Table 2.—Age Distribution by Sex in 1962 of Members of Latter Day Saints Families Compared With Age Distribution of Utah Mormons in 1971° Matec Age, yr Femeie Age, yr O14 18-44 4554 5564 65-74 #£75+ Total O14 144 4554 55-64 65-74 75+ Total Utah Mormons, *1,000, 1971 122.1 175.6 32.2 26.2 14.5 10.0 380.6 123.4 181.0 36.1 28.5 19.6 13.3 400.9 Southwestern Utter, 1962 533 956 260 187 110 56 2,131 §a1 937 248 14§ a7 38 1,004 18 47 35 20 14 3 134 18 43 25 13 6 0 105 Fatout-effects group, 1962 "Latter Day Saints families consisting of those listed in both 1961 and 1962 directories of communities in southweetemn Utah. Some of the families could! not be located in thia 1961 survey and are not included in this table or in the survey of cancer incidence. The study area inctudes St George, Kanab, Parowan, Paragonah, Gunlock, Hurricane, tvins, Leeds, Santa Clara, Weahington, and Veyo, Utah; Fredonia. Ariz; and Bunkerville, Nev. In some cases, reapondents were uncertain of exact age of some family members. and these were piaced in the broad age categories above by position in the family. Age distribution of Utah Mormons from Lyon JL, Gardner JM, Weet DW: Cancer incidence in Mormona and non-Mormona in Utan during 1967-1975 (JMC! 1980;65: 1085). population was too small to follow annual cancer incidence rates. Cancer incidence data for all Utah Mormonswereavailable for a nine-year period (1967 through 1975), and 2 nine-year periods were selected for study of the cohort. Assuming that subStantial radiation doses from fallout might have been sustained by the major part of the study population by 1953, an increase in leukemia incidence could be evident by 1958, and so an early period (1958 through 1966) was selected. Other cancers of more radiosensitive organs have longer latency periods, and the latest nine-year period available for study was selected (1972 through 1980). This design provided internal as well as external controls. Cancer incidence in the early period may be compared with cancerincidence in the later period, as well as with the control population (all Utah Mormons). Cancer incidence in the early period for the subgroup complaining of fallout effects may be compared with that for the subgroupin the later period, with the study population as a whole, and with all Utah Mormons. Furthermore, the ratios of cancer of more radiosensitive organs to all other cancer may be compared between the study and control populations. Cancer incidence in the study population, the fallout-effects group, and the comparison group (all Utah Mormons) were age adjusted by the direct method, with the 1970 US white population as a standard. Expected case numbers were calculated for each category by the average annual age-adjusted cancer incidence rate for Utah Mormons divided by the average annual age-adjusted cancer incidence rate for the study population, times the number of cases of cancer in the study population (Table 1). Although the hypothesis to be evaluated is that there may be an excess of cancer of the more radiosensitive organs in a population exposed to radioactive fallout, a more conservative two-tailed test of significance was selected (Table 1). Cancer registry data could not be used JAMA, Jan 13, 1984—Vol 251, No. 2 in this cohort study (1951 through 1980) because the registry was not established until 1966, fifteen years after the testing of nuclear weapons began. The registry could not be used to establish a specific high-risk cohort in 1951 and follow it through time. Registry data include larger populations that have immigrated into the area in recent years with smalier, more recent exposures to fallout and insufficient time to permit any latent cancers to appear. Many early residents would have left the area and had cancer develop elsewhere. RESULTS The age and sex distribution of the study population in 1962 is compared with that for all Utah Mormons in 1971 (Table 2). Although 59.3% of women in the study group were of childbearing age in 1962 or the decade before (1,183/1,994), compared with 53.8% for all Mormon women in Utah,only 27.5% of the study population are children, compared with 31% for the state, an apparent deficit in the fallout area for 1948 through 1962 of about 20%. Average annual age-adjusted cancer incidence rates in Mormons in southwestern Utah in 1958 through 1966 and in 1972 through 1980 are compared with all Utah Mormonsin _ 1967 through 1975 (Table 1). The ratio of cancers of the more radiosensitive organs (identified in studies of Japanese survivors) to other cancer in the study population in 1958 through 1966 is compared with that ratio in the same population in 1972 through 1980 and to that for all Utah Mormons in 1967 through 1975. In addition, cancer of certain other organs thought to be sensitive to radiation are considered separately. These are cancer of bone and brain and malignant melanoma. There was an excess of 42 cases (P=.01) of all cancer in the early period (1958 through 1966) and an excess of 67 cases (P=.01)" in the later period (1972 through 1980). The excess cases reflected principally an increase in the incidence of cancer of more radiosensitive organs. Most notably, there were 19 cases of leukemia (P=.01)" (3.6 expected) in the early period, and this persisted into the later period, with 12 cases observed (P=.01) (3.4 expected). An early excess of cancer of the thyroid gland (6/1.4, P=.01)" was followed by a sharp increase later (14/1.7, P=.01)."° There was a borderline increase in lymphoma in 1972 through 1980 (10/ 5.2, P==.040). The incidence of cancer of the breast was unremarkable in 1958 through 1966 (8/8.8), but a sharp increase was noted in 1972 through 1980 (27/14.2, P=.01)."° There was a persistent excess of cancer of the upper GI tract (4/1.5 and 4/2.0) that was not significant. The incidence of cancer of the stomach showed an early increase (9/1.8, P=.01)* that persisted later (5/2.8). An excess number of cases of cancer of colon occurred later (14/8.1, P=.087). No increase in incidence of cancer of the lung was observed. There were more brain tumors, cancer of bone, and melanoma than expected in the early period (5/1.6, P=.05; 3/0.3, P=.01"; and 3/1.9, respectively) and this persisted (4/2.3; 5/0.4, P=.01"; and 9/2.6, P=.01", respectively). The subgroup of239 persons (134 males and 105 females) who reported effects from fallout had a much higher incidence of cancer (Table 1). In the early period (1958 through 1966), cancer developed in 31 (P=.01)" (only five cases expected) and cancer developed Cancer and Radioactive Fallout—-Johnson 233