentary time in older Belgian and Chinese community dwellers: 759 potential pathways of influence and the moderating role of physical function

Land-use mixing and suburban mobility

Travel demand and the 3Ds: Density, diversity, and 764 design

Travel demand and the 3Ds: Density, diversity, and 766 design

Smart Location Database. Technical 768 Documentation and User Guide (Version 3.0)

Regression diagnostics

Factors associated with disease severity and 773 mortality among patients with COVID-19: A systematic review and meta-analysis

The role of the built environment in explaining 776 relationships between perceived and actual pedestrian and bicyclist safety

Association of neighborhood walkability 779 with change in overweight, obesity, and diabetes

Obesity and its implications for COVID-19 mortality

Travel and the built environment: A meta-analysis

Relationship between urban sprawl and 785 physical activity, obesity, and morbidity

COVID-19, the Built Environment, and Health

Urban sprawl and public health: Designing, 796 planning, and building for healthy communities

Patterns of COVID-19 related excess 798 mortality in the municipalities of Northern Italy during the first wave of the pandemic

An artificial immune network approach to multi-sensor 801 land use/land cover classification

Longitudinal analyses of the relationship between 803 development density and the COVID-19 morbidity and mortality rates: Early evidence 804 from 1,165 metropolitan counties in the United States

Does density aggravate the COVID-19 pandemic? 806 Early findings and lessons for planners

How the built environment affects physical 809 activity: views from urban planning

Exploring the neighbourhood-level correlates of Covid-19 deaths in London 811 using a difference across spatial boundaries method

Change in walking steps and association with built environments 813 during the COVID-19 state of emergency: a longitudinal comparison with the first half of 814 2019 in Yokohama

The impact of the built environment on health: an emerging field

Travel behavior in neo-traditional neighborhood 818 developments: A case study in USA

COVID-19 testing, case, and death rates and 821 spatial socio-demographics in New York City: An ecological analysis as of

Infection rates from Covid-19 in Great Britain by geographical 824 units: A model-based estimation from mortality data

Spatio-temporal Patterns of the COVID-19 Pandemic, and 826 Place-based Influential Factors at the Neighborhood scale in Tehran

Built environment and early infection of COVID-19 in urban 829 districts: A case study of

Association of built environment attributes with the spread of 831 COVID-19 at its initial stage in China

The link between obesity and the built environment

Evidence from an ecological analysis of obesity and vehicle miles of travel in California

Diverse and nonlinear influences of built environment factors on 836 COVID-19 spread across townships in China at its initial stage

Big data, traditional data and the tradeoffs 839 between prediction and causality in highway-safety analysis. Analytic methods in 840 accident research

Unobserved heterogeneity and the statistical 842 analysis of highway accident data

Antivirus-built environment: Lessons learned from 844 Covid-19 pandemic

Using 164 Million Google Street View Images to 846 Derive Built Environment Predictors of COVID-19 Cases

Poisson-based regression analysis of aggregate crime rates

Racial/ethnic minority and neighborhood 851 disadvantage leads to disproportionate mortality burden and years of potential life lost 852 due to COVID-19 in

Exploring the relationship between average speed, speed variation, and 854 accident rates using spatial statistical models and GIS

Smart location database

The built environment and health

Air pollution, physical activity and ischaemic 862 heart disease: a prospective cohort study of interaction effects

Can new urbanism encourage physical 864 activity?: Comparing a new Urbanist neighborhood with conventional suburbs

Built environment correlates of walking: a review

Built environment, physical activity, and obesity: findings 869 from the international physical activity and environment network (IPEN) adult study

Neighborhood built environment and 872 socioeconomic status in relation to physical activity, sedentary behavior, and weight 873 status of adolescents

Role of built environments in physical activity, 875 obesity, and cardiovascular disease

Heterogeneity in the relationship between the built environment and driving: 877 Focus on neighborhood type and travel purpose

Neighbourhood walkability and incidence of 880 hypertension: Findings from the study of 429,334 UK Biobank participants

A systematic review and meta-analysis of 883 physical activity and endometrial cancer risk

Coronavirus is making some people rethink where they want to 886 live

Discrete choice methods with simulation

The 17 Goals. Sustainable Development. Department of Economic and 890 Social Affairs

Physical Inactivity: A Global Public Health Problem

Introductory econometrics: A modern approach: Cengage learning

Built environment and obesity by urbanicity in the US

Examining spatial inequality in COVID-19 positivity 901 rates across New York City ZIP codes

Examine the associations between perceived neighborhood 903 conditions, physical activity, and mental health during the COVID-19 pandemic

Examining driver injury severity in two vehicle 906 crashes-A copula based approach