Search

Background: Existing studies on the relationships between impairments and activities of daily living (ADLs) in nursing home residents have serious limitations. This study examines the relationships among admission impairments, including pain, depression, incontinence, balance, and falls, and follow-up ADLs, as well as the effect of the nursing home on follow-up ADLs of extended-stay nursing home residents., Methods: This longitudinal cohort study consisted of 4,942 extended-stay residents who were admitted into 377 Minnesota nursing homes during 2004. General linear mixed models were used for all analyses, with 14 resident-level and 8 facility-level control variables., Results: Incontinence and balance function at admission were significantly associated with increases in ADL dependence at follow-up. Individual nursing homes had independent effects on all three ADL models. Similar findings were found after facility-level control variables were added., Conclusions: Incontinence predicts subsequent ADL functional levels. The relationship between balance dysfunction and subsequent ADL dependence could be causal. Future studies of the causal relationships between impairments and ADL should examine the effectiveness of impairment interventions on ADL as well as these relationships in different subgroups of nursing home residents.

In pesticide biomonitoring studies, researchers typically collect either single voids or daily (24-h) urine samples. Collection of 24-h urine samples is considered the "gold-standard", but this method places a high burden on study volunteers, requires greater resources, and may result in misclassification of exposure or underestimation of dose due to noncompliance with urine collection protocols. To evaluate the potential measurement error introduced by single void samples, we present an analysis of exposure and dose for two commonly used pesticides based on single morning void (MV) and 24-h urine collections in farmers and farm children. The agreement between the MV concentration and its corresponding 24-h concentration was analyzed using simple graphical and statistical techniques and risk assessment methodology. A consistent bias towards overprediction of pesticide concentration was found among the MVs, likely in large part due to the pharmacokinetic time course of the analytes in urine. These results suggest that the use of single voids can either over- or under-estimate daily exposure if recent pesticide applications have occurred. This held true for both farmers as well as farm children, who were not directly exposed to the applications. As a result, single void samples influenced the number of children exposed to chlorpyrifos whose daily dose estimates were above levels of toxicologic significance. In populations where fluctuations in pesticide exposure are expected (e.g., farm families), the pharmacokinetics of the pesticide and the timing of exposure events and urine collection must be understood when relying on single voids as a surrogate for longer time-frames of exposure.

This chapter highlights important steps in using correlation and simple linear regression to address scientific questions about the association of two continuous variables with each other. These steps include estimation and inference, assessing model fit, the connection between regression and ANOVA, and study design. Examples in microbiology are used throughout. This chapter provides a framework that is helpful in understanding more complex statistical techniques, such as multiple linear regression, linear mixed effects models, logistic regression, and proportional hazards regression.

The University of Minnesota Epilepsy Research and Education Program published two studies evaluating the use of antiepileptic drugs (AEDs) among nursing home (NH) elderly. The studies used a large, nongovernmental data set for studying this population. This chapter is a summary of those two studies. In the first study, a 1-day point prevalence study, 10.5% of the NH residents had one or more AED orders, a prevalence 10 times greater than that found in the community. In a multivariate analysis of factors associated with AED treatment, seizure indication was the most important factor, and age was inversely related to AED use. Phenytoin was the most commonly used AED, followed by carbamazepine, phenobarbital, and valproic acid. The most frequently used combination was phenytoin and phenobarbital. In the second study, evaluating NH admission data, 8% of newly admitted residents were already receiving one or more AEDs when they entered the NH. Factors associated with AED use in this group included epilepsy/seizure disorder, age, cognitive performance, and manic depression (bipolar disease). Among residents recently admitted who were not using an AED at entry, 3% were initiated on an AED within 3 months of admission. Among the factors associated with the initiation of AEDs during this period, the strongest association was with epilepsy/seizure disorder. Manic depression (bipolar disease) was also significantly associated with initiation of an AED after admission. In this group, there was an inverse relationship between age and initiation of an AED.

This chapter describes multiple linear regression, a statistical approach used to describe the simultaneous associations of several variables with one continuous outcome. Important steps in using this approach include estimation and inference, variable selection in model building, and assessing model fit. The special cases of regression with interactions among the variables, polynomial regression, regressions with categorical (grouping) variables, and separate slopes models are also covered. Examples in microbiology are used throughout.

Background: We examined the potential association between physical activity and lung cancer in women., Methods: In 1986, 36,929 women from Iowa, who were free of prior cancer, completed a questionnaire regarding physical activity, smoking, body mass index, and other life-style factors. Women were followed through 2002 for cancer incidence., Results: After adjusting for potential confounders, women with high physical activity levels were less likely [hazard ratio, 0.77; 95% confidence intervals (CI), 0.64-0.94] to develop lung cancer than women with low activity levels. This hazard ratio was 0.72 (95% CI, 0.55-0.94) in current smokers and 0.63 (95% CI, 0.43-0.92) in former smokers, but was not significant in never smokers., Conclusion: These results suggest that physical activity might reduce the risk of lung cancer in women who are current or former smokers.

There is a well-established, strong association between socioeconomic position and mortality. Public health mortality analyses thus routinely consider the confounding effect of socioeconomic position when possible. Particularly in the absence of personally reported data, researchers often use area-based measures to estimate the effects of socioeconomic position. Data are limited regarding the relative merits of measures based on US Census tract versus ZIP code (postal code). ZIP-code measures have more within-unit variation but are also more easily obtained. The current study reports on 293,138 middle-aged men screened in 14 states in 1973-1975 for the Multiple Risk Factor Intervention Trial and having 25-year mortality follow-up. In risk-adjusted proportional hazards models containing either ZIP-code-based or tract-based median household income, all-cause mortality hazard ratios were 1.16 (95% confidence interval: 1.14, 1.17) per Dollars 10,000 less ZIP-code-based income and 1.15 (95% confidence interval: 1.13, 1.16) per Dollars 10,000 less tract-based income; adding either income variable to a risk-adjusted model improved model fit substantially. Both were significant independent predictors in a combined model; tract-based income was a slightly stronger mortality predictor (hazard ratios = 1.05 and 1.11 for ZIP-code-based and tract-based income, respectively). These patterns held across various causes of death, for both Blacks and non-Blacks, and with or without adjustment for ZIP-code-based income diversity or tract-based proportion below poverty.

Purpose: Ginkgo biloba, among the most widely used herbs, possesses the capacity both to induce and to inhibit seizures. The purpose of this study was to describe the prevalence of ginkgo and other common herb prescribing in a sample of nursing home (NH) elderly diagnosed with epilepsy/seizure (Epi/Sz) disorder and to determine demographic, clinical, and functional factors associated with ginkgo use., Methods: This was a 1-year prevalence study of 68,403 NH residents living in 557 nursing facilities throughout the United States., Results: Overall, herb use in the NHs was very low (0.41%). Ginkgo was prescribed 162 times, more than any other herb. St. John's wort was prescribed 40 times; garlic, 29 times; and all others, <20 times, for a total of 307 herb orders. Among all residents with an herb order, ginkgo was prescribed for 61.9% of residents with an Epi/Sz diagnosis and 58.0% (p = 0.820) of residents without an Epi/Sz diagnosis. Dementia, educational level, and the interaction of age group with cognitive impairment were all significantly associated with herb use among Epi/Sz residents. Cognitive impairment and the interaction of Epi/Sz disorder with dementia were associated specifically with ginkgo use., Conclusions: Ginkgo is the most frequently prescribed herb in this population in which >50% of all herb orders were written for ginkgo. The concern with ginkgo use among elderly with Epi/Sz is the lack of standardization that characterizes ginkgo products. In the absence of standardization, the likelihood is increased that ginkgo products may be adulterated with the Ginkgo biloba plant parts most commonly associated with seizure provocation.

Latinos are now the largest minority in the United States, but their distinctive health needs and mortality patterns remain poorly understood. Proportional hazards regressions were used to compare Latino versus White risk- and income-adjusted mortality over 25 years' follow-up from 5,846 Latino and 300,647 White men screened for the Multiple Risk Factor Intervention Trial. Men were aged 35-57 years and residing in 14 states when screened in 1973-1975. Data on coronary heart disease risk factors, self-reported race/ethnicity, and home addresses were obtained at baseline; income was estimated by linking addresses to census data. Mortality follow-up through 1999 was obtained using the National Death Index. The fully adjusted Latino/White hazard ratio for all-cause mortality was 0.82 (95% confidence interval (CI): 0.77, 0.87), based on 1,085 Latino and 73,807 White deaths; this pattern prevailed over time and across states (thus, likely across Latino subgroups). Hazard ratios were significantly greater than one for stroke (hazard ratio = 1.30, 95% CI: 1.01, 1.68), liver cancer (hazard ratio = 2.02, 95% CI: 1.21, 3.37), and infection (hazard ratio = 1.69, 95% CI: 1.24, 2.32). A substudy found only minor racial/ethnic differences in the quality of Social Security numbers, birth dates, soundex-adjusted names, and National Death Index searches. Results were not likely an artifact of return migration or incomplete mortality data.

The objective of this study was to evaluate associations between herd-level factors and the isolation of Salmonella in calves from dairy farms in Minnesota, Wisconsin, Michigan and New York. Study farms were 129 conventional and organic farms enrolled without regard to previous history of Salmonella infection. Herds were sampled at 2-month intervals over a 1-year period. Salmonella was isolated in fecal samples from 176 (3.8%) of 4673 preweaned calves with 40 (31.0%) of 129 farms having at least one positive calf sample over the course of the study. Multivariable logistic regression using the generalized estimating equations approach was used to evaluate risk factors for Salmonella shedding after adjustment for effects of herd size, season, state of origin and the multiple sampling occasions per herd. Factors retained in the final model that were associated with an increased odds for Salmonella shedding were lack of routine feeding of milk replacer containing antimicrobials to preweaned calves (OR=2.8, 95% CI: 1.4, 5.8), use of maternity housing as a hospital area for sick cows more than once a month (OR=2.1, 95% CI: 1.1, 4.0), and cow prevalence level by visit, categorized into the following four-levels: > or =20% (OR=11.6, 95% CI: 5.7, 23.7), 10-19.9% (OR=4.7, 95% CI: 2.0, 11.5), 0.1-9.9% (OR=3.6, 95% CI: 1.5, 8.7) and 0% (reference level). Herd size was not associated with Salmonella shedding in the final multivariable model.

Objectives: We explored differences between Black and White men for cardiovascular disease (CVD) mortality across major risk factor levels., Methods: Major CVD risk factors were measured among 300,647 White and 20,223 Black men aged 35 to 57 years who were screened for the Multiple Risk Factor Intervention Trial (MRFIT). Hazard ratios for CVD deaths for Black and White men over 25 years of follow-up were calculated for subgroups stratified according to risk factor levels., Results: CVD was responsible for 2518 deaths among Black men and 30,772 deaths among White men. The age-adjusted Black-to-White CVD hazard ratio was 1.35 (95% confidence interval [CI]=1.29, 1.40); the risk- and income-adjusted ratio was 1.05 (95% CI=1.01, 1.10). CVD mortality rates were dramatically lower in cases of favorable risk profiles. However, fully adjusted Black-to-White CVD hazard ratios within groups at low, intermediate, high, and very high levels of overall risk were 1.76, 1.20, 1.10, and 0.94, respectively. Similar gradients were evident for individual risk factors., Conclusions: Higher CVD mortality rates among Black men were largely mediated by risk factors and income. These data underscore the need for sustained primordial risk factor prevention among Black men.

Objectives: To determine the prevalence of mono- and poly-antiepileptic drug (AED) therapy in a cohort of nursing home (NH) residents and to describe specific AED combinations used, as well as demographic, clinical, and functional factors associated with poly-AED therapy and the most common AED combination., Design: Retrospective, point prevalence study., Setting: All NHs owned/managed by Beverly Enterprises., Participants: All residents aged 65 and older residing in one of the study NHs on July 1, 1999, and receiving an AED (N=3,881)., Measurements: Data were gathered using two secondary source data sets: physicians' orders (AED use) and the Minimum Data Set (health status indicators)., Results: Of residents taking AEDs, 370 (9.5%) were taking two or more; 268 of those (72%) were taking problematic AED combinations (those with the potential of undesirable pharmacokinetic or pharmacodynamic interactions). Phenytoin (PHT) with phenobarbital (PB) was the most common combination (27.0%). Logistic regression indicated that poly-AED subjects were more likely to have a diagnosis of epilepsy/seizure (epi/sz) and less likely to have a diagnosis of cerebrovascular accident (CVA). Residents taking the PHT/PB combination were more likely to have an epi/sz diagnosis and longer NH stay. The association between CVA and PHT/PB polytherapy differed by presence or absence of aphasia., Conclusion: The overall prevalence of poly-AED therapy is less than 10% in NH residents, but 72% of those residents were receiving problematic polytherapy combinations, thereby exposing them to potential risk of adverse reactions and toxicity.

The generalized estimation equation (GEE) method is widely used in longitudinal data analysis, particularly when the outcome variable is non-Gaussian distributed. Under mild regulatory conditions, the parameter estimates are consistent and their asymptotic variances are efficient. In an observational study focusing on alcoholism patients, we applied the GEE method to longitudinal count data from medical utilization records from a large national managed care organization. The health services research question was whether there was a change in medical utilization for patients after engaging in alcoholism treatment as compared to before treatment. Thus, the main effect of interest was a time-varying covariate indicating whether the patient had undergone treatment yet or not. GEE under five different working correlations was employed and mixed results regarding the significance of the treatment effect were found. Because of the large sample size, i.e. 8485 patients with an average of 46 repeated measurements per patient, differences across the estimates produced by the different working correlation structures was suspicious. It is shown that these differences are maybe caused by the fact that the time-varying covariate in the marginal mean model is misspecified. A simulation study is performed to demonstrate that misspecification of the time-varying covariate in the marginal mean structure can cause differences in GEE results across various choices of working correlation structure.

The objective of this study was to evaluate associations between cattle-level factors and environmental samples with the isolation of Salmonella from dairy farms in Minnesota, Wisconsin, Michigan, and New York. The study farms included 129 conventional and organic farms enrolled without regard to previous history of Salmonella infection. Herds were sampled at two-month intervals over a one-year period. Cattle groups more likely to be associated with Salmonella shedding (compared to preweaned calves) were cows designated as sick by farm personnel (OR=2.5, 95% CI: 1.7, 3.7), cows within 14 days of calving (OR=1.8, 95% CI: 1.1, 2.8), and cows due for culling within 14 days (OR=1.9, 95% CI: 1.0, 3.4). State of origin was also associated with the presence of Salmonella in samples from cattle and the farm environment; Midwestern states were more likely to have Salmonella-positive samples compared to New York. Cattle treated with antimicrobials within 14 days of sampling were more likely to be Salmonella-negative compared with nontreated cattle (OR=2.0, 95% CI: 1.1, 3.4). Farms with at least 100 cows were more likely to have Salmonella-positive cattle compared with smaller farms (OR=2.6, 95% CI: 1.4, 4.6). Season was associated with Salmonella shedding in cattle, and compared to the winter period, summer had the highest odds for shedding (OR=2.4, 95% CI: 1.5, 3.7), followed by fall (OR=1.9, 95% CI: 1.2, 3.1) and spring (OR=1.8, 95% CI: 1.2, 2.6). Environmental samples significantly more likely to be Salmonella-positive (compared to bulk tank milk) included, in descending order, samples from sick pens (OR=7.4, 95% CI: 3.4, 15.8), manure storage areas (OR=6.4, 95% CI: 3.5, 11.7), maternity pens (OR=4.2, 95% CI: 2.2, 8.1), haircoats of cows due to be culled (OR=3.9, 95% CI: 2.2, 7.7), milk filters (OR=3.3, 95% CI: 1.8, 6.0), cow waterers (OR=2.8, 95% CI: 1.4, 5.7), calf pens (OR=2.7, 95% CI: 1.3, 5.3), and bird droppings from cow housing (OR=2.4, 95% CI: 1.3, 4.4). Parity, stage of lactation, and calf age were not associated with Salmonella shedding.

Study Objective: To discover if area level social cohesion, neighbourhood safety, and home safety are associated with current cigarette smoking among adults after adjustment for concentrations of poverty and low education., Design: Cross sectional survey of a random sample of adults, stratified by 19 geographical areas., Setting: SHAPE, Survey of the Health of Adults, the Population, and the Environment-conducted in 1998 by the Hennepin County Community Health Department and the Minneapolis Department of Health and Family Support in Minnesota., Participants: 5256 men and 4806 women, 18 years and older, randomly selected from 19 geographical areas in an urban county., Main Results: Overall, 21.2% of survey respondents reported current cigarette smoking. Both higher area level social cohesion (OR = 0.85, 95% CI = 0.74 to 0.98) and higher individual social cohesion (OR = 0.96, 95% CI = 0.92 to 0.99) were associated with lower likelihoods of smoking. Similar models were obtained for neighbourhood safety and home safety., Conclusions: These findings contribute to the growing literature on the important role of social cohesion and other area level characteristics on smoking behaviour among adults.

Analyses to compare non-randomized groups are more and more common in both post hoc analyses of randomized clinical trials data and in analyses of long-term observational data. In such cases, it is quite likely that there are unknown or uncollected sources of heterogeneity in event rates. Research has shown that an underlying source of heterogeneity in event rates which is not included in proportional hazards regression models leads to biased estimates for included covariate effect estimates and lower power to test them whether the source of heterogeneity is assumed to be fixed or random. We demonstrate here using several post hoc analyses of clinical trials data that a potentially common problem may be that the non-randomized groups which are to be compared have differential variability in their event rates. We then show through simulation that such underlying heterogeneity which varies across the groups, when ignored in the modelling, can lead to an attenuated regression effect estimate for comparing the two groups to each other, lower rejection rates for the effect, and Wald-based confidence intervals with potentially much lower coverage than nominal. When the groups are not significantly different, but heterogeneity differs between them, an analysis ignoring the heterogeneity can even result in a significant negative comparison., (Copyright 2004 John Wiley & Sons, Ltd.)

Background: The magnitude of coronary mortality risk associated with diabetes or prior myocardial infarction (MI) is debatable. Modulating effects of age, risk factors, and duration of follow-up may explain discrepancies in previous research. Associations with noncardiovascular mortality are little explored., Objectives: To compare mortality patterns in men with a history of diabetes or MI and to assess modulating effects on mortality of age, cardiovascular risk factors, and follow-up duration., Methods: We compared the 25-year mortality of 4809 men with diabetes only and 4625 men with MI only (all men aged 35-57 years)., Results: The adjusted hazard ratio (HR) for all-cause mortality for those with MI only vs those with diabetes only was 0.97 (95% confidence interval, 0.92-1.03; P =.32). The pattern of deaths was different: higher coronary mortality (HR = 1.37; P<.001) and lower mortality from noncardiovascular causes (HR = 0.66; P<.001) in those with MI only compared with those with diabetes only. This finding prevailed across all ages and levels of cardiovascular risk factors. Hazard ratios for coronary mortality significantly declined over follow-up (2.7, 1.7, 1.2, 1.1, and 1.0 for < or =5, 6-10, 11-15, 16-20, and >20 years of follow-up, respectively), whereas HRs for noncardiovascular mortality remained relatively constant., Conclusions: Overall, diabetes and MI were similarly strong predictors of total mortality. Higher mortality from noncardiovascular causes was observed in those with diabetes only, whereas prior MI was more strongly predictive of coronary mortality than diabetes at any age and level of cardiovascular risk factors. The difference in coronary mortality between the 2 groups was most evident in the first 10 years of follow-up.

The wild chimpanzees in Gombe National Park, Tanzania, fish for termites with flexible tools that they make out of vegetation, inserting them into the termite mound and then extracting and eating the termites that cling to the tool. Tools may be used in different ways by different chimpanzee communities according to the local chimpanzee culture. Here we describe the results of a four-year longitudinal field study in which we investigated how this cultural behaviour is learned by the community's offspring. We find that there are distinct sex-based differences, akin to those found in human children, in the way in which young chimpanzees develop their termite-fishing skills.

As part of the Minnesota Children's Pesticide Exposure Study we measured volatile organic compound (VOC) concentrations in a probability sample of households with children. The 6-day average concentrations for 10 common VOCs were obtained in urban and nonurban residences twice during this multiphase study: screening-phase indoor measurements were collected in 284 households, and in the intensive-phase matched outdoor (O), indoor (I), and personal (P) measurements were collected in a subset (N=72) of the screened households. Screening-phase households with smokers had significantly higher concentrations of benzene and styrene compared to nonsmoking households; households with an attached garage had significantly higher levels of benzene, chloroform, styrene, and m/p- and o-xylene compared to households without an attached garage; and nonurban residences, which had a greater prevalence of smokers and attached garages, had significantly higher 1,1,1-trichloroethane, styrene, and toluene and significantly lower tetrachloroethylene concentrations compared to urban households. The screening-phase weighted distributions estimate the mean and variability in indoor VOC concentrations for more than 45,000 households with children in the census tracts sampled. Overall, median indoor concentrations of most VOCs measured in this study were similar to or lower than indoor levels measured previously in the United States. Intensive-phase outdoor VOC concentrations were generally lower than other major metropolitan areas, but urban concentrations were significantly higher than nonurban concentrations for all compounds except 1,1,1-trichloroethylene. A consistent pattern of P>I>O was observed for nine of 10 VOCs, with 1,1,1-trichloroethylene (I>P>O) being the only exception to this pattern. For most children, the indoor at-home microevironment was strongly associated with personal exposure after controlling for important covariates, but the ratio of median to upper bound exposures was smaller than that observed in studies of adults. There are relatively little data on VOC exposures in children, so these results are useful for estimating the central tendency and distribution of VOC exposures in locations where children spend a majority of their time.

Background: During the design phase of a clinical trial, sample size estimates should take into account medical screening criteria, the 'healthy volunteer' effect, consequences of run-in phases, and secular trends in the event rate of interest. All of these have been shown to relate to subsequent event rates, and hence trial power to detect intervention effects. The Multiple Risk Factor Intervention Trial (MRFIT) used three successive screenings of 361 662 men to enroll 12 866; observed coronary heart disease (CHD) mortality after a mean of 6.9 years was substantially lower than projected during design. We explore factors which may have contributed to these mortality differences and whether they persisted throughout follow-up., Methods: Proportional hazards models were used to compare 25-year mortality according to trial eligibility, self-exclusions, medical exclusions, and participation., Results: After adjustment for baseline risk factors and age, there was higher mortality among men excluded for presence of disease [hazard ratio (HR) 1.52, 95% confidence interval (CI) 1.43-1.61, for total; HR 1.92, CI 1.75-2.11, for CHD] compared to those not excluded which persisted throughout follow-up. Volunteers had lower total (HR 0.82, CI 0.76-0.87) and CHD (HR 0.79, CI 0.70-0.88) mortality than those discontinuing participation. Men excluded with characteristics deemed likely to interfere with adherence had higher total (HR 1.19, CI 1.07-1.33) and non-cardiovascular disease (CVD) (HR 1.32, CI 1.14-1.53) mortality but no higher CVD (HR 1.04, CI 0.88-1.23) or CHD (HR 0.98, CI 0.80-1.20) mortality compared to those not excluded. Differences in mortality were stronger during the first five years, but declined only slightly over 25 years., Conclusions: 25-year mortality was significantly higher for non-volunteers and exclusions. Differences between observed and predicted six-year total mortality for trial participants were largely attributable to volunteers and exclusions, but there were additional differences for CHD mortality, which were likely due to downward secular trends. These results emphasize the importance of anticipating these factors during clinical trial design, even for trials of short duration.

Background: Patients who report use of herbs to their physicians may not be able to accurately describe the ingredients or recommended dosage because the products for the same herb may differ. The purpose of this study was to describe variations in label information of products for each of the 10 most commonly purchased herbs., Methods: Products for each of 10 herbs were surveyed in a convenience sample of 20 retail stores in a large metropolitan area. Herbs were those with the greatest sales dollars in 1998: echinacea, St John's wort, Ginkgo biloba, garlic, saw palmetto, ginseng, goldenseal, aloe, Siberian ginseng, and valerian., Results: Each herb had a large range in label ingredients and recommended daily dose (RDD) across available products. Strengths were not directly comparable because of ingredient variability. Among 880 products, 43% were consistent with a benchmark in ingredients and RDD, 20% in ingredients only, and 37% were either not consistent or label information was insufficient. Price per RDD was a significant predictor of consistency with the benchmark, but store type was not., Conclusions: Persons self-medicating with an herb may be ingesting ingredients substantially different from that recommended by a benchmark, both in quantity and content. Higher price per label RDD was the best predictor of consistency with a benchmark. This study demonstrates that health providers and consumers need to closely examine label ingredients of presumably the same or similar herbal products.

Background: Forced expiratory volume in 1 second (FEV(1)) may be useful for identifying smokers at higher risk of lung cancer. We examined the association of FEV(1) with lung cancer mortality (LCM) among cigarette smokers in the Multiple Risk Factor Intervention Trial (MRFIT)., Methods: In all, 6613 MRFIT baseline smokers alive at trial end in 1982 had acceptable FEV(1) measures and complete smoking history; men were classified as during-trial long-term quitters (N = 1292), intermittent quitters (1961), and never quitters (3360). Proportional hazards models for LCM were fit with quintiles of average FEV(1), adjusted for age, height, race, smoking history, and other risk factors., Results: For long-term, intermittent, and never quitters respectively, mean baseline cigarettes/ day was 28, 32, and 35; trial-averaged FEV(1) was 3201, 3146, and 3082 ml; and average decline in FEV(1) was -46.0, -54.6, and -62.5 ml/year. With median post-trial mortality follow-up of 18 years, there were 363 lung cancer deaths. Age-adjusted LCM rates varied across FEV(1) quintiles from 50 (lowest quintile) to 11 (highest quintile), 58 to 11, and 76 to 20, per 10 000 person-years, for long-term quitters, intermittent quitters, and never quitters, respectively. Multivariate adjusted hazard ratios for 100 ml higher FEV(1) were 0.92 [P = 0.004], 0.95 [P = 0.003], and 0.95 [P < 0.0001] respectively., Conclusions: These results demonstrate the strong predictive value of FEV(1) for lung cancer among cigarette smokers independent of smoking history; results did not differ by during-trial quit status. FEV(1) may be a biological marker for smoking dose or it may be that genetic susceptibilities to both decreased FEV(1) and lung cancer are associated.

Background: It remains unclear whether hypertriglyceridemia is an independent risk factor for coronary heart disease (CHD), and whether fasting and nonfasting triglyceride (TG) levels are equally predictive., Methods: A total of 2809 (of 12 866) men randomized during 1973 through 1975 into the Multiple Risk Factor Intervention Trial with fasting and nonfasting TG levels measured at baseline were followed up for CHD incidence and death. Proportional hazards regression models were used to assess associations of fasting and nonfasting TG levels with CHD., Results: Average fasting and nonfasting TG levels were 187 and 284 mg/dL (2.11 and 3.21 mmol/L), respectively. Prevalence of hypertriglyceridemia (200 mg/dL [2.26 mmol/L] or more) was 31% for fasting and 61% for nonfasting. There were 175 nonfatal or fatal CHD events during 8 years and 328 CHD deaths during 25 years. Compared with TG levels less than 200 mg/dL, risk factor-adjusted hazard ratios for CHD mortality for hypertriglyceridemia were 1.24 (P =.09) for fasting and 1.26 (P =.07) for nonfasting. For nonfatal or fatal CHD, fasting and nonfasting TG levels were similarly predictive with hazard ratios of 1.64 (P =.004) for fasting and 1.46 (P =.03) for nonfasting. These associations for fasting TG levels were assessed to be underestimated by 56% because of regression dilution bias, with attenuation likely greater for nonfasting TG levels., Conclusions: Greater ease of obtaining nonfasting than fasting measurements, greater prevalence of hypertriglyceridemia with nonfasting than fasting values, and similarly increased risk with each indicate that nonfasting TG levels may be more useful than fasting ones for risk stratification.

Objective: To report long-term risks for total, cardiovascular disease (CVD), and coronary heart disease (CHD) mortality associated with incident diabetes (using current diagnostic criteria) and with incident nonfatal CVD (NF-CVD)., Research Design and Methods: A total of 11645 participants without diabetes or CVD at baseline from the Multiple Risk Factor Intervention Trial who survived to the end of the trial were grouped by during-trial incident diabetes and/or NF-CVD events: neither diabetes nor NF-CVD, diabetes only, NF-CVD only, or both diabetes and NF-CVD. Incident diabetes was defined by use of hypoglycemic agents or fasting glucose >or=126 mg/dl at any time over the 6 trial years. Proportional hazards models tested group differences in mortality over 18 post-trial years., Results: Among 3859 total deaths were 1846 from CVD and 1277 from CHD, with death rates per 10000 person-years of 203, 97, and 67, respectively. Multivariate-adjusted hazard ratios (HRs) for total mortality were 2.75 (P < 0.0001) for those with NF-CVD and diabetes both, 1.92 (P < 0.0001) for those with NF-CVD only, and 1.49 (P < 0.0001) for those with diabetes only, relative to neither diabetes nor NF-CVD. NF-CVD was associated with a higher hazard of death than diabetes for total (HR 1.29, P = 0.0004), CVD (HR 1.76, P < 0.0001), and CHD (HR 1.88, P < 0.0001) mortality. Only the subgroup of participants on hypoglycemic agents showed an equivalent risk of total mortality relative to participants with NF-CVD (HR 0.93, P = 0.54)., Conclusions: Current diabetes diagnostic criteria conferred significantly increased total, CVD, and CHD mortality risks independent of the impact of NF-CVD. NF-CVD was more strongly predictive of mortality.

The ability of questionnaires to predict children's exposure to pesticides was examined as part of the Minnesota Children's Pesticide Exposure Study (MNCPES). The MNCPES focused on a probability sample of 102 children between the ages of 3 and 13 years living in either urban (Minneapolis and St. Paul, MN) or nonurban (Rice and Goodhue Counties in Minnesota) households. Samples were collected in a variety of relevant media (air, food, beverages, tap water, house dust, soil, urine), and chemical analyses emphasized three organophosphate insecticides (chlorpyrifos, diazinon, malathion) and a herbicide (atrazine). Results indicate that the residential pesticide-use questions and overall screening approach used in the MNCPES were ineffective for identifying and oversampling children/households with higher levels of individual target pesticides.

Because most cooperative societies are despotic, it has been difficult to test models of egalitarianism. Female African lions demonstrate a unique form of plural breeding in which companions consistently produce similar numbers of surviving offspring. Consistent with theoretical predictions from models of reproductive skew, female lions are unable to control each other's reproduction because of high costs of fighting and low access to each other's newborn cubs. A female also lacks incentives to reduce her companions' reproduction, because her own survival and reproduction depend on group territoriality and synchronous breeding. Consequently, female relationships are highly symmetrical, and female lions are "free agents" who only contribute to communal care when they have cubs of their own.

The Minnesota Children's Pesticide Exposure Study is a probability-based sample of 102 children 3-13 years old who were monitored for commonly used pesticides. During the summer of 1997, first-morning-void urine samples (1-3 per child) were obtained for 88% of study children and analyzed for metabolites of insecticides and herbicides: carbamates and related compounds (1-NAP), atrazine (AM), malathion (MDA), and chlorpyrifos and related compounds (TCPy). TCPy was present in 93% of the samples, whereas 1-NAP, MDA, and AM were detected in 45%, 37%, and 2% of samples, respectively. Measured intrachild means ranged from 1.4 microg/L for MDA to 9.2 microg/L for TCPy, and there was considerable intrachild variability. For children providing three urine samples, geometric mean TCPy levels were greater than the detection limit in 98% of the samples, and nearly half the children had geometric mean 1-NAP and MDA levels greater than the detection limit. Interchild variability was significantly greater than intrachild variability for 1-NAP (p = 0.0037) and TCPy (p < 0.0001). The four metabolites measured were not correlated within urine samples, and children's metabolite levels did not vary systematically by sex, age, race, household income, or putative household pesticide use. On a log scale, mean TCPy levels were significantly higher in urban than in nonurban children (7.2 vs. 4.7 microg/L; p = 0.036). Weighted population mean concentrations were 3.9 [standard error (SE) = 0.7; 95% confidence interval (CI), 2.5, 5.3] microg/L for 1-NAP, 1.7 (SE = 0.3; 95% CI, 1.1, 2.3) microg/L for MDA, and 9.6 (SE = 0.9; 95% CI, 7.8, 11) microg/L for TCPy. The weighted population results estimate the overall mean and variability of metabolite levels for more than 84,000 children in the census tracts sampled. Levels of 1-NAP were lower than reported adult reference range concentrations, whereas TCPy concentrations were substantially higher. Concentrations of MDA were detected more frequently and found at higher levels in children than in a recent nonprobability-based sample of adults. Overall, Minnesota children's TCPy and MDA levels were higher than in recent population-based studies of adults in the United States, but the relative magnitude of intraindividual variability was similar for adults and children.

Birth sex ratios of baboons in Gombe National Park, Tanzania, show an overall male bias of ca. 20%, but there is no obvious explanation for this trend. Individual females did not alter their sex ratios according to persistent levels of local resource competition. Sex ratios showed an unexpected relationship between age and rank: subordinate females had more sons when they were young; dominant females had more sons when they were old. The sex ratio of low-ranking females also varied with the severity of environmental conditions during pregnancy. Our findings suggest that mammalian sex ratios might be the product of more complex processes than is generally recognized or that sex-determining mechanisms impose sufficient constraints to prevent adaptive variation in all contexts.

The marked increase in popularity of Bayesian methods in statistical practice over the last decade owes much to the simultaneous development of Markov chain Monte Carlo (MCMC) methods for the evaluation of requisite posterior distributions. However, along with this increase in computing power has come the temptation to fit models larger than the data can readily support, meaning that often the propriety of the posterior distributions for certain parameters depends on the propriety of the associated prior distributions. An important example arises in spatial modelling, wherein separate random effects for capturing unstructured heterogeneity and spatial clustering are of substantive interest, even though only their sum is well identified by the data. Increasing the informative content of the associated prior distributions offers an obvious remedy, but one that hampers parameter interpretability and may also significantly slow the convergence of the MCMC algorithm. In this paper we investigate the relationship among identifiability, Bayesian learning and MCMC convergence rates for a common class of spatial models, in order to provide guidance for prior selection and algorithm tuning. We are able to elucidate the key issues with relatively simple examples, and also illustrate the varying impacts of covariates, outliers and algorithm starting values on the resulting algorithms and posterior distributions., (Copyright 2000 John Wiley & Sons, Ltd.)

Objectives: To test the hypothesis that Kaposi's sarcoma (KS) protects against four central nervous system (CNS) diseases in HIV-1-infected individuals., Study Population and Design: The study population of 9404 subjects included participants in Terry Beirn Community Programs for Clinical Research on AIDS (CPCRA) protocols who were enrolled between September 1990 and September 1998. This was an observational study., Methods: Proportional hazards regression was used to estimate adjusted relative risks for predictors of four central nervous system diseases. Covariates included occurrence of Kaposi's sarcoma, occurrence of other opportunistic infections or malignancies, baseline CD4+ count, and other baseline characteristics., Results: Among the 5944 participants without progression to AIDS at entry, 451 developed a CNS disease. The adjusted relative risk of any CNS disease for those who developed Kaposi's sarcoma versus those who did not develop any AIDS-defining event was 1.41 [95% confidence interval (CI), 0.98-2.03; P = 0.06]. In contrast, the adjusted relative risk of any CNS disease for those with Kaposi's sarcoma versus those with some other non-Kaposi's sarcoma AIDS-defining event was 0.37 (95% CI, 0.24-0.57; P < 0.0001). Among the 3460 participants with progression to AIDS at entry, the adjusted relative risk of any CNS disease for those with Kaposi's sarcoma versus those with some other non-Kaposi's sarcoma AIDS-defining event was 0.71 (95% CI, 0.40-1.25; P = 0.23)., Conclusions: Our analyses indicate that the risk of CNS disease associated with Kaposi's sarcoma depends strongly on the reference or control group chosen. When compared to individuals with other non-Kaposi's sarcoma AIDS-defining diseases, Kaposi's sarcoma is associated with a lower risk of CNS disease in HIV-1 positive individuals. However, when compared to individuals with no AIDS-defining disease or with a similarly mild AIDS-defining disease such as invasive candidiasis, Kaposi's sarcoma is associated with an equivalent risk of CNS disease.