Impact of Temperature Change on Health

The influence of the winter North Atlantic Oscillation index on hospital admissions through diseases of the circulatory system in Lisbon, Portugal[1]

The aim of this paper is to analyze the relationship between North Atlantic Oscillation (NAO), meteorological variables, air pollutants, and hospital admissions due to diseases of circulatory systems in Lisbon (Portugal) during winter months (2003–2012). This paper is one of the few studies analyzing the impact of NAO on health through its influence on thermal stress and air pollution and is the first to be conducted in Lisbon. This study uses meteorological data (synthetized into a thermal comfort index), air pollutant metrics, and the NAO index (all clustered in 10-day cycles to overcome daily variability of the NAO index). The relationship between morbidity, thermal comfort index, NAO index, and air pollutants was explored through several linear models adjusted to seasonality through a periodic function. The possible indirect effect between the NAO index and hospital admissions was tested, assuming that NAO (independent variable) is affecting hospital admissions (outcome variable) through thermal discomfort and/or pollution levels (tested as individual mediators). This test was conducted through causal mediation analysis and adjusted for seasonal variation. The results from this study suggest a possible indirect relationship between NAO index and hospital admissions. Although NAO is not significantly associated with hospital admissions, it is significantly associated with CO, PM2.5, NO, and SO2 levels, which in turn increase the probability of hospitalization. The discomfort index (built with temperature and relative humidity) is significantly associated with hospital admissions, but its variability is not explained by the NAO index. This study highlights the impacts of the atmospheric circulation patterns on health. Furthermore, understanding the influence of the atmospheric circulation patterns can support the improvement of the existing contingency plans.

This study assesses the influence of NAO on health in Lisbon (during the winter months from 2003 to 2012). The results from this analysis show a possible indirect relationship between NAO and hospital admissions from circulatory diseases. This relationship is established through the indirect effect of pollution: NAO is associated with CO, PM2.5, NO, and SO2 levels, which in turn increase the probability of hospitalization. Hospital morbidity significantly increases with thermal stress (the DI is significantly associated with hospital morbidity due to diseases of the circulatory system diseases), although, thermal stress is not explained by the NAO index. Previous studies conducted in Europe have shown the effects of NAO on health. Messner et al. (2003) found a consistent positive relation between increasing NAO index and an increase in acute myocardial mortality in Sweden. Hubálek (2005) analyzed the impact of NAO on the incidence of some infectious diseases in the Czech Republic and found significant correlations between them. McGregor (2005) found statistically significant inverse associations between mortality from ischemic heart disease and the climate index representing the interaction between the NAO and temperature across England. Pausata et al. (2013), assessing the particulate matter variability induced by NAO in Europe during winter and the potential impact on human health, found that positive shift in the mean winter NAO of one standard deviation would lead to about 5500 additional premature deaths in Mediterranean countries due to the increase in particulate matter concentration.

Nevertheless, as described above, we found significant positive linear associations between the NAO index and CO, PM2.5, NO, and NO2. Similar results were also mentioned by Jerez et al. (2013) and Christoudias et al. (2012) which reported lower concentration of pollutants in southern Europe during NAO negative phases. These results show the processes of transport and deposition of air pollutants through the effect of wind and precipitation.

Air pollutants are positively associated with hospital admissions (except O3 and NO2); previous studies analyzing the health impact of several air pollutants also reported important differences between them: Forastiere et al. (2005), in Rome, identified significant increases of out-of-hospital coronary deaths with CO and PM10, but not with NO2; Cendon et al. (2006) found stronger positive correlations between SO2 and daily hospitalizations for myocardial infarctions than between CO, O3, NO2, and PM10.

O3 is negatively associated with hospital admissions; however, there is no causal relationship between the decreasing levels of ozone and the hospital admission increase. This apparent protective effect results from the O3 increase on warm days, while the relation between hospital admissions and temperature is inverse (Moolgavkar et al. 1995, Ito et al. 2005, Medina-Ramón et al. 2006). Similar results were found in the previous research (Moolgavkar et al. 1995, Medina-Ramón et al. 2006).

The thermal comfort index is not significantly associated with the NAO index. Unlike the countries of Central and North Europe where a strong association is found between the NAO index and temperature (Osborn et al. 1999, Trigo et al. 2002, Hurrell et al. 2003), in Lisbon, this relationship was not recorded. The results of this study are supported by the study of Ulbrich et al. (2012), showing that the relationship between the NAO index and temperature is not linear in the Iberian Peninsula.

Summarizing, hospital morbidity is positively associated with the pollutant levels (except ozone and NO2, as reported previously) and is negatively associated with the thermal comfort index, although no significant direct association with NAO was found. A similar pattern was also found in previous studies addressing vulnerability to cold weather (Almendra et al. 2012, 2016, Vasconcelos et al. 2013) or exposure to high levels of air pollution (Borrego et al. 2009, Slezakova et al. 2011) in Portugal.

Strengths and limitations

This study is one of the few analyzing the impact of NAO on health and is the first to be conducted in Lisbon. Considering the geographical and socioeconomic context of Portugal and the high vulnerability to harmful environmental conditions is fundamental to have a better understanding of the relationship between atmospheric conditions and health to effectively assess environmental risks. Thus, this study represents an important contribution to the current body of literature.

However, the results of this study must be interpreted with caution. Time series analyses were carried out for one location as such; the results should not be derived to other regions with different geographic and socioeconomic frameworks.

The methods applied in this study tested for direct and indirect linear associations; however, the relationship between environmental conditions and health is often studied by nonlinear modeling providing better fitting models. Therefore, the linear mediation model is an alternative to (direct) nonlinear models and can also be employed as an exploratory and complementary tool to those models.

Conclusions

This study investigated the effects of NAO on emergency hospital admissions from diseases of the circulatory system during the winter months in Lisbon. It was found that the NAO influences human health through its impacts on atmospheric pollutants. Positive NAO phases are associated with higher levels of air pollutants. No significant association was found between NAO and the discomfort index (built with temperature and relative humidity).

Although it is not possible to extrapolate from this to other countries or other areas of Portugal, this study draws attention to the impacts of the patterns of atmospheric circulation in the North Atlantic on human health and to the vulnerability to environmental factors.

This article can provide insights to improve public health policies and alert systems. A better understanding of the relationship between the NAO and health can help improve existing contingency plans, develop more effective adaptation strategies, and ensure they are put into action in a timely manner, thereby helping to decrease the health impacts of harmful environmental conditions.

Temperature effects on health – current findings and future implications[2]

Heat waves and cold spells have both shown adverse effects on mortality. Moreover, a recent study by Gasparrini and colleagues estimated that 7.7% of the mortality was attributable to non-optimum temperature using data from 384 locations (Gasparrini et al. 2015). Cold was responsible for a higher proportion of deaths than was heat, while moderate high and low temperatures represented most of the total health burden. Although forecasting studies suggest the passage of (summertime) cold fronts will diminish in frequency in a warmer climate, this will not per se mean that cold effects will only have a very small effect on population health. Studies have already shown that health effects associated with temperature decreases in winter and summer can be similar in magnitude but are more pronounced in years with higher average temperatures. Therefore, the influence of unexpected temperature changes may be more relevant than the absolute temperature level itself (Shi et al., 2015, Wolf et al., 2009).

Populations worldwide are rapidly ageing. The age group 60 years and older is expected to comprise 21.1% of the population by 2050 (http://www.un.org/en/development/desa/population/publications/ageing/WorldPopulationAgeingReport2013.shtml). This increase in the elderly is expected to increase future temperature-related mortality and morbidity.

Recent reviews have contributed to the evidence on the association between air temperature and mortality and morbidity (Aström et al., 2011, Franchini and Mannucci, 2015, Phung et al., 2016, Ryti et al., 2016, Turner et al., 2012, Yu et al., 2012). However, these reviews were limited by either a focus on only heat or cold, a specific outcome, or did not include a meta-analysis. Only two of them presented age-specific results.

The purpose of the systematic review and meta-analyses by Bunker et al. was to present quantitative evidence on the effects of non-optimum high and low ambient temperatures on a range of cause-specific mortality and morbidity outcomes in the elderly (Bunker et al. 2016). Heat waves and cold spells were excluded because they are unique events with differing characteristics. Epidemiological time-series and case-crossover studies reporting quantitative associations between temperature and cause-specific, elderly mortality or morbidity (i.e. hospitalization, emergency room admissions, general practice visits, home visits) were considered.

The authors identified substantially elevated risks in the elderly for temperature-induced cerebrovascular, cardiovascular, and respiratory outcomes in particular. In their meta-analysis for morbidity, the authors showed that the effect estimates for respiratory causes were much larger than for cardiovascular causes with both, high and low temperatures – although for mortality, the effect estimates for cardiovascular causes were similar or slightly larger than for respiratory causes in case of high temperatures. This phenomenon has been already shown in previous studies (e.g. Michelozzi et al. 2009). However, the underlying mechanisms through which high temperatures may increase the risk of morbidity from respiratory causes are yet unclear. In addition, Bunker et al. found an increased risk for heat-induced diabetes, renal, and infectious disease morbidity all of which are likely to increase further with climate change and global aging.

The authors had to pool across different lags and threshold temperatures of the various studies. Therefore, the important fact that the effects of heat are most often immediate while the ones of cold become predominant with longer time lags is not obvious in their pooled effects tables. This limitation was overcome by sensitivity analyses (e.g. for certain lags) where possible.

The systematic review and meta-analysis by Bunker et al. is timely and helpful for pointing out further research needs and possible public health interventions especially for the elderly.

Climate change does not just affect air temperatures, but also a lot of other meteorological variables which might be as important to human health as temperature changes, such as humidity, barometric pressure and precipitation as well as UV-radiation. In addition, it is still unclear why outdoor temperature changes show consistent health effects across all geographical regions despite the fact that especially the elderly spend most of their time indoors. Moreover, the interplay with air pollution is quite complex and new multicenter studies as well as statistical approaches are needed. With an aging population, mental health and cognitive function in particular is an emerging field of concern (Lacruz et al. 2010) in which environmental stressors certainly play a role.

According to the European Environment Agency (http://www.eea.europa.eu/soer/synthesis/synthesis), about 75% of the European population lives in urban areas and is projected to increase to 80% by 2020. Urbanization has resulted in improved access to education, employment and health care. However, it also changed living conditions and influences lifestyles, behaviors and environmental conditions. Urban heat islands were identified as contributing significantly to the health impact of the 2003 heat wave in Paris (Laaidi et al. 2012). Further, it has been projected that a reduction of the green cover by 10% would increase urban temperatures by 8.2°C over the next 70 years (Gill et al. 2007) – highlighting the potential for improving health and well-being by measures improving the built environment.