| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||
1 Geriatrics, University of Modena and Reggio Emilia, Italy.
2 Epidemiology and Statistical Service, District 3, Azienda Unità Sanitaria Locale of Modena, Italy.
3 Municipality of Modena, Demographic Service Department, Italy.
4 ARPA (Regional Agency for Environmental Prevention in Emilia-Romagna), Italy.
5 District 3, Azienda Unità Sanitaria Locale of Modena, Italy.
6 Federazione Italiana Medici Medicina Generale (FIMMG), Rome, Italy.
Address correspondence to Chiara Mussi, MD, University of Modena and Reggio Emilia, Nuovo Ospedale Civile S. Agostino-Estense, Via Giardini 1135, 41100 Baggiovara di Modena, Italy. E-mail: cmussi{at}iol.it
 |
A |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Methods. We performed a retrospective, case–control study of a cohort of 394 older persons living in Modena, 197 dead (cases) and 197 survivors (controls). A questionnaire to collect information about demographic, social, environmental, and clinical characteristics and about causes of death was completed.
Results. Cases were more likely to be living in a nursing home and needing assistance (p =.024, and p <.001, respectively). Survivors were living on higher level floors (p =.046). Spending the summer in Modena was significantly related to poor outcomes (p <.01). A higher number of cases were using public health services (p <.001). Individuals who died had a greater degree of comorbidity and dependence (p <.001); they were cognitively impaired (p <.001), took a larger number of drugs (p <.001), and had a greater number of hospital admissions (p <.001). Multivariate analysis showed that patients who spent the summer in Modena had a higher mortality. Other predictors of death were the use of home public-integrated assistance, a higher comorbidity and a higher degree of disability; the loss of at least 1 Activity of Daily Living (ADL) represents the strongest risk factor of heat-related death.
Conclusions. Our study identifies the major risk factors of heat-related death in the elderly population. With the creation of an up-to-date database, when a new heat wave will come, it will be possible to identify frail persons for preventive targeted strategies.
A heat wave is an isolated episode of hot weather lasting 
3 consecutive days with air temperature above 32°C (5). Heat waves have a much greater health effect in cities than in surrounding suburban and rural areas; first, air temperatures have higher values in densely built areas, a phenomenon widely known as an "urban heat island"; second, air pollution is usually higher in urban areas and, in some studies, has demonstrated a synergistic effect with heat on mortality (5).
It is well known that adverse climatic conditions have a negative effect on global health, leading to a higher mortality, particularly in large urban areas and most of all in the frail subgroups of the population (children and elderly and ill people). This effect becomes clear after a few days after exposure (6); the older population is at significant risk for hyperthermia and hypothermia; concomitant diseases, multiple medications, and altered perceptions increase the risk of developing complications during heat waves (7).
The available studies on heat-related death are designed to evaluate the excess of mortality in nonselected populations, often without using a control group (8). Owing to increased climatic variability, it is likely that in the future analogous periods will recur, in particular in continental areas near the sea. This likelihood underpins our study (the Argento Project) to define risk factors for heat-related death in a specifically geriatric population in Modena during August 2003 (the hottest month of the year), a study controlling for the usual effects of meteorologic and air pollution variables. The results of our analysis could help in refining targeted heat-wave-response activities in Italy to prevent future heat-related deaths among high-risk populations.
|
METHODS |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
A total of 394 persons living in Modena were recruited, 197 deaths (case group) and 197 survivors (control group). All the participants and when not available, relatives, gave informed consent for collection and the use of data.
For each person a questionnaire to collect information about demographic, social, environmental, and clinical characteristics was completed (Table 1). Mortality data were obtained from the National Bureau for Health Statistics (ISTAT) (4). The diagnoses were coded using the International Classification of Diseases, 9th Revision (ICD-9) (12). For both cases and controls, clinical, demographic, social, and environmental data were collected by phone, with an interview to family physicians; in case of missing data, we also phoned the relatives to obtain more precise information to complete the questionnaire and more insights on the circumstances of death. To confirm the information received and in case of missing data, we used the AUSL database.
|
|
|
When the index is 24, discomfort appears, increasing progressively; at 28 or more, a physical and cognitive deterioration of general conditions starts. The heat shock threshold is 32 (13). We did not analyze the rate of rainfall, as no abnormal data with respect to other periods of 2003 or previous years were observed in the city of Modena. Data on air pollution were collected by five pollution stations located in five different zones of the town. For each participant, values of different pollutants (medium and maximum nitric oxide, particulate matter <10 nanometers (PM10), and ozone, monthly and daily values) were recorded from the nearest station to his or her house.
Statistical Analysis
Data were analyzed with SPSS (11th version; SPSS, Chicago, IL). The chi-square test was used to compare proportions in univariate analyses of dichotomous variables and to calculate the odds ratios (OR) and the 95% confidence intervals (CI). The 
test was used to study ordinal variables. The variables significantly associated with the outcome of interest in univariate analyses were entered into a multivariate logistic regression model (backward stepwise) to assess their independent association with the outcome (mortality).
|
RESULTS |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
|
|
Spending the summer in Modena (rather than at holiday resorts) was significantly related to poor outcomes (OR = 4.24, 95% CI, 2.36–7.62; p <.01). There were no differences between the two groups as far as use of social services was concerned, but a significantly higher number of cases used public health services (home public assistance) (OR = 3.77, 95% CI, 2.22–6.39; p <.01).
Individuals who died had a complex clinical profile: They were more likely to have cognitive problems and a high degree of comorbidity and dependence. Moreover, they took a higher number of drugs and had a greater number of hospital admissions and specialist visits in the year before death (Table 3). The results described in Tables 2 and 3 did not change after excluding persons who spent the summer at holiday resorts.
|
|
|
In the multivariate model, living on a floor at higher level, in a nursing home or with private personnel, the presence of cognitive decline, the use of more than four drugs, and the number of hospital admissions in the previous years were not associated with higher mortality.
The results shown in Table 5 did not change after running the analysis excluding persons who spent the summer on holiday resorts.
|
DISCUSSION |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
In Modena, the hottest month of the 2003 summer was August; globally, the effects seem to have been very quick after the temperature increased, the excess of deaths coinciding with peaks of temperature or with a lag of 1 day and going down with a lag of a few days after a long period of very high temperatures (Figure 1); this phenomenon is in line with at least another study (3).
The problem of heat-related deaths was addressed several years ago, after the great heat wave in Chicago in 1995; researchers found a close relationship between the risk of summer deaths and poor clinical conditions, disability, loneliness, and the lack of air-conditioning facilities (14); individuals experienced multiorgan dysfunction with neurologic impairment, moderate to severe renal insufficiency, disseminated intravascular coagulation, and acute respiratory distress syndrome (15). Zhao and colleagues (16) showed that people who lived in poor social conditions have a significantly higher risk of summer death.
In the literature, the profile of the patient at risk of heat-related death is well known: Advanced age, preexisting diseases, loneliness, indigence, living conditions such as a small apartment on a high floor and no air conditioning are important risk factors. Nevertheless, when the effects of chronic debilitating diseases in the elderly population are minimized, their heat tolerance and thermoregulatory responses are comparable with those of younger individuals (17).
Our data recognized the clinical pattern of an increase in the number of deaths of elderly people during the hot summer: high comorbidity, disability, spending the summer in Modena, and home public assistance. Older people are at increased risk from heat death for a number of reasons: The ability to compensate for extreme heat and humidity becomes less efficient with aging, the presence of several chronic diseases is more likely (and their impact is exacerbated by heat); and some medications to treat chronic diseases make people more prone to health hazards from hot weather. In our study, survivors were living on higher floors: This finding is in contrast with the literature, because lower level floors are cooler. Nevertheless, survivors who lived on higher floors were more likely to be independent, as many houses lacked elevators. The same concept can be expressed as far as living alone is concerned: Elderly persons who live alone are more likely to be independent and in healthier living conditions.
Our study confirms the importance of high comorbidity and disability as factors significantly associated with a high risk of death; nevertheless, multivariate analysis shows, unexpectedly, that home public assistance is significantly related to higher mortality. We do not think that our data on this topic suggest underlying frailty not captured by other measures, or different socioeconomic status. In our opinion, this result challenges the quality and appropriateness of health interventions aimed at older and frail people living at home in big towns. In our study, living in a nursing home was associated with higher mortality only in univariate analysis. More precisely, living in a nursing home does not cause a higher mortality; rather, it is caused by the higher comorbidity and frailty of patients living in this setting. Another possible explanation of the higher mortality in persons who were assisted by home public assistance is the more frequent presence of air conditioning in nursing homes than in private houses; air conditioning has been recognized as a powerful protective factor to prevent heat-related mortality (18). Because of missing data, we did not analyze the effect of air conditioning on reducing mortality.
Epidemiological studies show that health impacts of the heat wave range far beyond hyperthermia. Global, cardiovascular, and respiratory mortality are the indicators most often mentioned to show a significant relationship with high temperatures; heat-related deaths (dehydration, hyperthermia, heat stroke) are rarely diagnosed clinically as being due to heat and are recorded as such (19,20). In our study, heat shock or other heat-related diseases are not identified as the principal causes of death; in contrast, typical age-related chronic diseases such as cardiovascular and neurologic pathologies have a great impact in causing heat-related deaths. The excess of cardiovascular diseases as a cause of mortality occurred mostly in hospital; cardiovascular diseases seem to be a good indicator of risk of death for elderly patients. Epidemiological studies show that health impacts of the heat wave range much further than hyperthermia alone.
Many studies suggest a synergic effect between ozone and/or particulate pollution and heat on deaths and a high contribution by such pollution to the final toll (21,22). Nevertheless, the relationship between mortality in summer during heat waves seems to be closer than that between mortality and other environmental factors, such as atmospheric pollution (23). In our study we failed to find correlations between mortality and concentrations of PM10, ozone, and nitric oxide. This is in line with other studies; in France air pollution, longer term temporal trends, and seasonal fluctuations did not appear as the main factors affecting mortality during the heat wave (19).
In the presence of a functioning Heat Health Watch Warning System the prevention plan can reach those people most at risk of heat-related death. Heat-related diseases are preventable with adequate warning and a fast and well-organized response to the emergency. Any successful intervention requires identification of elderly persons at high risk (6). Preventive intervention should focus on factors recognized to be important in heat tolerance: aerobic fitness, body fat, and body weight, for instance, are fundamental to maintain heat tolerance with aging (17).
Unlike many North American cities, European cities are not well-prepared to deal with the problem of heat waves; most of the programs do not include any intervention apart from issuing a passive warning to the general public or to the local public health agencies (24). In Philadelphia, the preventive program includes activation of a heatline, involvement of neighbors or relatives, home visits by field teams of the Department of Health, increased medical emergency staffing, and facilitation of access by elderly persons to air-conditioned service facilities (25).
With the creation of an up-to-date database including data about elderly people, it will be possible to identify frail persons and implement preventive targeted strategies that involve different professional figures (e.g., physicians, nurses, therapists, social assistants, volunteer associations). The evidence and anticipation of adverse health effects will indicate priorities for planned adaptive strategies and, crucially, will strengthen the case for preventive policies (26). Alert and response systems based on monitoring of climate-related risks, emergency room activity, and mortality should be implemented, and the response capacity of the social and health services improved; the above-mentioned database could make the health care system more effective in exceptional climatic conditions.
The major limit of this study is the retrospective design; prospective studies are needed to more precisely identify risk factors of heat-related death and to test the usefulness of preventive programs. Moreover, data were collected only during summer, so we did not have the opportunity to compare our results with other possible outcomes obtained in other periods of the year. Future research that compares risk factors for mortality and causes of death during hot months with those during cooler months could help to clarify this limitation.
|
A |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
We thank the interviewed participants and relatives, particularly the family physicians.
Author contributions: Micaela Foroni, acquisition of subjects and database creation; Gianfranco Salvioli, designer of the study and preparation of the manuscript; Rita Rielli, statistical analysis; Carlo Alberto Goldoni, Giuliano Orlandi, and Andrea Guerzoni, database creation; Stefano Zauli Sajani, meteorological data and statistical analysis; Cristina Maccaferri, acquisition of subjects and database; Ghassan Daya, acquisition of subjects; Chiara Mussi, statistical analysis, interpretation of data, preparation of the manuscript.
|
F |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Received May 3, 2006
Accepted September 16, 2006
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
