Background: Ambient temperature is an important determinant of daily mortality that is of interest both in its own right and as a confounder of other determinants investigated using time-series ...regressions, in particular, air pollution. The temperature-mortality relationship is often found to be substantially nonlinear and to persist (but change shape) with increasing lag. We review and extend models for such nonlinear multilag forms. Development: Popular models for mortality by temperature at given lag include polynomial and natural cubic spline curves, and the simple but more easily interpreted linear thresholds model, comprising linear relationships for temperatures below and above thresholds and a flat middle section. Most published analyses that have allowed the relationship to persist over multiple lags have done so by assuming that spline or threshold models apply to mean temperature in several lag strata (eg, lags 0-1, 2-6, and 7-13). However, more flexible models are possible, and a modeling framework using products of basis functions ("cross-basis" functions) suggests a wide range, some used previously and some new. These allow for stepped or smooth changes in the model coefficients as lags increase. Applying a range of models to data from London suggest evidence for relationships up to at least 2 weeks' lag, with smooth models fitting best but lag-stratified threshold models allowing the most direct interpretation. Conclusions: A wide range of multilag nonlinear temperature-mortality relationships can be modeled. More awareness of options should improve investigation of these relationships and help control for confounding by them.
Full text
Available for:
BFBNIB, CMK, NMLJ, NUK, PNG, SAZU, UL, UM, UPUK
Time series regression studies have been widely used in environmental epidemiology, notably in investigating the short-term associations between exposures such as air pollution, weather variables or ...pollen, and health outcomes such as mortality, myocardial infarction or disease-specific hospital admissions. Typically, for both exposure and outcome, data are available at regular time intervals (e.g. daily pollution levels and daily mortality counts) and the aim is to explore short-term associations between them. In this article, we describe the general features of time series data, and we outline the analysis process, beginning with descriptive analysis, then focusing on issues in time series regression that differ from other regression methods: modelling short-term fluctuations in the presence of seasonal and long-term patterns, dealing with time varying confounding factors and modelling delayed ('lagged') associations between exposure and outcome. We finish with advice on model checking and sensitivity analysis, and some common extensions to the basic model.
The time stratified case cross-over approach is a popular alternative to conventional time series regression for analysing associations between time series of environmental exposures (air pollution, ...weather) and counts of health outcomes. These are almost always analyzed using conditional logistic regression on data expanded to case-control (case crossover) format, but this has some limitations. In particular adjusting for overdispersion and auto-correlation in the counts is not possible. It has been established that a Poisson model for counts with stratum indicators gives identical estimates to those from conditional logistic regression and does not have these limitations, but it is little used, probably because of the overheads in estimating many stratum parameters.
The conditional Poisson model avoids estimating stratum parameters by conditioning on the total event count in each stratum, thus simplifying the computing and increasing the number of strata for which fitting is feasible compared with the standard unconditional Poisson model. Unlike the conditional logistic model, the conditional Poisson model does not require expanding the data, and can adjust for overdispersion and auto-correlation. It is available in Stata, R, and other packages.
By applying to some real data and using simulations, we demonstrate that conditional Poisson models were simpler to code and shorter to run than are conditional logistic analyses and can be fitted to larger data sets than possible with standard Poisson models. Allowing for overdispersion or autocorrelation was possible with the conditional Poisson model but when not required this model gave identical estimates to those from conditional logistic regression.
Conditional Poisson regression models provide an alternative to case crossover analysis of stratified time series data with some advantages. The conditional Poisson model can also be used in other contexts in which primary control for confounding is by fine stratification.
Full text
Available for:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
Recent investigations have reported a decline in the heat-related mortality risk during the last decades. However, these studies are frequently based on modeling approaches that do not fully ...characterize the complex temperature-mortality relationship, and are limited to single cities or countries.
We assessed the temporal variation in heat-mortality associations in a multi-country data set using flexible modelling techniques.
We collected data for 272 locations in Australia, Canada, Japan, South Korea, Spain, the United Kingdom, and the United States, with a total 20,203,690 deaths occurring in summer months between 1985 and 2012. The analysis was based on two-stage time-series models. The temporal variation in heat-mortality relationships was estimated in each location with time-varying distributed lag nonlinear models, expressed through an interaction between the transformed temperature variables and time. The estimates were pooled by country through multivariate meta-analysis.
Mortality risk due to heat appeared to decrease over time in several countries, with relative risks associated to high temperatures significantly lower in 2006 compared with 1993 in the United States, Japan, and Spain, and a nonsignificant decrease in Canada. Temporal changes are difficult to assess in Australia and South Korea due to low statistical power, and we found little evidence of variation in the United Kingdom. In the United States, the risk seems to be completely abated in 2006 for summer temperatures below their 99th percentile, but some significant excess persists for higher temperatures in all the countries.
We estimated a statistically significant decrease in the relative risk for heat-related mortality in 2006 compared with 1993 in the majority of countries included in the analysis.
Gasparrini A, Guo Y, Hashizume M, Kinney PL, Petkova EP, Lavigne E, Zanobetti A, Schwartz JD, Tobias A, Leone M, Tong S, Honda Y, Kim H, Armstrong BG. 2015. Temporal variation in heat-mortality associations: a multicountry study. Environ Health Perspect 123:1200-1207; http://dx.doi.org/10.1289/ehp.1409070.
Full text
Available for:
CEKLJ, DOBA, IZUM, KILJ, NUK, OILJ, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK, VSZLJ
Few studies have examined variation in the associations between heat waves and mortality in an international context.
We aimed to systematically examine the impacts of heat waves on mortality with ...lag effects internationally.
We collected daily data of temperature and mortality from 400 communities in 18 countries/regions and defined 12 types of heat waves by combining community-specific daily mean temperature ≥90th, 92.5th, 95th, and 97.5th percentiles of temperature with duration ≥2, 3, and 4 d. We used time-series analyses to estimate the community-specific heat wave-mortality relation over lags of 0-10 d. Then, we applied meta-analysis to pool heat wave effects at the country level for cumulative and lag effects for each type of heat wave definition.
Heat waves of all definitions had significant cumulative associations with mortality in all countries, but varied by community. The higher the temperature threshold used to define heat waves, the higher heat wave associations on mortality. However, heat wave duration did not modify the impacts. The association between heat waves and mortality appeared acutely and lasted for 3 and 4 d. Heat waves had higher associations with mortality in moderate cold and moderate hot areas than cold and hot areas. There were no added effects of heat waves on mortality in all countries/regions, except for Brazil, Moldova, and Taiwan. Heat waves defined by daily mean and maximum temperatures produced similar heat wave-mortality associations, but not daily minimum temperature.
Results indicate that high temperatures create a substantial health burden, and effects of high temperatures over consecutive days are similar to what would be experienced if high temperature days occurred independently. People living in moderate cold and moderate hot areas are more sensitive to heat waves than those living in cold and hot areas. Daily mean and maximum temperatures had similar ability to define heat waves rather than minimum temperature. https://doi.org/10.1289/EHP1026.
Full text
Available for:
CEKLJ, DOBA, IZUM, KILJ, NUK, OILJ, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK, VSZLJ
Summary We used Comparative Risk Assessment methods to estimate the health effects of alternative urban land transport scenarios for two settings—London, UK, and Delhi, India. For each setting, we ...compared a business-as-usual 2030 projection (without policies for reduction of greenhouse gases) with alternative scenarios—lower-carbon-emission motor vehicles, increased active travel, and a combination of the two. We developed separate models that linked transport scenarios with physical activity, air pollution, and risk of road traffic injury. In both cities, we noted that reduction in carbon dioxide emissions through an increase in active travel and less use of motor vehicles had larger health benefits per million population (7332 disability-adjusted life-years DALYs in London, and 12 516 in Delhi in 1 year) than from the increased use of lower-emission motor vehicles (160 DALYs in London, and 1696 in Delhi). However, combination of active travel and lower-emission motor vehicles would give the largest benefits (7439 DALYs in London, 12 995 in Delhi), notably from a reduction in the number of years of life lost from ischaemic heart disease (10–19% in London, 11–25% in Delhi). Although uncertainties remain, climate change mitigation in transport should benefit public health substantially. Policies to increase the acceptability, appeal, and safety of active urban travel, and discourage travel in private motor vehicles would provide larger health benefits than would policies that focus solely on lower-emission motor vehicles.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UL, UM, UPCLJ, UPUK
Exposure to some perfluoroalkyl substances (PFAS), such as perfluorohexane sulfonate (PFHxS), perfluorooctanoate (PFOA), perfluorooctane sulfonate (PFOS), and perfluorononanoic acid (PFNA), may alter ...levels of sex hormones and insulin-like growth factor-1 (IGF-1) in animals. Human studies on this topic are scarce, and none have been conducted in young children.
We investigated the relationship between levels of PFAS and estradiol, total testosterone, and IGF-1 in 2,292 children (6-9 years of age) from the C8 Health Project who lived near a chemical plant in the Mid-Ohio Valley (USA) with local contamination from PFOA.
Serum samples were collected in 2005-2006 and analyzed for PFAS, sex hormones, and IGF-1. Results from regression models were expressed as the adjusted percentage difference (95% CI) per sex-specific interquartile range (IQR) increment of each PFAS serum concentration. Analyses by PFAS quartiles were also conducted.
Median concentrations of PFHxS, PFOA, PFOS, and PFNA were 8, 35, 22, and 1.7 ng/mL in boys and 7, 30, 21, and 1.7 ng/mL in girls. In boys, PFOA concentrations were significantly associated with testosterone levels (-4.9%; 95% CI: -8.7, -0.8%); PFOS with estradiol (-4.0%; 95% CI: -7.7, -0.1%), testosterone (-5.8%; 95% CI: -9.4, -2.0%), and IGF-1 (-5.9%; 95% CI: -8.3, -3.3%); and PFNA with IGF-1 (-3.5%; 95% CI: -6.0, -1.0%). In girls, significant associations were found between PFOS and testosterone (-6.6%; 95% CI: -10.1, -2.8%) and IGF-1 (-5.6%; -8.2, -2.9%); and PFNA and IGF-1 (-3.8%; 95% CI: -6.4, -1.2%). In both sexes, the magnitudes of the associations decreased monotonically across quartiles for both testosterone and IGF-1 in relation to PFOS, and for IGF-1 and PFNA in girls.
To our knowledge, this is the first study suggesting that PFAS are associated with lower levels of IGF-1 and sex hormones in young children.
Lopez-Espinosa MJ, Mondal D, Armstrong BG, Eskenazi B, Fletcher T. 2016. Perfluoroalkyl substances, sex hormones, and insulin-like growth factor-1 at 6-9 years of age: a cross-sectional analysis within the C8 Health Project. Environ Health Perspect 124:1269-1275; http://dx.doi.org/10.1289/ehp.1509869.
Full text
Available for:
CEKLJ, DOBA, IZUM, KILJ, NUK, OILJ, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK, VSZLJ
To inform potential pathophysiological mechanisms of air pollution effects on cardiovascular disease (CVD), we investigated short-term associations between ambient air pollution and a range of ...cardiovascular events from three national databases in England and Wales.
Using a time-stratified case-crossover design, over 400,000 myocardial infarction (MI) events from the Myocardial Ischaemia National Audit Project (MINAP) database, over 2 million CVD emergency hospital admissions and over 600,000 CVD deaths were linked with daily mean concentrations of carbon monoxide (CO), nitrogen dioxide (NO2), particulate matter less than 10 μm in aerodynamic diameter (PM10), particulate matter less than 2.5 μm in aerodynamic diameter (PM2.5) and sulfur dioxide (SO2), and daily maximum of 8-hourly running mean of O3 measured at the nearest air pollution monitoring site to the place of residence. Pollutant effects were modelled using lags up to 4 days and adjusted for ambient temperature and day of week.
For mortality, no CVD outcome analysed was clearly associated with any pollutant, except for PM2.5 with arrhythmias, atrial fibrillation and pulmonary embolism. With hospital admissions, only NO2 was associated with a raised risk: CVD 1.7% (95% CI 0.9 to 2.6), non-MI CVD 2.0% (1.1 to 2.9), arrhythmias 2.9% (0.6 to 5.2), atrial fibrillation 2.8% (0.3 to 5.4) and heart failure 4.4% (2.0 to 6.8) for a 10th-90th centile increase. With MINAP, only NO2 was associated with an increased risk of MI, which was specific to non-ST-elevation myocardial infarction (non-STEMIs): 3.6% (95% CI 0.4 to 6.9).
This study found no clear evidence for pollution effects on STEMIs and stroke, which ultimately represent thrombogenic processes, though it did for pulmonary embolism. The strongest associations with air pollution were observed with selected non-MI outcomes.
ObjectivesSeveral observational studies have suggested an association between high temperatures and all-cause mortality. However, estimates on more specific mortality outcomes are sparse, and ...frequently assessed in studies using different analytical methods.MethodsA time series analysis was performed on 10 regions in England and Wales during the summers (June–September) of 1993–2006. Average percentage linear increases in risk for a 1°C increase in temperature above region-specific thresholds and attributable deaths were computed by cause-specific mortality and age groups (0–64, 65–74, 75–84, 85+).ResultsThere was evidence of increased mortality with heat for almost all cause-of-death groups examined, with an overall increase in all-cause mortality of 2.1% (95% CI 1.6% to 2.6%) for a 1°C rise above the regional heat threshold. Among main causes, the steepest increase in risk was for respiratory mortality (+4.1% (3.5% to 4.8%) per 1°C). It was much smaller for cardiovascular causes (+1.8% (1.2% to 2.5%)) and myocardial infarction (+1.1% (0.7% to 1.5%)), but comparatively high for arrhythmias (+5.0% (3.2% to 6.9%)) and pulmonary heart disease (+8.3% (2.7% to 14.3%)). Among non- cardiorespiratory causes, the strongest effects were for genitourinary (+3.8% (2.9% to 4.7%)) and nervous system (+4.6% (3.7% to 5.4%)) disorders. 33.9% of heat deaths were attributable to cardiovascular causes, 24.7% to respiratory causes and 41.3% to all other causes combined.ConclusionsThese results suggest that the risk of heat-related mortality is distributed across a wide range of different causes, and that targeting of preventative actions based on pre-existing disease is unlikely to be efficient.
Full text
Available for:
BFBNIB, CMK, NMLJ, NUK, PNG, SAZU, UL, UM, UPUK
Distributed lag non-linear models (DLNMs) are a modelling tool for describing potentially non-linear and delayed dependencies. Here, we illustrate an extension of the DLNM framework through the use ...of penalized splines within generalized additive models (GAM). This extension offers built-in model selection procedures and the possibility of accommodating assumptions on the shape of the lag structure through specific penalties. In addition, this framework includes, as special cases, simpler models previously proposed for linear relationships (DLMs). Alternative versions of penalized DLNMs are compared with each other and with the standard unpenalized version in a simulation study. Results show that this penalized extension to the DLNM class provides greater flexibility and improved inferential properties. The framework exploits recent theoretical developments of GAMs and is implemented using efficient routines within freely available software. Real-data applications are illustrated through two reproducible examples in time series and survival analysis.
Full text
Available for:
BFBNIB, DOBA, FSPLJ, FZAB, GIS, IJS, INZLJ, IZUM, KILJ, NLZOH, NMLJ, NUK, OILJ, PILJ, PNG, SAZU, SBCE, SBMB, UILJ, UKNU, UL, UM, UPUK, ZRSKP
PDF
