Competing regression models for longitudinal data

The choice of an appropriate family of linear models for the analysis of longitudinal data is often a matter of concern for practitioners. To attenuate such difficulties, we discuss some issues that emerge when analyzing this type of data via a practical example involving pretestposttest longitudinal data. In particular, we consider log-normal linear mixed models (LNLMM), generalized linear mixed models (GLMM), and models based on generalized estimating equations (GEE). We show how some special features of the data, like a nonconstant coefficient of variation, may be handled in the three approaches and evaluate their performance with respect to the magnitude of standard errors of interpretable and comparable parameters. We also show how different diagnostic tools may be employed to identify outliers and comment on available software. We conclude by noting that the results are similar, but that GEE-based models may be preferable when the goal is to compare the marginal expected responses.

Evaluating links through spectral decomposition

Spectral decomposition has rarely been used to investigate complex networks. In this work we apply this concept in order to define two kinds of link-directed attacks while quantifying their respective effects on the topology. Several other kinds of more traditional attacks are also adopted and compared. These attacks had substantially diverse effects, depending on each specific network (models and real-world structures). It is also shown that the spectrally based attacks have special effects in affecting the transitivity of the networks.

Semi-supervised dimensionality reduction based on partial least squares for visual analysis of high dimensional data

Dimensionality reduction is employed for visual data analysis as a way to obtaining reduced spaces for high dimensional data or to mapping data directly into 2D or 3D spaces. Although techniques have evolved to improve data segregation on reduced or visual spaces, they have limited capabilities for adjusting the results according to user's knowledge. In this paper, we propose a novel approach to handling both dimensionality reduction and visualization of high dimensional data, taking into account user's input. It employs Partial Least Squares (PLS), a statistical tool to perform retrieval of latent spaces focusing on the discriminability of the data. The method employs a training set for building a highly precise model that can then be applied to a much larger data set very effectively. The reduced data set can be exhibited using various existing visualization techniques. The training data is important to code user's knowledge into the loop. However, this work also devises a strategy for calculating PLS reduced spaces when no training data is available. The approach produces increasingly precise visual mappings as the user feeds back his or her knowledge and is capable of working with small and unbalanced training sets.

Accuracy of positron emission tomography/computed tomography and clinical assessment in the detection of complete rectal tumor regression after neoadjuvant chemoradiation Long-Term Results of a Prospective Trial (National Clinical Trial 00254683)

BACKGROUND: Neoadjuvant chemoradiation (CRT) therapy may result in significant tumor regression in patients with rectal cancer. Patients who develop complete tumor regression have been managed by treatment strategies that are alternatives to standard total mesorectal excision. Therefore, assessment of tumor response with positron emission tomography/computed tomography (PET/CT) after neoadjuvant treatment may offer relevant information for the selection of patients to receive alternative treatment strategies. METHODS: Patients with clinical T2 (cT2) through cT4NxM0 rectal adenocarcinoma were included prospectively. Neoadjuvant therapy consisted of 54 grays of radiation and 5-fluorouracil-based chemotherapy. Baseline PET/CT studies were obtained before CRT followed by PET/CT studies at 6 weeks and 12 weeks after the completion of CRT. Clinical assessment was performed at 12 weeks after CRT completion. PET/CT results were compared with clinical and pathologic data. RESULTS: In total, 99 patients were included in the study. Twenty-three patients were complete responders (16 had a complete clinical response, and 7 had a complete pathologic response). The PET/CT response evaluation at 12 weeks indicated that 18 patients had a complete response, and 81 patients had an incomplete response. There were 5 false-negative and 10 false-positive PET/CT results. PET/CT for the detection of residual cancer had 93% sensitivity, 53% specificity, a 73% negative predictive value, an 87% positive predictive value, and 85% accuracy. Clinical assessment alone resulted in an accuracy of 91%. PET/CT information may have detected misdiagnoses made by clinical assessment alone, improving overall accuracy to 96%. CONCLUSIONS: Assessment of tumor response at 12 weeks after CRT completion with PET/CT imaging may provide a useful additional tool with good overall accuracy for the selection of patients who may avoid unnecessary radical resection after achieving a complete clinical response. Cancer 2012;35013511. (C) 2011 American Cancer Society.

Sudden Cardiac Death in Brazil: Study Based on Physicians' Perceptions of the Public Health Care System

Background: There are no available statistical data about sudden cardiac death in Brazil. Therefore, this study has been conducted to evaluate the incidence of sudden cardiac death in our population and its implications. Methods: The research methodology was based on Thurstone's Law of Comparative Judgment, whose premise is that the more an A stimulus differs from a B stimulus, the greater will be the number of people who will perceive this difference. This technique allows an estimation of actual occurrences from subjective perceptions, when compared to official statistics. Data were collected through telephone interviews conducted with Primary and Secondary Care physicians of the Public Health Service in the Metropolitan Area of Sao Paulo (MASP). Results: In the period from October 19, 2009, to October 28, 2009, 196 interviews were conducted. The incidence of 21,270 cases of sudden cardiac death per year was estimated by linear regression analysis of the physicians responses and data from the Mortality Information System of the Brazilian Ministry of Health, with the following correlation and determination coefficients: r = 0.98 and r2= 0.95 (95% confidence interval 0.81.0, P < 0.05). The lack of waiting list for specialized care and socioadministrative problems were considered the main barriers to tertiary care access. Conclusions: The incidence of sudden cardiac death in the MASP is high, and it was estimated as being higher than all other causes of deaths; the extrapolation technique based on the physicians perceptions was validated; and the most important bureaucratic barriers to patient referral to tertiary care have been identified. (PACE 2012; 35:13261331)

Utility Harmonic Impedance Measurement Based on Data Selection

Determination of the utility harmonic impedance based on measurements is a significant task for utility power-quality improvement and management. Compared to those well-established, accurate invasive methods, the noninvasive methods are more desirable since they work with natural variations of the loads connected to the point of common coupling (PCC), so that no intentional disturbance is needed. However, the accuracy of these methods has to be improved. In this context, this paper first points out that the critical problem of the noninvasive methods is how to select the measurements that can be used with confidence for utility harmonic impedance calculation. Then, this paper presents a new measurement technique which is based on the complex data-based least-square regression, combined with two techniques of data selection. Simulation and field test results show that the proposed noninvasive method is practical and robust so that it can be used with confidence to determine the utility harmonic impedances.

Dental fluorosis in the primary dentition and intake of manufactured soy-based foods with fluoride

Background & aims: To identify manufactured soy-based products more recommended by pediatricians and nutritionists; to determine fluoride concentrations in these products; to evaluate children concerning fluorosis in primary teeth and its association with the consumption of soy-based products. Methods: Pediatricians and Nutritionists answered a questionnaire about soy-based products they most recommended to children. Fluoride concentrations of the 10 products more cited were analyzed with the ion-specific electrode. Dental fluorosis exams were performed in 315 4e6-year-old children. Dean’s Index was used to assess fluorosis. Among the children examined, 26 had lactose intolerance. Their parents answered a questionnaire about children’s and family’s profile, besides permitting the identification of soy-based products use. Chi-squared and Multivariable Logistic Regression tests were used (p < 0.05). Results: Fluoride content in the analyzed products ranged from 0.03 to 0.50 mg F /mL. Dental fluorosis was detected in 11% of the children, with very mild and mild degrees. Dental fluorosis in primary teeth was associated with lactose intolerance (p < 0.05), but there was no significant association with the use of manufactured soy-based products. Conclusions: Isolated consumption of soy-based products recommended by health professionals to children do not offer risk of dental fluorosis in primary teeth, which had a low prevalence and severity.

Prevalence and correlates of dizziness in community-dwelling older people: a cross sectional population based study

Abstract Background Dizziness is a common complaint among older adults and has been linked to a wide range of health conditions, psychological and social characteristics in this population. However a profile of dizziness is still uncertain which hampers clinical decision-making. We therefore sought to explore the relationship between dizziness and a comprehensive range of demographic data, diseases, health and geriatric conditions, and geriatric syndromes in a representative sample of community-dwelling older people. Methods This is a cross-sectional, population-based study derived from FIBRA (Network for the Study of Frailty in Brazilian Elderly Adults), with 391 elderly adults, both men and women, aged 65 years and older. Elderly participants living at home in an urban area were enrolled through a process of random cluster sampling of census regions. The outcome variable was the self-report of dizziness in the last year. Several feelings of dizziness were investigated including vertigo, spinning, light or heavy headedness, floating, fuzziness, giddiness and instability. A multivariate logistic regression analysis was conducted to estimate the adjusted odds ratios and build the probability model for dizziness. Results The complaint of dizziness was reported by 45% of elderly adults, from which 71.6% were women (p=0.004). The multivariate regression analysis revealed that dizziness is associated with depressive symptoms (OR = 2.08; 95% CI 1.29–3.35), perceived fatigue (OR = 1.93; 95% CI 1.21-3.10), recurring falls (OR = 2.01; 95% CI 1.11-3.62) and excessive drowsiness (OR = 1.91; 95% CI 1.11–3.29). The discrimination of the final model was AUC = 0.673 (95% CI 0.619-0.727) (p< 0.001). Conclusions The prevalence of dizziness in community-dwelling elderly adults is substantial. It is associated with other common geriatric conditions usually neglected in elderly adults, such as fatigue and drowsiness, supporting its possible multifactorial manifestation. Our findings demonstrate the need to expand the design in future studies, aiming to estimate risk and identify possible causal relations.

The double task of preventing malnutrition and overweight: a quasi-experimental community-based trial

Background: The Maternal-Child Pastoral is a volunteer-based community organization of the Dominican Republic that works with families to improve child survival and development. A program that promotes key practices of maternal and child care through meetings with pregnant women and home visits to promote child growth and development was designed and implemented. This study aims to evaluate the impact of the program on nutritional status indicators of children in the first two years of age. Methods: A quasi-experimental design was used, with groups paired according to a socioeconomic index, comparing eight geographical areas of intervention with eight control areas. The intervention was carried out by lay health volunteers. Mothers in the intervention areas received home visits each month and participated in a group activity held biweekly during pregnancy and monthly after birth. The primary outcomes were length and body mass index for age. Statistical analyses were based on linear and logistic regression models. Results: 196 children in the intervention group and 263 in the control group were evaluated. The intervention did not show statistically significant effects on length, but point estimates found were in the desired direction: mean difference 0.21 (95%CI −0.02; 0.44) for length-for-age Z-score and OR 0.50 (95%CI 0.22; 1.10) for stunting. Significant reductions of BMI-for-age Z-score (−0.31, 95%CI −0.49; -0.12) and of BMI-for-age > 85th percentile (0.43, 95%CI 0.23; 0.77) were observed. The intervention showed positive effects in some indicators of intermediary factors such as growth monitoring, health promotion activities, micronutrient supplementation, exclusive breastfeeding and complementary feeding. Conclusions: Despite finding effect measures pointing to effects in the desired direction related to malnutrition, we could only detect a reduction in the risk of overweight attributable to the intervention. The findings related to obesity prevention may be of interest in the context of the nutritional transition. Given the size of this study, the results are encouraging and we believe a larger study is warranted.

Different approaches for dealing with rare variants in family-based genetic studies: application of a Genetic Analysis Workshop 17 problem

Rare variants are becoming the new candidates in the search for genetic variants that predispose individuals to a phenotype of interest. Their low prevalence in a population requires the development of dedicated detection and analytical methods. A family-based approach could greatly enhance their detection and interpretation because rare variants are nearly family specific. In this report, we test several distinct approaches for analyzing the information provided by rare and common variants and how they can be effectively used to pinpoint putative candidate genes for follow-up studies. The analyses were performed on the mini-exome data set provided by Genetic Analysis Workshop 17. Eight approaches were tested, four using the trait’s heritability estimates and four using QTDT models. These methods had their sensitivity, specificity, and positive and negative predictive values compared in light of the simulation parameters. Our results highlight important limitations of current methods to deal with rare and common variants, all methods presented a reduced specificity and, consequently, prone to false positive associations. Methods analyzing common variants information showed an enhanced sensibility when compared to rare variants methods. Furthermore, our limited knowledge of the use of biological databases for gene annotations, possibly for use as covariates in regression models, imposes a barrier to further research.

Accelerating benders decomposition with heuristicmaster problem solutions

In this paper, a general scheme for generating extra cuts during the execution of a Benders decomposition algorithm is presented. These cuts are based on feasible and infeasible master problem solutions generated by means of a heuristic. This article includes general guidelines and a case study with a fixed charge network design problem. Computational tests with instances of this problem show the efficiency of the strategy. The most important aspect of the proposed ideas is their generality, which allows them to be used in virtually any Benders decomposition implementation.