PEDIATRICS

May 1980, Volume 65, Pages 897-900

Acute fluoride poisoning in a New Mexico elementary school

by Hoffman R, Mann J, Calderone J, Trumbull J, Burkhart M.

ABSTRACT: Thirty-four persons became ill with acute fluoride poisoning shortly after drinking water in an elementary school in Los Lunas, NM. The water supply of the school was supplemented with a concentrated sodium fluoride solution designed to raise fluoride levels in drinking water to 1 to 5 ppm. Two water samples taken after the incident was reported showed elevated fluoride levels of 375 and 93.5 ppm. Malfunction of a relay switch controlling dilution of the fluoride solution produced toxic fluoride levels in the water. The symptoms of all patients were mild and generally of short duration, with two thirds of patients reporting complete resolution within 24 hours. The bitter, salty taste of the water deterred consumption of large amounts and prevented serious intoxication.

The controversy surrounding fluoridation of municipal water supply systems has generally focused on such long-term adverse health effects as cancer, mongolism, and skeletal fluorosis. (1-4) Acute fluoride intoxication resulting from the accidental occurrence of very high fluoride concentrations in drinking water has rarely been reported.

On Nov 17, 1978, the State of New Mexico Health Services Division was notified of an acute illness occurring in students in an elementary school in Los Lunas, NM. The cause of the illness was found to be an excessive fluoride concentration in the school drinking water supply. This report describes the investigation conducted by the Health Services Division into the cause and scope of the problem.

The affected elementary school is one of three grade schools in the Los Lunas rural school district, 15 miles south of Albuquerque. The school includes three grades – first, second, and third – along with a kindergarten. The total enrollment is 246 students. There are nine teachers and ten additional adult staff (principal, school nurse, secretary, physical education teacher, aides, custodian, and cooks). The school day starts at 8:45 AM and ends at 2:15 PM for grades one to three. The kindergarten is divided into morning and afternoon sessions.

Classrooms are located in two buildings. Two of the three first grade classes are in building A along with all three third grade classes. Each of these classrooms has an individual water fountain, and there is an additional fountain in the main hallway. Building B contains only one water fountain, and there is no water fountain located in the portable building where one second grade class meets.

The school’s water supply is obtained from an independent well located on the school property. The natural fluoride concentration of this well water is 0.33 ppm. In 1976 the school installed an electric fluoridator and pump designed to increase drinking water levels of fluoride to a range of 1 to 5 ppm. A concentrated NaF solution (4,220 ppm) is added to well water collected in two storage tanks, one for each of the school’s buildings. The well pump and the fluoridator are turned on simultaneously by a relay switch when pressure in the storage tanks decreases below a threshold level.

Water samples were tested for fluoride level on a regular basis. From September through Nov 15, 1978, fluoride levels taken on eight occasions ranged from 0.81 to 4.75 ppm, with a median of 1.51 ppm. None of these fluoride levels exceeded the expected upper limit. Each time water samples were taken, the electrical switch controlling the fluoridator and well pump was found to be functioning properly.

During the morning of Friday, Nov 17, 1978, 15 children were sent home from the school by the school nurse because of the acute onset of gastrointestinal symptoms. Several of the students commented that the water tasted bitter and salty, and the students, teachers, and administrators immediately assumed that the water was the responsible agent. Students became ill within 30 minutes after the start of the school day, and the principal quickly sent a message to all of the teachers forbidding anyone to use the water fountains.

An environmentalist was contacted, and water samples were obtained from the drinking fountains in the hallway of both school buildings. Fluoride levels in both samples were extremely high – 375 ppm in the sample from building A and 93.5 ppm in the sample from building B. The fluoridator pump was turned off immediately after results of the tests were received.

METHODS

The Epidemiology Unit of the Health Services Division was notified of the problem and immediately initiated an investigation. The names and telephone numbers of all students, teachers, and staff members of the school were obtained, and a telephone survey of all students and staff was conducted. A standard message was read to the parent of each child informing them of the incident, and a questionnaire was administered either directly to the schoolchildren or to their parents. Childrens’ statements were confirmed by their parents. Approximately 75% of the 265 students, teachers, and staff were contacted and interviewed by 11:00 PM on the day of the incident. The New Mexico Poison Control Center was contacted to determine whether any cases or questions concerning fluoride poisoning had been referred to that agency. To determine the duration of symptoms a follow-up telephone survey of ill persons was conducted one week after the incident. Also, school records for the two weeks before the outbreak were checked to determine the usual number of children who left school each day because of illness.

From the information obtained in the surveys, a case was defined as illness in a person meeting the following criteria:

1. The person attended school on Nov 17.
2. The person had the onset of symptoms after 8:45 AM on Nov 17.
3. The symptoms included at least one of the following: nausea, vomiting, or abdominal pain.
4. The person did not have a fever.

In addition to the surveys, the Dental Fluoridation Unit and a Los Lunas School District electrician investigated the school water system and the fluoridator on Friday evening, Nov 17. Their efforts continued throughout the weekend.

RESULTS

Absenteeism from school during the weeks of Oct 30 to Nov 3 and Nov 6-10 was examined to determine whether the fact that 15 students left school in one day because of illness was unusual. When students who left for reasons other than illness were excluded from the totals, we found that six students left school during each of the two “normal” weeks, for an average of 1.5 students leaving school because of sickness per day. Thus, 15 ill students in a single day (Nov 17) represented a marked increase in absenteeism.

In all, 207 of 265 (78.1%) persons were interviewed, and 34 of these (16.4%) met all of the case criteria and were considered to have acute fluoride poisoning. We ascertained the water drinking habits of 148 of 207 (71.5%) persons interviewed, and the association between illness and drinking water was significant (P < .001, chi square analysis). The most common symptoms in this outbreak were abdominal pain (79% of patients), nausea (68%), and vomiting (32%). Forty-one percent of the patients had only one of these symptoms, 38% had two, and 21% reported al three. Less common symptoms included weakness (18%), diarrhea (12%), and muscle twitching and excessive salivation (both 9%). No patient reported muscle spasms, tetany, convulsions, or shock. In addition, none of the children was thought to be ill enough to prompt a physician or emergency room visit. All of the patients became ill within two hours after the first possible exposure.

Twenty-seven of 34 patients were contacted one week later. Their symptoms were found to have been generally of short duration, with two thirds of patients reporting complete resolution within 24 hours. As an additional index of the brief duration of illness, all affected students returned to school on Monday, Nov 20.

The distribution of ill persons by grade is shown in the Table (teachers are included with their class). Analysis of the distribution of ill persons showed that the attack rate for the first graders (30.9%) was significantly higher (P < .01) than for children in kindergarten, second, or third grades. Further analysis showed that 13 of 17 ill first graders attended one classroom in building A. An interview with the teacher of this class revealed that she routinely had her students drink water each morning at 8:45 AM before beginning class. This practice was designed to reduce the number of interruptions during class. She also stated that the students did not use the water fountain inside each of the classrooms in building A because they were too high to reach. They instead used the hallway fountain, which had a step, thus enabling easy access.

The investigation of the fluoridator and the water system revealed that two components of the total system were not functioning properly. First, an electrical relay switch in the circuit from the pressure gauge for the storage tanks to the well pump and the fluoridator pump was found to be faulty; its malfunction permitted the fluoridator to pump concentrated NaF solution into the storage tanks without simultaneously pumping an appropriate amount of water into the tanks. Second, a toilet in the boys’ bathroom in building A on the opposite side of the wall from the hallway drinking fountain would not shut off after flushing. Therefore, water was continuously being drawn through the pipes to building A, while water was being used only intermittently in building B. A greater amount of the concentrated NaF solution was therefore drawn into the storage tank for building A than into the one for building A. This flow difference accounted for the difference in fluoride levels in the water samples taken from the fountain in building A (375 ppm) and the fountain in building B (93.5 ppm).

DISCUSSION

This is the third reported outbreak of acute fluoride poisoning caused by excess concentrations of fluoride in drinking water. (5) In all three outbreaks the high concentration resulted from an electrical or mechanic malfunction of the system controlling the fluoride feeder pump (fluoridator). The first report of waterborne fluoride poisoning occurred in 1974. Before that time documented instances of acute fluoride poisoning by ingestion of fluoride involved NaF in a powdered form (sodium fluoride) or H2SiF6 (hydrofluosilicic acid). (6-12) No deaths have been reported from fluoridated water outbreaks, but the fluoride ion is highly soluble in water and the potential that extremely high concentrations of fluoride may result in severe toxicity and even death must be recognized.

There are three major pathophysiologic methods through which ingestion of fluoride compounds my cause acute intoxication. NaF reacts with gastric HCl (NaF + HCl = NaCl + HF) to form hydrofluoric acid (HF), which has a direct corrosive effect on the gastric mucosa, especially when gastric acidity is high. As a result of the corrosive action of fluoride, the patient usually experiences nausea, vomiting, and abdominal pain within minutes after ingestion of the toxic fluoride compound. Second, the fluoride ion precipitates calcium and lowers the serum ionized calcium concentration; this results in paresthesias, tetany, convulsions, cardiac arrhythmias, and occasionally cardiovascular collapse. Finally, fluoride interferes with many enzyme systems including glycolytic enzymes, cholinesterases, and enzymes in which Mg and Mn are present.

The symptoms of patients in this outbreak were extremely mild. All of the symptoms can be attributed to the direct corrosive action of fluoride on the upper gastrointestinal tract, and there is little, if any, evidence that any person absorbed sufficient quantities of fluoride to cause systemic effects. The bitter, salty taste of the water probably prevented the children and teachers from drinking more than minimal quantities of the contaminated water. Estimates of the ingested dose of fluoride necessary to produce acute toxicity in adults vary from 7 to 70 mg (1 mg/liter = 1 ppm), and the lethal dose ranges from 70 to 140 mg/kg. Ill persons stated they drank between a few sips to several ounces of water. The estimated ingested dose would range from 1.4 mg (93.5 mg/liter x 0.015 liter) to 90 mg (375 mg/liter x 0.24 liter). These figures are consistent with the mild symptoms and relatively short duration of illness. Fortunately, a lethal dose for the younger children would have required the ingestion of approximately 1 liter of water from the fountain in building A, an unlikely possibility because of drinking fountain habits as well as the unpleasant taste of the water.

Fluoridation is considered a cornerstone of modern dental public health. Controversies surrounding fluoridation have involved cancer risks and other long-term health effects. The rarity of acute fluoride intoxication from drinking water attests to the safety of modern water fluoridation systems. The lack of fatalities suggests that the strange and unpleasant taste of water with a high fluoride concentration may function as a deterrent of water consumption, thereby reducing the quantity of fluoride ingested. The recent introduction of pleasant-tasting fluoridated mouthwashes containing concentrations of NaF from 200 to 900 ppm increases the potential for ingestion of toxic doses and emphasizes the need of physicians to become aware of the toxic effects of fluoride as well as its preventive effects.

ACKNOWLEDGMENTS

Attila Dogruel, MPA, assisted in the epidemiologic investigation. Kenneth Cable, BS, obtained water samples from the school drinking fountains for testing of fluoride content.

REFERENCES

1. Newbrun E: The safety of water fluoridation. J Am Dent Assoc 94: 301, 1977
2. Rogot E, Sharrett AR, Feinleib M, et al: Trends in urban mortality in relation to fluoridation status. Am J Epidemiol 107: 104, 1978
3. Erickson JD: Mortality in selected cities with fluoridated and non-fluoridated water supplies. N Engl J Med 298: 1112, 1978.
4. Anonymous: Fluoridation. Consumer Rep 43: 392, 480, 1978
5. Center for Disease Control: Acute fluoride poisoning – North Carolina. Morbidity Mortality Weekly Rep 23: 199, 1974
6. Gosselin RE, Hodge HC, Smith RP, et al: Clinical Toxicology of Commercial Products, ed 4. Baltimore, Williams & Wilkins, 1976, pp 159-163
7. Waldbott GL: Acute fluoride intoxication. Acta Med Scand Suppl 400: 5, 1963
8. Abukurah AR, Moser AM, Baird CL, et al: Acute sodum fluoride poisoning, JAMA 222: 816, 1972
9. Yolken R, Konecny P, McCarthy P: Acute fluoride poisoning. Pediatrics 58:90, 1976
10. Rabinowitch IM: Acute fluoride poisoning. Can Med Assoc J 52: 345, 1945
11. Maletz L: Report of a fatal case of fluoride poisoning. N Engl J Med 213: 370, 1935
12. Sharkey TP, Simpson WM: Accidental sodium fluoride poisoning. JAMA 100: 97, 1933