A meta-analysis of Hodgkin lymphoma reveals 19p13.3 TCF3 as a novel susceptibility locus.

Recent genome-wide association studies (GWAS) of Hodgkin lymphoma (HL) have identified associations with genetic variation at both HLA and non-HLA loci; however, much of heritable HL susceptibility remains unexplained. Here we perform a meta-analysis of three HL GWAS totaling 1,816 cases and 7,877 controls followed by replication in an independent set of 1,281 cases and 3,218 controls to find novel risk loci. We identify a novel variant at 19p13.3 associated with HL (rs1860661; odds ratio (OR)=0.81, 95% confidence interval (95% CI) = 0.76-0.86, P(combined) = 3.5 × 10(-10)), located in intron 2 of TCF3 (also known as E2A), a regulator of B- and T-cell lineage commitment known to be involved in HL pathogenesis. This meta-analysis also notes associations between previously published loci at 2p16, 5q31, 6p31, 8q24 and 10p14 and HL subtypes. We conclude that our data suggest a link between the 19p13.3 locus, including TCF3, and HL risk.

Fecal microbiota diversity in survivors of adolescent/young adult Hodgkin lymphoma: a study of twins.

BACKGROUND: Adolescent/young adult Hodgkin lymphoma (AYAHL) survivors report fewer exposures to infections during childhood compared with controls, and they have functional lymphocyte aberrations. The gut microbiota plays a central role in immunity. METHODS: We investigated whether fecal microbial diversity differed between 13 AYAHL survivors and their unaffected co-twin controls. Pyrosequencing of fecal bacterial 16S rRNA amplicons yielded 252 943 edited reads that were assigned to species-level operational taxonomic units (OTUs) and standardised for sequencing depth by random sampling. Microbial diversity was compared within vs between twin pairs and by case-control status. RESULTS: The number of unique OTUs was more similar within twin pairs compared with randomly paired participants (P=0.0004). The AYAHL cases had fewer unique OTUs compared with their co-twin controls (338 vs 369, P=0.015); this difference was not significant (169 vs 183, P=0.10) when restricted to abundant OTUs. CONCLUSION: In this small study, AYAHL survivors appear to have a deficit of rare gut microbes. Further work is needed to determine if reduced microbial diversity is a consequence of the disease, its treatment, or a particularly hygienic environment.

Evidence of genetic susceptibility to infectious mononucleosis: a twin study.

Infectious mononucleosis is a clinical manifestation of primary Epstein-Barr virus infection. It is unknown whether genetic factors contribute to risk. To assess heritability, we compared disease concordance in monozygotic to dizygotic twin pairs from the population-based California Twin Program and assessed the risk to initially unaffected co-twins. One member of 611 and both members of 58 twin pairs reported a history of infectious mononucleosis. Pairwise concordance in monozygotic and dizygotic pairs was respectively 12·1% [standard error (s.e.)=1·9%] and 6·1% (s.e.=1·2%). The relative risk (hazard ratio) of monozygotic compared to dizygotic unaffected co-twins of cases was 1·9 [95% confidence interval (CI) 1·1-3·4, P=0·03], over the follow-up period. When the analysis was restricted to same-sex twin pairs, that estimate was 2·5 (95% CI 1·2-5·3, P=0·02). The results are compatible with a heritable contribution to the risk of infectious mononucleosis.

Heritable breast cancer in twins.

Known major mutations such as BRCA1/2 and TP53 only cause a small proportion of heritable breast cancers. Co-dominant genes of lower penetrance that regulate hormones have been thought responsible for most others. Incident breast cancer cases in the identical (monozygotic) twins of representative cases reflect the entire range of pertinent alleles, whether acting singly or in combination. Having reported the rate in twins and other relatives of cases to be high and nearly constant over age, we now examine the descriptive and histological characteristics of the concordant and discordant breast cancers occurring in 2310 affected pairs of monozygotic and fraternal (dizygotic) twins in relation to conventional expectations and hypotheses. Like other first-degree relatives, dizygotic co-twins of breast cancer cases are at higher than usual risk (standardised incidence ratio (SIR)=1.7, CI=1.1-2.6), but the additional cases among monozygotic co-twins of cases are much more numerous, both before and after menopause (SIR=4.4, CI=3.6-5.6), than the 100% genetic identity would predict. Monozygotic co-twin diagnoses following early proband cancers also occur more rapidly than expected (within 5 years, SIR=20.0, CI=7.5-53.3). Cases in concordant pairs represent heritable disease and are significantly more likely to be oestrogen receptor-positive than those of comparable age from discordant pairs. The increase in risk to the monozygotic co-twins of cases cannot be attributed to the common environment, to factors that cumulate with age, or to any aggregate of single autosomal dominant mutations. The genotype more plausibly consists of multiple co-existing susceptibility alleles acting through heightened susceptibility to hormones and/or defective tumour suppression. The resultant class of disease accounts for a larger proportion of all breast cancers than previously thought, with a rather high overall penetrance. Some of the biological characteristics differ from those of breast cancer generally.

Development and representativeness of a large population-based cohort of native Californian twins.

We have established a large cohort of twins to facilitate studies of the role of genetics and environment in the development of disease. The cohort has been derived from all multiple births occurring in California between 1908-82 (256,616 in total). We report here on our efforts to contact these twins and their completion of a detailed 16 page risk factor questionnaire. Addresses of the individuals were obtained by linking the birth records with the California Department of Motor Vehicles (DMV) roster of licensees. To date this has been completed for twins born between 1908 and 1972 (200,589 individuals). The linkage has revealed 112,468 matches and, because of less complete DMV records in some years, was less successful in older females than in younger females and all males. Over 41,000 twins have participated by completing the questionnaire. Based on estimates of numbers of individuals receiving a questionnaire, we estimate our crude response rate to be between 42.2% and 49.6%, highest among females in their 40s (62.8%). We describe the representativeness of the twins in the original birth cohort, those identified by the linkage, and those completing the questionnaire. Compared to the 1990 resident population of California-born resident singletons, the respondents were of similar age, sex, race and residential distribution (for although we were able to locate fewer older females, they had a higher response rate), but were less likely to have been educated for more than 12 years. We provide a brief synopsis of studies nested within this cohort. We also elucidate our plans for expanding the cohort in the near future.

High constant incidence in twins and other relatives of women with breast cancer.

The incidence of breast cancer rises steeply between ages 25 and 50, and more slowly thereafter. In contrast, the incidence in the unaffected (contralateral) breast of women who have had breast cancer remains constant at about 0.7% per year for at least the next 20 years after diagnosis, irrespective of age at first diagnosis. The incidence in relatives of the patients seems to show a similar pattern. The incidence in a prospective study of monozygotic twins of patients was approximately constant at 1.3% per year (77 cases), again about 0.7% per breast. At ages older than a patient's age at diagnosis, her mother and sisters have an incidence of 0.3-0.4% per year. Above the index patient's age at diagnosis, the rate in relatives shows no temporal trend and is independent of the patient's age at diagnosis. A statistically simple explanation is that incidence in susceptible women increases to a high constant level by a predetermined age that varies between families, but this seems inconsistent with conventional models of carcinogenesis and susceptibility. The very high incidence in monozygotic twins of patients indicates that a high proportion, and perhaps the majority, of breast cancers arise in a susceptible minority of women.

Use of twins as mutual proxy respondents in a case-control study of breast cancer: effect of item nonresponse and misclassification.

A case-control study of breast cancer in twins diagnosed before 1988 was used to characterize the effects on odds ratios when proxy responses from co-twins are used. North American disease-discordant pairs were ascertained through advertisements, and mailed questionnaires were returned from both members of 671 pairs and from one member of 391 pairs. Biases from the proxy response were attributed to nonresponse or misclassification. Nonresponse varied according to type of exposure variable, depth of detail requested, joint exposure status of the pair, respondent's case-control status, zygosity, and social closeness of the pair. Misclassification was minimal, generally nondifferential, and a high degree of reliability between the proxy and self-report was indicated by the kappa statistic and the intraclass correlation coefficient. By using double-respondent pairs, a method was developed to adjust proxy responses for both sources of bias. These adjustments resulted in minor changes to the odds ratios for the variables studied (age at menarche, reproductive factors, and hormone use). A larger difference was observed between the odds ratios based on all pairs and those based on double-respondent pairs only. These findings demonstrate that, for these variables in this population, twins are reliable proxies for each other and that results from single-respondent pairs should be included.

Representativeness of a roster of volunteer North American twins with chronic disease.

To identify large numbers of twins affected by chronic disease as potential subjects for studies of environmental and genetic chronic disease determinants, we advertised for affected twins over the period 1980-91 in newspapers across North America. Responses were received from 17 245 twin pairs in which cases of cancer or other chronic disease had occurred. To assess the representativeness of affected twins identified by advertising, we evaluated the pattern of reporting, compared the cases identified to the number of cases estimated to be prevalent among all North American twins, compared the cases to population-based singleton case series, compared the healthy co-twins to population-based samples of healthy persons, assessed the impact on ascertainment of opinions about disease causation, compared the pattern of prospective to retrospective ascertainment of disease in the originally unaffected co-twins of cases, and compared the results of the prospective ascertainment of disease in co-twins to comparable published estimates. Youth, gender, zygosity, education, and disease concordance were found to be overall determinants of ascertainment. Disease-discordant DZ twins appeared to be modestly underascertained. While somewhat better educated, both concordant and discordant pairs were judged to be reasonably representative of affected non-Hispanic white North American twin pairs of comparable status, ie of comparable age, sex, race, and zygosity. If interpreted with caution, the concordance patterns of such twins can be used to generate genetic hypotheses, but should not be the basis of definitive heritability analyses. We conclude that advertising offers a method of identifying pairs of twins that can serve as subjects for studies designed to identify disease determinants.

The risk of angiosarcoma following primary breast cancer.

Lymphangiosarcoma of the upper extremity is a rare and aggressive tumour reported to occur following post-mastectomy lymphoedema (Stewart-Treves syndrome). Haemangiosarcoma, a related rare tumour, has occasionally been reported to occur in the breast following irradiation. We conducted a case-control study using the University of Southern California-Cancer Surveillance Program, the population-based cancer registry for Los Angeles County, to evaluate the relationship between invasive female breast cancer and subsequent upper extremity or chest lymphangiosarcoma and haemangiosarcoma together referred to as angiosarcoma. Cases were females diagnosed between 1972 and 1995 with angiosarcoma of the upper extremity (n = 20) or chest (n = 48) who were 25 years of age or older and residing in Los Angeles County when diagnosed. Other sarcomas at the same anatomic sites were also studied. Controls were females diagnosed with cancers other than sarcoma during the same time period (n = 266,444). Cases and controls were then compared with respect to history of a prior invasive epithelial breast cancer. A history of breast cancer increased the risk of upper extremity angiosarcoma by more than 59-fold (odds ratio [OR] = 59.3, 95% confidence interval [95% CI] = 21.9-152.8). A strong increase in risk after breast cancer was also observed for angiosarcoma of the chest and breast (OR = 11.6, 95% CI = 4.3-26.1) and for other sarcomas of the chest and breast (OR = 3.3, 95% CI = 1.1-1.7).

Impaired cytotoxic T lymphocyte activity in systemic lupus erythematosus following in vitro polyclonal T cell stimulation: a contributory role for non-T cells.

To determine whether non-T cells contribute to impaired generation of nonrestricted cytotoxic T lymphocyte (CTL) activity in human SLE, peripheral blood mononuclear cells (PBMC) and sort-purified T cells from normal subjects and SLE patients were stimulated with anti-CD3 mAb, maintained in IL2, and assayed for cytolytic activity against 51Cr-labeled Daudi target cells. In addition, T cell and non-T cell fractions were isolated from nine pairs of monozygotic (MZ) twins discordant for SLE, reconstituted in a criss-cross pattern, and stimulated and assayed for cytolytic activity. Cytolytic responses were significantly lower in SLE PBMC cultures than in normal PBMC cultures. Addition of SLE serum to normal PBMC cultures did not inhibit generation of normal cytolytic responses, and neither 'resting' SLE PBMC prior to stimulation nor addition of neutralizing anti-IL10 mAb or costimulating anti-CD28 mAb restored generation of SLE cytolytic responses to normal. Nevertheless, despite the significantly greater cytolytic responses in normal PBMC cultures than in SLE PBMC cultures, cytolytic responses in normal purified T cell cultures were only modestly and insignificantly greater than those in SLE purified T cell cultures. Moreover, substitution of 'healthy' non-T cells for SLE non-T cells in four of the nine MZ twin-pairs appreciably enhanced cytolytic responses, and substitution of SLE non-T cells for 'healthy' non-T cells in five of the seven twin-pairs tested appreciably diminished cytolytic responses. Taken together, these results indicate that, in addition to any inherent SLE T cell abnormalities, impaired function of SLE non-T cells contributes to impaired generation of nonrestricted CTL activity.

Estrogen-progestin replacement therapy and endometrial cancer.

BACKGROUND: It has been known for more than 20 years that estrogen replacement therapy substantially increases a woman's risk of developing endometrial cancer. To reduce this increased risk, progestins have been added to estrogen replacement therapy for between 5 and 15 days (usually 7 or 10 days) per "month" in a sequential fashion (sequential estrogen-progestin replacement therapy) or with each dose of estrogen replacement therapy (continuous combined replacement therapy). At the present time, however, little is known about the effects of varying the number of days that progestin is used in sequential estrogen-progestin replacement therapy. PURPOSE: We sought to determine the effects of sequential estrogen-progestin replacement therapy and continuous combined replacement therapy on a woman's risk of developing endometrial cancer. METHODS: A population-based, case-control study of 833 case subjects and 791 control subjects was conducted. Women were postmenopausal, white, and aged 50-74 years when first diagnosed with invasive endometrial cancer or were aged 50-74 years at the matching date for control subjects. All subjects were interviewed in person with the aid of a month-by-month calendar. Relative risks were estimated by odds ratios (ORs); ORs were adjusted simultaneously for the different forms of hormone replacement therapy and for the known endometrial cancer risk factors. RESULTS: The adjusted OR was 2.17 (95% confidence interval [CI] = 1.91-2.47) per 5 years of estrogen replacement therapy use (based on 422 users among the case subjects and 262 users among the control subjects). For women who received sequential estrogen-progestin replacement therapy with the progestin given for less than 10 days (effectively 7 days) per month, the adjusted OR was only slightly reduced to 1.87 (95% CI = 1.32-2.65) per 5 years of use (74 case subjects and 47 control subjects). However, when progestin was given for 10 or more days (effectively 10 days), there was essentially no increased risk (adjusted OR = 1.07 per 5 years of use; 95% CI = 0.82-1.41) (79 case subjects and 88 control subjects). Continuous combined replacement therapy was also associated with essentially no increased risk (adjusted OR = 1.07 per 5 years of use; 95% CI = 0.80-1.43) (94 case subjects and 81 control subjects). CONCLUSIONS: The progestin in sequential estrogen-progestin replacement therapy needs to be given for at least 10 days to block effectively any increased risk of endometrial cancer. Continuous combined estrogen-progestin therapy is similarly effective. Neither regimen reduces a woman's underlying risk of endometrial cancer. The sharp distinction between the effects of less than 10 days (effectively 7 days) and 10 or more days (effectively 10 days) of progestin use in sequential estrogen-progestin replacement therapy suggests that the extent of endometrial sloughing may play a critical role in determining endometrial cancer risk.

Smoking, alcohol use, dietary factors and risk of small intestinal adenocarcinoma.

To investigate the role of tobacco and alcohol use and dietary factors in the etiology of small intestinal adenocarcinoma, we analyzed data from a large population-based case-control study of multi-site cancers conducted in Los Angeles County between 1975 and 1984. The present analysis included interview information on 36 small intestinal adenocarcinoma patients and 998 population controls. After adjusting for age and ethnicity, men who smoked more than 100 cigarettes during their lifetimes were at a non-significantly 3-fold increased risk for small intestinal adenocarcinoma; this association was substantially weaker in women. In men and women combined, a significant 3-fold increased risk in heavy drinkers (80+ g ethanol/day) relative to more moderate drinkers and non-drinkers was observed. Although frequent (>6 times vs. less than 2 times of intake a week) intake of foods rich in heterocyclic aromatic amines (based on the combined intake of fried bacon and ham, barbecued and/or smoked meat and smoked fish) was associated with a significant 4.5-fold increased risk of small intestinal adenocarcinoma in men; this association was not present in women. Based on 2 questions that provided a crude assessment of sugar intake, risk of small intestinal adenocarcinoma in men and women appeared to be associated with adding sugar regularly in coffee or tea and daily intake of non-diet carbonated soft drinks. When we computed total sugar intake from tea, coffee and non-diet carbonated soft drinks, there was a consistent and significant trend of increasing sugar intake and risk of small intestinal adenocarcinomas. Compared with the lowest intake level a day (<5 g), medium (5-25 g) and high intakes (>25 g) were associated with ORs of 2.5 and 3.8, respectively.

Impaired recovery and cytolytic function of CD56+ T and non-T cells in systemic lupus erythematosus following in vitro polyclonal T cell stimulation. Studies in unselected patients and monozygotic disease-discordant twins.

OBJECTIVE: To determine whether there is impaired generation and cytolytic function of CD56+ T cells and non-T cells in human systemic lupus erythematosus (SLE). METHODS: Peripheral blood mononuclear cells (PBMC) were obtained from 73 patients with SLE, 39 normal controls, and 9 pairs of monozygotic (MZ) twins discordant for SLE. PBMC were stimulated with anti-CD3 monoclonal antibody, maintained in interleukin-2, and assayed for percentages of total CD56+ cells and CD56+ T cells by flow cytometry, and for cytolytic activity against 51Cr-labeled Daudi target cells. RESULTS: Despite normal total cell expansion, the percentages of recovered CD56+ T cells and total CD56+ cells were 1.6-fold and 1.8-fold lower, respectively, in patients with SLE compared with normal controls (P = 0.011 and P < 0.001, respectively). Cytolytic activities of isolated total CD56+ cells and CD56+ T cells and were also reduced in patients with SLE compared with normal controls (P = 0.033). These defects associated with SLE were independent of disease activity and immunosuppressive medications, and they reflected impaired maturation of cytolytic effector cells rather than a deficiency in precursor cell number. In MZ twins discordant for SLE, recovered percentages of CD56+ cells and cytolytic responses were very low in 4 of 8 and 6 of 9 co-twins with SLE, respectively. Cellmixing experiments with the PBMC of the MZ twins demonstrated that the E+ cell fractions (containing all T cells and CD56+ non-T cells) from the co-twins with SLE had decreased ability to generate cytolytic activity compared with the corresponding E+ cell fractions from the healthy co-twins. However, recovered percentages of CD56+ cells and non-T cells and cytolytic responses were also depressed in 4 of 8 and 4 of 9 healthy co-twins, respectively. CONCLUSION: Impaired CD56+ T cell and non-T cell responses are a feature of SLE and may antedate the onset of clinical disease.

Concordance for childhood cancer in twins.

The causes of most childhood cancer remain elusive; some children clearly have a genetic predisposition, but in the majority the relative contributions of environmental and host factors are not established. One approach to this question is through twin concordance studies, but only the most common malignancy, acute leukemia, has been studied to date, owing to the rarity of other forms of childhood cancer. The aim of the study was to determine the concordance rates for childhood cancer in twins, in order to clarify the importance of constitutional predisposition for a range of tumor types. Twins with cancer were ascertained through three cooperative clinical trials groups, a cancer-twin registry, and a large pediatric hospital. Subjects were sent a postal questionnaire requesting information on cancer concordance and zygosity. Data were obtained on 556 twins with cancer. Three twin pairs, out of 197 twin pairs (76 monozygous, MZ, twin pairs), were concordant for leukemia, giving an MZ case-wise concordance rate (5%) that is substantially lower than previously reported. The case-wise concordance for non-retinoblastoma solid tumors was 2.2%: Two twin pairs were concordant for CNS tumors, one was concordant for neuroblastoma, and two twin pairs were concordant for cancer but not for the type of cancer. The results of the present study, together with previous data from population studies of siblings and offspring, suggest that there is not in general a strong constitutional genetic component for childhood cancers other than retinoblastoma.

Proportional melanoma incidence and occupation among white males in Los Angeles County (California, United States).

A case-control analysis of cancer registry data was used to examine the hypothesis that occupational exposure to sunlight influences the risk of melanoma. Occupation at diagnosis was available for 3,527 cutaneous melanomas and 53,129 other cancers identified by the Los Angeles County (California, United States) Cancer Surveillance Program among non-Spanish-surnamed White males aged 20 to 65 years between 1972 and 1990. Occupational exposure to sunlight was assessed by blinded expert coding of job titles as indoor, outdoor, and mixed indoor/outdoor. Relative to indoor occupations, proportionate odds ratios (OR) adjusted for age, level of education, and birthplace were 1.16 (95 percent confidence interval [CI] = 1.07-1.27) for indoor/outdoor occupations and 1.15 (CI = 0.94-1.40) for outdoor occupations. However, increasing levels of the education or training required for the occupation was associated more strongly with increased melanoma occurrence (ORs adjusted for age, occupational sun exposure, and birthplace, were 1.0, 1.63, 2.09, 2.23, and 2.99 for low-skill occupation, high school, college, postgraduate, and doctoral levels, respectively). Analysis of melanoma occurrence by job titles confirmed a clear variation by the required education or training level but not by the category of occupational sunlight exposure. The findings suggest that lifestyle factors associated with higher levels of education may be more important determinants of melanoma risk than characteristics of the work environment.

Epidemiology of smallpox in west Pakistan. III. Outbreak detection and interlocality transmission. 1971.

During one year, 121 outbreaks of variola major were detected in 99 of the 1717 localities within a rural area of West Pakistan with a total population of approximately 1.2 million. Only 19% of the outbreaks, representing 36% of the 1040 investigated cases, were officially reported, although potential strengths in the government surveillance system were also identified. Persistence of smallpox within the area depended on introductions from the outside, and more than one-half of all outbreaks of known source could be ultimately traced to cities. Within the study area, outbreaks with the largest numbers of cases and those in the larger communities were the ones from which smallpox was most frequently transported. The frequency with which variola was introduced into localities was directly related to population size and to the presence of medical care facilities. Trips between localities by infected individuals were extremely rare events. They were made most often during the late fall and winter, primarily during the incubation period of the disease, and did not differ in purpose, means or destination from journeys unassociated with smallpox. Individuals at relatively high risk of becoming introducers included the unvaccinated (primarily children under five), the unschooled and those not native to the area. Vaccination priorities based on these findings could increase the efficiency of smallpox eradication efforts.

Concordance for Hodgkin's disease in identical twins suggesting genetic susceptibility to the young-adult form of the disease.

BACKGROUND: Relatives of young adults with Hodgkin's disease are at increased risk of Hodgkin's disease, and lines of evidence implicate both inheritance and environment. METHODS: We have identified and followed 432 sets of twins affected by Hodgkin's disease. The number of cases of Hodgkin's disease observed before the age of 50 years in the healthy monozygotic and dizygotic twins of the patients with Hodgkin's disease was compared with the number expected from national age-specific incidence rates. RESULTS: None of the 187 pairs of dizygotic twins became concordant for Hodgkin's disease, whereas 10 of the 179 pairs of monozygotic twins did; in 5 of these pairs, the second case appeared after the original ascertainment. During the observation period, 0.1 (monozygotic) and 0.1 (dizygotic) cases in the unaffected twins were expected. Monozygotic twins of patients with Hodgkin's disease thus had a greatly increased risk (standardized incidence ratio, 99; 95 percent confidence interval, 48 to 182), whereas no increase in the risk for dizygotic twins of patients with Hodgkin's was observed. CONCLUSIONS: Genetic susceptibility underlies Hodgkin's disease in young adulthood.

Sarcomas and other malignancies of soft tissue, retroperitoneum, peritoneum, pleura, heart, mediastinum, and spleen.

BACKGROUND: Malignant neoplasms of the structural tissues, consisting mostly of soft tissue sarcomas, are morphologically diverse and rarely treated for epidemiologic purposes as individual entities. Our understanding to date of the pattern of occurrence of sarcomas is based largely on reports of limited individual clinic experience or case-control studies, each driven by a single hypothesis, and there have been virtually no descriptions according to specific morphologic type. METHODS: The accumulated coverage of the SEER populations offers an opportunity to correct this deficit. Each of the diagnoses has been reported and coded using a single set of rules and described in relation to the population at risk in terms of age, sex, race, calendar period, anatomic location, and outcome. In addition, each morphologic type has been compared with each of the others with respect to the pattern of occurrence and survival. RESULTS: For most of the individual morphologic entities, the pattern of occurrence is specific and unlike other patterns. Differences according to anatomic site, age, sex, race, and period-specific survival were found. Partly because of changes in diagnostic criteria over the years, differences in secular trend, other than that for Kaposi's sarcoma, could not be verified. Although some types of sarcoma may have important genetic determinants, there is evidence of environmental causation in others; for some varieties both genetic and environmental factors may operate. There is no evidence of improvements in survival. CONCLUSIONS: The most likely basis for the observed patterns are morphology-specific differences in etiology and growth phase. Each of the entities should be considered etiologically distinct and should be studied individually whenever possible.

Epidemiology of multiple sclerosis in US veterans. 6. Population ancestry and surname ethnicity as risk factors for multiple sclerosis.

Previously, we studied the effect of population ancestry on the risk of multiple sclerosis (MS) in US veterans of World War II, comparing by state 1980 US census ancestry data with MS case/control ratios. Here, the joint effects of population ancestry and surname-derived ethnicity on MS risk are examined in the same series. Census data are used again to characterize the population ancestry of the state from which each subject entered active duty (EAD)--that is, the proportions of the populace reporting various ancestries--and subjects were also individually categorized into a single ethnic group, without knowledge of case/control status, based on surname. In this study population, categorized ethnicity was strongly correlated with population ancestry, as expected. Although univariate analyses showed statistically significant associations between MS risk and several surname-derived ethnicities and ethnic groups, when residence at EAD was accounted for as well, there was almost no ethnic variation in MS risk. A logistic regression analysis further showed that variations in MS risk are associated most strongly with latitude and population ancestry group; in particular, subjects who entered military service from states with higher proportions of Swedish or French ancestry had higher risks of MS. After adjustment for characteristics of place, the only significant individual ethnicity factor found was Southern European ethnicity. In general, we conclude that an individual's ethnicity seems to be of less relative importance in determining MS risk than is the population ancestry of the state of EAD. These findings underscore the fact that MS is a disease of place, with 'place' including not only attributes of the locale (e.g., latitude), but also of its populace (e.g., ancestry).