Declining trends in mortality from cerebrovascular disease at ages 10-65 years: a test of validity.

Recent reports of declining trends in mortality rates from cerebrovascular disease are based on underlying cause of death as stated on the death certificate, and may contain inaccuracies because death certificates are often completed without reference to all information in medical records, and because of changes in coding conventions and diagnostic fashion. This is a report of trends in mortality rates from cerebrovascular disease in Baltimore, Md., during 1950-1970 using data validated by reference to individual medical records from 19 hospitals. Overall accuracy of death certificate diagnoses did not change markedly during the study period. Death rates from subarachnoid hemorrhage increased in white women, but decreased in white men and in nonwhites of both sexes. Death rates from other cerebrovascular diseases declined in most sex and color groups, a result which cannot be attributed to errors in death certificate diagnoses. The increasing rates of subarachnoid hemorrhage observed in white women deserve further attention.

Risk factors for ischemic stroke: a prospective study in Rochester, Minnesota.

A cohort of 1,804 residents of Rochester, Minnesota, who were at least 50 years old, free of stroke, and who underwent examination at the Mayo Clinic in 1960, was followed for 13 years. During this period, there were 110 first ischemic strokes and 616 deaths without stroke. The time of onset, if available, or the time of diagnosis of potential risk factors was determined for all patients during the study and was used to construct a proportional hazards model of time to occurrence of stroke with time-dependent risk factors. The model included 8 risk factors (2 fixed and 6 time-dependent). For these, the individual relative risks are: 1.6 for age (per 10 years), 2.0 for males, 4.0 for definite hypertension, 3.9 for transient ischemic attacks, 2.2 for hypertensive heart disease, 2.2 for coronary heart disease, 1.7 for congestive heart failure, and 1.7 for diabetes mellitus. Atrial fibrillation was not a significant risk factor using time-dependent multivariate analysis.

Inter- and intra-pathologist variability in the diagnosis of gestational trophoblastic neoplasia.

All 190 cases of gestational trophoblastic neoplasia diagnosed in the Baltimore metropolitan area from 1975-1982 were identified. Histologic slides were requested and reviewed independently by two pathologists who agreed upon uniform criteria for the diagnosis of hydatidiform (complete) mole, invasive mole, and choriocarcinoma. A representative sample of the slides was selected and resubmitted to one of the study pathologists for a second review. The inter- and intra-pathologist variability in the diagnosis of gestational trophoblastic neoplasia was calculated using the kappa statistic (K). Our findings indicated that the variability in the diagnosis of gestational trophoblastic neoplasia was low whereas that for the related tumor of incomplete mole was high.

Risk factors for gestational trophoblastic neoplasia.

A case-control study to determine the gynecologic and reproductive risk factors for gestational trophoblastic neoplasia was conducted in the Baltimore Metropolitan Area. All cases (N = 190) that were pathologically diagnosed from 1975 to 1982 as hydatidiform mole, invasive mole, or choriocarcinoma were ascertained. Slides were independently reviewed by two pathologists. Cases were matched by age, race, and last menstrual period to controls who were delivered of normal pregnancies at term. In the analysis of medical record and interview data, factors found to be positively associated with gestational trophoblastic neoplasia included professional occupations (odds ratio = 2.56, p less than 0.0001), prior spontaneous abortions (odds ratio = 2.32, p = 0.02), and the mean number of months from the last pregnancy to the index pregnancy (cases = 35.9, controls = 28.2; p = 0.03). Factors found not to be associated with disease included contraceptive history, irradiation, ABO blood group, and smoking factors of the male partner. The findings suggest that gestational trophoblastic neoplasia may be part of a continuum of early (first-trimester) reproductive abnormalities.

Trends in incidence rates of ulcerative colitis and Crohn's disease.

Between 1960 and 1979, three studies were conducted in the Baltimore Standard Metropolitan Statistical Area to ascertain the incidence rates of first hospitalizations for ulcerative colitis and Crohn's disease. The age-adjusted rates per 100,000 population for the 1977-1979 survey for ulcerative colitis in white and nonwhite males and females were 2.92, 1.79, 1.29, and 2.90, respectively; the Crohn's disease rates were 3.39, 3.54, 1.29, and 4.08, respectively. In Baltimore the age-adjusted rate for Crohn's disease has increased to exceed the ulcerative colitis rate for whites of both sexes and nonwhite females. The ulcerative colitis and Crohn's disease rates for nonwhite males are similar. The rate for white males exceeds that for nonwhite males for both ulcerative colitis and Crohn's disease, but the converse is true for females. Females have higher rates than males for Crohn's disease in both color groups and for ulcerative colitis among nonwhites. White ulcerative colitis rates are higher for males than for females. From the first to the second surveys, the white male and female rates for ulcerative colitis converge with increasing male and decreasing female rates, but then both decline from the second to the third surveys. For Crohn's disease, the age-adjusted rates increased for whites of both sexes and nonwhite females from the first to second surveys. The Crohn's disease rates appeared to stabilize for whites of both sexes between the second and present surveys, but they increased for nonwhites of both sexes. Trends in age-adjusted rates for other areas are also discussed.

Practical limitations of epidemiologic methods.

Epidemiologic methods can be categorized into demographic studies of mortality and morbidity and observational studies that are either retrospective or prospective. Some of the limitations of demographic studies are illustrated by a review of one specific mortality study showing possible relationship of nuclear fallout to leukemia. Problems of accuracy of diagnosis or causes of death on death certificates, estimates of population, migration from areas of study, and the issue of "ecological fallacy" are discussed. Retrospective studies have such problems as recall of previous environmental exposure, selection bias and survivor bias. In environmental epidemiology, prospective studies have been used. The problems associated with these studies are illustrated by reviewing some of the details of the study of effects of microwave radiation on embassy employees in Moscow. The study population had to be reconstructed, individuals had to be located and information on exposure status had to be obtained by questionnaire. The relatively small size of the exposed group permitted the detection of only fairly large relative risks. Despite these limitations, epidemiologic studies have been remarkably productive in elucidating etiological factors. They are necessary since "the proper study of man is man."

Senile macular degeneration: a case-control study.

Senile macular degeneration, although a leading cause of visual loss in the United States, remains a poorly understood disease. To assess the effects of host and environmental factors on this condition, a study of 228 cases and 237 controls matched by age and sex, who had visited any of 34 Baltimore ophthalmologists between September 1, 1978 and March 31, 1980, was conducted. Study participants were interviewed for past medical, residential, occupational, smoking and family histories, as well as social and demographic factors. Diagnoses were validated by means of fundus photographs. The 162 cases and 175 controls who met the study diagnostic criteria for cases and controls were included in the analysis. Statistically significant associations were demonstrated between senile macular degeneration and family history of macular disease (odds ratio (OR) = 2.9), chemical work exposures (OR = 4.2), blue or medium pigmented eyes (OR = 3.5), history of one or more cardiovascular diseases (OR = 1.7), decreased hand grip strength, and hyperopia. The risk of macular degeneration in cigarette smokers was significant for males only (OR = 2.6). The study results suggest that the development of macular degeneration is mainly influenced by familial, genetic, and personal characteristics, rather than by the few environmental factors studied. Additional studies are needed to further evaluate the role of environmental factors.

The surgeon general's "epidemiologic criteria for causality": a criticism of Burch's critique.

In this paper, it has not been possible to critically and objectively assess each of the points made by Burch, since it would be time-consuming both for this author as well as for the reader. However, the major areas of concern expressed by Burch have been reviewed and evaluated. This has in many instances required going back to the original reports of studies to which Burch refers. In doing so, it came as a complete surprise to note that Burch has incompletely presented the findings of several studies or has omitted discussing others more fully. This is, needless to say, most disturbing. In viewing the issue of the causal interpretation of the cigarette smoking--lung cancer relationship, it would appear that Burch would not be satisfied unless there is a randomized controlled experiment in humans. However, since such an experiment seems hardly feasible, one must depend upon a synthesis of findings of epidemiologic observations, experimental work on animals and studies of chemical constituents of tobacco and cigarette smoke. However, these do not seem to satisfy Burch, who has developed a theoretical conceptual framework for the etiology of cancer and many other diseases, based upon statistical models and equations. Such an approach is extremely limited. Scientific information and inferences are in the public domain, thereby allowing criticism and evaluation. It is customary, in the experimental sciences, for a scientist who is dissatisfied with the findings of another scientist, to attempt to repeat the experiment to see if his findings are the same or different. This should also be expected in the cigarette smoking-lung cancer issue.(ABSTRACT TRUNCATED AT 250 WORDS)

Patterns of incidence and trends in diagnostic classification of cerebrovascular disease in Washington County, Maryland, 1969-1971 to 1974-1976.

A study was conducted to provide baseline data on patterns of cerebrovascular disease incidence in the community of Washington County, Maryland, over two time intervals--1969-1971 and 1974-1976--and to evaluate any temporal trends with regard to their possible relation to reported declines in stroke death rates. Among the statistically significant changes in incidence between 1969-1971 and 1974-1976 were a decline in the rate of cerebral thrombosis and a concomitant increase in frequency of cerebrovascular disease-unspecified, most likely a reflection of a change in diagnostic terminology. In addition, the findings of a significant decrease in incidence of cerebral hemorrhage (for all ages over 55 years and both sexes) and a non-significant decrease in incidence rates for all types combined among those in the oldest age category between the two study periods support the role of changing incidence rates in the stroke death rate decline. It is clear, however, that monitoring of this community over a longer period of time will be necessary before the validity of these trends can be established and their causes elucidated. The need for more complete recording of historical data and clinical findings on admission, as well as adherence to standardized diagnostic and classification protocols was underscored by this study.

Some aspects of cancer epidemiology.

Epidemiolgic studies have strongly suggested that a vast majority (80-90%) of cancers are caused by radiation, chemical and biologic agents; the remainder result from endogenous or genetic factors. Biologically, cancer is most probably the end result of a complex multistage process and therefore may be due to a sequence of exposures to different agents at each of these stages. This emphasizes the need to stress the study of interactions in epidemiologic studies to a greater extent than has been done thus far. Examples of the importance of interactions in several types of cancer are presented.

Incidence rates of ulcerative colitis and Crohn's disease in fifteen areas of the United States.

A study was carried out during 1973 to determine the incidence of first hospitalizations for ulcerative colitis and Crohn's disease in 15 areas of the United States, including communities of widely varied size, climatic, ethnic, racial, and socioeconomic characteristics. The following descriptions apply to incidence rates per 100,000 population for the aggregate of the 15 areas. Ulcerative colitis had a bimodal age distribution in white males (with peaks at ages 20-29 and 70-79 yr) and females (with peaks at ages 30-39 and 70-79 yr). Crohn's disease had a bi- or trimodal age distribution in white males (with peaks at ages 20-29, 50-59, and 70-79 yr) and females (with peaks at ages 20-29, 50-59, and 70-79 yr). The age, sex, and geographic distributions that were observed in this study may have important etiologic implications.

Patterns of mortality in the the Old Order Amish. I. Background and major causes of death.

The major causes of death were studied in the Old Order Amish people in three settlements in Indiana, Ohio and Pennsylvania to determine if lifestyle and genetic isolation altered their mortality risk compared to neighboring non-Amish. The Amish are a conservative religious group who live in farm settlements, use horses for work and travel, exercise vigorously, and avoid cigarettes and alcohol. They are reproductively isolated and highly inbred. Death certificates and Amish censuses were used to determine mortality risks, which were summarized using age-adjusted mortality ratio (MRs). Amish mortality patterns were not systematically higher or lower than those of the non-Amish, but differed by age, sex, and cause. Amish males had slightly higher all-cause MRs as children and significantly lower MRs over the age of 40, due primarily to lower rates of cancer (MR = 0.44, age 40-69), and cardiovascular diseases (MR = 0.65, age 40-69). Amish females MRs for all causes of death were lower from age 10-39, not different from 40-69, and higher over age 69. MRs were not significantly different for all cancer sites combined in Amish women and they had higher cardiovascular mortality ratio aged 70 and over (MR =1.34). Other major causes of death were also examined. Because the Amish and other farming groups have similar mortality patterns, it is suggested that lifestyle may be the primary determinant of the overall mortality patterns in the Amish.