Harmful Algae 41 (2003) Recreational exposure to aerosolized brevetoxins during Florida red tide events

Transcription

1 Harmful Algae 41 (2003) 1 10 Recreational exposure to aerosolized brevetoxins during Florida red tide events Lorraine C. Backer a,, Lora E. Fleming b,1, Alan Rowan c,2, Yung-Sung Cheng d,3, Janet Benson d,3, Richard H. Pierce e,4, Julia Zaias f,5, Judy Bean g,6, Gregory D. Bossart h,7, David Johnson c,8, Raul Quimbo c,9, Daniel G. Baden i,10 a National Center for Environmental Health, Centers for Disease Control and Prevention, 1600 Clifton Road NE, MS E-23, Atlanta, GA 30333, USA b NIEHS Marine and Freshwater Biomedical Sciences Center, University of Miami School of Medicine, 1801 NW 9th Avenue, Room 212J, Miami, FL 33136, USA c Florida Department of Health, 4052 Bald Cypress Way, Tallahassee, FL , USA d Inhalation Toxicology Laboratory, Lovelace Respiratory Research Institute, P.O. Box 5890, Albuquerque, NM 87185, USA e Mote Marine Laboratory, 1600 Ken Thompson Parkway, Sarasota, FL 34236, USA f Division of Comparative Pathology, University of Miami School of Medicine, 1600 NW 10th Avenue, Room 7101A, Miami, FL 33136, USA g Children s Hospital Medical Center, 3333 Burnet Avenue, Cincinnati, OH 45229, USA h Harbor Branch Oceanographic Institution, 5600 US 1 North, Fort Pierce, FL 34946, USA i Center for Marine Science Research, University of North Carolina at Wilmington, 5600 Marvin K. Moss Lane, Wilmington, NC 28409, USA Abstract During two separate Karenia brevis red tide events, we measured the levels of brevetoxins in air and water samples, conducted personal interviews, and performed pulmonary function tests on people before and after they visited one of two Florida beaches. One hundred and twenty-nine people participated in the study, which we conducted during red tide events in Sarasota and Jacksonville, FL, USA. Exposure was categorized into three levels: low/no exposure, moderate exposure, and high exposure. Lower respiratory symptoms (e.g. wheezing) were reported by 8% of unexposed people, 11% of the moderately exposed people, and 28% of the highly exposed people. We performed nasal pharyngeal swabs on people who experienced moderate or high exposure, and we found an inflammatory response in over 33% of these participants. We did not find any clinically significant changes in pulmonary function test results; however, the study population was small. In future epidemiologic studies, we plan to further investigate the human health impact of inhaled brevetoxins Published by Elsevier Science B.V. Keywords: Aerosols; Brevetoxins; Gymnodinium breve; Karenia brevis; Red tide; Bronchoconstriction Corresponding author. Tel.: ; fax: address: lfb9@cdc.gov (L.C. Backer). 1 Tel.: ; fax: Tel.: ; fax: Tel.: ; fax: Tel.: x226; fax: /03/$ see front matter 2003 Published by Elsevier Science B.V. 2 doi: /s (03) Tel.: ; fax: Tel.: ; fax: Tel.: x556; fax: Tel.: x2385; fax: Tel.: ; fax: Tel.: ; fax:

2 L.C. Backer et al. / Harmful Algae 41 (2003) Introduction Karenia brevis (formerly known as Gymnodinium breve) is the marine dinoflagellate responsible for extensive blooms (called red tides ) that form primarily in the Gulf of Mexico. K. brevis produces a suite of as many as nine potent toxins called polyether brevetoxins, designated PbTx-1 through PbTx-9, that have been responsible for killing millions of fish, birds (Forrester et al., 1977), mammals (Bossart et al., 1998), and other marine organisms during red tide blooms. Although considerable knowledge exists about the biochemical and neurophysiologic activities of the brevetoxins, very little information is available about human health effects from environmental exposures such as aerosols. During onshore winds with breaking surf, brevetoxins become incorporated into marine aerosol by bubble-mediated transport (Pierce et al., 1990). Anecdotal reports and limited references to human symptoms in the literature (Music et al., 1973; Asai et al., 1982; Baden, 1983; Steidinger and Baden, 1984; Bossart et al., 1998) have consistently cited acute respiratory irritation (including non-productive cough and nasal congestion) and burning of eyes, nose, and throat as typical responses from exposure to aerosolized brevetoxins. In addition, experimental work (Baden et al., 1982; Asai et al., 1982, 1984; Franz and DeClaire, 1989) demonstrated that inhaled brevetoxins can cause bronchoconstriction and smooth respiratory muscle response that could induce an asthma attack in susceptible individuals (e.g. asthmatics, people with chronic obstructive pulmonary disease). In any study of the human health effects from inhaled brevetoxins, verifying individual exposures is important to document that the toxin has had a biological impact. Investigators have developed a number of techniques potentially useful as markers of human exposure and biological effects. For example, Bossart et al. (1998) developed an immunohistochemistry method to quantify brevetoxin within macrophages that was used to evaluate brevetoxin exposure in Florida manatees during the 1996 epizootic. This technique could be applied to human specimens to develop a biological marker of exposure to brevetoxins (i.e. that the brevetoxins were actually inhaled). Cotran et al. (1999) indicated that the peak neutrophilic influx in an acute inflammatory response occurs about 24 h after exposure to an inflammatory 82 agent. Symptoms reported by individuals exposed 83 to red tide aerosols suggest that inhaled brevetoxins 84 induce an inflammatory response, and a technique to 85 quantitatively evaluate this response (e.g. the influx 86 of neutrophils) in human specimens would be useful 87 in developing a biological marker of the effect of 88 brevetoxins Materials and methods 90 An interdisciplinary team of researchers from fed- 91 eral, state, private, and local agencies 11 conducted pi- 92 lot studies during two separate Florida red tide events 93 to assess whether beach-goers exposed to aerosolized 94 brevetoxins reported adverse health symptoms and 95 experienced measurable changes in lung function 96 (spirometry) tests after spending time on the beach. 97 The first study was conducted in February 1999 dur- 98 ing an offshore red tide off the Gulf coast of Florida. 99 The second study was conducted in October during an onshore red tide on the northeast coast 101 of Florida. During both studies, we administered 102 questionnaires, collected nose and throat swabs to 103 evaluate potential biological markers of exposure and 104 effect, and conducted pulmonary function tests on 105 beach visitors before and after they spent time on 106 the beach. We also collected samples to measure the 107 concentration of K. brevis cells in seawater and the 108 concentrations of brevetoxins in seawater and ambient 109 air. 110 We conducted the studies on two beaches, one in 111 Sarasota where people were unlikely to be exposed to 112 brevetoxins because the red tide remained offshore and 113 there was little wind, and one in Jacksonville where 114 the strong onshore winds carried the aerosolized tox- 115 ins onto the beach. We collected data on two consec- 116 utive days at each beach. The protocol for the studies 117 received approval from the institutional review boards 118 of the Centers for Disease Control and Prevention, Charlotte County Parks and Recreation, Florida Department of Environmental Quality, Florida Department of Health, Lovelace Respiratory Research Institute, Mote Marine Laboratory, National Center for Environmental Health (Centers for Disease Control and Prevention), National Institute of Environmental Health Sciences, National Institute for Occupational Safety and Health, University of Miami, and the University of North Carolina at Wilmington.

3 the University of Miami School of Medicine, and the Florida Department of Health. Adults (at least 18 years of age) were recruited as they approached the beaches. Everyone who agreed to participate in study activities was interviewed and given spirometry tests before and after going to the beach. The pre- and post-exposure questionnaires included questions about demographics, pulmonary health history, amount of time spent on the beach, medications, information about potential confounders, and various symptoms (including diarrhea, a screening symptom not anticipated to be associated with aerosol exposure). Spirometry tests were done using portable flow spirometers (Spirometrics, Gray, ME) by personnel trained using the course developed by the National Institute for Occupational Safety and Health (NIOSH, 1997). The spirometry values of interest were peak flow, the forced expiratory volume in 1 s (FEV 1 ) and the ratio of FEV 1 to forced vital capacity (FVC). We used the reference values from Hankinson et al. (1999) and the interpretation recommendations from the American Thoracic Society (1991) to compute predicted spirometric values. All study participants had at least three reproducible spirograms before and after visiting the beach, and the best values from these three spirograms were used for data analysis (see American Thoracic Society, 1991). The data were considered adequate if they conformed to standard guidelines for the collection and interpretation of spirometry measurements (American Thoracic Society, 1995). In Jacksonville, we also collected nose and throat swabs from study participants before and after going to the beach. Samples were obtained by gently wiping the nose or throat with a cotton-tipped swab, smearing the material onto duplicate microscope slides, and fixing with a cytologic adhesive spray (Spray-Cyte, Beckton Dickinson and Company, Sparks, MD). One slide from each pair was stained using Diff Quik (Dade Behring Inc., Newark, DE) for cytologic evaluation of epithelial and inflammatory cells. The number of epithelial cells on each slide was classified as mild (approximately 10 epithelial cells per high power (400 magnification) microscope field), moderate (approximately 30 cells per high power field), and high (50 or more cells per high power field). The inflammatory response (the number of inflammatory cells in pro- L.C. Backer et al. / Harmful Algae 41 (2003) portion to the number of epithelial cells present) was 168 classified as mild (inflammatory cells outnumber the 169 epithelial cells by 1.5 times), moderate (the inflam- 170 matory cells outnumber the epithelial cells by two to 171 three times), or severe (inflammatory cells outnumber 172 the epithelial cells by four or more times). The in- 173 flammatory response was characterized according to 174 the percent of neutrophils and chronic inflammatory 175 cells (e.g. macrophages, lymphocytes, plasma cells); 176 acute ( 95% neutrophils and 5% chronic inflamma- 177 tory cells; subacute (85 95% neutrophils and 5 15% 178 chronic inflammatory cells), and chronic-active (>15% 179 chronic inflammatory cells). 180 The second slide from each pair was used to de- 181 termine the presence, abundance, and distribution of 182 brevetoxin in cells. We used the immunohistochemical 183 method of Bossart et al. (1998). The slides were scored 184 as positive (the presence of clearly defined brown in- 185 tracytoplasmic granules containing brevetoxin), nega- 186 tive (the absence of clearly defined brown intracyto- 187 plasmic granules), or equivocal (questionable or am- 188 biguous staining within the cells; see Bossart et al., ). 190 Seawater samples were collected in 1 l glass bot- 191 tles twice each day. In the field, the samples were mi- 192 croscopically examined for the presence and concen- 193 tration of K. brevis cells. In the laboratory, brevetox- 194 ins were extracted by passing the seawater through a 195 C-18 solid-phase extraction disk under vacuum (An- 196 sys Technologies Inc., Lake Forest, CA). The C disks were then rinsed with reverse osmosis water to 198 remove any remaining salts and eluted with methanol 199 according to the method of Pierce et al. (1992). Con- 200 centrated samples were injected in methanol onto a 201 Shimadzu SPDM6-A diode array HPLC. The mobile 202 phase was 1 ml/min isocratic 85:15 methanol:water us- 203 ing a 250 mm 4.6mm5 m C-18 column. Detection 204 was from 200 to 300 nm with quantification at 215 nm. 205 To obtain a grid sample matrix for brevetoxin distri- 206 bution over the beach areas, six high volume air sam- 207 plers (TE-5000; Tisch Environmental Inc., Village of 208 Cleaves, OH) were placed 65 m apart: two near the 209 surf, two near the dunes, and two in the beach parking 210 lots. The air samplers were fitted with 8 in. 10 in. 211 glass fiber filters (Whatman EPM2000; Maidstone, 212 UK) and allowed to run for approximately 5 h on October 1999, and 3.5 h on 10 October The 214 flow rate was calibrated using standard procedures and 215

4 L.C. Backer et al. / Harmful Algae 41 (2003) 1 10 was maintained at 1132 l/min (40 ft 3 /min) to capture a large quantity of airborne particles for analysis as described by Pierce et al. (in press). Filters were collected from the samplers, placed in glass jars, covered with dichloromethane (DCM), and placed on ice for transport to the lab according to the method of Pierce et al. (1990). The glass fiber filters were extracted at MML using a Soxhlet apparatus, in dichloromethane (DCM). Samples were allowed to extract overnight. The DCM was transferred to a round bottom flask and evaporated using a rotary evaporator. Samples were transferred to vials using methanol for HPLC analysis. The HPLC analysis was conducted as described for the water samples above. A portable, self-contained weather station was used at the sampling locations to monitor the air temperature, relative humidity, and wind speed and direction (Complete Weather Station, Davis Instruments, Hayward, CA). The weather station was solar-powered and was capable of storing data to download into a notebook computer. We performed statistical analyses using SAS statistical software (SAS, Cary, NC). We computed descriptive statistics and use McNemar s test (Kleinbaum et al., 1982) to compare pre- and post-beach-visit symptoms within exposure groups. We used a paired t-test to compare results from the pre- and postbeach-visit pulmonary function tests. 3. Results The results from the environmental sampling are presented in Table 1. During the low/no exposure days in Sarasota, the brevetoxin (i.e. PbTx-2 and PbTx-3) levels in the ambient air were below the limit of detection, although low to moderate concentrations of K. Table 1 Results from the analysis of air and water samples Analysis brevis were found in the water. Air and water samples 251 collected during the 2 days in Jacksonville contained 252 measurable levels of brevetoxin and very high concen- 253 trations of K. brevis cells. The brevetoxin levels were 254 high on the first day of the study in Jacksonville and 255 moderate on the second day. Thus, the study days in 256 Sarasota are referred to as low/no exposure (since 257 exposure was negligible), Jacksonville day 2 is re- 258 ferred to as moderate exposure, and Jacksonville 259 study day 1 is referred to as high exposure. 260 The environmental conditions at the beaches dur- 261 ing the study are summarized in Table 2. When com- 262 pared with the environmental conditions in Sarasota, 263 Jacksonville was somewhat hotter and less humid. The 264 wind speed in Sarasota was low in the morning, and the 265 wind direction was offshore, thus limiting the impact 266 of marine aerosol on the beach. In the afternoon, the 267 wind changed to onshore and may have carried breve- 268 toxin onto the beaches; however, the air brevetoxin 269 levels remained below the limit of detection. In Jack- 270 sonville, onshore winds were present on both days. 271 Stronger onshore winds on the first day contributed 272 to the higher exposures to aerosolized brevetoxin that 273 day. 274 One hundred and twenty-nine people who spent 275 from 10 min to nearly 8 h on the beach (average min) participated in the study. During the low/no 277 exposure periods, people spent an average of 85 min 278 on the beach doing some form of exercise (e.g. walk- 279 ing, running) compared with an average of 39 min 280 during the moderate exposure period and 55 min 281 during the high exposure period. 282 The characteristics of the study participants are 283 shown in Table 3. Forty-nine people (38.0%) reported 284 having allergies and seven (5.4%) reported having 285 asthma confirmed by a physician. Thirteen partici- 286 pants (10.0%) reported that they currently have or in 287 the past have had other respiratory illnesses, including Relative brevetoxin a,b levels in ambient air Low/no (<10 ng/m 3 ) Moderate (<10 36 ng/m 3 ) High (20 93 ng/m 3 ) Concentration of K. brevis cells in seawater (l 1 ) Concentration of brevetoxins a,b in seawater ( g/l) a PbTx-2 and PbTx-3. b Limits of detection: 10 ng/m 3 for PbTx-2 + PbTx-3 in air; 0.5 g/l for PbTx-2 + PbTx-3 in seawater.

5 Table 2 Environmental conditions on the study days Environmental parameter L.C. Backer et al. / Harmful Algae 41 (2003) Relative brevetoxin a levels in ambient air Low/no (<10 ng/m 3 ) Moderate (<10 36 ng/m 3 ) High (20 93 ng/m 3 ) Temperature ( F) 71.4 ± ± ± 0.3 Relative humidity (%) 76.4 ± ± ± 1.4 Wind speed (mile/h) 5.2 ± ± ± 1.4 Maximum wind speed (mile/h) 7.5 ± ± ± 2.0 Wind direction North (morning), southwest b (afternoon) Southeast b Southeast b a PbTx-2 and PbTx-3. b Onshore winds. emphysema and bronchitis. Forty (31.0%) participants reported current smoking, and 29 (22.5%) were past smokers. Participants reported being employed in one of 44 different jobs, being in school (17 or 13.3%), or being retired (51 or 40.0%). Nearly everyone (96% on the moderate exposure day to 100% on the high exposure day) reported having an air conditioner in their home, and most (over 94%) reported using it regularly. The results of the symptom surveys are presented in Table 4. Compared with those with low exposure, people with medium or high exposure reported more symptoms associated with both upper respira- Table 3 Demographics and self-reported respiratory illnesses for study participants in Sarasota and Jacksonville Characteristic Relative brevetoxin a levels in ambient air Low/no (<10 ng/m 3 ; N = 36) Moderate (<10 36 ng/m 3 ; N = 53) High (20 93 ng/m 3 ; N = 40) Mean age (range; in years) 49.8 (24 77) 45 (18 71) 39 (18 80) Gender (female) 12 (18.2) 21 (39.6) 19 (47.5) Race White 36 (100) 46 (86.8) 30 (75) Black 0 4 (7.55) 6 (15) Other 0 3 (5.7) 4 (10) Hispanic origin 0 3 (6.0) 2 (5.9) Self-reported respiratory illnesses Asthma 0 5 (9.4) 2 (5.0) Allergies 14 (40) 20 (37.7) 15 (37.5) Other respiratory illness 5 (14.7) 3 (6.1) 5 (12.5) Smoking history Current smoker 11 (30.6) 19 (35.8) 10 (25) Past smoker 10 (27.8) 9 (17.0) 10 (25) Values are number (percentage) unless otherwise stated. a PbTx-2 and PbTx-3. tory irritation (e.g. eye irritation, nasal congestion, 301 throat irritation, and cough) and lower respiratory 302 irritation (e.g. chest tightness, wheezing, shortness of 303 breath). 304 We grouped the symptom data into four categories: 305 upper respiratory irritation (URI, i.e. eye irritation, 306 nasal congestion, throat irritation, cough), lower res- 307 piratory irritation (LRI, i.e. chest tightness, wheez- 308 ing, shortness of breath), other symptoms (including 309 skin irritation and headache), and screening symp- 310 tom (diarrhea) (see Fig. 1). We compared the pre- 311 and post-exposure symptoms reported by each study 312 participant. Those who did not have these symptoms

6 L.C. Backer et al. / Harmful Algae 41 (2003) 1 10 Table 4 Symptoms reported by study participants before and after their going to the beach Symptom Relative brevetoxin a levels in ambient air Low/no (<10 ng/m 3 ; N = 36) Moderate (<10 36 ng/m 3 ; N = 53) High (20 93 ng/m 3 ; N = 40) Upper respiratory Eye irritation 0 8 (15) 5 (12.5) Nasal congestion 1 (2.8) 8 (15) 16 (40) Throat irritation 1 (2.8) 10 (18.9) 8 (20) Cough 1 (2.8) 13 (24.5) 6 (15) Lower respiratory Chest tightness 2 (5.6) 4 (7.5) 8 (20) Wheezing 0 1 (1.9) 6 (15) Shortness of breath 1 (2.8) 3 (5.7) 6 (15) Other symptoms Itchy skin 1 (2.8) 2 (3.8) 2 (5) Headache 2 (5.6) 3 (5.7) 3 (7.5) Other 1 (2.8) 5 (9.4) 2 (5) Screening symptom (not anticipated to be associated with aerosol exposure) Diarrhea Values are number (percentage) of people who did not report the symptoms before going onto the beach but who did report the symptoms after being on the beach. a PbTx-2 and PbTx-3. Statistically significant using McNemar s test (P <0.05). Statistically significant using McNemar s test (P <0.01). before going to the beach reported no increase in symptoms following exposure to low/no levels of aerosolized brevetoxin. A significant (P < 0.001) number of people reported at least one upper respiratory symptom, many people reported lower respira- tory irritation (P = 0.06) following exposure to mod- 318 erate levels, and a significant (P <0.05) number of 319 people reported at least one lower respiratory symp- 320 tom following exposure to high levels of aerosolized 321 brevetoxin. 322 Fig. 1. Percent of people reporting symptoms after going to the beach but not before going to the beach. Symptom categories are upper respiratory irritation (URI, i.e. eye irritation, nasal congestion, throat irritation, cough), lower respiratory irritation (LRI, i.e. chest tightness, wheeze, shortness of breath), other (including skin irritation, headache), and screening symptom (diarrhea).

7 Table 5 Results from the pulmonary function tests Parameter L.C. Backer et al. / Harmful Algae 41 (2003) Relative brevetoxin a levels in ambient air Low/no (<10 ng/m 3 ; N = 28) Moderate (<10 36 ng/m 3 ; N = 44) High (20 93 ng/m 3 ; N = 40) FVC b (l/s) 0.01 (0.05) 0.04 (0.03) 0.01 (0.03) FEV c 1 (l) 0.08 (0.11) 0.04 (0.03) 0.00 (0.03) FEV 1 /FVC (%) 1.24 (0.47) 0.03 (0.53) 0.16 (0.57) Peak flow (l/s) 0.03 (0.15) 0.29 (0.10) 1.73 (1.75) Values are the mean (standard error) changes (post-exposure pre-exposure) in each parameter. a PbTx-2 and PbTx-3. b Forced vital capacity. c Forced expiratory volume in 1 s. Statistically significant change using a paired t-test (P 0.05). The people with physician-confirmed asthma did not report increases in symptoms following a trip to the beach when the brevetoxins exposure was low/no or high. Three of four people with asthma in the moderate exposure group reported additional symptoms after going to the beach: one person reported lower respiratory irritation, and two reported upper respiratory irritation. Twenty-eight (78%) people in the low exposure group, 44 (83%) in the moderate exposure group, and 40 (100%) in the high exposure group successfully completed the spirometry tests. The results of the tests are presented in Table 5. Before going to the beach, the pulmonary function test results were below the normal range for four (14.3%) people in the low/no exposure group, eight (17.8%) people in the moderate exposure group, and eight (20%) people in the high exposure group. After going to the beach, a total of seven (21.2%) people with low/no exposure, seven (16.3%) people with moderate exposure, and nine (22.5%) people with high exposure had pulmonary function tests below the normal range. The average peak flow volumes were slightly higher after going to the beach, but were highly variable. No clinically important decreases in pulmonary function parameters were found. In Jacksonville, we collected nose and throat swabs from 41 highly exposed and 51 moderately exposed study participants before and after they spent time on the beach. On the high exposure day, 1 person (2.5%) had a marked increase, 2 (5%) had moderate increases, and 17 (42%) had mild increases in inflammatory response after exposure. On the moderate exposure day, no one had a marked increase, 5 (10%) had moderate increases, and 15 (29%) had mild increases in inflam- matory response after exposure. A total of 49 and 39% 357 of people sampled on the high and moderate exposure 358 days, respectively, had an increase in inflammation in 359 the nose and/or throat swab sample. The predominant 360 inflammatory response was acute to subacute, and we 361 observed primarily neutrophils in the samples. 362 Twenty swab samples were immunohistochemically 363 examined for the presence of brevetoxin: 13 (65%) 364 pairs of pre- and post-exposure samples were negative, (10%) pairs were negative in the pre-exposure and 366 positive in the post-exposure samples, and 5 (25%) 367 were negative in the pre-exposure and equivocal in the 368 post-exposure samples Discussion 370 Anecdotal reports of respiratory irritation during red 371 tide events have suggested that the neurotoxins (breve- 372 toxins) produced by a red tide organism, K. brevis, 373 may have irritant or inflammatory effects when inhaled 374 (Music et al., 1973; Asai et al., 1982; Baden, 1983; 375 Steidinger and Baden, 1984; Bossart et al., 1998) in 376 addition to the established gastrointestinal and neuro- 377 toxic effects when ingested. Our studies indicate that 378 people exposed to aerosolized brevetoxins experience 379 both upper and lower respiratory symptoms and per- 380 haps an inflammatory response. 381 During the study period in Sarasota, the water and 382 ambient air samples did not contain detectable lev- 383 els of brevetoxins. The red tide did not move on- 384 shore and few people reported symptoms after spend- 385 ing time on the beach, indicating that simply going to 386 the beach and being exposed to sea spray aerosols and 387

8 L.C. Backer et al. / Harmful Algae 41 (2003) 1 10 sand did not induce respiratory symptoms. However, in Jacksonville, strong onshore winds created breaking waves, brevetoxins were detected in the air on the beach, and people reported a number of symptoms. On the high exposure day in Jacksonville when the cell counts in water samples were high, the brevetoxin levels were high in water samples, and the strong onshore winds carried a higher concentration of aerosolized brevetoxins, people reported an increase in lower respiratory symptoms after visiting the beach. On the moderate exposure day when brevetoxin levels in the air and water were much lower, fewer people reported lower respiratory symptoms. However, on the moderate exposure day, people s reports of upper respiratory symptoms significantly increased after being on the beach. It is not clear why people reported fewer upper respiratory symptoms on the high exposure day; this may be an artifact of the small sample size of our study and needs to be re-evaluated in a larger study. The spirometry tests did not identify any decreases in pulmonary function following exposure to aerosolized brevetoxins. The results from the peak flow tests suggested that the average peak flow volume increased after people to the beach. However, the changes in peak flow were highly variable and most likely reflect people s becoming trained to perform the tests. To distinguish these training effects from actual physiologic effects, future studies should include successive daily spirometry tests. The levels of aerosolized brevetoxins on Florida beaches during the red tide event in Jacksonville were greater than the levels of brevetoxins that have produced adverse respiratory effects in laboratory animals. During the moderate and high exposure periods of our study, people were exposed to up to 36 or 93 ng/m 3, respectively, of brevetoxin in the air. If an average adult breathes in about 6 l of air per minute (Guyton, 1981), then people visiting the beaches during our studies were inhaling between 13 and 33 ng of brevetoxin each hour, or an inhaled dose of between 0.19 and 0.47 ng/kg (assuming an average weight of 70 kg) each hour. In laboratory studies, Singer et al. (1998) reported a significant and rapid increased respiratory resistance in unanesthetized asthmatic sheep after an inhalation challenge with increasing doses ( fg) of PbTx-3. Also, Wells et al. (1984) reported an increase in airway resistance in guinea pigs exposed to aerosolized brevetoxin. Thus, during red tide events, people, particularly those with underly- 436 ing respiratory disease, may be exposed to enough 437 aerosolized brevetoxin to induce adverse respiratory 438 responses, including respiratory resistance. 439 The physiologic impact of exposure to an aerosol 440 depends on the characteristics of the inhaled parti- 441 cles. In a similar study of red tide aerosols conducted 442 in Texas, the particles containing brevetoxin were be- 443 tween 2.9 and 15 m in mass median aerodynamic di- 444 ameter (Cheng, personal communication, November ). Most inhaled particles of this size would be 446 deposited in the upper respiratory tract (Schlesinger, ), and subsequent respiratory irritation could re- 448 sult from the presence of the particles themselves or 449 from toxins associated with the particles. This is con- 450 sistent with Wells et al. (1984), who reported a signifi- 451 cant increase in airway resistance in guinea pigs when 452 brevetoxin was inhaled as an aerosol or applied as 453 nose drops. In our study, the reported respiratory irri- 454 tation was most likely caused by exposure to brevetox- 455 ins because beach-goers who inhaled sea spray with 456 measurable levels of brevetoxin reported many more 457 symptoms than did beach-goers who inhaled sea spray 458 with little or no brevetoxin. 459 Individual exposures varied widely during our 460 study. For example, people spent about half as much 461 time on the beach when airborne brevetoxin levels 462 were moderate or high than they did when brevetoxin 463 exposure was negligible. In addition, the red tide in 464 Jacksonville had been offshore for about a week be- 465 fore we conducted our study, thus it is unclear whether 466 some of the reported symptoms were the result of 467 acute exposure on the day of the study or were the 468 result of previous periodic exposures. Further studies 469 of both acute and chronic exposure to brevetoxins 470 red tide events that include collecting baseline data 471 on study participants when they are not exposed are 472 needed to clarify these findings. 473 A marker of exposure and biological effect would be 474 very useful in assessing individual exposure. In con- 475 trolled studies examining the efficacy of using swabs 476 to collect material for evaluation, the swabbing itself 477 induced a mild inflammatory response in 23% (n = ) of those who participated. Also, we found a mild 479 inflammatory response in only 10% (n = 10) of those 480 who were examined before and after going to the beach 481 when there was no red tide (Zaias et al., unpublished 482 data). We found that exposure to aerosolized breve- 483

9 toxins resulted in a rapid (within hours) and quantifiable inflammatory response in more than a third of the people sampled. Although our results from the swab tests were not definitive, additional studies of the immunostaining technique and the differential response to the toxin by the different inflammatory cell types may validate these endpoints as biomarkers of exposure and effect. This study used self-reported data of respiratory and other symptoms. Although we know that the people in our study were on the beach during the time between the pre- and post-exposure data collections, individual exposures were likely variable. Although several of the exposure measures and analytic methods are still being developed, this is the first epidemiologic study to evaluate possible irritant and inflammatory respiratory effects of aerosolized brevetoxins exposure in human populations that included simultaneous air and water monitoring. The exposure and effect measures developed for these pilot studies will be used in future studies to evaluate the effects of these toxins on animal models and on various special human populations. 5. Conclusions Our data suggest that people can experience upper and lower respiratory irritation and some inflammatory response after inhaling aerosolized brevetoxins during red tide events. Further epidemiologic studies will improve our understanding of the mechanism of this effect in people. Acknowledgements The authors would like to thank the following individuals who participated in the data collection activities for these studies: Julie Jacobson, Adam Karpati, Luke Naeher, and Maria Mirabelli, Centers for Disease Control and Prevention; Amy Burns, Aaron Hilliard, and Antonio Nichols, Duval County Health Department; Gilbert Hoover, Florida Department of Health; Dennis Yazzie and Thomas Holmes, Lovelace Respiratory Research Institute; Mike Henry, Justine Lyons, and Steve Payne, Mote Marine Laboratory; Kathleen Shea, Brad Wholer Torres, and Dominick Squiccia- L.C. Backer et al. / Harmful Algae 41 (2003) rini, National Institute of Environmental Health Sci- 526 ences Center at the University of Miami; Raymond 527 Petsko and David Spainhour, National Institute for Oc- 528 cupational Safety and Health; and Mark Harrington 529 and Tammy Harrington from Twin Cities Hospital. We 530 would like to acknowledge Robert Tamer, Children s 531 Hospital Medical Center, for his assistance in data 532 analyses. We would like to acknowledge the Florida 533 Harmful Algal Blooms Task Force for providing fund- 534 ing for the project. 535 References 536 American Thoracic Society, Lung function testing: selection 537 of reference values and interpretive strategies. Am. Rev. Respir. 538 Dis. 144, American Thoracic Society, Standardization of spirometry update. Am. J. Respir. Crit. Care Med. 152, Asai, S., Krzanowski, J.J., Anderson, W.H., Martin, D.F., Polson, 542 J.B., Lockey, R.F., Bukantz, S.C., Szentivanyi, A., Effects 543 of the toxin or red tide, Ptychodiscus brevis, on canine tracheal 544 smooth muscle: a possible new asthma-triggering mechanism. 545 J. Allergy Clin. Immunol. 69, Asai, J.S., Krzanoswki, J., Lockey, R.F., Anderson, W.H., Martin, 547 D.F., Polson, J.B., Bukantz, S.C., Szentivanyi, A., Site 548 of Action of Ptychodiscus brevis toxin within parasympathetic 549 axonal sodium channel h gate in airway smooth muscle. J. 550 Allergy Clin. Immunol. 73, Baden, D.G., Marine food-borne dinoflagellate toxins. Int. 552 Rev. Cytol. 82, Baden, D.G., Mende, T.J., Bikhzi, G., Leung, I., Broncho- 554 constriction caused by Florida red tide toxins. Toxicon 20, Bossart, G.D., Baden, D.G., Ewing, R.Y., Roberts, B., Wright, 557 S., Brevetoxicosis in manatees (Trichechus manatus 558 latirostris) from the 1996 epizootic: gross, histologic, and 559 immunohistochemical features. Toxicol. Pathol. 26, Cotran, R.S., Kuman, V., Collins, T., Pathologic Basis of 561 Disease, 6th ed. Saunders, Philadelphia, PA, pp Forrester, D.J., Gaskin, J.M., White, F.H., Thompson, N.P., Quick, 563 J.A., Henderson, G.E., Woodard, J.C., Robertson, W.D., An epizootic of waterfowl associated with a red tide episode 565 in Florida. J. Wildl. Dis. 13, Guyton, A.C., Textbook of Medical Physiology. Saunders, 567 Philadelphia, PA, p Hankinson, J.L., Odencrantz, J.R., Fedan, K.B., Spirometric 569 reference values from a sample of the general US population. 570 Am. J. Respir. Crit. Care Med. 159, Kleinbaum, D.G., Kupper, L.L., Morgenstern, H., Epide- 572 miologic Research. Van Nostrand Reinhold, New York, 573 pp Music, S.I., Howell, J.T., Brumback, C.L., Red tide its public 575 health implications. J. Fla. Med. Soc. 60,

10 L.C. Backer et al. / Harmful Algae 41 (2003) 1 10 National Institute for Occupational Safety and Health (NIOSH), NIOSH Spirometry Training Guide. NIOSH, Morgantown, WV. Pierce, R.H., Henry, M.S., Proffitt, L.S., Hasbrouck, P.A., Red tide toxin (brevetoxin) enrichment in marine aerosol. In: Graneli, E., Sundstrim, B., Elder, L., Anderson, D. (Eds.), Toxic Marine Phytoplankton. Elsevier, Amsterdam, pp Pierce, R.H., Henry, M.S., Proffitt, L.S., derossett, A.J., Evaluation of solid sorbents for the recovery of polyether toxins (brevetoxins) in seawater. Bull. Environ. Contam. Toxicol. 49, Pierce, R.H., Henry, M.S., Blum, P.C., Lyons, J., Cheng, Y.-S., Yazzie, D., Zhou, U. Brevetoxin concentrations in marine aerosol: human exposure levels during a Karenia brevis harmful algal bloom. Bull. Environ. Contam. Toxicol., in press. Schlesinger, R.B., Comparative deposition of inhaled aero- 592 sols in experimental animals and humans: a review. J. Toxicol. 593 Environ. Health 15, Singer, L.J., Lee, T., Rosen, K.A., Baden, D.G., Abraham, W.M., Inhaled Florida red tide toxins induce bronchoconstriction 596 (BC) and airway hyperresponsiveness (AHR) in sheep. Am. J. 597 Respir. Crit. Care Med. 157, A Steidinger, K.A., Baden, D.G., Toxic marine dinoflagellates. 599 In: Spector, D.L. (Ed.), Dinoflagellates. Academy Press, New 600 York, pp Wells, J.H., Lerner, M.R., Martin, D.F., Strecker, R.A., Lockey, 602 R.F., The effect of respiratory exposure to red tide toxin 603 on airway resistance in conscious guinea pigs. J. Allergy Clin. 604 Immunol. 73, 128 (Abstract 79). 605