A time series approach for evaluating intra-city heat-related mortality

Small-area assessment of temperature-related mortality risks in England and Wales: a case time series analysis

BACKGROUND

Epidemiological literature on the health risks associated with non-optimal temperature has mostly reported average estimates across large areas or specific population groups. However, the heterogeneous distribution of drivers of vulnerability can result in local differences in health risks associated with heat and cold. We aimed to analyse the association between ambient air temperature and all-cause mortality across England and Wales and characterise small scale patterns in temperature-related mortality risks and impacts.

METHODS

We performed a country-wide small-area analysis using data on all-cause mortality and air temperature for 34 753 lower super output areas (LSOAs) within 348 local authority districts (LADs) across England and Wales between Jan 1, 2000, and Dec 31, 2019. We first performed a case time series analysis of LSOA-specific and age-specific mortality series matched with 1 × 1 km gridded temperature data using distributed lag non-linear models, and then a repeated-measure multivariate meta-regression to pool LAD-specific estimates using area-level climatological, socioeconomic, and topographical predictors.

FINDINGS

The final analysis included 10 716 879 deaths from all causes. The small-area assessment estimated that each year in England and Wales, there was on average 791 excess deaths (empirical 95% CI 611-957) attributable to heat and 60 573 (55 796-65 145) attributable to cold, corresponding to standardised excess mortality rates of 1·57 deaths (empirical 95% CI 1·21-1·90) per 100 000 person-years for heat and 122·34 deaths (112·90-131·52) per 100 000 person-years for cold. The risks increased with age and were highly heterogeneous geographically, with the minimum mortality temperature ranging from 14·9°C to 22·6°C. Heat-related mortality was higher in urban areas, whereas cold-related mortality showed a more nuanced geographical pattern and increased risk in areas with greater socioeconomic deprivation.

INTERPRETATION

This study provides a comprehensive assessment of excess mortality related to non-optimal outdoor temperature, with several risk indicators reported by age and multiple geographical levels. The analysis provides detailed risk maps that are useful for designing effective public health and climate policies at both local and national levels.

FUNDING

Medical Research Council, Natural Environment Research Council, EU Horizon 2020 Programme, National Institute of Health Research.

Outdoor ambient temperatures and occupational injuries and illnesses: Are there risk differences in various regions within a city?

Increased risk of occupational injuries and illnesses (OI) is associated with hot ambient temperatures. However, the existing evidence of risk estimation is limited to large regions at the city or provincial scales. For effective and localized occupational health risk management, spatio-temporal analysis should be carried out at the intra-city level to identify high-risk areas within cities. This study examined the exposure-response relationship between ambient temperatures and OI at the intra-city scale in Greater Adelaide, Australia. Vulnerable groups of workers, in terms of workers' characteristics, the nature of their work, and workplace characteristics were identified. Further, the projected risk of OI was quantified in various climate change scenarios. The temperature-OI association was estimated using a time-series study design combined with Distributed Lag Non-linear Models. Daily workers' compensation claims (2005-2018) were merged with 5 km gridded meteorological data of maximum temperature (°C) at Statistical Area Level 3 in Greater Adelaide. Region-wise subgroup analyses were conducted to identify vulnerable groups of workers. Future projections (2006-2100) were conducted using downscaled climate projections and the risk was quantified using log-linear extrapolation. The analyses were performed in R 4.1.0. The overall OI risk was 16.7% (95%CI: 10.8-23.0) at moderate heat (90th percentile) and increased to 25.0% (95%CI: 16.4-34.2) at extreme heat (99th percentile). Northern Adelaide had a higher risk of OI for all types of workers at moderate heat, while western regions had a high risk for indoor industries. Southern and eastern regions had a higher OI risk for males, older workers, and outdoor industries at extreme heat. The projected risk of OI is estimated to increase from 20.8% (95%CI: -0.2-46.3) in 2010s to 22.9% (95%CI: -8.0-64.1) by 2050s. Spatio-temporal risk assessment at the intra-city scale can help us identify high-risk areas, where targeted interventions can be efficiently employed to reduce the socio-economic burden of OI.

Exposure to Abnormally Hot Temperature and the Demand for Commercial Health Insurance

Using the China Health and Retirement Longitudinal Study, this paper studies the impact of abnormal hot temperature on residents' demand for commercial health insurance. The results show that for every 1°F rise in abnormal temperature, the probability of people buying commercial health insurance increased by 6%. Furthermore, the abnormal hot temperature has a more significant impact on the commercial health insurance demand of women, residents in the South and residents in the East. Channel analysis shows that abnormal hot temperature affects the demand for commercial health insurance through two channels: increasing residents' concern about climate risk and affecting health. This paper provides evidence for actively promoting sustainable development and improving the construction of medical security system.

Geospatial context of social and environmental factors associated with health risk during temperature extremes: Review and discussion

This study reviews forty-six publications between 2008 and 2017 dealing with socio-environmental impacts on adverse health effects of temperature extremes, in a geospatial context. The review showed that most studies focus on extremely hot weather but lack analysis of how spatial heterogeneity across a region can influence cold mortality/morbidity. There are limitations regarding the use of temperature datasets for spatial analyses. Only a few studies have applied air temperature datasets with high spatial resolution to health studies, but none of these studies have used anthropogenic heat as a factor for analysis of health risk. In addition, the elderly is generally recognized as a vulnerable group in most studies, but the interaction between old age and temperature risk varies by location. Other socio-demographic factors such as low income, low education and accessibility to community shelters may also need to be considered in the future. There are only a few studies which investigate the interaction between temperature and air pollution in a geospatial context, despite the fact that this is a known interaction that can influence health risk under extreme weather. In conclusions, although investigation of temperature effects on health risk is already at the "mature stage", studies of socio-environmental influences on human health under extreme weather in a geospatial context is still being investigated. A comprehensive assessment is required to analyse how the spatial aspects of the geophysical and social environments can influence human health under extreme weather, in order to develop a better community plan and health protocols for disaster preparedness.

Defining heat waves and extreme heat events using sub-regional meteorological data to maximize benefits of early warning systems to population health

BACKGROUND

Extreme heat events have been consistently associated with an increased risk of hospitalization for various hospital diagnoses. Classifying heat events is particularly relevant for identifying the criteria to activate early warning systems. Heat event classifications may also differ due to heterogeneity in climates among different geographic regions, which may occur at a small scale. Using local meteorological data, we identified heat waves and extreme heat events that were associated with the highest burden of excess hospitalizations within the County of San Diego and quantified discrepancies using county-level meteorological criteria.

METHODS

Eighteen event classifications were created using various combinations of temperature metric, percentile, and duration for both county-level and climate zone level meteorological data within San Diego County. Propensity score matching and Poisson regressions were utilized to ascertain the association between heat wave exposure and risk of hospitalization for heat-related illness and dehydration for the 1999-2013 period. We estimated both relative and absolute risks for each heat event classification in order to identify optimal definitions of heat waves and extreme heat events for the whole city and in each climate zone to target health impacts.

RESULTS

Heat-related illness differs vastly by level (county or zone-specific), definition, and risk measure. We found the county-level definitions to be systematically biased when compared to climate zone definitions with the largest discrepancy of 56 attributable hospitalizations. The relative and attributable risks were often minimally correlated, which exemplified that relative risks alone are not adequate to optimize heat waves definitions.

CONCLUSIONS

Definitions based on county-level defined thresholds do not provide an accurate picture of the observed health effects and will fail to maximize the potential effectiveness of heat warning systems. Absolute rather than relative risks are a more appropriate measure to define the set of criteria to activate early warnings systems and thus maximize public health benefits.

High Summer Temperatures and Mortality in Estonia

BACKGROUND

On-going climate change is predicted to result in a growing number of extreme weather events-such as heat waves-throughout Europe. The effect of high temperatures and heat waves are already having an important impact on public health in terms of increased mortality, but studies from an Estonian setting are almost entirely missing. We investigated mortality in relation to high summer temperatures and the time course of mortality in a coastal and inland region of Estonia.

METHODS

We collected daily mortality data and daily maximum temperature for a coastal and an inland region of Estonia. We applied a distributed lag non-linear model to investigate heat related mortality and the time course of mortality in Estonia.

RESULTS

We found an immediate increase in mortality associated with temperatures exceeding the 75th percentile of summer maximum temperatures, corresponding to approximately 23°C. This increase lasted for a couple of days in both regions. The total effect of elevated temperatures was not lessened by significant mortality displacement.

DISCUSSION

We observed significantly increased mortality in Estonia, both on a country level as well as for a coastal region and an inland region with a more continental climate. Heat related mortality was higher in the inland region as compared to the coastal region, however, no statistically significant differences were observed. The lower risks in coastal areas could be due to lower maximum temperatures and cooling effects of the sea, but also better socioeconomic condition. Our results suggest that region specific estimates of the impacts of temperature extremes on mortality are needed.

Relationships between maximum temperature and heat-related illness across North Carolina, USA

Heat kills more people than any other weather-related event in the USA, resulting in hundreds of fatalities each year. In North Carolina, heat-related illness accounts for over 2,000 yearly emergency department admissions. In this study, data on emergency department (ED) visits for heat-related illness (HRI) were obtained from the North Carolina Disease Event Tracking and Epidemiologic Collection Tool to identify spatiotemporal relationships between temperature and morbidity across six warm seasons (May-September) from 2007 to 2012. Spatiotemporal relationships are explored across different regions (e.g., coastal plain, rural) and demographics (e.g., gender, age) to determine the differential impact of heat stress on populations. This research reveals that most cases of HRI occur on days with climatologically normal temperatures (e.g., 31 to 35 °C); however, HRI rates increase substantially on days with abnormally high daily maximum temperatures (e.g., 31 to 38 °C). HRI ED visits decreased on days with extreme heat (e.g., greater than 38 °C), suggesting that populations are taking preventative measures during extreme heat and therefore mitigating heat-related illness.

Multiple Trigger Points for Quantifying Heat-Health Impacts: New Evidence from a Hot Climate

BACKGROUND

Extreme heat is a public health challenge. The scarcity of directly comparable studies on the association of heat with morbidity and mortality and the inconsistent identification of threshold temperatures for severe impacts hampers the development of comprehensive strategies aimed at reducing adverse heat-health events.

OBJECTIVES

This quantitative study was designed to link temperature with mortality and morbidity events in Maricopa County, Arizona, USA, with a focus on the summer season.

METHODS

Using Poisson regression models that controlled for temporal confounders, we assessed daily temperature-health associations for a suite of mortality and morbidity events, diagnoses, and temperature metrics. Minimum risk temperatures, increasing risk temperatures, and excess risk temperatures were statistically identified to represent different "trigger points" at which heat-health intervention measures might be activated.

RESULTS

We found significant and consistent associations of high environmental temperature with all-cause mortality, cardiovascular mortality, heat-related mortality, and mortality resulting from conditions that are consequences of heat and dehydration. Hospitalizations and emergency department visits due to heat-related conditions and conditions associated with consequences of heat and dehydration were also strongly associated with high temperatures, and there were several times more of those events than there were deaths. For each temperature metric, we observed large contrasts in trigger points (up to 22 °C) across multiple health events and diagnoses.

CONCLUSION

Consideration of multiple health events and diagnoses together with a comprehensive approach to identifying threshold temperatures revealed large differences in trigger points for possible interventions related to heat. Providing an array of heat trigger points applicable for different end-users may improve the public health response to a problem that is projected to worsen in the coming decades.

Heat and Humidity in the City: Neighborhood Heat Index Variability in a Mid-Sized City in the Southeastern United States

Daily weather conditions for an entire city are usually represented by a single weather station, often located at a nearby airport. This resolution of atmospheric data fails to recognize the microscale climatic variability associated with land use decisions across and within urban neighborhoods. This study uses heat index, a measure of the combined effects of temperature and humidity, to assess the variability of heat exposure from ten weather stations across four urban neighborhoods and two control locations (downtown and in a nearby nature center) in Knoxville, Tennessee, USA. Results suggest that trees may negate a portion of excess urban heat, but are also associated with greater humidity. As a result, the heat index of locations with more trees is significantly higher than downtown and areas with fewer trees. Trees may also reduce heat stress by shading individuals from incoming radiation, though this is not considered in this study. Greater amounts of impervious surfaces correspond with reduced evapotranspiration and greater runoff, in terms of overall mass balance, leading to a higher temperature, but lower relative humidity. Heat index and relative humidity were found to significantly vary between locations with different tree cover and neighborhood characteristics for the full study time period as well as for the top 10% of heat index days. This work demonstrates the need for high-resolution climate data and the use of additional measures beyond temperature to understand urban neighborhood exposure to extreme heat, and expresses the importance of considering vulnerability differences among residents when analyzing neighborhood-scale impacts.

Heterogeneity in individually experienced temperatures (IETs) within an urban neighborhood: insights from a new approach to measuring heat exposure

Urban environmental health hazards, including exposure to extreme heat, have become increasingly important to understand in light of ongoing climate change and urbanization. In cities, neighborhoods are often considered a homogenous and appropriate unit with which to assess heat risk. This manuscript presents results from a pilot study examining the variability of individually experienced temperatures (IETs) within a single urban neighborhood. In July 2013, 23 research participants were recruited from the South End neighborhood of Boston and equipped with Thermochron iButtons that measured the air temperatures surrounding individuals as they went about their daily lives. IETs were measured during a heat wave period (July 17-20), which included 2 days with excessive heat warnings and 1 day with a heat advisory, as well as a reference period (July 20-23) in which temperatures were below seasonal averages. IETs were not homogeneous during the heat wave period; mean IETs were significantly different between participants (p < 0.001). The majority of participants recorded IETs significantly lower than outdoor ambient temperatures (OATs), and on average, the mean IET was 3.7 °C below the mean OAT. Compared with IETs during the reference period, IETs during the heat wave period were 1.0 °C higher. More than half of participants did not experience statistically different temperatures between the two test periods, despite the fact that the mean OAT was 6.5 °C higher during the heat wave period. The IET data collected for this sample and study period suggest that (1) heterogeneity in individual heat exposure exists within this neighborhood and that (2) outdoor temperatures misrepresent the mean experienced temperatures during a heat wave period. Individual differences in attributes (gender, race, socioeconomic status, etc.), behaviors (schedules, preferences, lifestyle, etc.), and access to resources are overlooked determinants of heat exposure and should be better integrated with group- and neighborhood-level characteristics. Understanding IETs for the population at large may lead to innovative advances in heat-health intervention and mitigation strategies.

Geographic dimensions of heat-related mortality in seven U.S. cities

Spatially targeted interventions may help protect the public when extreme heat occurs. Health outcome data are increasingly being used to map intra-urban variability in heat-health risks, but there has been little effort to compare patterns and risk factors between cities. We sought to identify places within large metropolitan areas where the mortality rate is highest on hot summer days and determine if characteristics of high-risk areas are consistent from one city to another. A Poisson regression model was adapted to quantify temperature-mortality relationships at the postal code scale based on 2.1 million records of daily all-cause mortality counts from seven U.S. cities. Multivariate spatial regression models were then used to determine the demographic and environmental variables most closely associated with intra-city variability in risk. Significant mortality increases on extreme heat days were confined to 12-44% of postal codes comprising each city. Places with greater risk had more developed land, young, elderly, and minority residents, and lower income and educational attainment, but the key explanatory variables varied from one city to another. Regression models accounted for 14-34% of the spatial variability in heat-related mortality. The results emphasize the need for public health plans for heat to be locally tailored and not assume that pre-identified vulnerability indicators are universally applicable. As known risk factors accounted for no more than one third of the spatial variability in heat-health outcomes, consideration of health outcome data is important in efforts to identify and protect residents of the places where the heat-related health risks are the highest.

The effect of heat waves on mortality in susceptible groups: a cohort study of a mediterranean and a northern European City

BACKGROUND

Climate change is projected to increase the number and intensity of extreme weather events, for example heat waves. Heat waves have adverse health effects, especially for the elderly, since chronic diseases are more frequent in that group than in the population overall. The aim of the study was to investigate mortality during heat waves in an adult population aged 50 years or over, as well as in susceptible subgroups of that population in Rome and Stockholm during the summer periods from 2000 to 2008.

METHODS

We collected daily number of deaths occurring between 15th May and 15th September each year for the population above 50 as well as the susceptible subgroups. Heat wave days were defined as two or more days exceeding the city specific 95th percentile of maximum apparent temperature (MAT). The relationship between heat waves and all-cause non-accidental mortality was investigated through time series modelling, adjusting for time trends.

RESULTS

The percent increase in daily mortality during heat waves as compared to normal summer days was, in the 50+ population, 22% (95% Confidence Interval (CI): 18-26%) in Rome and 8% (95% CI: 3-12%) in Stockholm. Subgroup specific increase in mortality in Rome ranged from 7% (95% CI:-17-39%) among survivors of myocardial infarction to 25% in the COPD (95% CI:9-43%) and diabetes (95% CI:14-37%) subgroups. In Stockholm the range was from 10% (95% CI: 2-19%) for congestive heart failure to 33% (95% CI: 10-61%) for the psychiatric subgroup.

CONCLUSIONS

Mortality during heat waves increased in both Rome and Stockholm for the 50+ population as well as in the considered subgroups. It should be evaluated if protective measures should be directed towards susceptible groups, rather than the population as a whole.

Assessing variability in the impacts of heat on health outcomes in New York City over time, season, and heat-wave duration

While the impacts of heat upon mortality and morbidity have been frequently studied, few studies have examined the relationship between heat, morbidity, and mortality across the same events. This research assesses the relationship between heat events and morbidity and mortality in New York City for the period 1991-2004. Heat events are defined based on oppressive weather types as determined by the Spatial Synoptic Classification. Morbidity data include hospitalizations for heat-related, respiratory, and cardiovascular causes; mortality data include these subsets as well as all-cause totals. Distributed-lag models assess the relationship between heat and health outcome for a cumulative 15-day period following exposure. To further refine analysis, subset analyses assess the differences between early- and late-season events, shorter and longer events, and earlier and later years. The strongest heat-health relationships occur with all-cause mortality, cardiovascular mortality, and heat-related hospital admissions. The impacts of heat are greater during longer heat events and during the middle of summer, when increased mortality is still statistically significant after accounting for mortality displacement. Early-season heat waves have increases in mortality that appear to be largely short-term displacement. The impacts of heat on mortality have decreased over time. Heat-related hospital admissions have increased during this time, especially during the earlier days of heat events. Given the trends observed, it suggests that a greater awareness of heat hazards may have led to increased short-term hospitalizations with a commensurate decrease in mortality.