American Journal of Epidemiology

This study investigated the relation between weekly levels of influenza activity and the risk of acute influenza-like illness episodes among 8,323 healthy pregnant and postpartum women enrolled in a Puget Sound region, Washington, health maintenance organization, Group Health Cooperative, between June 1991 and December 1997. The authors classified weeks between October and May for isolate activity level based on surveillance data for influenza, respiratory syncytial virus, parainfluenza, and adenovirus infection. Influenza-like illness episodes were identified from medical encounters assigned a diagnostic code consistent with a symptomatic influenza infection. The authors compared the occurrence of influenza-like illness episodes within each pregnancy stage for periods with varying levels of influenza isolate detection in the community. Repeated-measures logistic regression methods accounted for time-dependent factors. The adjusted strength of association between influenza exposure and influenza-like illness episodes increased as the pregnancy stage progressed (first trimester odds ratio = 1.12, 95% confidence interval: 0.79, 1.59; second trimester odds ratio = 1.30, 95% confidence interval: 0.97, 1.73; third trimester odds ratio = 1.84, 95% confidence interval: 1.31, 2.59; postpartum period odds ratio = 2.28, 95% confidence interval: 1.42, 3.68). Pregnancy stage modified the association between influenza activity and influenza-like illness episodes. Findings estimate that 20–43 pregnant/postpartum women would need to be vaccinated with an 80% effective vaccine to prevent one influenza-like illness episode.

Consistent evidence has shown a positive association between particulate matter with an aerodiameter of less than or equal to 10 µm (PM10) and daily mortality. Less is known about the modification of this association by factors measured at the individual level. The authors examined this question in a case-crossover study of 20 US cities. Mortality events (1.9 million) were obtained for nonaccidental, respiratory, heart disease, and stroke mortality between 1989 and 2000. PM10 concentrations were obtained from the US Environmental Protection Agency. The authors examined the modification of the PM10–mortality association by sociodemographics, location of death, season, and secondary diagnoses. They found different patterns of PM10–mortality associations by gender and age but no differences by race. The level of education was inversely related to the risk of mortality associated with PM10. PM10-related, out-of-hospital deaths were more likely than were in-hospital deaths, as were those occurring during spring/fall versus summer/winter. A secondary diagnosis of diabetes modified the effect of PM10 for respiratory and stroke mortality. Pneumonia was a positive effect modifier for deaths from all causes and stroke, while secondary stroke modified the effects for all-cause and respiratory deaths. The findings suggest that more attention must be paid to population characteristics to identify greater likelihood of exposures and susceptibility and, as a result, to improve policy making for air pollution standards.

Using data on 4,813 children from the ALSPAC cohort in Bristol, United Kingdom, recontacted in 1998–1999, the authors investigated whether intrauterine growth restriction (indexed by birth weight and length) was associated with behavioral problems at age 7 years. Childhood behavioral problems were measured by using a brief behavioral screening questionnaire (the Strengths and Difficulties Questionnaire (parental completion)). For term singleton infants, a one standard deviation increase in birth weight was associated with an 11% reduction in the odds of behavioral problems at age 81 months. After adjustment for confounders and birth length, this association was no longer seen. The association with birth length remained after adjustment for confounders. A one standard deviation increase in birth length was associated with a 14% decrease in the odds of being in the top tertile of total behavioral difficulties at age 81 months (odds ratio = 0.86, 95% confidence interval: 0.79, 0.95) and was similarly associated with hyperactivity and conduct problems. Evidence was weak for an association between birth length and behavioral problems earlier in childhood. In summary, there was a weak association between intrauterine growth restriction, indexed by birth length (rather than weight), and childhood behavioral problems. Future work should focus on elucidating the biologic mechanisms that lead to variations in birth length and underlie this association.

The purpose of this analysis was to examine the efficacy of prophylactic hematopoietic colony-stimulating factors (CSFs) in pediatric cancer and to describe how a Bayesian meta-analysis can be conducted and then modified to incorporate information not readily included in a frequentist meta-analysis. Three Bayesian models were developed. The simplest model used the same data as a published frequentist meta-analysis. The second model included data that could not easily be incorporated into the frequentist meta-analysis, including data from different courses of chemotherapy and continuous outcomes that did not report variance estimates. The third model examined the effect of CSF type (granulocyte CSF vs. granulocyte-macrophage CSF). Compared with the frequentist model, the Bayesian model with the most data suggested a greater benefit of CSFs, with a 3.2-day reduction in duration of parenteral antibiotics (95% credible interval: –7.1, 0.7) in the expanded Bayesian model compared with a 0.8-day (95% confidence interval: –2.3, 0.7) reduction in the frequentist model. Bayesian meta-analysis also suggested that, compared with granulocyte-macrophage CSF, granulocyte CSF was associated with a 4.8-day decrease in the duration of parenteral antibiotics. Bayesian meta-analysis can readily include information not easily incorporated in a frequentist meta-analysis. Some treatment effect estimates were larger by a clinically important amount when additional data contributed to the pooled estimate.

Hepatoblastoma is a rare embryonal tumor with unknown etiology. The authors conducted a case-cohort study using public health surveillance data sets to examine perinatal risk factors for hepatoblastoma. Hepatoblastoma cases (n = 58) diagnosed between 1985 and 2001 were identified from the New York State Cancer Registry and were matched to electronic birth records for 1985&ndash;2001 from New York State, excluding New York City. Controls (n = 6,056) were selected from the birth cohorts for the same years. Having a birth weight less than 1,000 g was associated with a strongly increased risk of hepatoblastoma (relative risk (RR) = 56.9, 95% confidence interval (CI): 24.0, 130.7). After adjustment for birth weight, a moderately increased risk of hepatoblastoma was found for younger maternal age (<20 years vs. 20&ndash;29 years: RR = 2.5, 95% CI: 1.0, 5.5), presumptive use of infertility treatment (RR = 9.2, 95% CI: 2.1, 31.5), maternal smoking (RR = 2.1, 95% CI: 1.0, 4.2), and higher maternal prepregnancy body mass index (body mass index of 25&ndash;29 vs. 20&ndash;24: RR = 2.9, 95% CI: 1.2, 7.6).

Transforming growth factor alpha (TGFA) is a well-characterized mammalian growth factor. Since the first report of an association between DNA sequence variants at the TGFA genetic locus and nonsyndromic oral clefts, 47 studies have been carried out, producing conflicting results. In this review, the author synthesizes findings from published reports on the association between the TGFA gene and clefting in humans. Bias, lack of statistical power, and genuine population diversity can explain the diverse results. In the aggregate, TGFA is probably a genetic modifier of clefting in humans, which is consistent with the oligogenic model suggested for nonsyndromic oral clefts.

The replication of important findings by multiple independent investigators is fundamental to the accumulation of scientific evidence. Researchers in the biologic and physical sciences expect results to be replicated by independent data, analytical methods, laboratories, and instruments. Epidemiologic studies are commonly used to quantify small health effects of important, but subtle, risk factors, and replication is of critical importance where results can inform substantial policy decisions. However, because of the time, expense, and opportunism of many current epidemiologic studies, it is often impossible to fully replicate their findings. An attainable minimum standard is "reproducibility," which calls for data sets and software to be made available for verifying published findings and conducting alternative analyses. The authors outline a standard for reproducibility and evaluate the reproducibility of current epidemiologic research. They also propose methods for reproducible research and implement them by use of a case study in air pollution and health.

(No abstract is available for this citation)

(No abstract is available for this citation)

(No abstract is available for this citation)

(No abstract is available for this citation)

Vegetables and fruits are rich in carotenoids, a group of compounds thought to protect against cancer. Studies of diet-disease associations need valid and reliable instruments for measuring dietary intake. The authors present a measurement error model to estimate the validity (defined as correlation between self-reported intake and "true" intake), systematic error, and reliability of two self-report dietary assessment methods. Carotenoid exposure is measured by repeated 24-hour recalls, a food frequency questionnaire (FFQ), and a plasma marker. The model is applied to 1,013 participants assigned between 1995 and 2000 to the nonintervention arm of the Women's Healthy Eating and Living Study, a randomized trial assessing the impact of a low-fat, high-vegetable/fruit/fiber diet on preventing new breast cancer events. Diagnostics including graphs are used to assess the goodness of fit. The validity of the instruments was 0.44 for the 24-hour recalls and 0.39 for the FFQ. Systematic error accounted for over 22% and 50% of measurement error variance for the 24-hour recalls and FFQ, respectively. The use of either self-report method alone in diet-disease studies could lead to substantial bias and error. Multiple methods of dietary assessment may provide more accurate estimates of true dietary intake.

(No abstract is available for this citation)

The recent US decline in sudden infant death syndrome (SIDS) rates may be explained by a shift in how these deaths are classified or reported. To examine this hypothesis, the authors compared cause-specific mortality rates for SIDS, other sudden, unexpected infant deaths, and cause unknown/unspecified, and they evaluated trends in the age and month of death for these causes using 1989&ndash;2001 US linked birth/death certificate data. Reported deaths in state and national data were compared to assess underreporting or overreporting. SIDS rates declined significantly from 1989&ndash;1991 to 1995&ndash;1998, while deaths reported as cause unknown/unspecified and other sudden, unexpected infant deaths, such as accidental suffocation and strangulation in bed (ASSB), remained stable. From 1999&ndash;2001, the decline in SIDS rates was offset by increasing rates of cause unknown/unspecified and ASSB. Changes in the cause-specific age at death and month of death distributions suggest that cases once reported as SIDS are now being reported as ASSB and cause unknown/unspecified. Most of the decline in SIDS rates since 1999 is likely due to increased reporting of cause unknown/unspecified and ASSB. Standardizing data collection at death scenes and improving the reporting of cause of death on death certificates should improve national vital records data and enhance prevention efforts.

Residential proximity to applications of agricultural pesticides may be an important source of exposure to agents that have been classified as developmental toxins. Data on two case-control study populations of infants with neural tube defects (NTDs) and nonmalformed controls delivered in California between 1987 and 1991 were pooled to investigate whether maternal residential proximity to applications of specific pesticides or physicochemical groups of pesticides during early gestation increases the risk of these malformations. Maternal residential proximity within 1,000 m of pesticide applications was ascertained by linking mothers' addresses with agricultural pesticide use reports and crop maps. Odds ratios were computed by using conventional single- and multiple-pesticide and hierarchical multiple-pesticide logistic regression. In single-pesticide models, several pesticides were associated with NTDs after adjustment for study population, maternal ethnicity, educational level, cigarette smoking, and vitamin use. In a hierarchical multiple-pesticide model, effect estimates for only benomyl and methomyl suggested a possible association. Elevated risks of NTDs and anencephaly or spina bifida subtypes were also associated with exposures to chemicals classified as amide, benzimidazole, methyl carbamate, or organophosphorus pesticides and with increasing numbers of pesticides. These results suggest that ambient exposure to certain categories of agricultural pesticides may increase the risk of NTDs.

Understanding the course of back pain is important for clinicians and researchers, but analyses of longitudinal data from multiple time points are lacking. A prospective cohort study of consecutive back pain consulters from five general practices in the United Kingdom was carried out between 2001 and 2003 to identify groups defined by their pain pathways. Patients were sent monthly questionnaires for a year. Longitudinal latent class analysis was performed by using pain intensity scores for 342 consulters. Analysis yielded four clusters representing different pathways of back pain. Cluster 1 ("persistent mild"; n = 122) patients had stable, low levels of pain. Patients in cluster 2 ("recovering"; n = 104) started with mild pain, progressing quickly to no pain. Cluster 3 ("severe chronic"; n = 71) patients had permanently high pain. For patients in cluster 4 ("fluctuating"; n = 45), pain varied between mild and high levels. Distinctive patterns for each cluster were maintained throughout follow-up. Clusters showed statistically significant differences in disability, psychological status, and work absence (p < 0.001). This is the first time, to the authors' knowledge, that latent class analysis has been applied to longitudinal data on back pain patients. Identification of four distinct groups of patients improves understanding of the course of back pain and may provide a basis of classification for intervention.

Age at menopause has implications for fertility and risk of hormonally related chronic diseases. Some pesticides disrupt reproductive hormones or are toxic to the ovary, but little is known about the association between pesticide exposure and timing of menopause. Cox proportional hazards modeling was used to examine the association between use of pesticides and age at menopause among 8,038 women living and working on farms in Iowa and North Carolina. Premenopausal women aged 35&ndash;55 years were followed from enrollment (1993&ndash;1997) to the date of their last menstrual period, or their follow-up interview (1999&ndash;2003) if still premenopausal. Women who experienced surgical menopause were censored at the date of surgery. Approximately 62% of the women reported ever mixing or applying pesticides; women who had never used pesticides were the comparison group for all analyses. After control for age, smoking status, and past use of oral contraceptives, the median time to menopause increased by approximately 3 months for women who used pesticides (hazard ratio = 0.87, 95% confidence interval: 0.78, 0.97) and by approximately 5 months for women who used hormonally active pesticides (hazard ratio = 0.77, 95% confidence interval: 0.65, 0.92). Pesticide use may be associated with a later age at menopause.

The authors prospectively investigated whether working rotating night shifts was associated with the risk of Parkinson's disease among 84,794 female nurses who reported years of night shift work in 1988 (the US Nurses' Health Study). After 975,912 person-years of follow-up (1988&ndash;2000), 181 incident Parkinson's disease cases were documented. Compared with nurses who never worked rotating night shifts, those with 15 years or more of night shift work had a 50% lower risk of Parkinson's disease after adjustment for age and smoking (95% confidence interval: 0.26, 0.97; ptrend = 0.01). Sleep duration was positively associated with Parkinson's disease risk: The relative risk was 1.84 (95% confidence interval: 0.99, 3.42) when comparing nurses who reported 9 or more hours of sleep per day with those who slept 6 hours or less (ptrend = 0.005). These data suggest that working night shifts may be protective against Parkinson's disease or that low tolerance for night shift work is an early marker of Parkinson's disease. Conversely, habitual longer sleep duration may be an earlier marker of Parkinson's disease. Because of the novelty and the exploratory nature of these findings, confirmation is needed.

Factors that influence circulating sex hormones, such as physical activity, have been proposed to influence ovarian cancer risk; however, results from previous epidemiologic studies have been inconsistent. The authors examined the association among physical activity, sedentary behavior, and ovarian cancer risk in the American Cancer Society Cancer Prevention Study II Nutrition Cohort, a prospective study of cancer incidence and mortality, using information obtained at baseline in 1992. From 1992 to 2001, 314 incident ovarian cancer cases were identified among 59,695 postmenopausal women who were cancer free at enrollment. Cox proportional hazards modeling was used to compute hazard rate ratios while adjusting for potential confounders. No overall association was observed between measures of past physical activity or with recreational physical activity at baseline and risk of ovarian cancer in this study (for the highest category of physical activity compared with none: hazard rate ratio = 0.73, 95% confidence interval: 0.40, 1.34). However, a prolonged duration of sedentary behavior was associated with an increased risk (for &ge;6 vs. <3 hours per day: hazard rate ratio = 1.55, 95% confidence interval: 1.08, 2.22; ptrend = 0.01). Results from this study suggest that high levels of sedentary behavior may increase the risk of ovarian cancer, but they do not support a major impact of light and moderate physical activity on ovarian cancer risk.

The authors evaluated associations between occupational exposures in the textile industry and the risks of esophageal cancer and stomach cancer. The authors conducted a case-cohort study nested in a cohort of female textile workers in Shanghai, China. One hundred and two workers with incident esophageal cancer and 646 workers with incident stomach cancer diagnosed between 1989 and 1998 were compared with an age-stratified reference subcohort (n = 3,188). Work histories were ascertained for all study subjects from factory personnel records or interviews. Exposures were reconstructed for chemicals and dusts by linking work history data with a job-exposure matrix developed for the Shanghai textile industry. Hazard ratios and 95 percent confidence intervals were calculated with Cox proportional hazards modeling adapted for the case-cohort design. Risk of esophageal cancer was associated with long-term (&ge;10 years) exposure to silica dust (hazard ratio = 15.8, 95% confidence interval: 3.5, 70.6) and metals (hazard ratio = 3.7, 95% confidence interval: 1.9, 7.1). Cumulative exposure to endotoxin, a contaminant of cotton dust, was inversely related to risks of both esophageal cancer (p-trend = 0.01) and stomach cancer (p-trend < 0.001) when exposures were lagged 20 years. Endotoxin has not been previously reported to be a protective factor for either stomach cancer or esophageal cancer and therefore warrants further study.

Homocysteine levels are associated with peripheral arterial disease (PAD) in observational studies. Lead and cadmium are risk factors for PAD that affect thiol metabolism, and they may partly explain the association of homocysteine with PAD. To evaluate the roles of lead and cadmium exposure in confounding the association between homocysteine and PAD, the authors performed a cross-sectional study among 4,447 persons aged &ge;40 years who participated in the 1999&ndash;2002 National Health and Nutrition Examination Survey (NHANES). PAD was defined as an ankle-brachial blood pressure index less than 0.90 in at least one leg. After adjustment for sociodemographic variables, the odds ratio for PAD in the highest quintile of homocysteine compared with the lowest was 1.92 (ptrend = 0.004). Adjusting for blood lead and cadmium levels reduced this odds ratio to 1.37 (ptrend = 0.13), and further adjusting for estimated glomerular filtration rate and smoking reduced it to 0.89 (ptrend = 0.87). Adjustment for other risk factors did not affect this association. In the general population, the association of homocysteine level with PAD can be completely explained by confounding due to smoking, increased blood lead and cadmium levels, and impaired renal function. The association of lead and cadmium with PAD risk deserves further investigation.

The lignan enterolactone produced by the intestinal microflora from dietary precursors has been hypothesized to protect against coronary heart disease. The present study examined the association between serum enterolactone concentration and the risk of coronary heart disease. A prospective case-cohort study was conducted among male smokers randomized to receive a placebo supplement in the Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study (1986&ndash;1999). Serum enterolactone concentrations were measured by the gas chromatography-mass spectrometry method in serum collected at trial baseline from 340 men diagnosed with nonfatal myocardial infarction (n = 205) or coronary death (n = 135) during follow-up and from the randomly selected subcohort of 420 subjects. The classic risk factors-adjusted rate ratios for all coronary heart disease events in increasing quintiles of enterolactone were 1.00 (referent), 0.85 (95% confidence interval (CI): 0.51, 1.43), 0.59 (95% CI: 0.35, 1.00), 0.69 (95% CI: 0.40, 1.16), and 0.63 (95% CI: 0.33, 1.11), and the ptrend was 0.07. For the highest versus the lowest quintile of enterolactone, the rate ratios for nonfatal myocardial infarction and coronary death were 0.67 (95% CI: 0.37, 1.23; ptrend = 0.10) and 0.57 (95% CI: 0.26, 1.25; ptrend = 0.18), respectively. In conclusion, only weak support for the association between serum enterolactone concentration and coronary heart disease was found.

Despite the documented antioxidant and chemopreventive properties of selenium, studies of selenium intake and supplementation and cardiovascular disease have yielded inconsistent findings. The authors examined the effect of selenium supplementation (200 &micro;g daily) on cardiovascular disease incidence and mortality through the entire blinded phase of the Nutritional Prevention of Cancer Trial (1983&ndash;1996) among participants who were free of cardiovascular disease at baseline (randomized to selenium: n = 504; randomized to placebo: n = 500). Selenium supplementation was not significantly associated with any of the cardiovascular disease endpoints during 7.6 years of follow-up (all cardiovascular disease: hazard ratio (HR) = 1.03, 95% confidence interval (CI): 0.78, 1.37; myocardial infarction: HR = 0.94, 95% CI: 0.61, 1.44; stroke: HR = 1.02, 95% CI: 0.63, 1.65; all cardiovascular disease mortality: HR = 1.22, 95% CI: 0.76, 1.95). The lack of significant association with cardiovascular disease endpoints was also confirmed when analyses were further stratified by tertiles of baseline plasma selenium concentrations. These findings indicate no overall effect of selenium supplementation on the primary prevention of cardiovascular disease in this population.

(No abstract is available for this citation)

The use of biomarkers is of ever-increasing importance in clinical diagnosis of disease. In practice, a cutpoint is required for dichotomizing naturally continuous biomarker levels to distinguish persons at risk of disease from those who are not. Two methods commonly used for establishing the "optimal" cutpoint are the point on the receiver operating characteristic curve closest to (0,1) and the Youden index, J. Both have sound intuitive interpretations&mdash;the point closest to perfect differentiation and the point farthest from none, respectively&mdash;and are generalizable to weighted sensitivity and specificity. Under the same weighting of sensitivity and specificity, these two methods identify the same cutpoint as "optimal" in certain situations but different cutpoints in others. In this paper, the authors examine situations in which the two criteria agree or disagree and show that J is the only "optimal" cutpoint for given weighting with respect to overall misclassification rates. A data-driven example is used to clarify and demonstrate the magnitude of the differences. The authors also demonstrate a slight alteration in the (0,1) criterion that retains its intuitive meaning while resulting in consistent agreement with J. In conclusion, the authors urge that great care be taken when establishing a biomarker cutpoint for clinical use.

Data collected in a routine clinical setting are frequently used to compare antiretroviral treatments for human immunodeficiency virus (HIV). Differences in the frequency of measurement of HIV RNA levels and CD4-positive T-lymphocyte cell counts introduce a possible source of bias into estimates of the difference in effectiveness between treatments. The authors investigated the size of this bias when survival analysis methods are used to compare the initial efficacy of antiretroviral regimens. Data sets of clinical markers were simulated by use of differential equations that model the interaction between HIV and human T-cells. Cox proportional hazards and parametric models were fitted to the simulated data sets to evaluate the bias and coverage of 95% confidence intervals for the difference between regimens. The authors' results demonstrate that differences in the frequency of follow-up can substantially bias estimated treatment differences if methods do not correctly account for the intervals between measurements and if the statistical model chosen does not fit the data well. Analyses using methods applicable to interval-censored data reduce the bias. In the Athena cohort of HIV-infected individuals in the Netherlands from 1999 to 2003, there are differences in measurement frequency between current regimens that are of sufficient magnitude to conclude incorrectly that some regimens are more effective than others.

Between 2001 and 2003, the authors studied pregnancy outcomes and infant mortality among 202 married women in West Bengal, India. Reproductive histories were ascertained using structured interviews. Arsenic exposure during each pregnancy, including all water sources used, was assessed; this involved measurements from 409 wells. Odds ratios for spontaneous abortion, stillbirth, neonatal mortality, and infant mortality were estimated with logistic regression based on the method of generalized estimating equations. Exposure to high concentrations of arsenic (&ge;200 &micro;g/liter) during pregnancy was associated with a sixfold increased risk of stillbirth after adjustment for potential confounders (odds ratio (OR) = 6.07, 95% confidence interval (CI): 1.54, 24.0; p = 0.01). Arsenic-related skin lesions were found in 12 women who had a substantially increased risk of stillbirth (OR = 13.1, 95% CI: 3.17, 54.0; p = 0.002). The odds ratio for neonatal death was 2.81 (95% CI: 0.73, 10.8). No association was found between arsenic exposure and spontaneous abortion (OR = 1.01, 95% CI: 0.38, 2.70) or overall infant mortality (OR = 1.33, 95% CI: 0.43, 4.04). This study adds to the limited evidence that exposure to high concentrations of arsenic during pregnancy increases the risk of stillbirth. However, there was no indication of the increased rates of spontaneous abortion and overall infant mortality that have been reported in some studies.

In a 32-year follow-up study, the authors analyze how social circumstances during early life, childhood social participation, and school performance affect the risk of being admitted to a hospital or dying from a diagnosis closely related to drug or alcohol abuse in young adulthood. A total of 11,376 Danish males born in 1953, for whom data from birth certificates and conscription board examinations had been traced, were followed until 2002 through linkage to the Danish Psychiatric, National Patient, and Cause of Death registries. At age 12 years, 7,877 subjects completed a questionnaire on social participation and school performance. During follow-up, 12 percent of these were given a diagnosis indicating drug or alcohol abuse. Having a single mother and a working-class father were each associated with an increased risk of drug or alcohol abuse in adult life. At age 12 years, those who disliked school, scored low on a school test, or preferred to visit a youth club during leisure time showed a greater risk of adult substance abuse. These associations were slightly attenuated when adjusted for educational status at conscription. Deprived social circumstances during childhood, poor school performance in early adolescence, and attending a youth club seemed to be independent markers of substance abuse in adult life.

The authors evaluated the associations of ginseng use as a complementary therapy with survival and quality of life (QOL) in a cohort of 1,455 breast cancer patients who were recruited to the Shanghai Breast Cancer Study between August 1996 and March 1998 in Shanghai, China. Patients were followed through December 2002. Information on ginseng use before cancer diagnosis was collected at baseline recruitment and was linked to survival. Survivors' ginseng use after cancer diagnosis was obtained at the follow-up survey and was correlated to QOL at the same time. The Kaplan-Meier method and Cox regression models were applied to evaluate the association of ginseng use with overall and disease-free survival. The relation of ginseng use and QOL was evaluated by using multiple linear regression models. Approximately 27% of study participants were regular ginseng users before cancer diagnosis. Compared with patients who never used ginseng, regular users had a significantly reduced risk of death; adjusted hazard ratios associated with ginseng use were 0.71 (95% confidence interval: 0.52, 0.98) for total mortality and 0.70 (95% confidence interval: 0.53, 0.93) for disease-specific mortality/recurrence. Ginseng use after cancer diagnosis, particularly current use, was positively associated with QOL scores, with the strongest effect in the psychological and social well-being domains. Additionally, QOL improved as cumulative ginseng use increased.

Investigators have reported an inverse association between coffee consumption and risk of colorectal cancer in several case-control studies, but prospective studies, most of them involving small numbers of cases, have not supported such a relation. In this analysis, the authors prospectively examined the association of coffee consumption with colorectal cancer risk among participants from two population-based cohort studies: 61,433 women in the Swedish Mammography Cohort and 45,306 men in the Cohort of Swedish Men. Information about coffee consumption was obtained from food frequency questionnaires in 1987&ndash;1990 and 1997 for women and in 1997 for men. The authors used Cox proportional hazards modeling for cohort-specific multivariate analyses, and results were pooled using random-effects models. During 1,240,597 person-years of follow-up, 1,279 incident cases of colorectal cancer were diagnosed. Coffee consumption was not associated with risk of colorectal cancer, colon cancer, or rectal cancer in either women or men. For both cohorts combined, the multivariate rate ratio for colorectal cancer for each additional cup of coffee per day was 1.00 (95% confidence interval: 0.97, 1.04). The associations were not modified by colorectal cancer risk factors. The findings from these two large prospective cohort studies do not support the hypothesis that coffee consumption lowers the risk of colorectal cancer.

The authors examined whether a serum pepsinogen test (SPT) based on the combination of the serum pepsinogen I level and pepsinogen I/II ratio is a good predictor of gastric cancer occurrence in a general Japanese population. A total of 2,446 subjects aged &ge;40 years were classified into negative, positive, and strong-positive SPT groups and were followed prospectively for 14 years (1988&ndash;2002). Compared with that for the negative SPT group (26 men, 10 women), gastric cancer incidence increased significantly for both men (n = 17; age-adjusted hazard ratio = 4.56, 95% confidence interval: 2.42, 8.60) and women (n = 6; age-adjusted hazard ratio = 5.84, 95% confidence interval: 2.00, 17.11) in the strong-positive SPT group. It was also significantly higher in the positive SPT group for men (n = 23; age-adjusted hazard ratio = 3.91, 95% confidence interval: 2.23, 6.86). These associations did not attenuate even after adjustment for other comprehensive risk factors. Stratified analysis revealed significant associations between the SPT and development of intestinal-type gastric cancer as well as of cancer in both Helicobacter pylori&ndash;negative and &ndash;positive subjects. These findings suggest that the SPT can serve as a predictor of intestinal-type gastric cancer, irrespective of H. pylori infection.

Socioeconomic position consistently predicts coronary heart disease; however, the biologic mechanisms that may mediate this association are not well understood. The objective of this study was to determine whether socioeconomic position (measured as educational level) is associated with inflammatory risk factors for coronary heart disease, including C-reactive protein, interleukin-6, soluble intercellular adhesion molecule-1, monocyte chemoattractant protein-1, and P-selectin. The study sample included 2,729 participants (53.4% women; mean age, 62 &plusmn; 10 years) from the US Framingham Offspring Study cohort who attended examination cycles 3 (1984&ndash;1987) and 7 (1998&ndash;2001) and provided educational attainment data. Inflammatory markers were measured in fasting serum samples. Multivariable linear regression analyses were performed, adjusting for potential confounders including age, sex, and clinical risk factors. In age- and sex-adjusted analyses, educational attainment was significantly inversely associated with C-reactive protein (p < 0.0001), interleukin-6 (p < 0.0001), soluble intercellular adhesion molecule-1 (p < 0.0001), and monocyte chemoattractant protein-1 (p = 0.0004). After further adjustment for clinical risk factors, educational level remained significantly associated with C-reactive protein (p = 0.0002), soluble intercellular adhesion molecule-1 (p = 0.01), and monocyte chemoattractant protein-1 (p = 0.01). In conclusion, educational attainment is associated with inflammatory risk factors for coronary heart disease. The association provides evidence suggestive of a biologic pathway by which socioeconomic position may predispose to coronary heart disease.

An inverse association between height and risk of coronary heart disease (CHD) is well demonstrated, but it is not known whether this association is because of genetic factors, socioeconomic background, or other environmental factors. Four population-based twin cohorts with register-based follow-up data on CHD mortality from Denmark (1966&ndash;1996), Finland (1975&ndash;2001), and Sweden (1963&ndash;2001 and 1972&ndash;2001) were used to investigate this question; response rates varied between 65% and 86%. Together, the cohorts included 74,704 twin individuals (35,042 complete twin pairs) with 5,943 CHD deaths during 1.99 million person-years of follow-up. Cox and conditional logistic regression models were used. Per 1-standard deviation decrease in height, height was inversely associated with CHD mortality in men (hazard ratio = 1.08, 95% confidence interval (CI): 1.04, 1.12) and in women (hazard ratio = 1.06, 95% CI: 1.01, 1.10). A twin who had died from CHD was on average shorter than the co-twin within monozygotic pairs (odds ratio = 1.27, 95% CI: 1.12, 1.44, with no sex difference), whereas a weaker association was found within dizygotic pairs in men (odds ratio = 1.01, 95% CI: 0.91, 1.13) and in women (odds ratio = 1.14, 95% CI: 1.01, 1.28). The inverse association between height and CHD mortality found within monozygotic discordant twin pairs suggests that this association is because of environmental factors that directly affect height and CHD risk.

Genetic variants in coagulation factors are associated with myocardial infarction and may modify the association between hormone therapy and cardiovascular disease risk. This study assessed whether common variation in the prothrombin gene was associated with incident nonfatal myocardial infarction in postmenopausal women and whether this association differed according to current estrogen use. Eight variants representing 98% of common prothrombin variants were selected using publicly available genomic variation data. These variants and the functional G20210A variant were genotyped and used to infer haplotypes in a population-based Washington State case-control study of postmenopausal Caucasian women (1995&ndash;1999; 273 cases and 788 controls). Women carrying a nonsynonymous polymorphism in exon 6 (C5467T) had an increased risk of myocardial infarction (for each additional copy, relative to women with one fewer copy, odds ratio = 1.4, 95% confidence interval: 1.0, 1.8). Prothrombin haplotypes were also associated with myocardial infarction (with minimal adjustment, global p = 0.056; with full adjustment, p = 0.034). Associations between haplotypes and myocardial infarction were similar among users of hormone therapy and nonusers (global p = 0.61), though statistical power was limited. These preliminary results suggest that common genetic variants in the prothrombin gene or other variants in linkage disequilibrium are associated with myocardial infarction in postmenopausal women.

The authors hypothesized that individuals born in the early 20th century who experienced the hottest and driest summers during infancy would be more likely to have suffered severe infant diarrhea and dehydration, and consequently have had higher blood pressure in adulthood, than those who experienced cooler and wetter summers. In this context, these climate data act as an instrumental variable for the association of early-life dehydration with later blood pressure. For 3,964 randomly selected British women born between 1919 and 1940 and whose blood pressure was measured at age 60&ndash;79 years, a one standard deviation (1.3&deg;C) higher mean summer temperature in the first year of life was associated with a 1.12-mmHg (95% confidence interval: 0.33, 1.91) higher adult systolic blood pressure, and a one standard deviation higher mean summer rainfall (33.9 mm) was associated with lower systolic blood pressure (&ndash;1.65 mmHg, 95% confidence interval: &ndash;2.44, &ndash;0.85). Equivalent results for diastolic blood pressure were 0.11 (95% confidence interval: &ndash;0.65, 0.86) and &ndash;0.32 (95% confidence interval: &ndash;0.71, 0.05). The climate variables were not associated with potential confounding factors such as socioeconomic position or lifestyle risk factors. These findings provide some evidence in favor of the hypothesis that dehydration in infancy is associated with higher adult blood pressure.

Circumstances in which both randomized controlled trial and observational study data are available provide an important opportunity to identify biases and improve study design and analysis procedures. In addition, joint analyses of data from the two sources can extend clinical trial findings. The US Women's Health Initiative includes randomized controlled trials of use of estrogen by posthysterectomy women and of estrogen plus progestin by women with a uterus, along with corresponding observational study components. In this paper, for coronary heart disease, stroke, and venous thromboembolism, results are first presented from joint analysis of estrogen clinical trial and observational study data to show that residual bias patterns are similar to those previously reported for estrogen plus progestin. These findings support certain combined analyses of the observational data on estrogen and the estrogen plus progestin clinical trial and observational study data to give adjusted observational study estimates of estrogen treatment effects. The resulting treatment effect estimates are compared with corresponding clinical trial estimates, and parallel analyses are also presented for estrogen plus progestin. An application to postmenopausal hormone treatment effects on coronary heart disease among younger women is also provided.

A case-crossover study was conducted in 36 US cities to evaluate the effect of ozone and particulate matter with an aerodynamic diameter of &le;10 &micro;m (PM10) on respiratory hospital admissions and to identify which city characteristics may explain the heterogeneity in risk estimates. Respiratory hospital admissions and air pollution data were obtained for 1986&ndash;1999. In a meta-analysis based on the city-specific regression models, several city characteristics were evaluated as effect modifiers. During the warm season, the 2-day cumulative effect of a 5-ppb increase in ozone was a 0.27% (95% confidence interval (CI): 0.08, 0.47) increase in chronic obstructive pulmonary disease admissions and a 0.41% (95% CI: 0.26, 0.57) increase in pneumonia admissions. Similarly, a 10-&micro;g/m3 increase in PM10 during the warm season resulted in a 1.47% (95% CI: 0.93, 2.01) increase in chronic obstructive pulmonary disease at lag 1 and a 0.84% (95% CI: 0.50, 1.19) increase in pneumonia at lag 0. Percentage of households with central air conditioning reduced the effect of air pollution, and variability of summer apparent temperature reduced the effect of ozone on chronic obstructive pulmonary disease. The study confirmed, in a large sample of cities, that exposure to ozone and PM10 is associated with respiratory hospital admissions and provided evidence that the effect of air pollution is modified by certain city characteristics.

During the 2004&ndash;2005 influenza season, the supply of vaccine to the United States was significantly reduced. In response, the Centers for Disease Control and Prevention and the Advisory Committee on Immunization Practices issued interim recommendations for prioritizing vaccination. Given trends in racial/ethnic disparities in vaccination for influenza, the authors assessed the impact of the shortage on those historically less likely to be vaccinated. Using data from the Behavioral Risk Factor Surveillance System, they considered vaccination coverage among those non-Hispanic Whites, non-Hispanic Blacks, and Hispanics who had priority for being vaccinated during the 2004&ndash;2005 influenza season. The vaccine shortage had a significant negative effect on coverage among adults aged 65 years or older across the three racial/ethnic groups. Yet, the magnitude of the disparities in coverage did not change significantly from previous seasons. This finding may imply similar patterns of vaccine-seeking behavior during shortage and nonshortage years. No racial/ethnic differences were seen among adults aged 18&ndash;64 years, which likely reflects the higher percentage of health-care workers in this age group. Yearly monitoring of influenza vaccine coverage is important to assess the long-term impact of shortages on overall coverage and gaps in coverage between racial/ethnic groups.

Assessment of delay in age-appropriate vaccination provides more information about timeliness of vaccination than up-to-date vaccination coverage. The authors applied survival analysis methods to data from a vaccination coverage survey among children aged 13&ndash;59 months conducted in Argentina in 2002. By age 19 months, 43% of children (95% confidence interval (CI): 40, 46) were vaccinated with the fourth dose of diphtheria, tetanus, and pertussis (DTP4). By age 13 months, 55% of children (95% CI: 52, 57) were vaccinated with measles-containing vaccine. By age 7 months, 33% of children (95% CI: 27, 40) were vaccinated with the third dose of hepatitis B. Compared with firstborn children, third children were more likely to be delayed for DTP4 (relative risk (RR) = 1.41, 95% CI: 1.22, 1.62), measles-containing vaccine (RR = 1.54, 95% CI: 1.32, 1.78), and the third dose of hepatitis B (RR = 1.31, 95% CI: 1.03, 1.67). Children whose caregivers had completed secondary school were less likely to be delayed for DTP4 (RR = 0.68, 95% CI: 0.52, 0.90) compared with those whose caregivers had not completed primary school. Survival analysis methods were helpful in measuring vaccine uptake and should be considered in future surveys when assessing delay in age-appropriate vaccination.

Group B Streptococcus causes a variety of morbid and sometimes fatal conditions affecting individuals of all age groups. There are nine known serotypes of this Gram-positive coccus but few estimates of the incidence and duration of its colonization and none by serotype in the literature. In 2001, the authors conducted a prospective cohort study among 257 men and women living in a single dormitory in Ann Arbor, Michigan. The 3-week incidence with any serotype was 11.3% (&plusmn;3.9%) among women and 8.8% (&plusmn;3.0%) among men; 3-week incidence rates were highest for serotype V (4.7% for women and 3.5% for men) and type Ia (2.3% for women and 2.4% for men), with no significant differences by gender. The estimated average duration of any group B Streptococcus colonization was longer for women (13.7 weeks) than men (8.5 weeks); serotype Ia was carried an average of 6.5 weeks longer in women, and serotype III was carried 4.9 weeks longer. Colonization with more than one serotype occurred significantly less than would be expected by chance (p <<< 0.001). Based on the overall incidence, transmission occurred between roommate pairs at the rate expected. Group B Streptococcus colonization is frequent and dynamic, but it is not transmitted by casual contact.