To the Editors:
The dramatic volcanic eruption in Iceland has recently brought the possible health effects of volcanic ashes to the general attention of the media. Lessons learnt from other episodes may be of interest and could guide future public health action.
During Autumn 2002, a large eruption of Mount Etna (Sicily, Italy) occurred. It persisted for 3 months, showing peaks of intense volcanic activity characterised by explosions, jets hundreds of metres high and seismic activity. Local winds carried the ash columns for kilometres, causing discomfort to the population of Catania (nearly 370,000 inhabitants) and other inhabited areas near the volcano. Repeated episodes of rainfall occurred after the eruption, reducing the phenomenon of ash resuspension. The population was alerted by a general alarm caused by national and local mass media. National and local authorities enacted measures aimed at preventing health effects: schools were closed; motor vehicle circulation was prohibited; protective eye masks were distributed; the elderly and individuals affected by chronic diseases of the respiratory system were suggested to limit their outdoor exposure and to reduce physical activity.
An epidemiologic study 1 on acute health effects on the nearby population was conducted. During the study period (106 days of observation in 2002; 59 days of eruption and 47 days of noneruption) unusually high levels (>200 μg·m−3) of particulate matter with a 50% cut-off aerodynamic diameter <10 μm (PM10) were recorded (fig. 1). A 284% increase in PM10 and 385% increase in daily levels of sulfur dioxide (SO2) were observed with respect to the previous year. No changes in nitrogen dioxide (NO2) and carbon monoxide (CO) were detected. Earthquakes 2 started 1 day before the onset of the eruption, with at least two earthquakes per day (above magnitude 3) during the first week of eruption. As shown in figure 1, the intensity of seismic activity was highly correlated with PM10 levels.
Both mortality and hospital admissions data were collected for all residents of the city during eruption and noneruption days in 2002, using the corresponding days in 2001 as a control. Relative risks (RR) of health events during the eruption with respect to the control period were computed both for mortality and for hospital admissions, shifted 1 day forward in order to take into account the day of the week effect. No changes in all-cause (RR 0.98, 95% CI 0.86–1.12) and cardiovascular (RR 1.09, 95% CI 0.90–1.32) mortality were observed, whereas a significant decrease in mortality for respiratory diseases during the eruption period was detected (RR 0.46, 95% CI 0.46–0.82). Hospital admissions for cardiovascular diseases significantly increased (RR 1.18, 95% CI 1.08–1.29), particularly for ischaemic heart diseases (RR 1.31, 95% CI 1.10–1.56) and for cerebrovascular diseases (RR 1.24, 95% CI 1.05–1.47), especially among people aged >65 yrs. Also notable, although not statistically significant, were admissions for eye disorders, two times higher during the eruption period than in the previous year. However, a decrease in hospitalisations for trauma was observed (RR 0.9, 95% CI 0.8–1.0).
To evaluate the time relationship between daily PM10 levels and daily hospital admissions for cardiovascular diseases, a time-series analysis with Poisson regression was conducted (considering day of the week and holidays as confounders) and evaluating different time lags for PM10 concentrations ≤14 days. No statistically significant association appeared, although the effect increased with time lags suggesting a delayed effect on cardiovascular health.
Studies on health hazards related to volcanoes 3 have reported cardiovascular and respiratory problems possibly related to particulate matter and/or to emitted gases. In fact, volcanic ash is generally composed of particles of fragmented volcanic rock, most of which are within the coarse fraction range (2.5–10 μm). Other minor or temporary health effects have been reported in individuals who stay very close to craters for occupational reasons and in areas with ash contamination 4. In our study, we observed an increase in PM10 and SO2 following the eruption, but not in NO2 or CO. The relationship between daily PM10 level and cardiovascular admissions following Mount Etna's eruption shows that the effect was not immediate but, rather, delayed. Given the collinearity between the eruption and the earthquake, it is difficult to disentangle the responsible factor. However, previous studies strongly suggest that the stress was caused by the eruption and the frequent earthquakes experienced by residents could be responsible for our results. Reviews that have focused on the effects on cardiovascular diseases of unanticipated catastrophic natural event-related stress, such as earthquakes 5 or terrorist attacks 6, describe cardiac events as “indirect effects” of natural disasters, detectable in the long term 7, 8. Stress-related mortality from cardiovascular diseases was also reported among residents who experienced the Seveso dioxin accident 9 and in Los Angeles, CA (USA) during the 2 weeks before the turn of the millennium 10.
In our study, the preventive measures recommended to the population during the emergency period could explain the observed reduction of both the mortality for respiratory diseases and the decrease in hospital admissions for trauma: it is likely that people spent more time at home than during the control period, preventing car accidents and other minor trauma.
In conclusion, no increase in mortality was observed during Mount Etna's 2002 eruption. The temporary increase in cardiovascular morbidity, observed particularly among the elderly, may be a consequence of acute stress, as reported following other natural disasters. The decrease in mortality for respiratory diseases and the decrease in hospital admissions for trauma were likely a result of the precautionary measures taken during the emergency period. In any case, it is recommended that in future episodes, physicians closely monitor high-risk cardiac patients following natural disasters. Moreover, when planning public health measures during natural disasters, special attention should be given to risk communication and the consequences of general media alarm should not be neglected.
The authors wish to thank S. D’Amico (National Institute of Geophysics and Volcanology of Catania, Catania, Italy) for his help in the selection of the meaningful information on earthquakes, and M. Becker for her editorial help.
Statement of Interest
- ©ERS 2010