Marta Blangiardo

Spatial and spatio-temporal models with R-INLA.

During the last three decades, Bayesian methods have developed greatly in the field of epidemiology. Their main challenge focusses around computation, but the advent of Markov Chain Monte Carlo methods (MCMC) and in particular of the WinBUGS software has opened the doors of Bayesian modelling to the wide research community. However model complexity and database dimension still remain a constraint. Recently the use of Gaussian random fields has become increasingly popular in epidemiology as very often epidemiological data are characterised by a spatial and/or temporal structure which needs to be taken into account in the inferential process. The Integrated Nested Laplace Approximation (INLA) approach has been developed as a computationally efficient alternative to MCMC and the availability of an R package (R-INLA) allows researchers to easily apply this method. In this paper we review the INLA approach and present some applications on spatial and spatio-temporal data.

Aircraft noise and cardiovascular disease near Heathrow airport in London: small area study

Objective To investigate the association of aircraft noise with risk of stroke, coronary heart disease, and cardiovascular disease in the general population. Design Small area study. Setting 12 London boroughs and nine districts west of London exposed to aircraft noise related to Heathrow airport in London. Population About 3.6 million residents living near Heathrow airport. Risks for hospital admissions were assessed in 12 110 census output areas (average population about 300 inhabitants) and risks for mortality in 2378 super output areas (about 1500 inhabitants). Main outcome measures Risk of hospital admissions for, and mortality from, stroke, coronary heart disease, and cardiovascular disease, 2001-05. Results Hospital admissions showed statistically significant linear trends (P<0.001 to P<0.05) of increasing risk with higher levels of both daytime (average A weighted equivalent noise 7 am to 11 pm, LAeq,16h) and night time (11 pm to 7 am, Lnight) aircraft noise. When areas experiencing the highest levels of daytime aircraft noise were compared with those experiencing the lowest levels (>63 dB v ≤51 dB), the relative risk of hospital admissions for stroke was 1.24 (95% confidence interval 1.08 to 1.43), for coronary heart disease was 1.21 (1.12 to 1.31), and for cardiovascular disease was 1.14 (1.08 to 1.20) adjusted for age, sex, ethnicity, deprivation, and a smoking proxy (lung cancer mortality) using a Poisson regression model including a random effect term to account for residual heterogeneity. Corresponding relative risks for mortality were of similar magnitude, although with wider confidence limits. Admissions for coronary heart disease and cardiovascular disease were particularly affected by adjustment for South Asian ethnicity, which needs to be considered in interpretation. All results were robust to adjustment for particulate matter (PM10) air pollution, and road traffic noise, possible for London boroughs (population about 2.6 million). We could not distinguish between the effects of daytime or night time noise as these measures were highly correlated. Conclusion High levels of aircraft noise were associated with increased risks of stroke, coronary heart disease, and cardiovascular disease for both hospital admissions and mortality in areas near Heathrow airport in London. As well as the possibility of causal associations, alternative explanations such as residual confounding and potential for ecological bias should be considered.

Road traffic noise is associated with increased cardiovascular morbidity and mortality and all-cause mortality in London.

Aims Road traffic noise has been associated with hypertension but evidence for the long-term effects on hospital admissions and mortality is limited. We examined the effects of long-term exposure to road traffic noise on hospital admissions and mortality in the general population. Methods and results The study population consisted of 8.6 million inhabitants of London, one of Europes largest cities. We assessed small-area-level associations of day- (7:00–22:59) and nighttime (23:00–06:59) road traffic noise with cardiovascular hospital admissions and all-cause and cardiovascular mortality in all adults (≥25 years) and elderly (≥75 years) through Poisson regression models. We adjusted models for age, sex, area-level socioeconomic deprivation, ethnicity, smoking, air pollution, and neighbourhood spatial structure. Median daytime exposure to road traffic noise was 55.6 dB. Daytime road traffic noise increased the risk of hospital admission for stroke with relative risk (RR) 1.05 [95% confidence interval (CI): 1.02–1.09] in adults, and 1.09 (95% CI: 1.04–1.14) in the elderly in areas >60 vs. <55 dB. Nighttime noise was associated with stroke admissions only among the elderly. Daytime noise was significantly associated with all-cause mortality in adults [RR 1.04 (95% CI: 1.00–1.07) in areas >60 vs. <55 dB]. Positive but non-significant associations were seen with mortality for cardiovascular and ischaemic heart disease, and stroke. Results were similar for the elderly. Conclusions Long-term exposure to road traffic noise was associated with small increased risks of all-cause mortality and cardiovascular mortality and morbidity in the general population, particularly for stroke in the elderly.

Exposure to aircraft and road traffic noise and associations with heart disease and stroke in six European countries: a cross-sectional study

BackgroundAlthough a number of studies have found an association between aircraft noise and hypertension, there is a lack of evidence on associations with other cardiovascular disease. For road traffic noise, more studies are available but the extent of possible confounding by air pollution has not been established.MethodsThis study used data from the Hypertension and Environmental Noise near Airports (HYENA) study. Cross-sectional associations between self-reported ‘heart disease and stroke’ and aircraft noise and road traffic noise were examined using data collected between 2004 and 2006 on 4712 participants (276 cases), who lived near airports in six European countries (UK, Germany, Netherlands, Sweden, Greece, Italy). Data were available to assess potential confounding by NO2 air pollution in a subsample of three countries (UK, Netherlands, Sweden).ResultsAn association between night-time average aircraft noise and ‘heart disease and stroke’ was found after adjustment for socio-demographic confounders for participants who had lived in the same place foru2009≥u200920xa0years (odds ratio (OR): 1.25 (95% confidence interval (CI) 1.03, 1.51) per 10xa0dB (A)); this association was robust to adjustment for exposure to air pollution in the subsample. 24xa0hour average road traffic noise exposure was associated with ‘heart disease and stroke’ (OR: 1.19 (95% CI 1.00, 1.41), but adjustment for air pollution in the subsample suggested this may have been due to confounding by air pollution. Statistical assessment (correlations and variance inflation factor) suggested only modest collinearity between noise and NO2 exposures.ConclusionsExposure to aircraft noise over many years may increase risks of heart disease and stroke, although more studies are needed to establish how much the risks associated with road traffic noise may be explained by air pollution.

Bayesian hierarchical model for the prediction of football results

The problem of modelling football data has become increasingly popular in the last few years and many different models have been proposed with the aim of estimating the characteristics that bring a team to lose or win a game, or to predict the score of a particular match. We propose a Bayesian hierarchical model to fulfil both these aims and test its predictive strength based on data about the Italian Serie A 1991–1992 championship. To overcome the issue of overshrinkage produced by the Bayesian hierarchical model, we specify a more complex mixture model that results in a better fit to the observed data. We test its performance using an example of the Italian Serie A 2007–2008 championship.

Medication use in relation to noise from aircraft and road traffic in six European countries: results of the HYENA study

Objectives Studies on the health effects of aircraft and road traffic noise exposure suggest excess risks of hypertension, cardiovascular disease and the use of sedatives and hypnotics. Our aim was to assess the use of medication in relation to noise from aircraft and road traffic. Methods This cross-sectional study measured the use of prescribed antihypertensives, antacids, anxiolytics, hypnotics, antidepressants and antasthmatics in 4,861 persons living near seven airports in six European countries (UK, Germany, the Netherlands, Sweden, Italy, and Greece). Exposure was assessed using models with 1dB resolution (5dB for UK road traffic noise) and spatial resolution of 250×250m for aircraft and 10×10m for road traffic noise. Data were analysed using multilevel logistic regression, adjusting for potential confounders. Results We found marked differences between countries in the effect of aircraft noise on antihypertensive use; for night-time aircraft noise, a 10dB increase in exposure was associated with ORs of 1.34 (95% CI 1.14 to1.57) for the UK and 1.19 (1.02 to 1.38) for the Netherlands but no significant associations were found for other countries. For day-time aircraft noise, excess risks were found for the UK (OR 1.35; CI: 1.13 to 1.60) but a risk deficit for Italy (OR 0.82; CI: 0.71 to 0.96). There was an excess risk of taking anxiolytic medication in relation to aircraft noise (OR 1.28; CI: 1.04 to 1.57 for daytime and OR 1.27; CI: 1.01 to 1.59 for night-time) which held across countries. We also found an association between exposure to 24hr road traffic noise and the use of antacids by men (OR 1.39; CI 1.11 to 1.74). Conclusion Our results suggest an effect of aircraft noise on the use of antihypertensive medication, but this effect did not hold for all countries. Results were more consistent across countries for the increased use of anxiolytics in relation to aircraft noise.

Historic air pollution exposure and long-term mortality risks in England and Wales: prospective longitudinal cohort study

Introduction Long-term air pollution exposure contributes to mortality but there are few studies examining effects of very long-term (>25u2005years) exposures. Methods This study investigated modelled air pollution concentrations at residence for 1971, 1981, 1991 (black smoke (BS) and SO2) and 2001 (PM10) in relation to mortality up to 2009 in 367u2005658 members of the longitudinal survey, a 1% sample of the English Census. Outcomes were all-cause (excluding accidents), cardiovascular (CV) and respiratory mortality. Results BS and SO2 exposures remained associated with mortality decades after exposure—BS exposure in 1971 was significantly associated with all-cause (OR 1.02 (95% CI 1.01 to 1.04)) and respiratory (OR 1.05 (95% CI 1.01 to 1.09)) mortality in 2002–2009 (ORs expressed per 10 μg/m3). Largest effect sizes were seen for more recent exposures and for respiratory disease. PM10 exposure in 2001 was associated with all outcomes in 2002–2009 with stronger associations for respiratory (OR 1.22 (95% CI 1.04 to 1.44)) than CV mortality (OR 1.12 (95% CI 1.01 to 1.25)). Adjusting PM10 for past BS and SO2 exposures in 1971, 1981 and 1991 reduced the all-cause OR to 1.16 (95% CI 1.07 to 1.26) while CV and respiratory associations lost significance, suggesting confounding by past air pollution exposure, but there was no evidence for effect modification. Limitations include limited information on confounding by smoking and exposure misclassification of historic exposures. Conclusions This large national study suggests that air pollution exposure has long-term effects on mortality that persist decades after exposure, and that historic air pollution exposures influence current estimates of associations between air pollution and mortality.

Postnuclear disaster evacuation and chronic health in adults in Fukushima, Japan: a long-term retrospective analysis.

Objective Japans 2011 Fukushima Daiichi Nuclear Power Plant incident required the evacuation of over a million people, creating a large displaced population with potentially increased vulnerability in terms of chronic health conditions. We assessed the long-term impact of evacuation on diabetes, hyperlipidaemia and hypertension. Participants We considered participants in annual public health check-ups from 2008 to 2014, administrated by Minamisoma City and Soma City, located about 10–50u2005km from the Fukushima nuclear plant. Methods Disease risks, measured in terms of pre-incident and post-incident relative risks, were examined and compared between evacuees and non-evacuees/temporary-evacuees. We also constructed logistic regression models to assess the impact of evacuation on the disease risks adjusted for covariates. Results Data from a total of 6406 individuals aged 40–74u2005years who participated in the check-ups both at baseline (2008–2010) and in one or more post-incident years were analysed. Regardless of evacuation, significant post-incident increases in risk were observed for diabetes and hyperlipidaemia (relative risk: 1.27–1.60 and 1.12–1.30, respectively, depending on evacuation status and post-incident year). After adjustment for covariates, the increase in hyperlipidaemia was significantly greater among evacuees than among non-evacuees/temporary-evacuees (OR 1.18, 95% CI 1.06 to 1.32, p<0.01). Conclusions The singularity of this study is that evacuation following the Fukushima disaster was found to be associated with a small increase in long-term hyperlipidaemia risk in adults. Our findings help identify discussion points on disaster planning, including preparedness, response and recovery measures, applicable to future disasters requiring mass evacuation.

Traffic Air Pollution and Other Risk Factors for Respiratory Illness in Schoolchildren in the Niger-Delta Region of Nigeria

Background: Association of childhood respiratory illness with traffic air pollution has been investigated largely in developed but not in developing countries, where pollution levels are often very high. Objectives: In this study we investigated associations between respiratory health and outdoor and indoor air pollution in schoolchildren 7–14 years of age in low socioeconomic status areas in the Niger Delta. Methods: A cross-sectional survey was carried out among 1,397 schoolchildren. Exposure to home outdoor and indoor air pollution was assessed by self-report questionnaire. School air pollution exposures were assessed using traffic counts, distance of schools to major streets, and particulate matter and carbon monoxide measurements, combined using principal components analysis. Hierarchical logistic regression was used to examine associations with reported respiratory health, adjusting for potential confounders. Results: Traffic disturbance at home (i.e., traffic noise and/or fumes evident inside the home vs. none) was associated with wheeze [odds ratio (OR) = 2.16; 95% confidence interval (CI), 1.28–3.64], night cough (OR = 1.37; 95% CI, 1.03–1.82), phlegm (OR = 1.49; 95% CI, 1.09–2.04), and nose symptoms (OR = 1.40; 95% CI, 1.03–1.90), whereas school exposure to a component variable indicating exposure to fine particles was associated with increased phlegm (OR = 1.38; 95% CI, 1.09–1.75). Nonsignificant positive associations were found between cooking with wood/coal (OR = 2.99; 95% CI, 0.88–10.18) or kerosene (OR = 2.83; 95% CI, 0.85–9.44) and phlegm compared with cooking with gas. Conclusion: Traffic pollution is associated with respiratory symptoms in schoolchildren in a deprived area of western Africa. Associations may have been underestimated because of nondifferential misclassification resulting from limitations in exposure measurement.

Using Building Heights and Street Configuration to Enhance Intraurban PM10, NOX, and NO2 Land Use Regression Models

Land use regression (LUR) models have been widely used to provide long-term air pollution exposure assessment in epidemiological studies. However, models have rarely offered variables that account for the dispersion environment close to the source (e.g., street canyons, position and dimensions of buildings, road width). This study used newly available data on building heights and geometry to enhance the representation of land use and the dispersion field in LUR. Models were developed for PM10, NO(X), and NO2 for 2008-2011 for London, U.K. A separate set of models using traditional land use and traffic indicators (e.g., distance from road, area of housing within circular buffers) were also developed and their performance was compared with enhanced models. Models were evaluated using leave-one-out (n - 1) (LOOCV) and grouped (n - 25%) cross-validation (GCV). LOOCV R(2) values were 0.71, 0.50, 0.66 and 0.73, 0.79, 0.78 for traditional and enhanced PM10, NOX, and NO2 models, respectively. GCV R(2) values were 0.71, 0.53, 0.64 and 0.68, 0.77, 0.77 for traditional and enhanced PM10, NO(X), and NO2 models, respectively. Data on building volume within the area common to a 20 m road buffer within a 25 m circular buffer substantially improved the performance (R(2) > 13%) of NO(X) and NO2 LUR models.