The Holuhraun volcanic eruption (2014–2015) was the largest Icelandic eruption since Laki (1783–1784) (1). It emitted ≥11 million tons of SO2 (2). At low concentrations, SO2 exposure causes eye and respiratory tract irritation and, at high levels, triggers acute asthma symptoms, pulmonary oedema and respiratory distress in susceptible individuals. In epidemiological studies, urban SO2 was associated with increases in respiratory morbidity and mortality, and PM10 and cold weather increases the effects on asthmatic individuals (3). In volcanic environments, SO2 and acid aerosols are associated with increased risk of respiratory symptoms, (4) morbidity, and mortality (5). Occupational SO2 exposure is associated with persisting lung injury (6), bronchoconstriction and reactive airways dysfunction syndrome (RADS), ie, persisting asthma symptoms appearing within 24 hours of high exposure events. RADS overlaps somewhat with irritant-induced asthma, which is caused by exposure to lower SO2 concentrations where symptom-onset is up to a week from exposure (7). For SO2, the 10-minute occupational health limit is 500 μg/m3 (3), although UK occupational guidelines allow exposure of ≤13mg/m3 (5 ppm) for 15 minutes (8)
The objective of this study was to determine if exposure to volcanic SO2 was associated with changes in lung function, lung inflammation, and symptoms among professionals working at the Holuhraun eruption site.
Methods
The Holuhraun volcanic eruption began in a remote area of Iceland on 31 August 2014 and ended 27 February 2015. Personnel from the Meteorological Institute, the Earth Science Institute of the University of Iceland, and the Civil Protection Agency who worked at, or near, the Holuhraun eruption site either monitoring, measuring, or conducting safety activities were invited to participate in a health examination at either Landspitali – the National University Hospital in Reykjavík – or Akureyri Hospital in North Iceland from 31 January 2015 to 31 March 2015. During that time, 41 employees visited the eruption site, after which (being exposed) they were asked to attend a follow-up clinical examination as soon as possible. Exposure time was 4–5 days at a time, followed by a period of ≥4 days of non-exposure. Personnel working at the eruption site were instructed to wear gas protection (either half- or full-face) masks with the appropriate gas filters when personal SO2 monitors indicated SO2 levels ≥1 ppm. Masks were fitted and personnel was advised to keep facial hair to a minimum. SO2 monitors worn at Holuhraun measured 10-minute SO2 concentrations as high as 100 mg/m3 (35.59 ppm) and median daily 10-minute SO2 concentration was 360 μg/m3 (0.128 ppm).
At the baseline health examination, participants answered a questionnaire about background respiratory health (9), current symptoms, and symptoms during exposure. Lung function was measured using EasyOne spirometer according to a standardized protocol (10). Furthermore, the fraction of NO in exhaled air (FeNO) was measured using an electrochemical analyzer, NIOX MINO, among participants in Reykjavík. At the follow-up examination, testing was repeated and participants were asked again about current symptoms, symptoms during exposure, and to report any change in their health in free-text replies. Mixed linear models and McNemar‘s test were used to analyze if lung function outcomes or symptom rates were significantly different before and after exposure, analyses were performed in R Studio.
Results
Ultimately, 32 individuals (participation rate 78%) were enrolled into the study, answered the baseline questionnaires and underwent clinical examination. All had normal spirometry and FeNO was within the normal range. No one reported having underlying diseases. After visiting the eruption site, 17 individuals (participation rate 53%) were followed-up 1–6 (median 3) days after eruption site exposure and included in the study group. In the study group, 2 were smokers (12%), and 3 (18%) had doctor-diagnosed asthma, 2 (12%) had previously used asthma medication (1 of whom reported having doctor-diagnosed asthma) and 1 (6%) individual with doctor-diagnosed asthma also had hay fever. Additionally, 27 of the 32 baseline participants (84%) were exposed before the study began. For those, median time from exposure to reporting was 3–194 (median 25) days. Of the 27 pre-exposed participants, 4 (15%) were current smokers, 4 (15%) had asthma, 2 (7%) had previously used asthma medication (again 1 did not report doctor-diagnosed asthma), and 2 (7%) had hay fever.
In the spirometry analysis, all participants had normal lung function and FeNO levels both before and after exposure. Only ΔFEV1 was significantly lower after exposure, 4.6% vs -0.3%, P<0.001. There were no associations between FEV1, FVC or FeNO and days since exposure, and neither was different individuals with asthma.
Among the study participants there were no statistically significant difference between the frequency of symptoms at baseline and after exposure although absolute symptom rates were higher during exposure (table 1).
Table 1
Baseline | Follow-up | P-value a | ||
---|---|---|---|---|
|
|
|
||
Level (range) | Level (range) | |||
Lung function | ||||
FEV1 (% of predicted) | 107.7 (85-130) | 106.6 (92-130) | 0.47 | |
FVC (% of predicted) | 107.4 (92-131) | 107.4 (94-129) | 0.64 | |
ΔFEV1 (%points) | 4.6 (-1-14) | -0.3 (-5-2) | <0.001 | |
FeNO (ppb) b | 21.2 (11-40) | 19.3 (10-24) | 0.48 | |
|
||||
Unexposed (reported at baseline) | During exposure (reported at baseline) | Unexposed (reported at follow-up) | P-value c | |
|
|
|
|
|
N (%) | N (% | N (% | ||
|
||||
Symptoms | ||||
Eye irritation | 1 (6) | 4 (24) | 0 (0) | 0.37 |
Nasal irritation | 1 (6) | 4 (24) | 0 (0) | 0.37 |
Cough | 3 (18) | 1 (6) | 1 (6) | 0.48 |
Shortness of breath | 0 (0) | 1 (6) | 1 (6) | - |
Wheezing/whistling in chest | 0 (0) | 0 (0) | 0 (0) | - |
Furthermore, 27 individuals reported symptoms from previous exposure. The most commonly reported symptom during exposure was nasal irritation followed by eye irritation, reported by 14 (52%) and 13 (48%) respectively, significantly increased from baseline exam of 4 (15%) and 3 (11%) (P=0.004 for both). Cough was reported by 10 participants (37%) during exposure versus 6 (22%) at the clinical examination; 4 (15%) reported experiencing shortness of breath during exposure (1 had doctor-diagnosed asthma) versus 0 at the clinical examination; 3 (11%) had experienced wheezing versus 1 at the time of the clinical examination; an 1 (4%) had used inhalation drugs for asthma at previous exposure.
In open-ended questions, 4 subjects (15%) reported taste of SO2, sulphur or match sticks in the mouth, and throat or palate irritation. There were also individual reports of “more frequent coughing”, “stress and sleeping problems”, “headaches”, “increased sensitivity to SO2 so it can be tasted at lower concentrations now (January 2015) than back in September-October 2014”, and “weakness, concentration problems and headaches”.
Discussion
In this small study of healthy professional adults who were clinically examined before and after exposure to high SO2 levels at the active eruption site, we found no changes in lung function, airway restriction or inflammation after exposure. Reversibility of airway obstruction was statistically significantly decreased from 5% to 0% after the visit to the eruption site, but FEV1 changes of <10% points are not considered clinically significant. During exposure, reports of eye and nasal symptoms at the eruption site were common, similar to findings in other studies of SO2-exposed healthy volunteers (4), but in our study, the symptoms had resolved by the time of the clinical examination. Both before and after exposure, FeNO values were within normal range. Health effects of SO2 are enhanced or occur at lower concentration in mixtures with other air pollutants and in cold conditions and physical exercise, both likely phenomenon at Holuhraun (1–3). In addition to self-reported physical symptoms, symptoms such as stress, sleeping problems, weakness, and concentration problems were reported, and although similar symptoms are known in SO2 exposure (8), they could also be stress-related and associated with being in a potentially hazardous environment. In this study, many participants reported eye and nose symptoms during previous exposure in the early phase of the eruption, before the study period, correlating with higher SO2 emissions in earlier phases of the eruption (1, 2).
The strengths of this study include detailed, clinical assessment of professionals before and after volcanic gas exposure. Weaknesses include the small study size, and loss to follow-up. Also, we have no individual exposure measurements of SO2 concentration at the eruption site, nor do we know the extent to which study subjects followed the recommendations for mask-wearing, an effect-modifying factor of the SO2 exposure. The median delay from last exposure to the clinical examination was three days, which may be too long to capture any respiratory health effects of SO2, which are widely agreed to be transient (3). In the retrospective symptom reports, some participants reported their symptoms as long as 194 days after exposure, which may cause significant bias, both due to faulty recall, but also because a short, temporal comparison is preferable as the study period included both cold and flu seasons. A factor which limits the generalizability of the study is the extremely healthy workers who self-selected into the study. Furthermore, we have no information about individuals who declined participation or individuals who visited the eruption site early and experienced symptoms and consequently could have either avoided going back or sought health care elsewhere. This would not be recorded in the study protocol, and it is thus possible that our results present an underestimate of the true effect by including healthy individuals who endured repeated exposure to the volcanic gasses. However, individuals who visited the eruption site must have had sizeable exposure to SO2 as several individuals in the free-text replies reported SO2 odor, which is detectable at concentrations between 860–4000 µg/m3 (6). The reported SO2 odor also indicates that although masks were used, they were either not used at all times, did not fit perfectly, or had full filters.
In this study, respiratory symptoms were reported during volcanic SO2 exposure but resolved without clinical signs one to six days later, indicating that in healthy individuals using protective masks, respiratory health effects after exposure to volcanic gases are minimal. However, our findings may be biased toward the null due to a very healthy, self-selected study population.