STRONTIUM-90 IN BABY TEETH AS A FACTOR IN EARLY CHILDHOOD CANCER

Jay M. Gould, Ernest J. Sternglass, Janette D. Sherman,
Jerry Brown, William McDonnell, Joseph J. Mangano

International Journal of Health Services
Volume 30, Number 3, Pages 515-539, 2000
Copyright Baywood Publishing Co., Inc
Note: This is the draft of the article.

While RPHP teeth provide the only in vivo concentrations of radioactivity, comprehensive measurements of beta and alpha particle emitters in local waters taken by the New York State Department of Health can also be contrasted with trends in childhood cancer incidence. (21) One of the two areas sampled in Suffolk County by the state is located in the eastern part of the BNL where the Peconic River originates. The other is situated at Fisher's Island, in the far northeast part of the county, located 10 miles south-southeast of the Millstone nuclear complex in Connecticut.

State officials record monthly levels of total gross alpha, gross beta, and tritium activity in water (Figure 6). Gross beta activity is of particular interest, since Sr-90 is a beta emitter. At various sites around New York State, this isotope makes up between 20 and 50 percent of gross beta activity. Other radionuclides included in gross beta are the thyroid-seeking Iodine-131, and bone seekers Strontium-89, and Barium-140. Gross beta levels in the Peconic River and in the ponds near BNL in the 1980s and early 1990s are about five times greater than those in the water tested in Albany, which is not located near any major nuclear power plants. Annual averages vary due to unusually high readings for one or two months of the year. For example, 11 of 12 monthly measurements in the Peconic River in 1992 were between 2 and 7 picocuries of gross beta per liter of water; the other, taken on November 16, measured 130. No other testing site in New York State showed any similar rise in November, meaning that releases from the Brookhaven installation are the likely source. Relatively high activity levels in the water occurred in 1982, 1986, 1991, 1992, and 1993.

Gross Alpha and Beta Activity In Peconic River, 1984-93 Three Year Moving Averages

When gross beta trends in the Peconic River from 1984-93 are compared with cancer incidence age 0-4 from 1987-96 in Suffolk County, there is a strong correlation (r =.87, P<.0001), one greater than that found for Sr-90 concentration in Suffolk teeth (Figure 7). Again, three-year moving averages are used to enhance the statistical validity of both averages.

Gross Beta Activity in Peconic River, 1984-93 Vs. Suffolk County Cancer Incidence 0-4, 1987-96 Three Year Moving Averages

Gross alpha represents radionuclides such as plutonium-239 and other transuranic elements that are produced in nuclear reactors and have a high toxicity. From 1982 to 1993, high gross alpha readings in the Peconic River occurred during the same monitoring periods that produced peaks in gross beta, resulting in similar trends in average concentrations from 1984 to 1993. This finding indicates that alpha activity is not due to natural sources such as radium, radon, or thorium. Hence, gross alpha in the Peconic River is also correlated with cancer incidence age 0-4 in Suffolk (Figure 8), presuming a three-year latency period from exposure to diagnosis (r =.63, P<.001).

Gross Alpha Activity in Peconic River, 1984-93 vs. Suffolk County Cancer Incidence 0-4, 1987-96 Three Year Moving Averages

Tritium levels in the Peconic River are generally about 10 times greater than in Albany. No strong correlation between annual averages of tritium and cancer incidence age 0-4 exists. However, the same pattern of periodically elevated levels in the Peconic and nearby ponds, observed for gross beta and gross alpha, occur in tritium measurements. In 1992, all readings were below 4200 pCi/l, with two exceptions; the July 31 and September 14 measurements were 8600 and 16,200, respectively.

The state health department also measures Sr-90 concentrations in Peconic River fish each year. From 1982-93, an average of 243.2 pCi Sr-90/kg (entire body without the head) was found in 62 fish near Brookhaven, with a high of 1070 recorded on August 1, 1992. Sr-90 concentrations in Peconic River fish are 15 times greater than those found in 28 fish sampled in West Valley, 40 miles south of Buffalo, and the site of a nuclear waste dump. Not enough fish are available to attempt a meaningful year-by-year correlation with childhood cancer.

A simple reliability test of RPHP data can be made using Sr-90 readings from teeth from persons born during the same time that St. Louis teeth were being measured. Of the 20 teeth contributed by persons born 1957-70, 16 were from those who spent the gestation period in New York, New Jersey, or Connecticut. Sr-90 averages for both St. Louis and RPHP teeth show a rise to 1965-67 before dropping in 1968-70 (Table 2). St. Louis teeth register higher Sr-90 values, which is commensurate with higher Sr-90 values in the city's raw milk (about double that of New York City's) during these years. (8)

Discussion
For the first time in nearly two decades, in vivo levels of radioactivity have been measured in the U. S. by RPHP. Average concentrations of Sr-90 in baby teeth in New York's Long Island, the central part of New Jersey, and the Miami area in Florida all are far greater than expected, if initial decreases in St. Louis teeth during the years following large-scale atmospheric tests by the U.S., U.S.S.R., and Britain (1964-70) had continued. Yearly averages fluctuated, and rose slightly for children born in the 1980s and early 1990s; thus, the gap between observed and expected levels is growing.

This finding indicates that factors other than atmospheric bomb testing account for current levels of Sr-90 in American children. Because of the observed temporal and geographic patterns, the likeliest source is emissions from nuclear power reactors.

Just over one-quarter of the teeth contained expected low levels of Sr-90 concentrations, while 14% were 15 times or more than expected. This large variation suggests that differences in dietary intake of pregnant women - such as the amount of calcium in the diet - may affect fetal Sr-90 concentration. Moreover, the geographic origin of food and type of water consumed (public water, well water, or bottled water) are factors in determining Sr-90 content in deciduous teeth. Month of birth is also a factor; temporal variations in airborne reactor emissions from month to month that are inhaled will affect fetal concentrations of radioactivity, as well as accidents such as Three Mile Island and Chernobyl.

The connection between low-level radiation exposure and early childhood cancer is well documented. Small variations in background radiation levels and cancer in young children have been linked. (32) (33) Pelvic X-rays administered in utero raised the risk of dying from cancer before age 10, with an especially high risk for those X-rayed early in the pregnancy. (13) (14) Exposure to dietary radioactivity affected by bomb test fallout has been correlated with leukemia in children. (34) Elevated levels of childhood cancers have been found near U.S. nuclear power reactors (35) (36) (37), as well as those in other nations. (38) (39) (40) (41) (42) (43) (44). Decreased cancer incidence in children under five was demonstrated in the Sacramento CA area immediately after the 1989 shutdown of the Rancho Seco reactor. (45)

However, none of these studies had the benefit of radioactivity counts in the body, instead relying on degree of proximity to reactors, soil deposition or dietary levels. The correlation between Sr-90 in baby teeth and cancer before age five in Suffolk County is the first documented linkage between low-dose radioactivity levels in the human body due to fission products many years after exposure and elevated risk of disease.

Specifically related to Sr-90 is the extremely rare form of childhood cancer known as rhabdomyosarcoma. Bone, teeth, blood cells, and lymphoid tissue all originate from the mesenchymal layer during development. (10) Leukemia and sarcomas are malignancies derived from mesenchymal cells, (46) the latter accounting for five percent of pediatric cancers. (47) Rhabdomyosarcoma (RMS) has an annual incidence of 4.5 per million children under the age of 15, (48) so that the multiple cases reported in Suffolk County in recent years are most likely not the result of chance. (24) RMS has been induced by exposure to the beta radiation from Sr-90 applied to the skin of test animals. (49)

Observations about health risks posed by Sr-90 in Suffolk teeth are supported by additional correlations between cancer and gross beta/gross alpha activity in waters near Brookhaven National Laboratory and the nearby nuclear power plants. Sr-90 is just one beta emitter typically making up 20 to 50% of gross beta in New York State water. The higher correlation coefficient between gross beta levels and childhood cancer compared with that for Sr-90 indicates that multiple radionuclides, and not just Sr-90, affect fetal and infant health, and thus, adverse health effects in later life. Other bone-seeking isotopes emitting beta particles have similar biological actions and present similar health risks to humans as does strontium.

The finding that gross alpha activity has similar temporal trends as the beta activity and correlates with the rises and decline of childhood incidence is important because it further supports the hypothesis that nuclear reactors are the principal source of recent environmental radioactivity. The observed rises and declines cannot be due to naturally occurring alpha emitters such as radium and radon, but from reactor emissions.

A few gross alpha and beta readings in the Peconic River have been 10 or more times greater than the typical levels. This means that BNL reactors may be a major source of the fluctuations in observed alpha and beta activity, since the neither the county's only other monitoring station (Fisher's Island) nor any other site in New York State shows simultaneous sharp fluctuations. The water in the Peconic, along with all surface water in the county, is not used for drinking, as Long Islanders consume water drawn from the aquifers underlying the island. (50) Airborne releases are not only inhaled, but also enter wells via precipitation, affecting all consumers of both private and municipal well water in the county. Since the well water is also used locally to produce crops such as fruits, vegetables, and potatoes, food becomes an additional vector for uptake of radioactivity.

Brookhaven emissions should be viewed as only one contributor to radioactivity levels in Suffolk county baby teeth. The county is located near several other nuclear power reactors, a concentration with reported large airborne emissions unmatched anywhere else in the U.S. The western part of Suffolk lies only 40 miles southeast of Indian Point, which had extremely high airborne releases in 1985 and 1986. (23) Indian Point operators have had to shut its reactors periodically because of safety concerns; one reactor (Unit 3) was closed from 1993 to 1995. New York State Health Department monitoring data show that gross beta levels in the Hudson near Indian Point have frequently been two to three times greater than in the Peconic River. (31) Since many of these releases are airborne, they will reach Suffolk via prevailing winds and precipitation.

As shown in Figure 3, another highly troubled reactor (Oyster Creek) is situated 60 to 70 miles southwest of western Suffolk. Prevailing winds from the southwest place Suffolk directly downwind of the reactor, reported to have released 76.8 trillion picocuries of airborne long-lived fission products since 1970. Finally, in winter months windborne emissions from the equally troubled Haddam Neck and Millstone reactors in Connecticut reach eastern Suffolk county.

One important result of the present study is the confirmation that the fetus is much more sensitive than adults and even children to the harm caused by small exposures to radiation due to rapid cell growth. (51) (52) Energy released by alpha, beta, and gamma activity disrupts the cellular membranes, alters DNA, and changes hormones all crucial to normal growth. When elevated rates of disease occur in early childhood, they often reflect insults during the fetal period. In particular, early childhood cancers have been termed as delayed congenital defects, or late effects of in utero mutations. (53) The greater the insult to the developing fetus, the greater the risk of developing cancer and other disorders, particularly in the earliest stages of life. In New York State, with a majority of the population living within 50 miles of the troubled Indian Point reactors, cancer incidence age 0-1 from 1980-82 to 1991-93 increased 97.8%, compared to 35.2% for all children under five years of age. (54) Because of the small number of annual cases in Suffolk County under age 1, the State Department of Health declined to make this data available due to confidentiality concerns.

Demonstrating a statistical and clinical link between radioactivity and childhood cancer helps explain increases in a number of diseases among American infants and young children since the early 1980s because of the damage to their immune systems. From 1984 to 1992, rates of births under 2500 grams (55), congenital hypothyroidism (56), and acute ear infections age 0-5 (57) all rose, by 5.4%, 46.3%, and 75.9%, respectively. During this period, asthma - which had previously reached its highest incidence during the years of atmospheric testing - has become an epidemic affecting all Americans, particularly the youngest.

From about 1980 to the mid-1990s, asthma for children age 0-4 surged in terms of prevalence (58), physician office visits rate (59), and hospitalization rate (60), by 160.4%, 65.5%, and 60.5%, respectively. The increase in the asthma hospitalization rate (86.2%) during the first year of life was especially large (61), as was the increase in hospitalizations for infant bronchiolitis (141.9%). (62) Radiation may interact synergistically with the increased indoor use of pesticides, particularly organophosphates, which also produce both immediate and delayed pulmonary symptoms.

This study is limited in several ways. Sr-90 is a single radionuclide out of dozens found in the environment, so that absolute determination of total radiation doses received is impossible. It is not feasible to measure concentrations of short-lived isotopes in vivo on a large scale, and measurements of isotopes seeking soft tissue such as the gamma emitting Cs-137 using whole body counters is also quite difficult and costly. Another limitation of measurement of Sr-90 levels in baby teeth is that the measured levels cannot be attributed to releases from individual sources such as a particular reactor, except in cases such as Miami which has only a single nuclear plant within 100 miles.

At present, the study is also limited by the relatively small number of teeth in all areas except for Suffolk County. Teeth from children born from 1984 to 1992 account for about 85% of teeth received thus far, preventing any meaningful analysis of very recent trends. However, continued tooth collection will permit more detailed examinations of trends, of patterns in small areas, and relationships of radioactivity to infant and child health. Many submissions in 1999 represent children born in the early 1990s, as they begin to shed their baby teeth.

This report highlights the critical need for expanded measurement of radioactivity in the diet and in thousands of baby teeth, both from the ZIP code areas studied here and elsewhere. Continued emissions of fission products and their entry into the environment make this process an essential public health function, and more detailed analyses should be conducted to determine the extent of the role of presently permitted nuclear plant emissions play in increasing disease risks to the developing infant and child.

Conclusion
Sr-90 levels in deciduous teeth for children born during the 1980s and early 1990s are currently equal to those observed in the mid-1950s and are much greater than levels expected after the cessation of all atmospheric weapons tests. The observation that Sr-90 levels in deciduous teeth often rose and declined following known large radioactive releases into the environment measured independently by state and federal agencies means that they do not reflect past bomb test fallout but recent reactor releases.

Moreover, the evidence linking the changing levels of Sr-90 in baby teeth with changes in childhood cancer both during atmospheric nuclear testing and when only reactor releases occurred indicates that the developing embryo, fetus and young infant are far more strongly affected by reactor emissions than expected on the basis of all past studies of adults. Thus, any extrapolations from high doses to adults, such as those used to establish present permissible releases based on the study of Hiroshima and Nagasaki bomb survivors underestimate the risk of childhood leukemia, cancer or other illnesses from lower doses of fission products released into the environment by hundreds to thousands of times.

At the same time, the present results using in vivo measurements of a long-lived fission product like Sr-90 as a marker for radiation exposure at the time of birth strongly support the hypothesis that the recent epidemic rises in the incidence of immune- and hormone-related diseases among children such as cancer, asthma, low weight births, hypothyroidism, ear infections, and bronchiolitis documented in recent publications is most likely due to the unexpectedly severe effect of nuclear reactor releases, often synergistically increased by chemicals and air pollutants.