Marquet et al. International Journal of International Journal of Health Geographics (2023) 22:3 https://doi.org/10.1186/s12942-023-00325-8 Health Geographics RESEARCH Open Access Comparison of static and dynamic exposures to air pollution, noise, and greenness among seniors living in compact‑city environments Oriol Marquet1,2*, Jose Tello‑Barsocchini2, Daniel Couto‑Trigo2, Irene Gómez‑Varo2 and Monika Maciejewska2,3 Abstract GPS technology and tracking study designs have gained popularity as a tool to go beyond the limitations of static exposure assessments based on the subject’s residence. These dynamic exposure assessment methods offer high potential upside in terms of accuracy but also disadvantages in terms of cost, sample sizes, and types of data gen‑ erated. Because of that, with our study we aim to understand in which cases researchers need to use GPS-based methods to guarantee the necessary accuracy in exposure assessment. With a sample of 113 seniors living in Barce‑ lona (Spain) we compare their estimated daily exposures to air pollution (PM2.5, PM10, NO2), noise (dB), and green‑ ness (NDVI) using static and dynamic exposure assessment techniques. Results indicate that significant differences between static and dynamic exposure assessments are only present in selected exposures, and would thus suggest that static assessments using the place of residence would provide accurate-enough values across a number of exposures in the case of seniors. Our models for Barcelona’s seniors suggest that dynamic exposure would only be required in the case of exposure to smaller particulate matter (PM2.5) and exposure to noise levels. The study signals to the need to consider both the mobility patterns and the built environment context when deciding between static or dynamic measures of exposure assessment. Highlights Static and dynamic exposures were estimated for a set of basic exposures (NO2, PM2.5, PM10, dB, NDVI) among 113 seniors in Barcelona. Static exposures were based on residential location while 7-day GPS tracking was used for dynamic exposures. Dynamic assessments using GPS-tracking are not providing enough accuracy across all the range of expo- sures. PM2.5, and noise exposures significantly benefited from GPS-tracking use. NO2, PM10, and exposures to greenness did not lose significant accuracy using only residential static exposures with Barcelona’s seniors *Correspondence: Oriol Marquet oriol.marquet@uab.cat Full list of author information is available at the end of the article © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​ mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 2 of 16 Keywords Exposure assessment, Air pollution, Greenness, Noise, Older adults, Dynamic exposure Introduction on individuals’ time-activity patterns [21] and they Research on how environmental exposures directly affect require intensive engagement from study participants human health has gained attention in recent years. The which usually leads to small sample groups, vulner- assessment of how exposure to different environmental able to participation rates, and study-abandoning [22]. features can harm or promote health has greatly ben- Participants in studies requiring location tracking and efitted from more accurate data both on the spatial dis- high spatio-temporal precision also report concerns tribution of exposures and on human positioning and with data privacy and surveillance. Because of that, it travel behavior. To date, available scientific evidence has is important to understand when it is necessary to use amply demonstrated the associations between daily aver- dynamic exposures and in which situations we can age exposures to factors such as air pollution, noise, and expect high-accuracy exposure assessments from static greenness and cardiovascular disease, mental health, and measures alone. Deciding between static or dynamic even wellbeing [1–3]. exposure measurements is a complex task that will likely depend on the subject of analysis and the type of exposure of interest. Static vs dynamic exposure assessment This need is even more important when dealing with Traditionally, most environmental exposure assessments specific population groups such as seniors or children have taken the residence location or the workplace as a for whom wearing a specific tracking device may be proxy for people’s environmental exposures [4, 5]. Using more difficult than the adult population. In addition, geolocated addresses has been a major step forward from real exposures have been found to vary widely between using just the administrative neighborhood [6]. However, social groups. That is mainly because of differences in in real life, most people move beyond their residential location patterns and daily mobility habits. Previous areas during their everyday travel, and thus static resi- research has found socioeconomically disadvantaged dential neighborhoods cannot capture the entire context groups to be more frequently in contact with hazard- of exposure [7–9]. Some have suggested that the litera- ous exposures [8]. Because of the link between travel ture needs to move beyond notions of contextual influ- and exposure, sociodemographic groups with differ- ence that rely on using such specific fixed locations [10, ent travel patterns are likely also to differ in their daily 11]. GPS and tracking technologies have been viewed as exposures. General assessments of exposure to the the optimal solution to go beyond where people live to overall population are, for example, not often applica- where people visit and how much time they spend at each ble to seniors due to their distinct spatial practices and particular location [12, 13]. GPS location and tracking travel behavior. trajectories can accurately identify people’s space–time While studying seniors’ exposure levels to air pol- trajectories, frequency, and duration which can provide lutants, noise or greenness is particularly important dynamic measures of exposure measures that stand in because of their larger chronic diseases’ prevalence contrast to traditional static address-based measures. [23, 24], these assessments often require specific study These dynamic measures of exposure can greatly improve designs given seniors’ special relationship with their the accuracy of exposure assessments either at the neighborhood and their most near built environment momentary level or at the activity space daily aggregate [25]. Seniors spend higher amounts of time in the close level [14–16]. vicinity of their residences [26–28] which would sug- Dynamic assessments of exposures also allow for gest that residential-based exposure assessments would avoiding commonly identified limitations of neighbor- provide accurate-enough exposure measures, but at the hood effects research, such as the uncertain geographic same time, they also tend to engage in higher amounts context problem (UGCoP) [17] or the neighbor- of walking trips [29] which increases their sensitivity to hood effect averaging problem (NEAP) [18]. Such environmental exposures. While the consensus seems approaches, however, are far more resource-inten- to be that seniors tend to have smaller and more com- sive, require more specific research designs [19], and pact activity spaces [30, 31] they cannot be considered may be affected by other common spatial biases such home-bound, as their daily mobility is complex and as the selective daily mobility bias [20]. Most impor- they show great variance in their activity spaces sizes tantly, studies using raw GPS data usually need to use and shapes [32]. Previous studies have found that given additional methods to add contextual information the proper built environment conditions, seniors can Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 3 of 16 significantly extend their living spaces and even spend Vegetation Index (NDVI) [61, 62, 65–67] can provide more time outside of their residential neighborhood optimal places to walk or recreate while also being neg- than inside of it [33, 34]. That is especially true in walk- atively correlated with other negative urban exposures able and compact cities such as Barcelona where sen- such as air or noise pollution. Among seniors, exposure iors most often used modes of transport -walking and to green space has been linked to many health outcomes, public transport- can take them far within the city. including mortality, social capital, obesity, and most fre- quently with physical activity [68–70]. Common environmental exposure assessments Given the high number of biases that affect neighbor- Choosing between static or dynamic exposure assess- hood effects research, more light is needed to define ment methods, however, will not only be dependent on what is the appropriate scale at which to measure expo- the population group, but also the nature of the expo- sure among specific population groups. Considering this sures. Studies comparing the spatial distribution of envi- need for a better understanding of when to use dynamic ronmental exposures such as carbon dioxide (CO2), Fine vs static research designs in studies involving the senior particulate matter PM2.5 and PM10, nitrogen dioxide population, this study sets to estimate exposure to air (NO2), or sulfur dioxide (SO2) have found little to no pollutants, noise, and greenness in a sample of 113 older spatial correlation between them [35–37] suggesting that people living in Barcelona and participating in a GPS- the decision on whether to use dynamic or static meas- tracking study for 7 days. The study sets out to answer urements would have to be also exposure dependent. whether using static exposure techniques based on the To date, studies estimating the effects of environmental residential address is enough to represent the totality exposures on seniors have focused on air pollution, noise of exposures experienced throughout a day. That is to pollution, and greenness [38–41]. understand if continuing to use static exposure measure- Studies estimating the effects of air pollution typically ments instead of dynamic-based ones will compromise use a combination of NO2, PM10, and PM25 as air qual- the reliability of exposure assessment studies among ity indicators. These pollutants are mainly emitted by seniors. internal combustion vehicles and are the most prevalent air pollutants in urban environments [42, 43], including Methods in Barcelona [44, 45]. Their public health dangers are well This study was set in the context of the RecerCaixa Pro- documented [46] and seniors are considered one of the ject (“Ciudad, calidad de vida y movilidad activa en la most at-risk populations [47]. Ambient particulate matter tercera edad. Un análisis multi-metodológico a través de is recognized as one of the main environmental risk fac- Tracking Living Labs”) and took place in the municipality tors for chronic respiratory diseases [48, 49] and cogni- of Barcelona. The project aimed to explore basic mobility tive decline among them [50, 51]. patterns of seniors (> 64 years old) living in the Barcelona Exposure to unwanted sound from industry, trans- metropolitan area along with quantifying the environ- port, or other urban activities is commonly referred to as mental, social and health issues that impact their daily noise pollution [52–54]. Noise pollution is a public health mobility choices. issue that has been is gaining prominence as scientific evidence linking noise and health increases [55]. Among Study design and population sample these links, studies have found noise to be associated With more than one-fifth of its population being over with sleep disturbances, hypertension, cardiovascular 65 years old, Barcelona is usually defined as a compact, risks, chronic stress, or disturbances in mental health [56, walkable, and vital city [71–74]. Barcelona’s morphologi- 57]. Recent analyses have also found seniors to be more cal conditions are representative of other historical Med- affected by noise than the rest of the population even iterranean cities, combining high population density and when exposed to a similar level of measured noise [58]. land use mix that create walkable environments but also Most worrisome, seniors also tend to live in areas with maintaining high car-use levels due to extensive motori- higher concentrations of noise pollution [59] something zation rates along with significant car-dependency in that Lagonigro et al. [60] also found in Barcelona. parts of the metropolitan area [75]. Due to its density and However, there are also positive environmental expo- high vehicle use, Barcelona has also high average levels of sures for seniors, such as exposure to greenness. Being in air pollution and noise pollution [76–78]. Despite hav- contact with nature, which includes having visual access ing a low ratio of green-space per inhabitant compared to green space, has been associated with a wide range of to other European cities (18 ­m2 including the peri-urban positive health effects, from physical to mental health forest of Collserola), small parks, streets, boulevards, and restorative processes [61–64]. Urban greenness, and plazas with trees often compensate for the need commonly measured using the Normalized Difference for spaces for outdoor activities and provide an optimal Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 4 of 16 distribution of NDVI (Normalized Difference Vegetation Participants were also asked to fill in a questionnaire dis- Index) levels [79, 80]. closing their age, gender, self-reported health, and per- Participants were recruited between June 2016 and ceived characteristics of their residential neighborhood. June 2017. Researchers contacted 39 senior centers scat- The Physical Activity Location Measurement System tered through the Barcelona metropolitan area to recruit (PALMS) v.R4 was used to aggregate data extracted from seniors to participate in a tracking-GPS data gather- GPS and accelerometer devices into 15 s intervals [16]. ing. Participants had to be 65 years old or above and Spatial exposure to air quality components was based not have specific mobility impairments. After being on the air quality inmission maps provided by the Bar- informed written and orally about the study, and pro- celona municipality with data from 2019 (Fig. 1). These vided with research protocols and instructions, 269 par- maps provide data on nitrogen dioxide (NO2), and sus- ticipants gave informed consent, out of those, we focused pended particulates PM10 and PM2.5 based on annual on those that lived within the limits of the Barcelona averages and are calculated at the street section level. The municipality (n = 113). Confidentiality was ensured by spatial modeling of the dispersion of pollutants is per- using random identification numbers and data censor- formed by the municipality and offers a complete map ing. The study was approved by the Ethics Committee of average immission levels at the street-section scale on Animal and Human Experimentation at Universitat (Fig. 2L). Similar to Chum and Ocampo [81] and Nyhan Autònoma de Barcelona (UAB; CEEAH-3656). Consid- et al. [12] we used average daily NO2, PM10, and PM2.5 ering that seniors’ outdoor behavior in relation to the concentrations as a proxy of exact exposures. built environment presents differences according to their Data on noise exposure were based on the noise con- characteristics, we classified our sample by age [younger tour maps from the strategic noise map of the city of (< 75 years old), and older (≥ 75 years old) seniors] and Barcelona, obtained through the Barcelona Open Data gender. portal with data from 2017 (Fig. 2). These are the result of the collaboration between the Barcelona city council Data collection and the Barcelona Public Health Agency (ASPB) and rep- To collect data regarding the routine travel behavior of resent noise levels on a daily average (7 am to 9 pm). The seniors, participants were asked to wear a GPS device spatial distribution of noise exposure is represented in (QStarz BT-Q1000X; QStarz International Co., Ltd., Fig. 2 (left). Taiwan, R.O.C.) and a wrist-worn accelerometer (Acti- Finally, we used NDVI to represent the levels of urban graph GT3X + ; ActiGraph LLC, Pensacola, Florida USA) greenness in Barcelona. Due to healthy green vegetation for seven consecutive days. Valid days included at least reflecting more infrared radiation and absorbing more four wearing days and ten hours of device wear-time. energy in the red wavelength compared to unhealthy Fig. 1 Daily air pollution levels spatial distribution in Barcelona: PM10 (left), PM2.5 (center), NO2 (right). Source: Ajuntament de Barcelona Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 5 of 16 Fig. 2 Daily noise levels (left) and greenness (right) spatial distribution in Barcelona vegetation or surfaces without vegetation, NDVI is com- before in Barcelona [61, 62]. The distribution of NDVI monly used to assess vegetation levels. The NDVI data- levels in Barcelona is presented in Fig. 2 (right). base of Catalonia 2019 available on the portal of the The spatial resolution of environmental exposures Cartographic and Geological Institute of Catalonia [82] maps allowed us to pair average environmental condi- was used. NDVI orthoimages are generated from images tions to GPS points using GIS spatial analysis techniques. obtained by an aerial photogrammetric camera, capa- These allowed us to assess the variability of intensity of ble of obtaining information not only from the visible exposures based on spatial location and movements of range of the electromagnetic spectrum but also from seniors. It was, however, not possible to assess temporal the near-infrared, and with a pixel size of about 25.cm variability as environmental exposures represented daily on the ground. NDVI scores can vary from -1 to 1 when averages and were thus not disaggregated by hour or time the result is < 0 corresponds to areas without vegetation of day. cover corresponding to water or artificial surfaces, for A series of covariates were extracted from the par- values 0 < NDVI < 0.2 it is expected to find bare soil or ticipants baseline survey such as gender and age dead vegetation, between 0.2 < NDVI < 0, 4 corresponds (mean = 74.9; sd = 7.56). Based on their place of residence to soil with sparse or not very vigorous vegetation, in we also calculated their income and population density cases of 0.4 < NDVI < 0.6, areas with vigorous and abun- based on census track data. In order to ease interpreta- dant vegetation are estimated, while those exceeding < 0.6 tion, we also grouped seniors per types of built environ- correspond to areas with dense and vigorous vegeta- ment with historic district representing the older parts tion [82]. NDVI use is very common in studies assess- of the city, Expansion representing the compact develop- ing greenness exposure [83–86] and has also been used ment starting around 1850 and sprawl areas representing Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 6 of 16 the more recent development areas characterized by low [5, 14]. We chose a 20-m buffer because the GPS accu- densities and high car dependency. racy cannot distinguish which side of the street the par- ticipant is using and thus covering the whole street width GIS processing was deemed necessary. Using the self-reported home address of each participant Once residential and dynamic exposure ranges were we created a street network buffer of 600 m. Within these defined, we used spatial join to average the exposures 600 m, the second buffer of 20 m was estimated, calcu- within each buffer of 600 m, thus creating daily aver- lated by including only those roads that are walkable. age exposures for NO2, PM10, PM2.5, noise, and NDVI Thus, the residential area of exposure is the result of a respectively (Fig. 4). All processes were conducted using network buffer of 20 m around the roads where walking ArcMap 10.7.1. is possible and that falls within 600 m from the place of residence. to define the residential extent of exposures Statistical analysis (Fig. 3, left). The 600-m buffer was estimated based on Using daily averages of exposure, we tested the differ- previous studies on older adult mobility [71, 87, 88] and ences between home and dynamic average daily expo- corresponds to approximately 10 min of walking for that sures. Pairwise correlations between exposures are age group which has been a walkable distance for the available in Additional file 1: Table S1. To do so we used population group. The 20-m buffer corresponds to the paired sample t-test to examine home-dynamic averages average street width in Barcelona [79]. To capture the that were significantly different from each other. NO2, range of dynamic exposure we used a 20-m buffer around PM10, PM2.5, noise, and NDVI daily average exposure all walking tracks accumulated by each participant measurements were compared using t-tests to assess throughout each participating day (Fig. 3, right) following the overall level of agreement. To visualize exposure Fig. 3 Residential (left) and dynamic (right) exposure buffers of one day of data collection Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 7 of 16 Fig. 4 Example of noise exposure calculation based on residential (left) and dynamic (right) buffers differences, we used boxplots while stratifying the results Results by age group and gender of the participants. Average exposure levels for the main three air quality To examine the role of any sociodemographic or con- pollutants -PM10, PM2.5, NO2-, are shown in Table 1. textual variable driving the differences in exposure, we Differences between home and dynamic exposure are used a 5 model multivariable analysis in which we use tested using paired samples t-tests. multilevel Poisson models estimating the differences In terms of PM10, the average participant was exposed between residential and dynamic environmental expo- in a single day to areas with 25.64 µg/m3 annual immis- sures. Each model uses the difference between home sion levels, while their dynamic exposure was 25.9 µg/m3. and dynamic exposure levels for each environmental While a statistically significant difference, the estimated exposure type (model 1: PM10; model 2: PM2.5; model value using GPS tracking was only 1.0% different than 3: NO2; model 4: dB; model 5: NDVI). Each model esti- the estimated value using the home exposure (p = 0.006). mates the effect of gender and age of the participant and When stratified by population groups, the larger statisti- the urban layout and average income level of the partici- cally significant differences between home and dynamic pant’s home neighborhood while also controlling for the PM10 exposure were found in those participants that number of walking trips conducted each day. To account were older than 75 years old, who registered dynamic for the nested nature of our data, we employ multilevel exposures 1.92% higher than home exposure (p < 0.001). models using the participant’s ID as a random effect. All Other statistically significant differences were found analysis were run using Stata v16. among men (1.74%; p < 0.001), those living in the Expan- sion area of the city (1.53%; p < 0.001), and those living in Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 8 of 16 Table 1 Home and dynamic daily exposures to PM10, PM2.5, and NO2 per population group PM10 PM2.5 NO2 Homea Dynamicb Diff. %c pd Homea Dynamicb Diff. %c pd Homea Dynamicb Diff. %c pd Total (N = 113) 25.64 25.90 1.00 0.006 16.17 17.16 5.77 0.000 37.55 38.05 1.33 0.036 Sex Men (N = 55) 25.71 26.17 1.74 0.001 16.27 17.21 5.47 0.000 37.23 37.91 1.77 0.022 Women (N = 58) 25.57 25.64 0.29 0.559 16.07 17.11 6.05 0.000 37.84 38.19 0.92 0.357 Age 65 to 74 y.o. (N = 58) 25.85 25.89 0.14 0.808 16.17 17.07 5.25 0.000 37.93 38.29 0.96 0.315 75 + y.o. (N = 55) 25.41 25.91 1.92 0.000 16.16 17.25 6.32 0.000 37.14 37.79 1.73 0.036 Urban layout Expansion (N = 91) 25.80 26.20 1.53 0.000 16.50 17.50 5.69 0.000 38.05 38.60 1.41 0.042 Sprawl (N = 6) 23.69 22.48 − 5.37 0.259 14.16 15.20 6.81 0.022 31.20 31.41 0.67 0.838 Historic District (N = 16) 25.43 25.42 − 0.03 0.923 15.01 15.94 5.89 0.000 36.98 37.39 1.09 0.227 Income High (N = 68) 25.75 26.05 1.15 0.000 15.68 16.52 5.13 0.000 39.08 39.37 0.76 0.384 Low (N = 45) 25.48 25.68 0.79 0.173 16.88 18.08 6.61 0.000 35.33 36.14 2.24 0.001 Density High (N = 107) 25.74 26.08 1.30 0.000 16.28 17.26 5.72 0.000 37.89 38.41 1.36 0.025 Low (N = 6) 23.69 22.48 − 5.37 0.259 14.16 15.20 6.81 0.022 31.20 31.41 0.67 0.838 a Residence-based exposure measured on a 600 m street-network buffer from the geocoded participant’s address b Dynamic-based exposure measured on walking Daily Path Areas c Difference between Home and Dynamic exposures as a percentage of home exposure d Paired samples t-test high-income areas (1.15%; p < 0.001), although no differ- 11%. The larger differences are found among those who ence was greater than 2%. live in the Expansion area (− 11.06; p < 0.001) or the his- In terms of exposure to PM2.5, the average daily expo- toric district (− 11.1; p < 0.001) and among those living in sure difference was 5.77% higher when calculated using high-income areas (− 11; p < 0.001). dynamic GPS tracking than using the home method Greenness for its part did not show major significant (p < 0.001). All population groups registered statistically differences when assessed using the home range or the significant differences, with no differences below 5%. dynamic exposure (0.44%; p = 0.708). Only those living The maximum differences were found among women in the historic district of the city recorded a 5% differ- (6.05%; p < 0.001), the older population (6.32%; p < 0.001), ence between the home and the dynamic exposure, with and those living in sprawl or low-density areas (6.81%; the dynamic exposure being higher (home = 0.190 vs p = 0.022). dynamic = 0.201; p = 0.004). The NO2 exposure assessment registered a daily expo- The graphical representation of the differences between sure difference between the home and dynamic method residence and dynamic exposures (Fig. 5) shows there are of 1.33 on average (1.33%; p = 0.036). The only statistically higher mean differences in the case of PM2.5, noise, and significant differences were found among men (1.77%; NDVI. When focusing on gender and age, women and p = 0.022), the older population (1.73%; p = 0.036), those those older than 75 portray higher mean differences in living in the Expansion (1,41%; p = 0.042) and most sig- noise and NDVI, while men and elders’ younger than 75 nificantly in low-income areas (2.24%; p = 0.001). show higher mean differences in PM2.5 and NO2, thus in Average exposure levels for noise (dB) and greenness air pollution. (NDVI) are shown in Table 2. In terms of noise, the aver- When using multilevel Poisson models to examine the age participant lived in an area with an average of 58.6 role of individual sociodemographic or contextual vari- daily db. In contrast, the average participant was exposed ables driving the differences in exposure, we spot signifi- to areas with 64.6db during his/her daily walking trips. cant differences between the types of exposure (Table 3). The estimated dynamic exposure was thus 9.35% lower Model 1 estimates differences in PM10 to be signifi- than the home exposure (p < 0.001). Almost all popula- cantly higher for those living in sprawled areas when tion groups share that same difference between 9 and compared to those living in historic districts and acting Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 9 of 16 Table 2 Home and dynamic daily exposures to noise (dB) and greenness (NDVI) per population group dB NDVI Homea Dynamicb Diff. %c pd Homea Dynamicb Diff. %c pd Total (N = 113) 58.66 64.62 − 9.35 0.000 0.237 0.24 0.44 0.708 Sex Men (N = 55) 58.61 64.42 − 9.03 0.000 0.236 0.23 − 1.53 0.340 Women (N = 58) 58.54 68.40 − 9.66 0.000 0.238 0.24 2.21 0.189 Age 65 to 74 y.o. (N = 58) 65.37 59.29 − 10.26 0.000 0.225 0.235 − 0.23 0.875 75 + y.o. (N = 55) 63.81 57.80 − 10.39 0.000 0.249 0.25 1.07 0.539 Urban layout Expansion (N = 91) 65.48 58.60 − 11.06 0.000 0.239 0.24 0.47 0.697 Sprawl (N = 6) 58.92 57.50 − 2.48 0.197 0.336 0.31 − 8.92 0.002 Historic District (N = 16) 65.48 58.96 − 11.1 0.000 0.190 0.20 5.44 0.004 Income High (N = 68) 66.12 59.56 − 11.00 0.000 0.216 0.21 − 1.25 0.272 Low (N = 45) 62.45 57.14 − 9.29 0.000 0.268 0.27 2.33 0.160 Density High (N = 107) 64.93 58.63 − 10.74 0.000 0.232 0.23 15.48 0.330 Low (N = 6) 58.93 57.5 − 2.48 0.197 0.336 0.31 − 8.92 0.002 a Residence-based exposure measured on a 600 m street-network buffer from the geocoded participant’s address b Dynamic-based exposure measured on walking Daily Path Areas c Difference between Home and Dynamic exposures as a percentage of home exposure d Paired samples t-test as a reference. Model 2 finds no significant association dynamic PM2.5 exposure [21, 89, 90]. Both NO2 and between PM2.5 and any of the introduced variables. PM10 presented discrepancies below 1.5% leading us Model 3—focused on NO2—finds expected differences to believe that the potential gains in accuracy derived in home-dynamic exposure to be 70% higher in those liv- from the use of GPS tracking would not outweigh its ing in sprawl areas when compared to those living in his- challenges and burdens. Our results suggest that among torical districts, once the rest of the covariates are taken seniors, only studies dealing with PM10 and NO2 expo- into account. Model 4 and Model 5 dedicated to esti- sure among very specific population sub-groups would mating differences in home-dynamic exposures to noise require a GPS-tracking methodology. An example of that and greenness respectively, did not find any associations would be studies focused on the older age range (sen- between home-based measures of exposure and GPS- iors over 75 years old), or in low-income areas, in which track-based dynamic measures of exposure once all the discrepancies between static and dynamic exposures control variables are taken into account. are closer to 2%. In the most complete study to date on the differences between static and dynamic measures of Discussion and conclusions traffic-related exposure set in Shenzen, China, Yu et al. Our results indicate that significant differences between [40, 41] found the static measures to overrepresent expo- static and dynamic exposure assessments are only pre- sures by almost 30%. Our results confirm Yu’s findings sent in selected exposures and would thus suggest that for the specific case of seniors and the major discrepan- dynamic assessments using GPS-tracking are not provid- cies found when assessing PM2.5. Our models, however, ing superior accuracy across all the ranges of exposures. seem to indicate that in the context of Barcelona static Regarding air quality—measured using NO2, PM10, measures are underrepresenting real exposures and not and PM2.5—our data suggest that dynamic exposure overrepresenting them. would only be recommended in the case of the smaller Regarding noise exposure, our estimates suggest that particulate matter (PM2.5) for which a discrepancy of studies focusing on seniors and noise would greatly ben- almost 6% between residential and dynamic exposures efit from dynamic exposure assessment and the use of was detected. This is consistent with previous findings GPS tracking. According to our data, static home-based that have found major differences between static and measures tended to overestimate noise exposure by 10%, Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 10 of 16 Fig. 5 Distribution boxplots for differences between residence and dynamic exposure per gender, age When we interact the differences by gender and age (Fig. 6) we can spot similar differences with very little changes in the means but with interquartile ranges being consistently wider in the case of males, especially in the case of NDVI and that overestimation was higher in areas with heavy In part, these results might be explained by the high traffic, such as the example or historic districts. Interpre- density of curbside street trees in Barcelona [79]. Linear tation of dynamic vs static measures of sound exposure in methods calculated around pedestrian tracks might be the literature is diverse. While Kou et al. [91] argue that positively affected by the large presence of curbside trees. mobility constrained groups such as seniors might find In contrast, because Barcelona has a lack of parks and more difficulties in adjusting their behavior to the pres- open green spaces, home exposure to NDVI might also ence of high noise pollution, others such as Ma et al. [89, be limited to curbside trees. Previous studies have also 90] have argued that residents of high-noise pollution failed to find a significant difference between dynamic areas might be self-selecting their mobility routes and and static exposures to greenness [92]. Differences destinations to travel to less noisy areas which may lower in terms of NDVI exposure in Barcelona seem highly their average dynamic-exposure. In that specific case, dependent on the residential location, suggesting that in the fact that seniors don’t have mobility patterns that are this case, it might be worth considering the kind of urban strictly fixed by the presence of the workplace would con- area that is being studied before deciding on whether to tribute to their flexibility to self-select for less noisy areas. use dynamic exposures. In the specific case of greenness Finally, studies interested in the exposure of seniors to and seniors, home-based exposure is even more impor- greenness and NDVI, would not seem to benefit from tant given the links between home-based measures of dynamic exposures in urban areas similar to Barcelona. greenness and mental health [67]. Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 11 of 16 Fig. 6 Distribution boxplots for differences between residence and dynamic exposure per the interaction between gender and age Overall, our study failed to uncover large discrepancies capacities may find it difficult to lower their exposures to between static and dynamic exposures. This might be hazardous local conditions. In the specific case of Bar- explained by several factors. First, the study population— celona however, previous research has demonstrated seniors above 65 years old—are also one of the popula- that the seniors increased use of proximity and the local tion groups that tend to use their local neighborhood the neighborhood is a matter of preference and potential of most, with travel patterns that are usually concentrated the built environment [28, 96] rather than a case of spa- around their home residence and with activity spaces tial entrapment caused by limited mobility options [97]. that cover fewer distances away from home [29]. Taking The specific characteristics of the local built environ- most trips in the vicinities of home would likely attenuate ment in Barcelona may also contribute to explaining the the expected accuracy benefits from a dynamic exposure general low accuracy gains from dynamic exposures, as approach. This would confirm the previous hypothesis low variance in the built environment characteristics that states that daily exposures estimates obtained from has also been known to affect other similar study set- the two approaches only differ substantially if an indi- tings [98]. Studies dealing with more diverse urban envi- vidual’s time spent away from home is large [21, 93–95]. ronments might accrue additional gains from dynamic These findings, supported by the study by Yu et al. [40, exposure, suggesting that the decision on whether or not 41] in China suggest that dynamic exposure assessment to use a tracking methodology and dynamic measures may only be warranted when studying those population would depend not only on the population group that is groups that tend to spend more time in out-of-home being studied but also be location dependent. activities in nonresidential neighborhoods and are thus In the interpretation of results a couple of limita- exposed to considerable different conditions over the day. tions need to be considered. For once, mobility data In practical terms, the fact that most seniors have lower and exposure data came from different sources as par- mobility rates makes them potentially more vulnerable ticipants were not equipped with specific devices to to the conditions of their local neighborhoods. As such, capture exposure while they were moving. Thus, the seniors with higher mobility ranges may be able to select opportunistic use of city council exposure data creates for the less polluted environment during their daily eve- a time-gap between the moment on which the tracking ryday mobility, while seniors with more limited mobility took place (2017) and the exposure data (2018). This Table 3 Multilevel Poisson models estimating the differences between residential and dynamic environmental exposures among seniors Model 1 Model 2 Model 3 Model 4 Model 5 PM10 PM2.5 NO2 dB NDVI Coef. P>z 95% Conf. Coef. P>z 95% Conf. Coef. P>z 95% Conf. Coef. P>z 95% Conf. Coef. P>z 95% Conf. Marquet et al. International Journal of Health Geographics Gender Women = ref. = ref. = ref. = ref. = ref. Men 0.537 0.163 − 0.22; 1.29 0.279 0.385 − 0.79; 0.31 0.031 0.928 − 0.64; 0.70 0.084 0.735 − 0.40; 0.57 0.098 0.64 − 0.31; 0.51 Urban layout (2023) 22:3 Historic district = ref. = ref. = ref. = ref. Sprawl 2.684 0.001 1.07; 4.30 0.643 0.448 − 0.72; 1.75 1.741 0.02 0.28; 3.20 0.386 0.569 − 0.94; 0.71 0.066 0.896 − 0.93; 1.06 Expansion 0.618 0.402 − 0.83;2.06 0.418 0.963 − 0.80; 0.84 0.456 0.447 − 0.72; 1.63 0.355 0.371 − 0.42; 1.13 − 0.194 0.527 − 0.80; 0.41 Income High = ref. = ref. = ref. Low − 0.004 0.991 − 0.76; 0.75 0.291 0.851 − 0.63; 0.52 − 0.081 0.819 − 0.77; 0.61 − 0.276 0.297 − 0.80; 0.24 0.226 0.304 − 0.21; 0.66 Age 64–74 y.o = ref. = ref. = ref. 75 y.o − 0.407 0.286 − 1.15; 0.34 0.275 0.911 − 0.57; 0.51 − 0.396 0.247 − 1.07; 0.28 0.133 0.594 − 0.36; 0.62 0.261 0.218 − 0.15; 0.68 N Trips − 0.017 0.825 − 0.17;0.14 0.056 0.658 − 0.14; 0.08 − 0.006 0.927 − 0.14; 0.13 − 0.001 0.978 − 0.10; 0.09 − 0.05 0.245 − 0.14; 0.03 Cons − 3.994 0.000 − 5.60; − 2.38 0.48 0.000 − 3.25; − 1.37 − 3.233 0.000 − 4.56; − 1.91 − 2.611 0.000 − 3.51; 1.71 − 1.907 0.000 − 2.61; − 1.21 Page 12 of 16 Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 13 of 16 may create small accuracy inconsistencies that need to Availability of data and materials The datasets used and/or analyzed during the current study are available from be considered, however in recent years Barcelona has the corresponding author upon reasonable request. failed to produce a significant change in the levels of emissions, noise or greenness [99, 100] and that makes Declarations us confident that baseline conditions did not change significantly between 2017 and 2018. Secondly, the use Ethics approval and consent to participate The study received the approval of the Ethics Committee on Animal and of NDVI as a measure of greenness of exposure differs Human Experimentation at the Universitat Autònoma de Barcelona on 2 in nature to the other measures of exposure as NDVI in February 2017 (CEEAH-3656). All participants signed informed consent docu‑ not a physical measure but a simplified indicator which ments. All procedures have been performed in accordance with the Declara‑ tion of Helsinki. is highly sensor dependent. This could affect future rep- licability of the study’s findings. Consent for publication This study is the first to assess the need for dynamic All authors have approved and reviewed the final manuscript and approved its publication. exposure assessment when studying exposures of seniors. In this case, we use seniors as an example of a popula- Competing interests tion group with low mobility levels among which using The authors declare that they have no competing interests. dynamic measures of exposure would not accrue sig- Author details nificant accuracy gains in environmental exposure 1 Institute of Environmental Science and Technology (ICTA‐UAB), Universitat assessments. This qualifies the often extended idea that Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Barcelona, Spain. 2 Research Group on Mobility, Transportation and Territory (GEMOTT), Depart‑ dynamic measures of exposure are greatly needed across ment of Geography, Universitat Autònoma de Barcelona, 08193 Cerdanyola all kinds of studies [101] and would point to the need del Vallès, Barcelona, Spain. 3 Cosmopolis Centre for Urban Research, Vrije to adapt the use of dynamic vs static measures of expo- Universiteit Brussels, Brussels, Belgium. sure to the kind of mobility patterns of the population Received: 2 November 2022 Accepted: 13 January 2023 group of interest. Generally, studies on environmental exposures need to consider human mobility and spatial variations of exposures if they want to avoid misinter- pretations. Accurate estimates of exposures are key for References policymakers to identify spatial or social inequities in 1. Callaghan A, McCombe G, Harrold A, McMeel C, Mills G, Moore-Cherry exposures and to design interventions that can alleviate N, Cullen W. The impact of green spaces on mental health in urban set‑ tings: a scoping review. J Ment Health. 2021;30(2):179–93. https://​doi.​ them. The search for accuracy cannot hide, however, the org/​10.​1080/​09638​237.​2020.​17550​27. need to use more efficient and cost-effective methods for 2. Dominski FH, Lorenzetti Branco JH, Buonanno G, Stabile L, Gameiro each research question. da Silva M, Andrade A. Effects of air pollution on health: a mapping review of systematic reviews and meta-analyses. Environ Res. 2021;201: 111487. https://​doi.​org/​10.​1016/j.​envres.​2021.​111487. Supplementary Information 3. Münzel T, Sørensen M, Daiber A. Transportation noise pollution and The online version contains supplementary material available at https://​doi.​ cardiovascular disease. Nat Rev Cardiol. 2021;18(9):Article 9. https://​doi.​ org/​10.​1186/​s12942-​023-​00325-8. org/​10.​1038/​s41569-​021-​00532-5. 4. Kwan MP. Beyond space (as we knew it): toward temporally integrated geographies of segregation, health, and accessibility: space-time Additional file 1: Table S1. Bi-variate correlations between static and integration in geography and GIScience. Ann Assoc Am Geogr. dynamic measures of exposure. 2013;103(5):1078–86. https://​doi.​org/​10.​1080/​00045​608.​2013.​792177. 5. Marquet O, Floyd MF, James P, Glanz K, Jennings V, Jankowska MM, Kerr J, Hipp JA. Associations between worksite walkability, greenness, Author contributions and physical activity around work. Environ Behav. 2020;52(2):139–63. OM and MM conceptualized the study and developed the conceptual frame‑ https://​doi.​org/​10.​1177/​00139​16518​797165. work. IG-V, OM, MM collected the GPS data. JT and DT led the data processing 6. Petrović A, Manley D, van Ham M. Freedom from the tyranny of neigh‑ and analysis. OM supervised the data processing and interpretation of results. bourhood: rethinking sociospatial context effects. Prog Hum Geogr. OM and MM led the writing of the manuscript. All authors read and approved 2020;44(6):1103–23. https://​doi.​org/​10.​1177/​03091​32519​868767. the final manuscript. 7. Helbich M. Toward dynamic urban environmental exposure assess‑ ments in mental health research. Environ Res. 2018;161(June Funding 2017):129–35. https://​doi.​org/​10.​1016/j.​envres.​2017.​11.​006. This project has received funding from the European Union’s Horizon 2020 8. Tao Y, Chai Y, Zhang X, Yang J, Kwan MP. Mobility-based environmental research and innovation programme under the Marie Sklodowska-Curie grant justice: understanding housing disparity in real-time exposure to air agreement No 845570, funding from the LaCaixa Foundation project STEPP pollution and momentary psychological stress in Beijing, China. Soc (Social Transport Equity by Planning for Proximity) (SR22-00147) and the Span‑ Sci Med. 2021;287(September): 114372. https://​doi.​org/​10.​1016/j.​socsc​ ish Ministry of Science and Innovation ECOMOV (The ecological transition imed.​2021.​114372. in mobility and transport. The role of urban proximity) project (TED2021- 9. Wang F. Why public health needs GIS: a methodological overview. Ann 129280B-I00) Oriol Marquet is funded by a Ramón y Cajal fellowship (RYC- GIS. 2020;26(1):1–12. https://​doi.​org/​10.​1080/​19475​683.​2019.​17020​99. 2020-029441-I) awarded by the Spanish Ministry of Science and Innovation. 10. Kwan MP. The limits of the neighborhood effect: contextual uncertain‑ ties in geographic, environmental health, and social science research. Ann Am Assoc Geogr. 2018;108(6):1482–90. https://​doi.​org/​10.​1080/​ 24694​452.​2018.​14537​77. Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 14 of 16 11. Sheller M, Urry J. The new mobilities paradigm. Environ Plan A. transportation over ten years. J Transp Health. 2017;7:190–201. https://​ 2006;38(2):207–26. https://​doi.​org/​10.​1068/​a37268. doi.​org/​10.​1016/j.​jth.​2017.​09.​006. 12. Nyhan M, Grauwin S, Britter R, Misstear B, McNabola A, Laden F, Barrett 30. Douma L, Steverink N, Meijering L. Geographical life-space and subjec‑ SRH, Ratti C. ’Exposure track’—the impact of mobile device based tive wellbeing in later life. Health Place. 2021;70(February): 102608. mobility patterns on quantifying population exposure to air pollution. https://​doi.​org/​10.​1016/j.​healt​hplace.​2021.​102608. Environ Sci Technol. 2016. https://​doi.​org/​10.​1021/​acs.​est.​6b023​85. 31. Prescott M, Miller WC, Routhier F, Mortenson WB. Factors affecting the 13. Park YM, Kwan MP. Individual exposure estimates may be erroneous activity spaces of people who use mobility devices to get around the when spatiotemporal variability of air pollution and human mobility are community. Health Place. 2020;64(June): 102375. https://​doi.​org/​10.​ ignored. Health Place. 2017;43(November 2016):85–94. https://​doi.​org/​ 1016/j.​healt​hplace.​2020.​102375. 10.​1016/j.​healt​hplace.​2016.​10.​002. 32. Kestens Y, Thierry B, Shareck M, Chaix B. Integrating activity spaces in 14. James P, Jankowska M, Marx C, Hart JE, Berrigan D, Kerr J, Hurvitz PM, health research: comparing the VERITAS activity space questionnaire Hipp JA, Laden F. “Spatial energetics”: integrating data from GPS, accel‑ with 7-day GPS tracking and prompted recall. Spatial Spatio-Temporal erometry, and GIS to address obesity and inactivity. Am J Prev Med. Epidemiol. 2018. https://​doi.​org/​10.​1016/j.​sste.​2017.​12.​003. 2016;51(5):792–800. https://​doi.​org/​10.​1016/j.​amepre.​2016.​06.​006. 33. Mariotti I, Burlando C, Landi S. Is local public transport unsuitable for 15. Jankowska M, Jiue-An Y, Nana L, Chad S, Tarik B. Accounting for space, elderly? Exploring the cases of two Italian cities. Res Transp Bus Manag. time, and behavior using GPS derived dynamic measures of environ‑ 2021;40(January): 100643. https://​doi.​org/​10.​1016/j.​rtbm.​2021.​100643. mental exposure. Health Place. 2021. https://​doi.​org/​10.​1016/j.​healt​ 34. York Cornwell E, Cagney KA. Aging in activity space: results from hplace.​2021.​102706. smartphone-based GPS-tracking of urban seniors. J Gerontol Ser B, 16. Jankowska MM, Schipperijn J, Kerr J. A framework for using GPS data Psychol Sci Soc Sci. 2017;72(5):864–75. https://​doi.​org/​10.​1093/​geronb/​ in physical activity and sedentary behavior studies. Exerc Sport Sci Rev. gbx063. 2014;43(1):48–56. https://​doi.​org/​10.​1249/​JES.​00000​00000​000035. 35. Gelb J, Apparicio P. Cyclists’ exposure to atmospheric and noise pollu‑ 17. Kwan MP. How GIS can help address the uncertain geographic context tion: a systematic literature review. Transp Rev. 2021. https://​doi.​org/​10.​ problem in social science research. Ann GIS. 2012;18(4):245–55. https://​ 1080/​01441​647.​2021.​18953​61. doi.​org/​10.​1080/​19475​683.​2012.​727867. 36. Hatzopoulou M, Weichenthal S, Barreau G, Goldberg M, Farrell W, 18. Kwan MP. The neighborhood effect averaging problem (NEAP): an Crouse D, Ross N. A web-based route planning tool to reduce cyclists’ elusive confounder of the neighborhood effect. Int J Environ Res Public exposures to traffic pollution: a case study in Montreal, Canada. Environ Health. 2018. https://​doi.​org/​10.​3390/​ijerp​h1509​1841. Res. 2013;123(2):58–61. https://​doi.​org/​10.​1016/j.​envres.​2013.​03.​004. 19. Campbell M, Marek L, Hobbs M. Reconsidering movement and 37. Pattinson W, Kingham S, Longley I, Salmond J. Potential pollution expo‑ exposure: towards a more dynamic health geography. Geogr Compass. sure reductions from small-distance bicycle lane separations. J Transp 2021. https://​doi.​org/​10.​1111/​gec3.​12566. Health. 2017;4:40–52. https://​doi.​org/​10.​1016/j.​jth.​2016.​10.​002. 20. Shrestha S, Kestens Y, Thomas F, El Aarbaoui T, Chaix B. Spatial access 38. Ali MJ, Rahaman M, Hossain SI. Urban green spaces for elderly human to sport facilities from the multiple places visited and sport practice: health: a planning model for healthy city living. Land Use Policy. assessing and correcting biases related to selective daily mobility. Soc 2022;114: 105970. https://​doi.​org/​10.​1016/j.​landu​sepol.​2021.​105970. Sci Med. 2019;236: 112406. https://​doi.​org/​10.​1016/j.​socsc​imed.​2019.​ 39. Yang Y, Tang R, Qiu H, Lai P-C, Wong P, Thach T-Q, Allen R, Brauer M, Tian 112406. L, Barratt B. Long term exposure to air pollution and mortality in an 21. Yoo EH, Rudra C, Glasgow M, Mu L. Geospatial estimation of indi‑ elderly cohort in Hong Kong. Environ Int. 2018;117:99–106. https://​doi.​ vidual exposure to air pollutants: moving from static monitoring to org/​10.​1016/j.​envint.​2018.​04.​034. activity-based dynamic exposure assessment. Ann Assoc Am Geogr. 40. Yu X, Ivey C, Huang Z, Gurram S, Sivaraman V, Shen H, Eluru N, Hasan 2015;105(5):915–26. https://​doi.​org/​10.​1080/​00045​608.​2015.​10542​53. S, Henneman L, Shi G, Zhang H, Yu H, Zheng J. Quantifying the impact 22. Miyasaka T, Oba A, Akasaka M, Tsuchiya T. Sampling limitations in using of daily mobility on errors in air pollution exposure estimation using tourists’ mobile phones for GPS-based visitor monitoring. J Leis Res. mobile phone location data. Environ Int. 2020;141(April): 105772. 2018;49(3–5):298–310. https://​doi.​org/​10.​1080/​00222​216.​2018.​15425​ https://​doi.​org/​10.​1016/j.​envint.​2020.​105772. 26. 41. Yu Y, Paul K, Arah OA, Mayeda ER, Wu J, Lee E, Shih I-F, Su J, Jerrett M, 23. Christensen K, Doblhammer G, Rau R, Vaupel JW. Ageing populations: Haan M, Ritz B. Air pollution, noise exposure, and metabolic syn‑ the challenges ahead. Lancet. 2009;374(9696):1196–208. https://​doi.​ drome—a cohort study in elderly Mexican-Americans in Sacramento org/​10.​1016/​S0140-​6736(09)​61460-4. area. Environ Int. 2020;134: 105269. https://​doi.​org/​10.​1016/j.​envint.​ 24. Enssle F, Kabisch N. Urban green spaces for the social interaction, 2019.​105269. health and well-being of older people—an integrated view of urban 42. Cai Y, Hansell AL, Blangiardo M, Burton PR, De Hoogh K, Doiron D, For‑ ecosystem services and socio-environmental justice. Environ Sci Policy. tier I, Gulliver J, Hveem K, Mbatchou S, Morley DW, Stolk RP, Zijlema WL, 2020;109(March):36–44. https://​doi.​org/​10.​1016/j.​envsci.​2020.​04.​008. Elliott P, Hodgson S. Long-termexposure to road traffic noise, ambient 25. Van Cauwenberg J, De Bourdeaudhuij I, Clarys P, Nasar J, Salmon J, air pollution, and cardiovascular risk factors in the HUNTand lifelines Goubert L, Deforche B. Street characteristics preferred for transporta‑ cohorts. Eur Heart J. 2017;38(29):2290–6. https://​doi.​org/​10.​1093/​eurhe​ tion walking among older adults: a choice-based conjoint analysis artj/​ehx263. with manipulated photographs. Int J Behav Nutr Phys Act. 2016;13(1):6. 43. de Hoogh K, Chen J, Gulliver J, Hoffmann B, Hertel O, Ketzel M, Bauwel‑ https://​doi.​org/​10.​1186/​s12966-​016-​0331-8. inck M, van Donkelaar A, Hvidtfeldt UA, Katsouyanni K, Klompmaker 26. Barnett DW, Barnett A, Nathan A, Van Cauwenberg J, Cerin E. Built J, Martin RV, Samoli E, Schwartz PE, Stafoggia M, Bellander T, Strak M, environmental correlates of older adults’ total physical activity and Wolf K, Vienneau D, et al. Spatial PM 2.5, NO2, O3 and BC models for walking: a systematic review and meta-analysis. Int J Behav Nutr Phys Western Europe—evaluation of spatiotemporal stability. Environ Int. Act. 2017;14(1):1–24. https://​doi.​org/​10.​1186/​s12966-​017-​0558-z. 2018;120(2):81–92. https://​doi.​org/​10.​1016/j.​envint.​2018.​07.​036. 27. Cerin E, Nathan A, Van Cauwenberg J, Barnett DW, Barnett A. The 44. Querol X, Alastuey A, Rodriguez S, Plana F, Ruiz CR, Cots N, Massagué G, neighbourhood physical environment and active travel in older adults: Puig O. PM10 and PM2.5 source apportionment in the Barcelona Met‑ a systematic review and meta-analysis on behalf of the Council on ropolitan area, Catalonia, Spain. Atmos Environ. 2001;35(36):6407–19. Environment and Physical Activity (CEPA)—older adults working https://​doi.​org/​10.​1016/​S1352-​2310(01)​00361-2. group. Int J Behav Nutr Physical Act. 2017. https://​doi.​org/​10.​1186/​ 45. Sunyer J, Esnaola M, Alvarez-Pedrerol M, Forns J, Rivas I, López-Vicente s12966-​017-​0471-5. M, Suades-González E, Foraster M, Garcia-Esteban R, Basagaña X, Viana 28. Marquet O, Miralles-Guasch C. Neighbourhood vitality and physical M, Cirach M, Moreno T, Alastuey A, Sebastian-Galles N, Nieuwenhuijsen activity among the elderly: the role of walkable environments on active MJ, Querol X. Association between traffic-related air pollution in schools ageing in Barcelona, Spain. Soc Sci Med. 2015;135:24–30. https://​doi.​ and cognitive development in primary school children: a prospective org/​10.​1016/j.​socsc​imed.​2015.​04.​016. cohort study. PLoS Med. 2015;12(3):1–24. https://​doi.​org/​10.​1371/​journ​ 29. Marquet O, Hipp JA, Miralles-Guasch C. Neighborhood walkability al.​pmed.​10017​92. and active ageing: a difference in differences assessment of active Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 15 of 16 46. Sun Z, Zhu D. Exposure to outdoor air pollution and its human-related difference vegetation indices. Int J Environ Res Public Health. 2019. health outcomes: an evidence gap map. BMJ Open. 2019;9(12):1–18. https://​doi.​org/​10.​3390/​ijerp​h1605​0852. https://​doi.​org/​10.​1136/​bmjop​en-​2019-​031312. 66. Rhew IC, Vander Stoep A, Kearney A, Smith NL, Dunbar MD. Validation 47. Simoni M, Baldacci S, Maio S, Cerrai S, Sarno G, Viegi G. Adverse effects of the normalized difference vegetation index as a measure of neigh‑ of outdoor pollution in the elderly. J Thorac Dis. 2015;7(1):34–45. borhood greenness. Ann Epidemiol. 2011;21(12):946–52. https://​doi.​ https://​doi.​org/​10.​3978/j.​issn.​2072-​1439.​2014.​12.​10. org/​10.​1016/j.​annep​idem.​2011.​09.​001. 48. Chen T, Chen F, Wang K, Ma X, Wei X, Wang W, Huang P, Yang D, Xia 67. Roberts H, Helbich M. Multiple environmental exposures along daily Z, Zhao Z. Acute respiratory response to individual particle exposure mobility paths and depressive symptoms: a smartphone-based track‑ (PM1.0, PM2.5 and PM10) in the elderly with and without chronic ing study. Environ Int. 2021;156(April): 106635. https://​doi.​org/​10.​1016/j.​ respiratory diseases. Environ Pollut. 2021;271: 116329. https://​doi.​org/​ envint.​2021.​106635. 10.​1016/j.​envpol.​2020.​116329. 68. Dempsey S, Lyons S, Nolan A. Urban green space and obesity in older 49. Hu Y, Yao M, Liu Y, Zhao B. Personal exposure to ambient PM2.5, PM10, adults: Evidence from Ireland. SSM - Population Health. 2018;4(Novem‑ O3, NO2, and SO2 for different populations in 31 Chinese provinces. ber 2017):206–15. https://​doi.​org/​10.​1016/j.​ssmph.​2018.​01.​002. Environ Int. 2020;144(2): 106018. https://​doi.​org/​10.​1016/j.​envint.​2020.​ 69. Hong A, Sallis JF, King AC, Conway TL, Saelens B, Cain KL, Fox EH, 106018. Frank LD. Linking green space to neighborhood social capital in older 50. Ailshire JA, Clarke P. Fine particulate matter air pollution and cognitive adults: The role of perceived safety. Social Science and Medicine, function among US older adults. J Gerontol Ser B Psychol Sci Soc Sci. 2018;207(April):38–45. https://​doi.​org/​10.​1016/j.​socsc​imed.​2018.​04.​051. 2015;70(2):322–8. https://​doi.​org/​10.​1093/​geronb/​gbu064. 70. Petersen E, Schoen G, Liedtke G, Zech A. Relevance of urban green 51. Ailshire J, Brown LL. The importance of air quality policy for older adults space for physical activity and health-related quality of life in older and diverse communities. Public Policy Aging Rep. 2021;31(1):33–7. adults. Quality in Ageing and Older Adults, 2018;19(3):158–66. https://​ https://​doi.​org/​10.​1093/​ppar/​praa0​36. doi.​org/​10.​1108/​QAOA-​01-​2018-​0002. 52. Basner M, Babisch W, Davis A, Brink M, Clark C, Janssen S, Stansfeld 71. Akinci ZS, Delclòs-Alió X, Vich G, Miralles-Guasch C. Neighborhood S. Auditory and non-auditory effects of noise on health. Lancet. urban design and outdoor later life: an objective assessment of out- 2014;383(9925):1325–32. https://​doi.​org/​10.​1016/​S0140-​6736(13)​ of-home time and physical activity among older adults in Barcelona. J 61613-X. Aging Physical Act. 2021. https://​doi.​org/​10.​1123/​japa.​2020-​0254. 53. Stansfeld SA, Matheson MP. Noise pollution: non-auditory effects on 72. Delclòs-Alió X, Gutiérrez A, Miralles-Guasch C. The urban vitality condi‑ health. Br Med Bull. 2003;68:243–57. https://​doi.​org/​10.​1093/​bmb/​ tions of Jane Jacobs in Barcelona: residential and smartphone-based ldg033. tracking measurements of the built environment in a Mediterranean 54. Stansfeld S, Haines M, Brown B. Noise and health in the urban environ‑ metropolis. Cities. 2018. https://​doi.​org/​10.​1016/j.​cities.​2018.​09.​021. ment. Rev Environ Health. 2000;15(3):43–82. 73. Marquet O, Miralles-Guasch C. Walking short distances. The socioeco‑ 55. Héroux ME, Babisch W, Belojevic G, Brink M, Janssen S, Lercher P, Paviotti nomic drivers for the use of proximity in everyday mobility in Barcelona. M, Pershagen G, Waye KP, Preis A, Stansfeld S, van den Berg M, Verbeek Transp Res Part A: Policy Pract. 2014;70:210–22. https://​doi.​org/​10.​ J. WHO environmental noise guidelines for the European Region. Euro‑ 1016/j.​tra.​2014.​10.​007. noise. 2020;2015:2589–93. 74. Marquet O, Miralles-Guasch C. The Walkable city and the importance 56. Clark C, Paunovic K. Who environmental noise guidelines for the Euro‑ of the proximity environments for Barcelona’s everyday mobility. Cities. pean region: a systematic review on environmental noise and quality of 2015;42:258–66. https://​doi.​org/​10.​1016/j.​cities.​2014.​10.​012. life, wellbeing and mental health. Int J Environ Res Public Health. 2018. 75. Miralles-Guasch C, Martinez-Melo M, Marquet Sarda O. On user percep‑ https://​doi.​org/​10.​3390/​ijerp​h1511​2400. tion of private transport in Barcelona Metropolitan area: an experience 57. Hegewald J, Schubert M, Freiberg A, Starke KR, Augustin F, Riedel- in an academic suburban space. J Transp Geogr. 2014;36:24–31. https://​ Heller SG, Zeeb H, Seidler A. Traffic noise and mental health: a doi.​org/​10.​1016/j.​jtran​geo.​2014.​02.​009. systematic review and meta-analysis. Int J Environ Res Public Health. 76. Barceló MA, Varga D, Tobias A, Diaz J, Linares C, Saez M. Long term 2020;17(17):1–26. https://​doi.​org/​10.​3390/​ijerp​h1717​6175. effects of traffic noise on mortality in the city of Barcelona, 2004–2007. 58. Tao Y, Chai Y, Kou L, Kwan MP. Understanding noise exposure, noise Environ Res. 2016;147:193–206. https://​doi.​org/​10.​1016/j.​envres.​2016.​ annoyance, and psychological stress: incorporating individual mobility 02.​010. and the temporality of the exposure-effect relationship. Appl Geogr. 77. de Bont J, Casas M, Barrera-Gómez J, Cirach M, Rivas I, Valvi D, Álvarez M, 2020;125(September): 102283. https://​doi.​org/​10.​1016/j.​apgeog.​2020.​ Dadvand P, Sunyer J, Vrijheid M. Ambient air pollution and overweight 102283. and obesity in school-aged children in Barcelona, Spain. Environ Int. 59. van Kamp I, Davis H. Noise and health in vulnerable groups: a review. 2019;125(September 2018):58–64. https://​doi.​org/​10.​1016/j.​envint.​ Noise Health. 2013;15(64):153–9. 2019.​01.​048. 60. Lagonigro R, Martori JC, Apparicio P. Environmental noise ineq‑ 78. Marquez-Ballesteros MJ, Mora-López L, Lloret-Gallego P, Sumper A, uity in the city of Barcelona. Transp Res Part D: Transp Environ. Sidrach-de-Cardona M. Measuring urban energy sustainability and its 2018;63(June):309–19. https://​doi.​org/​10.​1016/j.​trd.​2018.​06.​007. application to two Spanish cities: Malaga and Barcelona. Sustain Cities 61. Gascon M, Cirach M, Martínez D, Dadvand P, Valentín A, Plasència A, Soc. 2019;45:335–47. https://​doi.​org/​10.​1016/j.​scs.​2018.​10.​044. Nieuwenhuijsen MJ. Normalized difference vegetation index (NDVI) as 79. Vich G, Marquet O, Miralles-Guasch C. Green exposure of walking routes a marker of surrounding greenness in epidemiological studies: the case and residential areas using smartphone tracking data and GIS in a of Barcelona city. Urban For Urban Green. 2016;19:88–94. https://​doi.​ Mediterranean city. Urban For Urban Green. 2019;40(April 2019):275–85. org/​10.​1016/j.​ufug.​2016.​07.​001. https://​doi.​org/​10.​1016/j.​ufug.​2018.​08.​008. 62. Gascon M, Triguero-Mas M, Martínez D, Dadvand P, Rojas-Rueda D, 80. Vich G, Marquet O, Miralles-Guasch C. Green streetscape and walking: Plasència A, Nieuwenhuijsen MJ. Residential green spaces and mortal‑ exploring active mobility patterns in dense and compact cities. J Transp ity: a systematic review. Environ Int. 2016;86:60–7. https://​doi.​org/​10.​ Health. 2019;12(November 2018):50–9. https://​doi.​org/​10.​1016/j.​jth.​ 1016/j.​envint.​2015.​10.​013. 2018.​11.​003. 63. Nieuwenhuijsen MJ, Khreis H, Triguero-Mas M, Gascon M, Dadvand P. 81. Chum A, O’Campo P. Contextual determinants of cardiovascular Fifty shades of green. Epidemiology. 2017;28(1):63–71. https://​doi.​org/​ diseases: overcoming the residential trap by accounting for non-resi‑ 10.​1097/​EDE.​00000​00000​000549. dential context and duration of exposure. Health Place. 2013;24:73–9. 64. Triguero-Mas M, Dadvand P, Cirach M, Martínez D, Medina A, Mompart https://​doi.​org/​10.​1016/j.​healt​hplace.​2013.​07.​007. A, Basagaña X, Gražulevičiene R, Nieuwenhuijsen MJ. Natural outdoor 82. ICGC. NDVI. Normalized Difference Vegetation Index. Institut Cartogràfic environments and mental and physical health: relationships and i Geològic de Catalunya. Retrieved January 27, 2021. https://​www.​icgc.​ mechanisms. Environ Int. 2015;77:35–41. https://​doi.​org/​10.​1016/j.​ cat/​es/​Admin​istra​cion-​yempr​esa/​Desca​rgas/​Fotog​rafias-​aereas-​y-​ortof​ envint.​2015.​01.​012. otos/​NDVI. 65. Helbich M. Spatiotemporal contextual uncertainties in green space 83. Almanza E, Jerrett M, Dunton G, Seto E, Ann Pentz M. A study of exposure measures: exploring a time series of the normalized community design, greenness, and physical activity in children using Marquet et al. International Journal of Health Geographics (2023) 22:3 Page 16 of 16 satellite, GPS and accelerometer data. Health Place. 2012;18(1):46–54. over Barcelona city (Spain). Geosci Model Dev. 2019;12(7):2811–35. https://​doi.​org/​10.​1016/j.​healt​hplace.​2011.​09.​003. https://​doi.​org/​10.​5194/​gmd-​12-​2811-​2019. 84. Dadvand P, Wright J, Martinez D, Basagaña X, McEachan RRC, Cirach 100. Casquero-Vera JA, Lyamani H, Titos G, Borrás E, Olmo FJ, Alados- M, Gidlow CJ, de Hoogh K, Gražulevičiene R, Nieuwenhuijsen MJ. Arboledas L. Impact of primary NO2 emissions at different urban Inequality, green spaces, and pregnant women: roles of ethnicity and sites exceeding the European NO2 standard limit. Sci Total Environ. individual and neighbourhood socioeconomic status. Environ Int. 2019;646:1117–25. https://​doi.​org/​10.​1016/j.​scito​tenv.​2018.​07.​360. 2014;71:101–8. https://​doi.​org/​10.​1016/j.​envint.​2014.​06.​010. 101. Lu Y. Beyond air pollution at home: assessment of personal exposure 85. James P, Hart JE, Banay RF, Laden F. Exposure to greenness and mortal‑ to PM2.5 using activity-based travel demand model and low-cost air ity in a nationwide prospective cohort study of women. Environ Health sensor network data. Environ Res. 2021;201(June): 111549. https://​doi.​ Perspect. 2015;1344(9):1344–52. https://​doi.​org/​10.​1289/​ehp.​15103​63. org/​10.​1016/j.​envres.​2021.​111549. 86. McMorris O, Villeneuve PJ, Su J, Jerrett M. Urban greenness and physical activity in a national survey of Canadians. Environ Res. 2015;137:94–100. https://​doi.​org/​10.​1016/j.​envres.​2014.​11.​010. Publisher’s Note 87. Adams MA, Frank LD, Schipperijn J, Smith G, Chapman J, Christiansen Springer Nature remains neutral with regard to jurisdictional claims in pub‑ LB, Coffee N, Salvo D, du Toit L, Dygrỳn J, Hino AAF, Lai PC, Mavoa S, lished maps and institutional affiliations. Pinzón JD, Van de Weghe N, Cerin E, Davey R, Macfarlane D, Owen N, Sallis JF. International variation in neighborhood walkability, transit, and recreation environments using geographic information systems: the IPEN adult study. Int J Health Geogr. 2014;13(1):1–18. https://​doi.​org/​10.​ 1186/​1476-​072X-​13-​43. 88. Carlson JA, Sallis JF, Conway TL, Saelens BE, Frank LD, Kerr J, Cain KL, King AC. Interactions between psychosocial and built environ‑ ment factors in explaining older adults’ physical activity. Prev Med. 2012;54(1):68–73. https://​doi.​org/​10.​1016/j.​ypmed.​2011.​10.​004. 89. Ma J, Tao Y, Kwan MP, Chai Y. Assessing mobility-based real-time air pollution exposure in space and time using smart sensors and GPS tra‑ jectories in Beijing. Ann Am Assoc Geogr. 2020;110(2):434–48. https://​ doi.​org/​10.​1080/​24694​452.​2019.​16537​52. 90. Ma X, Li X, Kwan MP, Chai Y. Who could not avoid exposure to high levels of residence-based pollution by daily mobility? Evidence of air pollution exposure from the perspective of the neighborhood effect averaging problem (neap). Int J Environ Res Public Health. 2020;17(4):1– 19. https://​doi.​org/​10.​3390/​ijerp​h1704​1223. 91. Kou L, Kwan MP, Chai Y. Living with urban sounds: understanding the effects of human mobilities on individual sound exposure and psycho‑ logical health. Geoforum. 2021;126(April 2020):13–25. https://​doi.​org/​ 10.​1016/j.​geofo​rum.​2021.​07.​011. 92. Wang R, Feng Z, Pearce J, Zhou S, Zhang L, Liu Y. Dynamic greenspace exposure and residents’ mental health in Guangzhou, China: from over- head to eye-level perspective, from quantity to quality. Landsc Urban Plann. 2021;215(January): 104230. https://​doi.​org/​10.​1016/j.​landu​ rbplan.​2021.​104230. 93. Kim J, Kwan MP. How neighborhood effect averaging might affect assessment of individual exposures to air pollution: a study of ozone exposures in Los Angeles. Ann Am Assoc Geogr. 2021;111(1):121–40. https://​doi.​org/​10.​1080/​24694​452.​2020.​17562​08. 94. Li M, Gao S, Lu F, Tong H, Zhang H. Dynamic estimation of individual exposure levels to air pollution using trajectories reconstructed from mobile phone data. Int J Environ Res Public Health. 2019. https://​doi.​ org/​10.​3390/​ijerp​h1622​4522. 95. Yoo EH, Pu Q, Eum Y, Jiang X. The impact of individual mobility on long-term exposure to ambient pm2.5: assessing effect modification by travel patterns and spatial variability of pm2.5. Int J Environ Res Public Health. 2021;18(4):1–16. https://​doi.​org/​10.​3390/​ijerp​h1804​2194. 96. Akinci ZS, Marquet O, Delclòs-Alió X, Miralles-Guasch C. Urban vitality and seniors’ outdoor rest time in Barcelona. J Transp Geogr. 2022;98(December 2021): 103241. https://​doi.​org/​10.​1016/j.​jtran​geo.​ Ready to submit your research ? Choose BMC and benefit from: 2021.​103241. 97. Matthews SA, Yang T-C. Spatial polygamy and contextual exposures • fast, convenient online submission (SPACEs): promoting activity space approaches in research on place and health. Am Behav Sci. 2013;57(8):1057–81. https://​doi.​org/​10.​1177/​ • thorough peer review by experienced researchers in your field 00027​64213​487345. • rapid publication on acceptance 98. Chaix B, Méline J, Duncan S, Merrien C, Karusisi N, Perchoux C, Lewin A, • support for research data, including large and complex data types Labadi K, Kestens Y. GPS tracking in neighborhood and health studies: a step forward for environmental exposure assessment, a step backward • gold Open Access which fosters wider collaboration and increased citations for causal inference? Health Place. 2013;21:46–51. https://​doi.​org/​10.​ • maximum visibility for your research: over 100M website views per year 1016/j.​healt​hplace.​2013.​01.​003. 99. Benavides J, Snyder M, Guevara M, Soret A, Pérez García-Pando C, At BMC, research is always in progress. Amato F, Querol X, Jorba O. CALIOPE-Urban v1.0: coupling R-LINE with a mesoscale air quality modelling system for urban air quality forecasts Learn more biomedcentral.com/submissions