Observational Study Open Access
Copyright ©The Author(s) 2023. Published by Baishideng Publishing Group Inc. All rights reserved.
World J Gastroenterol. Jun 21, 2023; 29(23): 3688-3702
Published online Jun 21, 2023. doi: 10.3748/wjg.v29.i23.3688
Spatial cluster mapping and environmental modeling in pediatric inflammatory bowel disease
Mielle Michaux, Justin M Chan, Kevan Jacobson, Department of Pediatrics, Division of Gastroenterology, Hepatology and Nutrition, Faculty of Medicine, British Columbia Children’s Hospital, University of British Columbia, Vancouver V6H 3V4, British Columbia, Canada
Mielle Michaux, Justin M Chan, Kevan Jacobson, British Columbia Children’s Hospital Research Institute, British Columbia Children’s Hospital, Vancouver V5Z 4H4, British Columbia, Canada
Luke Bergmann, Brian Klinkenberg, Department of Geography, University of British Columbia, Vancouver V6T 1Z2, British Columbia, Canada
Luis F Chaves, Department of Environmental and Occupational Health, School of Public Health, Indiana University, Bloomington, IN 47405, United States
Kevan Jacobson, Department of Cellular and Physiological Sciences, University of British Columbia, Vancouver V6T 1Z3, British Columbia, Canada
ORCID number: Mielle Michaux (0000-0003-4742-3126); Justin M Chan (0000-0002-8298-3594); Luis F Chaves (0000-0002-5301-2764); Brian Klinkenberg (0000-0002-3357-1446); Kevan Jacobson (0000-0001-7269-8557).
Author contributions: Michaux M participated in designing the study, acquisition, analysis, and interpretation of the data, and drafted the initial manuscript; Chan JM participated in designing the study, acquisition, analysis, and interpretation of the data, and revised the article critically for important intellectual content; Bergmann L and Chaves LF participated in the acquisition, analysis, and interpretation of the data, and revised the article critically for important intellectual content; Klinkenberg B participated in designing the study and revised the article critically for important intellectual content; Jacobson K was the guarantor and participated in designing the study, analysis, and interpretation of the data, and revised the article critically for important intellectual content.
Institutional review board statement: The study was reviewed and approved by the University of British Columbia Children’s and Women’s Research Ethics Board (Vancouver), No. H19-00739.
Conflict-of-interest statement: Dr. Jacobson reports other from BC Children’s Hospital Research Institute Clinician Scientist Awards Program Award, grants from Janssen, non-financial support from adMare Bioinnovations, other from Engene, outside the submitted work; and has served on the advisory boards of Janssen, AbbVie, Merck, Amgen, Mylan Inc, and McKesson.
Data sharing statement: Data is available upon reasonable request to the corresponding author subject to research ethics board approval, at kjacobson@cw.bc.ca.
STROBE statement: The authors have read the STROBE Statement—checklist of items, and the manuscript was prepared and revised according to the STROBE Statement—checklist of items.
Open-Access: This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: https://creativecommons.org/Licenses/by-nc/4.0/
Corresponding author: Kevan Jacobson, AGAF, FRCPC, MBChB, Professor, Senior Scientist, Department of Pediatrics, Division of Gastroenterology, Hepatology and Nutrition, Faculty of Medicine, British Columbia Children’s Hospital, University of British Columbia, 4480 Oak Street, Room K4-184, Vancouver V6H 3V4, British Columbia, Canada. kjacobson@cw.bc.ca
Received: February 7, 2023
Peer-review started: February 7, 2023
First decision: March 20, 2023
Revised: March 31, 2023
Accepted: April 23, 2023
Article in press: April 23, 2023
Published online: June 21, 2023
Processing time: 128 Days and 14.8 Hours

Abstract
BACKGROUND

Geographical (geospatial) clusters have been observed in inflammatory bowel disease (IBD) incidence and linked to environmental determinants of disease, but pediatric spatial patterns in North America are unknown. We hypothesized that we would identify geospatial clusters in the pediatric IBD (PIBD) population of British Columbia (BC), Canada and associate incidence with ethnicity and environmental exposures.

AIM

To identify PIBD clusters and model how spatial patterns are associated with population ethnicity and environmental exposures.

METHODS

One thousand one hundred eighty-three patients were included from a BC Children’s Hospital clinical registry who met the criteria of diagnosis with IBD ≤ age 16.9 from 2001–2016 with a valid postal code on file. A spatial cluster detection routine was used to identify areas with similar incidence. An ecological analysis employed Poisson rate models of IBD, Crohn’s disease (CD), and ulcerative colitis (UC) cases as functions of areal population ethnicity, rurality, average family size and income, average population exposure to green space, air pollution, and vitamin-D weighted ultraviolet light from the Canadian Environmental Health Research Consortium, and pesticide applications.

RESULTS

Hot spots (high incidence) were identified in Metro Vancouver (IBD, CD, UC), southern Okanagan regions (IBD, CD), and Vancouver Island (CD). Cold spots (low incidence) were identified in Southeastern BC (IBD, CD, UC), Northern BC (IBD, CD), and on BC’s coast (UC). No high incidence hot spots were detected in the densest urban areas. Modeling results were represented as incidence rate ratios (IRR) with 95%CI. Novel risk factors for PIBD included fine particulate matter (PM2.5) pollution (IRR = 1.294, CI = 1.113-1.507, P < 0.001) and agricultural application of petroleum oil to orchards and grapes (IRR = 1.135, CI = 1.007-1.270, P = 0.033). South Asian population (IRR = 1.020, CI = 1.011-1.028, P < 0.001) was a risk factor and Indigenous population (IRR = 0.956, CI = 0.941-0.971, P < 0.001), family size (IRR = 0.467, CI = 0.268-0.816, P = 0.007), and summer ultraviolet (IBD = 0.9993, CI = 0.9990–0.9996, P < 0.001) were protective factors as previously established. Novel risk factors for CD, as for PIBD, included: PM2.5 air pollution (IRR = 1.230, CI = 1 .056-1.435, P = 0.008) and agricultural petroleum oil (IRR = 1.159, CI = 1.002-1.326, P = 0.038). Indigenous population (IRR = 0.923, CI = 0.895–0.951, P < 0.001), as previously established, was a protective factor. For UC, rural population (UC IRR = 0.990, CI = 0.983-0.996, P = 0.004) was a protective factor and South Asian population (IRR = 1.054, CI = 1.030–1.079, P < 0.001) a risk factor as previously established.

CONCLUSION

PIBD spatial clusters were identified and associated with known and novel environmental determinants. The identification of agricultural pesticides and PM2.5 air pollution needs further study to validate these observations.

Key Words: Inflammatory bowel diseases; Crohn disease; Ulcerative colitis; Pesticides; Air pollution; South Asian people

Core Tip: Utilizing spatial mapping methodology, high and low incidence clusters of pediatric inflammatory bowel disease (IBD) were identified in British Columbia, Canada. Associating geographical location with IBD, rurality was negatively associated with ulcerative colitis. Notably, no high incidence hot spots were detected in the densest urban areas, suggesting unexplored urban protective factors. Novel risk factors for PIBD and specifically Crohn’s disease included fine particulate matter pollution and agricultural applications of petroleum oil to orchards and grapes. Spatial distribution was partially explained by rurality, population ethnicity, family size, pesticide applications, air pollution, ultraviolet exposure, and residential greenness.



INTRODUCTION

Canada has one of the world’s highest rates of pediatric inflammatory bowel disease (IBD), which includes Crohn’s disease (CD) and ulcerative colitis (UC)[1]. Incidence within countries can be quite varied, but local patterns of IBD incidence are unknown for most countries[1]. Results from studies in Finland, Norway, Northern France, and Manitoba, Canada suggest that IBD may have a clustered spatial distribution, but more research is necessary for understanding local spatial patterns[2-5]. To date, no study has focused on empirically detecting and evaluating disease clusters in Canadian or North American pediatric IBD (PIBD) populations. Notably, detecting local spatial clusters of IBD would allow for a better understanding of clinical populations and identify areas where services are needed. Moreover, spatial epidemiology can be used to identify novel environmental risk and protective factors. Although it is evident that environmental factors are important determinants for disease development, additional study is necessary.

A variety of population and environmental factors are thought to influence IBD risk. High incidence has been observed in Canadian populations of Jewish ethnicity and South Asian pediatric populations in British Columbia (BC), and low incidence has been reported in Canadian Indigenous communities[4,6,7]. Higher socioeconomic status has been associated with IBD[4,8]. A variety of environmental exposures, including rural residence, green space, ultraviolet (UV) radiation, and air pollution have also been studied[9-13]. No study known to us has examined pesticides as a potential determinant of PIBD. Pesticides can be present in food and the environment, and a variety of pesticides have been linked with changes to the gut microbiota which could have implications for IBD development[14]. Based on Canadian immigration and rural residence studies of IBD, exposures during early life appear to be important[9,15].

The aims of this exploratory study were (1) To determine spatial patterning of PIBD and identify location of disease hot and cold spots in the Canadian province of BC; and (2) to model the association between IBD case counts and population-level ethnicity, average income, rural residence, and known as well as novel environmental determinants. We hypothesized that we would identify IBD clusters that would be associated with ethnicity and environmental exposures. Modeling potential population risk factors provided context to the spatial analysis and helped identify areas where additional novel environmental risk or protective factors might have meaningfully affected disease incidence.

MATERIALS AND METHODS
Study area

BC is Canada’s westernmost province. It is divided into five Regional Health Authorities, which can be further subdivided into 89 Local Health Areas (LHAs) (see Figure 1 and Supplementary Figure 1). The majority of BC’s population live in urban areas located in the Vancouver Coastal and Fraser Health Authorities near the United States border[16]. Northern, Interior, and Vancouver Island Health serve largely more rural populations.

Figure 1
Figure 1  British Columbia's five Health Authorities and 89 Local Health Areas (black boundaries).
Participants

Patients for this study were selected from a clinical registry of IBD patients maintained by the BC Children’s Hospital (BCCH) Division of Gastroenterology, Hepatology and Nutrition who were diagnosed with or received care for IBD at BCCH in Vancouver[17]. Author Jacobson K is data steward for this registry. As BCCH is the only tertiary care pediatric institution with academic pediatric gastroenterologists in the province, it is where most children with IBD are diagnosed. Our recent study comparing PIBD incidence derived from the BCCH registry with incidence derived from population-wide provincial health administrative data between 1996 and 2008 found similar overall rates, particularly from 2001 onwards, reaffirming the validity of the registry as a reflection of population-based cases[18]. Notably, a small number of cases are diagnosed in the community, use health services from a different province or, in the case of older patients (> 16.9 years), are diagnosed by adult gastroenterology physicians. Registry patients were excluded from the study if they were diagnosed outside the study period (2001 to 2016) or over age 16.9 years, did not have a valid postal code on file, or had a postal code associated with BCCH as this was likely not their permanent address. Postal codes were cleaned to provide consistent formatting. See Table 1 for BC pediatric incidence, and Supplementary Table 1 for incidence by patient age. Patient six-digit postal code point location at diagnosis was associated with latitude and longitude coordinates from DMTI Spatial Inc. obtained from the Canadian Urban Environmental Health Research Consortium (CANUE)[19]. Incidence was pooled for the full study period to maximize the number of cases in each analysis.

Table 1 British Columbia average pediatric incidence for inflammatory bowel disease, Crohn’s disease, ulcerative colitis, and inflammatory bowel disease-unclassified from 2001–2016.
Diagnosis
Health authority
Cases
Incidence per 100000
95%CI for incidence
All IBDAll British Columbia11839.228.79.76
Fraser55810.9510.0611.9
Interior1437.156.038.43
Northern646.174.757.88
Vancouver Coastal2549.288.1710.49
Vancouver Island1648.367.139.74
CDAll British Columbia7806.085.666.52
Fraser3566.996.287.74
Interior964.83.895.85
Northern403.862.765.24
Vancouver Coastal1666.065.187.06
Vancouver Island1226.225.177.43
UCAll British Columbia2882.241.992.52
Fraser1512.962.513.48
Interior321.61.092.26
Northern161.540.882.51
Vancouver Coastal622.271.742.9
Vancouver Island271.380.912
IBD-UAll British Columbia1150.90.741.08
Fraser5110.751.32
Interior150.750.421.24
Northern80.770.331.52
Vancouver Coastal260.950.621.39
Vancouver Island150.760.431.26

This study was approved by the University of British Columbia Children's and Women's Research Ethics Board, No. H19-00739.

Methods

Guidelines from the REporting of studies Conducted using Observational Routinely-collected health Data statement were adapted for this ecological study. The statistical methods of this study were reviewed by biostatistician Jeffrey Bone from the BCCH Research Institute. Data is summarized in Table 2, with additional details on environmental exposures presented in the Supplementary materials. Ethnicity and family size (married or common-law spouses, single parents, and at least one child) data from the Canadian census were obtained at the Dissemination Area level and resampled to LHAs by population-weighted overlay[16,20,21]. Average family income for 2005 and 2015 was available for LHAs[22]. To approximate average environmental exposures for each LHA, census population age 0-19 was used to population-weight postal code exposure data in a process that captured 92.6% of BC’s youth population (excluded population resided in areas not covered by single link postal codes). Environmental data provided by CANUE for six digit postal codes included Normalized Difference Vegetation Index (NDVI)[19,23-26] greenness[27,28], vitamin D UV[19,29,30], nitrogen dioxide (NO2)[19,31,32], ozone (O3)[33-37], and fine particulate matter (PM2.5)[19,38]. Data on metam, petroleum oil, and glyphosate pesticides was resampled from Global Pesticide Grids[39,40]. The least populous LHA was excluded from regression modeling due to missing data.

Table 2 Data used in environmental exposure modeling.
Variable
Description
Data source and database
IncomeMean of 2005 (adjusted to 2015 dollars) and 2015 census average family income for Local Health AreasStatistics Canada, BC Community Health Atlas
EthnicityCensus percent of the population of South Asian, Indigenous, Chinese, Jewish, and Non-Jewish European ethnic originStatistics Canada, University of Toronto Computing in the Humanities and Social Sciences Data Centre
Family sizeCensus average family sizeStatistics Canada, University of Toronto Computing in the Humanities and Social Sciences Data Centre
RuralityPercent of residents in each Local Health Area who lived outside a census metropolitan area or census agglomerationStatistics Canada
NDVI greennessGreen vegetation cover calculated from land satellite imagery of surface reflection. Maximum and average growing season (May – August) NDVI within 250 m from postal code location. Multiplied by 100 so changes could be assessed as percentagesUnited States Geological Survey, Canadian Urban Environmental Health Research Consortium
UV vitamin DLong-term stable monthly mean daily vitamin D dose from solar UV radiation adjusted for postal code elevation (J/m-2) which was averaged for winter (December through February) and summer (June through August)Environment Canada and Cancer Care Ontario, Canadian Urban Environmental Health Research Consortium
NO2 air pollutionAnnual average NO2 concentration in parts per billionCanadian Urban Environmental Health Research Consortium
O3 air pollutionAverage of O3 taken from the highest rolling 8-hour daily average concentration (parts per billion)Environment and Climate Change Canada, Canadian Urban Environmental Health Research Consortium
PM2.5 air pollutionAnnual average surface PM2.5 (μg/m3) concentrationAtmospheric Composition Analysis Group, Canadian Urban Environmental Health Research Consortium
Metam pesticide2015 metam pesticide applied to fruits and vegetables (kg/ha/yr)Global Pesticide Grids, Version 1.01, based on the United States Geological Survey’s Pesticide National Synthesis Project and the Food and Agriculture Organization Corporate Statistical Database pesticide databases
Petroleum oil pesticide2015 petroleum oil employed for grapes and orchards (kg/ha/yr)Global Pesticide Grids, Version 1.01, based on the United States Geological Survey’s Pesticide National Synthesis Project and the Food and Agriculture Organization Corporate Statistical Database pesticide databases
Glyphosate pesticide2015 total glyphosate employed in major crops (wheat, corn, alfalfa, and others) in (kg/ha/yr)Global Pesticide Grids, Version 1.01, based on the United States Geological Survey’s Pesticide National Synthesis Project and the Food and Agriculture Organization Corporate Statistical Database pesticide databases
Population age 0-16.9Yearly estimates of population age 0-16.9 for Local Health Areas averaged for 2001–2016Statistics Canada, BC Stats
Spatial cluster analysis

Standardized incidence ratios (SIRs) for each of BC’s LHAs were calculated from incident registry cases using BC incidence as the reference rate[41]. DataBC provided geographic data[42]. To reduce the instability in disease rates caused by low case counts and populations, rates were averaged for the study period. In addition, spatial linear empirical Bayes estimation was used to smooth SIRs in areas with average pediatric populations under 10000[43]. The smoothing process calculated averages between a LHA’s SIR and the average SIR value of adjacent LHAs. Spatial relationships for smoothing and clustering were defined by direct adjacency (queen contiguity) to approximately model geographic connectivity.

Two forms of spatial analysis were used to identify spatial patterns in PIBD. The Global Moran’s I statistic measured the degree to which spatial patterning of IBD, CD, and UC in BC was clustered, dispersed, or had no detectable spatial pattern. The Local Moran’s I statistic was used as a Local Indicator of Spatial Association (LISA) to identify the location of hot spots (clusters of comparatively high SIRs) and cold spots (clusters of comparatively low SIRs) among LHAs[44]. To approximate the likelihood that a cluster would arise by chance, we used 999 Monte Carlo simulations to compute a pseudo-P value[44], which was then adjusted with a Holm correction for multiple comparisons. We chose LISA statistics over spatial and spatiotemporal SCAN methods given the irregular nature of the spatial areal units under analysis and the relative rarity of the disease, conditions under which LISA statistics have offered high sensitivity and specificity in cluster detection[45,46]. Data was aggregated over time (i.e. no spatio-temporal cluster analyses) given the aforementioned relative rarity of PIBD cases in BC.

Count regression modeling

Poisson Rate generalized linear models (PR-GLM) were used in an ecological analysis of LHAs to quantify the impact of population ethnicity, income, family size, rurality, air pollution, greenness, UV, and pesticide exposure on raw (unsmoothed) IBD, CD, and UC case counts. These models adjust for the population of each LHA and have an equal mean and variance. Model selection was based on the minimization of the Akaike Information Criterion using a stepwise algorithm combining backward elimination and forward addition[47]. PR-GLMs performed better than Negative Binomial, zero-inflated Poisson, and hurdle Poisson models. Considerations for reducing collinearity, conducting model selection, testing spatial independence of residuals, examining best model diagnostics, and considerations around the mapping of Incidence Rate Ratios (IRRs) are presented in the Supplementary materials. Multivariate models and high-quality Canadian census and exposure data were used to minimize sources of potential bias.

RESULTS
Spatial clustering results

The registry consisted of 1232 eligible patients diagnosed at or before age 16.9. Of these, 49 were associated with invalid or BCCH postal codes. In the spatial cluster modeling of the remaining 1183 PIBD patients, we observed a statistically significant (pseudo-P < 0.05) clustered distribution with a Moran’s I statistic of 0.65 (P = 0.001). For the 780 CD patients we observed a Moran’s I of 0.56 (P = 0.001) and for the 288 UC patients a Moran’s I of 0.28 (P = 0.001). See Figure 2 for local cluster locations. Hot spots were observed in the lower mainland, the main urban centre of BC, for IBD (P ≤ 0.031), CD (P ≤ 0.031), and UC (P ≤ 0.007), in the Okanagan region for IBD (P ≤ 0.029) and CD (P = 0.030), and on Vancouver Island for CD (P = 0.034). A cold spot cluster was detected in southeastern BC for IBD (P values < 0.001), CD (P ≤ 0.0499), and UC (P ≤ 0.040), in Northern BC for IBD (P = 0.026) and CD (P = 0.038), and on BC’s coast for UC (P = 0.036). We observed no LHA which had a significantly different trend in incidence than neighboring LHAs (spatial outliers) for IBD, CD, or UC.

Figure 2
Figure 2 Statistically significant standardized incidence ratio spatial clusters, 2001–2016. A: Inflammatory bowel disease; B: Crohn’s disease; C: Ulcerative colitis.
Environmental modeling results

Summary statistics for variables are presented in Table 3 for BC and Supplementary Table 2 by Health Authority (see Figure 1 for reference and Table 1 for incidence). Variables (Table 3) were modeled individually and combined in a multivariate model. Lack of spatial clustering observed in the residuals indicates that the model assumption of spatial independence was met, and suggests that within the parameters of the analysis there were either no important missing variables with spatial patterning or that the existing predictors capture similar spatial variability of important unmeasured variables. The final models included variables shown in Table 4 and logged average population as an offset variable.

Table 3 Summary statistics for variables of interest for the 88 local health areas used for modeling.
Explanatory variable
Minimum LHA value
Maximum LHA value
Mean of LHA values
Chinese ethnic origin (%)0.0047.214.35
Indigenous ethnic origin (%)1.0793.2214.38
Jewish ethnic origin (%)0.002.820.45
Non-Jewish European ethnic origin (%)10.6790.9870.91
South Asian ethnic origin (%)0.0031.192.71
Average family income ($)60265.01216769.6091884.26
Family size2.253.132.76
Population density (per square km)0.019443.44553.95
Rural population (%)0.00100.0053.05
NDVI maximum0.590.800.70
NDVI mean0.230.620.45
NO2 (ppb)0.1023.617.71
O3 (ppb)17.4939.5430.14
PM2.5 (μg m3)2.738.326.02
UV vitamin D summer (J/m-2)4488.807272.526053.15
UV vitamin D winter (J/m-2)102.02449.6274.98
Glyphosate used in common crops (kg/ha/yr)0.006.710.61
Glyphosate used in alfalfa crops (kg/ha/yr)0.000.170.02
Glyphosate used in corn crops (kg/ha/yr)0.006.060.49
Glyphosate used in wheat crops (kg/ha/yr)0.000.210.02
Metam used in fruits and vegetables (kg/ha/yr)0.008.820.60
Petroleum oil used in orchards and grapes (kg/ha/yr)0.004.750.28
Table 4 Incidence rate ratios and 95%CIs for the best model explaining inflammatory bowel disease, Crohn’s disease, and ulcerative colitis rates in local health areas of British Columbia, Canada (n = 88).
VariableInflammatory bowel disease
Crohn’s disease
Ulcerative colitis
IRR and 95%CI
P value
IRR and 95%CI
P value
IRR and 95%CI
P value
Average family size0.467 (0.268–0.816)0.007b
Average family income0.912 (0.827–1.006)0.064
Glyphosate used in alfalfa crops0.001 (0.000–0.253)0.024a
Glyphosate used in corn crops0.828 (0.707–0.974)0.021a
Glyphosate used in wheat crops121.196 (2.252–6671.461)0.019a
Indigenous ethnic origin0.956 (0.941–0.971)< 0.001c0.923 (0.895–0.951)< 0.001c
Metam used in fruits and vegetables0.931 (0.892–0.970)0.001b
NDVI Maximum at 250 m (× 100)1.060 (1.040–1.082)< 0.001c1.038 (1.014–1.062)0.002b
NDVI Mean at 250 m (× 100)1.043 (1.013–1.075)0.006b
NO20.952 (0.919–0.986)0.006b
O30.947 (0.909–0.987)0.010a1.095 (1.039–1.154)0.001b
Petroleum oil used in orchards and grapes1.135 (1.007–1.270)0.033a1.159 (1.002–1.326)0.038a
PM2.51.294 (1.113–1.507)0.001b1.230 (1.056–1.435)0.008b
Rural population0.990 (0.983–0.996)0.004b
South Asian ethnic origin1.020 (1.011–1.028)< 0.001c1.054 (1.030–1.079)< 0.001c
UV Summer0.9993 (0.9990–0.9996)< 0.001c0.9997 (0.9993–1.0001)0.091
UV Winter0.996 (0.992–1.000)0.067

For the whole PIBD population, Indigenous ethnic origin, average family size, summer UV radiation, and metam pesticide application were identified as significant protective factors. In contrast, South Asian origin, greenness, PM2.5, and petroleum oil application were significant risk factors. When broken down by disease subtype, Indigenous ethnic origin, NO2, and O3 were statistically significantly negatively correlated with CD, while maximum growing season greenness, PM2.5, and petroleum oil application were significant risk factors. Rurality, and alfalfa and corn glyphosate were statistically significant protective factors for UC, while significant risk factors included South Asian origin, mean growing season greenness, O3 pollution, and wheat glyphosate.

A 1% increase in the Indigenous population of a LHA was associated with a 4.4% decrease in the number of IBD cases (IRR = 0.956, P < 0.001) and a 7.7% decrease in CD (IRR = 0.923, P < 0.001), while a 20% increase in Indigenous population was associated with a 59% decrease in IBD. Each additional family member added to family size was associated with a 53.3% decrease in IBD cases (IRR = 0.467, P = 0.007). An increase of 1 J/m-2 in summer UV was associated with 0.001% decrease in IBD cases (IRR = 0.999, P < 0.001) while an increase of 500 J/m-2 was associated with a 29.4% decrease. A 1% increase in rural population was associated with a 1% decrease in UC (IRR = 0.990, P = 0.004). Each unit increase of metam pesticide was associated with a 6.9% decrease in IBD cases (IRR = 0.931, P = 0.001 for 1 kg/ha). Each ppb increase in O3 was associated with a 9.5% increase in UC (IRR = 1.095, P = 0.001) and a 5.3% decrease in CD (IRR = 9.47, P = 0.010). Each ppb increase in NO2 was associated with a 4.8% decrease in CD cases (IRR = 0.952, P = 0.006). Each unit increase of glyphosate applied to wheat was associated with an increase in UC (IRR = 121.196, P = 0.019 for 1 kg/ha), while the same pesticide applied to corn (IRR = 0.828, P = 0.021 for 1 kg/ha) and alfalfa (IRR = 0.001, P = 0.024 for 1 kg/ha) was a significant protective factor. It is important to note that glyphosate applications to corn (range 0.00–6.06, mean 0.49 kg/ha) were much higher than those to alfalfa (range 0.00–0.17, mean 0.02 kg/ha) and wheat (range 0.00–0.21, mean 0.02 kg/ha), and the extremely high and low IRRs for glyphosate are more reflective of the small data values than actual effect size.

A 1% increase in the percent of South Asian residents was associated with a 2% increase in IBD (IRR = 1.020, P < 0.001) and a 5.4% increase in UC (IRR = 1.054, P < 0.001), while a 20% increase in South Asian residents was associated with a 47.5% increase in IBD. A 1% increase in maximum growing season greenness was associated with a 6% increase in IBD and a 3.8% increase in CD (IRR = 1.060, P < 0.001, and IRR = 1.038, P = 0.002, respectively). A 1% increase in mean growing season greenness was associated with a 4.3% increase in UC (IRR = 1.043, P = 0.006). Each 1 μg/m3 concentration increase in PM2.5 air pollution was associated with a 29.4% increase in IBD cases (IRR = 1.294, P = 0.001) and a 23% increase in CD cases (IRR = 1.230, P = 0.008). Finally, a 1 kg/ha increase in the application of petroleum oil in grapes and orchards was associated with a 13.5% increase in IBD (IRR = 1.135, P = 0.033) and a 15.9% increase in CD (IRR = 1.159, P = 0.038).

DISCUSSION

Based on our exploratory analysis describing spatial patterning of PIBD in BC, incidence varied substantially across the province in both rural and urban areas, likely reflective of BC’s diverse population and environments. The spatial distribution of IBD, CD, and UC was significantly clustered during the period of 2001 to 2016, with substantial overlap between cluster locations for each of the three. The lower mainland hotspot and southeastern BC cold spot were consistent across the study period in the IBD population. As previously reported in PIBD, UC represented a lower percentage of IBD cases than CD and in addition, displayed a less clustered distribution with fewer shared clusters. It is unlikely that these observed clusters resulted from chance. Notably, patients living near the eastern border (Alberta) may receive care in that province, potentially explaining the cold spot in southeastern BC.

In our modeling of the association between IBD case counts and population and environment variables, higher proportions of South Asians tended to be associated with higher IBD and UC case counts. In Ontario, Canada, similar incidence of IBD has been observed for children of immigrants of South Asian origin born in Canada and children of non-immigrants[15]. However, higher rates of IBD have been documented in South Asian populations in BC, the United States, the United Kingdom, New Zealand, and Singapore and Malaysia[7,48-51]. Our observation of lower IBD and CD cases associated with higher proportions of Indigenous residents is consistent with previous Canadian studies[4,52]. When interpreting modeling results, locations where IBD rates were well predicted by population ethnicity should still be considered places where environmental risk or protective factors were also present. Larger family size has been associated with a protective effect on CD, which is consistent with our results for IBD but not CD[6]. Higher socioeconomic status is an established risk factor[4,8], but average income quantile was not a significant predictor in any model (though it improved CD model fit). While several identified hot spots were located in the highest income areas of BC and cold spots in the lowest income areas, there were also hot spots with low income and cold spots with high income. It may be that average household income is less relevant than other socioeconomic indicators in BC, or that an effect was not observable for aggregated populations.

Our results suggest that at the LHA level, residential environmental exposures at diagnosis may also be significant potential determinants. Links with vitamin D and sunlight have been inconsistent, due in part to variability in study design and exposure assessment[12]. Our findings indicate that summer vitamin D UV radiation may confer a protective effect on IBD development. An individual unit of UV (1 J/m-2) is quite small which resulted in a low measured IRR. Mean UV in BC is significantly higher in summer (6053.15 J/m-2) than winter (274.98 J/m-2) and varies within the province (summer maximum of 7272.52 and minimum of 4488.80 J/m-2). An increase in UV of 500 J/m-2, which is more representative of actual geographic and seasonal UV variation, was associated with a 29.4% decrease in IBD cases. In the analysis stratified by disease type, summer and winter UV vitamin D were nonsignificant protective factors that improved model fit for CD and UC, respectively. Smaller sample size in stratified analysis may have contributed to nonsignificant results. Low winter UV across BC could be particularly relevant for people of South Asian descent who have a higher level of skin melanin and require substantially more UV exposure to synthesize sufficient vitamin D[30]. PM2.5 air pollution was a significant risk factor for PIBD and CD. Italian (IBD) and Chinese (UC but not CD) studies of middle and older aged adults have identified PM2.5 as a risk factor for incident cases as well as IBD and UC hospitalizations in China[53-55]. In contrast, an Ontario pediatric study found no association and a European adult study found a negative association with PM2.5[13,56]. Though population exposure for most LHAs met 2020 national and provincial air quality objectives, BC experiences seasonal wildfire events which can cause short-term high PM2.5 concentrations that would be obscured in the yearly average values used in this study[57]. Regular high exposure events should be investigated further, especially as climate change is projected to increase wildfire potential[58]. O3 air pollution was a significant risk factor for UC which is consistent with a Chinese study which measured an association between O3 and IBD and UC hospital visits[54]. We observed statistically significant negative associations with CD for NO2 and O3. This is in contrast to the lack of association observed for IBD in Ontario and Europe, a United Kingdom study which found a positive association between NO2 pollution and CD onset before age 23, and a Chinese study which found a positive association between NO2 and UC incidence in middle and older aged adults[13,55,56,59]. Differential effects on CD and UC have been observed for environmental exposures such as smoking and appendectomy[60].

BC had few areas of low residential greenness and high incidence (Supplementary Figures 3–5), and we observed statistically significant positive associations between IBD, CD, UC, and greenness. Our findings contrast with a pediatric cohort study in Ontario, which found a protective effect for maximum growing season NDVI at 250 m[11]. This inconsistency could be due to methodological differences between the two studies. NDVI is a measure of vegetation greenness only and may not capture other elements of green spaces, such as vegetation composition, environmental microbiome, or pesticide and herbicide applications, which may differ in BC. Measures of greenness in BC are highly dependent on specific indicators, as a Metro Vancouver study comparing green space metrics found that NDVI at 250 m from postal codes diverged significantly from other measures such as street tree density, total percentage of green space, and park quality[61].

A previous Canadian study found a protective effect of rurality on pediatric CD and UC, while our study only observed an effect for UC[9]. BC may have more diverse environments than other provinces; for example, the largely rural Interior Health Authority had many of the highest average PM2.5 and O3 exposures while the majority-urban lower mainland region included significant sections of Agricultural Land Reserve. A hot spot near Vancouver included mostly suburban LHAs rather than the main urban center, which is similar to results observed in Oslo, Norway[62]. Perhaps some suburban and rural areas lack protective effects conferred by other rural regions while also missing potential health-promoting features of dense urban areas (e.g. public transportation, public parks, and access to amenities). Some rural and peri-urban areas can also be associated with potential risk factors such as petroleum pesticides.

A novel result in this study is the measured associations between pesticides and PIBD. Petroleum oil applied to grapes and orchards was a significant risk factor for IBD and CD. Indeed, exposure to agricultural petroleum oil has been previously associated with systemic autoimmunity (measured with antinuclear antibodies) and rhinitis in a prospective cohort of pesticide applicators in the United States, suggesting that petroleum pesticide products may have inflammatory properties[63,64]. As numerous other pesticides have been linked with dysbiosis, and petroleum is used as a fungicide, agricultural applications of petroleum oil should be investigated for potential impacts on the gut microbiome which could induce CD[14,39]. Interestingly, glyphosate application appeared to have a different effect on UC depending on which crop it was applied to. Wheat crop glyphosate was a significant risk factor, while alfalfa and corn crop glyphosate were significant protective factors. Differences in agricultural practices between alfalfa, corn, and wheat may be responsible for the negative or both associations observed for UC. The unexpected negative association between IBD and metam pesticide also warrants further investigation. This association could have resulted from an unmeasured confounder such as diet. For example, proximity to fruit and vegetable crops where metam is applied could be correlated with access to these food groups which are known to lower IBD risk[65].

A key strength of this study is the use of high resolution national environmental exposure metrics which increased the quality of the study and will facilitate comparisons between our results and future research. In addition, the clinical registry only contained patients with a confirmed IBD diagnosis, confirmed diagnosis of CD or UC, and accurate date of diagnosis which minimizes risk of misclassification. Moreover, we have demonstrated that our registry data is representative of the BC IBD population[18]. The use of aggregate case counts and reliable population data in both analyses and areal geographic analysis combined with spatial smoothing in the cluster analysis reduced the instability of disease rates caused by low cases numbers. The use of Monte Carlo simulation and a multiple comparison correction for statistical significance increased confidence in the spatial results. Finally, future healthcare service planning in BC would be implemented for health administrative units such as LHAs, so our scale of analysis would allow this research to be directly integrated into planning and intervention efforts.

Despite these strengths, there are several limitations which warrant discussion. Some patients were excluded due to missing data and it is possible that cases were missing from our clinical registry, either of which may have altered spatial clustering or biased modeling results. However, the registry likely included the vast majority of PIBD patients diagnosed in BC during the study period[18]. Patient data was not initially collected for research purposes, which may have affected available variables and may have contributed to missing data. Small numbers of cases could have produced large variation in incidence in sparsely populated areas. In addition, missing early environmental exposure data for several LHAs may have impacted the results, though all but one of the modeled LHAs included environment data from the majority of the study period. An important unmeasured variable in this analysis is diet, and missing potential confounding variables may have biased the results. This was not a birth cohort study, so there is uncertainty about exposures at gestation or early life. However, specific critical periods for many environmental exposures have not been established; consequently, further prospective studies are required. Previous environmental studies have used average childhood exposure[11,13] or did not have a standardized lookback period dating from diagnosis[56,59]. Accordingly, we used broad metrics of average population-level exposure during the study period. This was an ecological analysis and our findings should not be used to make claims about individual risk.

CONCLUSION

To our knowledge, this is the first spatial hot spot analysis focused on PIBD in North America and the first study identifying an association with pesticides. Spatial cluster detection was a valuable method for exploring patterns of IBD, and we identified PM2.5 air pollution, petroleum oil, glyphosate, and metam pesticides as novel determinants of PIBD. Given the inconsistency of IBD incidence in urban areas and relatively high incidence in some suburban and rural areas, future research should move beyond binary urban-rural classifications and use specific characteristics such as built environment and pollutant exposures to characterize environments. Expanded regional and global studies are needed to validate these results and to determine the relationship between timing of exposure and clinical onset of disease. Furthermore, the inclusion of other immune-mediated inflammatory diseases will likely uncover potential shared disease clusters and environmental determinants.

ARTICLE HIGHLIGHTS
Research background

Geospatial patterning has been observed in inflammatory bowel disease (IBD) incidence and linked to environmental determinants of disease. However, knowledge of North American IBD spatial patterns is limited, and unknown in pediatric IBD (PIBD). A further understanding of geospatial patterns of IBD will help guide distribution of healthcare services and aid in identifying potential environmental risk and protective factors and populations at risk.

Research motivation

There is a lack of knowledge of the spatial distribution and environmental exposures relevant to PIBD in Canada and specifically in the Canadian province of British Columbia (BC).

Research objectives

The main objectives of this study were (1) To determine spatial patterning of PIBD and identify location of disease hot and cold spots in the Canadian province of BC during the period of 2001–2016; and (2) to model the association between IBD case counts and population-level ethnicity, average income, rural residence, and known as well as novel environmental determinants. Both objectives were addressed using the methods described below.

Research methods

The Moran’s I statistic was used as a Local Indicator of Spatial Association to measure the degree, location, and type of geographic clustering of PIBD incidence, a method which improves on visual analysis of mapped incidence by empirically quantifying clustering. Statistical significance of observed clusters was approximated using Monte Carlo simulation. Case counts of IBD, Crohn’s disease (CD), and ulcerative colitis (UC) were modeled in Poisson rate models as a function of average population characteristics and average population environmental exposures to assess associations between IBD and rurality, ethnicity, income, family size, and air pollution, green space, ultraviolet (UV) light, and pesticide exposures. Data sources included a BCCH clinical registry of patients diagnosed with IBD ≤ age 16.9, high-quality national environmental exposure datasets developed for health research, and Canadian census data.

Research results

No high incidence hot spots were detected in the densest urban areas, suggesting unexplored urban protective factors. Rurality was negatively associated with UC. Novel risk factors for PIBD and specifically CD included fine particulate matter (PM2.5) pollution and agricultural applications of petroleum oil to orchards and grapes. Spatial distribution was partially explained by rurality, population ethnicity, family size, pesticide applications, air pollution, UV exposure, and residential greenness.

Research conclusions

Pesticide and PM2.5 exposure are linked to the development of PIBD. Suburban and low-density urban areas of BC appear to lack protective exposures conferred by rural and dense urban areas.

Research perspectives

Exploring geographic patterns of PIBD facilitated the identification of novel environmental determinants, which has prompted followup studies of environmental exposures and IBD onset.

ACKNOWLEDGEMENTS

Mielle Michaux was supported as a MSc student by the University of British Columbia Graduate Support Initiative and International Tuition Award and is currently a research assistant supported by the Moffat Foundation. Justin M Chan is a PhD candidate supported by the BCCH Research Institute Studentship and the Lutsky Foundation. Luke Bergmann acknowledges support by the Canada Research Chairs Program and the Canada Foundation for Innovation. Luis F Chaves acknowledges funding from Indiana University. Kevan Jacobson is a Senior Clinician Scientist supported by the Children with Intestinal and Liver Disorders (CHILD) Foundation and the BCCH Research Institute Clinician Scientists Award Program, University of British Columbia. NDVI metrics, nitrogen dioxide data, calculated ozone metrics, and PM2.5 metrics were indexed to DMTI Spatial Inc. postal codes and provided by CANUE (Canadian Urban Environmental Health Research Consortium). Long-term monthly UV data were accessed via the CANUE data portal (https://canuedata.ca). Portions of this methodology were developed for Mielle Michaux’s MSc thesis.

Footnotes

Provenance and peer review: Unsolicited article; Externally peer reviewed.

Peer-review model: Single blind

Corresponding Author's Membership in Professional Societies: Canadian Association of Gastroenterology.

Specialty type: Gastroenterology and hepatology

Country/Territory of origin: Canada

Peer-review report’s scientific quality classification

Grade A (Excellent): 0

Grade B (Very good): 0

Grade C (Good): C, C

Grade D (Fair): 0

Grade E (Poor): 0

P-Reviewer: Maric I, Croatia; Wang LH, China S-Editor: Li L L-Editor: A P-Editor: Cai YX

References
1.  Benchimol EI, Fortinsky KJ, Gozdyra P, Van den Heuvel M, Van Limbergen J, Griffiths AM. Epidemiology of pediatric inflammatory bowel disease: a systematic review of international trends. Inflamm Bowel Dis. 2011;17:423-439.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 640]  [Cited by in F6Publishing: 675]  [Article Influence: 51.9]  [Reference Citation Analysis (0)]
2.  Aamodt G, Jahnsen J, Bengtson MB, Moum B, Vatn MH; IBSEN Study Group. Geographic distribution and ecological studies of inflammatory bowel disease in southeastern Norway in 1990-1993. Inflamm Bowel Dis. 2008;14:984-991.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 20]  [Cited by in F6Publishing: 16]  [Article Influence: 1.0]  [Reference Citation Analysis (0)]
3.  Declercq C, Gower-Rousseau C, Vernier-Massouille G, Salleron J, Baldé M, Poirier G, Lerebours E, Dupas JL, Merle V, Marti R, Duhamel A, Cortot A, Salomez JL, Colombel JF. Mapping of inflammatory bowel disease in northern France: spatial variations and relation to affluence. Inflamm Bowel Dis. 2010;16:807-812.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 21]  [Cited by in F6Publishing: 18]  [Article Influence: 1.3]  [Reference Citation Analysis (0)]
4.  Green C, Elliott L, Beaudoin C, Bernstein CN. A population-based ecologic study of inflammatory bowel disease: searching for etiologic clues. Am J Epidemiol. 2006;164:615-23; discussion 624.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 99]  [Cited by in F6Publishing: 102]  [Article Influence: 5.7]  [Reference Citation Analysis (0)]
5.  Nikkilä A, Auvinen A, Kolho KL. Clustering of pediatric onset inflammatory bowel disease in Finland: a nationwide register-based study. BMC Gastroenterol. 2022;22:512.  [PubMed]  [DOI]  [Cited in This Article: ]  [Reference Citation Analysis (0)]
6.  Bernstein CN, Rawsthorne P, Cheang M, Blanchard JF. A population-based case control study of potential risk factors for IBD. Am J Gastroenterol. 2006;101:993-1002.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 209]  [Cited by in F6Publishing: 210]  [Article Influence: 11.7]  [Reference Citation Analysis (0)]
7.  Pinsk V, Lemberg DA, Grewal K, Barker CC, Schreiber RA, Jacobson K. Inflammatory bowel disease in the South Asian pediatric population of British Columbia. Am J Gastroenterol. 2007;102:1077-1083.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 117]  [Cited by in F6Publishing: 112]  [Article Influence: 6.6]  [Reference Citation Analysis (0)]
8.  Bernstein CN, Burchill C, Targownik LE, Singh H, Roos LL. Events Within the First Year of Life, but Not the Neonatal Period, Affect Risk for Later Development of Inflammatory Bowel Diseases. Gastroenterology. 2019;156:2190-2197.e10.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 41]  [Cited by in F6Publishing: 50]  [Article Influence: 10.0]  [Reference Citation Analysis (0)]
9.  Benchimol EI, Kaplan GG, Otley AR, Nguyen GC, Underwood FE, Guttmann A, Jones JL, Potter BK, Catley CA, Nugent ZJ, Cui Y, Tanyingoh D, Mojaverian N, Bitton A, Carroll MW, deBruyn J, Dummer TJB, El-Matary W, Griffiths AM, Jacobson K, Kuenzig ME, Leddin D, Lix LM, Mack DR, Murthy SK, Sánchez JNP, Singh H, Targownik LE, Vutcovici M, Bernstein CN. Rural and Urban Residence During Early Life is Associated with Risk of Inflammatory Bowel Disease: A Population-Based Inception and Birth Cohort Study. Am J Gastroenterol. 2017;112:1412-1422.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 63]  [Cited by in F6Publishing: 80]  [Article Influence: 11.4]  [Reference Citation Analysis (0)]
10.  Soon IS, Molodecky NA, Rabi DM, Ghali WA, Barkema HW, Kaplan GG. The relationship between urban environment and the inflammatory bowel diseases: a systematic review and meta-analysis. BMC Gastroenterol. 2012;12:51.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 91]  [Cited by in F6Publishing: 102]  [Article Influence: 8.5]  [Reference Citation Analysis (0)]
11.  Elten M, Benchimol EI, Fell DB, Kuenzig ME, Smith G, Kaplan GG, Chen H, Crouse D, Lavigne E. Residential Greenspace in Childhood Reduces Risk of Pediatric Inflammatory Bowel Disease: A Population-Based Cohort Study. Am J Gastroenterol. 2021;116:347-353.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 14]  [Cited by in F6Publishing: 29]  [Article Influence: 9.7]  [Reference Citation Analysis (0)]
12.  Holmes EA, Rodney Harris RM, Lucas RM. Low Sun Exposure and Vitamin D Deficiency as Risk Factors for Inflammatory Bowel Disease, With a Focus on Childhood Onset. Photochem Photobiol. 2019;95:105-118.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 9]  [Cited by in F6Publishing: 9]  [Article Influence: 1.5]  [Reference Citation Analysis (0)]
13.  Elten M, Benchimol EI, Fell DB, Kuenzig ME, Smith G, Chen H, Kaplan GG, Lavigne E. Ambient air pollution and the risk of pediatric-onset inflammatory bowel disease: A population-based cohort study. Environ Int. 2020;138:105676.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 23]  [Cited by in F6Publishing: 28]  [Article Influence: 7.0]  [Reference Citation Analysis (0)]
14.  Jin Y, Wu S, Zeng Z, Fu Z. Effects of environmental pollutants on gut microbiota. Environ Pollut. 2017;222:1-9.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 383]  [Cited by in F6Publishing: 405]  [Article Influence: 57.9]  [Reference Citation Analysis (0)]
15.  Benchimol EI, Mack DR, Guttmann A, Nguyen GC, To T, Mojaverian N, Quach P, Manuel DG. Inflammatory bowel disease in immigrants to Canada and their children: a population-based cohort study. Am J Gastroenterol. 2015;110:553-563.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 158]  [Cited by in F6Publishing: 151]  [Article Influence: 16.8]  [Reference Citation Analysis (0)]
16.  Statistics Canada  2011 census profiles files / 2016 census profiles files / Profile of census dissemination areas; n.d. Database: Canadian Census Analyser. 2011. [cited 1 December 2020]. Available from: http://dc1.chass.utoronto.ca/cgi-bin/census/2011/displayCensus.cgi?year=2011&geo=da.  [PubMed]  [DOI]  [Cited in This Article: ]
17.  Michaux M  Exploring spatio-temporal patterns and environmental determinants of pediatric inflammatory bowel disease in British Columbia. M.Sc. Thesis, University of British Columbia. 2020. Available from: https://open.library.ubc.ca/collections/ubctheses/24/items/1.0394067.  [PubMed]  [DOI]  [Cited in This Article: ]
18.  Chan JM, Carroll MW, Smyth M, Hamilton Z, Evans D, McGrail K, Benchimol EI, Jacobson K. Comparing Health Administrative and Clinical Registry Data: Trends in Incidence and Prevalence of Pediatric Inflammatory Bowel Disease in British Columbia. Clin Epidemiol. 2021;13:81-90.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 3]  [Cited by in F6Publishing: 7]  [Article Influence: 2.3]  [Reference Citation Analysis (0)]
19.   DMTI Spatial Inc. CanMap postal code suite; n.d. [cited 24 March 2021]. Available from: https://canue.ca/wp-content/uploads/2019/09/CANUE-Browser-Metadata-PostalCodes.pdf.  [PubMed]  [DOI]  [Cited in This Article: ]
20.  Statistics Canada  2011 National Household Survey (NHS) profiles files / Profile of census dissemination areas; n.d. Database: Canadian Census Analyser. 2011. [cited 1 December 2020]. Available from: http://dc1.chass.utoronto.ca/cgi-bin/census/2011nhs/displayCensus.cgi?year=2011&geo=da.  [PubMed]  [DOI]  [Cited in This Article: ]
21.  Statistics Canada  Profile of dissemination areas: 2001 census / 2006 census / Cumulative profiles; n.d. 2011. [cited 1 December 2020]. Available from: Database: Canadian Census Analyser. Available from: http://dc1.chass.utoronto.ca/cgi-bin/census/2001/displayCensusDA.cgi.  [PubMed]  [DOI]  [Cited in This Article: ]
22.  Statistics Canada  Average family income (dollars), 2005, 2015 n.d. Database: BC Community Health Atlas. [cited 1 December 2020]. Available from: http://maps.gov.bc.ca/ess/hm/cha/.  [PubMed]  [DOI]  [Cited in This Article: ]
23.  Google Earth Engine  Landsat 5 TM annual greenest-pixel TOA reflectance composite, 1984 to 2012; n.d. Database: Google Earth Engine. [cited 20 July 2017]. Available from: https://explorer.earthengine.google.com/#detail/LANDSAT%2FLT5_L1T_ANNUAL_GREENEST_TOA.  [PubMed]  [DOI]  [Cited in This Article: ]
24.  Google Earth Engine  Landsat 8 annual greenest-pixel TOA reflectance composite, 2013 to 2015; n.d. Database: Google Earth Engine. [cited 20 July 2017]. Available from: https://explorer.earthengine.google.com/#detail/LANDSAT%2FLC8_L1T_ANNUAL_GREENEST_TOA.  [PubMed]  [DOI]  [Cited in This Article: ]
25.  Google Earth Engine  USGS Landsat 5 TM TOA reflectance (orthorectified), 1984 to 2011; n.d. Database: Google Earth Engine. [cited 20 July 2007]. Available from: https://explorer.earthengine.google.com/#detail/LANDSAT%2FLT5_L1T_TOA.  [PubMed]  [DOI]  [Cited in This Article: ]
26.  Google Earth Engine  USGS Landsat 8 TOA reflectance (orthorectified), 2013 to 2017; n.d. Database: Google Earth Engine. [cited 20 July 2007]. Available from: https://explorer.earthengine.google.com/#detail/LANDSAT%2FLC8_L1T_TOA.  [PubMed]  [DOI]  [Cited in This Article: ]
27.  Gorelick N, Hancher M, Dixon M, Ilyushchenko S, Thau D, Moore R. Google Earth Engine: Planetary-scale geospatial analysis for everyone. Remote Sens Environ. 2017;202:18-27.  [PubMed]  [DOI]  [Cited in This Article: ]
28.  Robinson N, Allred B, Jones M, Moreno A, Kimball J, Naugle D, Erickson T, Richardson A. A dynamic landsat derived normalized difference vegetation index (NDVI) product for the conterminous United States. Remote Sens. 2017;9:863.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 120]  [Cited by in F6Publishing: 74]  [Article Influence: 10.6]  [Reference Citation Analysis (0)]
29.  Fioletov VE, Kimlin MG, Krotkov N, McArthur LJB, Kerr JB, Wardle DI, Herman JR, Meltzer R, Mathews TW, Kaurola J. UV index climatology over the United States and Canada from ground-based and satellite estimates. J Geophys Res Atmospheres. 2004;109:1-13.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 69]  [Cited by in F6Publishing: 69]  [Article Influence: 3.5]  [Reference Citation Analysis (0)]
30.  Fioletov VE, McArthur LJ, Mathews TW, Marrett L. Estimated ultraviolet exposure levels for a sufficient vitamin D status in North America. J Photochem Photobiol B. 2010;100:57-66.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 45]  [Cited by in F6Publishing: 48]  [Article Influence: 3.4]  [Reference Citation Analysis (0)]
31.  Hystad P, Setton E, Cervantes A, Poplawski K, Deschenes S, Brauer M, van Donkelaar A, Lamsal L, Martin R, Jerrett M, Demers P. Creating national air pollution models for population exposure assessment in Canada. Environ Health Perspect. 2011;119:1123-1129.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 179]  [Cited by in F6Publishing: 157]  [Article Influence: 12.1]  [Reference Citation Analysis (0)]
32.  Weichenthal S, Pinault LL, Burnett RT. Impact of Oxidant Gases on the Relationship between Outdoor Fine Particulate Air Pollution and Nonaccidental, Cardiovascular, and Respiratory Mortality. Sci Rep. 2017;7:16401.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 60]  [Cited by in F6Publishing: 76]  [Article Influence: 10.9]  [Reference Citation Analysis (0)]
33.  Environment and Climate Change Canada  CHRONOS_Ground-Level_O3_NA_2002.nc to CHRONOS_Ground-Level_O3_NA_2009.nc inclusive, generated. July 2017. [cited 18 April 2023]. Available from: https://www.canuedata.ca/tmp/CANUE_METADATA_O3CHG_A_YY.pdf.  [PubMed]  [DOI]  [Cited in This Article: ]
34.  Environment and Climate Change Canada  GEMMACH_Ground Level_O3_NA_2010.nc to GEMMACH_Ground-Level_O3_NA_2015.nc inclusive, generated. Jul 2017. [cited 18 April 2023]. Available from: https://www.canuedata.ca/tmp/CANUE_METADATA_O3CHG_A_YY.pdf.  [PubMed]  [DOI]  [Cited in This Article: ]
35.  Robichaud A, Ménard R, Zaïtseva Y, Anselmo D. Multi-pollutant surface objective analyses and mapping of air quality health index over North America. Air Qual Atmos Health. 2016;9:743-759.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 39]  [Cited by in F6Publishing: 20]  [Article Influence: 2.5]  [Reference Citation Analysis (0)]
36.  Robichaud A, Ménard R. Multi-year objective analyses of warm season ground-level ozone and PM 2.5 over North America using real-time observations and Canadian operational air quality models. Atmospheric Chem Phys. 2014;14:1769-800.  [PubMed]  [DOI]  [Cited in This Article: ]
37.  Canadian Urban Environmental Health Research Consortium  Canue Metadata - Air Quality Ozone (O3). 2021. [cited 18 April 2023]. Available from: https://www.canuedata.ca/tmp/CANUE_METADATA_O3CHG_A_YY.pdf.  [PubMed]  [DOI]  [Cited in This Article: ]
38.  Hammer MS, van Donkelaar A, Li C, Lyapustin A, Sayer AM, Hsu NC, Levy RC, Garay MJ, Kalashnikova OV, Kahn RA, Brauer M, Apte JS, Henze DK, Zhang L, Zhang Q, Ford B, Pierce JR, Martin RV. Global Estimates and Long-Term Trends of Fine Particulate Matter Concentrations (1998-2018). Environ Sci Technol. 2020;54:7879-7890.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 294]  [Cited by in F6Publishing: 289]  [Article Influence: 72.3]  [Reference Citation Analysis (0)]
39.  Maggi F, Tang FHM, la Cecilia D, McBratney A. PEST-CHEMGRIDS, global gridded maps of the top 20 crop-specific pesticide application rates from 2015 to 2025. Sci Data. 2019;6:170.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 139]  [Cited by in F6Publishing: 93]  [Article Influence: 18.6]  [Reference Citation Analysis (0)]
40.  Maggi F, Tang FHM, la Cecilia D, McBratney A.   Global Pesticide Grids (PEST-CHEMGRIDS); Database: NASA Socioeconomic Data and Applications Center (SEDAC) 2020.  [PubMed]  [DOI]  [Cited in This Article: ]
41.  BC Stats  British Columbia - Population estimates; 2020. [cited 1 December2020]. Available from: https://bcstats.shinyapps.io/popApp/.  [PubMed]  [DOI]  [Cited in This Article: ]
42.  Ministry of Health  Local health area boundaries; Database: BC Data Catalogue. 2019. [cited 26 May 2020]. Available from: https://catalogue.data.gov.bc.ca/dataset/Local-health-area-boundaries.  [PubMed]  [DOI]  [Cited in This Article: ]
43.  Marshall RJ. Mapping disease and mortality rates using empirical Bayes estimators. J R Stat Soc Ser C Appl Stat. 1991;40:283-294.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 146]  [Cited by in F6Publishing: 146]  [Article Influence: 4.4]  [Reference Citation Analysis (0)]
44.  Anselin L. Local indicators of spatial association-LISA. Geogr Anal. 1995;27:93-115.  [PubMed]  [DOI]  [Cited in This Article: ]
45.  Kulldorff M, Nagarwalla N. Spatial disease clusters: detection and inference. Stat Med. 1995;14:799-810.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 971]  [Cited by in F6Publishing: 833]  [Article Influence: 28.7]  [Reference Citation Analysis (0)]
46.  Moraga P, Montes F. Detection of spatial disease clusters with LISA functions. Stat Med. 2011;30:1057-1071.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 24]  [Cited by in F6Publishing: 24]  [Article Influence: 1.8]  [Reference Citation Analysis (0)]
47.  Venables WN, Ripley BD.   Modern Applied Statistics with S. 4th ed. New York: Springer, 2002: 1-498.  [PubMed]  [DOI]  [Cited in This Article: ]
48.  Malhotra R, Turner K, Sonnenberg A, Genta RM. High prevalence of inflammatory bowel disease in United States residents of Indian ancestry. Clin Gastroenterol Hepatol. 2015;13:683-689.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 28]  [Cited by in F6Publishing: 30]  [Article Influence: 3.3]  [Reference Citation Analysis (0)]
49.  Misra R, Faiz O, Munkholm P, Burisch J, Arebi N. Epidemiology of inflammatory bowel disease in racial and ethnic migrant groups. World J Gastroenterol. 2018;24:424-437.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in CrossRef: 34]  [Cited by in F6Publishing: 27]  [Article Influence: 4.5]  [Reference Citation Analysis (2)]
50.  Rajasekaran V, Evans HM, Andrews A, Bishop JR, Lopez RN, Mouat S, Han DY, Alsweiler J, Roberts AJ. Rising Incidence of Inflammatory Bowel Disease in South Asian children in New Zealand - A Retrospective Population-Based Study. J Pediatr Gastroenterol Nutr. 2023;.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 2]  [Reference Citation Analysis (0)]
51.  Huang JG, Wong YKY, Chew KS, Tanpowpong P, Calixto Mercado KS, Reodica A, Rajindrajith S, Chang KC, Ni YH, Treepongkaruna S, Lee WS, Aw MM. Epidemiological characteristics of Asian children with inflammatory bowel disease at diagnosis: Insights from an Asian-Pacific multi-centre registry network. World J Gastroenterol. 2022;28:1830-1844.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in CrossRef: 7]  [Cited by in F6Publishing: 5]  [Article Influence: 2.5]  [Reference Citation Analysis (0)]
52.  Torabi M, Bernstein CN, Yu BN, Wickramasinghe L, Blanchard JF, Singh H. Geographical Variation and Factors Associated With Inflammatory Bowel Disease in a Central Canadian Province. Inflamm Bowel Dis. 2020;26:581-590.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 1]  [Cited by in F6Publishing: 2]  [Article Influence: 0.5]  [Reference Citation Analysis (0)]
53.  Adami G, Pontalti M, Cattani G, Rossini M, Viapiana O, Orsolini G, Benini C, Bertoldo E, Fracassi E, Gatti D, Fassio A. Association between long-term exposure to air pollution and immune-mediated diseases: a population-based cohort study. RMD Open. 2022;8.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 2]  [Cited by in F6Publishing: 32]  [Article Influence: 16.0]  [Reference Citation Analysis (0)]
54.  Ding S, Sun S, Ding R, Song S, Cao Y, Zhang L. Association between exposure to air pollutants and the risk of inflammatory bowel diseases visits. Environ Sci Pollut Res Int. 2022;29:17645-17654.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 14]  [Cited by in F6Publishing: 12]  [Article Influence: 6.0]  [Reference Citation Analysis (0)]
55.  Li FR, Wu KY, Fan WD, Chen GC, Tian H, Wu XB. Long-term exposure to air pollution and risk of incident inflammatory bowel disease among middle and old aged adults. Ecotoxicol Environ Saf. 2022;242:113835.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 1]  [Cited by in F6Publishing: 5]  [Article Influence: 2.5]  [Reference Citation Analysis (0)]
56.  Opstelten JL, Beelen RMJ, Leenders M, Hoek G, Brunekreef B, van Schaik FDM, Siersema PD, Eriksen KT, Raaschou-Nielsen O, Tjønneland A, Overvad K, Boutron-Ruault MC, Carbonnel F, de Hoogh K, Key TJ, Luben R, Chan SSM, Hart AR, Bueno-de-Mesquita HB, Oldenburg B. Exposure to Ambient Air Pollution and the Risk of Inflammatory Bowel Disease: A European Nested Case-Control Study. Dig Dis Sci. 2016;61:2963-2971.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 41]  [Cited by in F6Publishing: 40]  [Article Influence: 5.0]  [Reference Citation Analysis (0)]
57.  Matz CJ, Egyed M, Xi G, Racine J, Pavlovic R, Rittmaster R, Henderson SB, Stieb DM. Health impact analysis of PM(2.5) from wildfire smoke in Canada (2013-2015, 2017-2018). Sci Total Environ. 2020;725:138506.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 68]  [Cited by in F6Publishing: 60]  [Article Influence: 15.0]  [Reference Citation Analysis (0)]
58.  Wang X, Thompson DK, Marshall GA, Tymstra C, Carr R, Flannigan MD.   Increasing frequency of extreme fire weather in Canada with climate change. Clim Change 2015; 130: 573-586.  [PubMed]  [DOI]  [Cited in This Article: ]
59.  Kaplan GG, Hubbard J, Korzenik J, Sands BE, Panaccione R, Ghosh S, Wheeler AJ, Villeneuve PJ. The inflammatory bowel diseases and ambient air pollution: a novel association. Am J Gastroenterol. 2010;105:2412-2419.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 158]  [Cited by in F6Publishing: 176]  [Article Influence: 12.6]  [Reference Citation Analysis (0)]
60.  Piovani D, Danese S, Peyrin-Biroulet L, Nikolopoulos GK, Lytras T, Bonovas S. Environmental Risk Factors for Inflammatory Bowel Diseases: An Umbrella Review of Meta-analyses. Gastroenterology. 2019;157:647-659.e4.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 265]  [Cited by in F6Publishing: 410]  [Article Influence: 82.0]  [Reference Citation Analysis (0)]
61.  Rugel EJ, Henderson SB, Carpiano RM, Brauer M. Beyond the Normalized Difference Vegetation Index (NDVI): Developing a Natural Space Index for population-level health research. Environ Res. 2017;159:474-483.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 71]  [Cited by in F6Publishing: 58]  [Article Influence: 8.3]  [Reference Citation Analysis (0)]
62.  Moum B, Vatn MH, Ekbom A, Aadland E, Fausa O, Lygren I, Stray N, Sauar J, Schulz T. Incidence of Crohn's disease in four counties in southeastern Norway, 1990-93. A prospective population-based study. The Inflammatory Bowel South-Eastern Norway (IBSEN) Study Group of Gastroenterologists. Scand J Gastroenterol. 1996;31:355-361.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 116]  [Cited by in F6Publishing: 118]  [Article Influence: 4.2]  [Reference Citation Analysis (0)]
63.  Slager RE, Poole JA, LeVan TD, Sandler DP, Alavanja MC, Hoppin JA. Rhinitis associated with pesticide exposure among commercial pesticide applicators in the Agricultural Health Study. Occup Environ Med. 2009;66:718-724.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 56]  [Cited by in F6Publishing: 43]  [Article Influence: 2.9]  [Reference Citation Analysis (0)]
64.  Parks CG, Santos ASE, Lerro CC, DellaValle CT, Ward MH, Alavanja MC, Berndt SI, Beane Freeman LE, Sandler DP, Hofmann JN. Lifetime Pesticide Use and Antinuclear Antibodies in Male Farmers From the Agricultural Health Study. Front Immunol. 2019;10:1476.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 29]  [Cited by in F6Publishing: 26]  [Article Influence: 5.2]  [Reference Citation Analysis (0)]
65.  Milajerdi A, Ebrahimi-Daryani N, Dieleman LA, Larijani B, Esmaillzadeh A. Association of Dietary Fiber, Fruit, and Vegetable Consumption with Risk of Inflammatory Bowel Disease: A Systematic Review and Meta-Analysis. Adv Nutr. 2021;12:735-743.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 22]  [Cited by in F6Publishing: 51]  [Article Influence: 12.8]  [Reference Citation Analysis (0)]