Cancer Risk From Low Level Radiation: A Review of Recent Evidence

Fig. 1. Excess deaths from solid tumors per 100 "expected" among Japanese A-bomb survivors (1950-1990) vs Dose.(11) Error bars are 95% confidence limits. Lines represent a proposed analysis into an LNT component (solid line) plus a contribution at low dose from biological defense mechanisms (BDM --- small dot line) to give a resultant behavior at low dose shown by large dot line, merging into the LNT line above 50 cSv.

TABLE 1: LEUKEMIA DEATHS FROM IARC STUDY(12)

Dose(sVc)	Observed	Expected	O/E
0-1	72	75.7	0.95
1-2	23	21.2	1.08
2-5	20	21.8	0.92
5-10	12	11.3	1.06
10-20	9	7.8	11.15
20-40	4	5.5	0.73
>40	6	2.6	2.3

Fig. 2. Excess deaths from leukemia per 100 "expected" among Japanese A-bomb survivors (1950-1990) vs dose.(11) Error bars are 95% confidence limits.

Fig. 3. Relative risk of mortality from lung cancer vs dose to lung, with 95% confidence limits. In lower figure with expanded vertical scale, circles are from Howe(14) and diamond is from Davis.(15) In upper figure, solid line connects data from Canadian fluoroscopy patients, and dashed line connects data from A-bomb survivors.(14)

Fig. 4. Data on dial painters, chemists, and others exposed to ingested radium. Ordinate is percent in each dose category that had tumors in the bone or head, and abscissa is the dose in cGy (rad) to the skeleton. For doses above 1000 cGy, error bars are one standard deviation. There were no tumors for doses below 1000 cGy; asterisks show the ordinate if there had been one tumor in the dose category. A higher dose data point at 20,000 cGy with ordinate 38% (+/-13 % ) is off the plot.(16)

Fig. 5. Lung cancer mortality rates (age-adjusted) vs average radon level in homes for U.S. Counties.(19) Fig 5b, d are lung cancer rates corrected for smoking prevalence. See explanations in text.