June E. Eichner

Predictions of coronary artery stenosis by artificial neural network

Data from angiography patient records comprised 14 input variables of a neural network. Outcomes (coronary artery stenosis or none) formed both supervisory and output variables. The network was trained by backpropagation on 332 records, optimized on 331 subsequent records, and tested on final 100 records. If 0.40 was chosen as the output distinguishing stenosis from no stenosis, 81 patients who had stenosis would have been identified, while 9 of 19 patients who did not have stenosis might have been spared angiography. The results demonstrated that artificial neural networks could identify some patients who do not need coronary angiography.

Blood Pressure Screening of School Children in a Multiracial School District: The Healthy Kids Project

BACKGROUND There are few studies of the prevalence of elevated blood pressure (BP) that include American Indian school children. Therefore, the intent of this study was to examine the relationships between BP and risk factors in a multiracial, predominantly American Indian, school district. METHODS A total of 1,829 American Indian, white, Hispanic, and African American students, 5-17 years old, were included in this study. The mean of two BP measurements, taken at the initial screening with an electronic BP monitor, were categorized as normal, prehypertensive, or hypertensive using the 2004 Fourth Report BP screening recommendations. Prevalence of prehypertensive and hypertensive BP measurements by race, gender, age, and body mass index (BMI) were determined, and their associations analyzed with logistic regression. RESULTS The prevalence of prehypertensive measurements was 16.7% and prevalence of hypertensive measurements was 13.8% at first screen. Obesity was a risk factor for elevated BP for females and males (adjusted odds ratio (OR) = 4.01 and 4.33, respectively). Older age was also significantly associated with prehypertensive and hypertensive BP measurements, especially among males (adjusted OR = 6.91). Among females, American Indian race was protective against elevated BP (adjusted OR = 0.61). CONCLUSIONS Obesity was consistently associated with an increased risk for elevated BP. American Indian race was associated with decreased risk of elevated BP measurements in females. The high prevalence of obesity-related hypertensive measurements in this population that includes American Indian children was similar to levels found in other multiracial populations of school children when using BP measurements from a single assessment.

Hypothalamic-Pituitary-Adrenal Axis Function: Relative Contributions of Perceived Stress and Obesity in Women

OBJECTIVE A range of behavioral and psychosocial factors may contribute to a chronically stimulated hypothalamic-pituitary-adrenal (HPA) axis and subsequently altered diurnal patterns. The goal of this cross-sectional study was to examine associations among diurnal cortisol levels, perceived stress, and obesity patterns. METHODS Seventy-eight women (aged 24-72 years) employed in a rural public school system completed the perceived stress scale, collected diurnal saliva samples, and underwent anthropometric assessments. Reduced peak-to-nadir cortisol values across the day were considered a sign of impairment in HPA function. A series of linear regression models determined the best predictors of diurnal cortisol variation. RESULTS There was a marginal linear trend in stress levels across body mass index (BMI) categories, with obese women reporting the highest levels of stress (p = 0.07). Perceived stress was the only significant predictor of the degree of flattening of the diurnal cortisol curve in the sample as a whole (beta = -0.042, R(2) = 0.11, F = 8.6, p = 0.005), indicating reduction in the normal diurnal pattern. Among overweight women (BMI = 25-29.9 kg/m(2)), stress and waist circumference combined predicted 35% of the variability in diurnal cortisol. In contrast, among obese women (BMI > or = 30 kg/m(2)), BMI predicted 31% of the variability in diurnal cortisol (F = 13.8, p = 0.001), but stress was no longer significantly related to diurnal cortisol. CONCLUSIONS Psychological stress predicts a significant portion of HPA axis functioning. In overweight women, perceived stress and waist circumference were of approximately equal importance in predicting adrenal cortisol secretion. However, among obese women, a major portion of the diurnal cortisol variation was predicted by BMI alone, not stress or waist circumference. This may help elucidate the mechanisms linking obesity to increased risk of cardiovascular disease (CVD).

Overweight and obesity in an ethnically diverse rural school district: the Healthy Kids Project.

Objective: Data on Native American children and adolescents are rarely reported along with other racial and ethnic groups. The Healthy Kids Project is part of an effort to describe the prevalence of overweight and obesity in a racially mixed rural area where Native American, Hispanic, African American, and white children reside.

Iron measures in coronary angiography patients

Excess iron has been postulated as a risk factor for coronary artery disease (CAD) because of its presence in atherosclerotic lesions, its ability to oxidize low density lipoprotein cholesterol (LDLc), and its promotion of oxygen reperfusion damage after an ischemic event. Whether iron, indirectly measured by its storage protein ferritin and its transport protein transferrin, is related to CAD was examined in a consecutive series of white male (n = 457) and female (n = 114) cardiac patients. Atherosclerosis measures were analyzed in patients grouped by tertiles of ferritin. A similar analysis was done with tertiles of transferrin. Contrary to expectations, men in the third tertile of ferritin had a smaller mean number of stenoses than men in the two lower tertiles (4.9 versus 5.6 and 5.9; P = 0.027); otherwise, there were no statistically significant differences in either number of lesions or extent of arterial narrowing based on tertiles of either measure. Separate multiple logistic regression models with age, fibrinogen, LDLc and triglycerides as covariates provided no evidence that ferritin (odds ratio = 0.88 with 95% C.I. = 0.72-1.07 for men and odds ratio = 0.79 with 95% C.I. = 0.54-1.16 for women) or transferrin (odds ratio = 0.60 with 95% C.I. = 0.31-1.16 for men and odds ratio = 1.33 with 95% C.I. 0.52-3.42 for women) were important correlates of the presence of atherosclerosis in this study.

Neural network predictions of significant coronary artery stenosis in men

OBJECTIVE A neural network system was designed to predict whether coronary arteriography on a given patient would reveal any occurrence of significant coronary stenosis (>50%), a degree of stenosis which often leads to coronary intervention. METHODOLOGY A dataset of 2004 records from male cardiology patients was derived from a national cardiac catheterization database. The catheterizations selected for analysis from the database were first-time and elective, and they were precipitated by chest pain. Eleven patient variables were used as inputs in an artificial neural network system. The network was trained on the earliest 902 records in the dataset. The next 902 records formed a cross-validation file, which was used to optimize the training. A third file composed of the next 100 records facilitated the choice of a cutoff number between 0 and 1. The cutoff number was applied to the last 100 records, which comprised a test file. RESULTS When a cutoff of 0.25 was compared to the network outputs of all 100 records in the test file, 12 of 46 (specificity=26%) patients without significant stenosis had outputs<or=0.25, but all patients with significant stenosis had outputs>0.25 (sensitivity=100%). Therefore, the network identified a fraction of the patients in the test file who did not have significant coronary artery stenosis, while at the same time the network identified all of the patients in the test file who had significant stenosis capable of causing chest pain. CONCLUSION Artificial neural networks may be helpful in reducing unnecessary cardiac catheterizations.

Evaluation of a community-based participatory physical activity promotion project: effect on cardiovascular disease risk profiles of school employees

BackgroundThe efficacy of physical activity in improving cardiovascular disease (CVD) risk profiles has been well established. However, the effectiveness of health promotion programs implemented at the community level remains controversial. This study evaluated a school-based work-site physical activity program.MethodsUsing a community-based participatory research model, a work-site wellness intervention was implemented in a rural public school system in Southwestern Oklahoma. During the 2005-2006 school year, 187 participants (mean age 45 years) completed a pre intervention screening for CVD risk factors followed by a physical activity promotion program. Post intervention screening was conducted after a 6 month period. During both screening sessions, body composition, blood pressure, lipids, glucose and self-reported physical activity levels were assessed. The focus of the intervention was on promoting physical activity. Opportunities for in school physical activity were created by marking hallways, adding a treadmill in each school, and allowing teachers to use planning periods for physical activity.ResultsDuring the post intervention screening, compared to pre intervention levels, participants had lower total, low, and high density lipoprotein-cholesterol (t = 5.9, p < 0.0001, t = 2.6, p = 0.01, and t = 13.2, p < 0.0001 respectively), lower systolic blood pressure (t = 2.9, p = 0.004), and higher self-reported physical activity levels (Sign t = -1.901, p = 0.06).ConclusionsA successful participatory program was associated with improvements in several CVD risk factors among school employees. Limitations of this study such as seasonal variation in the outcome variables and lack of a control group limit our ability to draw solid conclusions about the effectiveness of the intervention.

Tobacco use among American Indians in Oklahoma: An epidemiologic view

Objectives. With the exception of national surveys that sample the entire U.S. population, little information exists on tobacco habits among American Indians. This study is a comparison of tobacco use findings in the 1990s among American Indians in Oklahoma, a state with a large and diverse American Indian population (39 tribes). Methods. Data on current tobacco use are presented from two statewide surveys, the Oklahoma Youth Tobacco Survey and the Native American Behavioral Risk Factor Survey, as well as two large epidemiologic studies of chronic disease among American Indians—the Cherokee Diabetes Study and the Strong Heart Study. Three of these four sources of data involve research/surveys exclusively about American Indians. Results. Nontraditional use of tobacco by American Indians occurs frequently, according to each instrument. Initiation to this habit begins in middle school and increases dramatically during high school. After age 50, reporting by individuals that they currently smoke declines steadily. Conclusions. Despite sampling different individuals for the surveys and different tribes for the epidemiologic research, results were comparable in age groups that overlapped. These findings support national data indicating that American Indians have higher prevalence rates of smoking than other racial/ethnic groups. American Indians report smoking on average about a half a pack of cigarettes per day. Individuals reporting using tobacco solely for ceremonial purposes were far fewer than habitual users. Buying tobacco products in American Indian smoke shops helps tribal economies; this fact needs to be considered for prevention programs to succeed.

Angiotensin-converting enzyme gene polymorphism in a cohort of coronary angiography patients

An association between a polymorphism of the angiotensin-converting enzyme (ACE) gene and myocardial infarction (MI) in men has been previously reported. The present study examines the association between ACE genotype, atherosclerosis, MI, hypertension and other cardiovascular risk factors in Caucasian men (n=576) and women (n=124) who have undergone coronary angiography. Gene frequencies are also reported for African-American men (n=56). Genotype determination was based on the presence (allele I) or absence (allele D) of a 287 nucleotide Alu sequence in intron 16 of the ACE gene. Genotype frequencies for DD, ID and II were: 30.9, 47.7, 21.4% for Caucasian men; 28.2, 48.4, 23.4% for Caucasian women; and 30.4, 46.4, 23.2% for African-American men. There were no statistically significant associations between ACE genotype and number of plaques (> or =10% obstruction), lipid variables, or body mass index (BMI) for Caucasian men. Caucasian women with the DD genotype had on average fewer plaques, but this was accounted for by their younger ages. In Caucasian males, the DD genotype independently contributed to the presence of hypertension (odds ratio=1.8, 95% CI 1.1-2.9) after adjusting for age and BMI. In Caucasian males with total cholesterol levels less than 200 mg/dl (n=237), the DD (odds ratio=2.5, 95% CI 1.2-5.4) and ID genotypes (odds ratio=2.2, 95% CI 1.1-4.4) were associated with a history of MI.

Role of Factor V Leiden Mutation in Patients with Angiographically Demonstrated Coronary Artery Disease

The study sought to determine whether coagulation factor V Leiden (FV Leiden) plays a role in the pathogenesis of coronary artery disease and/or myocardial infarction. Association of FV Leiden with venous thromboembolism is well established in the literature, but the role of the mutation in arterial thrombotic events is controversial. Some studies have documented an association between the mutation and myocardial infarction and stroke in juveniles. Few studies have explored its possible contribution to coronary atherosclerosis. We screened FV genotype in 850 predominantly white coronary angiography patients. Coronary artery disease risk factors and history of myocardial infarction were then analyzed by genotype. The FV Leiden mutation occurred in 54 (6.4%) patients. There was one homozygote; a 37-year-old, white male smoker with a history of myocardial infarction. Gene frequencies for white males and females were similar: 0.965 for the normal allele and 0.035 for FV Leiden. Gene frequencies for both genders were in Hardy-Weinberg equilibrium. FV Leiden was not a useful predictor (p=0.23) of the presence of clinically defined atherosclerosis (> or = 50% stenosis) in a logistic regression model adjusting for age, lipoprotein (a), total cholesterol, triglycerides, high density lipoprotein cholesterol, and fibrinogen. In addition, there was no difference in frequency of FV Leiden among those with and without medical histories of myocardial infarction (p=0.51). Allelic frequencies of FV Leiden in this patient group do not differ significantly from those reported for white populations. The FV Leiden mutation in its heterozygous state is not independently associated with coronary artery disease or myocardial infarction.