Showing papers in "Epidemiology in 1992"
TL;DR: It is shown that adjustment for the intermediate variable, which is the most common method of estimating direct effects, can be biased, and that, even in a randomized crossover trial of exposure, direct and indirect effects cannot be separated without special assumptions.
Abstract: We consider the problem of separating the direct effects of an exposure from effects relayed through an intermediate variable (indirect effects). We show that adjustment for the intermediate variable, which is the most common method of estimating direct effects, can be biased. We also show that even in a randomized crossover trial of exposure, direct and indirect effects cannot be separated without special assumptions; in other words, direct and indirect effects are not separately identifiable when only exposure is randomized. If the exposure and intermediate never interact to cause disease and if intermediate effects can be controlled, that is, blocked by a suitable intervention, then a trial randomizing both exposure and the intervention can separate direct from indirect effects. Nonetheless, the estimation must be carried out using the G-computation algorithm. Conventional adjustment methods remain biased. When exposure and the intermediate interact to cause disease, direct and indirect effects will not be separable even in a trial in which both the exposure and the intervention blocking intermediate effects are randomly assigned. Nonetheless, in such a trial, one can still estimate the fraction of exposure-induced disease that could be prevented by control of the intermediate. Even in the absence of an intervention blocking the intermediate effect, the fraction of exposure-induced disease that could be prevented by control of the intermediate can be estimated with the G-computation algorithm if data are obtained on additional confounding variables.
1,593 citations
TL;DR: The methodology for obtaining confidence interval estimates of Relative excess risk due to interaction, the proportion of disease among those with both exposures that is attributable to their interaction, and the synergy index utilizing routinely available output from multiple logistic regression software is presented.
Abstract: Relative excess risk due to interaction, the proportion of disease among those with both exposures that is attributable to their interaction, and the synergy index have been proposed as measures of interaction in epidemiologic studies. This paper presents the methodology for obtaining confidence int
1,025 citations
TL;DR: The relation between vitamin C intake and mortality in the First National Health and Nutrition Examination Survey (NHANES I) Epidemiologic Follow-up Study cohort is examined, with no clear relation for individual cancer sites, except possibly an inverse relation for esophagus and stomach cancer among males.
Abstract: We examined the relation between vitamin C intake and mortality in the First National Health and Nutrition Examination Survey (NHANES I) Epidemiologic Follow-up Study cohort. This cohort is based on a representative sample of 11,348 noninstitutionalized U.S. adults age 25-74 years who were nutritionally examined during 1971-1974 and followed up for mortality (1,809 deaths) through 1984, a median of 10 years. An index of vitamin C intake has been formed from detailed dietary measurements and use of vitamin supplements. The relation of the standardized mortality ratio (SMR) for all causes of death to increasing vitamin C intake is strongly inverse for males and weakly inverse for females. Among those with the highest vitamin C intake, males have an SMR (95% confidence interval) of 0.65 (0.52-0.80) for all causes, 0.78 (0.50-1.17) for all cancers, and 0.58 (0.41-0.78) for all cardiovascular diseases; females have an SMR of 0.90 (0.74-1.09) for all causes, 0.86 (0.55-1.27) for all cancers, and 0.75 (0.55-0.99) for all cardiovascular diseases. Comparisons are made relative to all U.S. whites, for whom the SMR is defined to be 1.00. There is no clear relation for individual cancer sites, except possibly an inverse relation for esophagus and stomach cancer among males. The relation with all causes of death among males remains after adjustment for age, sex, and 10 potentially confounding variables (including cigarette smoking, education, race, and disease history).
548 citations
TL;DR: The analytic approach proposed in this paper will be necessary to control bias in any epidemiologic study in which there exists a time-dependent risk factor for death, such as pneumocystis carinii pneumonia history, that influences subsequent exposure to the agent under study, for example, prophylaxis therapy.
Abstract: AIDS Clinical Trial Group Randomized Trial 002 compared the effect of high-dose with low-dose 3-azido-3-deoxythymidine (AZT) on the survival of AIDS patients. Embedded within the trial was an essentially uncontrolled observational study of the effect of prophylaxis therapy for pneumocystis carinii pneumonia on survival. In this paper, we estimate the causal effect of prophylaxis therapy on survival by using the method of G-estimation to estimate the parameters of a structural nested failure time model (SNFTM). Our SNFTM relates a subject's observed time of death and observed prophylaxis history to the time the subject would have died if, possibly contrary to fact, prophylaxis therapy had been withheld. We find that, under our assumptions, the data are consistent with prophylaxis therapy increasing survival by 16% or decreasing survival by 18% at the alpha = 0.05 level. The analytic approach proposed in this paper will be necessary to control bias in any epidemiologic study in which there exists a time-dependent risk factor for death, such as pneumocystis carinii pneumonia history, that (A1) influences subsequent exposure to the agent under study, for example, prophylaxis therapy, and (A2) is itself influenced by past exposure to the study agent. Conditions A1 and A2 will be true whenever there exists a time-dependent risk factor that is simultaneously a confounder and an intermediate variable.
337 citations
TL;DR: The observation described here has implications for public health and medical treatment and for statistical interpretation; different cutpoints should be examined to determine the optimal cutpoint in terms of policy and/or treatment decisions.
Abstract: Dichotomizing a continuous outcome variable casts that variable in traditional epidemiologic terms (that is, disease, no disease). One consequence is overall reduced statistical power. A more fundamental concern is that the magnitude of various measures of association (for example, prevalence ratio, odds ratio) and statistical power depend on the cutpoint used to dichotomize the variable. The phenomenon is illustrated with a hypothetical situation assuming a two-level predictor variable and a normally distributed outcome variable. As the cutpoint is increased from lower to higher values, the prevalence ratio increases steadily, the odds ratio is described by a U-shaped curve, and statistical power is described by an inverted U-shaped curve. Furthermore, the extent of these effects depends on the difference between the means of the continuous outcome variable for the two levels of the predictor variable. An empirical example is given using data on education and blood pressure (dichotomized to create a high blood pressure vs low blood pressure variable). Except at each end of the distribution, the results follow the hypothetical example. The observation has implications for public health and medical treatment; different cutpoints should be examined to determine the optimal cutpoint in terms of policy and/or treatment decisions. The observation described here also has implications for statistical interpretation; statements about the magnitude of association or statistical significance have limited meaning unless both the cutpoint and the distribution of the outcome variable are specified.
248 citations
TL;DR: An increased relative risk of both Crohn's disease and ulcerative colitis associated with consumption of fast foods was found and coffee seemed to provide a protective effect for both diseases.
Abstract: We conducted a population-based case-control study of inflammatory bowel disease and dietary habits in Stockholm during 1984-1987. We obtained retrospective information about food intake 5 years previously by a postal questionnaire for 152 cases with Crohn's disease, 145 cases with ulcerative colitis, and 305 controls. The relative risk of Crohn's disease was increased for subjects who had a high (55 gm or more per day) intake of sucrose (relative risk = 2.6, 95% confidence interval = 1.4-5.0) and was decreased for subjects who had a high (15 gm or more per day) intake of fiber (relative risk = 0.5, 95% confidence interval = 0.3-0.9). The most striking finding was an increased relative risk of both Crohn's disease and ulcerative colitis associated with consumption of fast foods: the relative risk associated with consumption of fast foods at least two times a week was estimated at 3.4 (95% confidence interval = 1.3-9.3) for Crohn's disease and 3.9 (95% confidence interval = 1.4-10.6) for ulcerative colitis. Although coffee seemed to provide a protective effect for both diseases, there are reasons to consider this finding an artifact.
214 citations
TL;DR: The findings of this registry-based case-control study raise concerns that female smokers may assume a proportionally greater burden of lung cancer morbidity and mortality in the future.
Abstract: We conducted a registry-based case-control study to examine the relation between smoking and lung cancer by gender and histologic type. Our analyses were based on 14,596 cases and 36,438 age-matched controls. Relative risk associated with ever-smoking, and level of smoking was consistently higher in females than males for all lung cancers combined (ever-smoking odds ratios: 12.7 for females and 9.1 for males) and for each histologic type except adenocarcinoma. Female-male differences in relative risk were larger in younger age groups. The largest estimates of the attributable fraction due to smoking were observed for small cell carcinoma (97% in females and 91% in males); conversely, the smallest value was noted for adenocarcinoma (86% in females). Although our study was unable to measure absolute risk, our findings, other recent studies, and contemporary female smoking patterns raise concerns that female smokers may assume a proportionally greater burden of lung cancer morbidity and mortality in the future.
193 citations
TL;DR: In a sample of the United States population from the Census Bureau's Current Population Surveys, the authors compared demographic characteristics with those recorded on the death certificate for the 43,000 decedents in the sample followed from 1979 to 1985.
Abstract: In a sample of the United States population from the Census Bureau's Current Population Surveys, we compared demographic characteristics with those recorded on the death certificate for the 43,000 decedents in the sample followed from 1979 to 1985. Overall percentage agreements were: Sex 99.5, Race
190 citations
TL;DR: R residence in municipalities where chloroform concentrations were ≥10 μg/liter was associated with an increased risk for intrauterine growth retardation and the ascertainment and classification of exposures to trihalomethanes, including such issues as the imprecision of using aggregate municipal measures for classifying exposure at the level of the individual.
Abstract: The potential reproductive effects of long-term, low-dose exposure to chloroform have received little attention despite the known, acute toxicity of high exposures and the wide-spread occurrence of low concentrations in drinking water. We studied the association of waterborne chloroform with low birthweight (less than 2,500 gm), prematurity (less than 37 weeks gestation), and intrauterine growth retardation (less than 5th percentile of weight for gestational age). Cases were not mutually exclusive, but each outcome was analyzed independently. Birth certificates from January 1, 1989, to June 30, 1990, were used to identify cases and randomly selected controls. All were live, singleton infants born to non-Hispanic, white women from Iowa towns with 1,000-5,000 inhabitants. Exposures to chloroform and other trihalomethanes were ecologic variables based on maternal residence and a 1987 municipal water survey. After adjustment for maternal age, parity, adequacy of prenatal care, marital status, education, and maternal smoking by multiple logistic regression, residence in municipalities where chloroform concentrations were greater than or equal to 10 micrograms/liter was associated with an increased risk for intrauterine growth retardation (odds ratio = 1.8, 95% confidence interval = 1.1-2.9). The major limitations of this study involve the ascertainment and classification of exposures to trihalomethanes, including such issues as the imprecision of using aggregate municipal measures for classifying exposure at the level of the individual, the potential misclassification due to residential mobility, and the fluctuation of trihalomethane levels.
153 citations
TL;DR: The nonindependent misclassification described in this article will result in a positive bias in the odds ratio and is therefore of prime concern when questioning the validity of an observed effect.
Abstract: When misclassification of exposure and disease is nondifferential but not independent of one another, bias away from the null can result. For dichotomous variables, misclassification is nonindependent when the probability of misclassification of one variable is dependent on the correctness of classi
141 citations
TL;DR: A food frequency questionnaire on usual food intake is developed, which includes a set of pictures to estimate the portion size of 23 different dishes, which found a tendency toward overestimation of portion size by those who ate smaller portions and underestimation, by Those who ate larger portions.
Abstract: In the context of the planning phase of a large cohort study on dietary habits and cancer, we have developed a food frequency questionnaire on usual food intake, which includes a set of pictures to estimate the portion size of 23 different dishes. The validity of the estimation of portion size of recently consumed foods has been questioned by other researchers. To validate the use of pictures, we organized a field trial with 103 volunteers. They were invited to a dinner where standard Italian dishes were offered (total 17 foods); all of the portions they chose were recorded and weighed. The following day, we interviewed the volunteers on what they consumed during the dinner, and we compared the weight of the food actually eaten with the weight of the food represented in the pictures. The volunteers overestimated the portion size by more than 20% for six foods and underestimated the portion size by more than 20% for four foods. In addition, we found a tendency toward overestimation of portion size by those who ate smaller portions and underestimation, by those who ate larger portions ("flat slope syndrome").
TL;DR: Evidence is presented that post-study power calculations have little value and should be replaced by a more informative method using the upper (1 - alpha)% confidence limit of the point estimate that touches the value of the relative risk of interest.
Abstract: Frequently, after an epidemiologic study is completed, statistical power to detect a relative risk of interest is recalculated using data obtained during the course of the study. A negative study may then be dismissed on the grounds that its power was too low. However, post hoc power calculations ig
TL;DR: Data was analyzed from a large case-control study of spontaneous abortion that ascertained maternal alcohol consumption before and during pregnancy, as well as paternal consumption, to calculate a weighted average of the amount consumed weekly during the first trimester.
Abstract: Maternal alcoholism can lead to the fetal alcohol syndrome in offspring, but the effect of more moderate alcohol consumption during pregnancy remains an issue of concern. Therefore, we analyzed data from a large case-control study of spontaneous abortion (626 cases, 1,300 controls) that ascertained maternal alcohol consumption before and during pregnancy, as well as paternal consumption. Asking when in pregnancy alcohol consumption changed allowed us to calculate a weighted average of the amount consumed weekly during the first trimester. The odds ratio for consumption of seven or more drinks per week was 1.9 [95% confidence interval (CI) = 1.1-3.4] when adjusted for maternal smoking, passive smoking, and maternal age. Data were too sparse to examine higher consumption levels. There was some evidence that cases may have had less opportunity than controls to decrease consumption during their shorter pregnancies, potentially biasing the odds ratio upward. The adjusted odds ratio for any paternal alcohol consumption was 1.2 (CI = 0.93-1.5), with no dose-response effect seen. Among pregnancies in which the mother did not drink, there was no association with paternal drinking.
TL;DR: Data from the Health Care Financing Administration on 151,986 discharges listing a diagnosis of vertebral fracture over a 4-year period suggests that the race-sex differences in vertebral fractures discharge rates may be due to differences in the incidence of vertebra fracture.
Abstract: Whereas fractures related to osteoporosis have become a pressing public health concern, relatively few epidemiologic studies have focused on vertebral fractures. To shed further light on the occurrence of this injury, we collected data from the Health Care Financing Administration on 151,986 discharges listing a diagnosis of vertebral fracture over a 4-year period. After adjusting for age, white women experienced the highest rates of discharge, at 17.1 per 10,000 per year, followed by white men (9.9 per 10,000), black women (3.7 per 10,000), and black men (2.5 per 10,000). Among white women, discharge rates rose exponentially from 5.3 discharges per 10,000 population at age 65 to nearly 47.8 per 10,000 at age 90. White men, black women, and black men experienced less dramatic age-related increases in discharge rates. The similarity of these patterns to discharge rates for hip fracture suggests that the race-sex differences in vertebral fracture discharge rates may be due to differences in the incidence of vertebral fracture.
TL;DR: Evidence indicates that early life events and conditions, possibly extending to the intrauterine stages of life, and including energy restriction in early life, affect the risk of breast cancer, and can explain certain epidemiologic findings that cannot be accounted for by more established breast cancer risk factors.
Abstract: Evidence indicates that early life events and conditions, possibly extending to the intrauterine stages of life, and including energy restriction in early life, affect the risk of breast cancer. The mechanism of this effect is likely to be through a reduction in mammary gland mass and, inferentially, the total number of ductal stem cells. The evidence derives from epidemiologic and animal studies. It can explain certain epidemiologic findings that cannot be accounted for by more established breast cancer risk factors, including the more frequent occurrence of breast cancer in the left breast and the higher incidence of this disease among caucasian women than among Asian women in Asia.
TL;DR: The data suggest the interchangeability of the recall and record methods and their preference over the food frequency questionnaire for mean estimates of group nutrient intake, and the need to develop and evaluate short methods of diet assessment in specific populations of interest.
Abstract: We compared estimates of nutrient intake by three diet assessment methods (24-hour recall, 3-day food record, food frequency questionnaire) in a random sample of 73 females and 77 males from the Framingham Offspring/Spouse Study. The results differed according to analytic method. Estimates of group mean intake from the 24-hour recall and 3-day records were similar in both women and men, with differences of less than 10% for most nutrients. The estimates of mean intake calculated from the food frequency questionnaire generally differed from those obtained by the other methods, with higher estimated intakes in women and generally lower estimated intakes in men. Spearman rank correlations between the individuals' nutrient intakes estimated by the three diet assessment methods were modest (r = 0.08-0.68, most below 0.50) and comparable in the comparisons of the 24-hour recall or food frequency questionnaire with the 3-day records. Our data suggest the interchangeability of the recall and record methods and their preference over the food frequency questionnaire for mean estimates of group nutrient intake. The food frequency questionnaire appears to be of some utility in ranking individuals according to the usual intake, although these data are not informative in comparing this method with multiple days of recall or records beyond 3 days. This research underscores the need to develop and evaluate short methods of diet assessment in specific populations of interest.
TL;DR: Taken as a whole, the evidence is compelling that arsenic and smoking act in a synergistic manner to produce lung cancer.
Abstract: We assembled data from numerous studies to examine whether active smoking and occupational exposure to arsenic act synergistically (more than additively) to increase the risk of lung cancer. Although several smaller studies lacked the power to reject simple additive relations, the joint effect from both exposures consistently exceeded the sum of the separate effects by about 70 to 130%. The only study not showing a greater than additive effect appeared to have inadequate data to address this question. We calculated the excess fractions for the synergism; these showed that a minimum of between 30% and 54% of lung cancer cases among those with both exposures could not be attributed to either one or the other exposure alone. Previous authors addressing the synergism between arsenic exposure and smoking have evaluated deviations from a multiplicative model, which is inappropriate for this purpose. Reports of no interaction or "negative" interaction have therefore been misleading. Taken as a whole, the evidence is compelling that arsenic and smoking act in a synergistic manner to produce lung cancer. Substantial reductions in the lung cancer burden of smokers occupationally exposed to arsenic could be achieved by reductions in either exposure. The mechanism for the synergism is unclear.
TL;DR: The data indicate that smokers with a high intake of foods rich in fat and animal protein or who have a preference for cured meats are at increased risk of lung cancer.
Abstract: We conducted a population-based case-control study of the association of dietary cholesterol and fat with lung cancer between 1983 and 1985 on Oahu, Hawaii. The study population included 226 men and 100 women with lung cancer, and 597 male and 268 female community controls matched for age (±5 years)
TL;DR: Age showed a particularly strong association with the risk of GI hospitalization, and RRs were the highest in indomethacin users, and the lowest in sulindac users (3.1, 2.3–4.2).
Abstract: We evaluated the association between individual nonsteroidal antiinflammatory drugs (NSAIDs) and gastrointestinal (GI) toxicity in a retrospective cohort study aimed at examining and comparing the incidence of serious gastrointestinal disorders among NSAIDs users. We observed 2,302 GI hospitalizatio
TL;DR: Both reporting bias and selection bias are shown to be algebraically equivalent to bias arising from exposure misclassification, and the choice of study design involves balancing these two sources of bias.
Abstract: Retrospective studies of congenital malformations frequently rely on exposures reported by study subjects. Differential error in exposure reporting by cases and controls, which has alternatively been referred to as "recall bias" and "reporting bias," may result in a biased effect measure. Some authors have attempted to avoid reporting bias by comparing exposures between two malformed groups, rather than between cases and nonmalformed controls. This approach, however, may introduce its own bias, which we call selection bias. Both reporting bias and selection bias are shown to be algebraically equivalent to bias arising from exposure misclassification. The magnitudes of these biases are compared for a range of plausible parametric values. The case-control design is sensitive to both differential reporting and selection bias, and the choice of study design involves balancing these two sources of bias.
TL;DR: In this paper, the authors examined the risk factors for nonepithelial ovarian tumors by combining data from four case-control studies conducted in the United States and compared personal characteristics of 38 germ cell cases and 45 stromal cases, respectively, with 1,142 and 2,617 general population controls.
Abstract: Nonepithelial ovarian cancers are rare, and little is known about their etiology. Of particular interest are the effects of oral contraceptive use and pregnancy, both of which are associated with large decreases in risk for epithelial ovarian cancer. We examined the risk factors for nonepithelial ovarian tumors by combining data from four case-control studies conducted in the United States. We compared personal characteristics of 38 germ cell cases and 45 stromal cases, respectively, with 1,142 and 2,617 general population controls. All subjects were over age 18 years. For germ cell tumors, there was a weak negative association with parity but no consistent pattern of decreasing risk with increasing parity. In contrast, relative to nulligravid women, gravid nulliparous women were at increased risk of developing a germ cell cancer [odds ratio (OR) = 4.8, 95% confidence interval (CI) = 1.2-18.6]. The use of oral contraceptives was also associated with elevated risk (OR = 2.0, 95% CI = 0.77-5.1); however, no clear trends in risk were observed. For stromal tumors, oral contraceptive use was associated with decreased risk (OR = 0.37, 95% CI = 0.16-0.83), whereas pregnancy was associated with a small elevation in risk. A trend of increasing risk with increasing age at first term pregnancy was observed, with an odds ratio of 3.6 (95% CI = 1.0-12.5) for a first birth after age 29 years. Risk factors for nonepithelial ovarian cancers do not appear to parallel each other or those for epithelial ovarian cancer.
TL;DR: A cohort of 1,904 vegetarians and persons leading a health-conscious life-style in the Federal Republic of Germany was identified in 1978 and mortality from from all causes was reduced by one-half compared with the general population after a follow-up of 11 years.
Abstract: A cohort of 1,904 vegetarians and persons leading a health-conscious life-style in the Federal Republic of Germany was identified in 1978. After a follow-up of 11 years, mortality from all causes was reduced by one-half compared with the general population [the standardized mortality ratio (SMR) was 0.44 for men, 0.53 for women]. Among the 858 men, 111 deaths were observed, with 255 expected; among the 1,046 women, 114 deaths were observed, with 215 expected. The lowest mortality was found for cardiovascular diseases (SMR = 0.39 for men, 0.46 for women); in particular, for ischemic heart diseases, mortality was reduced to one-third of that expected. Cancer mortality was reduced by one-half in men (SMR = 0.48), but only by one-quarter in women (SMR = 0.74). The deficit in cancer deaths was mainly observed for lung cancer and gastrointestinal cancers in males and for gastrointestinal cancers in females. Deaths from diseases of the respiratory and digestive systems were also reduced by about 50%. An excess of deaths occurred only for anemia. When the strict and the moderate vegetarians were analyzed separately, the strongest differential was found for ischemic heart diseases, which were much less frequent among strict vegetarians for both sexes. Some nondietary factors, such as higher socioeconomic status, virtual absence of smoking, and lower body mass index, may also have contributed to the lower mortality of the study participants.
TL;DR: Examination of subtypes of moderately preterm delivery in blacks and whites using North Carolina birth certificate data for 1988–1989 showed that more highly educated blacks still had higher risks of moderately and very preterm deliveries than less educated whites.
Abstract: The differences in preterm birth between blacks and whites are poorly understood. Our study examined subtypes of moderately preterm delivery (34–36 completed weeks of gestation) and very preterm delivery (20–33 weeks) in blacks and whites using North Carolina birth certificate data for 1988–1989. We
TL;DR: Self-reported alcohol consumption throughout adult life was reported with precision sufficient to make the ranking of subjects' intake consistent between interviews.
Abstract: We examined the reliability of self-reported alcohol consumption in past age periods of women's lives. As part of a case-control study of breast cancer conducted in Massachusetts and Wisconsin in 1988-1991, the same questionnaire was administered for a second time to 211 controls (mean age = 54 years) after an interval of 6-12 months. The Spearman correlation coefficients between the average number of grams of alcohol consumed daily reported in the two interviews, by age period of consumption, were: 16-19 years, r = 0.81; 20-29 years, r = 0.84; 30-39 years, r = 0.75; and for recent consumption, r = 0.77. Self-reported alcohol consumption throughout adult life was reported with precision sufficient to make the ranking of subjects' intake consistent between interviews.
TL;DR: No temporal trend in diabetes incidence was observed overall or by ethnic group, and the increasing insulin-dependent diabetes incidence reported by registries in Europe during this time period was not observed in Colorado.
Abstract: We examined the incidence of insulin-dependent diabetes mellitus in children 0-17 years of age in Colorado from 1978 to 1988. Cases were ascertained from a statewide registry based on physician surveillance. A total of 1,376 children were diagnosed during this interval in a population averaging 860,000 children. The degree of ascertainment was estimated to be 93.3%. The age-adjusted incidence rate of insulindependent diabetes mellitus was 14.8/100,000 person-years. The rate was lower in individuals of Spanish origin (Hispanics) (8.7/100,000 person-years) compared with non-Hispanic individuals (15.5/100,000 person-years) (incidence rate ratio=0.6, 95% confidence interval=0.4-0.8). Incidence rates were higher in winter and lower in summer for children 5- 17 years old. Children diagnosed before the age of 5 years showed no significant seasonal pattern, although peak incidences were observed in autumn and spring. No temporal trend in diabetes incidence was observed overall or by ethnic group. The increasing insulin-dependent diabetes incidence reported by registries in Europe during this time period was not observed in Colorado. (Epidemiology 1992;3:232-238)
TL;DR: Analysis of lifetime occupational histories revealed a small number of noteworthy associations between occupation and oral and pharyngeal cancer, and increased risk was increased among male carpet installers and tended to rise with longer duration of employment.
Abstract: We studied the relation between occupation and oral and pharyngeal cancer with a population-based case-control study conducted in four areas of the United States. The study group included 1,114 incident male and female cases and 1,268 frequency-matched controls. After adjustment for age, race, smoki
TL;DR: The motivation behind the Pandora project was to use METAFONT as a design tool rather than as a production tool in the creation of a typeface.
Abstract: The motivation behind the Pandora project was to use METAFONT as a design tool rather than as a production tool in the creation of a typeface. Pandora was developed by using generalized descriptions of the visual relationships between parts of characters, characters in a font, and fonts in a typeface family. Pandora transforms from one font in her family to the next through different parameter settings applied to a single framework. It is not important that all variations look good, rather that a reasonable set can be found within the framework. A rich description allows a designer to quickly look at a number of possibilities. 1.1 Background When I first heard about METRFONT in 1980, the idea seemed intriguing-using computers to design type. In 1984 when actually faced with the first generation of SUN computers, an experimental V operating system with a cryptic boot command, and version O.** of METAFONT, I was not sure what it could really door rather, what 1 could convince it to do. The idea of designing type that was modifiable by the proverbial \"turning of knobs\" was seductive. And the idea that you, the designer, could choose those knobs was even more so. It was not very confidence-inspiring to know that I had no computer experience and there was no manual yet to use as a guide. However, I did have some ideas of what I would like METAFONT to do and lots of help from the authors of the language. Designing begins with ideas of how to solve a problem. What is the problem, what tools can be used to help solve the problem, how are those tools best used, what palette of solutions presents itself? The problem of type design involves identifying the requirements of a design. Such requirements are outlined in a document called a design brief. For what purpose is the design being considered? Is it a text or a display face, or intended primarily for use as symbols? What visual flavor is it to have? What feeling does it convey? Will it have a single weight or will a family of weights need to be considered? Type is intended to be reproduced without deviation within a given range of tolerance. So, which technologies will be used in reproducing it? Additionally for METAFONT-what lund of flexibility do you want to build into the design to make it adaptable to different marlung …
TL;DR: It is concluded that women who have a recent history of migraine may be at higher risk of pregnancy-induced hypertension.
Abstract: This analysis assesses the relation between a history of migraine and the risk of preeclampsia or gestational hypertension. Cases (172 women with preeclampsia and 254 with gestational hypertension) and controls (505) were primiparae with no history of hypertension before pregnancy. Information on migraine attacks in the year before pregnancy was obtained after delivery. Migraine was reported by 16% of preeclamptic women, 12% of women with gestational hypertension, and 8% of the controls. Adjusted odds ratios (95% confidence interval) of preeclampsia and gestational hypertension were 2.44 (1.42-4.20) and 1.70 (1.02-2.85), respectively. We conclude that women who have a recent history of migraine may be at higher risk of pregnancy-induced hypertension.
TL;DR: This paper examined factors associated with the subsequent development of AIDS-related Kaposi's sarcoma in a cohort of 353 homosexual men infected with human immunodeficiency virus (HIV).
Abstract: We examined factors associated with the subsequent development of AIDS-related Kaposi's sarcoma in a cohort of 353 homosexual men infected with human immunodeficiency virus (HIV). Cumulative incidence curves for the development of Kaposi's sarcoma and opportunistic infection were stratified over a wide range of variables at enrollment, including those related to demographics, sexual behavior, illicit drug use, and medical history. We found no strong associations between any of these variables and the development of opportunistic infection, but two were related to Kaposi's sarcoma: use of nitrite inhalants (relative risk, 2.3; 95% confidence interval, 1.0-5.0) and high numbers of sexual contacts during the period 1978-1982 in the AIDS epidemic centers of San Francisco, Los Angeles, and/or New York (relative risk, 3.5; 95% confidence interval, 1.6-7.6). The latter variables remained independently associated with risk of Kaposi's sarcoma even after multivariate adjustment for a number of classical HIV risk factors. These results are consistent with the hypothesis that Kaposi's sarcoma is caused by a sexually transmitted cofactor that has remained more prevalent in the original epidemic centers. The effect of nitrites could be due to an independent biological mechanism or to enhancement of transmission of the cofactor. (Epidemiology 1992;3:203-209)
TL;DR: No important association was found between malathion exposure and spontaneous abortion, intrauterine growth retardation, stillbirth, or most categories of congenital anomalies, and the statistical analysis compared each of the adverse pregnancy outcome groups against an appropriate control group using logistic regression or survival time regression approaches.
Abstract: We studied reproductive outcomes in a cohort of 7,450 pregnancies identified through three Kaiser-Permanente facilities in the San Francisco Bay Area, in relation to exposure to the pesticide malathion, applied aerially to control an infestation by the Mediterranean fruit fly. We included in the cohort all women over age 17 who were registered at these facilities and who were confirmed as pregnant during the spraying period. Residence histories throughout the pregnancy were obtained by mailed questionnaire or telephone interview from 933 women with adverse outcomes and a sample of 1,000 women with normal outcomes, and were converted to geographical coordinates. We linked the coordinates for malathion spraying corridors with the residence coordinates to create individual exposure indices for each week of pregnancy. The statistical analysis compared each of the adverse pregnancy outcome groups against an appropriate control group using logistic regression or survival time regression approaches. After adjustment for various confounders, no important association was found between malathion exposure and spontaneous abortion, intrauterine growth retardation, stillbirth, or most categories of congenital anomalies. Gastrointestinal anomalies were related to second trimester exposure (odds ratio = 2.6), based on 13 cases and not specific to any particular International Classification of Diseases code. (Epidemiology 1992;3:32–39)