Radiation andPublic Health Project
Home About RPHP Projects Publications & Reports Journal Press Room


International Journal of Health Services
March, 2006

Joseph J. Mangano, MPH, MBA

Previous reports document a short latency of cancer onset in young children exposed to low doses of radioactivity. The Standard Mortality Ratio for cancer in children dying before age ten rose in the period 6-10 years after the Three Mile Island and Chernobyl accidents in populations most exposed to fallout. SMRs near most nuclear power plants were elevated 6-10 years after startup, particularly for leukemia. Cancer incidence in children under age ten living near New York and New Jersey nuclear plants increased 4-5 years after increases in average Strontium-90 in baby teeth, and declined 4-5 years after Sr-90 averages dropped. The assumption that Sr-90 and childhood cancer are correlated is best supported for a supralinear dose-response, meaning the greatest per-dose risks are at the lowest doses. Findings document that the very young are especially susceptible to adverse effects of radiation exposure, even at relatively low doses.

The latency period between radiation exposure and the onset of cancer has been documented to be as long as several decades. However, some radiation-induced cancer occurs after a much shorter period. Perhaps the first evidence of a short latency was documented in the 1950s, with high rates of thyroid and other cancers typically within ten years of X-ray irradiation to infants and young children. (1-3) Leukemia rates among Hiroshima and Nagasaki survivors were elevated beginning five years after the 1945 bombings, reaching a peak ten years after. (4-6) Adults treated with therapeutic radiation for ankylosing spondylitis demonstrated increases in mortality from leukemia within two years, and prostate cancer, pancreatic cancer, and extracranial tumors (each after five years). (7) Lung cancer after the same treatment became elevated beginning nine years after exposure. (8) Irradiation treatment for cervical cancer resulted in elevated pancreatic cancer and leukemia 1-4 years after exposure. (9-11) Peak levels of bone cancer after injection of Radium-224 occurred eight years after treatment. (12)

Even at relatively low doses, irradiated adults are at greater risk for cancer just several years after exposure. A peak of chronic myeloid leukemia incidence was observed 6-10 years after Xrays to the back, gastrointestinal tract, and kidneys. (13) Mormon families in Utah living directly downwind of atmospheric nuclear weapons tests in Nevada were found to have significantly higher incidence of all cancers combined and certain radiosensitive tumors 7-15 years after the tests began. (14) Four to five years after the Chernobyl accident, thyroid cancer among adults in the Czech Republic and Poland increased. (15-16)

The developing fetus and infant has demonstrated a predisposition to cancer from various types of low-dose radiation exposure within a decade. Pelvic X-rays to pregnant women in the 1950s initially was linked to a near-doubling of the risk of cancer death before age ten. (17-18) Subsequent reports on larger populations confirmed this excess, both for leukemia and other childhood cancers. (19-21)

Elevated levels of radiosensitive cancers in the young shortly after exposure to atmospheric nuclear weapons test fallout have also been documented. Peaks in acute myeloid leukemia deaths in U.S. children age 5-9 occurred in 1962 and 1968, about five years after the peak testing periods of the late 1950s and early 1960s. (22) From 1948-1952 to 1958-1862, the number of Utah residents under age 30 who had their cancerous thyroid gland removed surgically rose from 6 to 30, much faster than the national increase. (23) In five Nordic countries, leukemia incidence in children under age five peaked during the highest periods of fallout from bomb tests. (24)

More recently, an elevation in leukemia diagnosed in the first year of life was seen in children born in 1986 and 1987, just after the accident at Chernobyl, representing a latency period of less than two years between in-utero exposure and diagnosis. These elevations were documented in multiple nations, including Belarus (25), Greece (26), Scotland (27), the U.S. (28), Wales (29), and West Germany (30), plus a grouping of European countries. (31) A latency beginning just four years between the accident at Chernobyl and elevated thyroid cancer rates in children has been reported in Belarus and the Ukraine. (32-34) Rising thyroid cancer incidence in children has also been reported within ten years of the accident in the moderately exposed areas of Belgium (35), East Hungary (36), and northern England (37). While some reports have found no excess in non-thyroid cancers in children irradiated by Chernobyl fallout, elevated rates within ten years of exposure have been documented in the Ukraine (38) and Turkey. (39)

Other reports have found unexpectedly high rates of childhood cancer, often leukemia and typically diagnosed before age ten, near nuclear installations. Early childhood cancer near nuclear plants likely represents effects of exposures in utero and in infancy. In the United Kingdom alone, at least eleven such reports representing different nuclear plants exist. (40-50) Similar results were observed in Canada (51), France (52), Germany (53), and the former Soviet Union. (54) Reports on this topic from the U.S. have been limited to several examining populations near a single facility at least two decades ago. (55-59) Data from a 1990 National Cancer Institute report show that cancer incidence age 0-9 near each of four U.S. reactors exceed the state rate. (60) A recent analysis shows that cancer incidence age 0-9 within 30 miles of each of 14 U.S. plants exceeds the national average for 1988-1997, based on 3669 cases. (61)

The many reports documenting a 5-10 year lag between radiation exposure and childhood cancer onset, plus elevated childhood cancer near nuclear power plants, illustrate the heightened sensitivity of the fetus and infant to toxins. This report will further examine this susceptibility by analyzing temporal trends in childhood cancer in populations exposed to low-dose nuclear power plant emissions 5-10 years after initial exposure.

The first part of this report will analyze changes in childhood cancer mortality in four U.S. populations exposed to radioactivity from nuclear reactor emissions. Mortality is used since it is easily available for each U.S. county for each year from 1979-2002. Deaths to children before age ten are used due to the heightened sensitivity to the fetus and infant, and the expected latency of 5-10 years.

Because of the expected brief latency between exposure and disease onset, cancer deaths age 0-9 in the periods 1-5 and 6-10 years after startup (used in the 1990 National Cancer Institute study of 52 U.S. nuclear power plants) can serve as controls and cases, respectively. Temporal changes in the Standard Mortality Ratio (SMR), representing the ratio of observed to expected (local vs. national) rates, will be examined. Significance of differences in observed and expected changes will be tested using a standard z-score test.

The four exposed groups in the study are:

1. Three Mile Island
On March 28, 1979, reactor unit 2 at the Three Mile Island nuclear installation in Pennsylvania experienced a partial core meltdown from loss of cooling water. The damaged reactor emitted elevated (but still relatively low) levels of radioactivity; the total of 14.2 curies of airborne iodine-131 and effluents (all radioisotopes with a half life of over eight days) was about 400 times greater than the average annual emissions from the plant to that time. (62) The majority of fallout from the accident traveled with prevailing winds, in a north-northeasterly direction, being detected in elevated levels in the environment in distant locations such as Albany NY (63) and Portland ME. (64)

Cancer mortality for children age 0-9 residing in the 34 contiguous counties north and northeast of Three Mile Island will be studied. (see Appendix 1) Of these, 28 are in Pennsylvania and 6 in New Jersey, and all lie within 130 miles of the plant. SMRs in the period 1979-1983 (1-5 years after the accident) and 1984-1988 (6-10 years after) will be compared. ICD-9 diagnosis codes 140.0 – 239.9 are used to identify all cancers combined, in all four study groups in this report. SMR changes for leukemia (ICD-9 204.0 – 208.9) and all other cancers combined will also be reviewed.

2. Chernobyl
On April 26, 1986, reactor unit 4 experienced at the Chernobyl plant in the Ukraine experienced a total core meltdown. Fallout from the disaster was propelled well into the stratosphere and across the globe. In the U.S., elevated but relatively low levels of environmental radioactivity were observed beginning May 5, as precipitation returned fallout to earth. Short lived radioisotopes remained elevated during the remainder of May and June; and long lived isotopes did not return to pre-accident levels for another three years. (65)

U.S. government measurements during May and June identified areas of the country that received the greatest levels of Chernobyl fallout. The upper Midwest and Pacific northwest, along with New York City, Washington DC, and Maine, had the highest concentrations of iodine-131 (half life of 8.05 days) in pasteurized milk from May 6 – June 30. (Table 1)


Table 1
U.S. Sites with Highest Average Concentrations
Of Iodine-131 in Pasteurized Milk
After the Chernobyl Accide
Site No. Samples Average I-131*
Boise ID
Spokane WA
Helena MT
Rapid City SD
Salt Lake City UT
Seattle WA
Wichita KS
Sacramento CA
Portland OR
Minneapolis MN
San Francisco CA
Des Moines IA
Grand Rapids MI
Las Vegas NV
Omaha NE
New York NY
Oklahoma City OK
Minot ND
Portland ME
Washington DC
Detroit MI
Other U.S. sites

* Average picocuries of iodine-131/liter of pasteurized milk, May 6–June 30 1986
Source: Office of Radiation Programs. Environmental Radiation Data. Quarterly reports. Montgomery AL: U.S. Environmental Protection Agency.

The change in SMR for cancer age 0-9 from the periods 1986-1990 and 1991-1995 (1-5 and 6-10 years after the accident) for 17 states and the District and Columbia will be compared with the remaining U.S. states. In New York, only the New York City area was included, since the average I-131 concentration in Buffalo and Syracuse was 8.0, well below the New York City average (14.0). In California, only the 29 northern counties were included, as I-131 averages for Sacramento and San Francisco (19.6 and 17.2) were well above that for Los Angeles (6.7). See appendix 2 for a list of states and counties studied.

3. Counties near New Nuclear Plants – Startup Before 1982
The 1990 study by the National Cancer Institute examined cancer mortality before and after startup of 52 nuclear power plants. The 1990 report calculated SMRs for five-year intervals (1-5 years before and after startup, 6-10 years before and after startup, etc.) for various age groups. This report will examine the change in SMR from 1-5 years after startup to 6-10 years after startup for children age 0-9 living near plants. Included will be data from each of the 20 areas near nuclear plants (defined by the 1990 study) with the largest populations, which account for 89% of annual cancer deaths age 0-9 near the 52 plants. Over 18.3 million persons lived in these counties in 2000 (Table 2).

Table 2
U.S. Nuclear Plants Started Before 1982
And Proximate Counties with Largest 2000 Populations
As Defined by the National Cancer Institute
Plant Startup Counties Population
Yankee Rowe
Indian Point
San Onofre
Oyster Creek
Quad Cities
Turkey Point
Duane Arnold
Rancho Seco
Three Mile Island
Fort St. Vrain
Total 20 Areas
Beaver PA, Hancock WV
Grundy, Will IL
Berkshire, Franklin MA
Rockland, Westchester NY
Orange, San Diego CA
Monroe MI
Ocean NJ
New London CT
Plymouth MA
Rock Island, Whiteside IL
Dade FL
Kenosha WI, Lake IL
Benton, Linn IA
Amador, Sacramento, San Joaquin CA
Dauphin, Lancaster, York PA
Berrien MI
Boulder, Larimer, Weld CO
New Castle DE, Salem NJ
Hamilton TN
Gaston, Lincoln, Mecklenberg NC
Source:Bureau of the Census, 2000 Census of the United States, www.census.gov, State/County quick facts

4. Counties Near New Nuclear Plants – Startup Since 1982
Beginning in 1982, a total of 23 U.S. nuclear plants began operations at installations with no existing nuclear reactors. These were not included in the 1990 study because of the late startup date. For purposes of this report, proximate areas were defined as those counties situated completely or mostly within 30 miles of the plant.

Of the areas proximate to these 23 plants, the most populated 14 (with 88% of the childhood cancer deaths a decade after startup) were selected for study. One of these, near the Catwaba plant in South Carolina, was excluded from the analysis, since it lies close to the McGuire plant, which began operations four years before Catawba startup, and is included in the previous analysis. Over 17.5 million Americans lived in counties proximate to these plants in 2000 (Table 3). The SMR for childhood cancer age 0-9 for the periods 1-5 years and 6-10 years after startup will be compared near each plant. If a plant began operations in 1982, the periods 1983-1987 and 1988-1992 will be used.

Table 3
U.S. Nuclear Plants and Proximate Counties
With Largest Populations
Nuclear Plants Started Since 1982
At Sites With No Existing Reactors
Plant Startup Counties Population
Summer 1982 Chester, Fairfield, Lexington, Newberry, Richland, Union SC 660,202
Susquehanna 1982 Carbon, Columbia, Luzerne, Montour, Schuylkill, Sullivan, Wyoming PA 645,411
Diablo Canyon 1984 San Luis Obispo, Santa Barbara CA 586,028
Limerick 1984 Berks, Bucks, Chester, Montgomery, Lehigh PA 2,466,961
Byron 1985 Boone, DeKalab, Ogle, Stephenson,Winnebago (IL), Rock, Walworth WI 755,250
Fermi 2 1985 Lenawee, Monroe, Washtenaw, Wayne MI 2,628,892
Palo Verde 1985 Maricopa AZ 3,072,149
River Bend
1985 E./W. Baton Rouge, E./W. Feliciana, Pointe Coupee LA, Wilkinson MS 503,999
Waterford 1985 Ascension, Jefferson, LaFourche, Orleans,St. Chas, St. Jas, St. J. the Baptist LA
Perry 1986 Ashtabula, Cuyahoga, Geauga, Lake OH 1,815,112
Braidwood 1987 Grundy, Kankakee, Kendall, Will IL 698,178
Harris 1 1987 Chatham, Durham, Harnett, Lee, Orange, Wake NC 1,168,781
Seabrook 1990 Rockingham, Strafford NH, Essex MA, York ME 1,299,753
Total 13 Areas     17,529,789
Source:Bureau of the Census, 2000 Census of the United States, www.census.gov, State/County quick facts.


The second part of this report examines the effects of radioactive emissions, as detected in the bodies of children. The average Strontium-90 concentration in baby teeth was measured for over 4,000 American children, most residing near nuclear power plants. The ratio of Sr-90 per gram of calcium at birth in each baby tooth was measured in a radiochemistry laboratory, using a scintillation counting technique.

Average Sr-90 concentrations were analyzed by the birth year of the tooth donor, since much of the Sr-90 uptake in deciduous teeth occurs during pregnancy and early infancy. Temporal trends in Sr-90 averages were compared with trends in cancer incidence for children under age ten in counties near nuclear plants with the largest numbers of teeth. These plants include Suffolk County NY (near the Brookhaven National Laboratories); Monmouth and Ocean Counties NJ (near the Oyster Creek plant); and Putnam, Rockland, and Westchester NY Counties (near the Indian Point plant). The correlation between these two trends will be assessed using a Poisson regression analysis testing the hypothesis that they are related. Linear and quadratic correlations will be tested, using the actual value, square root, and fourth root of Sr-90 averages.

The specific methodology to calculate Sr-90 concentrations for each tooth has been described previously (66) (67). Teeth from Suffolk County were analyzed using a Wallac WDY 1220X Quantulus low-level scintillation spectrometer, while a Perkin-Elmer 1220-003 Quantulus Ultra Low-Level Liquid Scintillation Spectrometer was used for other teeth. In addition, the method used to clean teeth before testing differed between Suffolk and other teeth; a more sophisticated preparation for non-Suffolk teeth, plus use of a different counter, allowed more Sr-90 to be detected. However, results for each area are internally consistent, allowing Sr-90 patterns and trends to be analyzed.

Sr-90 results are compared with cancer incidence diagnosed in children age 0-9 who resided in counties near nuclear plants at the time of diagnosis. Cancer registries from the states of New Jersey and New York provided counts of incident cases, while the U.S. Census Bureau counts and inter-censal estimates for resident population were used. Three-year moving averages, rather than individual years, are used for both Sr-90 and cancer rates, to increase statistical power of the comparison.


1. Three Mile Island
In the 34 downwind (north and northeast) counties closest to of Three Mile Island, the SMR for cancer in children age 0-9 rose 23.8% (0.87 to 1.08) from 1979-1983 to 1984-1988, the periods 1-5 years and 6-10 years after the accident. The crude cancer mortality rate age 0-9 in the 34 counties increased 3.6%, compared to a national decline of 16.4%. Because the number of local deaths in each five-year period (127 and 135) was relatively small, the rise in SMR is of borderline significance at p<.09. (Table 4) While the SMR for leukemia fell from 0.95 to 0.88, the ratio for all other cancers combined rose from 0.83 to 1.17, statistically significant at p<.03.

Table 4
Change in Standard Mortality Ratio, Children Age 0-9
After the 1979 Accident, 1979-1983 vs. 1984-1988
34 Counties North/Northeast and Closest to Three Mile Island

SMR (Deaths)

Type of Cancer 1979-1983 1984-1988
% Change SMR
All Cancers Combined 0.87 (127) 1.08 (135) +23.8 p<.09
Leukemia 0.95 ( 48) 0.88 ( 35) - 6.8 p<.90
All Other Cancers 0.83 ( 79) 1.17 (100) +41.0 p<.03
Source: U.S. Centers for Disease Control and Prevention, http://wonder.cdc.gov, underlying cause of death. Uses ICD-9 codes 140.0-239.9.

2. Chernobyl
From 1986-1990 to 1991-1995 (1-5 years and 6-10 years after the accident, the SMR for cancers age 0-9 in the 18 states with the most fallout from the Chernobyl accident rose from 0.97 to 1.06, a significant increase (p<.02). The crude cancer death rate age 0-9 declined 6.6% in the 18 states, compared to a reduction of 14.0% elsewhere in the U.S. The SMR rise for leukemia (0.90 to 1.01) exceeded that for all other cancers (1.00 to 1.07). Neither increase achieved statistical significance (p<.10 and p<.13). (Table 5)


Table 5
Change in Standard Mortality Ratio, Children Age 0-9
After the Chernobyl Accident (May/June 1986), 1986-1990 vs. 1991-1995
18 States With Sites With Highest Average I-131 Measurements

SMR (Deaths)

Type of Cancer 1986-1990 1991-1995 % Change SMR
All Cancers Combined 0.97 (1501) 1.06 (1466) + 8.7 p<.02
Leukemia 0.90 ( 434) 1.01 ( 422) +11.5 p<.10
All Other Cancers 1.00 (1067) 1.07 (1040) + 7.0 p<.13
Source: U.S. Centers for Disease Control and Prevention, http://wonder.cdc.gov, underlying cause of death. Uses ICD-9 codes 140.0-239.9.

3. Counties Near Nuclear Plants (startup before 1982)
The SMR for all cancers in children dying before their 10th birthday in the most populated 20 areas near nuclear power plants cited in the 1990 National Cancer Institute report increased for 17 of the 20 areas from 1-5 to 6-10 years after plant startup. Table 6 shows the total SMR rose from 0.99 to 1.18. Because of the large number of deaths in each period (587 and 590), the change was statistically significant at p<.003. Only one of the 20 changes near individual plants (Shippingport) was statistically significant. The increase in SMR for leukemia (1.00 to 1.22) exceeded that for all other cancers (0.98 to 1.15). Both increases achieved statistical significance (p<.03 and p<.05, respectively).

Table 6
Change in Cancer Mortality, Children Age 0-9
1-5 Years vs. 6-10 Years After Startup
20 Most Populated Areas Near Nuclear Power Plants
That Began Operations Before 1982
SMR (Deaths)

Type of Cancer

1-5 Yrs Aft

6-10 Yrs Aft

% Change SMR

All Cancers Combined

0.99 (587)

1.18 (590)




1.00 (276)

1.22 (264)



All Other Cancers

0.98 (311)

1.15 (326)



All Cancers by Plant


0.84 ( 20)

1.47 ( 29)




1.00 ( 22)

1.26 ( 26)



Yankee Rowe

0.65 ( 11)

1.23 ( 17)



Indian Point

0.98 ( 75)

1.22 ( 79)



San Onofre

1.07 (186)

1.11 (153)

+ 3.4


Fermi 1

0.68 (7)

1.18 ( 10)



Oyster Creek

1.12 ( 15)

0.69 ( 8)

- 38.6



1.34 ( 17)

0.60 ( 5)

- 55.6



1.02 ( 19)

1.10 ( 16)

+ 8.3


Quad Cities

1.03 ( 11)

1.48 ( 12)



Turkey Point

0.94 ( 48)

1.12 ( 49)




0.74 ( 18)

1.01 ( 20)



Duane Arnold

1.06 ( 8)

1.29 ( 8)



Rancho Seco

1.14 ( 44)

1.43 ( 55)



Three Mile Island
0.87 ( 28)
1.29 ( 36)


1.35 ( 9)

1.54 ( 8)



Fort St. Vrain

0.67 ( 10)

1.11 ( 16)




0.79 ( 12)

1.01 ( 13)




1.60 ( 13)

1.51 ( 10)

- 5.6



0.78 ( 14)

1.16 ( 20)




0.99 (587)

1.18 (590)



Source: National Cancer Institute. Cancer in Populations Near Nuclear Facilities. Washington DC: U.S. Government Printing Office, 1990.

4. Counties Near Nuclear Plants (startup since 1982)
Table 7 shows that the cancer SMR for age 0-9 in the 13 most populated areas near nuclear plants started since 1982 rose from 0.92 to 1.05, which is of borderline significance (p<.08). The ratio rose in 9 of 13 areas near nuclear plants, declined near three, and was essentially unchanged in another. The crude rate near the 13 plants fell just 1.6%, compared to larger declines nationwide. The SMR increase for leukemia (0.85 to 1.04) was roughly double that of all other cancers (0.96 to 1.06). Neither of these changes achieved statistical significance (p<.12 and p<.28, respectively).

Table 7
Change in Cancer Mortality, Children Age 0-9
1-5 Years vs. 6-10 Years After Startup
13 Most Populated Areas Near Nuclear Power Plants
That Began Operations Since 1982
SMR (Deaths)

Type of Cancer

1-5 Yrs Aft

6-10 Yrs Aft

% Change SMR

All Cancers Combined
0.92 (353)
1.05 (368)


0.85 (115)

1.04 (124)



All Other Cancers

0.96 (238)

1.06 (244)



All Cancers by Plant



0.95 ( 14)

0.76 ( 10)

- 19.8



0.41 ( 6)

0.87 ( 11)



Diablo Canyon

0.54 ( 7)

0.77 ( 10)

+ 42.9



0.76 ( 39)

0.99 ( 48)

+ 30.8



0.59 ( 10)

1.26 ( 19)



Fermi 2

0.94 ( 64)

1.20 ( 73)

+ 28.2


Palo Verde

1.01 ( 55)

0.89 ( 49)

- 11.6


River Bend

0.83 ( 11)

1.18 ( 13)

+ 41.7



0.69 ( 24)

0.94 ( 26)

+ 35.6



1.10 ( 47)

1.31 ( 48)

+ 18.6



0.57 ( 8)

0.71 ( 10)

+ 25.3


Harris 1

1.67 ( 31)

1.06 ( 19)

- 36.7



1.37 ( 37)

1.37 ( 32)

- 0.3



0.92 (353)

1.05 (368)

+ 14.7

Source: U.S. Centers for Disease Control and Prevention, available at http://wonder.cdc.gov, underlying cause of death. Uses ICD-9 codes 140.0-239.9.

5. Strontium-90 Trends and Childhood Cancer Incidence

Figures 1, 2, and 3 illustrate the comparisons of average Sr-90 in baby teeth and cancer incidence under age ten near three nuclear plants. Each represents between 10 and 14 three-year periods (moving average) covering persons born in the 1980s and the early 1990s. The analyses include a large number of teeth and cancer cases (453 and 390 for Suffolk County, 167 and 434 for Monmouth and Ocean Counties, and 239 and 371 for Putnam, Rockland, and Westchester Counties).

Each of the three indicates a similarity of trends in Sr-90 and childhood cancer with a four-year latency between the two in New York and a five-year latency in New Jersey. For example, the average Sr-90 level in Suffolk County teeth steadily rose from 0.97 to 1.68 picocuries of Sr-90 per gram of calcium, from 1981-1983 to 1984-1986. The rate of Suffolk children 0-9 diagnosed with cancer steadily rose from 1.518 to 2.075 cases per 10,000 persons from 1985-1987 to 1988-1990.

The correlation was statistically significant (p<.05) for Monmouth/Ocean and Suffolk Counties, but fell short of significance for Putnam/Rockland/Westchester). It was also significant for all three areas combined, after taking into account that there may be confounding factors. P-values were similar whether the actual value, square root, or fourth root of the Sr-90 measurements was used. However, the quadratic (fourth root) of Sr-90 best fits the assumption that the two variables are related; the Incidence Rate Ratio (IRR) is highest or each area when the fourth root is used for Sr-90 (Table 8).



Table 8
Poisson Regression Results
Average Sr-90 Concentration and
Cancer Incidence Age 0-9
Areas Near New York and New Jersey Nuclear Power Plants



95% CI


Monmouth/Ocean NJ


Actual Sr-90 value


1.005 1.201


Square root of Sr-90


1.020 2.003


Fourth root of Sr-90


1.058 6.716



Suffolk NY


Actual Sr-90 value


1.011 2.029


Square root of Sr-90


1.002 4.908


Fourth root of Sr-90


0.978 28.886



Putnam/Rockland/Westchester NY


Actual Sr-90 value


0.912 1.146


Square root of Sr-90


0.688 1.755


Fourth root of Sr-90


0.346 5.029



All Areas Combined


Actual Sr-90 value


1.005 1.064


Square root of Sr-90


1.017 1.223


Fourth root of Sr-90


1.041 1.650


The latency period between radiation exposure to the fetus and infant and onset of cancer has often been documented to be about 5-10 years. This latency includes various types of radiation exposure (from Xrays, nuclear weapons test fallout, the Chernobyl accident) and various types of cancer (leukemia, thyroid cancer, and other malignancies).

In the United States, the issue of whether nuclear reactor operations have affected childhood cancer risk is largely unexamined. This is a pertinent area of study, since atmospheric and subterranean weapons tests ceased in 1963 and 1992, respectively. The 103 U.S. nuclear power reactors now in operation represent nearly one-fourth of the world’s total, and include some of the oldest reactors.

This report analyzes cancer mortality in children exposed to radioactivity from nuclear power reactors who died before their tenth birthday. Because the lag between exposure and diagnosis is often 5-10 years, the periods 1-5 years and 6-10 years after initial exposure were compared. Excess cancer deaths among children during the first five years after exposure would not be expected, and thus represent a control group, while an elevated level of cancer deaths 6-10 years after exposure would be expected.

In areas of the U.S. exposed to the greatest levels of fallout from accidents at Three Mile Island and Chernobyl, and areas proximate to newly-started nuclear reactors, increases in the Standard Mortality Ratio 6-10 years after initial exposure in children under age ten were observed. Increases in SMR ranged from 8.7% to 23.8% (see Figure 1); each of these temporal changes achieved or approached statistical significance. For each of the four areas studied other than the area near Three Mile Island, the SMR increase for leukemia exceeded that for all other cancers. SMRs were all less than 1.00 in the period 1-5 years after initial exposure, and were greater than 1.00 in the period 6-10 years after; this indicates that populations with cancer rates below the national average changed to those above the national standard in just a few years.

In addition, the report examines the relationship between temporal trends in-body radioactivity (i.e. Sr-90 in baby teeth at birth) and childhood cancer incidence near three U.S. nuclear installations. For each area, the pattern of childhood cancer increasing 4-5 years after a rise in Sr-90 (and decreasing 4-5 years after a Sr-90 decline) was consistent. While the relationship achieved statistical significance in just two of the three areas, plus all three areas combined, the link between fetal/infant exposures from nuclear plant emissions and cancer in childhood is suggested. Much of the Sr-90 in deciduous teeth for children living near nuclear plants probably represents emissions from the plant that is ingested in air and food. (67)

An important finding in the analysis of Sr-90 and childhood cancer trends is that the quadratic (fourth root) value of Sr-90 in baby teeth provides the highest incidence rate ratio, and thus supports the theory that a quadratic of Sr-90 fits the assumption of a link better than linearity. Thus, the upward supralinear dose-response best describes the relationship between in-body Sr-90 and childhood cancer risk. This relationship indicates that the greatest per-dose risk occurs at the lowest dose levels, which is a critical aspect of understanding health risks of radioactive environmental emissions routinely released from nuclear facilities.

This report represents an in-depth examination of temporal childhood cancer patterns near U.S. nuclear plants. The findings are important in several ways. They support the pattern of a relatively short lag period between exposures early in life and disease onset. The pattern of children exposed to radiation being especially susceptible to leukemia as opposed to other types of cancer is consistent with many earlier findings.

Perhaps the most important aspect of the report is documentation of an apparent childhood cancer risk at relatively low levels of exposure. Many previous studies involved considerably larger doses, including fallout from atomic bomb tests and radiation from the Chernobyl accident. Radioactivity in the U.S. from the Three Mile Island and Chernobyl accidents were considerably less than that in Belarus/Ukraine after Chernobyl. While environmental emissions of fission products from nuclear plants vary, they are typically lower than those involved in major accidents or bomb test fallout.

Results indicate that ongoing exposure to radioactivity may present an increased health risk to infants and children not previously understood. Exposures like Hiroshima and X-rays represent a single dose, while nuclear plant emissions are continuous, and long lived isotopes from Three Mile Island/Chernobyl remained in the U.S. food chain for several years.

The study has limitations that should be addressed in subsequent research efforts. Perhaps the most important of these is the need to continue to improve dose estimates for exposures from nuclear plant emissions; and the need to further explore epidemiological comparisons of health risk. A case-control comparison of in-body doses of radioactivity in children with and without a disease such as cancer living proximate to nuclear facilities would be useful to fill this need. This report isolates only one specific type of cancer (leukemia). It examines only potential effects on young children, not adolescents or adults. It examines patterns of cancer mortality only in the first decade after initial exposure, and not thereafter. Not all increases in SMR, or the correlation between Sr-90 in baby teeth and childhood cancer incidence, are statistically significant.

Despite these shortcomings, the epidemiological findings documented here represent an important contribution to the understanding of radiation risks to the very young. With tens of millions of Americans living proximate to nuclear reactors, more detailed studies should be pursued forthwith.

The author wishes to thank Araceli Busby, PhD, for her assistance with statistical significance testing for this manuscript.

Appendix 1
Pennsylvania and New Jersey Counties
Located North/Northeast of Three Mile Island
And Within 130 Miles of Plant
Included in Analysis in Table 4
Pennsylvania Counties
Populations Age 0-9


PA/NJ Counties








New Jersey Counties



Appendix 2
States and Counties with Highest Iodine-131 Averages in Milk
Measured During May and June 1986, After Chernobyl Accident
Included in Analysis in Table 5
States Counties
District of Columbia
New York*
North Dakota
South Dakota
Contra Costa
Del Norte
El Dorado
San Francisco
Populations Age 0-9

States with High I-131






Other U.S






New York
New York


1. Duffy BJ, Fitzgerald PJ. Thyroid cancer in childhood and adolescence: report on 28 cases. Cancer 3:1018-1032, 1950.

2. Clark DE. Association of irradiation with cancer of the thyroid in children and adolescents. Journal of the American Medical Association 159(10):1007-1009, 1955.

3. Simpson CL, Hempelmann LH, Fuller LM. Neoplasia in children treated with X-rays in infancy for thymic enlargement. Radiology 64:840-845, 1955.

4. Heyssel R, Brill AB, Woodbury LA, et al. Leukemia in Hiroshima atomic bomb survivors. Blood 15:313-331, 1960.

5. Bizzozero OJ, Johnson KG, Ciocco A. Radiation-related leukemia in Hiroshima and Nagasaki, 1946-64. I. Distribution, incidence, and appearance time. New England Journal of Medicine 274(20):1095-1101, 1966.

6. Preston DLH, Kato K, Kopecky KJ, et al. Life Span Study Report 10. Part 1. Cancer mortality among A-bomb survivors in Hiroshima and Nagasaki, 1950-82. Technical Report RERF TR 1-86. Hiroshima: Radiation Effects Research Foundation.

7. Darby SC, Doll R, Gill SK, et al. Long-term mortality after a single treatment course with X-rays in patients tested for ankylosing spondylitis. British Journal of Cancer 55:179-190, 1987.

8. Smith PG, Doll R. Mortality among patients with ankylosing spondylitis after a single treatment course with X-rays. British Medical Journal 284:449-460, 1982.

9. Boice JD, Day NE, Andersen LA, et al. Second cancers following radiation treatment for cervical cancer. An international collaboration among cancer registries. Journal of the National Cancer Institute 74:955-975, 1985.

10. Boice JD, Blettner RA, Kleinerman M, et al. Radiation dose and leukemia risk in patients treated for cancer of the cervix. Journal of the National Cancer Institute 79:1295-1311, 1987.

11. Boice JD, Enghohm RA, Kleinerman M, et al. Radiation dose and second cancer risk in patients treated for cancer of the cervix. Radiation Research 116:3-55, 1988.

12. National Research Council, Committee on the Biological Effects of Ionizing Radiations. Health Risks of Radon and Other Internally Deposited Alpha-Emitters (BEIR IV). Washington DC: National Academy Press, 1988, 602.

13. Preston-Martin S, Thomas DC, Yu C, et al. Diagnostic radiography as a risk factor for chronic myeloid and monocytic leukemia. British Journal of Cancer 59(4):639-644, 1989.

14. Johnson CJ. Cancer incidence in an area of radioactive fallout downwind from the Nevada Test Site. Journal of the American Medical Association 251(2):230-236, 1984.

15. Murbeth S, Rousarova M, Scherb H, et al. Thyroid cancer has increased in the adult populations of countries moderately affected by Chernobyl fallout. Medical Science and Monitoring 10(7):CR 300-306, 2004.

16. Roszkowska H, Gorynski P. Thyroid cancer in Poland in 1980-2000. Przegl Epidemiol. 58(2):369-376, 2004.

17. Stewart A, Webb J, Giles D, et al. Malignant disease in childhood and diagnostic irradiation in utero. Lancet 2:447, 1956.

18. Stewart A, Webb J, Hewitt D. A survey of childhood malignancies. British Medical Journal 1:1495-1508, 1958.

19. MacMahon B. Prenatal x-ray exposure and childhood cancer. Journal of the National Cancer Institute 28:1173-1192, 1962.

20. Bithell JF, Stewart AM. Pre-natal irradiation and childhood malignancy: a review of British data from the Oxford survey. British Journal of Cancer 31:271-287, 1975.

21. Monson RR, MacMahon B. Prenatal x-ray exposure and cancer in children. In Radiation Carcinogenesis: Epidemiology and Biological Significance. Boice JD, and Fraumeni JF, eds. New York: Raven Press, 1984.

22. Archer VE. Association of nuclear fallout with leukemia in the U.S. Archives of Environmental Health 42:263-271, 1987.

23. Weiss ES, Olsen RE, Thompson C, et al. Surgically treated thyroid disease among young people in Utah, 1948-1962. American Journal of Public Health 57(10):1807-1814, 1967.

24. Darby SC, Olsen JH, Doll R, et al. Trends in childhood laukaemia in the Nordic countries in relation to fallout from atmospheric nuclear weapons testing. British Medical Journal 304:1005-1009, 1992.

25. Ivanov E, Tolochko GV, et al. Infant leukemia in Belarus after the Chernobyl accident. Radiation and Environmental Biophysics 37:53-55, 1998.

26. Petridou E, Trichopolous D, Dessypris N, et al. Infant leukaemia after in utero exposure to radiation from Chernobyl. Nature 382:352-353, 1996.

27. Gibson BE, Eden OB, Barrett A, et al. Leukaemia in young children in Scotland. Lancet ii:630, 1988.

28. Mangano JJ. Childhood leukaemia in US may have risen due to fallout from Chernobyl. British Medical Journal 314:1200, 1997.

29. Busby C, Scott Cato M. Increases in leukaemia in infants in Wales and Scotland following Chernobyl: evidence for errors in statutory risk estimates. Energy Environment 11:127-139, 2000.

30. Steiner M, BurkhartW, Grosche B, et al. Trends in infant leukaemia in West Germany in relation to in utero exposure due to the Chernobyl accident. Radiation and Environmental Biophysics 37:87-93, 1998.

31. Parkin DM, Clayton D, Black RJ, et al. Childhood leukaemia in Europe after Chernobyl: 5 year follow-up. British Journal of Cancer 73:1006-1012, 1996.

32. Karakov VS, Demidchik EP, Astakhova LN. Thyroid cancer after Chernobyl. Nature 359:21, 1992.

33. Likhtarev IA, Sobolev BG, Kairo IA, et al. Thyroid cancer in the Ukraine. Nature 375:365, 1995

34. United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR). Sources and Effects of Ionizing Radiation. 2000 Report to the General Assembly, with Scientific Annexes. New York: United Nations, 2000.

35. Blackburn DJ, Michel LA, Rosiere A, et al. Occurrence of thyroid papillary carcinoma in young patients. A Chernobyl connection? Journal of Pediatric Endocrinology and Metabolism 14(5):503-506, 2001.

36. Lukacs GL, Szakall S, Kozma I, et al. Changes in the epidemiological parameters of radiation-induced illnesses in East Hungary 10 years after Chernobyl. Langenbecks Arch Chir Suppl Kongressbd 114:375-377, 1997.

37. Cotterill SJ, Pearce MS, Parker L. Thyroid cancer in children and young adults in the North of England. Is increasing incidence related to the Chernobyl accident? European Journal of Cancer 37(8):945-947, 2001.

38. Noshchenko AG, Zamostyan PV, Bondar OY, et al. Radiation-induced leukemia risk among those aged 0-20 at the time of the Chernobyl accident: a case-control study in the Ukraine. International Journal of Cancer 99(4):609-618, 2002.

39. Gunay U, Meral A, Sevinir B. Pediatric malignancies in Bursa, Turkey. Journal of Environmental Pathology, Toxicology, and Oncology 15(2-4):263-265, 1996.

40. Sharp L, McKiney PA, Black RJ. Incidence of childhood brain and other non-haematopoietic neoplasms near nuclear sites in Scotland, 1975-94. Occupational and Environmental Medicine 56(5):308-314, 1999.

41. Busby C, Cato MS. Death rates from leukaemia are higher than expected in areas around nuclear sites in Berkshire and Oxfordshire. British Medical Journal 315(7103):309, 1997.

42. Black RJ, Sharp L, Harkness EF, et al. Leukaemia and non-Hodgkin’s lymphoma: incidence in children and young adults resident in the Dounreay area of Carthness, Scotland in 1968-91. Journal of Epidemiology and Community Health 48(3):232-236, 1994.

43. Draper GJ, Stiller CA, Cartwright RA, et al. Cancer in Cumbria and in the vicinity of the Sellafield nuclear installation, 1963-90. British Medical Journal 306(6870):89-94, 1993.

44. Goldsmith JR. Nuclear installations and childhood cancer in the UK: mortality and incidence for 0-9 year-old children, 1971-1980. The Science of the Total Environment 127(1-2):13-35, 1992.

45. Kinlen LJ, Hudson CM, Stiller CA. Contacts between adults as evidence for an infective origin of childhood leukaemia: an explanation for the excess near nuclear establishments in west Berkshire? British Journal of Cancer 64(3):549-554, 1991.

46. Ewings PD, Bowie C, Phillips MJ, et al. Incidence of leukemia in young people in the vicinity of Hinkley Point nuclear power station, 1959-86. British Medical Journal 299(6694):289-293, 1989.

47. Cook-Mozaffari PJ, Darby SC, Doll R, et al. Geographical variation in mortality from leukemia and other cancers in England and Wales in relation to proximity to nuclear installations, 1969-78. British Journal of Cancer 59(3):476-485, 1989.

48. Roman E, Beral V, Carpenter L, et al. Childhood leukaemia in the West Berkshire and Basingstoke and North Hampshire District Health Authorities in relation to nuclear establishments in the vicinity. British Medical Journal (Clinical Research Edition) 294(6572):597-602, 1987.

49. Forman D, Cook-Mozaffari P, Darby S, et al. Cancer near nuclear installations. Nature 329(6139):499-505, 1987.

50. Heasman MA, Kemp IW, Urquhart JD, et al. Childhood leukemia in northern Scotland. Lancet 1(8475):266, 1986.

51. McLauthlin JR, Clarke EA, Nishri ED, et al. Childhood leukemia in the vicinity of Canadian nuclear facilities. Cancer Causes and Control 4(1):51-58, 1993.

52. Viel JF, Pobel D, Carre A. Incidence of leukaemia in young people around the La Hague nuclear waste reprocessing plant: a sensitivity analysis. Statistical Medicine 14(21-22):2459-2472, 1995.

53. Hoffmann W, Dieckmann H, Schmitz-Feuerhake I. A cluster of childhood leukemia near a nuclear reactor in northern Germany. Archives of Environmental Health 52(4):275-280, 1997.

54. Zaridze DG, Li N, Men T, et al. Childhood cancer incidence in relation to distance from the former nuclear testing site in Semipalatinsk, Kazakhstan. International Journal of Cancer 59(4):471-475, 1994.

55. Johnson CJ. Cancer incidence in an area contaminated with radionuclides near a nuclear installation. Ambio 59(4):471-475, 1981.

56. Hatch MC, Beyea J, Nieves JW, et al. Cancer near the Three Mile Island nuclear plant: radiation emissions. American Journal of Epidemiology 132(3):397-412, 1990.

57. Goldsmith JR. Childhood leukemia mortality before 1970 among populations near two United States nuclear installations. Lancet 1(8641):793, 1989.

58. Enstrom JE. Cancer mortality patterns around the San Onofre nuclear power plant. American Journal of Public Health 73(1):1218, 1983.

59. Johnson CJ. Cancer and infant mortality around a nuclear power plant. American Journal of Public Health 73(10):1218, 1983.

60. Jablon S, Hrubec Z, Boice JD, et al. Cancer in Populations Living Near Nuclear Facilities. National Cancer Institute. NIH Pub. No. 90-874. Washington DC: U.S. Government Printing Office, 1990.

61. Mangano JJ, Sherman J, Chang C, et al. Elevated childhood cancer incidence proximate to U.S. nuclear power plants. Archives of Environmental Health 58(2):74-83, 2003.

62. Tichler J, Doty K, Lucadamo K. Radioactive Materials Released from Nuclear Power Plants: annual reports. Upton NY: Brookhaven National Laboratory. NUREG/CR-2907.

63. Wahlen M, Kunz, CO, Matuszek JM, Mahoney WE, Thompson RC. Radioactive plume from the Three Mile Island accident: Xenon-133 in air at a distance of 375 kilometers. Science 207:639-640, 1980.

64. Armentrout C. Report in In Re Three Mile Island Litigation Cases Consolidated II. Civil Actioln No. 1:CV-88-1452. Harrisburg PA, United States District Court for the Middle District of Pennsylvania, January 5, 1996.

65. Office of Radiation Programs. Environmental Radiation Data. Quarterly reports. Montgomery AL: U.S. Environmental Protection Agency.

66. Gould JM, Sternglass EJ, Sherman JD, et al. Strontium-90 in deciduous teeth as a factor in early childhood cancer. International Journal of Health Services 30(3):515-539, 2000.

67. Mangano JJ, Gould JM, Sternglass EJ, Sherman JD, McDonnell W. An unexpected rise in strontium-90 in US deciduous teeth in the 1990s. The Science of the Total Environment 317:37-51, 2003.

Joseph J. Mangano MPH MBA is National Coordinator of the Radiation and Public Health Project, a research group based in New York.

< Back to Spotlight