Showing papers in "American Journal of Epidemiology in 1995"

Reliability of the Youth Risk Behavior Survey Questionnaire

Abstract: The Centers for Disease Control and Prevention's Youth Risk Behavior Survey (YRBS) has been used on a biennial basis since 1990 to measure health risk behaviors of high school students nationwide. The YRBS measures behaviors related to intentional and unintentional injury, tobacco use, alcohol and other drug use, sexual activity, diet, and physical activity. The authors present the results from a test-retest reliability study of the YRBS, conducted by administering the YRBS questionnaire to 1,679 students in grades 7 through 12 on two occasions 14 days apart. The authors computed a kappa statistic for each of 53 self-report items and compared group prevalence estimates across the two testing occasions. Kappas ranged from 14.5% to 91.1%; 71.7% of the items were rated as having "substantial" or higher reliability (kappa = 61-100%). No significant differences were found between the prevalence estimates at time 1 and time 2. Responses of seventh grade students were less consistent than those of students in higher grades, indicating that the YRBS is best suited for students in grade 8 and above. Except for a few suspect items, students appeared to report personal health risk behaviors reliably over time. Reliability and validity issues in health behavior assessment also are discussed.

A Critical Look at Methods for Handling Missing Covariates in Epidemiologic Regression Analyses

Abstract: Epidemiologic studies often encounter missing covariate values. While simple methods such as stratification on missing-data status, conditional-mean imputation, and complete-subject analysis are commonly employed for handling this problem, several studies have shown that these methods can be biased under reasonable circumstances. The authors review these results in the context of logistic regression and present simulation experiments showing the limitations of the methods. The method based on missing-data indicators can exhibit severe bias even when the data are missing completely at random, and regression (conditional-mean) imputation can be inordinately sensitive to model misspecification. Even complete-subject analysis can outperform these methods. More sophisticated methods, such as maximum likelihood, multiple imputation, and weighted estimating equations, have been given extensive attention in the statistics literature. While these methods are superior to simple methods, they are not commonly used in epidemiology, no doubt due to their complexity and the lack of packaged software to apply these methods. The authors contrast the results of multiple imputation to simple methods in the analysis of a case-control study of endometrial cancer, and they find a meaningful difference in results for age at menarche. In general, the authors recommend that epidemiologists avoid using the missing-indicator method and use more sophisticated methods whenever a large proportion of data are missing.

Body Size and Fat Distribution as Predictors of Coronary Heart Disease among Middle-aged and Older US Men

Abstract: Obesity, android fat distribution, and other anthropometric measures have been associated with coronary heart disease in long-term prospective studies. However, fluctuations in weight due to age-related hormonal changes and changes in lifestyle practices may bias relative risk estimates over a long follow-up period. The authors prospectively studied the association between body mass index (BMI) (kg/m2), waist-to-hip ratio, and height as independent predictors of incident coronary heart disease in a 3-year prospective study among 29,122 US men aged 40-75 years in 1986. The authors documented 420 incident coronary events during the follow-up period. Body mass index, waist-to-hip ratio, short stature, and weight gain since age 21 were associated with an increased risk of coronary heart disease. Among men younger than 65, after adjusting for other coronary risk factors, the relative risk was 1.72 (95% confidence interval (CI) 1.10-2.69) for men with BMI of 25-28.9, 2.61 (95% CI 1.54-4.42) for BMI of 29.0-32.9, and 3.44 (95% CI 1.67-7.09) for obese men with BMI > or = 33 compared with lean men with BMI or = 65 years of age, the association between BMI and risk of coronary heart disease was much weaker. However, in this age group, the waist-to-hip ratio was a much stronger predictor of risk (relative risk = 2.76, 95% CI 1.22-6.23 between extreme quintiles). These results suggest that for younger men, obesity, independent of fat distribution, is a strong risk factor for coronary heart disease. For older men, measures of fat distribution may be better than body mass index at predicting risk of coronary disease.

Meta-analysis of Pap test accuracy

Abstract: A literature search identified 62 studies published by August 1992 comparing Papanicolaou (Pap) test results with histology. Critical appraisal revealed that 82% of these had potential for verification bias and that only 37% stated that cytology and histology were independently assessed. Estimates of sensitivity and specificity ranged from 11 to 99% and 14 to 97%, respectively, and were highly negatively correlated (r = -0.63). Meta-analysis was used to combine data from 59 studies to estimate the accuracy of the Pap test using a summary receiver operating characteristic curve and to examine the effect of study quality. The summary receiver operating characteristic curve suggests that the Pap test may be unable to achieve concurrently high sensitivity and specificity. For example, specificity in the 90-95% range corresponds to sensitivity in the 20-35% range. Pap test accuracy was not associated with reported study characteristics or dimensions of quality. Future primary studies should pay more attention to methodologic standards for the conduct and reporting of diagnostic test evaluations.

Bayesian Estimation of Disease Prevalence and the Parameters of Diagnostic Tests in the Absence of a Gold Standard

Abstract: It is common in population screening surveys or in the investigation of new diagnostic tests to have results from one or more tests investigating the same condition or disease, none of which can be considered a gold standard. For example, two methods often used in population-based surveys for estimating the prevalence of a parasitic or other infection are stool examinations and serologic testing. However, it is known that results from stool examinations generally underestimate the prevalence, while serology generally results in overestimation. Using a Bayesian approach, simultaneous inferences about the population prevalence and the sensitivity, specificity, and positive and negative predictive values of each diagnostic test are possible. The methods presented here can be applied to each test separately or to two or more tests combined. Marginal posterior densities of all parameters are estimated using the Gibbs sampler. The techniques are applied to the estimation of the prevalence of Strongyloides infection and to the investigation of the diagnostic test properties of stool examinations and serologic testing, using data from a survey of all Cambodian refugees who arrived in Montreal, Canada, during an 8-month period.

Prospective Study of Intentional Weight Loss and Mortality in Never-Smoking Overweight US White Women Aged 40–64 Years

Abstract: Although 40% of US women indicate they are currently trying to lose weight, the association between intentional weight loss and longevity is unknown. The authors analyzed prospective data from 43,457 overweight, never-smoking US white women aged 40-64 years who in 1959-1960 completed a questionnaire that included questions on weight change direction, amount, time interval, and intentionality. Vital status was determined in 1972. Proportional hazards regression was used to estimate mortality rate ratios for women who intentionally lost weight compared with women who had no change in weight. Women who died within the first 3 years of follow-up were excluded. Analyses were stratified by preexisting illness and adjusted for age, beginning body mass index, alcohol intake, education, physical activity, and health conditions. In women with obesity-related health conditions (n = 15,069), intentional weight loss of any amount was associated with a 20% reduction in all-cause mortality, primarily due to a 40-50% reduction in mortality from obesity-related cancers; diabetes-associated mortality was also reduced by 30-40% in those who intentionally lost weight. In women with no preexisting illness (n = 28,388), intentional weight loss of > or = 20 lb (> or = 9.1 kg) that occurred within the previous year was associated with about a 25% reduction in all-cause, cardiovascular, and cancer mortality; however, loss of or = 1 year was generally associated with small to modest increases in mortality. The association between intentional weight loss and longevity in middle-aged overweight women appears to depend on their health status. Intentional weight loss among women with obesity-related conditions is generally associated with decreased premature mortality, whereas among women with no preexisting illness, the association is equivocal.

Air Pollution and Hospital Admissions for Cardiovascular Disease in Detroit, Michigan

Abstract: In the December 1952 smog disaster in London, a substantial increase in mortality was closely associated with the increase in air pollution. Deaths from cardiovascular causes were elevated as well as respiratory deaths. The increase was greatest in the elderly. Hospital admissions were increased for both respiratory and cardiovascular diseases. Since then, many studies have reported associations between lower concentrations of air pollution and daily mortality. Little attention has been paid to the question of hospital admissions for cardiovascular illness, however. This study examined the association between air pollution and cardiovascular hospital admissions for persons aged 65 years and older in the Detroit, Michigan, metropolitan area during the years 1986-1989. After controlling for seasonal and other long-term temporal trends, temperature, and dew point temperature, the particulate matter with an aerodiameter of < or = 10 microns (PM10) was associated with daily admissions for ischemic heart disease (relative risk (RR) = 1.018, 95% confidence interval (CI) 1.005-1.032 for an interquartile range (32 micrograms/m3) increase in pollution). SO2, CO, and ozone made no independent contribution to ischemic heart disease admissions. Both PM10 (RR = 1.024, 95% CI 1.004-1.044) and CO (RR = 1.022, 95% CI 1.010-1.034 for an interquartile range (1.28 ppm) increase in pollution) showed independent associations with heart failure admissions. These results were robust to alternate methods of estimation and weather control.

Relation of Serum Uric Acid to Mortality and Ischemic Heart Disease The NHANES I Epidemiologic Follow-up Study

Abstract: Although hyperuricemia is frequently found among persons with ischemic heart disease, its importance as a risk factor remains uncertain. The authors examined this relation among 5,421 persons in the First National Health and Nutrition Examination Survey (NHANES I) Epidemiologic Follow-up Study; baseline data were collected in 1971-1975 and follow-up was through 1987. No associations were seen among men, but, among women, the serum uric acid level was predictive of mortality from all causes and from ischemic heart disease. These associations persisted even after excluding the first 10 years of follow-up and were independent of use of antihypertensive agents and diuretics, diastolic blood pressure, overweight, and other characteristics. A dose-response relation was evident for mortality from ischemic heart disease: each 1-mg/dl change in uric acid (about two thirds of the standard deviation) among women increased the rate by 1.48 (95% confidence interval 1.3-1.7). Furthermore, as compared with women who had a uric acid level <4 mg/dl, those with a level ≥7 mg/dl had a 4.8-fold (95% confidence interval 1.9-12) higher rate of ischemic heart disease mortality. In contrast, the uric acid level showed a weaker relation with disease incidence among women, with a rate ratio of 1.14 for each 1-mg/dl change. Although the biologic mechanism is unclear, further investigation into the possible role of uric acid in the development of ischemic heart disease is needed

Reproducibility and Validity of an Epidemiologic Questionnaire to Assess Past Year Physical Activity in Adolescents

Abstract: The reproducibility and validity of a past year physical activity questionnaire was determined in a sample of 100 adolescents aged 15-18 years, randomly selected from a population-based cohort. Subjects completed four 7-day recalls of activity approximately 3 months apart. The average of the four 7-day recalls of activity was utilized as the "gold standard" against which the past year questionnaire was compared to evaluate validity. The questionnaire was also validated against objective measures, such as physical fitness and body mass index. Interscholastic team rosters were utilized to directly validate the reporting of specific activities. One-month and one-year test-retest reproducibility of the questionnaire were determined. For different measures of activity, the Spearman correlations between the questionnaire and the average of the 7-day recalls ranged from 0.55 to 0.67 in males and 0.73 to 0.83 in females, all significant at p < 0.01. In general, although there was no association between the past year activity questionnaire results and objective measures, there was a significant, albeit weak association between the physical activity questionnaire and time to complete a 1-mile (1.61-km) run (r = -0.47) in females. Subjects reported participating in specific interscholastic sports with an accuracy of 100%, 86%, and 95% for the fall, winter, and spring sports, respectively. Test-retest reproducibility was higher over one month (r = 0.79) than over one year (r = 0.66). These data provide evidence that the questionnaire yields a reasonable estimate of past year or "habitual" physical activity in adolescents.

Public Drinking Water Contamination and Birth Outcomes

Abstract: The effects of public drinking water contamination on birth outcomes were evaluated in an area of northern New Jersey. After excluding plural births and chromosomal defects, 80,938 live births and 594 fetal deaths that occurred during the period 1985-1988 were studied. Information on birth outcome status and maternal risk factors was obtained from vital records and the New Jersey Birth Defects Registry. Monthly exposures during pregnancy were estimated for all births using tap water sample data. Odds ratios of > or = 1.50 were found for the following: total trihalomethanes with small for gestational age, central nervous system defects, oral cleft defects, and major cardiac defects; carbon tetrachloride with term low birth weight, small for gestational age, very low birth weight, total surveillance birth defects, central nervous system defects, neural tube defects, and oral cleft defects; trichloroethylene with central nervous system defects, neural tube defects, and oral cleft defects; tetrachloroethylene with oral cleft defects; total dichloroethylenes with central nervous system defects and oral cleft defects; benzene with neural tube defects and major cardiac defects; and 1,2-dichloroethane with major cardiac defects. Total trihalomethane levels > 100 ppb reduced birth weight among term births by 70.4 g. By itself, this study cannot resolve whether the drinking water contaminants caused the adverse birth outcomes; therefore, these findings should be followed up utilizing available drinking water contamination databases.

The Frequency of Idiopathic Parkinson's Disease by Age, Ethnic Group, and Sex in Northern Manhattan, 1988–1993

Abstract: Sex and ethnic differences in the frequency of Parkinson's disease have become increasingly important, because putative genetic and environmental risk factors have been identified. The authors estimated the prevalence and incidence of Parkinson's disease in a culturally diverse community in New York City over a 4-year period (January 1, 1988-December 31, 1991) using a disease registry substantiated, for older individuals, by a subsequent survey of a random sample of Medicare recipients between January 1, 1992, and December 31, 1993. The prevalence rate was 107 per 100,000 persons, and over a 3-year period the average incidence rate was 13 per 100,000 person-years. Age-adjusted prevalence rates were lower for women than for men in each ethnic group and were lower for blacks than for whites and Hispanics. Incidence rates were highest among black men, but they were otherwise comparable across the sex and ethnic groups. The estimated cumulative incidence of Parkinson's disease up to age 90 years was lower for women than for men, which could partially explain the lower prevalence rate. By ethnic group, the cumulative incidence was higher for blacks than for whites and Hispanics, but more deaths occurred among incident black cases. Discrepant prevalence and incidence rates of Parkinson's disease among blacks and women warrant further investigation. While selective mortality could partially account for this paradox, it is also possible that a delay in diagnosis due to limited access to appropriate health services among these individuals could have resulted in the observed discordant rates of disease.

Gamma-glutamyltransferase: Determinants and Association with Mortality from Ischemic Heart Disease and All Causes

Abstract: The association of serum levels of gamma-glutamyltransferase (GGT) with cardiovascular disease risk factors, and with mortality from all causes, cardiovascular disease, and non-cardiovascular diseases, has been examined in a prospective study of 7,613 middle-aged British men followed for 11.5 years. GGT levels were strongly associated with all-cause mortality, largely due to a significant increase in deaths from ischemic heart disease and other non-cardiovascular disease causes, i.e., non-cancer deaths, in the top quintile of the GGT distribution. No association was seen with cancer mortality. However, GGT was significantly (positively) associated with alcohol intake, body mass index, smoking, preexisting ischemic heart disease, diabetes mellitus, antihypertensive medication, systolic and diastolic blood pressure, total and high density lipoprotein cholesterol, heart rate, and blood glucose, and negatively associated with physical activity and lung function (forced expiratory volume in 1 second (FEV1)). After adjustment for these personal characteristics and biologic variables, elevated GGT (highest quintile > or = 24 unit/liter vs. the rest) was still associated with a significant increase in mortality from all causes (relative risk (RR) = 1.22, 95% confidence interval (CI) 1.01-1.42; n = 818 deaths) and from ischemic heart disease (RR = 1.42, 95% CI 1.12-1.80; n = 332 deaths). The increase in other non-cardiovascular disease causes was of marginal significance (RR = 1.45, 95% CI 0.95-2.20; n = 127 deaths). When examined separately by the presence or absence of preexisting ischemic heart disease, the increased risk of ischemic heart disease mortality was more marked in those with evidence of ischemic heart disease at screening, particularly in those with previous myocardial infarction (RR = 1.67, 95% CI 1.03-2.69; n = 84 deaths). The increased risk of other non-cardiovascular disease deaths was only seen in men without preexisting ischemic heart disease, largely due to an excess of hepatic cirrhosis. In summary, many factors other than alcohol intake are associated with increased levels of GGT, in particular body mass index, diabetes mellitus, and serum total cholesterol. The finding of increased risk of ischemic heart disease mortality seen in men with preexisting ischemic heart disease is related to the severity of the underlying myocardial damage. The biologic significance of raised GGT in men with preexisting ischemic heart disease merits further study.

Recall accuracy for prescription medications: self-report compared with database information

Abstract: A methodological study was performed in 1992 to evaluate the accuracy of self-reported use of nonsteroidal antiinflammatory drugs (NSAIDs) and noncontraceptive estrogens that had been dispensed during the previous 12 years. A sample of 560 individuals dispensed NSAIDs or estrogens, and 140 individuals without NSAID/estrogen dispensations were selected from the Group Health Cooperative pharmacy database. Demographic, behavioral, and drug information was ascertained by telephone interview for 356 persons with and 98 persons without NSAID/estrogen dispensations. Of those with only a single NSAID dispensation, 41% (95% confidence interval (CI) 32-50%) were able to recall any NSAID use compared with 85% (95% CI 76-94%) for those with multiple NSAID dispensations. Thirty percent (95% CI 24-36%) recalled the NSAID name, and 15% (95% CI 10-20%) recalled both the name and dose. For estrogens, 78% (95% CI 70-86%) recalled the name, but only 26% (95% CI 17-34%) recalled the name and dose. Age, but not sex, appeared to influence recall accuracy: Persons 50-65 years of age recalled the NSAID name more accurately than those aged 66-80 (odds ratio (OR) = 1.8, 95% confidence interval (CI) 1.0-3.4). A similar advantage was noted for 50- to 65-year-old women in recalling the estrogen name (OR = 1.5, 95% CI 0.6-3.9). Drug name was recalled more frequently for exposures stopped 2-3 years prior to interview than for those stopped 7-11 years prior (OR = 3.0, 95% CI 1.6-5.7, and OR = 2.4, 95% CI 0.9-6.7, for NSAIDs and estrogens, respectively). Specificity was consistently high, ranging from 92% to 100%. This study suggests significant underascertainment of self-reported prescription drug exposure but little evidence that exposures are overreported.

Age-related Macular Degeneration Is Associated with Atherosclerosis The Rotterdam Study

Abstract: Age-related macular degeneration is the most frequent cause of blindness in the elderly. A vascular basis of the disease has been suggested, but not confirmed. The association between atherosclerosis and this type of macular degeneration was investigated in 104 subjects with and 1,324 subjects without macular degeneration as part of the population-based Rotterdam Study. The study was performed between March 1990 and July 1993 in a suburb of Rotterdam, the Netherlands. Macular degeneration was assessed on fundus photographs. Carotid atherosclerosis was ultrasonographically evaluated by measurement of the common carotid intima-media thickness and by assessment of the presence of atherosclerotic plaques. Atherosclerosis in arteries of the lower extremities was studied by determination of the ankle-arm systolic blood pressure ratio. In subjects younger than age 85 years, plaques in the carotid bifurcation were associated with a 4.7 times increased prevalence odds of macular degeneration (95% confidence interval (CI) 1.8-12.2); those with plaques in the common carotid artery showed an increased prevalence odds of 2.5 (95% CI 1.4-4.5). The intima-media thickness of the common carotid arteries was not significantly different. Lower extremity arterial disease (ankle-arm index less than 0.90 on at least one side) was associated with a 2.5 times increased prevalence odds (95% CI 1.4-4.5). These findings suggest that atherosclerosis may be involved in the etiology of age-related macular degeneration.

Irritable Bowel Syndrome in a Community: Symptom Subgroups, Risk Factors, and Health Care Utilization

Abstract: The clinical relevance of subdividing the irritable bowel syndrome (IBS) into subgroups based on bowel habit is largely unknown. We therefore obtained an age- and sex-stratified random sample of Olmsted County, Minnesota, residents aged 20-95 years. All subjects were mailed a valid self-report questionnaire during the years 1988-1993; the response rate was 74% (n = 3,022). Among subjects with IBS (n = 536), four symptom-based subgroups of similar size were identified: constipation predominant, diarrhea predominant, alternating constipation and diarrhea, and neither. The prevalence of IBS was significantly greater in females, primarily because of a higher prevalence of constipation-predominant IBS in women. Of persons > or = 60 years of age, 23% reported the initial onset of IBS in the previous year compared with 10% in younger subjects; the age at onset of IBS was similar among the subgroups. Marital status, education level, smoking, and alcohol use were not significantly different among the subgroups. Of those with IBS, 25% reported visiting a physician for abdominal pain or disturbed defecation in the prior year compared with only 8% of persons without IBS. Female sex, an increased number of Manning's symptom criteria, and the individual IBS subgroups were not associated with higher rates of physician visits. We conclude that the onset of IBS may not be limited to early adulthood and that subgroups of IBS based on bowel patterns may not identify clinically distinct entities.

Associations between Ambient Particulate Sulfate and Admissions to Ontario Hospitals for Cardiac and Respiratory Diseases

Abstract: The association of daily cardiac and respiratory admissions to 168 acute care hospitals in Ontario, Canada, with daily levels of particulate sulfates was examined over the 6-year period 1983-1988. Sulfate levels were recorded at nine monitoring stations in regions of southern and central Ontario spanned by three monitoring networks. A 13-micrograms/m3 increase in sulfates recorded on the day prior to admission (the 95th percentile) was associated with a 3.7% (p < 0.0001) increase in respiratory admissions and a 2.8% (p < 0.0001) increase in cardiac admissions. Increases were observed for all age groups examined. Admissions for cardiac diseases increased 2.5% for those under 65 years and 3.5% for those 65 years and older. After adjusting for ambient temperature and ozone, similar increases in respiratory admissions were observed in the period from April to September (3.2%) and in the period from October to March (2.8%). A 3.2% increase was observed for cardiac admissions in the period from April to September, and a 3.4% increase was observed in the period from October to March after adjusting for ambient temperature and ozone.

Control Sampling Strategies for Case-Crossover Studies: An Assessment of Relative Efficiency

Abstract: The case-crossover study design is a method to assess the effect of transient exposures on the risk of onset of acute events. Control information for each case is based on his/her past exposure experience, and a self-matched analysis is conducted. Empiric evaluation of five approaches to the analysis of case-crossover data from a study of heavy physical exertion and acute myocardial infarction onset is shown. The data presented are from the Onset Study, a case-crossover study of the determinants of myocardial infarction onset conducted in 45 centers from August 1989 to October 1992. In model 1, exactly one control period (matched on clock-time) was sampled per case. In models 2-4, up to 25 control periods were sampled, and the effect of clock-time on the baseline hazard of infarction was modeled. In model 5, a census of the person-time experienced by each subject over the year preceding the infarction was sampled. The 95% confidence interval for model 1 was 2.7 times wider, and the relative efficiency, defined as v infinity/vM, where vM represents the asymptotic variance estimate of the estimated log relative risk with M control periods sampled per case, was only about 14% of model 5. In models 2-4, the efficiency increased with the number of control periods, regardless of the modeling assumptions. Even with many control periods sampled, models 2-4 achieved only half the efficiency of model 5. The control sampling strategy in any given case-crossover study should be selected with the trade-offs between precision and potential biases of the estimates in mind.

Aneurysms of the Abdominal Aorta in Older Adults The Rotterdam Study

Abstract: To assess the age- and sex-specific prevalence and risk factors for aneurysms of the abdominal aorta, the authors performed a population-based study in 5,419 subjects (42% men, 58% women) aged 55 years and over. The proximal and distal diameter of the abdominal aorta were measured by ultrasound. An aneurysm was defined as a distal aortic diameter of 35 mm or more or a dilatation of the distal part of the the abdominal aorta of 50% or more. The mean distal and proximal aortic diameter increased 0.7 mm and 0.3 mm, respectively, with every 10 years of age. In 2.1% (95% confidence interval (CI) 1.7-2.5) of the study population, an aneurysm was present, or in 4.1% (95% CI 3.2-4.9) of the men and 0.7% (95% CI 0.4-1.0) of the women. Subjects with an abdominal aneurysm were more likely to be smokers and they had higher serum cholesterol levels and higher prevalence of cardiovascular disease compared with subjects without an aneurysm. The authors conclude that the ultrasound diameter of the abdominal aorta clearly increases with age in both men and women and that the prevalence of aneurysms of the abdominal aorta in older adults in relatively high, especially in men.

Nonsteroidal Anti-inflammatory Drugs and the Incidence of Hospitalizations for Peptic Ulcer Disease in Elderly Persons

Abstract: To determine the incidence rate of serious ulcer disease among users and nonusers of nonsteroidal anti-inflammatory drugs (NSAIDs), a retrospective cohort study was done on 103,954 elderly Tennessee Medicaid recipients with 209,068 person-years of follow-up from 1984 to 1986. There were 1,371 patients hospitalized with peptic ulcer disease or upper gastrointestinal hemorrhage identified by Medicaid hospital claims and verified by review of the medical record. Ulcer hospitalization rates by NSAID exposure category, duration of use, and daily dose were determined. The rates of ulcer hospitalization among nonusers and current users of NSAIDs were 4.2 and 16.7 per 1,000 person-years, respectively, an excess rate among current users of 12.5 (95% confidence interval (CI) 11.4-13.6) per 1,000 person-years. Among new users, the ulcer hospitalization rates were 26.3 per 1,000 person-years during the first 30 days of use and 20.9 per 1,000 person-years over the next 31-180 days, representing excess ulcer hospitalization rates of 22.1 (95% CI 18.6-25.6) and 16.7 (95% CI 13.1-20.1) per 1,000 person-years, respectively. For long-term users (180 days or more of continuous NSAID use), the ulcer hospitalization rate remained elevated at 15.3, an excess of 12.0 (95% CI 10.3-13.6) hospitalizations per 1,000 person-years. The excess hospitalization rates per 1,000 person-years increased with increasing dose from 6.0 (95% CI 4.0-8.0) for the lowest dose category to 17.8 (95% CI 15.5-20.1) for the highest. The excess rate of ulcer hospitalization for elderly NSAID users is high. These drugs should be used with caution in elderly persons, and alternatives to NSAID therapy should be strongly considered.

Cardiovascular Disease Risk Factors among American Indians The Strong Heart Study

Abstract: The Strong Heart Study, a study of cardiovascular disease among American Indians, was conducted to determine cardiovascular disease rates and the prevalence of risk factors among members of 13 tribal groups in South Dakota/North Dakota (SD/ND), southeastern Oklahoma, and Arizona. From 1989 to 1992, 4,549 tribal members aged 45-74 years (62% of eligible participants) were surveyed and examined for cardiovascular disease and its risk factors. Mean total cholesterol concentrations were over 20 mg/dl lower among the men and 27 mg/dl lower among the women than national mean levels for the same age groups. Cholesterol levels varied by tribal group; Arizona Indians had mean levels more than 20 mg/dl lower than those of SD/ND Indians. The prevalence of hypercholesterolemia was almost twice as high among SD/ND Indians as among Arizona Indians, but the rates for all three groups were much lower than total US rates (all races). Mean levels of high density lipoprotein cholesterol were lower among Indian men and women than in the US population as a whole. The prevalence of hypertension among Arizona and Oklahoma Indians was higher than that for the entire United States. SD/ND Indians had significantly lower mean blood pressures and prevalence rates of hypertension than Oklahoma and Arizona Indians and the United States as a whole. The prevalence of cigarette smoking was higher for all Indian groups except Arizona women in comparison with US rates. Smoking rates were highest in SD/ND and lowest in Arizona. Indian smokers smoked fewer cigarettes per day than the average US smoker. Arizona Indians had the highest prevalence of diabetes mellitus; over 60% of those participants were diabetic. In Oklahoma and SD/ND, one third of the men and over 40% of the women were diabetic. In addition, 13-20% of the participants had impaired glucose tolerance. Proteinuria was also a common problem; almost half of the Arizona Indians had micro- or macroalbuminuria, and 20% of Oklahoma and SD/ND Indians had significant proteinuria. The prevalence of obesity was high in all three groups, with Arizona Indians having the highest rates and the highest mean body mass indices. The prevalence of current alcohol use was lower among Indians than in the nation as a whole, but binge drinking was common among those who used alcohol. These results indicate that cardiovascular disease risk factors vary significantly among tribal groups. Prevention programs tailored toward decreasing the prevalence of risk factors are recommended for long-term reduction of cardiovascular disease rates in American Indian communities.

Ingested Arsenic and Internal Cancer: A Historical Cohort Study Followed for 33 Years

Abstract: A historical cohort study was conducted to investigate the long-term effect of exposure to ingested arsenic. The 454 residents who had been identified in a list made in 1959 were followed until 1992. They lived in an arsenic-polluted area, called Namiki-cho, Nakajo-machi, in Niigata Prefecture, Japan, and used well water containing inorganic arsenic. The exposure period was estimated to be about 5 years (1955-1959). Death certificates for the people who died between 1959 and 1992 were examined, and a total of 113 of the 454 residents were estimated to have drunk well water containing a high dose of arsenic (> or = 1 ppm). The standardized mortality rate ratios of these 113 residents were 15.69 for lung cancer (observed/expected = 8/0.51; 95% confidence interval (CI) 7.38-31.02) and 31.18 for urinary tract cancer (observed/expected = 3/0.10; 95% CI 8.62-91.75). Cox's proportional hazard analyses demonstrated that the hazard ratios of the highest exposure level group (> or = 1 ppm) versus the background exposure level group (0.001 ppm) were 1.74 (95% CI 1.10-2.74) for all deaths and 4.82 (95% CI 2.09-11.14) for all cancers. The analysis according to the skin signs of chronic arsenicism in 1959 showed that they were useful risk indicators for subsequent cancer development. In the development of lung cancer, there was evidence of synergism between arsenic intake and smoking habit.

Family History and Prostate Cancer Risk in Black, White, and Asian Men in the United States and Canada

Abstract: Increased risk of prostate cancer in men with a family history of the disease has been observed consistently in epidemiologic studies. However, most studies have been confined to white men; little is known about familial aggregation of prostate cancer in populations with unusually high incidence, such as African Americans, or in populations with low incidence, such as Asian-Americans. The authors report results from a population-based case-control study of prostate cancer among blacks, whites, and Asian-Americans in the United States and Canada. Controls were matched to cases on age (5-year groups), ethnicity (black, white, Chinese-American, Japanese-American), and region of residence (Los Angeles, San Francisco, Hawaii, Vancouver, Toronto). In the combined group of participants, 5% of controls and 13% of cases reported a father, brother, or son with prostate cancer. These prevalences were somewhat lower among Asian-Americans than among blacks or whites. A positive family history was associated with a statistically significant two- to threefold increase in risk in each of the three ethnic groups. The overall odds ratio associated with such a family history, adjusted for age and ethnicity, was 2.5 (95% confidence interval 1.9-3.3). This odds ratio varied by neither ethnicity nor age of the participants. Sera from 1,087 controls were used to examine the relations between family history and serum concentrations of androgens and prostate-specific antigen. The concentrations of sex hormone-binding globulin were slightly higher in men with than without a positive family history. Prostate-specific antigen concentrations were unrelated to family history.

Magnetic Field Exposure in Relation to Leukemia and Brain Cancer Mortality among Electric Utility Workers

Abstract: Reports of leukemia and brain cancer among men in electrical occupations suggest a small increase in risk, but most previous studies have failed to classify magnetic field exposure accurately or to consider potential confounders. The authors conducted an historical cohort mortality study of 138,905 men employed at five large electric power companies in the United States between 1950 and 1986 with at least 6 months of work experience. Exposure was estimated by linking individual work histories to data from 2,842 workshift magnetic field measurements. Mortality follow-up identified 20,733 deaths based on 2,656,436 person-years of experience. Death rates were analyzed in relation to magnetic field exposure history with Poisson regression. Total mortality and cancer mortality rose slightly with increasing magnetic field exposure. Leukemia mortality, however, was not associated with indices of magnetic field exposure except for work as an electrician. Brain cancer mortality was modestly elevated in relation to duration of work in exposed jobs and much more strongly associated with magnetic field exposure indices. Brain cancer risk increased by an estimated factor of 1.94 per microtesla-year of magnetic field exposure in the previous 2-10 years, with a mortality rate ratio of 2.6 in the highest exposure category. In contrast to other studies, these data do not support an association between occupational magnetic field exposure and leukemia but do suggest a link to brain cancer.

Fish Consumption and Cardiovascular Disease in the Physicians' Health Study: A Prospective Study

Abstract: The authors examined the association between dietary intake of fish and omega 3 fatty acids from seafood and the risk of cardiovascular disease in a prospective cohort study of 21,185 US male physicians who are participants in the Physicians' Health Study. In 4 years of follow-up, there were 281 incident cases of total (fatal and nonfatal) myocardial infarction, 173 cases of stroke, and 121 cardiovascular deaths. There was no evidence for association between dietary intake of fish and any cardiovascular endpoint, including myocardial infarction, stroke, and cardiovascular death. The relative risks of total myocardial infarction, adjusted for age and randomized treatment assignment, for categories of fish intake were: 1.0 for or = 5 fish meals/week; chi 2 for trend = 0.9, p = 0.34. The relative risks were similar for omega 3 fatty acid intake and for specific types of fish, and did not change after adjustment for history of hypertension, hypercholesterolemia, diabetes mellitus, or angina pectoris, parental history of myocardial infarction before age 60 years, obesity, exercise, smoking, alcohol use, saturated fat intake, and vitamin supplement use. These data do not support the hypothesis that moderate fish consumption lowers the risk of cardiovascular disease.

Hostility, Incidence of Acute Myocardial Infarction, and Mortality in a Sample of Older Danish Men and Women

Abstract: An abbreviated version of the Cook-Medley Hostility Scale (ACM) was a predictor of documented acute myocardial infarction and total mortality. The sample consisted of 409 men and 321 women, residents of Glostrup, Denmark, who were 50 years old at the initiation of the study in 1964. Follow-up continued through 1991. Although not significant (relative risk (RR) = 1.22) in a model that contained only age and sex as covariates, the hostility scale scores were associated with increased risk of myocardial infarction in models controlling for traditional risk factors. A two standard deviation difference was associated with a RR of 1.53 (95% confidence interval (CI) 1.04-2.25), an effect that remained after eliminating the data of those with signs of ischemia at baseline. Hostility was also predictive of total mortality with controls for age and sex (RR = 1.35, CI 1.07-1.71), with controls for traditional risk factors (RR = 1.44, CI 1.13-1.83), and with additional controls for baseline ischemia and pulmonary function (RR = 1.36, CI 1.06-1.75). There were no sex differences in effect sizes. In cross-sectional analyses, high hostility was associated with higher body mass index, more physical activity at work, and poorer pulmonary function. These results constitute a rigorous test of the relation between hostility and health and increase the known generality of the phenomenon across sexes, age groups, and cultures.

Physical Activity Assessment Using a Pedometer and Its Comparison with a Questionnaire in a Large Population Survey

Abstract: The aim of this study was to evaluate the use of the pedometer in epidemiologic research on physical activity. Within the framework of a health examination survey in 1988-1989, physical activity was assessed in a representative population sample of 493 men and women aged 25-74 years who were residents of Switzerland. They wore a pedometer for 1 week at work and during leisure time, and the results, converted into steps per day, were compared with answers to a questionnaire. The average number of steps per day decreased from 11,900 to 6,700 and from 9,300 to 7,300 for men and women, respectively, in the youngest to the oldest age groups. For men, categorized according to type of physical activity at work, there was a highly significant difference in the number of steps (p < 0.001), whereas in women the results were associated with leisure-time physical activity (p = 0.003). For both sexes, practicing sports more than once a week was associated with an important increase in steps per day. Analyzing the number of steps according to the day of the week and occupational category produced an unexpected result : Men with a physically active job engaged in more leisure-time physical activity on the weekend. The pedometer proved to be useful in assessing physical activity in a large, free-living population.

Capture-recapture and multiple-record systems estimation II: Applications in human diseases

Abstract: This article evaluates the potential epidemiologic uses of capture-recapture, which include the primary area of determining disease frequency. Capture-recapture may be a means to effectively \"count\" new cases (incidence) or count existing cases (prevalence). Specific applications of capture-recapture in epidemiology are presented, one of which is its use in estimating death rates in a region close to Calcutta, India. The method also has considerable potential to assess suicides, and it may be the only technique to assess disease frequency in developing countries. In addition to generating an estimate of population size, another application of capture-recapture is to assess the costs of ascertainment relative to the degree of accuracy. This approach provides a formal means for assessing the cost-benefits of lists for the identification of cases. The authors believe that with careful and appropriate use, capture-recapture methods will provide a new approach that can considerably improve our ability to monitor disease.

Serum Fatty Acids and the Risk of Coronary Heart Disease

Abstract: To examine the relation between serum fatty acids and coronary heart disease (CHD), the authors conducted a nested case-control study of 94 men with incident CHD and 94 men without incident CHD who were enrolled in the Usual Care group of the Multiple Risk Factor Intervention Trial between December 1973 and February 1976. After confirming the stability of the stored serum samples, the authors measured serum fatty acid levels by gas-liquid chromatography and examined their association with CHD. In all multivariate models, levels of the cholesterol ester saturated fatty acid palmitic acid (16:0) were directly associated with CHD risk (standardized odds ratio = 1.68; 95% confidence interval 1.10-2.55 in the model that adjusted for total plasma cholesterol level). Levels of the phospholipid omega-3 fatty acid docosapentaenoic acid (22:5) were inversely associated with CHD risk in the two multivariate models that controlled for the effects of total plasma cholesterol level or high density lipoprotein cholesterol to total plasma cholesterol ratio (standardized odds ratio = 0.58; 95% confidence interval 0.38-0.89 in the first model that controlled for total plasma cholesterol level). In contrast to the first two multivariate models, levels of the docosahexaenoic acid (22:6) were inversely associated with CHD risk in a third multivariate model that controlled for the effects of high density lipoprotein cholesterol to low density lipoprotein cholesterol ratio (standardized odds ratio = 0.57; 95% confidence interval 0.36-0.90). These findings are consistent with other evidence indicating that saturated fatty acids are directly correlated with CHD and that omega-3 polyunsaturated fatty acids are inversely correlated with CHD. Because these associations were present after adjustment for blood lipid levels, other mechanisms, such as a direct effect on blood clotting, may be involved.

Nitrous Oxide and Spontaneous Abortion in Female Dental Assistants

Abstract: The relation between anesthetic gas exposure and spontaneous abortion remains unresolved. We examined the effect of nitrous oxide on spontaneous abortion among female dental assistants. Questionnaires were sent to 7,000 dental assistants aged 18-39 years who were registered in California in 1987; 4,856 (69%) responded. Analysis was based on 1,465 respondents whose most recent pregnancy was conceived while working full time. Women were asked how many hours a week they worked with nitrous oxide during this pregnancy and whether the excess gas was scavenged (vented). Relative risk of spontaneous abortion (through week 20) was calculated using a person-week model. This allowed women with current pregnancies (13%) or induced abortions (10%) to be included for appropriate time periods of risk. A total of 101 pregnancies (7%) ended as spontaneous abortions. An elevation in risk of spontaneous abortion was seen among women who worked with nitrous oxide for 3 or more hours per week in offices not using scavenging equipment (relative risk = 2.6, 95% confidence interval 1.3-5.0, adjusted for age, smoking, and number of amalgams prepared per week), but not among those using nitrous oxide in offices with scavenging equipment. This relation changed little when analyses were restricted to confirmed pregnancies or examined for several types of potential bias. Scavenging equipment appears to be important in protecting the reproductive health of women working with nitrous oxide.