Genome-wide association study using high-density single nucleotide polymorphism arrays and whole-genome sequences for clinical mastitis traits in dairy cattle.

Mastitis is a mammary disease that frequently affects dairy cattle. Despite considerable research on the development of effective prevention and treatment strategies, mastitis continues to be a significant issue in bovine veterinary medicine. To identify major genes that affect mastitis in dairy cattle, 6 chromosomal regions on Bos taurus autosome (BTA) 6, 13, 16, 19, and 20 were selected from a genome scan for 9 mastitis phenotypes using imputed high-density single nucleotide polymorphism arrays. Association analyses using sequence-level variants for the 6 targeted regions were carried out to map causal variants using whole-genome sequence data from 3 breeds. The quantitative trait loci (QTL) discovery population comprised 4,992 progeny-tested Holstein bulls, and QTL were confirmed in 4,442 Nordic Red and 1,126 Jersey cattle. The targeted regions were imputed to the sequence level. The highest association signal for clinical mastitis was observed on BTA 6 at 88.97 Mb in Holstein cattle and was confirmed in Nordic Red cattle. The peak association region on BTA 6 contained 2 genes: vitamin D-binding protein precursor (GC) and neuropeptide FF receptor 2 (NPFFR2), which, based on known biological functions, are good candidates for affecting mastitis. However, strong linkage disequilibrium in this region prevented conclusive determination of the causal gene. A different QTL on BTA 6 located at 88.32 Mb in Holstein cattle affected mastitis. In addition, QTL on BTA 13 and 19 were confirmed to segregate in Nordic Red cattle and QTL on BTA 16 and 20 were confirmed in Jersey cattle. Although several candidate genes were identified in these targeted regions, it was not possible to identify a gene or polymorphism as the causal factor for any of these regions.

Scope of radiological protection control measures.

In this report, the Commission recommends approaches to national authorities for their definition of the scope of radiological protection control measures through regulations, by using its principles of justification and optimisation. The report provides advice for deciding the radiation exposure situations that should be covered by the relevant regulations because their regulatory control can be justified, and, conversely, those that may be considered for exclusion from the regulations because their regulatory control is deemed to be unamenable and unjustified. It also provides advice on the situations resulting from regulated circumstances but which may be considered by regulators for exemption from complying with specific requirements because the application of these requirements is unwarranted and exemption is the optimum option. Thus, the report describes exclusion criteria for defining the scope of radiological protection regulations, exemption criteria for planned exposure situations, and the application of these concepts in emergency exposure situations and in existing exposure situations. The report also addresses specific exposure situations such as exposure to low-energy or low-intensity adventitious radiation, cosmic radiation, naturally occurring radioactive materials, radon, commodities, and low-level radioactive waste. The quantitative criteria in the report are intended only as generic suggestions to regulators for defining the regulatory scope, in the understanding that the definitive boundaries for establishing the situations that can be or need to be regulated will depend on national approaches.

Trends in cancer risks among Swedish agricultural workers.

The Swedish Cancer-Environment Register was used to study time-related trends in relative risks (RRs) of cancer between 1961 and 1979 in a cohort of 254,417 Swedish men who were employed in agriculture in 1960. The reference cohort consisted of 1,725,845 Swedish men who were gainfully employed in economic activities other than agriculture or forestry in 1960. Altogether 24,763 cancers were observed in the study cohort and 146,900 in the reference cohort, giving an estimated RR for the entire study period of 0.82 (95% confidence interval: 0.81-0.83). The RR for all sites combined increased from 0.80 in 1961-73 to 0.84 in 1974-79 (P less than .01). The RR also increased over time for primary liver cancer (P less than .01), prostate cancer (P less than .01), cancer of other genital organs (P less than .01), cancer of urinary organs (P less than .01), lip cancer (P less than .05), and cancer of the nose and nasal cavities (P less than .05). For most of these sites the RR remained lower than unity. For prostate cancer, however, the RR was unity at the end of the study period. A decrease in the RR over time was observed for skin carcinomas of the trunk and limbs (P less than .05) and malignant tumors of the nervous system (P less than .05). For 27 of the 48 analyzed tumor sites the RR for the entire period 1961-79 was significantly lower than unity. The lowest RRs were seen for cancer of the pleura (0.25), cancer of the larynx (0.35), lung cancer (0.36), cancer of the hypopharynx (0.36), cancer of the floor of mouth (0.40), primary liver cancer (0.44), and cancer of the kidney pelvis (0.49). RRs significantly higher than unity were found for cancer of the lip (1.92), malignant melanoma, and carcinoma of the skin in the head and neck region (1.39 and 1.15, respectively), multiple myeloma (1.20), and cancer of the stomach (1.07).

Soft tissue sarcoma risk in Swedish agricultural and forestry workers.

The risk of soft tissue sarcoma following possible exposure to phenoxy acid herbicides was studied in 354,620 Swedish men, who were employed in agriculture or forestry according to a national census in 1960. This cohort was further divided into six subcohorts, on assumed exposure to phenoxy acid herbicides. The most commonly used phenoxy acid in Sweden was (4-chloro-2-methylphenoxy)acetic acid (CAS: 94-74-6). The reference cohort encompassed 1,725,845 Swedish men employed in other industries. All persons were followed up in the cancer-environment register during the period 1961-79. A total of 331 cases of soft tissue sarcomas was observed in the study cohort and there were 1,508 cases in the reference group [relative risk (RR), 0.9; 95% confidence interval, 0.8-1.0]. No subcohort of agricultural or forestry workers showed any significantly increased RR, nor was there any significant difference in RR between the subcohorts. Despite the greatly increased use of phenoxy acid herbicides from 1947 to 1970, no time-related increase in the RR of soft tissue sarcoma was found in the total cohort or in any of the subcohorts.

Incidence of malignant thyroid tumors in humans after exposure to diagnostic doses of iodine-131. II. Estimation of thyroid gland size, thyroid radiation dose, and predicted versus observed number of malignant thyroid tumors.

The size of the thyroid glands was analyzed for 10% of the patients in a selected group that had been exposed to diagnostic doses of 131I. The mean thyroid gland weight +/- SD was 50 +/- 33 g for patients 20 or more years of age and 10 +/- 5 g for patients less than 20 years of age. With the present follow-up, diagnostic doses of 131I appeared not to be associated with an increased risk for later development of malignant thyroid tumors. Possible reasons for the difference between the observed number of such tumors and the number expected (47-124) on the basis of risk estimates of the United Nations Scientific Committee on the Effects of Atomic Radiation are discussed.

Incidence of malignant thyroid tumors in humans after exposure to diagnostic doses of iodine-131. I. Retrospective cohort study.

Between 1952 and 1965, 10,133 patients received an average of 60 mu Cl of 131I for diagnostic examinations of the thyroid gland. The examinations were mainly done on adults, and only 5% of the patients were younger than 20 years at the time of examination. Of the patients, 9 appeared in the Swedish Cancer Registry between 1958 and 1977 with a diagnosis of malignant tumor of the thyroid gland more than 5 years after the administration of 131I. Inasmuch as the expected number of malignant thyroid tumors computed from Swedish cancer incidence figures was 8.3, there was no elevation of the incidence of malignant thyroid tumors in these patients receiving diagnostic doses of 131I. The patients were examined because of a suspected thyroid dysfunction and therefore represented a selected group. The investigation is being continued to analyze further parameters that could have a bearing on these results.

Cancer incidence and cancer mortality in relation to serum cholesterol.

We studied cancer incidence and mortality from cancer and coronary heart disease in relation to serum cholesterol levels in 92,710 individuals followed-up in the nationwide Swedish Cancer Register and the Swedish Cause of Death Register for 18-20 years. According to Cox's proportional hazard model, total cancer incidence and total cancer mortality were negatively correlated to serum cholesterol level (P less than .001). The negative correlations were most pronounced during the first years of follow-up. Cancer mortality data showed a stronger negative association to cholesterol than did incidence data during the first 10 years of follow-up (P less than .05). Mortality from coronary heart disease was positively correlated to serum cholesterol (P less than .001) during the entire follow-up. In contrast to most studies that were based on mortality data, our results of the comparison of incidence and mortality data of the same cohort are in agreement with those of a cholesterol-lowering effect of a preclinical cancer. Efforts by investigators and clinicians to lower serum cholesterol to prevent cardiovascular disease are, according to the present findings, not likely to increase cancer mortality risks but would extend life, irrespective of cause of death.

Cancer incidence after radiotherapy for skin hemangioma: a retrospective cohort study in Sweden.

The cancer incidence was studied in 18,030 patients (33% males, 67% females) with skin hemangioma who were admitted to Radiumhemmet, Karolinska Hospital, Stockholm, Sweden, 1920-1959. Radium-226 sources were used in 12,821 patients, x-ray therapy was used in 2,515 patients, and no radiotherapy was given to 2,694 patients. Cancer incidence in the cohort was searched by record linkage with the Swedish Cancer Register for the period 1958-1982. The median age was 6 months for the treated patients and 8 months for the patients not receiving radiotherapy. In the group treated with radium-226 or orthovoltage x rays (greater than or equal to 100-kV peak), 224 cancers were observed [relative risk (RR) = 1.18; 95% confidence interval (CI) = 1.03-1.35]. In patients given contact x rays, 10 cancers were observed (RR = 0.71; 95% CI = 0.34-1.30). In patients not treated with ionizing radiation, 34 cancers were observed (RR = 0.93; 95% CI = 0.64-1.29). In patients treated with radium-226 or orthovoltage x rays, an RR of 1.65 was observed for breast cancer (95% CI = 1.26-2.13) and an RR of 2.73 was found for soft tissue tumors (95% CI = 1.18-5.38). Patients with brain tumors, thyroid cancers, and bone tumors had received radiotherapy close to the tumor site more often than expected. For patients with breast cancer, no such difference was found. For cancers of the breast and thyroid, the RR was higher in patients given more than one treatment.

Treatment failure and dietary habits in women with breast cancer.

BACKGROUND: Epidemiological and experimental evidence suggests that breast cancer risk can be reduced by dietary measures. Study of the relationships between dietary habits and prognosis in patients with breast cancer is essential to the design of diet intervention trials. PURPOSE: Our purpose was to determine whether dietary habits are associated with disease-free survival in patients with breast cancer who have undergone treatment. METHODS: We interviewed 240 women about their dietary histories. These women were 50-65 years old and had pathological stage I-II breast cancer with subsequent follow-up for 4 years; 209 of these women were postmenopausal. Differences in dietary variables between groups of patients were analyzed with bivariate and multivariate statistical methods. RESULTS: Cancers were classified as estrogen receptor (ER) rich (> or = 0.10 fmol/micrograms of DNA) in 149 patients and as ER poor (< 0.10 fmol/micrograms of DNA) in 71 patients. Fifty-two patients had treatment failure during follow-up. The 30 patients with ER-rich tumors who had treatment failure reported higher intakes of total fat, saturated fatty acids, and polyunsaturated fatty acids than did the 119 patients with ER-rich tumors who did not have treatment failure. The multiple-odds ratio (OR) for treatment failure in these women was 1.08 for each 1% increment in percentage of total energy (E%) from total fat. For treatment failure within the first 2 years, the OR was 1.19 for each 1-mg increase in vitamin E intake per 10 megajoules of energy. In women with treatment failure 2-4 years after diagnosis, ORs were 1.13 and 1.23 for each E% increment in total fat or saturated fatty acids, respectively. No association between dietary habits and treatment failure was found for women with ER-poor cancers. There was a tendency to a dose-response relationship (in quartiles) between intake of saturated fatty acids and disease-free survival, but the observed differences were not statistically significant. CONCLUSIONS: Dietary habits at the time of diagnosis may affect prognosis, at least for patients with ER-rich breast cancers. Dietary fat may have an effect on growth or spread of breast cancer, both of which may vary according to type of fat. Total fat and saturated fatty acids were the dietary parameters most strongly associated with risk for treatment failure. IMPLICATIONS: Dietary intervention might serve as an adjuvant treatment to improve breast cancer prognosis.

Dietary habits and prognostic factors in breast cancer.

The relationship between dietary habits and prognostic factors for breast cancer was studied in 240 women aged 50-65 years who had surgery for breast cancer between 1983 and 1986. A dietary history interview was conducted within the 4 months following resection of the primary tumor. In the stepwise multivariate analysis, the multiple-odds ratio (OR) for having a tumor greater than or equal to 20 mm in diameter was 0.95 (95% confidence interval, 0.91-0.99) for each 1-g increase in fiber intake per 10 MJ of energy intake. Compared with patients having tumors poor in estrogen receptor (ER), those having ER-rich tumors (greater than or equal to 0.10 fmol/microgram of DNA) were older (P less than .01) and reported carbohydrate intake yielding higher E% (percentage of total energy intake) (P less than .01) and higher retinol intake per 10 MJ (P less than .05). The OR for having an ER-rich tumor was 1.58 (95% confidence interval, 1.08-2.31) for each 1-mg increase in retinol intake per 10 MJ; 1.09 (95% confidence interval, 1.02-1.16) for each additional year of age; and 1.08 (95% confidence interval, 1.02-1.13) for each 1% increment in E% from carbohydrates. These results suggest that the dietary patterns of the western world (e.g., high fat intake and low intake of carbohydrates and fiber) affect certain prognostic factors in breast cancer, such as tumor size and ER content of the tumor.

Cancer risk after iodine-131 therapy for hyperthyroidism.

Cancer incidence was studied in 10,552 patients (mean age, 57 years) who received 131I therapy (mean dose, 506 MBq) for hyperthyroidism between 1950 and 1975. Follow-up on these patients was continued for an average of 15 years. Record linkage with the Swedish Cancer Register for the period 1958-1985 identified 1543 cancers occurring 1 year or more after 131I treatment, and the standardized incidence ratio (SIR) was 1.06 (95% confidence interval = 1.01-1.11). Significantly increased SIRs were observed for cancers of the lung (SIR = 1.32; n = 105) and kidney (SIR = 1.39; n = 66). Among 10-year survivors, significantly elevated risks were seen for cancers of the stomach (SIR = 1.33; n = 58), kidney (SIR = 1.51; n = 37), and brain (SIR = 1.63; n = 30). Only the risk for stomach cancer, however, increased over time (P less than .05) and with increasing activity administered (P = not significant). The risk for malignant lymphoma was significantly below expectation (SIR = 0.53; n = 11). Overall cancer risk did not increase with administered 131I dose or with time since exposure. The absence of any increase in leukemia adds further support to the view that a radiation dose delivered gradually over time is less carcinogenic than the same total dose received over a short time. Only for stomach cancer was a possible radiogenic excess suggested.

Thyroid cancer after diagnostic doses of iodine-131: a retrospective cohort study.

The incidence of thyroid cancer was evaluated in 35,074 patients examined for suspected thyroid disorders between 1951 and 1969 with an average of 1.92 megabecquerel [(MBq) 52 microCi] of 131I. The radiation dose to the thyroid gland was, on the average, approximately 0.5 Gy. The mean age at the time of examination was 44 years; 5% were under age 20. Patients were followed for an average of 20 years. Record linkage with the Swedish Cancer Register identified 50 thyroid cancers occurring 5 years or more after the initial 131I examination, in contrast to 39.4 expected based on general population rates [standardized incidence ratio (SIR) = 1.27, 95% confidence interval = 0.94-1.67]. Risk was highest among males (SIR = 2.70, n = 10), patients followed 5-9 years (SIR = 2.22, n = 23), and patients receiving more than 74 microCi or 2.74 MBq of 131I (SIR = 2.04, n = 17). However, these observations were confounded by the fact that patients examined for a suspected thyroid tumor received the highest 131I exposures and were at highest overall risk (SIR = 2.77, n = 34). Patients given 131I for reasons other than a suspected tumor were not at increased risk (SIR = 0.62, n = 16). Patients anticipated to be at highest risk, i.e., women (SIR = 1.12, n = 40) and those observed for 10 years or more (SIR = 0.93, n = 27), showed no evidence of a dose response. Overall, these data provide little proof that 131I is carcinogenic in humans and support the notion that the carcinogenic potential of internal 131I beta particles might be as low as four times less than external x rays or gamma rays.

Cancer risk in population examined with diagnostic doses of 131I.

Previously, we conducted a study of 35,074 patients receiving diagnostic doses of 131I for suspected thyroid disorders between 1951 and 1969. We reported that, between 1958 and 1984, the incidence of thyroid cancers in these patients was insignificantly greater than the incidence expected in the general population. This increase was attributed to the underlying condition that prompted the examination and not to the administration of 131I. The purpose of the present study was to analyze the total cancer risk in the same cohort of patients examined with diagnostic doses of 131I. To further evaluate the underlying risk of disease in these patients, we compared the incidence of all cancers with that expected in the general population. The average radiation dose was approximately 500 mGy to the thyroid and less than 10 mGy to other organs. In the 35,074 patients, 3,746 cancers occurred following the first 5 years after examination, and the resulting standardized incidence rate (SIR) was 1.01 (95% confidence interval = 0.98-1.04). SIRs were significantly increased for endocrine tumors other than thyroid cancer (1.93) and for lymphomas (1.24), leukemias (1.34), and nervous system tumors (1.19). The risk of leukemia was similar for chronic lymphocytic leukemia (CLL) (SIR = 1.30) and non-CLL (SIR = 1.34). SIR was significantly decreased for cancers of the female genital organs (0.86). The risk for cancer of all sites and types combined was highest 5-9 years after examination (SIR = 1.07) and did not differ from unity thereafter. With greater than or equal to 10 years of follow-up, risk was not statistically associated with the dose of 131I. Overall, the data exclude cancer risk increments greater than 5% (SIR = 1.05) with 95% confidence. The significant increase in the risk of non-CLL, a prominent radiogenic malignancy, however, warrants special attention. We are continuing our study to determine the possible factors involved in the significant increase in the risk of leukemia.

Detection of genetic variation affecting milk coagulation properties in Danish Holstein dairy cattle by analyses of pooled whole-genome sequences from phenotypically extreme samples (pool-seq).

Rennet-induced milk coagulation is an important trait for cheese production. Recent studies have reported an alarming frequency of cows producing poorly coagulating milk unsuitable for cheese production. Several genetic factors are known to affect milk coagulation, including variation in the major milk proteins; however, recent association studies indicate genetic effects from other genomic regions as well. The aim of this study was to detect genetic variation affecting milk coagulation properties, measured as curd-firming rate (CFR) and milk pH. This was achieved by examining allele frequency differences between pooled whole-genome sequences of phenotypically extreme samples (pool-seq).. Curd-firming rate and raw milk pH were measured for 415 Danish Holstein cows, and each animal was sequenced at low coverage. Pools were created containing whole genome sequence reads from samples with "extreme" values (high or low) for both phenotypic traits. A total of 6,992,186 and 5,295,501 SNP were assessed in relation to CFR and milk pH, respectively. Allele frequency differences were calculated between pools and 32 significantly different SNP were detected, 1 for milk pH and 31 for CFR, of which 19 are located on chromosome 6. A total of 9 significant SNP, which were selected based on the possible function of proximal candidate genes, were genotyped in the entire sample set ( = 415) to test for an association. The most significant SNP was located proximal to , explaining 33% of the phenotypic variance. , coding for κ-casein, is the most studied in relation to milk coagulation due to its position on the surface of the casein micelles and the direct involvement in milk coagulation. Three additional SNP located on chromosome 6 showed significant associations explaining 7, 3.6, and 1.3% of the phenotypic variance of CFR. The significant SNP on chromosome 6 were shown to be in linkage disequilibrium with the SNP peaking proximal to ; however, after accounting for the genotype of the peak SNP within this QTL, significant effects (-value < 0.1) could still be detected for 2 of the SNP accounting for 2 and 1% of the phenotypic variance. These 2 interesting SNP were located within introns or proximal to the candidate genes-solute carrier family 4 (sodium bicarbonate cotransporter), member 4 () and LIM and calponin homology domains 1 (), respectively-making them interesting targets for further analysis.

A microsatellite linkage map of rainbow trout (Oncorhynchus mykiss) characterized by large sex-specific differences in recombination rates.

We constructed a genetic linkage map for a tetraploid derivative species, the rainbow trout (Oncorhynchus mykiss), using 191 microsatellite, 3 RAPD, 7 ESMP, and 7 allozyme markers in three backcross families. The linkage map consists of 29 linkage groups with potential arm displacements in the female map due to male-specific pseudolinkage arrangements. Synteny of duplicated microsatellite markers was used to identify and confirm some previously reported pseudolinkage arrangements based upon allozyme markers. Fifteen centromeric regions (20 chromosome arms) were identified with a half-tetrad analysis using gynogenetic diploids. Female map length is approximately 10 M, but this is a large underestimate as many genotyped segments remain unassigned at a LOD threshold of 3.0. Extreme differences in female:male map distances were observed (ratio F:M, 3.25:1). Females had much lower recombination rates (0.14:1) in telomeric regions than males, while recombination rates were much higher in females within regions proximal to the centromere (F:M, 10:1). Quadrivalent formations that appear almost exclusively in males are postulated to account for the observed differences.

DNA pattern and dietary habits in patients with breast cancer.

An association between dietary fat, micronutrients and breast cancer aetiology and prognosis has been found in studies of experimental animals and in epidemiological studies. The relationship between dietary habits and the nuclear DNA content of breast cancer cells was studied in 82 women aged 50-65 years. A dietary history interview was conducted within 4 months following surgery. Patients having tumours with euploid DNA pattern reported lower mean intake of saturated fatty acids (FA) in absolute terms, lower mean intake of total fat, saturated FA, and monounsaturated FA, in percentage of total energy intake (E%), a higher E% from protein, and a higher intake of vitamin D, and selenium per 10 MJ than did patients having tumours with aneuploid DNA pattern. In the stepwise logistic regression analysis, the multivariate odds ratios (OR) for having a tumour with aneuploid DNA pattern was 1.16 (95% confidence interval, 1.04-1.28) for each 1 g increase in intake of total fat (E%) and 0.95 (95% confidence interval, 0.92-0.99) for each mg increase in selenium intake per 10 MJ. When total fat was substituted with types of fat, the OR for having a tumour with aneuploid DNA pattern was 1.30 (95% confidence interval, 1.07-1.59) for each 1 g increase in intake of saturated FA (E%). These results suggest a correlation between a diet rich in fat and protein and the DNA content of breast cancer cells.

Cancer occurring after radiotherapy and chemotherapy.

Radiotherapy and chemotherapy can effectively control cancer but can also cause new cancers to develop as long-term complications. Almost all types of cancer have been associated with radiotherapy. The breast, thyroid, and bone marrow are the organs most susceptible to radiation carcinogenesis. The bone marrow is also most frequently involved by chemotherapy and the leukemia risk is much higher than after radiotherapy. The combination of intensive radiotherapy and chemotherapy is particularly leukemogenic. The latent period between radiotherapy/chemotherapy and the appearance of a second primary cancer ranges from a few years to several decades. The risk for a second primary cancer following radiotherapy or chemotherapy emphasizes the need for life long follow-up of patients receiving such treatments. This is particularly the case in individuals with long life expectancy, for example, patients treated for childhood neoplasms. The benefits of radiotherapy and chemotherapy in oncology exceed the risks for second primary cancers. Efforts should be directed towards identifying those patients who will benefit from the treatments so that only they are exposed to the risk.

Changing annual incidence of hypothyroidism after iodine-131 therapy for hyperthyroidism, 1951-1975.

The incidence of hypothyroidism was analyzed in 4,553 hyperthyroid patients treated with I-131 between 1951 and 1975. The average annual rate of hypothyroidism during the first 7 yr after therapy increased continuously for each 5-yr period of treatment, from 3.6% for patients treated between 1951 and 1955 to 7.7% for patients treated during the period from 1971 though 1975 (p less than 0.001). The increase in the incidence of hypothyroidism was seen before the introduction of TSH assays and when allowance was made for thyroid gland size, the age of the patient, and the experience of the radiotherapist. The use of TSH assays has probably resulted in an earlier recognition of hypothyroidism, which may explain why the most marked rise in the incidence of hypothyroidism was observed after the introduction of TSH assays.

Incidence of hypothyroidism occurring long after iodine-131 therapy for hyperthyroidism.

We have studied the long-term incidence of hypothyroidism in 4,473 formerly hyperthyroid patients given I-131 therapy between 1951 and 1975. The mean age at the first I-131 treatment was 56 yr. Six percent developed hypothyroidism within one year of therapy, and 72% within 26 yr. Prior antithyroid medication did not affect the incidence of hypothyroidism. Patients cured with one dose of I-131 had a lower cumulative long-term incidence of hypothyroidism than those requiring more than one dose.

Mortality from leukemia after irradiation in infancy for skin hemangioma.

From 1920 through 1959, 14,624 infants were exposed to ionizing radiation for skin hemangioma at Radiumhemmet. They were all less than 18 months old (mean 6 months) at the time of first exposure. The irradiated hemangiomas were located all over the body. The weighted bone marrow dose was on average 0.13 Gy (range < 0.01-4.6 Gy). During the period 1920-1986, 20 deaths from leukemia (11 childhood and 9 adult) were observed in the cohort compared with 17 expected during the same period. There were no significant associations between childhood leukemia and radiation dose. Despite the relatively large number of infants studied, the low average dose to bone marrow limited the possibility of detecting a small radiation risk as might be predicted from other radiation studies.