Co-Benefit Assessment of Active Transportation in Delhi, Estimating the Willingness to Use Nonmotorized Mode and Near-Roadway-Avoided PM2.5 Exposure

This study aims to estimate the avoided mortalities and morbidities and related economic impacts due to adopting the nonmotorized transportation (NMT) policy in Delhi, India. To this aim, an integrated quantitative assessment framework is developed to estimate the expected environmental, health, and economic co-benefits from replacing personal motorized transport with NMT in Delhi, taking into account the inhabitants’ willingness to use NMT (walking and cycling) mode. The willingness to accept NMT is estimated by conducting a cross-sectional survey in Delhi, which is further used to estimate the expected health benefits from both increased physical activity and near-roadway-avoided PM2.5 exposure in selected traffic areas in 11 major districts in Delhi. The value of a statistical life (VSL) and cost of illness methods are used to calculate the economic benefits of the avoided mortalities and morbidities from NMT in Delhi. The willingness assessment indicates that the average per capita time spent walking and cycling in Delhi is 11.054 and 2.255 min, respectively. The results from the application of the NMT in Delhi show the annual reduction in CO2 and PM2.5 to be 121.5 kilotons and 138.9 tons, respectively. The model estimates the expected co-benefits from increased physical activities and reduced PM2.5 exposure at 17,529 avoided cases of mortality with an associated savings of about USD 4870 million in Delhi.


Physical Inactivity and Health Impacts
Physical inactivity and air pollution are two major public health issues currently being faced in India. Nonmotorized transport (NMT) especially walking and cycling, can reduce air pollution and physical inactivity, save lives, and lessen the effects of climate change. Physical inactivity is responsible for about 5 million annual deaths, while motorized transportation-related emissions are responsible for 74 thousand premature deaths, and Delhi has the highest transportation-related death rates among India's major cities [1,2]. Sedentary lifestyles are one of the main factors increasing the risk of mortality from noncommunicable diseases (NCDs), as 71% of all deaths worldwide, 41 million per year, are caused by NCDs. Cardiovascular diseases (17.9 million), cancers (9.3 million), respiratory diseases (4.1 million), and diabetes (1.2 million) are major NCD-related mortalities each year [3]. Studies have shown that being physically inactive raises the risk of psychological disorders, colon cancer, breast cancer, type 2 diabetes, coronary heart disease, and musculoskeletal conditions by about 15% to 20% [4].

Nonmotorized Transport in Delhi
The urban transportation system endangers human health through accidents, air pollution, and physical inactivity. Such health issues have demanded significant health impact analyses in order to facilitate the creation of walking and bicycling-friendly infrastructure to improve public health. India's urbanization rate has led to substantial increases in the nation's energy consumption, travel demand, and transportation-related emissions. Due to rising ambient air pollution in highly populated megacities such as Delhi, most Indian cities struggle to keep their inhabitants' air livable. In the past three decades, Delhi's population has grown by a factor of four, but the city's motor vehicle population has increased by 28 [14]. There are currently more than 11.4 million registered automobiles in Delhi [15,16]. Transportation emissions have recently drawn much attention since many people are exposed to exhaust gases and other particles from moving vehicles. The contribution of transportation is believed to be as high as 45 percent in Delhi's air pollution [17]. Private cars and two-wheelers make up most of the Delhi transportation sector's emissions [18]. PM 2.5 concentrations in Delhi vary throughout the day, with the highest levels during the evening rush hours and night [19]. The average contribution of the transportation sector to PM 2.5 concentrations ranges from 17 to 28 percent, making it the main contributor to the city's poor air quality [20]. Studies have indicated that the mortality burden of PM 2.5 in Delhi can significantly decrease if exposure levels are reduced below the present levels [21]. Several sustainable national policies have been implemented to reduce transportation sector emissions and Delhi-specific pollution control measures [22]. However, Delhi's air quality has not improved and has become a severe public health concern. Sustainable urban transportation policies may decrease ambient air pollution in New Delhi by putting it in place [23].
The commuters in Delhi are forced to use individualized modes of transportation due to the lack of a satisfactory alternative public transportation system [14]. As a result, the vast majority of adults were physically sedentary. The total physical activity in Delhi is estimated to be about 400.3 min per week for men and 265.3 min per week for women [4]. Being a low-emission and space-efficient mode of transportation, active transportation (mostly walking and bicycling) has grown in popularity in the transportation and environmental sectors in India. NMT or active transportation can provide both transportation (access to goods and activities) and recreation, while users may view a specific journey to serve both aims.

Co-Benefits of Nonmotorized and Active Transport
Co-benefits have been widely advocated in tandem with mitigation measures for greenhouse gas (GHG) reduction. Especially in developing nations such as India, where urbanization and climate change adaptation are all significant challenges, the health and economic co-benefits of nonmotorized transportation become particularly important [24]. Given its long-term benefits, NMT appears to be a feasible alternative to traditional fossilfuel-based modes of public transportation in Indian cities. Walking and bicycling for transportation provide significant health benefits to users by increasing physical activity [25]. Benefits include improved accessibility, more affordable travel, less congestion, cheaper infrastructure costs for parking lots and roads, energy conservation, decreased air and noise pollution, and decreased accidents for other drivers [26]. Additionally, it is an affordable form of transportation for millions of people with low incomes, particularly in Delhi.
The quantification of the co-benefits from climate change mitigation in the transportation sector has been investigated in previous studies, introducing the A-S-I (avoid-shiftimprove) approach, assessing not just the avoided local air pollution and carbon emissions but also the advantages of improved air quality and its influence on public health. The results indicated a considerable decrease in CO 2 and other local air pollutant emissions by 12.9 and 1.4 million tonnes, respectively, the prevention of almost 10,000 mortality cases, and a savings of more than USD 35 million by 2030 [27]. Pathak et al. assessed the sustainable and low-carbon mobility in India and quantified co-benefits, using the scenario analysis method, under different scenarios, including fuel transition, decarbonized electricity, modal change, and travel demand substitution, indicating that low-carbon transportation measures can offer significant development co-benefits that include a 74% reduction in NOx and an 83% reduction in PM 2.5 from the passenger transportation sector compared with BAU in 2035 [28].
A review of thirty health impact assessment studies from Europe, the United States, Australia, and New Zealand concluded that health benefits from increased PA outweighed the adverse effects of traffic accidents and exposure to air pollution by a large margin. It is estimated that an increase in the median daily walking and bicycling time from 4 to 22 min can reduce 14% of GHG emissions in California, which results in decreasing the corresponding burden of cardiovascular disease and diabetes and could avoid 32,466 disability-adjusted life years (DALYs) [29]. Increased physical activity and reduced local air pollution from vehicle emissions can save 122 lives and net savings of about USD 200 million annually in New Zealand [30]. A study in Adelaide, Australia [31], showed that increased cycling, public, and active transportation could reduce road traffic-related CO 2 emissions annually, ranging from 191,313 to 954,503 million tons and 160 to 542 deaths. Therefore, 2113 to 7674 DALYs could be avoided due to improved air quality, increased physical activity, and avoided traffic injury. Another study conducted in six European cities has shown that increases in cycling and walking trips to 35% and 50% of total city trips, respectively, will cut GHG emissions in the six cities by 1139 to 26,423 metric tons per year, including varying degrees of health benefits [32]. A recent study has indicated cost savings of 15 billion euros per year in Europe for a 10% shift to active mobility modes [33].
As transportation-related emissions rise in Delhi, NMT is becoming increasingly popular among policymakers and environmentalists as a practical substitute for motorized transportation. Studies have addressed the significant impacts of improving bicycle and bus infrastructure on lowering CO 2 emissions [34] and increasing PA in 5% of the population [35]. Using the travel demand and health impact modeling approach, a study showed that active transportation can reduce the annual health burden (90,000 DALYs) in India [36]. Another study in India assessed alternative scenarios, including increased active travel, lower-carbon-emission motor vehicles, and a combination of the two, utilizing comparative risk assessment techniques [37]. The results indicated that the most significant advantages would come from combining active transportation with low-emission motor vehicles, which can reduce 12,995 DALYs in Delhi.

What Will Be Elucidated in This Research: Model Development
While there are inadequate studies in India conducted to estimate the co-benefits of NMT transportation, most of these studies have mainly focused on traffic injuries and air pollution and associated health benefits rather than the benefits of extra PA [38]. By developing an integrated quantitative assessment modeling framework of environmentalhealth-economic benefits (Figure 1), this paper seeks to quantify the expected health and economic co-benefits of both increased PA and improved air quality resulting from implementing the NMT policy in Delhi, India. As shown in Figure 1, the willingnessbased trip demand model is used to estimate the total kilometers and time the inhabitants are willing to spend on NMT due to choosing to walk and bike instead of driving cars and motorized two-wheelers. The impact of the potential factors such as age, gender, education, income, and travel distance on the probability of the willingness to use NMT is analyzed, utilizing a logistic regression model on the collected data from a cross-sectional interview with 250 inhabitants in Delhi. The results of the trip demand model are used in two sub-models: (1) The physical activity sub-model converts the extra daily minutes of walking and cycling to metabolic equivalents (METs), which is further used in a health risk assessment to estimate the relationship between physical activity and the beneficial impact of lowering the occurrence of specific health outcomes in the selected traffic areas.
(2) The near-road air pollution sub-model is used to calculate the potential reduction in motorized vehicle kilometers due to travel distance covered by walking and cycling and also its associated near-roadway-avoided PM 2.5 exposure in 11 major districts in Delhi. The relationship between changes in PM 2.5 concentrations and the occurrence of specific health outcomes in the chosen traffic areas is estimated, using the concentration-response function (CRF) for several diseases in order to establish a connection between the avoided exposure of PM 2.5 and health benefits. The relative risk (RR) level, which predicts the likelihood of an adverse health outcome among the population exposed to a higher level of ambient air pollution than a lower level of ambient air pollution, served as the basis for the CRF coefficient values used in this study. The values of the RR used in the study were taken from a thorough meta-analysis of earlier research. In order to achieve this, a meta-analysis and a review of articles were carried out to determine the correlation between alterations in PM 2.5 concentrations and alterations in the incidence of each health endpoint. The data for the avoided mortalities and morbidities from both sub-models are finally used to calculate the economic value of the avoided health burden from both the avoided emissions and increase in physical activity using the value of a statistical life (VSL) and cost of illness methods. The findings are generalized to the whole city of Delhi.

PA Estimation Model: Estimation of the Weekly Time Spent Walking and Cycling for Transportation
The counting model is first used to calculate the number of walks and bike trips an individual makes throughout an average day in order to estimate the weekly time spent walking and bicycling for transportation, which is expressed as follows [39]: where , is the daily minutes spent traveling using mode c (walking or bicycle) for individual i; , is the expected daily number of trips taken using mode c for individual i; , is the probability that a trip taken by individual i using mode c and , is the trip duration for a trip taken by individual i using mode c.
, expresses the willingness to adopt active transport (walking and bicycling), which can be further estimated by developing the following logistic regression model: In the above equation, is the willingness to adopt active transport of the ( = 1,2, ⋯ , ) individual.
= ( , , ⋯ , , ⋯ , ) is the × 1 vector of covariates corresponding to the individual. represents the vector of independent regressors, including age, gender, education, income, and travel distance, which affect the individual's willingness to use a walking or bicycle mode for transport purposes. Furthermore, = ( , , ⋯ , , ⋯ , ) is the corresponding × 1 vector of regression coefficients, which can be estimated through the maximum likelihood estimation approach. The likelihood function of interest takes the following form:

PA Estimation Model: Estimation of the Weekly Time Spent Walking and Cycling for Transportation
The counting model is first used to calculate the number of walks and bike trips an individual makes throughout an average day in order to estimate the weekly time spent walking and bicycling for transportation, which is expressed as follows [39]: where M c,i is the daily minutes spent traveling using mode c (walking or bicycle) for individual i; N c,i is the expected daily number of trips taken using mode c for individual i; Pr c,i is the probability that a trip taken by individual i using mode c and t c,i is the trip duration for a trip taken by individual i using mode c. Pr c,i expresses the willingness to adopt active transport (walking and bicycling), which can be further estimated by developing the following logistic regression model: In the above equation, Y i is the willingness to adopt active transport of the i th T is the p × 1 vector of covariates corresponding to the i th individual. X i represents the vector of independent regressors, including age, gender, education, income, and travel distance, which affect the individual's willingness to use a walking or bicycle mode for transport purposes. Furthermore, β = β 1 , β 2 , · · · , β j , · · · , β p T is the corresponding p × 1 vector of regression coefficients, which can be estimated through the maximum likelihood estimation approach. The likelihood function of interest takes the following form:

of 22
The estimated coefficients of the above model are interpreted through odds ratios (ORs). The OR for the covariate x ij can be estimated from the estimates of regression coefficients as given below.
The daily PA can be calculated by combining the time spent cycling and walking multiplied by the intensity of each activity, as determined by metabolic equivalents (METs), which is explained as follows [39]: where DPA i is the daily physical activity (from walking and cycling) for individual i in MET hours. For transportation, the MET values for walking and cycling are 3.5 and 6.8, respectively [40].

Near-Roadway-Avoided PM 2.5 Exposure Model
The concentrations of air pollutants at particular locations are believed to represent population exposure levels. It is estimated that 55% of the population living within 500 m of major roads in Delhi are exposed to hazardous emissions from transportation, where PM 2.5 is primarily produced by vehicle emissions [1]. It is crucial to use appropriate models with the vehicle and meteorological data to estimate near-roadway PM 2.5 exposure, since monitoring PM 2.5 concentrations cannot be achieved for all near-roadway regions. This research uses an air dispersion modeling tool called CALRoads View 6.5 (Lakes Environmental Software) to forecast how mobile sources affect air quality near roads and intersections. The CALINE-4 model is a fourth-generation line source air quality dispersion model that uses a mixing zone concept to characterize pollutant dispersion near roadways. Using the Gaussian dispersion methodology, the model predicts pollutant concentrations for receptors located within 150 m on either side of the roadways using input parameters such as site geometry and characteristics: where C is the concentration (g/m 3 ), y 1 and y 2 are the finite line source endpoint y coordinates (m), u is the wind speed (m/s), σ y and σ z are the horizontal and vertical Gaussian dispersion parameters which are a function of downwind distance x (m), q is the avoided PM 2.5 emissions from replacing the distance traveled of the vehicle category τ (g/s), which can be calculated, as follows: where MC walking , MC biking , K walking and K biking are the yearly time spent walking and cycling (s) and the average distance covered by walking and cycling in Delhi (km/s), respectively. EF τ is the PM 2.5 emission factor of the vehicle category τ (g/km) MC walking , MC biking are calculated by Equation (1) and K walking and K biking are given values.

Health Impact Assessment Model
Excess deaths or illnesses (EDI) caused by an increase in PM 2.5 concentration can be estimated as follows [41]: where I is the annual baseline mortality rate, P is the total population, and PAF (population attributable fraction) indicates the percentage of illness burden attributable to pollution. The PAF can then be calculated using the formula below [42]: where P represents the population's exposure rate. The values of RR for physical activity and near-roadway PM 2.5 exposure are estimated in different ways. In order to accurately assess the correct values of RR, the health impact assessment requires highly aggregated epidemiological data. To this aim, in this study, a comprehensive meta-analysis was carried out based on a systematic quantitative review, where all directly relevant empirical data that satisfy the eligibility standards and criteria were gathered, combined, and statistically analyzed to produce a pooled estimate that was as close as possible to the RR values specified by Delhi. The following equation is used to estimate the final RR of willingnessbased physical activity [39]: where RR PA is relative risk estimated from walking and bicycling physical activity; RR m is the relative risk estimated from the meta-analysis of the collected data from 61 studies on 5 physical activity-related health outcomes (all-cause mortalities, type 2 diabetes, coronary heart diseases, cancers, and depressive disorders) to estimate the dose-response function and PA-related morbidities and mortalities as the functions of physical transportation activity (Appendix A, Figures A2-A12). The meta-analysis was based on collecting data from selected studies with reported values of RR of moderate-intensity PA (α ranges from 3 to 6 h). The current global recommendation for adults' physical activity is 150 min (α = 2.5 h) of moderate-intensity aerobic physical activity per week [43]. The pooled values of the RR collected from 64 studies related to total mortality, COPD, cardiovascular mortality, respiratory mortality, morbidity, and related hospital admission were used to calculate the relationship between a change in air pollutant concentration (in this study, PM 2.5 ) and a change in health effects, i.e., concentration-response functions (CRFs): where RR AP is the relative risk estimated from the near-roadway-avoided PM 2.5 exposure, and β is a coefficient that evaluates the response of a health outcome to a change in pollutant concentration. C and C o are the pollutant concentrations in the baseline and intervention scenarios. Table 1 represents the baseline incident rates of diseases in Delhi used in this study.  1 66 Cancer prevalence rate 4 98.8 Lung cancer prevalence rate (LC) 3 9.06

Economic Impact Assessment Model
The value of a statistical life (VSL) approach and the cost of an emergency room visit (ERV) were used in this study to calculate the mortality cost of PM 2.M exposure [41]. The cost of illness (COI) approach was used to determine the cost of treatment of coronary heart diseases, type 2 diabetes, cancer and depressive disorders, and hospital admissions due to respiratory diseases, COPD, respiratory mortality, cardiovascular mortality, and LC [44][45][46][47][48][49][50]. The values for the various health endpoints considered in this study are shown in Table 2.

Willingness of People in Delhi to Use NMT
A cross-sectional survey was conducted from 2 July to 8 July 2022, in Delhi, India, to estimate peoples' willingness to accept active transportation as their usual travel mode. The survey encompassed different areas of Delhi, namely Ashok Vihar, Nand Nagri, Dilshad Garden, Jhilmil, and Seema Puri. About 250 inhabitants were physically interviewed during the survey, while 50 responses were collected online. People were physically approached and were asked questions during morning hours while leaving for work and during evening hours while returning, especially near bus stops, public parks, and pedestrian crossings. For the online survey, the questionnaires were shared with only those currently living in Delhi. The detailed questionnaire used in this research is described in Appendix B ,  Table A2.
The questionnaire, consisting of 11 questions, initiated the most important survey inquiry, whether the respondent was willing to use active transport in Delhi. Then, separate questions were asked for walking and bicycling. The later part of the questionnaire asked about the mandatory and nonmandatory travel modes of the inhabitants of Delhi, India, along with their preferable distance to be covered by walking and bicycling. Finally, the survey enquired about the respondents' socio-economic condition (income and education) and demographic characteristics (age and gender). After discarding all the missing and erroneous data, a total of 283 respondents' information was finally analyzed. All these variables are described in Appendix B, Table A1, which depicts that 30.35% and 38.52% of the respondents were willing to use walking and bicycling, respectively, as their regular transportation modes. All the selected covariates were categorized to distinguish the probability of willingness for the different categories of the corresponding predictors. Among the set of covariates, regular travel mode, monthly income, education level, and age were categorized into more than two categories. Such categorical variables require special attention, unlike continuous and binary predictors, which can be incorporated into the regression model without any modification. In the case of the covariates with more than two categories, one category needs to be considered as a reference category with which the remaining categories can be compared. This reference category is selected based on the purpose of comparison. Then, the remaining categories are transformed into binary variables and compared with the reference category of interest. The results of the model are presented in Table 3. Regular travel mode, monthly income, and education level were found to be significant factors in the willingness to walk in Delhi, India. On the other hand, the LR model for the willingness to bicycle revealed significant associations between this willingness and regular travel mode, distance to cover, education level, gender, and age group, which are depicted in Table 3. The OR for private car users was 0.241, which is less than one. Hence, those who used private cars as their regular travel mode in Delhi had significantly (1 − 0.241) × 100% = 75.9% lower odds of preferring to walk than public transport users. Similarly, motorcycle users had 57.4% lower odds of willingness to walk regularly as a travel mode, compared with those who used public transport. A lower likelihood of bicycling was observed for both the regular users of private cars and motorcycles than those of public transport. Distance to cover could significantly influence only the willingness to bicycle. For longer distances (more than 2 km), people preferred (OR = 3.228 > 1 and p < 0.10) to use bicycling as a regular travel mode more than for shorter distances. Compared with the medium-income group, people with high incomes showed more tendency to choose walking as their travel mode in daily life. Both illiterate and highly educated people preferred walking more than those who completed their primary or secondary education. On the other hand, only illiterate people were more likely to bicycle than primary-to-secondary-educated ones. Both middle-aged and older people had a lower preference for bicycling than the younger group. On the other hand, compared with their female counterparts, males had a higher willingness to use bicycles in their daily lives. The p-values of the Hosmer-Lemeshow test for both the LR models were greater than 0.05, which indicates that the corresponding models did not provide a poor fit to the data [59].
Based on the NMT willingness analysis findings, the average per capita time spent on walking and bicycling were estimated as 11.1 and 2.3 min, respectively, which is equal to covering an extra walking and cycling distance of 1.18 km per day based on the average walking and cycling speed in Delhi. These results were further analyzed to evaluate the environmental and health impacts achieved by improving the infrastructure so that people in Delhi can adopt active transport as per their willingness.

Near-Roadway-Avoided PM 2.5 Exposure from Replacing the Distance Traveled by Private Vehicles with Walking and Cycling in Delhi
The near-roadway concentration model described in Section 2.2 was applied to selected traffic zones (0.2 × 1 km 2 area) in all 11 districts of Delhi while considering the population density in each district, in order to evaluate the impact of the avoided PM 2.5 exposure from replacing the distance traveled by private vehicles with walking and cycling. The PM 2.5 concentration was estimated by using the CalRoads software and was found to be 0.2 km downwind from vehicles to the near-roadway passengers. The heatmaps of the avoided PM 2.5 exposure in the selected traffic zones in different districts in Delhi are shown in Figure 2. The per-hour-avoided vehicle kilometers (VKM) due to an increase in NMT and the average near-roadway-avoided PM 2.5 exposure in the different districts in Delhi are shown in Table 4.  Figure 3a shows the avoided mortalities and morbidities related to increased PA per km 2 in the different districts of Delhi. Northeast Delhi had the highest health benefits due to high population density, while the New Delhi district had the lowest health benefits due to less population density. More specifically, the NMT (walking and bicycling) use in   Figure 3a shows the avoided mortalities and morbidities related to increased PA per km 2 in the different districts of Delhi. Northeast Delhi had the highest health benefits due to high population density, while the New Delhi district had the lowest health benefits due to less population density. More specifically, the NMT (walking and bicycling) use in areas with high population density, such as the northeast (36,155 people per kilometer), led to a more significant reduction in the number of morbidity and mortality cases. Figure 3b shows the avoided mortalities related to avoiding PM 2.5 exposure near roadways in the various districts. The pooled values of the RR used in health impact assessment for both PA and near-roadway-avoided PM 2.5 exposure are reported in Tables 5 and 6. Table 7 includes the detailed avoided morbidities per km 2 in the different districts of Delhi. As previously stated, the avoided health burden in each district depends on both traffic and the population density along the nearest roadway. Therefore, despite the fact that the southeast, Shahdara, and north districts had the highest avoided PM 2.5 emission and exposure due to their heavy traffic conditions, the northeast and central districts had the highest expected avoided health cases due to higher near-roadway population densities. previously stated, the avoided health burden in each district depends on both traffic and the population density along the nearest roadway. Therefore, despite the fact that the southeast, Shahdara, and north districts had the highest avoided PM2.5 emission and exposure due to their heavy traffic conditions, the northeast and central districts had the highest expected avoided health cases due to higher near-roadway population densities.

Public Health and Economic Co-Benefits from Replacing Distance Traveled by Private Vehicles with Walking and Cycling in Delhi
(a) (b)   Table 6. RR values per 10 μg reduction in PM2.5 concentration extracted from the meta-analysis.   Table 6. RR values per 10 µg reduction in PM 2.5 concentration extracted from the meta-analysis. These findings were then generalized to the entire Delhi, taking into account the total area of the route network length of 16,200 km in Delhi [58][59][60][61][62][63]. The total avoided CO 2 emissions and PM 2.5 exposure were estimated at 121.5 kilotons per year and 138.9 tons per year, respectively (the emission factors and avoided VKM are given in Appendix C, Table A3). Finally, the total avoided health and economic impacts of physical activity and the associated reduction in near-roadway PM 2.5 exposure are given in Table 8. The results indicate that compared with the near-roadway-avoided PM 2.5 exposure, increased PA had a higher impact on preventing annual mortalities and morbidities in the different districts of Delhi. This can be attributed to the higher values of the relative risks of PA, leading to higher health impacts. In addition, health benefits were seen as more prominent in the districts where the population density is higher, both in terms of increased PA and decreased air pollution. These results emphasize that an increase of 11.2 min of physical activity or 1.18 km of extra distance walked per day can significantly reduce noncommunicable diseases related to mortalities and morbidities in Delhi. The findings from this study are comparable with those of previous studies, which have predicted significant health and economic benefits of NMT policy implementation in India and other countries, emphasizing that the health benefits of active transport can outweigh the costs of NMT infrastructure requirements [63]. Similar studies in India have shown that active transportation can lower India's health burden by 90,000 DALYs annually [35], and integrating active transportation with low-emission vehicles can save 12,995 DALYs in Delhi [36]. While considering economic impacts, a 10% shift to active mobility options has been indicated to save 15 billion euros annually in Europe [32].

Conclusions
Sustainable transportation can be integrated with development objectives such as health and well-being, clean energy, and sustainable cities. Climate change strategies are crucial, especially in developing nations such as India. In order to support the actions to mitigate climate change and other human development goals, it is essential to identify concrete co-benefits. Due to their lower carbon footprint and significant economic advantages in terms of preventing health effects, NMT can be a crucial part of such a strategy in Delhi. It can lower individual healthcare and transportation costs while enhancing public health and the environment. To this aim, this study introduced the integrated environmental, health, and economic benefits of implementing the NMT scenario in Delhi based on the willingness to walk and bike.
The results revealed that increased physical activity and avoiding exposure to PM 2.5 near roadways are expected to reduce the mortality rate by 40,018 cases in addition to reducing other morbidities, as indicated in this study, while physical activity was revealed to play a significant role in reducing mortalities and morbidities. The associated cost savings from morbidities and mortalities were estimated to be approximately USD 4869.8 million annually, which will positively impact Delhi's local government's finances. On the other hand, the local government may face difficulties and high costs when creating NMT facilities in Delhi due to the need to build and improve bike lanes, paths, and crosswalks as well as design safer roads for NMT transportation. However, massive savings in annual health costs may outweigh investments in infrastructure in just a few years.

Appendix A
This meta-analysis included 64 + 61 studies based on the inclusion criteria using the RAVMAN (5.4.1 version) software. The following tables show the pooled risk ratio (RR) for total mortality (PM2.5), COPD, cardiovascular mortality, respiratory mortality, LC, and respiratory morbidities related to hospital admissions, cancer, type 2 diabetes, depression, total mortalities (PA), and coronary heart diseases from 64 studies related to PM2.5 and health impacts, and 61 studies that were included in the meta-analysis.
(a) (b) Figure A1. The geographical distribution of the studies that were chosen for the RR meta-analysis: (a) physical activity and health impacts; (b) PM2.5 and health impacts.

Appendix B
The questionnaire used for the estimation of willingness for NMT in Delhi.

Appendix B
The questionnaire used for the estimation of willingness for NMT in Delhi.