[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
Purchase Options:
[Skip to Content Landing]
Views 984
Citations 0
From the Centers for Disease Control and Prevention
July 18, 2012

Exposure to Nitrogen Dioxide in an Indoor Ice Arena—New Hampshire, 2011

JAMA. 2012;308(3):230-232. doi:

MMWR. 2012;61:139-142

1 figure, 1 table omitted

In January 2011, the New Hampshire Department of Health and Human Services (NHDHHS) investigated acute respiratory symptoms in a group of ice hockey players. The symptoms, which included cough, shortness of breath, hemoptysis, and chest pain or tightness, were consistent with exposure to nitrogen dioxide gas (NO2), a byproduct of combustion. Environmental and epidemiologic investigations were begun to determine the source of the exposure and identify potentially exposed persons. This report summarizes the results of those investigations, which implicated a local indoor ice arena that had hosted two hockey practice sessions during a 24-hour period when the arena ventilation system was not functioning. A total of 43 exposed persons were interviewed, of whom 31 (72.1%) reported symptoms consistent with NO2 exposure. The highest attack rate was among the hockey players (87.9%). After repair of the ventilation system, no additional cases were identified. To prevent similar episodes, ice arena operators should ensure ventilation systems and alarms are operating properly and that levels of NO2 and carbon monoxide (CO) are monitored continuously for early detection of increased gas levels.

On January 4, 2011, NHDHHS was notified that a previously healthy male aged 19 years was hospitalized for sudden onset of cough, shortness of breath, and hemoptysis shortly after a team ice hockey practice. His physical examination was notable for crackles heard in both lung bases, and his oxygen saturation was decreased to 88%-91% on room air (normal: >95%). Bilateral infiltrates and nodules were observed on chest computed tomography. Investigation revealed that other members of his team (team A) and at least one player from another local team (team B) were experiencing similar symptoms and independently had been directed to local emergency departments. Both teams had practiced in the same ice arena on the evening of January 3.

Further investigation revealed that on the morning of January 3, the ventilation system circuit board failed while being serviced, making the ventilation system inoperable. Two arena workers then spent 60-90 minutes resurfacing the ice using propane-powered equipment, finishing at approximately 11:30 a.m. The workers later reported observing a yellow haze over the ice, but neither reported any symptoms. Team A's practice was held from 6:00 to 8:00 p.m., and team B's practice was from 8:00 to 10:00 pm. The yellow haze was noted by players, coaches, and spectators at both practices. The next morning, January 4, at 9 a.m., the circuit board was replaced, and the ventilation system began operating normally. No other exposures had occurred during the time the ventilation system was not functional.

The arena housed a standard-sized ice hockey rink and was owned by a private school. The rink ice was maintained using propane-powered ice-resurfacing machines. The arena had an air monitoring system for CO and carbon dioxide (CO2), but not for NO2.

On the evening of January 4, NHDHHS staff members began interviewing all 33 players and five coaches who were present at practices on January 3. From these initial interviews, case finding was expanded to include four practice spectators and the two arena workers who operated the resurfacing equipment, for a total of 44 exposed persons. Questionnaires that assessed symptoms, exposures, and environmental observations were administered by NHDHHS staff members in person or by telephone. All but one of the 44 exposed persons completed the questionnaire.

A case was defined as the onset of cough, hemoptysis, chest pain, chest tightening, shortness of breath, headache, dizziness, nausea, or vomiting within 48 hours of being in the ice arena from 11:00 a.m. on Monday, January 3, to 9:00 a.m. on Tuesday, January 4. Illnesses with symptoms consistent with the common cold (e.g., runny nose, fever, and head congestion) were not counted as cases. Using this definition, 31 cases were identified among the 43 persons interviewed: 29 among the 33 players (87.9%) and two among the five coaches (40%). None of the four spectators had illness consistent with the case definition, nor did the one arena worker who completed the questionnaire. Most patients (90.3%) had two or more symptoms (Table). Although 10 nonplayers (coaches, spectators, and arena personnel) were exposed, players were nearly four times as likely to become ill (87.9% versus 20.0%, risk ratio [RR] = 4.39, 95% confidence interval [CI] = 1.26-15.28). Compared with nonplayers, players also were more likely to have spent more time on the ice (defined as >1 hour versus ≤1 hour) (84.8% versus 40.0%, RR = 2.12, CI = 0.98-4.59). As time spent on the ice increased, so did the attack rate and amount of hemoptysis.

On January 5, the New Hampshire Department of Environmental Services (NHDES) and the New Hampshire State Fire Marshal's Office (NHFMO) inspected the ice arena. Measurements for CO and NO2 were taken before running the resurfacing equipment (baseline conditions) and while operating the equipment, and recorded at breathing zone (where persons on the ice would be exposed) as well as adjacent to the equipment exhaust pipe. Air sampling was performed for NO using a Gastec piston hand pump equipped with Sensidyne colorimetric gas detector tubes. A TSI Q-Trak indoor air quality monitor was used to obtain direct readings for CO. While the ice resurfacer was operating in the arena, the NO2 concentration in the breathing zone increased, reaching 0.5 parts per million, the level at which corrective action must be taken according to regulations in states that regulate indoor air quality in ice arenas. These measurements did not simulate actual conditions in the arena on January 3 because the arena ventilation system had been fully functional for approximately 24 hours at the time of sampling.

Beginning January 20, a follow-up questionnaire was administered to exposed persons to assess late-onset and persistent symptoms. Thirty-nine (90.7%) of the original 43 persons interviewed responded to the follow-up questionnaire. No new cases were identified; however, two of the original patients reported late onset of hemoptysis (at 5 days and 21 days postexposure) and were advised to seek medical evaluation. Six patients (20%) reported persistent symptoms: shortness of breath on exertion (four cases), cough (two cases), and fatigue (one case).

NHDHHS, in consultation with the Northern New England Poison Center, recommended that all exposed persons seek medical evaluation, even if asymptomatic, preferably at a designated occupational health clinic. Ultimately, 39 (90.6%) complied (30 of 31 [96.8%] patients and nine of 12 [75.0%] of persons without symptoms). After these initial medical evaluations, the need for follow-up was determined on a case-by-case basis, dependent on severity. NHDES and NHFMO recommended that the arena include an NO2 sensor in the air monitoring system, establish alarm set points for CO and NO2 in line with air action level recommendations, and test this system at least monthly. The arena also was advised to conduct maintenance and tailpipe emissions testing on all ice resurfacing equipment at the beginning of the ice arena season and at least once during the season, and consider installing catalytic converters to reduce emissions. However, the most reliable way to prevent exposure in this setting is to replace propane-powered equipment with electric equipment, which should be considered as a long-term solution.

Reported by:

Antonia Altomare, DO, Kathryn Kirkland, MD, Robert McLellan, MD, Elizabeth Talbot, MD, Dartmouth-Hitchcock Medical Center, Lebanon; Christine Adamski, Sharon Alroy-Preis, MD, Elizabeth R. Daly, MPH, Jodie Dionne-Odom, MD, Maureen Macdonald, MSN, Darlene Morse, MEd, New Hampshire Dept of Health and Human Svcs; Teresa Ferrara, MPH, Richard Rumba, MPH, New Hampshire Dept of Environmental Svcs; Leslie Cartier, Maxim Schultz, New Hampshire Fire Marshal's Office. Karen E Simone, PharmD, Northern New England Poison Center, Portland, Maine. Steffany J. Cavallo, MPH, CDC/CSTE Applied Epidemiology Fellow. Corresponding contributor: Steffany J. Cavallo, steffany.cavallo@dhhs.state.nh.us, 603-271-7397.

CDC Editorial Note:

Respiratory illness caused by NO2 in indoor hockey rinks has been documented infrequently in the literature. Hazardous levels of NO2 in ice arenas often result from malfunction of propane-fueled ice resurfacing equipment or arena ventilation systems.15

Most ice arenas are designed to minimize natural ventilation in an effort to keep warm air away from the ice surface and the ice temperature near freezing. This can create a thermal inversion in which cold air and gases (especially NO2, which is denser than air) become trapped over the ice.6 The protective glass between spectator stands and the ice rink creates an additional barrier to airflow. In this episode, exposure was made worse by prolonged use of propane-powered ice resurfacers while the ventilation system was off.

Nitrogen dioxide is a yellow to reddish brown gas that irritates the upper and lower respiratory tracts and can cause short-term central nervous system symptoms.6 Severity of symptoms is related to duration of NO2 exposure,5,6 although exertion with increased frequency and depth of respiration might have made the hockey players more susceptible than the spectators or coaches to the effects of the gas. This has been reported during other exposures.4 No specific antidote for NO2 toxicity exists, and therapy is focused on supportive care and prolonged monitoring.6 The long-term consequences of acute NO2 exposure are not well understood, but in this instance, six of 31 persons had persistent symptoms up to 4 weeks postexposure. Other studies document self-reported symptoms several weeks after exposure,4 6 months postexposure,1 and even 5 years postexposure.7 However, tests of pulmonary function (e.g., spirometry and bronchoprovocation) at 10 days, 2 months, and 6 months postexposure have provided little objective evidence of compromised lung function.1,4 The small but unpredictable potential for delayed development of life-threatening conditions such as bronchiolitis obliterans warrants follow-up of exposed persons.6

The findings in this report are subject to at least two limitations. First, a broad case definition was used to ensure complete case finding and appropriate follow-up; however, this might have led to inflation of the attack rate. Second, with the exception of the index case, symptom data were based on self-report, which also might have inflated the attack rate.

No federal regulations exist for indoor air quality in ice arenas, and only three states have enacted regulations (Minnesota, Rhode Island, and Massachusetts). Only Minnesota and Massachusetts specify limits for NO2 levels. After this incident, NHFMO sent an informational bulletin to all indoor ice arenas in the state based, in part, on recommendations from the U.S. Environmental Protection Agency and the regulations existing in other states. Without legislated regulations, however, direct education of the public about signs and symptoms of NO2 exposure and education of arena staff about the risk of NO2 toxicity is important for prevention.


Tamas Peredy, MD, Northern New England Poison Center, Portland, Maine. Claudia Alvarado, Jill Drouin, Kenneth Dufault, MaryLee Greaves, Pamela Hill, Sarah Krycki, Jose Montero, MD, Karin Salome, New Hampshire Dept of Health and Human Svcs.

What is already known on this topic?

Combustion byproducts are a known threat to indoor air quality in ice arenas; however, nitrogen dioxide gas (NO2) is monitored less frequently than carbon monoxide, and signs and symptoms of NO2 intoxication are less well known than those of carbon monoxide.

What is added by this report?

The use of propane-powered ice-resurfacing equipment for 60-90 minutes in an indoor ice arena without an operating ventilation system caused symptoms of NO2 intoxication in 31 of 43 exposed persons, including 31 of 42 persons who first entered the arena more than 6 hours after the ice resurfacing had been completed.

What are the implications for public health practice?

Because exposure to NO2 can occur more frequently than is recognized, public health agencies should consider educating ice arena operators about the importance of arena ventilation, air monitoring for combustion gases, and maintenance of propane-powered equipment, if use of electric ice resurfacing equipment is not feasible. Additionally, ice arena operators as well as ice hockey players and coaches who use indoor rinks should be familiar with the signs and symptoms of NO2 toxicity.

Kahan ES, Martin UJ, Spungen S, Ciccolella D, Criner GJ. Chronic cough and dyspnea in ice hockey players after an acute exposure to combustion products of a faulty ice resurfacer.  Lung. 2007;185(1):47-54PubMedArticle
Centers for Disease Control (CDC).  Nitrogen dioxide and carbon monoxide intoxication in an indoor ice arena--Wisconsin, 1992.  MMWR Morb Mortal Wkly Rep. 1992;41(21):383-385PubMed
Pelham TW, Holt LE, Moss MA. Exposure to carbon monoxide and nitrogen dioxide in enclosed ice arenas.  Occup Environ Med. 2002;59(4):224-233PubMedArticle
Hedberg K, Hedberg CW, Iber C,  et al.  An outbreak of nitrogen dioxide-induced respiratory illness among ice hockey players.  JAMA. 1989;262(21):3014-3017PubMedArticle
Rosenlund M, Bluhm G. Health effects resulting from nitrogen dioxide exposure in an indoor ice arena.  Arch Environ Health. 1999;54(1):52-57PubMedArticle
Greenberg MI, ed, Hamilton RJ, ed, Phillips SD, ed, McCluskey GJ, edOccupational, industrial, and environmental toxicology. 2nd ed. Philadelphia, PA: Mosby; 2003
Rosenlund M, Jungnelius S, Bluhm G, Svartengren M. A 5-year follow-up of airway symptoms after nitrogen dioxide exposure in an indoor ice arena.  Arch Environ Health. 2004;59(4):213-217PubMedArticle