1946-47	First Trinity test in New Mexico in 1945.
1951-52	First Soviet test in 1949 and first Nevada test site tests in 1951.
1956-57	First large thermonuclear tests after 1954.
1962-63	Huge Soviet thermonuclear tests in 1961 and 1962.
1970-71	Start-up of first Millstone reactor in 1970.
1976-77	Large recorded accidental Millstone releases in 1975 and 1976.
1983-84	Emissions from the Three Mile Island reactor in 1979, and nearby Pilgrim reactor in 1982.
1987-88	Arrival of Chernobyl radiation in 1986.
1992-93 and 1995-96	Malfunctions at Millstone reactors and temporary closings.

All these peaks are too large to be due to chance. As in our earlier study, there is a significant degree of correlation between a three year moving average of the Sr-90 levels in about 500 Suffolk county teeth and childhood cancer 0-4, but with a somewhat longer lag of 4 years, as in Figure 3.

Here the four year lag for cancer incidence represents the fact that after a massive exposure it will take a few years for the cancers diagnosed in children age 0 to 4 to be associated with the massive insult . The shorter lag observed for cancer incidence associated with atmospheric bomb tests may be associated with the fact that bomb test fallout ascends to stratospheric levels unlike reactor emissions and comes down after about two years, representing a delay of approximately two years.

Epidemiological data demonstrate the high correlation between particular radioactive reactor releases and its subsequent adverse health effects. Thus as long ago as 1972, Sternglass pointed to the significant difference in high infant mortality rates in those counties downwind of the malfunctioning Dresden reactor releases in 1966 as compared with nearby upwind counties in Illinois (6).

In 1971, large accidental releases from the Savannah River plant in late 1970 were followed by significant increases in infant mortality in South Carolina in 1972. (7 )

In 1975 and 1976, huge accidental releases from a Connecticut Millstone reactor (comparable with recorded releases from Three Mile Island in 1979), were supported by official radioactivity readings which were highest close to Waterford but decreased systematically in all New England states with distance from Millstone, soon followed by increases in cancer rates is areas closest to Millstone.(8)

Similar examples are available for the Indian Point reactors in Westchester NY which started up in 1962, with initial releases high enough to contribute to an 138 percent increase in infant mortality over the period 1962-65 in the nearby town of Peekskill. The Peekskill infant mortality rate had previously quadrupled--from 5.3 deaths per 1000 births in 1973 to 23.6 in 1976 after a reported NRC release of 2.3 trillion picocuries of long-lived Indian Point airborne fission products in 1974-76. The largest Indian Point release ever recorded by the NRC-- -18 trillion picocuries for the years 1985-86--- was followed by a 49 percent increase in the sensitive Peekskill infant mortality rate, from 10.7 infant deaths per 1000 births in 1985 to 15.9 in 1987. (9)

The most significant of all radiation-induced increases in total mortality rate, including infant and cancer, followed the TMI 1979 accident in 10 counties closest to the stricken reactor (10 ). This could be traced in almost every state of the nation as well, very much like the case of Chernobyl radiation. (11)

The signing of the Partial Test Ban in 1963 represented a nearly universal acknowledgement of the harmful nature of nuclear fallout. In 1982, Congress passed legislation calling for the Department of Health and Human Services to develop methods to estimate I-131 exposure to the American people from the Nevada atmospheric nuclear bomb tests from 1951 to 1963, when such Soviet and US tests were terminated.

Accordingly, by 1999, the National Cancer Institute published estimates of such exposure in rads for every county in the nation. NCI asserted that

"Doses varied widely according to geographic area. In general, persons living in western states to the north and east of the test site had higher doses than those living in West Coast, Southern and Eastern states. In the 24 counties with the highest exposures, estimated average cumulative doses ranged from 9 to 16 rads."

The average for most counties was found to be about 2 rads. NCI stated that

"In 1993, researchers at the University of Utah…published results of an NCI-sponsored study of children living in parts of Utah and Arizona that had high I-131 fallout levels and a comparative number living in parts of Arizona that had relatively little fallout. The average dose for children in the study was estimated at about 10 rads....."

The researchers found a statistically significant association between I-131 exposure and all thyroid neoplasms. (12)

The National Cancer Institute maintains a file of age-adjusted cancer mortality rates for all counties since 1950. Since I-131 is always accompanied by bone-seeking strontium-90 as well as other isotopes associated with cancers with latency periods much longer than those for childhood thyroid cancer, those files were consulted to trace (Table 2) the secular change since 1950-54 in the ratios of the number of observed to expected white female breast cancer deaths for those counties with high I-131 exposures.

Two such groups were defined; those with exposures greater than 9 rads and about 100 counties with exposures greater than 7 rads. All are rural counties mainly in Idaho, Illinois, Missouri, Montana, Nebraska, North and South Dakota, and Utah, with far fewer observed deaths than expected in the initial 5 year period 1950-54, but by 1993-97, the latest 5 year period available, the Observed over Expected (O/E) ratios are close to the national average.

If the underlying data are correct these calculations suggest that particular concentrations of bomb test fallout after about 40 years may have contributed to the epidemic rise in breast cancer mortality observed since 1980 in those areas, as well as children's cancers observed soon after the radioactive insult.

Table 2 WHITE FEMALE BREAST CANCER MORTALITY, 1950-1997 Counties Most Exposed to Nevada Test Site Fallout
5 Year Period	24 counties exposed to more than 9 rads	101 counties exposed to more than 7 rads
1950-54
Observed deaths	85	523
Expected deaths	230	721
O/E	0.37	0.71
1980-84
Observed deaths	114	815
Expected deaths	186	945
O/E	0.61	0.86
1993-97
Observed deaths	171	931
Expected deaths	179	1004
O/E	0.96	0.93
Source: National Cancer Institute White Female Age-adjusted Breast Cancer Mortality

The most recent evidence of the importance of proximity to a point source of low-level radiation is shown by Mangano who found significant improvements in infant mortality within two years in counties close to seven reactors that ceased operations in recent years, as well as significant increases in infant mortality within a few years of their start-up dates. (13)