Dr. Lawrence Frank, Professor, University of British Columbia & President, Urban Design 4 Health, Inc. On 350 calories one apple tart or a special slice of Ray's Pizza a cyclist can travel 10 miles, a pedestrian 3.5 miles, and an automobile 100 feet. Transportation Alternatives, Bicycle Blueprint, 1998 Carbonless Footprints Paper in Preventive Medicine TRANSPORTATION ENERGY INDEX Dr. Lawrence Frank 1
CONCEPTUAL MODEL Land Use Patterns Built Environment Travel Choice Physical Activity & Dietary Patterns Body Mass Index Path I - Behavioral Chronic Disease Onset Health Care Utilization Patterns & Transportation Investments Travel Patterns Vehicle Emissions Respiratory Function Health Care Costs Path II - Exposure Note: Diet and nutrition, age, gender, income, genetics, and other factors also impact weight and chronic disease and to the extent possible are controlled in analyses. Vehicle age and climate impacts emissions and air quality, and respiratory function is also impacted by a variety of factors Dr. Lawrence Frank Quality of Life Environmental Quality Air Quality and Greenspace Human Behavior Travel Patterns and Physical Activity Built Environment Transportation Investments and Land Use Dr. Lawrence Frank 2
Why Should we care? Climate change impacts of travel patterns Obesity epidemic links with community design and related chronic diseases Lack of energy to sustain current economic model Aging population and requirement to retrofit urban environments Global migration to urban areas Dr. Lawrence Frank 3
Health and Built Environment Evidence to Date 1) built environments play a role in shaping health outcomes and disparities a) behaviors and b) exposures 2) Environments relate with health outcomes independent of preferences or self selection - relationship is at likely least partially causal 3) Meeting physical activity guidelines and reduce risk of obesity are associated with transit use and more bikable and walkable environments a) reduces odds of chronic disease onset for several morbidities b) logically reduces demands on health care system and associated costs 4) It is both possible and timely to monetize these costs Proximity. 2 KM 1 KM Connect- ivity Dr. Lawrence Frank 4
Built Environment Data Sources Assessors / Parcel data Residential density, land use mix, retail FAR Public Health Food locations GIS Data Layers Roads, trails, bicycle facilities, sidewalks BUILT ENVIRONMENT MEASURES (Independent variables) Ministry of Education Schools Transit Agencies Transit stops and mode Other Farmers markets, crime Census Demographic covariates Comparing Two Communities Dr. Lawrence Frank 5
Made up of: Residential density, retail Floor Area Ratio, intersection i density, land use mix Regional walkability distribution, by block group Toronto Walkability Index Dr. Lawrence Frank 6
Transit use was significantly associated with greater odds of meeting physical activity recommendations (OR=3.42; CI=2.40-4.87) 487) 4.87) by walking for transportation The odds of meeting 30 minute Physical Activity Recommendation is negative for additional trips as a car driver when compared to moderate walking (OR =.87; CI=0.76-0.99) 0.99) Source: LaChapelle and Frank JPAH, 2009 1) Evidence suggests that built environments likely play a role in shaping health outcomes and disparities a) behaviors and b) exposures 2) Recent evidence further clarifies that environments relate with health outcomes independent of preferences Charlotte NC evidence is among the first to suggest the relationship is potentially causal: Riding the rails can leave users an average of 6.5 pounds lighter than others, and 81 percent less likely to become obese over time, according to the study, published in the American Journal of Preventive Medicine 3) Increased odds of meeting physical activity guidelines and reduce risk of obesity are associated with transit use and more walkable environments > reduced risk of chronic disorders Dr. Lawrence Frank 7
2005 study found that residents of the most walkable areas in Atlanta t were 2.4 times more likely to get the recommended amounts of physical activity 18 percent in the lowest Versus 37 Percent in the highest levels of walkability got recommended levels of physical activity Frank LD, Schmid TL, Sallis JF, Chapman JE, Saelens BE. 2005. Linking Objective Physical Activity Data with Objective Measures of Urban Form. American Journal of Preventive Medicine. Volume 28, No. 2 (Suppl2): pp. 117-125. Built environment characteristics explaining walking in adults Adult Findings - Walking Any walk trip Non- Work/school walk trip work/school walk trip Higher residential density +++ +++ +++ Higher street connectivity +++ +++ +++ Higher commercial density +++ +++ +++ Higher mix of land uses ++ + ++ More nearby parks and open spaces +++ + +++ Higher overall neighbourhood walkability +++ ++ +++ NS = not significant, '+' = 95% significant; '++' = 99% significant, '+++' = 99.9% significant Devlin and Frank, 2009 Dr. Lawrence Frank 8
Predictors of Obesity Coefficient t-ratio P-Value Age 0.012 6.00 0.000 Education -0.080-4.71 0.000 Income -0.057-4.75 0.000 Walk Distance -0.049-2.04 0.034 Car Time 0.001 2.875 0.003 Land Use Mix -2.035-5.65 0.000 Black Male 0.311 3.930 0.000 Black Female 0.372 5.09 0.000 White Female -0.871-11.3 0.000 Constant -0.497-2.22 0.026 Frank, L., Andresen, M., and Schmid, T., Obesity Relationships With Community Design, Physical Activity, and Time Spent in Cars. American Journal of Preventive Medicine. June 2004. Every additional hour per day in a car translates into a 6 percent increase in the likelihood of obesity Time spent driving increases as walkability decreases Every additional Kilometer (.6 miles) walked translates into 4.8 percent reduction in the likelihood of being obese Distances walked increases with walkability Frank, L., Andresen, M., and Schmid, T., Obesity Relationships With Community Design, Physical Activity, and Time Spent in Cars. American Journal of Preventive Medicine. June 2004. Dr. Lawrence Frank 9
Dr. Lawrence Frank 10
LOGISTIC REGRESSION ANALYSES PREDICTING THE ODDS OF WALKING AT LEAST ONCE OVER 2-DAYS YOUTH Age Range 5-8 years OR (95% CI) 9-11 years OR (95% CI) 12-15 years OR (95% CI) 16-20 years OR (95% CI) N=847 N=632 N=867 N=815 Intersection highest tertile (vs lowest) Density highest tertile (vs lowest) 1.7 (1.0-2.9) 1.8 (1.0-3.1) Mixed land use (vs no mix) 1.5 (0.9-2.4) At least 1 commercial land use (vs 0) At least 1 recreation/open space land use (vs 0) *p<.05, **p<.01, ***p<.001 1.5 (0.9-2.4) 2.1 (1.3-3.4)*** 1.3 (0.8-2.3) 2.3 (1.2-4.3)** 1.5 (0.9-2.5) 1.6 (1.0-2.5) 1.8 (1.1-2.9)* controlling for socio-demographics and stratified by age group (Averaged over a two day period) 1.7 (1.1-2.8)* 3.7 (2.2-6.4)*** 2.5 (1.6-3.8)*** 2.6 (1.7-4.0)*** 2.5 (1.7-3.6)*** 2.0 (1.1-3.6)* 2.0 (1.0-4.1) 1.9 (1.0-3.2)* 1.7 (1.0-3.1) 1.8 (1.1-2.9)** Compared with youth from households with 3 or more cars (99.9% confidence level): Youth from households with 2 cars were 1.4 times more likely to walk at least once over a two day period Youth from households with 1 car were 2.6 times more likely to walk at least once over a two period and 2.2 times more likely to walk more than a ½ mile per day Youth from households with no cars were 7.7 times more likely to walk at least once a two day period and 6.8 times more likely to walk more than a ½ mile per day. Dr. Lawrence Frank 11
The Silver Tsunami By 2032, California will see its 65-plus population more than double in the next 25 years, from 3.5 million in 2000 (10.6 percent of the state's population) to 8.2 million in 2030 (17.8 percent). Where will the 65-plus population live? Access to services, Less reliance on driving for safety reasons, Maintaining Independence Lack of Affordable Housing in Walkable Places Cost of service delivery with Aging in central / walkable versus peripheral / unwalkable unbikeable places Increased walkability was related with more walking (OR 2.02), less time spent traveling in a car (OR.53), and lower odds of being overweight (OR.68). Those with 1 or no cars were more likely to walk (OR 2.9) and spend less time in cars (OR.53); but also less likely to get recommended levels of PA (OR.55). Visiting a fast food outlet was associated with increased odds of obesity (OR 1.81). Dr. Lawrence Frank 12
Air Pollution and Walkability Nitric Oxide Marshall, Brauer, and Frank 2008 The Hidden Health Costs of Transportation - Frank et al 2010 American Public Health Association Dr. Lawrence Frank 13
Kavage, Frank, and Kolian 2010 American Public Health Association Dr. Lawrence Frank 14
Residential Density and Mix of Housing Type Dr. Lawrence Frank 15
Prefers a Walkable Community Design Low Walkability Neighbo orhood Prefe erences 1 2 3 Built Environment 4 High Walkability Prefers Auto - Based Community Design Walk 2 1.5 Preference (Subjectiv ve) Auto Quadrant 1: Unmatched Walkability -- Low Preference -- Walk 1 0.5-1 Quadrant 2: Matched Walkability -- High Preference -- Walk 0-6 -4-2 0 2 4 6-0.5 Quadrant 3: Matched Walkability -- Low Preference -- Auto Low -1.5-2 Neighborhood Walkability (Objective) Quadrant 4: Unmatched Walkability -- High Preference -- Auto High Dr. Lawrence Frank 16
Walkability & Preference Groups Preference for Neighborhood Type Walkability of Current Neighborhood I High Low II III High Low High Low IV Low High Percent Average Daily Taking a Vehicle Miles Walk Trip Traveled (n) (n) 16.0% 36.6 (188) (188) 33.9% 25.8 (446) (446) 3.3% 43.0 (246) (246) 7.0% 25.7 (43) (43) If We Want More Evidence Based Practice We Need More Practice Based Evidence Dr. Larry Green Dr. Lawrence Frank 17
Data preparation The Scenario Planning Process Loading data Define the analysis area Build the future (change) scenario(s) Report existing/future conditions Review results and adjust scenarios as necessary Tool Development Dr. Lawrence Frank 18
Evidence based research on relationships (regression models) Urban Form Patterns Residential Density Land Use Mix Street Network Connectivity Retail Floor Area Ratio Outcomes Physical activity Obesity / Body Mass Index Transportation patterns Greenhouse gas -- CO2 Integrate findings into existing software Calculated Outcome Changes value Z Health Outcome value Y value X Base Scenario A Scenario B Neighbourhood Design Feature Dr. Lawrence Frank 19
Case Study West Don Lands (area shown in orange) Precinct Plan Boundary Source: Dr. Lawrence Frank 20
West Don Lands (Toronto) Example Pilot study site for software tool application: Substantial planning already done 80 acres significant changes in built environment dense/mixed use development 6000 6500 housing units 1 million sq ft of office/retail 2 new streetcar stops new park space Redevelopment is part of revitalizing Toronto's waterfront Site of athlete s village for Pan American Games (2015) Steps 1 2. Prepare and Add Data Data preparation Loading data Define the analysis area Build the future (change) scenario(s) Report existing/future conditions Review results and adjust scenarios as necessary Data type Postal code boundaries with built environment measures Used to Measure Residential density Land use mix Retail floor area ratio Office floor area ratio Walkability index Bicycle facilities i Metres of bicycle facilities i Trail network Metres of trails Metres of bicycle facilities and trails Food retail locations Supermarket density Convenience store density Farmer market density Restaurant density Take out restaurant density Parks Park area Road network School locations Sidewalk network Transit stops Metres of all walkable roads Number mberof intersections Intersection density Crow fly distance to nearest major arterial Network distance to nearest school School density Sidewalk coverage Network distance to nearest transit stop Transit stop density Dr. Lawrence Frank 21
Place Types Contain assumptions and information about land use in study area Starter place types can be customized by the user 5 Place Types created for West Don Lands (coloured highlights). Used high level control totals (housing units, retail/office sq ft, park space) to fine tune place type attributes Place Type Single Family Low SF Med SF High Townhomes Downtown Lofts Transit, rail station (not Yonge St) Transit, rail station (Yonge St) City Center/ Central Business District Office Park Shopping Center Industrial Park Park Front Street Neighbourhood Mill Street Neighbourhood River Square Neighbourhood Don River Park Neighbourhood Don River Park Neighbourhood with large park Description Low Density single family lots Medium density single family lots High density single family lots High density Townhomes & Multifamily Buildings Multifamily High density mixed use High density mixed use High density mixed use Single Use Office Park Single Use Commercial Single Use Industrial Single Use Conservation / Recreation Mix of apartment buildings (small, mid, highrise), retail. Predominantlyresidential residential (towers) 8 and 10 storey apartment buildings, some employment/production uses Townhouses and apartment buildings Townhouses and apartment buildings, with 18 acre park Comparing Scenarios West Don Lands Example West Don Lands Place Types Undeveloped postal code Mill Street neighbourhood Front Street neighbourhood Don River Park neighbourhood Don River Park neighbourhood (park) River Square neighbourhood Built environment features Transit stop School Road Bicycle facility Trail Dr. Lawrence Frank 22
Outcome Changes West Don Lands Outcome Study Area* Impacted Area** City** Base Future Base Future Base average active trips/person/day 0.2 0.4 0.4 0.4 0.1 average transit trips/person/day p y 0.6 0.7 0.7 0.8 0.5 average automobile trips/person/day 1.0 0.6 0.6 0.6 1.3 average trip kilometers/person/day 18.2 15.9 14.6 14.3 22.6 average CO2 generated (kg/hh/day) 3.4 2.5 2.3 2.3 4.2 walking for exercise monthly freq. 14.4 14.6 12.0 12.0 10.7 walk to work/school monthly freq. 7.8 9.8 10.8 11.2 5.6 bicycle for exercise monthly freq. 1.1 1.4 1.2 1.2 0.6 bicycle to work/school monthly freq. 0.8 1.1 1.0 1.1 0.3 daily energy expenditure (kcal/kg/day) 2.7 3.2 2.8 2.9 2.4 body mass index 24.3 24.2 24.0 24.0 24.6 high blood pressure (likelihood) 0.1 0.1 0.1 0.1 0.1 *Average of postal code values ** Population weighted average of postal code values Dr. Lawrence Frank 23
Case study 1 Palomar Gateway Neighborhood-scale, using a parcel-level tool Located just east of I-5 in southern Chula ua Vista 100 acres of vacant, retail, and industrial land near Palomar St, with residential to the north and south Identified in the City s 2005 General Plan as one of the top locations for infill and redevelopment Case study will test health impacts of potential Specific Plan alternatives Dr. Lawrence Frank 24
Case study 1 Palomar Gateway Case study 2 South Bay BRT Regional-scale, using an MGRA-level tool Analysis will identify the health impacts of introducing BRT and transit-oriented development along the South Bay BRT corridor Dr. Lawrence Frank 25
CO2 & Neighbourhood Design mean daily pe erson CO2 (KG) -- pe -- mean daily p person CO2 (KG) p 12.5 12 11.5 11 10.5 10 12.5 12 11.5 11 10.5 10 9.5 0-4 4-7 7-10 10-15 15+ Net Residential Density (housing units per residential acre) 0 1-2 3-9 10-29 30-165 CO2 (KG) -- mean daily pe person 13 12 11 10 9 8 0-0.1 0.1-0.2 0.2-0.3 0.3-0.4 0.4+ Intersections per acre # of Neighborhood Retail Parcels Source: LUTAQH final report, King County ORTP, 2005 Final Map of CO2 emissions from transportation Includes: Local urban form (land use mix, intersection density, retail FAR) Regional location (auto travel time Transit accessibility & travel time Demographics LFC, Inc. May. 19, 2009 Dr. Lawrence Frank 26
Driving 1/3 As Much in 2050 Brookings Draft Report King County John Snow 1857 Dr. Lawrence Frank 27
The Broad Street Pump. Change is Inevitable. In a progressive Country change is constant. Benjamin Disraeli, 1867 Dr. Lawrence Frank 28
Thank You! Dr. Lawrence Frank 29