Identifying Co-occurring Neighbourhood Environmental Patterns and Their Association with Health Behaviours in a Dutch Urban Population at High Cardiometabolic Risk

Identifying Co-occurring Neighbourhood Environmental Patterns and Their Association with Health Behaviours in a Dutch Urban Population at High Cardiometabolic Risk | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Identifying Co-occurring Neighbourhood Environmental Patterns and Their Association with Health Behaviours in a Dutch Urban Population at High Cardiometabolic Risk Anne K Smit, Rimke C Vos, Laura A van der Velde, Marcel R Haas, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9051455/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 11 You are reading this latest preprint version Abstract Background Most studies examine environmental determinants of health in isolation, neglecting the complex interplay between neighbourhood characteristics. Furthermore, research has primarily focused on general populations, leaving a gap in understanding how environmental factors relate to health behaviours in high-risk groups. This study aimed to (1) identify patterns of co-occurring neighbourhood characteristics related to the food environment, green space and walkability in a highly urbanised Dutch city; and (2) explore whether these neighbourhood patterns are associated with weight status, diet quality and physical activity among patients at high cardiovascular risk. Methods This cross-sectional study used baseline data from the Healthy Heart study in The Hague, Netherlands (N = 475 participants from 73 postal codes). Twenty-three indicators across food environment, green space and walkability domains were assessed at two spatial scales. Correlation analyses examined intercorrelations among indicators and with contextual characteristics. Neighbourhood patterns were identified using principal component analysis (PCA), and associations with health outcomes were examined using multivariable regression, adjusted for individual and neighbourhood confounders. Results Several environmental indicators were strongly intercorrelated and closely linked to contextual neighbourhood characteristics, particularly population density and absolute food retailer densities, even within the most highly urbanised areas. PCA identified four neighbourhood patterns. Only the “Relative food environment advantage” pattern was robustly associated with lower odds of being overweight, but also counterintuitively with lower fruit and vegetable diet quality. There was some indication of effect modification by neighbourhood socioeconomic position for this pattern. Conclusions Neighbourhood characteristics cluster into structured patterns in urban settings. Food environment measures remained coupled to population density even within high-density areas, fundamentally challenging independent effect estimation. Modest, paradoxical associations suggest limited environmental influence on lifestyle behaviours and weight in medically supervised high-risk populations. Built environment food environment green space walkability cardiovascular disease Figures Figure 1 Figure 2 Background Cardiovascular disease (CVD) remains a leading cause of morbidity and mortality globally, with a substantial proportion of the population at high risk for developing CVD. [ 1 ] Specific environmental features have been linked to cardiovascular outcomes. For instance, the density of fast-food outlets has been linked to poor diet quality, higher BMI, and less physical activity [ 2 – 4 ], all of which contribute to CVD risk. Similarly, exposure to green spaces has been associated with better cardiovascular outcomes, including reduced incidence of CVD events, lower BMI, and higher levels of physical activity. [ 2 , 4 ] However, as pointed out in several reviews, these associations often remain weak and are complicated by methodological issues. [ 2 , 5 – 7 ] One of the primary challenges is that most studies examine the impact of single environmental exposures in isolation, such as the density of fast-food outlets or proximity to green spaces, neglecting the complex interplay between neighbourhood characteristics. For example, obesogenic environments are often clustered in areas with lower socioeconomic position (SEP), which may intensify their negative impact on health outcomes. [ 8 – 10 ] Another challenge is the lack of consensus on how to define and measure these environmental exposures. For the food environment, studies have used density measures (e.g., number of outlets in a buffer), accessibility measures (e.g., network distance to nearest retailer), and relative measures such as the Modified Retail Food Environment Index (MRFEI), which compares the healthy to unhealthy availability. [ 3 , 7 ] Similarly, for green space exposure, measures include density (e.g., percentage of green space in buffer), accessibility (e.g., distance to nearest park), and quality indicators such as the Normalised Difference in Vegetation Index (NDVI). [ 11 , 12 ] Studies comparing different operationalisations often find divergent results, suggesting these measures may capture distinct environmental dimensions rather than interchangeable alternatives. [ 13 – 15 ] This inconsistency complicates comparisons across studies and limits the ability to draw definitive conclusions about the impact of the environment on (cardiovascular) health. Beyond these measurement challenges, most existing research has focused on the general population. This leaves a significant gap in our understanding of the interplay between environmental factors and lifestyle behaviours in high-risk groups. From a public health perspective, population-wide interventions are logical, as they yield larger absolute health gains. However, a substantial and growing portion of the population is already at high risk for developing CVD. For these individuals, understanding which environmental factors are most salient could inform targeted interventions. Studies in high-risk populations are therefore crucial for determining which factors warrant prioritisation, particularly those linked to adverse health outcomes in both the general and high-risk population. [ 16 , 17 ] Given these complexities, a data-driven approach is needed to understand how different aspects of the neighbourhood environment cluster together and whether these patterns, rather than individual indicators, are meaningfully associated with health behaviours. This study has two aims: (1) to identify patterns of co-occurring neighbourhood characteristics related to food environment, green spaces, and walkability in a highly urbanised area; and (2) to explore whether these neighbourhood patterns are associated with weight status, diet quality, and physical activity in primary care patients at high risk for CVD in an urban setting. By including multiple operationalisations of the same environmental constructs, we additionally examine to what extent different measurement approaches capture similar or distinct dimensions of the neighbourhood environment. Methods Study design and setting This study was conducted as a secondary analysis using baseline data from the Healthy Heart study, a non-randomized cluster stepped-wedge trial investigating the effect of a lifestyle program on patients in the primary cardiovascular prevention care protocol in general practice in the Hague. [ 18 ] The Hague is the third largest city of the Netherlands, with approximately 540.000 residents and characterized by dense urban development. The Healthy Heart study was approved by the Ethics Committee of the Leiden University Medical Centre (P17.079). All participants provided written informed consent. Study area and spatial unit definition Environmental characterization was conducted for 73 four-digit postal codes where participants resided: 48 in The Hague (65.8%) and 25 (34.2%) in directly adjacent municipalities, reflecting typical catchment areas of general practices in the Hague. Because postal codes vary in size, environmental indicators were first calculated at the address level using residential address locations from the national Registration of Addresses and Buildings (BAG) [ 19 ], and then averaged at the postal code level. We included indicators at two spatial scales: within 300 meters (immediate neighbourhood) and within 300–1000 meters (extended neighbourhood, calculated as a donut-shaped buffer). Environmental exposure measures Geospatial analyses were performed using Qgis version 3.18.3 to calculate the 23 indicators presented in Table 1 . For an extensive description of the geospatial analyses, see Additional File 1. Table 1 Included environmental exposure variables. Category Variable Calculation Variable Name Food environment Density of core food retailers in buffers Number of core retailers in buffer Core_300m, Core_donut Density of non-core food retailers in buffers Number of non-core retailers in buffer Noncore_300m, Noncore_donut Modified retail food environment index (MRFEI) in buffers* [ 22 ] (Number of core retailers + 1) / ((Number of core retailers + 1) + (Number of non-core retailers + 1)) × 100 MRFEI_300m, MRFEI_donut Food environment health index (FEHI) in buffers [ 23 ] Sum of FEHI-scores of food retailers in buffer FEHI_300m, FEHI_donut Spatial access to core-retailers in buffers Accessibility score to core retailers Spac_core_300m, Spac_core_donut Spatial access to non-core retailers in buffers Accessibility score to non-core retailers Spac_noncore_300m, Spac_noncore_donut Ratio of spatial access scores in buffers* [ 24 , 25 ] (Spatial access to core retailers + 1) / ((Spatial access to core + 1) + (Spatial access to non-core retailers + 1)) Rel_spac_300m, Rel_spac_donut Green spaces Density of accessible green spaces in buffers Percentage accessible green space in buffer Perc_access_green_300m, Perc_access_green_donut Density of other green spaces in buffers Percentage other green space in buffer Perc_other_green_300m, Perc_other_green_donut Network distance to nearest entrance of accessible green space Shortest walking distance (in meters) to nearest entrance of accessible green space Nearest_park_network Mean Normalized Difference Vegetation Index (NDVI) in buffers Average NDVI value in buffer area NDVI_300m, NDVI_donut Walkability Walkability score in 4 digit postal code [ 26 ] Composite walkability index score Walkability_250m, Walkability_1000m *To avoid division by zero in areas with no food retailers, a score of one was added to the number of core and non-core retailers. Food-environment Food-environment data, published in 2019, was obtained from Locatus, a company that collects information on business locations through regular field audits. [ 20 ] Following previous example we calculated: density of core and non-core food retailers [ 21 ], total Food Environment Healthiness Index (FEHI) score [ 21 ], spatial accessibility to core and non-core retailers, the Modified Retail Food Environment Index (MRFEI) and relative spatial accessibility. [ 22 , 23 ] Green spaces Four indicators were assessed using Dutch Land Use data (2017) [ 24 ] and Landsat 8 satellite imagery [ 25 ]: percentage of accessible green space (e.g. parks, forests) in buffer, percentage of other green space (e.g. sports fields, agricultural land), network distance to nearest accessible green space entrance, and mean Normalised Difference Vegetation Index (NDVI) within buffer. Walkability As an additional environmental factor, the Dutch walkability index developed by Lam et. al. [ 26 ] obtained through the Geoscience and Health Cohort Consortium (GECCO) was used. [ 27 , 28 ] This composite score, which can be used to assess the degree to which an area is conducive to walking as a mode of transportation, includes population density, density of retail and service destinations, land-use mix, street connectivity, green space and sidewalk presence. The walkability index was calculated by averaging the z-scores of these components, and then scaling the result to range from 0 to 100. The index was previously positively associated with walking behaviour in Dutch adults. [ 26 ] Data, for the year 2017, was pre-processed on address-level by GECCO and aggregated on average score in a 250m and 1000m buffer. All 23 indicators were included in the PCA, including multiple operationalisations of the same constructs (e.g., density, accessibility, and relative measures of the food environment) and two spatial scales (300 m and 300–1000 m buffer). The PCA thus serves as an empirical test of whether these operationalisations capture the same or different environmental dimensions: indicators measuring the same construct will load on the same component, while those capturing distinct aspects will separate. For further details on the included indicators, see Additional File 1. Correlation analysis To explore relationships between environmental indicators and contextual neighbourhood characteristics, a correlation analysis was conducted. Pairwise Spearman correlations were calculated between all environmental indicators and several contextual neighbourhood characteristics: urbanicity class (five levels, defined by Statistics Netherlands (CBS) [ 29 ]: very strongly urban (> 2500 addresses per km²), strongly urban (between 1500–2500 addresses per km²), moderately urban (between 1000 and 1500 addresses per km²), slightly urban (between 500–1000 addresses per km²), non-urban/rural (< 500 addresses per km²)), population density (continuous, using addresses per km² as a proxy; CBS 2017 [ 29 ]), and neighbourhood socioeconomic position (SES-WOA). The SES-WOA score is a comprehensive measure of socioeconomic position at the local level, based on a combination of data regarding financial prosperity, educational level and recent employment history of the households. [ 30 ] Principal component analysis Prior to PCA, variables were Box-Cox transformed to approximate normality and z-standardised (see Additional File 1 for details). To address multicollinearity issues and identify clusters of co-occurring neighbourhood features, a PCA with varimax rotation at the postal code level (n = 73) was performed, using the 23 standardized environmental indicators. To check sampling adequacy, Kaiser-Meyer-Olkin measure and Bartlett’s test of sphericity were used. The number of extracted components was based on eigenvalues greater than 1 and the scree plot. Varimax rotation was applied and rotated component scores were assigned to participants. For clarity, throughout the remainder of this article, the varimax rotated principal component scores will be referred to as the principal components (PC). Interpretation of PCA components To explore how the environmental indicators relate to broader contextual neighbourhood factors, Spearman correlations were calculated between each continuous PC score (PC1–PC4) and two contextual neighbourhood variables that were not included in the PCA: Population density, and neighbourhood socioeconomic position (SES-WOA). Exploratory health associations Having characterised environmental patterns in the neighbourhoods, we conducted exploratory analyses to examine whether these patterns associated with health behaviours in a sample of residents at high cardiovascular risk. Study population Patients at high risk for cardiovascular disease from 56 general practices in the Hague, Netherlands were invited to participate in the Healthy Heart study during a primary care consultation with their practice nurse or general practitioner (GP). For the current analysis, we included only baseline data of patients included during the control period (before intervention implementation) who had a valid postal code, to avoid potential bias due to intervention participation. During the control period patients received care as usual. Identification of high-risk patients is carried out as part of usual care in GP practices, without any involvement from the researchers. High-risk patients were defined as having a 10-year cardiovascular risk of ≥ 10% according to Dutch guidelines at the moment of inclusion (2017–2019). [ 18 , 31 ] As part of standard care, these patients have one to four individual consultations annually with a trained practice nurse. During these sessions, they receive lifestyle counselling, undergo lab work, and have physical examinations, all in line with Dutch Primary Care standards. Patients with an ICPC coded diagnosis of diabetes mellitus or cardiovascular disease were excluded from participation in the general CVD prevention program, since they were already enrolled in separate care programs. Data on sociodemographics, lifestyle behaviour and BMI were collected using self-reported online questionnaires. Health outcome variables Three outcomes were assessed: (1) diet quality, measured using the Dutch Healthy Diet index (DHD-index, range 0–160) [ 32 , 33 ] based on the DHD-FFQ; (2) physical activity, measured using the SQUASH questionnaire [ 34 ] as total minutes per week; and (3) weight status, based on self-reported BMI categorised as healthy weight, overweight, or obese using age-specific cut-offs from the Dutch Nutrition Centre [ 35 ] (see Additional File 1 for more details). Outliers >3SD were removed. Missing DHD scores (6.7%) were addressed through zero imputation of subcomponents when no more than two of 16 were missing. [ 36 ] No imputation was performed for BMI (< 1% missing) or physical activity (< 2.5% missing). Potential confounders The following potential individual- and neighbourhood-level confounders were used in the regression analyses to adjust for associations between principal components and health outcomes. Individual confounders included: age (years), sex (male/female), level of education (lower, middle, higher) and migration background (yes/no). Neighbourhood level confounders included: population density (using continuous addresses per km² as a proxy) and SEP of the postal code (SES-WOA). Statistical analysis of health associations Sample characteristics were provided using descriptive statistics, summarized as percentage for categorical variables and mean (standard deviation) or median (IQR) for continuous variables, depending on distribution of the variable. Intraclass correlation coefficients for all outcomes were < 0.05 (weight status = 0.02; DHD ≈ 0.00; physical activity = 0.01), indicating minimal clustering within postal codes; standard regression was therefore used without multilevel modelling. For interpretability, the PCs were divided into quartiles prior to performing the regression analyses, as a one-unit change in PC score is not inherently meaningful. This decision was further supported by the absence of clear linear relationships between the PCs and outcomes. The association between PC quartile and the DHD-score and total minutes of physical activity per week (square root transformed due to heteroscedasticity of the untransformed variable) were analysed using univariate and multivariate linear regression. The association between the PCs and weight status was analysed using a multinomial logistic regression (as the proportional odds assumption for an ordinal logistic regression was violated), with healthy weight set as the reference category. Assumptions were checked before conducting the analyses. Although population density correlated strongly with PC1 (continuous: r = 0.86, VIF = 8.6), quartile coding of PC1 reduced multicollinearity (VIF = 4.5). In the univariate models, the PC quartiles were included individually (a separate model for each PC, without any further adjustments). The first multivariate model (model 1) also incorporated individual and neighbourhood level confounders. In the final model (model 2), all PCs were included simultaneously and adjusted for the individual and neighbourhood level confounders. Due to sample size considerations, the PC of interest was included in the model as a categorical (dummy or factor) variable, with the lowest quartile used as the reference category. This approach was consistent with how the PC was handled in the univariate model and model 1. In contrast, the other PCs that were used for adjustment (in model 2) were added as ordinal quartile scores (1–4). To establish dose-response relationships, linear trends (P-trend) across the quartiles were examined using the median values of each quartile as continuous variable in the analyses. Additional Analyses Additional analyses of dietary and physical activity subcomponents, as well as effect modification by neighbourhood socioeconomic position, are described in Additional File 2. Statistical analyses were performed in Python version 3.8 and SPSS version 29. Results Sample Characteristics Characteristics of the 475 participants are presented in Table 2. Participants were on average 65.6 years of age and 55.4% were female. A minority had a migration background (14.7%), most were middle to higher educated (68.7%), most participants below the age of 65 were employed (72.6%) and almost all participants lived in a (very) strong urban area (94.5%). Furthermore, 53.9% were either overweight or obese, and only 37.9% met the physical activity guideline of 150 minutes of moderate to heavy physical activity per week. The median address density was 2836 addresses per km² (range: 362–8499), corresponding to urbanicity class very strongly urban. The median neighbourhood SES-WOA score was 0.09 (range: –0.98 to 0.58). Table 2. Characteristics of participants (n=475) Characteristics Age (yr), mean (SD) 65.6 (9.3) Sex, female, n (%) 263 (55.4%) Migration background, n (%) 70 (14.7%) Educational level, n (%) Low Middle High 131 (27.6%) 146 (30.7%) 184 (38.7%) Employed (<65yr), n (%) 151 (72.6%) Weight (kg), mean (SD) Weight status, n (%) - Healthy weight - Overweight - Obese 81.0 (16.1) 213 (44.8%) 156 (32.8%) 100 (21.1%) BMI (kg/m2), mean (SD) 27.1 (4.1) Total DHD-score, mean (SD) Total minutes of physical activity per week, mean (SD) Meets physical activity guideline, n (%) Address density, median (IQR) Neighbourhood SES, median (IQR) 98.4 (17.1) 2045.5 (1200.1) 180 (37.9%) 2836 (1865, 4130) 0.09 (-0.13, 0.20) * Measure of socioeconomic status at the local level, based on a combination of data regarding financial prosperity, educational level and recent employment history of the households. [30] Correlation analyses Strong intercorrelations were observed among variables within the same exposure domains (Additional File 2, Figure A1). Densities of core and non-core food retailers were very highly correlated (ρ ≈ 0.8), indicating that neighbourhoods with many non-core outlets tend to also have more healthy food retailers. Walkability was also strongly related to food-retailer density, particularly within the 1000m buffer for walkability (ρ ≈ 0.8). Relative food environment measures were strongly intercorrelated but showed lower correlations with absolute food density (not significantly correlated with core retailer density, significant negative correlation (ρ ≈ -0.4) with non-core retailer density and significant positive correlation with inverse FEHI (ρ ≈ 0.3-0.4)). In contrast, green-space density and quality measures (but not network distance to nearest accessible green space) showed consistent negative correlations with food-environment and walkability indicators, reflecting that greener areas generally coincide with fewer retail destinations. Network distance to nearest accessible green space showed lower and non-significant correlations with other green space measures. When examining contextual characteristics, population density showed much stronger correlations with environmental indicators than the ordinal classification. Population density correlated strongly with core retailer density (ρ ≈ 0.9), non-core density (ρ ≈ 0.8-0.9), and walkability (ρ ≈ 0.7-0.9), and inversely with NDVI and green-space indicators (between ρ = -0.3 and -0.8). Even within the “very strongly urban” CBS category (> 2,500 addresses /km²), substantial correlation remained: higher population density continued to be associated with higher retail density (between ρ = 0.7 and 0.9) (Figure 1). Population density was not significantly correlated with the relative food environment measures. Neighbourhood socioeconomic position showed weaker relationships with the environmental indicators: lower-SES areas tended to have higher retail density (between ρ = -0.4 and -0.6) with lower FEHI-score and higher walkability (ρ = -0.6, ρ = -0.4) but less green space (although associations were low and only the association with accessible green space in donut buffer was significant (ρ = 0.2)). Comparing the two spatial scales revealed that environmental characteristics in the immediate (300 m) and extended (300–1000 m, donut) buffers were generally highly correlated, indicating that neighbourhood features largely co-occur across scales in this urban area. The relative food environment measures showed lower correlations across scales (ρ ≈ 0.6). Principal Component Analysis The Kaiser-Meyer-Olkin (KMO) measure confirmed sampling adequacy (KMO = 0.842), and Bartlett’s test of sphericity was significant (p 1 and were retained based on the scree plot, together explaining approximately 87% of the total variance (Table 3). Component structure and interpretation The rotated component loadings (Figure 2) revealed four distinct environmental patterns. For interpretability, component names describe the environmental characteristics of areas scoring in the highest quartile. PC1 – “High food, low FEHI, high walkability” This component had high positive loadings for both core and non-core food retailer densities, walkability, and inversed FEHI and average negative loadings for NDVI. It reflects highly urbanised environments characterized by dense food retailer presence (both healthy and unhealthy) and high walkability, but limited green space. PC2 – “Relative food environment advantage” This component loaded strongly on the relative measures (MRFEI, relative spatial access). It represents neighbourhoods where healthier food options dominate relative to unhealthy ones, independent of overall retail density. PC3 – “Low density and high distance to accessible green” PC3 was defined by positive loadings for network distance to nearest accessible greenspace and negative loadings for accessible green space density, indicating low-density areas with poor accessibility to nearby green accessible space. PC4 – “Low-quality, agricultural/inaccessible green” This component had high negative loadings for inaccessible (‘other’, mainly agricultural) green and NDVI, reflecting areas characterized by low vegetation quality and non-accessible agricultural green space. Component correlations with contextual variables. To explore how these principal components related to broader neighbourhood context, Spearman correlations were computed between component scores and population density and neighbourhood socioeconomic position (See Additional File 2, Figure A2). Population density showed a very strong positive correlation with PC1 (ρ = 0.9), while neighbourhood socioeconomic position was inversely correlated (ρ = –0.6), indicating that the “High food, low FEHI, high walkability” pattern coincides with densely populated, lower-SEP areas. Other components were only weakly associated with neighbourhood socioeconomic position or population density (|ρ| < 0.4). These findings suggest that while the first component largely reflects a population density gradient, the remaining components capture neighbourhood environments that are largely independent of population density. Table 3. Explained variance of varimax rotated principal components PC1 PC2 PC3 PC4 Variance 9.621 4.148 2.321 3.838 Proportional variance 0.418 0.180 0.101 0.167 Cumulative variance 0.418 0.599 0.700 0.866 Association with Weight Status In both the univariate and the multivariate multinomial logistic regression analyses (Table 4), patients in higher quartiles of the “ Relative food environment advantage ” component showed statistically significant lower odds of being overweight (for Q3 (but not Q4) in multivariate model 2: OR=0.45, 95% CI [0.22, 0.95], P-trend=0.19). Patients in the third quartile of “ Relative food environment advantage ” also showed a lower risk of being obese in the univariate model and model 1, but this was no longer the case in model 2. These results suggest that neighbourhoods with relatively higher proportions of healthy food retailers were associated with a lower likelihood of being overweight, although no significant p for trend was found. Univariate results also showed statistically significant higher odds of obesity for patients in higher quartiles of the “ High food, low FEHI, high walkability ” component (Q4: OR = 1.98, 95% CI [1.04, 3.78]), however, this association was no longer significant in the adjusted models. Associations with Diet Quality In the linear regression analysis (Table 5), a statistically significant association was observed for the “Low quality, agricultural/inaccessible green” component. In the fully adjusted model (model 2), participants in the second quartile (Q2) had lower DHD scores compared with those in the lowest quartile (β = −6.00, 95% CI [−11.58, −0.41]). This association was not observed for higher quartiles, and no significant trend across quartiles was found (P-trend = 0.37). For the other components no statistically significant associations with the DHD index were identified in either univariate or multivariate analyses. Associations with Physical Activity The linear regression analyses (Table 6) indicated a significant negative association between the highest quartile of the “Relative food environment advantage” component and total minutes of physical activity per week in the univariate analysis (β = −3.85, 95% CI [−7.44, −0.26]). This suggests that patients living in neighbourhoods with a relatively higher proportion of healthy food retailers reported fewer minutes of physical activity per week. However, this association was no longer statistically significant after adjustment for individual and neighbourhood-level confounders in models 1 and 2. For the other components, no statistically significant associations were observed in either univariate or multivariate models. Table 4. Result of univariate and multivariate multinomial logistic regression analysis between varimax rotated principal components and weight status with healthy weight as the reference category Weight status Rotated Components Quartiles Univariate analyses Multivariate model 1 Multivariate model 2 Overweight OR 95% CI OR 95% CI OR 95% CI P for trend High food, low FEHI, high walkability Q2 Q3 Q4 1.12 1.15 0.72 0.64, 1.97 0.66, 2.02 0.39, 1.33 1.25 1.15 1.04 0.66, 2.40 0.58, 2.29 0.41, 2.61 1.35 1.21 1.02 0.67, 2.72 0.51, 2.85 0.33, 3.16 0.86 Relative food environment advantage Q2 Q3 Q4 0.86 0.50* 0.72 0.48, 1.53 0.27, 0.93 0.39, 1.30 0.79 0.48* 0.68 0.41, 1.52 0.24, 0.94 0.34, 1.37 0.82 0.45* 0.66 0.42, 1.59 0.22, 0.94 0.32, 1.36 0.19 Low density and high distance to accessible green Q2 Q3 Q4 1.08 1.10 1.00 0.60, 1.91 0.62, 1.96 0.55, 1.80 0.93 1.51 0.83 0.49, 1.79 0.80, 2.84 0.43, 1.62 1.04 1.59 0.90 0.51, 2.13 0.83, 3.05 0.42, 1.90 0.93 Low quality, low inaccessible, agricultural green Q2 Q3 Q4 0.86 0.76 0.74 0.49, 1.52 0.43, 1.35 0.41, 1.31 0.90 0.97 1.20 0.46, 1.17 0.49, 1.93 0.56, 2.58 1.09 1.15 1.35 0.48, 2.50 0.47, 2.81 0.50, 3.65 0.55 Obese High food, low FEHI, high walkability Q2 Q3 Q4 1.07 1.12 1.98* 0.53, 2.18 0.56, 2.27 1.04, 3.78 1.31 1.10 1.61 0.60, 2.82 0.49, 2.48 0.60, 4.29 1.44 1.31 1.83 0.627, 3.285 0.488, 3.497 0.555, 5.997 0.36 Relative food environment advantage Q2 Q3 Q4 0.60 0.49* 0.53 0.31, 1.16 0.25, 0.95 0.27, 1.05 0.46* 0.43* 0.46 0.22, 0.95 0.21, 0.90 0.21, 1.01 0.50 0.48 0.49 0.23, 1.07 0.21, 1.08 0.21, 1.11 0.13 Low density and high distance to accessible green Q2 Q3 Q4 1.08 1.15 1.05 0.55, 2.10 0.60, 2.25 0.53, 2.07 1.03 1.40 0.72 0.49, 2.13 0.68, 2.87 0.34, 1.54 1.34 1.54 1.00 0.61, 2.95 0.74, 3.21 0.43, 2.33 0.88 Low-quality, agricultural/inaccessible green Q2 Q3 Q4 1.21 1.24 1.48 0.60, 2.43 0.62, 2.47 0.76, 2.91 0.92 1.14 1.51 0.42, 2.03 0.51, 2.56 0.65, 3.55 1.379 1.772 2.144 0.524, 3.628 0.638, 4.921 0.715, 6.427 0.16 *p<0.05. Multivariate model 1 included the individual principal components per model corrected for age, sex, educational level, migration background, SES-WOA score of 4-digit postal code and population density. Multivariate model 2 included all principal components corrected for the aforementioned confounders. The first quartile of the principal components was set as the reference category. Table 5. Result of univariate and multivariate regression analysis between varimax rotated principal components and total DHD-index score Rotated Components Univariate analyses Multivariate model 1 Multivariate model 2 β 95% CI β 95% CI β 95% CI P for trend High food, low FEHI, high walkability Q2 Q3 Q4 -0.70 -0.78 -1.64 -5.05, 3.66 -5.12, 3.55 -6.00, 2.71 -1.01 -1.09 -0.17 -5.47, 3.45 -5.79, 3.62 -6.29, 5.95 -2.10 -3.94 -3.94 -6.87, 2.68 -9.74, 1.87 -11.36, 3.48 0.22 Relative food environment advantage Q2 Q3 Q4 1.34 1.76 -0.83 -2.96, 5.64 -2.67, 6.19 -5.30, 3.64 0.55 -0.13 -3.02 -3.79, 4.90 -4.59, 4.34 -7.71, 1.65 -0.04 -0.97 -3.60 -4.52, 4.44 -5.84, 3.90 -8.47, 1.27 0.12 Low density and high distance to accessible green Q2 Q3 Q4 -0.72 0.63 -2.05 -5.04, 3.61 -3.70, 4.96 -6.48, 2.37 -1.28 0.11 -0.64 -5.62, 3.06 -4.18, 4.41 -5.08, 3.80 -1.33 -0.33 -1.78 -6.02, 3.36 -4.70, 4.03 -6.75, 3.20 0.56 Low-quality, agricultural/inaccessible green Q2 Q3 Q4 -4.25 -2.38 -4.24 -8.61, 0.10 -6.74, 1.98 -8.59, 0.11 -4.11 -2.29 -2.81 -8.66, 0.45 -7.00, 2.41 -7.94, 2.31 -6.00* -4.87 -6.23 -11.6, -0.41 -10.9, 1.12 -12.9, 0.39 0.37 *p<0.05. Multivariate model 1 included the individual principal components per model corrected for age, sex, educational level, migration background, SES-WOA score of 4-digit postal code and population density. Multivariate model 2 included all principal components corrected for the aforementioned confounders. Table 6. Result of univariate and multivariate regression analysis between varimax rotated principal components and total minutes of physical activity per week (square root transformed) Univariate analyses Multivariate model 1 Multivariate model 2 β 95% CI β 95% CI β 95% CI P for trend High food, low FEHI, high walkability Q2 Q3 Q4 -0.29 -2.55 1.25 -3.77, 3.18 -6.00, 0.90 -2.23, 4.73 -0.46 0.08 3.55 -3.90, 2.98 -3.53, 3.69 -1.18, 8.28 -0.60 0.60 4.43 -4.27, 3.08 -3.86, 5.05 -1.33, 10.19 0.23 Relative food environment advantage Q2 Q3 Q4 -1.75 -1.96 -3.85* -5.18, 1.68 -5.48, 1.56 -7.44, -0.26 -0.20 -0.04 -1.32 -3.55, 3.14 -3.48, 3.41 -4.97, 2.33 0.23 1.22 -0.61 -3.21, 3.66 -2.50, 4.95 -4.38, 3.16 0.80 Low density and high distance to accessible green Q2 Q3 Q4 -1.93 -1.08 -0.96 -5.40, 1.55 -4.54, 2.39 -4.52, 2.59 -2.23 -1.40 -2.45 -5.59, 1.13 -4.69, 1.90 -5.87, 0.97 -1.56 -1.12 -1.37 -5.18, 2.07 -4.47, 2.23 -5.19, 2.45 0.52 Low-quality, agricultural/inaccessible green Q2 Q3 Q4 0.36 0.72 3.19 -3.14, 3.85 -2.78, 4.21 -0.28, 6.66 -0.94 -2.31 1.80 -4.45, 2.57 -5.93, 1.30 -2.15, 5.74 -0.83 -1.76 2.32 -5.13, 3.48 -6.39, 2.87 -2.79, 7.42 0.26 *p<0.05. Multivariate model 1 included the individual principal components per model corrected for age, sex, educational level, migration background, SES-WOA score of 4-digit postal code and population density. Multivariate model 2 included all principal components corrected for the aforementioned confounders. Additional Outcomes and Sensitivity Analyses Analyses of additional health behaviours are presented in Additional File 2. Analysis of the fruit and vegetable subcomponent of the DHD revealed a significant association with the “ Relative food environment advantage” component. In both multivariate models, participants in the highest quartile had significantly lower fruit and vegetable diet quality scores compared with those in the lowest quartile (model 2: β=-1.70, 95% CI [-3.10, -0.31], P-trend=0.01). No significant associations were observed between fruit and vegetable intake and the other three PCs (Table S1). For the unhealthy food choices subcomponent, the “ Low density and high distance to accessible green ” component showed a significant association in univariate and model 1 (model 1: OR=0.56, 95% CI [0.33, 0.96]) analyses. However, the association was no longer statistically significant after full adjustment in model 2. No other PCs showed a significant association with unhealthy food choices (Table S2). Binary analysis of meeting the physical activity guideline (≥150 minutes of moderate to vigorous activity per week) revealed a significant univariate association for the “ Low quality, agricultural/inaccessible green ” component. Participants in Q3 had higher odds of meeting the guidelines compared with Q1 (OR=1.72, 95% CI [1.01, 2.90]). However, this association was no longer significant after adjustment for confounders in models 1 and 2. No other PCs showed significant associations with meeting the physical activity guidelines in any model (Table S3). Effect modification for SES-WOA score and the PCs was tested by including an interaction term in model 2 of the multinomial logistic regression model for weight status and the linear models for total DHD-score and total minutes of weekly physical activity (Table S4-S6). In none of the models was the interaction term statistically significant, except for the interaction between the “ Relative food environment advantage ” component and neighbourhood socioeconomic position for obesity (P<0.02), though not for overweight. However, the neighbourhood socioeconomic position main effect showed an extremely low OR likely reflecting model instability due to sparse data (only 11 obese participants in highest tertile of SES-WOA). Discussion Examination of Environmental Patterns The correlation analyses and PCA revealed several important patterns in how neighbourhood environmental characteristics cluster together in urban areas, with implications for environmental health research. Food environment An important finding was the very strong positive correlation between food retailer density and population density. Crucially, this strong correlation persisted even within the "very strongly urban" category (> 2,500 addresses per km²). This implies that simply adjusting for urbanicity category (as frequently done in environmental health research) might not adequately control for population density-related confounding. Previous research in compact Dutch urban settings has shown that population density itself is associated with multiple health-relevant pathways, underscoring its role as a broader structuring factor of urban environments. [ 37 ] Our approach of adjusting for continuous population density addresses this limitation, but introduces a different concern: population density correlated very strongly with PC1 (r = 0.86), meaning that adjusting for it may have removed much of the very variation this component captures. This raises a fundamental issue: failing to adjust for population density risks confounding, yet adjusting for it may obscure real associations between the food environment and health outcomes that operate precisely through the urban density gradient. Densities of core and non-core food retailers were very highly correlated (ρ > 0.9), reflecting the commercial clustering typical of urban environments. Additionally, relative and absolute food environment measures loaded on separate principal components and showed weak correlations with each other. This independence has been noted previously by Pinho et. al. [ 22 ] and indicates these measures capture fundamentally different dimensions of the food environment. Notably, PC2 (relative measures) showed no significant correlation with population density, while PC1 (absolute measures) remained strongly correlated. This confirms that relative measures capture food environment quality independent of the urbanicity gradient. However, even though calls have been made to use relative measures more widely [ 22 , 38 , 39 ], this independence also presents interpretive challenges. Identical relative scores can represent fundamentally different food environments depending on absolute availability. For example, a neighbourhood with 2 supermarkets and 1 fast-food outlet (67% healthy) scores better than one with 20 supermarkets and 30 fast-food outlets (40% healthy), despite the latter offering far more absolute healthy options. Moreover, as previously described [ 40 ], ratio measures can be volatile in low-density areas, where a single retailer opening or closing can shift the ratio by 10 percentage points, while the same change in a dense area produces less than a 1-percentage-point shift. Future research should consider: restricting analyses to areas exceeding minimum retail density thresholds [ 40 ] and/or stratifying by absolute density (not urbanicity) to test whether associations are consistent across contexts. Green space Our analysis revealed substantial heterogeneity in green space measures. NDVI and other, mostly agricultural, green space loaded together (PC4) but showed weak correlations with accessible green spaces (PC3), reflecting that these measures capture fundamentally different constructs. NDVI primarily captures vegetation exposure correlating with larger agricultural land and might be more relevant to air quality and heat mitigation, while accessible green spaces capture opportunities for physical activity and social interaction. [ 41 ] Besides, green space showed expected inverse relationships with population density. Relationships with neighbourhood socioeconomic position were weaker and this finding aligns with previous Dutch research showing that socioeconomic differences in greenness are smaller in highly urbanised municipalities. [ 42 ] Scale and Spatial Considerations Our data showed high correlations between environmental characteristics at 300m and 300-1000m scales. Neighbourhoods with high retail or green space density immediately around residences (300m) generally also have high density in their extended surroundings (300-1000m donut buffer). However, an important exception emerged. Relative food environment measures showed lower cross-scale correlations (ρ ≈ 0.6), suggesting that the relative composition of healthy versus unhealthy food retailers can vary between immediate and extended neighbourhoods. This has implications for mechanism: if the relative food environment primarily influences health through daily shopping patterns, immediate neighbourhood exposure (300m, walkable distance) may matter more than extended area composition. Conversely, if mechanisms operate through neighbourhood norms or perceived availability as previously proposed [ 43 – 45 ], extended area characteristics might be more relevant. Association with Weight Status and Lifestyle Behaviours Despite interesting environmental variation captured by the four principal components, associations with weight status and lifestyle behaviours were limited and showed mostly inconsistent patterns. The " Relative food environment advantage " component provided one of the more robust findings. This is consistent with a systematic review by Cobb. et. al. which noted that studies utilizing relative measures are more likely to see significant and expected results with weight status than absolute measures. [ 3 ] This component showed consistent associations with lower odds of being overweight across all models for the third quartile (univariate, multivariate model 1, and fully adjusted model 2). Similar protective associations were observed for obesity in univariate and model 1 analyses, though these attenuated in the fully adjusted model. However, no clear dose-response pattern emerged (P-trend = 0.19), suggesting potential threshold effects or that these associations may be influenced by other unmeasured neighbourhood characteristics. This finding is complicated by a counterintuitive pattern: higher levels of this same component were associated with lower fruit and vegetable intake in the highest quartile (including a significant trend). A systematic review by Turner et. al., studying the association between the retail environment and fruit and vegetable consumption, found mixed results. [ 46 ] However, most of the included studies focused on absolute measures. Previous studies investigating the association between relative measures and fruit/vegetable intake in the general population found results in the opposite direction, i.e. reporting that individuals living in areas with a higher proportion of healthier food stores reported greater fruit and vegetable consumption. [ 38 , 39 , 47 ] A possible explanation could be that high access to supermarkets may reduce the need for detailed planning of grocery trips, compared to those with lower access. For example, research has previously identified planning is one of the factors that influence the translation of intentions into actual behaviour related to fruit and vegetable consumption. [ 48 ] For the other principal components, associations were weak or inconsistent. The " High food, low FEHI, high walkability " component showed a univariate association with obesity that disappeared after adjustment for confounders, suggesting this pattern primarily reflects broader neighbourhood disadvantage rather than specific environmental effects. No consistent associations emerged for the green space components (PC3, PC4) with any outcome, nor were meaningful associations observed for total physical activity. The few significant findings in supplementary analyses of dietary and physical activity subcomponents similarly failed to show consistent patterns or dose-response relationships and did not survive full adjustment for confounders. The inconsistent results might be due to chance, residual confounding and/or limited sample size. Moreover, the limited associations observed may also reflect characteristics specific to high-risk populations. Individuals already at high cardiovascular risk may have reduced environmental sensitivity due to established dietary patterns that are less responsive to neighbourhood influences. Additionally, high-risk patients in our study were engaged with healthcare providers who provided dietary guidance as part of standard Dutch primary care protocols, potentially creating alternative pathways to healthy behaviours that are independent of neighbourhood characteristics. Further large-scale confirmatory research is needed to explore these findings more robustly. Effect Modification We found no evidence that associations between neighbourhood characteristics and health outcomes varied by neighbourhood socioeconomic position, except for a significant interaction between the "Relative food environment advantage" component and neighbourhood socioeconomic position for obesity, suggesting that food environment characteristics may be more strongly associated with obesity in neighbourhoods with a lower socioeconomic position. This finding is consistent with recent Dutch evidence. [ 49 ] However, model instability was observed due to insufficient data within strata, necessitating replication in larger samples before drawing substantive conclusions. Strengths and limitations A major strength of this study was the comprehensive assessment of the neighbourhood environment, including 23 indicators across food environment, green space, and walkability domains. By applying PCA, we were able to reduce multicollinearity and explore how clusters of co-occurring neighbourhood characteristics relate to BMI, diet quality, and physical activity. This data-driven approach revealed distinct environmental patterns and their intercorrelations, providing insights into how neighbourhood characteristics cluster in dense urban settings. Furthermore, to our knowledge, this is the first study that investigates the association between neighbourhood characteristics and BMI and lifestyle behaviour in a primary care population at high risk for CVD. While the identified environmental patterns likely reflect general characteristics of the urban landscape applicable across populations, the associations with health behaviours were examined specifically in this high-risk group. Whether environmental patterns play a similar role in the general population remains an open question. As previously mentioned, high-risk populations may differ from the general populations in important ways which could modify how they interact with their neighbourhood environment. Nonetheless, studying this population offers preliminary insights relevant for population health management and risk stratification in primary care settings. Several limitations should be noted. First, the cross-sectional exploratory design precludes any conclusions regarding causation. Additionally, the large number of statistical tests conducted across multiple outcomes and components increases the probability of chance findings. No correction for multiple testing was applied, consistent with the exploratory aims of this study; however, this means that individual significant associations should be interpreted with caution. Second, despite adjusting for several neighbourhood confounders, residual confounding can remain. Although we adjusted for continuous population density, density was strongly tied to several environmental characteristics. This reflects that, in dense urban settings, many neighbourhood features are inherently structured by the density gradient itself. Consequently, the influence of specific environmental patterns cannot be fully separated from the broader high-density urban context. The third limitation is that due to privacy considerations, only 4-digit postal code of the participants was available, exposure results are less sensitive than when address level data would be available. The fourth limitation is that outcomes were measured by self-report questionnaires, which can lead to participants giving more socially acceptable answers that may lead to an underestimation of the observed associations. Lastly, the PCA was conducted using neighbourhood data linked to the residential postal codes of participants in our sample, rather than based on all neighbourhoods in the city The Hague. While in theory this limits the representativeness of the identified environmental patterns for the entire urban area, the included postal codes cover a very large and diverse portion of the Hague (79% of all postal codes in the Hague) and directly adjacent municipalities, capturing substantial environmental variation and sociodemographic backgrounds. Conclusions This study identified four distinct neighbourhood environmental patterns in a dense urban area in the Netherlands. The patterns reflected systematic co-occurrence of food environment, walkability, and green space characteristics, with population density emerging as a key underlying structuring feature. Correlation analyses confirmed that population density remained strongly tied to multiple environmental indicators, with particularly strong correlations for food retailer densities. This strong correlation persisted even within very strongly urban areas, demonstrating methodological challenges for isolating independent effects of food environment on health behaviours using conventional observational approaches. Of the identified patterns, only one, the “ Relative food environment advantage ” pattern, showed a more robust association with a lower odds of being overweight. However, this association was modest, showed no consistent dose–response, and was accompanied by a paradoxical association with lower fruit and vegetable dietary quality. No robust associations were observed for physical activity or other dietary outcomes. There was some indication that the relative food environment may be more strongly associated with obesity in neighbourhoods with lower socioeconomic position, suggesting possible effect modification that warrants further exploration. These findings highlight the importance of examining neighbourhood characteristics as interrelated patterns rather than isolated exposures, especially in highly urbanised contexts. Larger, confirmatory studies, using pattern-based approaches, are needed to investigate the preliminary associations with health outcomes and assess their relevance for population health management. Abbreviations BAG Basisregistratie Adressen en Gebouwen (Registration of Addresses and Buildings) BMI Body mass index CBS Centraal Bureau voor de Statistiek (Statistics Netherlands) CI Confidence interval CVD Cardiovascular disease DHD-FFQ Dutch Healthy Diet Food Frequency Questionnaire DHD-index Dutch Healthy Diet index FEHI Food Environment Healthiness Index GECCO Geoscience and Health Cohort Consortium GP General practitioner ICPC International Classification of Primary Care IQR Interquartile range KMO Kaiser-Meyer-Olkin MRFEI Modified Retail Food Environment Index NDVI Normalised Difference Vegetation Index NHG Nederlands Huisartsen Genootschap (Dutch College of General Practitioners) PC Principal component PCA Principal component analysis SD Standard deviation SEP Socioeconomic position SES-WOA Sociaaleconomische status op basis van welvaart, opleidingsniveau en arbeidspositie SPSS Statistical Package for the Social Sciences SQUASH Short Questionnaire to Assess Health-enhancing physical activity VIF Variance inflation factor. Declarations Ethics approval and informed consent This study was approved by the Ethics Committee of Leiden University Medical Centre (P17.079) and conducted in accordance with the Declaration of Helsinki and Dutch medical research regulations (WMO). All participants provided written informed consent prior to participation. Consent for publication Not applicable Availability of data and materials Data from the Healthy Heart study may be available from the corresponding author upon reasonable request, subject to ethics approval (Leiden UMC, P17.079). Environmental data were obtained from GECCO (walkability index, available upon request via https://www.gecco.nl) and Locatus (commercial data, available under licence). Green space data from the Dutch Land Use Database (CBS) and Landsat 8 satellite imagery (USGS) are publicly available. Competing Interests The authors declare no conflict of interest. Acknowledgements and funding This publication is part of the project "ECOTIP" (with project number NWA.1518.22.151) of the research programme "Dutch Research Agenda Routes by Consortia" (NWA-ORC 2022) which is financed by the Dutch Research Council (NWO). The Healthy Heart Study was sponsored by ZonMW grant number 531 001 203. The walkability index was collected as part of the Geoscience and Health Cohort Consortium (GECCO), which was financially supported by the Netherlands Organisation for Scientific Research (NWO), the Netherlands Organisation for Health Research and Development (ZonMw), and Amsterdam UMC. More information on GECCO can be found on www.gecco.nl Data protection All data were handled according to GDPR and Dutch privacy legislation. AI Use Statement During the preparation of this work the authors used Claude (Anthropic) in order to perform language editing, including spelling and grammar checks, and to assist with word count reduction. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication. Authors’ contributions A.S., R.V., L.vdV. and J.K. wrote the draft of the manuscript. T.B. and M.N. were involved in the design and execution of the study. A.S., M.H. and J.K. were involved in data curation, formal analysis and methodology. A.S., R.V., L.vdV., M.H., M.B. T.B., M.N. and J.K. critically revised and approved the final manuscript. References World Health Organization. Cardiovascular diseases (CVDs) [Internet]. [cited 2025 May 23]. Available from: https://www.who.int/en/news-room/fact-sheets/detail/cardiovascular-diseases Dixon BN, Ugwoaba UA, Brockmann AN, Ross KM. Associations between the built environment and dietary intake, physical activity, and obesity: A scoping review of reviews. Obes Rev. 2021;22:e13171. https://doi.org/10.1111/obr.13171 . Cobb LK, Appel LJ, Franco M, Jones-Smith JC, Nur A, Anderson CAM. The relationship of the local food environment with obesity: A systematic review of methods, study quality, and results. Obesity. 2015;23(7):1331–44. Malambo P, Kengne AP, De Villiers A, Lambert EV, Puoane T. Built environment, selected risk factors and major cardiovascular disease outcomes: A systematic review. PLoS ONE 2016;11(11). Nguyen PY, Astell-Burt T, Rahimi-Ardabili H, Feng X. Green space quality and health: A systematic review. Int J Environ Res Public Health. 2021;18(21):11028. Lytle LA, Sokol RL. Measures of the food environment: A systematic review of the field, 2007–2015. Health Place. 2017;44:18–34. Meijer P, Numans H, Lakerveld J. Associations between the neighbourhood food environment and cardiovascular disease: a systematic review. Eur J Prev Cardiol. 2023;00:1–11. Leal C, Bean K, Thomas F, Chaix B. Multicollinearity in Associations Between Multiple Environmental Features and Body Weight and Abdominal Fat: Using Matching Techniques to Assess Whether the Associations are Separable. Am J Epidemiol. 2012;175(11):1152–62. Feuillet T, Charreire H, Roda C, Ben Rebah M, Mackenbach JD, Compernolle S, et al. Neighbourhood typology based on virtual audit of environmental obesogenic characteristics. Obes Rev. 2016;17:19–30. Molina-García J, Queralt A, Adams MA, Conway TL, Sallis JF. Neighborhood built environment and socio economic status in relation to multiple health outcomes in adolescents. Prev Med (Baltim). 2017;105:88–94. Liu XX, Ma XL, Huang WZ, Luo YN, He CJ, Zhong XM, et al. Green space and cardiovascular disease: A systematic review with meta-analysis. Environ Pollut. 2022;301:118990. Luo YN, Huang WZ, Liu XX, Markevych I, Bloom MS, Zhao T, et al. Greenspace with overweight and obesity: A systematic review and meta-analysis of epidemiological studies up to 2020. Obes Rev. 2020;21(11):e13078. Bernsdorf KA, Bøggild H, Aadahl M, Toft U. Measuring associations between the food environment and dietary habits: comparing the proportion and density of food outlets. BMC Public Health. 2024;24(1):1–14. Maguire ER, Burgoine T, Penney TL, Forouhi NG, Monsivais P. Does exposure to the food environment differ by socioeconomic position? Comparing area-based and person-centred metrics in the Fenland Study, UK. Int J Health Geogr. 2017;16(1):1–14. Klompmaker JO, Hoek G, Bloemsma LD, Gehring U, Strak M, Wijga AH, et al. Green space definition affects associations of green space with overweight and physical activity. Environ Res. 2018;160:531–40. Mackenbach JP, Lingsma HF, Van Ravesteyn NT, Kamphuis CBM. The population and high-risk approaches to prevention: quantitative estimates of their contribution to population health in the Netherlands, 1970–2010. Eur J Public Health. 2013;23(6):909–15. Rose G. The Strategy of Preventive Medicine. Oxford University Press; 1992. Bonten TN, Verkleij SM, van der Kleij RM, Busch K, Hout WB, van den, Chavannes NH, et al. Selective prevention of cardiovascular disease using integrated lifestyle intervention in primary care: protocol of the Healthy Heart stepped-wedge trial. BMJ Open. 2021;11(7):e043829. BAG producten -. Kadaster.nl zakelijk [Internet]. [cited 2024 Jun 13]. Available from: https://www.kadaster.nl/zakelijk/registraties/basisregistraties/bag/bag-producten Home. - Locatus [Internet]. [cited 2024 Jun 13]. Available from: https://locatus.com/en/ Poelman MP, Nicolaou M, DIjkstra SC, MacKenbach JD, Lu M, Karssenberg D, et al. Does the neighbourhood food environment contribute to ethnic differences in diet quality? Results from the HELIUS study in Amsterdam, the Netherlands. Public Health Nutr. 2021;24(15):5101–12. Pinho MGM, Mackenbach JD, Oppert JM, Charreire H, Bárdos H, Rutter H, et al. Exploring absolute and relative measures of exposure to food environments in relation to dietary patterns among European adults. Public Health Nutr. 2019;22(6):1037–47. Centers for Disease Control and Prevention. (2011) Census Tract Level State Maps of the Modified Retail Food Environment Index (mRFEI). Children’s Food Environment State Indicator Report, 2011. Atlanta, GA: CDC. Bestand bodemgebruik | CBS [Internet]. [cited 2024 Nov 28]. Available from: https://www.cbs.nl/nlnl/dossier/nederland-regionaal/geografische-data/bestand-bodemgebruik EarthExplorer [Internet]. [cited 2024 Nov 28]. Available from: https://earthexplorer.usgs.gov/ Lam TM, Wang Z, Vaartjes I, Karssenberg D, Ettema D, Helbich M, et al. Development of an objectively measured walkability index for the Netherlands. Int J Behav Nutr Phys Act. 2022;19(1):1–16. Lakerveld J, Wagtendonk A, Vaartjes I, Karssenberg D, Penninx B, Beulens J et al. Deep phenotyping meets big data: the Geoscience and hEalth Cohort COnsortium (GECCO) data to enable exposome studies in The Netherlands. Int J Health Geogr. 2020;19(1). Timmermans EJ, Lakerveld J, Beulens JWJ, Boomsma DI, Kramer SE, Oosterman M et al. Cohort profile: the Geoscience and Health Cohort Consortium (GECCO) in the Netherlands. BMJ Open. 2018;8(6). Stedelijk gebied | CBS [Internet]. [cited 2025 May 23]. Available from: https://www.cbs.nl/nl-nl/onze diensten/methoden/begrippen/stedelijk-gebied Seswoa. Sociaaleconomische statusscores huishoudens | CBS [Internet]. [cited 2024 Apr 26]. Available from: https://www.cbs.nl/nl-nl/onze-diensten/maatwerk-en-microdata/microdata-zelf-onderzoek doen/microdatabestanden/seswoa-sociaaleconomische-statusscores-huishoudens NHG-Standaard Cardiovasculair risicomanagement [internet]. [cited 2025 May 23]. Available from: https://richtlijnen.nhg.org/standaarden/cardiovasculair-risicomanagement Looman M, Feskens EJM, De Rijk M, Meijboom S, Biesbroek S, Temme EHM, et al. Development and evaluation of the Dutch Healthy Diet index 2015. Public Health Nutr. 2017;20(13):2289–99. van Lee L, Feskens EJM, Meijboom S, Van Hooft EJC, Van ’T, Veer P, De Vries JHM, et al. Evaluation of a screener to assess diet quality in the Netherlands. Br J Nutr. 2016;115(3):517–26. Wendel-Vos GCW, Schuit AJ, Saris WHM, Kromhout D. Reproducibility and relative validity of the short questionnaire to assess health-enhancing physical activity. J Clin Epidemiol. 2003;56(12):1163–9. BMI berekenen |. Voedingscentrum [Internet]. [cited 2024 Jun 24]. Available from: https://www.voedingscentrum.nl/nl/bmi-meter.aspx#70jaar Ichikawa M, Hosono A, Tamai Y, Watanabe M, Shibata K, Tsujimura S, et al. Handling missing data in an FFQ: multiple imputation and nutrient intake estimates. Public Health Nutr. 2019;22(8):1351. Beenackers MA, Oude Groeniger J, Kamphuis CBM, Van Lenthe FJ. Urban population density and mortality in a compact Dutch city: 23-year follow-up of the Dutch GLOBE study. Health Place. 2018;53:79–85. Mason KE, Bentley RJ, Kavanagh AM. Fruit and vegetable purchasing and the relative density of healthy and unhealthy food stores: evidence from an Australian multilevel study. J Epidemiol Community Health. 2013;67(3):231–6. Clary CM, Ramos Y, Shareck M, Kestens Y. Should we use absolute or relative measures when assessing foodscape exposure in relation to fruit and vegetable intake? Evidence from a wide-scale Canadian study. Prev Med (Baltim). 2015;71:83–7. Thornton LE, Lamb KE, White SR. The use and misuse of ratio and proportion exposure measures in food environment research. Int J Behav Nutr Phys Act. 2020;17(1):118. Markevych I, Schoierer J, Hartig T, Chudnovsky A, Hystad P, Dzhambov AM, et al. Exploring pathways linking greenspace to health: Theoretical and methodological guidance. Environ Res. 2017;158:301–17. de Vries S, Buijs AE, Snep RPH. Environmental Justice in The Netherlands: Presence and Quality of Greenspace Differ by Socioeconomic Status of Neighbourhoods. Sustainability 2020, Vol 12, Page 5889. 2020;12(15):5889. Flint E, Cummins S, Matthews S. Do perceptions of the neighbourhood food environment predict fruit and vegetable intake in low-income neighbourhoods? Health Place. 2013;24:11–5. van Rongen S, Poelman MP, Thornton L, Abbott G, Lu M, Kamphuis CBM, et al. Neighbourhood fast food exposure and consumption: The mediating role of neighbourhood social norms. Int J Beh Nutr Phys Act. 2020;17(1):61. Tak NI, J te Velde SJ, Kamphuis CB, Ball K, Crawford D, Brug J, et al. Associations between neighbourhood and household environmental variables and fruit consumption: exploration of mediation by individual cognitions and habit strength in the GLOBE study. Public Health Nutr. 2013;16(3):505–14. Turner G, Green R, Alae-Carew C, Dangour AD. The association of dimensions of fruit and vegetable access in the retail food environment with consumption; a systematic review. Glob Food Sec. 2021;29:100528. McGuirt JT, Pitts SBJ, Gustafson A. Association between Spatial Access to Food Outlets, Frequency of Grocery Shopping, and Objectively-Assessed and Self-Reported Fruit and Vegetable Consumption. Nutrients. 2018;10(12):1974. Kreausukon P, Gellert P, Lippke S, Schwarzer R. Planning and self-efficacy can increase fruit and vegetable consumption: A randomized controlled trial. J Behav Med. 2012;35(4):443–51. van Erpecum CPL, van Zon SKR, Bültmann U, Smidt N. The association between the presence of fast food outlets and BMI: the role of neighbourhood socio-economic status, healthy food outlets, and dietary factors. BMC Public Health. 2022;22(1):1–12. Additional Declarations No competing interests reported. Supplementary Files AdditionalFile1.docx AdditionalFile2.docx Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 21 Apr, 2026 Reviews received at journal 20 Apr, 2026 Reviews received at journal 20 Apr, 2026 Reviewers agreed at journal 01 Apr, 2026 Reviewers agreed at journal 31 Mar, 2026 Reviewers agreed at journal 31 Mar, 2026 Reviewers invited by journal 28 Mar, 2026 Editor invited by journal 13 Mar, 2026 Editor assigned by journal 12 Mar, 2026 Submission checks completed at journal 12 Mar, 2026 First submitted to journal 06 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9051455","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":616198120,"identity":"277e2ae2-c782-4e7a-be8c-09e2445d7be4","order_by":0,"name":"Anne K Smit","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAsElEQVRIiWNgGAWjYFACNhAhwcDPDuHKEK1FQrIZwuUhVguDhMFhYrXwN7AlMP7MsagzPsxjuuEHwx3CWiQOsB1g5t0mIWF2mMfsZg/DMyIcdoC9gZkRquU2A8NhwlrkgVoYfwK1GDcTq8UA6DAGkMMMmInVYniYLeEwUIvkjMNsZTd7DIjQIne8zfDhz211/Pztzdtu/Kg4LEdQCwMzMASQ3ElYwygYBaNgFIwCIgAA6xkxAGrox8AAAAAASUVORK5CYII=","orcid":"","institution":"Health Campus The Hague","correspondingAuthor":true,"prefix":"","firstName":"Anne","middleName":"K","lastName":"Smit","suffix":""},{"id":616198122,"identity":"b1c39f75-2fb3-4e20-8d19-2c04fb951447","order_by":1,"name":"Rimke C Vos","email":"","orcid":"","institution":"Health Campus The Hague","correspondingAuthor":false,"prefix":"","firstName":"Rimke","middleName":"C","lastName":"Vos","suffix":""},{"id":616198125,"identity":"8993b431-a582-4277-8c80-9e7bfa58af28","order_by":2,"name":"Laura A van der Velde","email":"","orcid":"","institution":"Health Campus The Hague","correspondingAuthor":false,"prefix":"","firstName":"Laura","middleName":"A van der","lastName":"Velde","suffix":""},{"id":616198126,"identity":"552f7f77-5bb4-4cc4-9bda-b6b4c6078319","order_by":3,"name":"Marcel R Haas","email":"","orcid":"","institution":"Health Campus The Hague","correspondingAuthor":false,"prefix":"","firstName":"Marcel","middleName":"R","lastName":"Haas","suffix":""},{"id":616198127,"identity":"6e0a25a9-d5e9-4871-ab8f-75f4409daea2","order_by":4,"name":"Mattijs E Numans","email":"","orcid":"","institution":"Health Campus The Hague","correspondingAuthor":false,"prefix":"","firstName":"Mattijs","middleName":"E","lastName":"Numans","suffix":""},{"id":616198128,"identity":"23a0b137-8d5f-421b-ab2c-8aac93edad1a","order_by":5,"name":"Mariëlle A Beenackers","email":"","orcid":"","institution":"Erasmus MC","correspondingAuthor":false,"prefix":"","firstName":"Mariëlle","middleName":"A","lastName":"Beenackers","suffix":""},{"id":616198129,"identity":"b9285611-5810-46f9-b980-2994531a6c71","order_by":6,"name":"Tobias N Bonten","email":"","orcid":"","institution":"Health Campus The Hague","correspondingAuthor":false,"prefix":"","firstName":"Tobias","middleName":"N","lastName":"Bonten","suffix":""},{"id":616198130,"identity":"7f266767-874e-4847-a89a-5ea7162e7316","order_by":7,"name":"Jessica C Kiefte-de Jong","email":"","orcid":"","institution":"Health Campus The Hague","correspondingAuthor":false,"prefix":"","firstName":"Jessica","middleName":"C Kiefte-de","lastName":"Jong","suffix":""}],"badges":[],"createdAt":"2026-03-06 14:09:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9051455/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9051455/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105993218,"identity":"04e6bcb4-0da5-4d99-b30e-7a05ccf4ec8d","added_by":"auto","created_at":"2026-04-02 08:44:14","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":92993,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCorrelation table within highly urban areas\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e*For definition of the environmental indicators see Table 1. Pop_density (population density) defined as addresses per km\u003csup\u003e2\u003c/sup\u003e. SES: Measure of socioeconomic status at the local level, based on a combination of data regarding financial prosperity, educational level and recent employment history of the households. [30]\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9051455/v1/5a090a7ea2278d583af750d5.jpg"},{"id":105993220,"identity":"eb9636b1-18ea-4dff-b921-7c9bea664fed","added_by":"auto","created_at":"2026-04-02 08:44:14","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":189441,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eVariable loadings on the four varimax rotated principal components\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e* NDVI = Normalised difference in Vegetation Index; perc_other_green = percentage of other green spaces, perc_access_green = percentage of accessible green spaces; nearest_park_network = shortest distance to entrance of nearest park; core = density of core retailers; noncore = density of noncore retailers; MRFEI = Modified Retail Food-Environment Index; inversed_FEHI = the inverse of Food Environment Healthiness Index; spac_noncore = spatial accessibility to noncore retailers; spac_core = spatial accessibility to core retailers; rel_spac = ratio of spatial access scores.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9051455/v1/6797d485a2614a03caec45e0.jpg"},{"id":106095769,"identity":"f28d71d4-02c1-4b46-9178-1a9e95475dca","added_by":"auto","created_at":"2026-04-03 11:50:53","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1953955,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9051455/v1/703fb897-4218-4a23-8f9f-e0aebb93ce6b.pdf"},{"id":105993221,"identity":"9d0be9ca-a473-41e9-8464-650ba5301ec1","added_by":"auto","created_at":"2026-04-02 08:44:14","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":81459,"visible":true,"origin":"","legend":"","description":"","filename":"AdditionalFile1.docx","url":"https://assets-eu.researchsquare.com/files/rs-9051455/v1/08851b821599e85c12f297de.docx"},{"id":106094750,"identity":"8478b061-1775-48d3-a0dc-2b2699310d71","added_by":"auto","created_at":"2026-04-03 11:43:12","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":403444,"visible":true,"origin":"","legend":"","description":"","filename":"AdditionalFile2.docx","url":"https://assets-eu.researchsquare.com/files/rs-9051455/v1/32d12412997f8240f830ca09.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Identifying Co-occurring Neighbourhood Environmental Patterns and Their Association with Health Behaviours in a Dutch Urban Population at High Cardiometabolic Risk","fulltext":[{"header":"Background","content":"\u003cp\u003eCardiovascular disease (CVD) remains a leading cause of morbidity and mortality globally, with a substantial proportion of the population at high risk for developing CVD. [\u003cspan class=\"CitationRef\"\u003e1\u003c/span\u003e] Specific environmental features have been linked to cardiovascular outcomes. For instance, the density of fast-food outlets has been linked to poor diet quality, higher BMI, and less physical activity [\u003cspan class=\"CitationRef\"\u003e2\u003c/span\u003e–\u003cspan class=\"CitationRef\"\u003e4\u003c/span\u003e], all of which contribute to CVD risk. Similarly, exposure to green spaces has been associated with better cardiovascular outcomes, including reduced incidence of CVD events, lower BMI, and higher levels of physical activity. [\u003cspan class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e4\u003c/span\u003e] However, as pointed out in several reviews, these associations often remain weak and are complicated by methodological issues. [\u003cspan class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e5\u003c/span\u003e–\u003cspan class=\"CitationRef\"\u003e7\u003c/span\u003e]\u003c/p\u003e \u003cp\u003eOne of the primary challenges is that most studies examine the impact of single environmental exposures in isolation, such as the density of fast-food outlets or proximity to green spaces, neglecting the complex interplay between neighbourhood characteristics. For example, obesogenic environments are often clustered in areas with lower socioeconomic position (SEP), which may intensify their negative impact on health outcomes. [\u003cspan class=\"CitationRef\"\u003e8\u003c/span\u003e–\u003cspan class=\"CitationRef\"\u003e10\u003c/span\u003e] Another challenge is the lack of consensus on how to define and measure these environmental exposures. For the food environment, studies have used density measures (e.g., number of outlets in a buffer), accessibility measures (e.g., network distance to nearest retailer), and relative measures such as the Modified Retail Food Environment Index (MRFEI), which compares the healthy to unhealthy availability. [\u003cspan class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e7\u003c/span\u003e] Similarly, for green space exposure, measures include density (e.g., percentage of green space in buffer), accessibility (e.g., distance to nearest park), and quality indicators such as the Normalised Difference in Vegetation Index (NDVI). [\u003cspan class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e12\u003c/span\u003e] Studies comparing different operationalisations often find divergent results, suggesting these measures may capture distinct environmental dimensions rather than interchangeable alternatives. [\u003cspan class=\"CitationRef\"\u003e13\u003c/span\u003e–\u003cspan class=\"CitationRef\"\u003e15\u003c/span\u003e] This inconsistency complicates comparisons across studies and limits the ability to draw definitive conclusions about the impact of the environment on (cardiovascular) health.\u003c/p\u003e \u003cp\u003eBeyond these measurement challenges, most existing research has focused on the general population. This leaves a significant gap in our understanding of the interplay between environmental factors and lifestyle behaviours in high-risk groups. From a public health perspective, population-wide interventions are logical, as they yield larger absolute health gains. However, a substantial and growing portion of the population is already at high risk for developing CVD. For these individuals, understanding which environmental factors are most salient could inform targeted interventions. Studies in high-risk populations are therefore crucial for determining which factors warrant prioritisation, particularly those linked to adverse health outcomes in both the general and high-risk population. [\u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e]\u003c/p\u003e \u003cp\u003eGiven these complexities, a data-driven approach is needed to understand how different aspects of the neighbourhood environment cluster together and whether these patterns, rather than individual indicators, are meaningfully associated with health behaviours. This study has two aims: (1) to identify patterns of co-occurring neighbourhood characteristics related to food environment, green spaces, and walkability in a highly urbanised area; and (2) to explore whether these neighbourhood patterns are associated with weight status, diet quality, and physical activity in primary care patients at high risk for CVD in an urban setting. By including multiple operationalisations of the same environmental constructs, we additionally examine to what extent different measurement approaches capture similar or distinct dimensions of the neighbourhood environment.\u003c/p\u003e "},{"header":"Methods","content":"\u003ch3\u003eStudy design and setting\u003c/h3\u003e\u003cp\u003eThis study was conducted as a secondary analysis using baseline data from the Healthy Heart study, a non-randomized cluster stepped-wedge trial investigating the effect of a lifestyle program on patients in the primary cardiovascular prevention care protocol in general practice in the Hague. [\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e] The Hague is the third largest city of the Netherlands, with approximately 540.000 residents and characterized by dense urban development.\u003c/p\u003e\u003cp\u003e The Healthy Heart study was approved by the Ethics Committee of the Leiden University Medical Centre (P17.079). All participants provided written informed consent.\u003c/p\u003e\u003ch2\u003eStudy area and spatial unit definition\u003c/h2\u003e\u003cp\u003eEnvironmental characterization was conducted for 73 four-digit postal codes where participants resided: 48 in The Hague (65.8%) and 25 (34.2%) in directly adjacent municipalities, reflecting typical catchment areas of general practices in the Hague. Because postal codes vary in size, environmental indicators were first calculated at the address level using residential address locations from the national Registration of Addresses and Buildings (BAG) [\u003cspan class=\"CitationRef\"\u003e19\u003c/span\u003e], and then averaged at the postal code level. We included indicators at two spatial scales: within 300 meters (immediate neighbourhood) and within 300–1000 meters (extended neighbourhood, calculated as a donut-shaped buffer).\u003c/p\u003e\u003ch3\u003eEnvironmental exposure measures\u003c/h3\u003e\u003cp\u003eGeospatial analyses were performed using Qgis version 3.18.3 to calculate the 23 indicators presented in Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e. For an extensive description of the geospatial analyses, see Additional File 1.\u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\"\u003e\u003c/div\u003e\u003ctable id=\"Tab1\" border=\"1\"\u003e \u003ccaption\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eIncluded environmental exposure variables.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003c/colgroup\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\"\u003e \u003cp\u003eCategory\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\"\u003e \u003cp\u003eCalculation\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\"\u003e \u003cp\u003eVariable Name\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eFood environment\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eDensity of core food retailers in buffers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eNumber of core retailers in buffer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eCore_300m, Core_donut\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eDensity of non-core food retailers in buffers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eNumber of non-core retailers in buffer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eNoncore_300m, Noncore_donut\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eModified retail food environment index (MRFEI) in buffers* [\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003e(Number of core retailers + 1) / ((Number of core retailers + 1) + (Number of non-core retailers + 1)) × 100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eMRFEI_300m, MRFEI_donut\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eFood environment health index (FEHI) in buffers [\u003cspan class=\"CitationRef\"\u003e23\u003c/span\u003e]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eSum of FEHI-scores of food retailers in buffer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eFEHI_300m, FEHI_donut\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eSpatial access to core-retailers in buffers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eAccessibility score to core retailers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eSpac_core_300m, Spac_core_donut\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eSpatial access to non-core retailers in buffers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eAccessibility score to non-core retailers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eSpac_noncore_300m, Spac_noncore_donut\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eRatio of spatial access scores in buffers* [\u003cspan class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003e(Spatial access to core retailers + 1) / ((Spatial access to core + 1) + (Spatial access to non-core retailers + 1))\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eRel_spac_300m, Rel_spac_donut\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eGreen spaces\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eDensity of accessible green spaces in buffers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003ePercentage accessible green space in buffer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003ePerc_access_green_300m, Perc_access_green_donut\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eDensity of other green spaces in buffers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003ePercentage other green space in buffer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003ePerc_other_green_300m, Perc_other_green_donut\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eNetwork distance to nearest entrance of accessible green space\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eShortest walking distance (in meters) to nearest entrance of accessible green space\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eNearest_park_network\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eMean Normalized Difference Vegetation Index (NDVI) in buffers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eAverage NDVI value in buffer area\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eNDVI_300m, NDVI_donut\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eWalkability\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eWalkability score in 4 digit postal code [\u003cspan class=\"CitationRef\"\u003e26\u003c/span\u003e]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eComposite walkability index score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\"\u003e \u003cp\u003eWalkability_250m, Walkability_1000m\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/table\u003e\u003c/div\u003e\u003cp\u003e \u003cem\u003e*To avoid division by zero in areas with no food retailers, a score of one was added to the number of core and non-core retailers.\u003c/em\u003e \u003c/p\u003e\u003ch3\u003eFood-environment\u003c/h3\u003e\u003cp\u003eFood-environment data, published in 2019, was obtained from Locatus, a company that collects information on business locations through regular field audits. [\u003cspan class=\"CitationRef\"\u003e20\u003c/span\u003e] Following previous example we calculated: density of core and non-core food retailers [\u003cspan class=\"CitationRef\"\u003e21\u003c/span\u003e], total Food Environment Healthiness Index (FEHI) score [\u003cspan class=\"CitationRef\"\u003e21\u003c/span\u003e], spatial accessibility to core and non-core retailers, the Modified Retail Food Environment Index (MRFEI) and relative spatial accessibility. [\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e23\u003c/span\u003e]\u003c/p\u003e\u003ch3\u003eGreen spaces\u003c/h3\u003e\u003cp\u003eFour indicators were assessed using Dutch Land Use data (2017) [\u003cspan class=\"CitationRef\"\u003e24\u003c/span\u003e] and Landsat 8 satellite imagery [\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e]: percentage of accessible green space (e.g. parks, forests) in buffer, percentage of other green space (e.g. sports fields, agricultural land), network distance to nearest accessible green space entrance, and mean Normalised Difference Vegetation Index (NDVI) within buffer.\u003c/p\u003e\u003ch3\u003eWalkability\u003c/h3\u003e\u003cp\u003eAs an additional environmental factor, the Dutch walkability index developed by Lam et. al. [\u003cspan class=\"CitationRef\"\u003e26\u003c/span\u003e] obtained through the Geoscience and Health Cohort Consortium (GECCO) was used. [\u003cspan class=\"CitationRef\"\u003e27\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e28\u003c/span\u003e] This composite score, which can be used to assess the degree to which an area is conducive to walking as a mode of transportation, includes population density, density of retail and service destinations, land-use mix, street connectivity, green space and sidewalk presence. The walkability index was calculated by averaging the z-scores of these components, and then scaling the result to range from 0 to 100. The index was previously positively associated with walking behaviour in Dutch adults. [\u003cspan class=\"CitationRef\"\u003e26\u003c/span\u003e] Data, for the year 2017, was pre-processed on address-level by GECCO and aggregated on average score in a 250m and 1000m buffer.\u003c/p\u003e\u003cp\u003e All 23 indicators were included in the PCA, including multiple operationalisations of the same constructs (e.g., density, accessibility, and relative measures of the food environment) and two spatial scales (300 m and 300–1000 m buffer). The PCA thus serves as an empirical test of whether these operationalisations capture the same or different environmental dimensions: indicators measuring the same construct will load on the same component, while those capturing distinct aspects will separate.\u003c/p\u003e\u003cp\u003eFor further details on the included indicators, see Additional File 1.\u003c/p\u003e\u003ch2\u003eCorrelation analysis\u003c/h2\u003e\u003cp\u003eTo explore relationships between environmental indicators and contextual neighbourhood characteristics, a correlation analysis was conducted. Pairwise Spearman correlations were calculated between all environmental indicators and several contextual neighbourhood characteristics: urbanicity class (five levels, defined by Statistics Netherlands (CBS) [\u003cspan class=\"CitationRef\"\u003e29\u003c/span\u003e]: very strongly urban (\u0026gt; 2500 addresses per km²), strongly urban (between 1500–2500 addresses per km²), moderately urban (between 1000 and 1500 addresses per km²), slightly urban (between 500–1000 addresses per km²), non-urban/rural (\u0026lt; 500 addresses per km²)), population density (continuous, using addresses per km² as a proxy; CBS 2017 [\u003cspan class=\"CitationRef\"\u003e29\u003c/span\u003e]), and neighbourhood socioeconomic position (SES-WOA). The SES-WOA score is a comprehensive measure of socioeconomic position at the local level, based on a combination of data regarding financial prosperity, educational level and recent employment history of the households. [\u003cspan class=\"CitationRef\"\u003e30\u003c/span\u003e]\u003c/p\u003e\u003ch3\u003ePrincipal component analysis\u003c/h3\u003e\u003cp\u003ePrior to PCA, variables were Box-Cox transformed to approximate normality and z-standardised (see Additional File 1 for details). To address multicollinearity issues and identify clusters of co-occurring neighbourhood features, a PCA with varimax rotation at the postal code level (n = 73) was performed, using the 23 standardized environmental indicators. To check sampling adequacy, Kaiser-Meyer-Olkin measure and Bartlett’s test of sphericity were used. The number of extracted components was based on eigenvalues greater than 1 and the scree plot. Varimax rotation was applied and rotated component scores were assigned to participants. For clarity, throughout the remainder of this article, the varimax rotated principal component scores will be referred to as the principal components (PC).\u003c/p\u003e\u003ch3\u003eInterpretation of PCA components\u003c/h3\u003e\u003cp\u003eTo explore how the environmental indicators relate to broader contextual neighbourhood factors, Spearman correlations were calculated between each continuous PC score (PC1–PC4) and two contextual neighbourhood variables that were not included in the PCA: Population density, and neighbourhood socioeconomic position (SES-WOA).\u003c/p\u003e\u003ch2\u003eExploratory health associations\u003c/h2\u003e\u003cp\u003eHaving characterised environmental patterns in the neighbourhoods, we conducted exploratory analyses to examine whether these patterns associated with health behaviours in a sample of residents at high cardiovascular risk.\u003c/p\u003e\u003ch2\u003eStudy population\u003c/h2\u003e\u003cp\u003ePatients at high risk for cardiovascular disease from 56 general practices in the Hague, Netherlands were invited to participate in the Healthy Heart study during a primary care consultation with their practice nurse or general practitioner (GP). For the current analysis, we included only baseline data of patients included during the control period (before intervention implementation) who had a valid postal code, to avoid potential bias due to intervention participation. During the control period patients received care as usual.\u003c/p\u003e\u003cp\u003eIdentification of high-risk patients is carried out as part of usual care in GP practices, without any involvement from the researchers. High-risk patients were defined as having a 10-year cardiovascular risk of ≥ 10% according to Dutch guidelines at the moment of inclusion (2017–2019). [\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e] As part of standard care, these patients have one to four individual consultations annually with a trained practice nurse. During these sessions, they receive lifestyle counselling, undergo lab work, and have physical examinations, all in line with Dutch Primary Care standards.\u003c/p\u003e\u003cp\u003ePatients with an ICPC coded diagnosis of diabetes mellitus or cardiovascular disease were excluded from participation in the general CVD prevention program, since they were already enrolled in separate care programs. Data on sociodemographics, lifestyle behaviour and BMI were collected using self-reported online questionnaires.\u003c/p\u003e\u003ch2\u003eHealth outcome variables\u003c/h2\u003e\u003cp\u003eThree outcomes were assessed: (1) diet quality, measured using the Dutch Healthy Diet index (DHD-index, range 0–160) [\u003cspan class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e33\u003c/span\u003e] based on the DHD-FFQ; (2) physical activity, measured using the SQUASH questionnaire [\u003cspan class=\"CitationRef\"\u003e34\u003c/span\u003e] as total minutes per week; and (3) weight status, based on self-reported BMI categorised as healthy weight, overweight, or obese using age-specific cut-offs from the Dutch Nutrition Centre [\u003cspan class=\"CitationRef\"\u003e35\u003c/span\u003e] (see Additional File 1 for more details). Outliers \u0026gt;3SD were removed. Missing DHD scores (6.7%) were addressed through zero imputation of subcomponents when no more than two of 16 were missing. [\u003cspan class=\"CitationRef\"\u003e36\u003c/span\u003e] No imputation was performed for BMI (\u0026lt; 1% missing) or physical activity (\u0026lt; 2.5% missing).\u003c/p\u003e\u003ch2\u003ePotential confounders\u003c/h2\u003e\u003cp\u003eThe following potential individual- and neighbourhood-level confounders were used in the regression analyses to adjust for associations between principal components and health outcomes. Individual confounders included: age (years), sex (male/female), level of education (lower, middle, higher) and migration background (yes/no). Neighbourhood level confounders included: population density (using continuous addresses per km² as a proxy) and SEP of the postal code (SES-WOA).\u003c/p\u003e\u003ch2\u003eStatistical analysis of health associations\u003c/h2\u003e\u003cp\u003eSample characteristics were provided using descriptive statistics, summarized as percentage for categorical variables and mean (standard deviation) or median (IQR) for continuous variables, depending on distribution of the variable.\u003c/p\u003e\u003cp\u003eIntraclass correlation coefficients for all outcomes were \u0026lt; 0.05 (weight status = 0.02; DHD ≈ 0.00; physical activity = 0.01), indicating minimal clustering within postal codes; standard regression was therefore used without multilevel modelling.\u003c/p\u003e\u003cp\u003eFor interpretability, the PCs were divided into quartiles prior to performing the regression analyses, as a one-unit change in PC score is not inherently meaningful. This decision was further supported by the absence of clear linear relationships between the PCs and outcomes.\u003c/p\u003e\u003cp\u003eThe association between PC quartile and the DHD-score and total minutes of physical activity per week (square root transformed due to heteroscedasticity of the untransformed variable) were analysed using univariate and multivariate linear regression. The association between the PCs and weight status was analysed using a multinomial logistic regression (as the proportional odds assumption for an ordinal logistic regression was violated), with healthy weight set as the reference category. Assumptions were checked before conducting the analyses. Although population density correlated strongly with PC1 (continuous: r = 0.86, VIF = 8.6), quartile coding of PC1 reduced multicollinearity (VIF = 4.5).\u003c/p\u003e\u003cp\u003eIn the univariate models, the PC quartiles were included individually (a separate model for each PC, without any further adjustments). The first multivariate model (model 1) also incorporated individual and neighbourhood level confounders. In the final model (model 2), all PCs were included simultaneously and adjusted for the individual and neighbourhood level confounders. Due to sample size considerations, the PC of interest was included in the model as a categorical (dummy or factor) variable, with the lowest quartile used as the reference category. This approach was consistent with how the PC was handled in the univariate model and model 1. In contrast, the other PCs that were used for adjustment (in model 2) were added as ordinal quartile scores (1–4). To establish dose-response relationships, linear trends (P-trend) across the quartiles were examined using the median values of each quartile as continuous variable in the analyses.\u003c/p\u003e\u003ch2\u003eAdditional Analyses\u003c/h2\u003e\u003cp\u003eAdditional analyses of dietary and physical activity subcomponents, as well as effect modification by neighbourhood socioeconomic position, are described in Additional File 2.\u003c/p\u003e\u003cp\u003eStatistical analyses were performed in Python version 3.8 and SPSS version 29.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003e\u003cem\u003eSample Characteristics\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCharacteristics of the 475 participants are presented in Table 2. Participants were on average 65.6 years of age and 55.4% were female. A minority had a migration background (14.7%), most were middle to higher educated (68.7%), most participants below the age of 65 were employed (72.6%) and almost all participants lived in a (very) strong urban area (94.5%). Furthermore, 53.9% were either overweight or obese, and only 37.9% met the physical activity guideline of 150 minutes of moderate to heavy physical activity per week.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe median address density was 2836 addresses per km\u0026sup2; (range: 362\u0026ndash;8499), corresponding to urbanicity class very strongly urban. The median neighbourhood SES-WOA score was 0.09 (range: \u0026ndash;0.98 to 0.58).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eTable 2. Characteristics of participants (n=475)\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"387\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 236px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCharacteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eAge (yr), mean (SD)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 180px;\"\u003e\n \u003cp\u003e65.6 (9.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eSex, female, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 180px;\"\u003e\n \u003cp\u003e263 (55.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eMigration background, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 180px;\"\u003e\n \u003cp\u003e70 (14.7%)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eEducational level, n (%)\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Low\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Middle\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;High\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 180px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e131 (27.6%)\u003c/p\u003e\n \u003cp\u003e146 (30.7%)\u003c/p\u003e\n \u003cp\u003e184 (38.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eEmployed (\u0026lt;65yr), n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 180px;\"\u003e\n \u003cp\u003e151 (72.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eWeight (kg), mean (SD)\u003c/p\u003e\n \u003cp\u003eWeight status, n (%)\u003c/p\u003e\n \u003cp\u003e- Healthy weight\u003c/p\u003e\n \u003cp\u003e- Overweight\u003c/p\u003e\n \u003cp\u003e- Obese\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 180px;\"\u003e\n \u003cp\u003e81.0 (16.1)\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e213 (44.8%)\u003c/p\u003e\n \u003cp\u003e156 (32.8%)\u003c/p\u003e\n \u003cp\u003e100 (21.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eBMI (kg/m2), mean (SD)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 180px;\"\u003e\n \u003cp\u003e27.1 (4.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eTotal DHD-score, mean (SD)\u003c/p\u003e\n \u003cp\u003eTotal minutes of physical activity per week, mean (SD)\u003c/p\u003e\n \u003cp\u003eMeets physical activity guideline, n (%)\u003c/p\u003e\n \u003cp\u003eAddress density, median (IQR)\u003c/p\u003e\n \u003cp\u003eNeighbourhood SES, median (IQR)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 180px;\"\u003e\n \u003cp\u003e98.4 (17.1)\u003c/p\u003e\n \u003cp\u003e2045.5 (1200.1)\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e180 (37.9%)\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e2836 (1865, 4130)\u003c/p\u003e\n \u003cp\u003e0.09 (-0.13, 0.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 180px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 208px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 28px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 151px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e*\u0026nbsp;Measure of socioeconomic status at the local level, based on a combination of data regarding financial prosperity, educational level and recent employment history of the households. [30]\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorrelation analyses\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eStrong intercorrelations were observed among variables within the same exposure domains (Additional File 2, Figure A1). Densities of core and non-core food retailers were very highly correlated (\u0026rho; \u0026asymp; 0.8), indicating that neighbourhoods with many non-core outlets tend to also have more healthy food retailers. Walkability was also strongly related to food-retailer density, particularly within the 1000m buffer for walkability (\u0026rho; \u0026asymp; 0.8).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRelative food environment measures were strongly intercorrelated but showed lower correlations with absolute food density (not significantly correlated with core retailer density, significant negative correlation (\u0026rho; \u0026asymp; -0.4) with non-core retailer density and significant positive correlation with inverse FEHI (\u0026rho; \u0026asymp; 0.3-0.4)).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn contrast, green-space density and quality measures (but not network distance to nearest accessible green space) showed consistent negative correlations with food-environment and walkability indicators, reflecting that greener areas generally coincide with fewer retail destinations. Network distance to nearest accessible green space showed lower and non-significant correlations with other green space measures.\u003c/p\u003e\n\u003cp\u003eWhen examining contextual characteristics, population density showed much stronger correlations with environmental indicators than the ordinal classification. Population density correlated strongly with core retailer density (\u0026rho; \u0026asymp; 0.9), non-core density (\u0026rho; \u0026asymp; 0.8-0.9), and walkability (\u0026rho; \u0026asymp; 0.7-0.9), and inversely with NDVI and green-space indicators (between \u0026rho; = -0.3 and -0.8). Even within the \u0026ldquo;very strongly urban\u0026rdquo; CBS category (\u0026gt; 2,500 addresses /km\u0026sup2;), substantial correlation remained: higher population density continued to be associated with higher retail density (between \u0026rho; = 0.7 and 0.9) (Figure 1). Population density was not significantly correlated with the relative food environment measures.\u003c/p\u003e\n\u003cp\u003eNeighbourhood socioeconomic position showed weaker relationships with the environmental indicators: lower-SES areas tended to have higher retail density (between \u0026rho; = -0.4 and -0.6) with lower FEHI-score and higher walkability (\u0026rho; = -0.6, \u0026rho; = -0.4) but less green space (although associations were low and only the association with accessible green space in donut buffer was significant (\u0026rho; = 0.2)).\u003c/p\u003e\n\u003cp\u003eComparing the two spatial scales revealed that environmental characteristics in the immediate (300 m) and extended (300\u0026ndash;1000 m, donut) buffers were generally highly correlated, indicating that neighbourhood features largely co-occur across scales in this urban area. The relative food environment measures showed lower correlations across scales (\u0026rho; \u0026asymp; 0.6).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePrincipal Component Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe Kaiser-Meyer-Olkin (KMO) measure confirmed sampling adequacy (KMO = 0.842), and Bartlett\u0026rsquo;s test of sphericity was significant (p \u0026lt; 0.001), indicating that the data were suitable for dimensionality reduction.\u003cbr\u003e\u0026nbsp;Four components had eigenvalues \u0026gt;1 and were retained based on the scree plot, together explaining approximately 87% of the total variance (Table 3).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eComponent structure and interpretation\u003c/strong\u003e\u003cbr\u003e\u0026nbsp;The rotated component loadings (Figure 2) revealed four distinct environmental patterns. For interpretability, component names describe the environmental characteristics of areas scoring in the highest quartile.\u003c/p\u003e\n\u003cul type=\"disc\"\u003e\n \u003cli\u003ePC1 \u0026ndash; \u0026ldquo;High food, low FEHI, high walkability\u0026rdquo;\u003cbr\u003e\u0026nbsp;This component had high positive loadings for both core and non-core food retailer densities, walkability, and inversed FEHI and average negative loadings for NDVI.\u003cbr\u003e\u0026nbsp;It reflects highly urbanised environments characterized by dense food retailer presence (both healthy and unhealthy) and high walkability, but limited green space.\u003c/li\u003e\n \u003cli\u003ePC2 \u0026ndash; \u0026ldquo;Relative food environment advantage\u0026rdquo;\u003cbr\u003e\u0026nbsp;This component loaded strongly on the relative measures (MRFEI, relative spatial access). It represents neighbourhoods where healthier food options dominate relative to unhealthy ones, independent of overall retail density.\u003c/li\u003e\n \u003cli\u003ePC3 \u0026ndash; \u0026ldquo;Low density and high distance to accessible green\u0026rdquo;\u003cbr\u003e\u0026nbsp;PC3 was defined by positive loadings for network distance to nearest accessible greenspace and negative loadings for accessible green space density, indicating low-density areas with poor accessibility to nearby green accessible space.\u003c/li\u003e\n \u003cli\u003ePC4 \u0026ndash; \u0026ldquo;Low-quality, agricultural/inaccessible green\u0026rdquo;\u003cbr\u003e\u0026nbsp;This component had high negative loadings for inaccessible (\u0026lsquo;other\u0026rsquo;, mainly agricultural) green and NDVI, reflecting areas characterized by low vegetation quality and non-accessible agricultural green space.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eComponent correlations with contextual variables.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo explore how these principal components related to broader neighbourhood context, Spearman correlations were computed between component scores and population density and neighbourhood socioeconomic position (See Additional File 2, Figure A2). \u0026nbsp;\u003c/p\u003e\n\u003cp\u003ePopulation density showed a very strong positive correlation with PC1 (\u0026rho; = 0.9), while neighbourhood socioeconomic position was inversely correlated (\u0026rho; = \u0026ndash;0.6), indicating that the \u0026ldquo;High food, low FEHI, high walkability\u0026rdquo; pattern coincides with densely populated, lower-SEP areas. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOther components were only weakly associated with neighbourhood socioeconomic position or population density (|\u0026rho;| \u0026lt; 0.4). \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThese findings suggest that while the first component largely reflects a population density gradient, the remaining components capture neighbourhood environments that are largely independent of population density.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eTable 3. Explained variance of varimax rotated principal components\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePC1\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePC2\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePC3\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePC4\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eVariance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e9.621\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e4.148\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e2.321\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e3.838\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eProportional variance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.418\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.180\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.101\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.167\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eCumulative variance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.418\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.599\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.700\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.866\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eAssociation with Weight Status\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn both the univariate and the multivariate multinomial logistic regression analyses (Table 4), patients in higher quartiles of the \u0026ldquo;\u003cem\u003eRelative food environment advantage\u003c/em\u003e\u0026rdquo; component showed statistically significant lower odds of being overweight (for Q3 (but not Q4) in multivariate model 2: OR=0.45, 95% CI [0.22, 0.95], P-trend=0.19). Patients in the third quartile of \u0026ldquo;\u003cem\u003eRelative food environment advantage\u003c/em\u003e\u0026rdquo; also showed a lower risk of being obese in the univariate model and model 1, but this was no longer the case in model 2. \u0026nbsp;These results suggest that neighbourhoods with relatively higher proportions of healthy food retailers were associated with a lower likelihood of being overweight, although no significant p for trend was found.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eUnivariate results also showed statistically significant higher odds of obesity for patients in higher quartiles of the \u0026ldquo;\u003cem\u003eHigh food, low FEHI, high walkability\u003c/em\u003e\u0026rdquo; component (Q4: OR = 1.98, 95% CI [1.04, 3.78]), however, this association was no longer significant in the adjusted models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssociations with Diet Quality\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn the linear regression analysis (Table 5), a statistically significant association was observed for the \u003cem\u003e\u0026ldquo;Low quality, agricultural/inaccessible green\u0026rdquo;\u003c/em\u003e component. In the fully adjusted model (model 2), participants in the second quartile (Q2) had lower DHD scores compared with those in the lowest quartile (\u0026beta; = \u0026minus;6.00, 95% CI [\u0026minus;11.58, \u0026minus;0.41]). This association was not observed for higher quartiles, and no significant trend across quartiles was found (P-trend = 0.37). For the other components no statistically significant associations with the DHD index were identified in either univariate or multivariate analyses.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssociations with Physical Activity\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe linear regression analyses (Table 6) indicated a significant negative association between the highest quartile of the \u003cem\u003e\u0026ldquo;Relative food environment advantage\u0026rdquo;\u003c/em\u003e component and total minutes of physical activity per week in the univariate analysis (\u0026beta; = \u0026minus;3.85, 95% CI [\u0026minus;7.44, \u0026minus;0.26]). This suggests that patients living in neighbourhoods with a relatively higher proportion of healthy food retailers reported fewer minutes of physical activity per week. However, this association was no longer statistically significant after adjustment for individual and neighbourhood-level confounders in models 1 and 2. For the other components, no statistically significant associations were observed in either univariate or multivariate models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eTable 4. Result of univariate and multivariate multinomial logistic regression analysis between varimax rotated principal components and weight status with healthy weight as the reference category\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" align=\"\" width=\"749\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003eWeight status\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003eRotated Components\u003c/p\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003cp\u003e\u003cem\u003eQuartiles\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87px;\"\u003e\n \u003cp\u003eUnivariate analyses\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 74px;\"\u003e\n \u003cp\u003eMultivariate model 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eMultivariate model 2\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 72px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cem\u003eOverweight\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87px;\"\u003e\n \u003cp\u003e\u003cem\u003eOR\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u003cem\u003e95% CI\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 74px;\"\u003e\n \u003cp\u003e\u003cem\u003eOR\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u003cem\u003e95% CI\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u003cem\u003eOR\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 72px;\"\u003e\n \u003cp\u003e\u003cem\u003e95% CI\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u003cem\u003eP for trend\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003eHigh food, low FEHI, high walkability\u003c/p\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.12\u003c/p\u003e\n \u003cp\u003e1.15\u003c/p\u003e\n \u003cp\u003e0.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.64, 1.97\u003c/p\u003e\n \u003cp\u003e0.66, 2.02\u003c/p\u003e\n \u003cp\u003e0.39, 1.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 74px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.25\u003c/p\u003e\n \u003cp\u003e1.15\u003c/p\u003e\n \u003cp\u003e1.04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.66, 2.40\u003c/p\u003e\n \u003cp\u003e0.58, 2.29\u003c/p\u003e\n \u003cp\u003e0.41, 2.61\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.35\u003c/p\u003e\n \u003cp\u003e1.21\u003c/p\u003e\n \u003cp\u003e1.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 72px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.67, 2.72\u003c/p\u003e\n \u003cp\u003e0.51, 2.85\u003c/p\u003e\n \u003cp\u003e0.33, 3.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e0.86\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003eRelative food environment advantage \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; Q2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.86\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e0.50*\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e0.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.48, 1.53\u003c/p\u003e\n \u003cp\u003e0.27, 0.93\u003c/p\u003e\n \u003cp\u003e0.39, 1.30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 74px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.79\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e0.48*\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e0.68\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.41, 1.52\u003c/p\u003e\n \u003cp\u003e0.24, 0.94\u003c/p\u003e\n \u003cp\u003e0.34, 1.37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.82\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e0.45*\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e0.66\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 72px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.42, 1.59\u003c/p\u003e\n \u003cp\u003e0.22, 0.94\u003c/p\u003e\n \u003cp\u003e0.32, 1.36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e0.19\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003eLow density and high distance to accessible green\u003c/p\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.08\u003c/p\u003e\n \u003cp\u003e1.10\u003c/p\u003e\n \u003cp\u003e1.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.60, 1.91\u003c/p\u003e\n \u003cp\u003e0.62, 1.96\u003c/p\u003e\n \u003cp\u003e0.55, 1.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 74px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.93\u003c/p\u003e\n \u003cp\u003e1.51\u003c/p\u003e\n \u003cp\u003e0.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.49, 1.79\u003c/p\u003e\n \u003cp\u003e0.80, 2.84\u003c/p\u003e\n \u003cp\u003e0.43, 1.62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.04\u003c/p\u003e\n \u003cp\u003e1.59\u003c/p\u003e\n \u003cp\u003e0.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 72px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.51, 2.13\u003c/p\u003e\n \u003cp\u003e0.83, 3.05\u003c/p\u003e\n \u003cp\u003e0.42, 1.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e0.93\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003eLow quality, low inaccessible, agricultural green\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.86\u003c/p\u003e\n \u003cp\u003e0.76\u003c/p\u003e\n \u003cp\u003e0.74\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.49, 1.52\u003c/p\u003e\n \u003cp\u003e0.43, 1.35\u003c/p\u003e\n \u003cp\u003e0.41, 1.31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 74px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.90\u003c/p\u003e\n \u003cp\u003e0.97\u003c/p\u003e\n \u003cp\u003e1.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.46, 1.17\u003c/p\u003e\n \u003cp\u003e0.49, 1.93\u003c/p\u003e\n \u003cp\u003e0.56, 2.58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.09\u003c/p\u003e\n \u003cp\u003e1.15\u003c/p\u003e\n \u003cp\u003e1.35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 72px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.48, 2.50\u003c/p\u003e\n \u003cp\u003e0.47, 2.81\u003c/p\u003e\n \u003cp\u003e0.50, 3.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e0.55\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cem\u003eObese\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87px;\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 74px;\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 72px;\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003eHigh food, low FEHI, high walkability\u003c/p\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e1.07\u003c/p\u003e\n \u003cp\u003e1.12\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e1.98*\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.53, 2.18\u003c/p\u003e\n \u003cp\u003e0.56, 2.27\u003c/p\u003e\n \u003cp\u003e1.04, 3.78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 74px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.31\u003c/p\u003e\n \u003cp\u003e1.10\u003c/p\u003e\n \u003cp\u003e1.61\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.60, 2.82\u003c/p\u003e\n \u003cp\u003e0.49, 2.48\u003c/p\u003e\n \u003cp\u003e0.60, 4.29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.44\u003c/p\u003e\n \u003cp\u003e1.31\u003c/p\u003e\n \u003cp\u003e1.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 72px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.627, 3.285\u003c/p\u003e\n \u003cp\u003e0.488, 3.497\u003c/p\u003e\n \u003cp\u003e0.555, 5.997\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e0.36\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003eRelative food environment advantage \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; Q2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.60\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e0.49*\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e0.53\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.31, 1.16\u003c/p\u003e\n \u003cp\u003e0.25, 0.95\u003c/p\u003e\n \u003cp\u003e0.27, 1.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 74px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e0.46*\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e0.43*\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e0.46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.22, 0.95\u003c/p\u003e\n \u003cp\u003e0.21, 0.90\u003c/p\u003e\n \u003cp\u003e0.21, 1.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.50\u003c/p\u003e\n \u003cp\u003e0.48\u003c/p\u003e\n \u003cp\u003e0.49\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 72px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.23, 1.07\u003c/p\u003e\n \u003cp\u003e0.21, 1.08\u003c/p\u003e\n \u003cp\u003e0.21, 1.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e0.13\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003eLow density and high distance to accessible green\u003c/p\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.08\u003c/p\u003e\n \u003cp\u003e1.15\u003c/p\u003e\n \u003cp\u003e1.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.55, 2.10\u003c/p\u003e\n \u003cp\u003e0.60, 2.25\u003c/p\u003e\n \u003cp\u003e0.53, 2.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 74px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.03\u003c/p\u003e\n \u003cp\u003e1.40\u003c/p\u003e\n \u003cp\u003e0.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.49, 2.13\u003c/p\u003e\n \u003cp\u003e0.68, 2.87\u003c/p\u003e\n \u003cp\u003e0.34, 1.54\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.34\u003c/p\u003e\n \u003cp\u003e1.54\u003c/p\u003e\n \u003cp\u003e1.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 72px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.61, 2.95\u003c/p\u003e\n \u003cp\u003e0.74, 3.21\u003c/p\u003e\n \u003cp\u003e0.43, 2.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e0.88\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003eLow-quality, agricultural/inaccessible green \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; Q2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.21\u003c/p\u003e\n \u003cp\u003e1.24\u003c/p\u003e\n \u003cp\u003e1.48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.60, 2.43\u003c/p\u003e\n \u003cp\u003e0.62, 2.47\u003c/p\u003e\n \u003cp\u003e0.76, 2.91\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 74px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.92\u003c/p\u003e\n \u003cp\u003e1.14\u003c/p\u003e\n \u003cp\u003e1.51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.42, 2.03\u003c/p\u003e\n \u003cp\u003e0.51, 2.56\u003c/p\u003e\n \u003cp\u003e0.65, 3.55\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.379\u003c/p\u003e\n \u003cp\u003e1.772\u003c/p\u003e\n \u003cp\u003e2.144\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 72px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.524, 3.628\u003c/p\u003e\n \u003cp\u003e0.638, 4.921\u003c/p\u003e\n \u003cp\u003e0.715, 6.427\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e0.16\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003e*p\u0026lt;0.05. Multivariate model 1 included the individual principal components per model corrected for age, sex, educational level, migration background, SES-WOA score of 4-digit postal code and population density. Multivariate model 2 included all principal components corrected for the aforementioned confounders. The first quartile of the principal components was set as the reference category.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eTable 5. Result of univariate and multivariate regression analysis between varimax rotated principal components and total DHD-index score\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" align=\"\" width=\"718\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003eRotated Components\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 41px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003eUnivariate analyses\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003eMultivariate model 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003eMultivariate model 2\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 41px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e\u0026beta;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e\u0026beta;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e\u0026beta;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003eP for trend\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003eHigh food, low FEHI, \u0026nbsp;high walkability\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 41px;\"\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e-0.70\u003c/p\u003e\n \u003cp\u003e-0.78\u003c/p\u003e\n \u003cp\u003e-1.64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-5.05, 3.66\u003c/p\u003e\n \u003cp\u003e-5.12, 3.55\u003c/p\u003e\n \u003cp\u003e-6.00, 2.71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-1.01\u003c/p\u003e\n \u003cp\u003e-1.09\u003c/p\u003e\n \u003cp\u003e-0.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-5.47, 3.45\u003c/p\u003e\n \u003cp\u003e-5.79, 3.62\u003c/p\u003e\n \u003cp\u003e-6.29, 5.95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-2.10\u003c/p\u003e\n \u003cp\u003e-3.94\u003c/p\u003e\n \u003cp\u003e-3.94\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-6.87, 2.68\u003c/p\u003e\n \u003cp\u003e-9.74, 1.87\u003c/p\u003e\n \u003cp\u003e-11.36, 3.48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e0.22\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003eRelative food environment advantage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 41px;\"\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e1.34\u003c/p\u003e\n \u003cp\u003e1.76\u003c/p\u003e\n \u003cp\u003e-0.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-2.96, 5.64\u003c/p\u003e\n \u003cp\u003e-2.67, 6.19\u003c/p\u003e\n \u003cp\u003e-5.30, 3.64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.55\u003c/p\u003e\n \u003cp\u003e-0.13\u003c/p\u003e\n \u003cp\u003e-3.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-3.79, 4.90\u003c/p\u003e\n \u003cp\u003e-4.59, 4.34\u003c/p\u003e\n \u003cp\u003e-7.71, 1.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-0.04\u003c/p\u003e\n \u003cp\u003e-0.97\u003c/p\u003e\n \u003cp\u003e-3.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-4.52, 4.44\u003c/p\u003e\n \u003cp\u003e-5.84, 3.90\u003c/p\u003e\n \u003cp\u003e-8.47, 1.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e0.12\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003eLow density and high distance to accessible green\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 41px;\"\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e-0.72\u003c/p\u003e\n \u003cp\u003e0.63\u003c/p\u003e\n \u003cp\u003e-2.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-5.04, 3.61\u003c/p\u003e\n \u003cp\u003e-3.70, 4.96\u003c/p\u003e\n \u003cp\u003e-6.48, 2.37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-1.28\u003c/p\u003e\n \u003cp\u003e0.11\u003c/p\u003e\n \u003cp\u003e-0.64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-5.62, 3.06\u003c/p\u003e\n \u003cp\u003e-4.18, 4.41\u003c/p\u003e\n \u003cp\u003e-5.08, 3.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-1.33\u003c/p\u003e\n \u003cp\u003e-0.33\u003c/p\u003e\n \u003cp\u003e-1.78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-6.02, 3.36\u003c/p\u003e\n \u003cp\u003e-4.70, 4.03\u003c/p\u003e\n \u003cp\u003e-6.75, 3.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e0.56\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003eLow-quality, agricultural/inaccessible green\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 41px;\"\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e-4.25\u003c/p\u003e\n \u003cp\u003e-2.38\u003c/p\u003e\n \u003cp\u003e-4.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-8.61, 0.10\u003c/p\u003e\n \u003cp\u003e-6.74, 1.98\u003c/p\u003e\n \u003cp\u003e-8.59, 0.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-4.11\u003c/p\u003e\n \u003cp\u003e-2.29\u003c/p\u003e\n \u003cp\u003e-2.81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-8.66, 0.45\u003c/p\u003e\n \u003cp\u003e-7.00, 2.41\u003c/p\u003e\n \u003cp\u003e-7.94, 2.31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e-6.00*\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e-4.87\u003c/p\u003e\n \u003cp\u003e-6.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-11.6, -0.41\u003c/p\u003e\n \u003cp\u003e-10.9, 1.12\u003c/p\u003e\n \u003cp\u003e-12.9, 0.39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e0.37\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003e*p\u0026lt;0.05. Multivariate model 1 included the individual principal components per model corrected for age, sex, educational level, migration background, SES-WOA score of 4-digit postal code and population density. Multivariate model 2 included all principal components corrected for the aforementioned confounders.\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eTable 6. Result of univariate and multivariate regression analysis between varimax rotated principal components and total minutes of physical activity per week (square root transformed)\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" align=\"\" width=\"707\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003eUnivariate analyses\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003eMultivariate model 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003eMultivariate model 2\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e\u0026beta;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026beta;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026beta;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003eP for trend\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003eHigh food, low FEHI, high walkability\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-0.29\u003c/p\u003e\n \u003cp\u003e-2.55\u003c/p\u003e\n \u003cp\u003e1.25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e-3.77, 3.18\u003c/p\u003e\n \u003cp\u003e-6.00, 0.90\u003c/p\u003e\n \u003cp\u003e-2.23, 4.73\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e-0.46\u003c/p\u003e\n \u003cp\u003e0.08\u003c/p\u003e\n \u003cp\u003e3.55\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-3.90, 2.98\u003c/p\u003e\n \u003cp\u003e-3.53, 3.69\u003c/p\u003e\n \u003cp\u003e-1.18, 8.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e-0.60\u003c/p\u003e\n \u003cp\u003e0.60\u003c/p\u003e\n \u003cp\u003e4.43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-4.27, 3.08\u003c/p\u003e\n \u003cp\u003e-3.86, 5.05\u003c/p\u003e\n \u003cp\u003e-1.33, 10.19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.23\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003eRelative food environment advantage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-1.75\u003c/p\u003e\n \u003cp\u003e-1.96\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e-3.85*\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e-5.18, 1.68\u003c/p\u003e\n \u003cp\u003e-5.48, 1.56\u003c/p\u003e\n \u003cp\u003e-7.44, -0.26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e-0.20\u003c/p\u003e\n \u003cp\u003e-0.04\u003c/p\u003e\n \u003cp\u003e-1.32\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-3.55, 3.14\u003c/p\u003e\n \u003cp\u003e-3.48, 3.41\u003c/p\u003e\n \u003cp\u003e-4.97, 2.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e0.23\u003c/p\u003e\n \u003cp\u003e1.22\u003c/p\u003e\n \u003cp\u003e-0.61\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-3.21, 3.66\u003c/p\u003e\n \u003cp\u003e-2.50, 4.95\u003c/p\u003e\n \u003cp\u003e-4.38, 3.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.80\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003eLow density and high distance to accessible green\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-1.93\u003c/p\u003e\n \u003cp\u003e-1.08\u003c/p\u003e\n \u003cp\u003e-0.96\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e-5.40, 1.55\u003c/p\u003e\n \u003cp\u003e-4.54, 2.39\u003c/p\u003e\n \u003cp\u003e-4.52, 2.59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e-2.23\u003c/p\u003e\n \u003cp\u003e-1.40\u003c/p\u003e\n \u003cp\u003e-2.45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-5.59, 1.13\u003c/p\u003e\n \u003cp\u003e-4.69, 1.90\u003c/p\u003e\n \u003cp\u003e-5.87, 0.97\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e-1.56\u003c/p\u003e\n \u003cp\u003e-1.12\u003c/p\u003e\n \u003cp\u003e-1.37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-5.18, 2.07\u003c/p\u003e\n \u003cp\u003e-4.47, 2.23\u003c/p\u003e\n \u003cp\u003e-5.19, 2.45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.52\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003eLow-quality, agricultural/inaccessible green\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.36\u003c/p\u003e\n \u003cp\u003e0.72\u003c/p\u003e\n \u003cp\u003e3.19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 83px;\"\u003e\n \u003cp\u003e-3.14, 3.85\u003c/p\u003e\n \u003cp\u003e-2.78, 4.21\u003c/p\u003e\n \u003cp\u003e-0.28, 6.66\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e-0.94\u003c/p\u003e\n \u003cp\u003e-2.31\u003c/p\u003e\n \u003cp\u003e1.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-4.45, 2.57\u003c/p\u003e\n \u003cp\u003e-5.93, 1.30\u003c/p\u003e\n \u003cp\u003e-2.15, 5.74\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e-0.83\u003c/p\u003e\n \u003cp\u003e-1.76\u003c/p\u003e\n \u003cp\u003e2.32\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-5.13, 3.48\u003c/p\u003e\n \u003cp\u003e-6.39, 2.87\u003c/p\u003e\n \u003cp\u003e-2.79, 7.42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.26\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003e*p\u0026lt;0.05. Multivariate model 1 included the individual principal components per model corrected for age, sex, educational level, migration background, SES-WOA score of 4-digit postal code and population density. Multivariate model 2 included all principal components corrected for the aforementioned confounders.\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAdditional Outcomes and Sensitivity Analyses\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAnalyses of additional health behaviours are presented in Additional File 2.\u003c/p\u003e\n\u003cp\u003eAnalysis of the fruit and vegetable subcomponent of the DHD revealed a significant association with the \u0026ldquo;\u003cem\u003eRelative food environment advantage\u0026rdquo;\u003c/em\u003e component. In both multivariate models, participants in the highest quartile had significantly lower fruit and vegetable diet quality scores compared with those in the lowest quartile (model 2: \u0026beta;=-1.70, 95% CI [-3.10, -0.31], P-trend=0.01). No significant associations were observed between fruit and vegetable intake and the other three PCs \u0026nbsp;(Table S1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFor the unhealthy food choices subcomponent, the \u0026ldquo;\u003cem\u003eLow density and high distance to accessible green\u003c/em\u003e\u0026rdquo; component showed a significant association in univariate and model 1 (model 1: OR=0.56, 95% CI [0.33, 0.96]) analyses. However, the association was no longer statistically significant after full adjustment in model 2. No other PCs showed a significant association with unhealthy food choices (Table S2).\u003c/p\u003e\n\u003cp\u003eBinary analysis of meeting the physical activity guideline (\u0026ge;150 minutes of moderate to vigorous activity per week) revealed a significant univariate association for the \u0026ldquo;\u003cem\u003eLow quality, agricultural/inaccessible green\u003c/em\u003e\u0026rdquo; component. Participants in Q3 had higher odds of meeting the guidelines compared with Q1 (OR=1.72, 95% CI [1.01, 2.90]). However, this association was no longer significant after adjustment for confounders in models 1 and 2. No other PCs showed significant associations with meeting the physical activity guidelines in any model (Table S3).\u003c/p\u003e\n\u003cp\u003eEffect modification for SES-WOA score and the PCs was tested by including an interaction term in model 2 of the multinomial logistic regression model for weight status and the linear models for total DHD-score and total minutes of weekly physical activity (Table S4-S6). In none of the models was the interaction term statistically significant, except for the interaction between the \u0026ldquo;\u003cem\u003eRelative food environment advantage\u003c/em\u003e\u0026rdquo; component and neighbourhood socioeconomic position for obesity (P\u0026lt;0.02), though not for overweight. However, the neighbourhood socioeconomic position main effect showed an extremely low OR likely reflecting model instability due to sparse data (only 11 obese participants in highest tertile of SES-WOA).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cdiv id=\"Sec27\" class=\"Section2\"\u003e \u003ch2\u003eExamination of Environmental Patterns\u003c/h2\u003e \u003cp\u003eThe correlation analyses and PCA revealed several important patterns in how neighbourhood environmental characteristics cluster together in urban areas, with implications for environmental health research.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec28\" class=\"Section2\"\u003e \u003ch2\u003eFood environment\u003c/h2\u003e \u003cp\u003eAn important finding was the very strong positive correlation between food retailer density and population density. Crucially, this strong correlation persisted even within the \"very strongly urban\" category (\u0026gt;\u0026thinsp;2,500 addresses per km\u0026sup2;). This implies that simply adjusting for urbanicity category (as frequently done in environmental health research) might not adequately control for population density-related confounding. Previous research in compact Dutch urban settings has shown that population density itself is associated with multiple health-relevant pathways, underscoring its role as a broader structuring factor of urban environments. [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e] Our approach of adjusting for continuous population density addresses this limitation, but introduces a different concern: population density correlated very strongly with PC1 (r\u0026thinsp;=\u0026thinsp;0.86), meaning that adjusting for it may have removed much of the very variation this component captures. This raises a fundamental issue: failing to adjust for population density risks confounding, yet adjusting for it may obscure real associations between the food environment and health outcomes that operate precisely through the urban density gradient.\u003c/p\u003e \u003cp\u003eDensities of core and non-core food retailers were very highly correlated (ρ\u0026thinsp;\u0026gt;\u0026thinsp;0.9), reflecting the commercial clustering typical of urban environments. Additionally, relative and absolute food environment measures loaded on separate principal components and showed weak correlations with each other. This independence has been noted previously by Pinho et. al. [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e] and indicates these measures capture fundamentally different dimensions of the food environment. Notably, PC2 (relative measures) showed no significant correlation with population density, while PC1 (absolute measures) remained strongly correlated. This confirms that relative measures capture food environment quality independent of the urbanicity gradient.\u003c/p\u003e \u003cp\u003eHowever, even though calls have been made to use relative measures more widely [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e], this independence also presents interpretive challenges. Identical relative scores can represent fundamentally different food environments depending on absolute availability. For example, a neighbourhood with 2 supermarkets and 1 fast-food outlet (67% healthy) scores better than one with 20 supermarkets and 30 fast-food outlets (40% healthy), despite the latter offering far more absolute healthy options. Moreover, as previously described [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e], ratio measures can be volatile in low-density areas, where a single retailer opening or closing can shift the ratio by 10 percentage points, while the same change in a dense area produces less than a 1-percentage-point shift. Future research should consider: restricting analyses to areas exceeding minimum retail density thresholds [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e] and/or stratifying by absolute density (not urbanicity) to test whether associations are consistent across contexts.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec29\" class=\"Section2\"\u003e \u003ch2\u003eGreen space\u003c/h2\u003e \u003cp\u003eOur analysis revealed substantial heterogeneity in green space measures. NDVI and other, mostly agricultural, green space loaded together (PC4) but showed weak correlations with accessible green spaces (PC3), reflecting that these measures capture fundamentally different constructs. NDVI primarily captures vegetation exposure correlating with larger agricultural land and might be more relevant to air quality and heat mitigation, while accessible green spaces capture opportunities for physical activity and social interaction. [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e] Besides, green space showed expected inverse relationships with population density. Relationships with neighbourhood socioeconomic position were weaker and this finding aligns with previous Dutch research showing that socioeconomic differences in greenness are smaller in highly urbanised municipalities. [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eScale and Spatial Considerations\u003c/h3\u003e\n\u003cp\u003eOur data showed high correlations between environmental characteristics at 300m and 300-1000m scales. Neighbourhoods with high retail or green space density immediately around residences (300m) generally also have high density in their extended surroundings (300-1000m donut buffer). However, an important exception emerged. Relative food environment measures showed lower cross-scale correlations (ρ\u0026thinsp;\u0026asymp;\u0026thinsp;0.6), suggesting that the relative composition of healthy versus unhealthy food retailers can vary between immediate and extended neighbourhoods. This has implications for mechanism: if the relative food environment primarily influences health through daily shopping patterns, immediate neighbourhood exposure (300m, walkable distance) may matter more than extended area composition. Conversely, if mechanisms operate through neighbourhood norms or perceived availability as previously proposed [\u003cspan additionalcitationids=\"CR44\" citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e], extended area characteristics might be more relevant.\u003c/p\u003e \u003cdiv id=\"Sec31\" class=\"Section2\"\u003e \u003ch2\u003eAssociation with Weight Status and Lifestyle Behaviours\u003c/h2\u003e \u003cp\u003eDespite interesting environmental variation captured by the four principal components, associations with weight status and lifestyle behaviours were limited and showed mostly inconsistent patterns.\u003c/p\u003e \u003cp\u003eThe \"\u003cem\u003eRelative food environment advantage\u003c/em\u003e\" component provided one of the more robust findings. This is consistent with a systematic review by Cobb. et. al. which noted that studies utilizing relative measures are more likely to see significant and expected results with weight status than absolute measures. [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e] This component showed consistent associations with lower odds of being overweight across all models for the third quartile (univariate, multivariate model 1, and fully adjusted model 2). Similar protective associations were observed for obesity in univariate and model 1 analyses, though these attenuated in the fully adjusted model. However, no clear dose-response pattern emerged (P-trend\u0026thinsp;=\u0026thinsp;0.19), suggesting potential threshold effects or that these associations may be influenced by other unmeasured neighbourhood characteristics. This finding is complicated by a counterintuitive pattern: higher levels of this same component were associated with lower fruit and vegetable intake in the highest quartile (including a significant trend). A systematic review by Turner et. al., studying the association between the retail environment and fruit and vegetable consumption, found mixed results. [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e] However, most of the included studies focused on absolute measures. Previous studies investigating the association between relative measures and fruit/vegetable intake in the general population found results in the opposite direction, i.e. reporting that individuals living in areas with a higher proportion of healthier food stores reported greater fruit and vegetable consumption. [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e] A possible explanation could be that high access to supermarkets may reduce the need for detailed planning of grocery trips, compared to those with lower access. For example, research has previously identified planning is one of the factors that influence the translation of intentions into actual behaviour related to fruit and vegetable consumption. [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]\u003c/p\u003e \u003cp\u003eFor the other principal components, associations were weak or inconsistent. The \"\u003cem\u003eHigh food, low FEHI, high walkability\u003c/em\u003e\" component showed a univariate association with obesity that disappeared after adjustment for confounders, suggesting this pattern primarily reflects broader neighbourhood disadvantage rather than specific environmental effects. No consistent associations emerged for the green space components (PC3, PC4) with any outcome, nor were meaningful associations observed for total physical activity. The few significant findings in supplementary analyses of dietary and physical activity subcomponents similarly failed to show consistent patterns or dose-response relationships and did not survive full adjustment for confounders. The inconsistent results might be due to chance, residual confounding and/or limited sample size. Moreover, the limited associations observed may also reflect characteristics specific to high-risk populations. Individuals already at high cardiovascular risk may have reduced environmental sensitivity due to established dietary patterns that are less responsive to neighbourhood influences. Additionally, high-risk patients in our study were engaged with healthcare providers who provided dietary guidance as part of standard Dutch primary care protocols, potentially creating alternative pathways to healthy behaviours that are independent of neighbourhood characteristics. Further large-scale confirmatory research is needed to explore these findings more robustly.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec32\" class=\"Section2\"\u003e \u003ch2\u003eEffect Modification\u003c/h2\u003e \u003cp\u003eWe found no evidence that associations between neighbourhood characteristics and health outcomes varied by neighbourhood socioeconomic position, except for a significant interaction between the \"Relative food environment advantage\" component and neighbourhood socioeconomic position for obesity, suggesting that food environment characteristics may be more strongly associated with obesity in neighbourhoods with a lower socioeconomic position. This finding is consistent with recent Dutch evidence. [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e] However, model instability was observed due to insufficient data within strata, necessitating replication in larger samples before drawing substantive conclusions.\u003c/p\u003e \u003cdiv id=\"Sec33\" class=\"Section3\"\u003e \u003ch2\u003eStrengths and limitations\u003c/h2\u003e \u003cp\u003eA major strength of this study was the comprehensive assessment of the neighbourhood environment, including 23 indicators across food environment, green space, and walkability domains. By applying PCA, we were able to reduce multicollinearity and explore how clusters of co-occurring neighbourhood characteristics relate to BMI, diet quality, and physical activity. This data-driven approach revealed distinct environmental patterns and their intercorrelations, providing insights into how neighbourhood characteristics cluster in dense urban settings. Furthermore, to our knowledge, this is the first study that investigates the association between neighbourhood characteristics and BMI and lifestyle behaviour in a primary care population at high risk for CVD. While the identified environmental patterns likely reflect general characteristics of the urban landscape applicable across populations, the associations with health behaviours were examined specifically in this high-risk group. Whether environmental patterns play a similar role in the general population remains an open question. As previously mentioned, high-risk populations may differ from the general populations in important ways which could modify how they interact with their neighbourhood environment. Nonetheless, studying this population offers preliminary insights relevant for population health management and risk stratification in primary care settings.\u003c/p\u003e \u003cp\u003eSeveral limitations should be noted. First, the cross-sectional exploratory design precludes any conclusions regarding causation. Additionally, the large number of statistical tests conducted across multiple outcomes and components increases the probability of chance findings. No correction for multiple testing was applied, consistent with the exploratory aims of this study; however, this means that individual significant associations should be interpreted with caution. Second, despite adjusting for several neighbourhood confounders, residual confounding can remain. Although we adjusted for continuous population density, density was strongly tied to several environmental characteristics. This reflects that, in dense urban settings, many neighbourhood features are inherently structured by the density gradient itself. Consequently, the influence of specific environmental patterns cannot be fully separated from the broader high-density urban context.\u003c/p\u003e \u003cp\u003eThe third limitation is that due to privacy considerations, only 4-digit postal code of the participants was available, exposure results are less sensitive than when address level data would be available. The fourth limitation is that outcomes were measured by self-report questionnaires, which can lead to participants giving more socially acceptable answers that may lead to an underestimation of the observed associations.\u003c/p\u003e \u003cp\u003eLastly, the PCA was conducted using neighbourhood data linked to the residential postal codes of participants in our sample, rather than based on all neighbourhoods in the city The Hague. While in theory this limits the representativeness of the identified environmental patterns for the entire urban area, the included postal codes cover a very large and diverse portion of the Hague (79% of all postal codes in the Hague) and directly adjacent municipalities, capturing substantial environmental variation and sociodemographic backgrounds.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThis study identified four distinct neighbourhood environmental patterns in a dense urban area in the Netherlands. The patterns reflected systematic co-occurrence of food environment, walkability, and green space characteristics, with population density emerging as a key underlying structuring feature. Correlation analyses confirmed that population density remained strongly tied to multiple environmental indicators, with particularly strong correlations for food retailer densities. This strong correlation persisted even within very strongly urban areas, demonstrating methodological challenges for isolating independent effects of food environment on health behaviours using conventional observational approaches.\u003c/p\u003e \u003cp\u003eOf the identified patterns, only one, the \u0026ldquo;\u003cem\u003eRelative food environment advantage\u003c/em\u003e\u0026rdquo; pattern, showed a more robust association with a lower odds of being overweight. However, this association was modest, showed no consistent dose\u0026ndash;response, and was accompanied by a paradoxical association with lower fruit and vegetable dietary quality. No robust associations were observed for physical activity or other dietary outcomes. There was some indication that the relative food environment may be more strongly associated with obesity in neighbourhoods with lower socioeconomic position, suggesting possible effect modification that warrants further exploration.\u003c/p\u003e \u003cp\u003eThese findings highlight the importance of examining neighbourhood characteristics as interrelated patterns rather than isolated exposures, especially in highly urbanised contexts. Larger, confirmatory studies, using pattern-based approaches, are needed to investigate the preliminary associations with health outcomes and assess their relevance for population health management.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eBAG\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eBasisregistratie Adressen en Gebouwen (Registration of Addresses and Buildings)\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eBMI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eBody mass index\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCBS\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eCentraal Bureau voor de Statistiek (Statistics Netherlands)\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eConfidence interval\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCVD\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eCardiovascular disease\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eDHD-FFQ\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eDutch Healthy Diet Food Frequency Questionnaire\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eDHD-index\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eDutch Healthy Diet index\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eFEHI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eFood Environment Healthiness Index\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eGECCO\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eGeoscience and Health Cohort Consortium\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eGP\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eGeneral practitioner\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eICPC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eInternational Classification of Primary Care\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eIQR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eInterquartile range\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eKMO\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eKaiser-Meyer-Olkin\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eMRFEI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eModified Retail Food Environment Index\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eNDVI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eNormalised Difference Vegetation Index\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eNHG\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eNederlands Huisartsen Genootschap (Dutch College of General Practitioners)\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ePC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ePrincipal component\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ePCA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ePrincipal component analysis\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eSD\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eStandard deviation\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eSEP\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eSocioeconomic position\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eSES-WOA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eSociaaleconomische status op basis van welvaart, opleidingsniveau en arbeidspositie\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eSPSS\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eStatistical Package for the Social Sciences\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eSQUASH\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eShort Questionnaire to Assess Health-enhancing physical activity\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eVIF\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eVariance inflation factor.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and informed consent\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Ethics Committee of Leiden University Medical Centre (P17.079) and conducted in accordance with the Declaration of Helsinki and Dutch medical research regulations (WMO).\u003c/p\u003e\n\u003cp\u003eAll participants provided written informed consent prior to participation.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData from the Healthy Heart study may be available from the corresponding author upon reasonable request, subject to ethics approval (Leiden UMC, P17.079). Environmental data were obtained from GECCO (walkability index, available upon request via https://www.gecco.nl) and Locatus (commercial data, available under licence). Green space data from the Dutch Land Use Database (CBS) and Landsat 8 satellite imagery (USGS) are publicly available.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no conflict of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements and funding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis publication\u0026nbsp;is part of the project \u0026quot;ECOTIP\u0026quot; (with project number NWA.1518.22.151) of the research programme \u0026quot;Dutch Research Agenda Routes by Consortia\u0026quot; (NWA-ORC 2022) which is financed by the Dutch Research Council (NWO).\u003c/p\u003e\n\u003cp\u003eThe Healthy Heart Study was sponsored by ZonMW grant number 531 001 203.\u003c/p\u003e\n\u003cp\u003eThe walkability index was collected as part of the Geoscience and Health Cohort Consortium (GECCO), which was financially supported by the Netherlands Organisation for Scientific Research (NWO), the Netherlands Organisation for Health Research and Development (ZonMw), and Amsterdam UMC. More information on GECCO can be found on www.gecco.nl\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData protection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data were handled according to GDPR and Dutch privacy legislation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAI Use Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDuring the preparation of this work the authors used Claude (Anthropic) in order to perform language editing, including spelling and grammar checks, and to assist with word count reduction. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA.S., R.V., L.vdV. and J.K. wrote the draft of the manuscript. T.B. and M.N. were involved in the design and execution of the study. A.S., M.H. and J.K. were involved in data curation, formal analysis and methodology. A.S., R.V., L.vdV., M.H., M.B. T.B., M.N. and J.K. critically revised and approved the final manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eWorld Health Organization. Cardiovascular diseases (CVDs) [Internet]. [cited 2025 May 23]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.who.int/en/news-room/fact-sheets/detail/cardiovascular-diseases\u003c/span\u003e\u003cspan address=\"https://www.who.int/en/news-room/fact-sheets/detail/cardiovascular-diseases\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDixon BN, Ugwoaba UA, Brockmann AN, Ross KM. Associations between the built environment and dietary intake, physical activity, and obesity: A scoping review of reviews. Obes Rev. 2021;22:e13171. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/obr.13171\u003c/span\u003e\u003cspan address=\"10.1111/obr.13171\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCobb LK, Appel LJ, Franco M, Jones-Smith JC, Nur A, Anderson CAM. The relationship of the local food environment with obesity: A systematic review of methods, study quality, and results. Obesity. 2015;23(7):1331\u0026ndash;44.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMalambo P, Kengne AP, De Villiers A, Lambert EV, Puoane T. Built environment, selected risk factors and major cardiovascular disease outcomes: A systematic review. PLoS ONE 2016;11(11).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNguyen PY, Astell-Burt T, Rahimi-Ardabili H, Feng X. Green space quality and health: A systematic review. Int J Environ Res Public Health. 2021;18(21):11028.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLytle LA, Sokol RL. Measures of the food environment: A systematic review of the field, 2007\u0026ndash;2015. Health Place. 2017;44:18\u0026ndash;34.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeijer P, Numans H, Lakerveld J. Associations between the neighbourhood food environment and cardiovascular disease: a systematic review. Eur J Prev Cardiol. 2023;00:1\u0026ndash;11.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeal C, Bean K, Thomas F, Chaix B. Multicollinearity in Associations Between Multiple Environmental Features and Body Weight and Abdominal Fat: Using Matching Techniques to Assess Whether the Associations are Separable. Am J Epidemiol. 2012;175(11):1152\u0026ndash;62.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFeuillet T, Charreire H, Roda C, Ben Rebah M, Mackenbach JD, Compernolle S, et al. Neighbourhood typology based on virtual audit of environmental obesogenic characteristics. Obes Rev. 2016;17:19\u0026ndash;30.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMolina-Garc\u0026iacute;a J, Queralt A, Adams MA, Conway TL, Sallis JF. Neighborhood built environment and socio economic status in relation to multiple health outcomes in adolescents. Prev Med (Baltim). 2017;105:88\u0026ndash;94.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu XX, Ma XL, Huang WZ, Luo YN, He CJ, Zhong XM, et al. Green space and cardiovascular disease: A systematic review with meta-analysis. Environ Pollut. 2022;301:118990.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuo YN, Huang WZ, Liu XX, Markevych I, Bloom MS, Zhao T, et al. Greenspace with overweight and obesity: A systematic review and meta-analysis of epidemiological studies up to 2020. Obes Rev. 2020;21(11):e13078.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBernsdorf KA, B\u0026oslash;ggild H, Aadahl M, Toft U. Measuring associations between the food environment and dietary habits: comparing the proportion and density of food outlets. BMC Public Health. 2024;24(1):1\u0026ndash;14.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaguire ER, Burgoine T, Penney TL, Forouhi NG, Monsivais P. Does exposure to the food environment differ by socioeconomic position? Comparing area-based and person-centred metrics in the Fenland Study, UK. Int J Health Geogr. 2017;16(1):1\u0026ndash;14.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlompmaker JO, Hoek G, Bloemsma LD, Gehring U, Strak M, Wijga AH, et al. Green space definition affects associations of green space with overweight and physical activity. Environ Res. 2018;160:531\u0026ndash;40.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMackenbach JP, Lingsma HF, Van Ravesteyn NT, Kamphuis CBM. The population and high-risk approaches to prevention: quantitative estimates of their contribution to population health in the Netherlands, 1970\u0026ndash;2010. Eur J Public Health. 2013;23(6):909\u0026ndash;15.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRose G. The Strategy of Preventive Medicine. Oxford University Press; 1992.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBonten TN, Verkleij SM, van der Kleij RM, Busch K, Hout WB, van den, Chavannes NH, et al. Selective prevention of cardiovascular disease using integrated lifestyle intervention in primary care: protocol of the Healthy Heart stepped-wedge trial. BMJ Open. 2021;11(7):e043829.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBAG producten -. Kadaster.nl zakelijk [Internet]. [cited 2024 Jun 13]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.kadaster.nl/zakelijk/registraties/basisregistraties/bag/bag-producten\u003c/span\u003e\u003cspan address=\"https://www.kadaster.nl/zakelijk/registraties/basisregistraties/bag/bag-producten\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHome. - Locatus [Internet]. [cited 2024 Jun 13]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://locatus.com/en/\u003c/span\u003e\u003cspan address=\"https://locatus.com/en/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePoelman MP, Nicolaou M, DIjkstra SC, MacKenbach JD, Lu M, Karssenberg D, et al. Does the neighbourhood food environment contribute to ethnic differences in diet quality? Results from the HELIUS study in Amsterdam, the Netherlands. Public Health Nutr. 2021;24(15):5101\u0026ndash;12.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePinho MGM, Mackenbach JD, Oppert JM, Charreire H, B\u0026aacute;rdos H, Rutter H, et al. Exploring absolute and relative measures of exposure to food environments in relation to dietary patterns among European adults. Public Health Nutr. 2019;22(6):1037\u0026ndash;47.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCenters for Disease Control and Prevention. (2011) Census Tract Level State Maps of the Modified Retail Food Environment Index (mRFEI). Children\u0026rsquo;s Food Environment State Indicator Report, 2011. Atlanta, GA: CDC.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBestand bodemgebruik | CBS [Internet]. [cited 2024 Nov 28]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cbs.nl/nlnl/dossier/nederland-regionaal/geografische-data/bestand-bodemgebruik\u003c/span\u003e\u003cspan address=\"https://www.cbs.nl/nlnl/dossier/nederland-regionaal/geografische-data/bestand-bodemgebruik\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEarthExplorer [Internet]. [cited 2024 Nov 28]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://earthexplorer.usgs.gov/\u003c/span\u003e\u003cspan address=\"https://earthexplorer.usgs.gov/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLam TM, Wang Z, Vaartjes I, Karssenberg D, Ettema D, Helbich M, et al. Development of an objectively measured walkability index for the Netherlands. Int J Behav Nutr Phys Act. 2022;19(1):1\u0026ndash;16.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLakerveld J, Wagtendonk A, Vaartjes I, Karssenberg D, Penninx B, Beulens J et al. Deep phenotyping meets big data: the Geoscience and hEalth Cohort COnsortium (GECCO) data to enable exposome studies in The Netherlands. Int J Health Geogr. 2020;19(1).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTimmermans EJ, Lakerveld J, Beulens JWJ, Boomsma DI, Kramer SE, Oosterman M et al. Cohort profile: the Geoscience and Health Cohort Consortium (GECCO) in the Netherlands. BMJ Open. 2018;8(6).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStedelijk gebied | CBS [Internet]. [cited 2025 May 23]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cbs.nl/nl-nl/onze diensten/methoden/begrippen/stedelijk-gebied\u003c/span\u003e\u003cspan address=\"https://www.cbs.nl/nl-nl/onze diensten/methoden/begrippen/stedelijk-gebied\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeswoa. Sociaaleconomische statusscores huishoudens | CBS [Internet]. [cited 2024 Apr 26]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cbs.nl/nl-nl/onze-diensten/maatwerk-en-microdata/microdata-zelf-onderzoek doen/microdatabestanden/seswoa-sociaaleconomische-statusscores-huishoudens\u003c/span\u003e\u003cspan address=\"https://www.cbs.nl/nl-nl/onze-diensten/maatwerk-en-microdata/microdata-zelf-onderzoek doen/microdatabestanden/seswoa-sociaaleconomische-statusscores-huishoudens\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNHG-Standaard Cardiovasculair risicomanagement [internet]. [cited 2025 May 23]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://richtlijnen.nhg.org/standaarden/cardiovasculair-risicomanagement\u003c/span\u003e\u003cspan address=\"https://richtlijnen.nhg.org/standaarden/cardiovasculair-risicomanagement\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLooman M, Feskens EJM, De Rijk M, Meijboom S, Biesbroek S, Temme EHM, et al. Development and evaluation of the Dutch Healthy Diet index 2015. Public Health Nutr. 2017;20(13):2289\u0026ndash;99.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Lee L, Feskens EJM, Meijboom S, Van Hooft EJC, Van \u0026rsquo;T, Veer P, De Vries JHM, et al. Evaluation of a screener to assess diet quality in the Netherlands. Br J Nutr. 2016;115(3):517\u0026ndash;26.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWendel-Vos GCW, Schuit AJ, Saris WHM, Kromhout D. Reproducibility and relative validity of the short questionnaire to assess health-enhancing physical activity. J Clin Epidemiol. 2003;56(12):1163\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBMI berekenen |. Voedingscentrum [Internet]. [cited 2024 Jun 24]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.voedingscentrum.nl/nl/bmi-meter.aspx#70jaar\u003c/span\u003e\u003cspan address=\"https://www.voedingscentrum.nl/nl/bmi-meter.aspx#70jaar\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIchikawa M, Hosono A, Tamai Y, Watanabe M, Shibata K, Tsujimura S, et al. Handling missing data in an FFQ: multiple imputation and nutrient intake estimates. Public Health Nutr. 2019;22(8):1351.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeenackers MA, Oude Groeniger J, Kamphuis CBM, Van Lenthe FJ. Urban population density and mortality in a compact Dutch city: 23-year follow-up of the Dutch GLOBE study. Health Place. 2018;53:79\u0026ndash;85.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMason KE, Bentley RJ, Kavanagh AM. Fruit and vegetable purchasing and the relative density of healthy and unhealthy food stores: evidence from an Australian multilevel study. J Epidemiol Community Health. 2013;67(3):231\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClary CM, Ramos Y, Shareck M, Kestens Y. Should we use absolute or relative measures when assessing foodscape exposure in relation to fruit and vegetable intake? Evidence from a wide-scale Canadian study. Prev Med (Baltim). 2015;71:83\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThornton LE, Lamb KE, White SR. The use and misuse of ratio and proportion exposure measures in food environment research. Int J Behav Nutr Phys Act. 2020;17(1):118.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarkevych I, Schoierer J, Hartig T, Chudnovsky A, Hystad P, Dzhambov AM, et al. Exploring pathways linking greenspace to health: Theoretical and methodological guidance. Environ Res. 2017;158:301\u0026ndash;17.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Vries S, Buijs AE, Snep RPH. Environmental Justice in The Netherlands: Presence and Quality of Greenspace Differ by Socioeconomic Status of Neighbourhoods. Sustainability 2020, Vol 12, Page 5889. 2020;12(15):5889.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFlint E, Cummins S, Matthews S. Do perceptions of the neighbourhood food environment predict fruit and vegetable intake in low-income neighbourhoods? Health Place. 2013;24:11\u0026ndash;5.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Rongen S, Poelman MP, Thornton L, Abbott G, Lu M, Kamphuis CBM, et al. Neighbourhood fast food exposure and consumption: The mediating role of neighbourhood social norms. Int J Beh Nutr Phys Act. 2020;17(1):61.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTak NI, J te Velde SJ, Kamphuis CB, Ball K, Crawford D, Brug J, et al. Associations between neighbourhood and household environmental variables and fruit consumption: exploration of mediation by individual cognitions and habit strength in the GLOBE study. Public Health Nutr. 2013;16(3):505\u0026ndash;14.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTurner G, Green R, Alae-Carew C, Dangour AD. The association of dimensions of fruit and vegetable access in the retail food environment with consumption; a systematic review. Glob Food Sec. 2021;29:100528.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcGuirt JT, Pitts SBJ, Gustafson A. Association between Spatial Access to Food Outlets, Frequency of Grocery Shopping, and Objectively-Assessed and Self-Reported Fruit and Vegetable Consumption. Nutrients. 2018;10(12):1974.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKreausukon P, Gellert P, Lippke S, Schwarzer R. Planning and self-efficacy can increase fruit and vegetable consumption: A randomized controlled trial. J Behav Med. 2012;35(4):443\u0026ndash;51.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Erpecum CPL, van Zon SKR, B\u0026uuml;ltmann U, Smidt N. The association between the presence of fast food outlets and BMI: the role of neighbourhood socio-economic status, healthy food outlets, and dietary factors. BMC Public Health. 2022;22(1):1\u0026ndash;12.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-public-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pubh","sideBox":"Learn more about [BMC Public Health](http://bmcpublichealth.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/pubh/default.aspx","title":"BMC Public Health","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Built environment, food environment, green space, walkability, cardiovascular disease","lastPublishedDoi":"10.21203/rs.3.rs-9051455/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9051455/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eMost studies examine environmental determinants of health in isolation, neglecting the complex interplay between neighbourhood characteristics. Furthermore, research has primarily focused on general populations, leaving a gap in understanding how environmental factors relate to health behaviours in high-risk groups. This study aimed to (1) identify patterns of co-occurring neighbourhood characteristics related to the food environment, green space and walkability in a highly urbanised Dutch city; and (2) explore whether these neighbourhood patterns are associated with weight status, diet quality and physical activity among patients at high cardiovascular risk.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eThis cross-sectional study used baseline data from the Healthy Heart study in The Hague, Netherlands (N\u0026thinsp;=\u0026thinsp;475 participants from 73 postal codes). Twenty-three indicators across food environment, green space and walkability domains were assessed at two spatial scales. Correlation analyses examined intercorrelations among indicators and with contextual characteristics. Neighbourhood patterns were identified using principal component analysis (PCA), and associations with health outcomes were examined using multivariable regression, adjusted for individual and neighbourhood confounders.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eSeveral environmental indicators were strongly intercorrelated and closely linked to contextual neighbourhood characteristics, particularly population density and absolute food retailer densities, even within the most highly urbanised areas. PCA identified four neighbourhood patterns. Only the \u0026ldquo;Relative food environment advantage\u0026rdquo; pattern was robustly associated with lower odds of being overweight, but also counterintuitively with lower fruit and vegetable diet quality. There was some indication of effect modification by neighbourhood socioeconomic position for this pattern.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eNeighbourhood characteristics cluster into structured patterns in urban settings. Food environment measures remained coupled to population density even within high-density areas, fundamentally challenging independent effect estimation. Modest, paradoxical associations suggest limited environmental influence on lifestyle behaviours and weight in medically supervised high-risk populations.\u003c/p\u003e","manuscriptTitle":"Identifying Co-occurring Neighbourhood Environmental Patterns and Their Association with Health Behaviours in a Dutch Urban Population at High Cardiometabolic Risk","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-02 08:44:04","doi":"10.21203/rs.3.rs-9051455/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-21T04:14:10+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-21T01:04:03+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-20T09:10:14+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"106386810051619256217129185663293710785","date":"2026-04-01T22:30:40+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"248050627379299152941968683804664803097","date":"2026-03-31T14:50:42+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"49926614108431785926346709752907295380","date":"2026-03-31T07:32:04+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-29T02:30:09+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-13T06:18:41+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-12T06:32:17+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-12T06:32:15+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Public Health","date":"2026-03-06T13:59:57+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-public-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pubh","sideBox":"Learn more about [BMC Public Health](http://bmcpublichealth.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/pubh/default.aspx","title":"BMC Public Health","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"299a7bb5-b92a-4497-beb5-c635382f88d0","owner":[],"postedDate":"April 2nd, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-04-21T04:24:32+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-02 08:44:04","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9051455","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9051455","identity":"rs-9051455","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}