Explanatory Variables In Regression Analysis Research Articles

Bivariate Log-Symmetric Regression Models Applied to Newborn Data

This paper introduces bivariate log-symmetric models for analyzing the relationship between two variables, assuming a family of log-symmetric distributions. These models offer greater flexibility than the bivariate lognormal distribution, allowing for better representation of diverse distribution shapes and behaviors in the data. The log-symmetric distribution family is widely used in various scientific fields and includes distributions such as log-normal, log-Student-t, and log-Laplace, among others, providing several options for modeling different data types. However, there are few approaches to jointly model continuous positive and explanatory variables in regression analysis. Therefore, we propose a class of generalized linear model (GLM) regression models based on bivariate log-symmetric distributions, aiming to fill this gap. Furthermore, in the proposed model, covariates are used to describe its dispersion and correlation parameters. This study uses a dataset of anthropometric measurements of newborns to correlate them with various biological factors, proposing bivariate regression models to account for the relationships observed in the data. Such models are crucial for preventing and controlling public health issues.

Imprinting methylation predicts hippocampal volumes and hyperintensities and the change with age in later life

Epigenetic imprinting is important for neurogenesis and brain function. Hippocampal volumes and brain hyperintensities in late life have been associated with early life circumstances. Epigenetic imprinting may underpin these associations. Methylation was measured at 982 sites in 13 imprinted locations in blood samples from a longitudinal cohort by bisulphite amplicon sequencing. Hippocampal volumes and hyperintensities were determined at age 64y and 72y using MRI. Hyperintensities were determined in white matter, grey matter and infratentorial regions. Permutation methods were used to adjust for multiple testing. At 64y, H19/IGF2 and NESPAS methylation predicted hippocampal volumes. PEG3 predicted hyperintensities in hippocampal grey matter, and white matter. GNASXL predicted grey matter hyperintensities. Changes with age were predicted for hippocampal volume (MEST1, KvDMR, L3MBTL, GNASXL), white matter (MEST1, PEG3) and hippocampal grey matter hyperintensities (MCTS2, GNASXL, NESPAS, L3MBTL, MCTS2, SNRPN, MEST1). Including childhood cognitive ability, years in education, or socioeconomic status as additional explanatory variables in regression analyses did not change the overall findings. Imprinting methylation in multiple genes predicts brain structures, and their change over time. These findings are potentially relevant to the development of novel tests of brain structure and function across the life-course, strategies to improve cognitive outcomes, and our understanding of early influences on brain development and function.

Morbidity and mortality in critically ill patients with invasive group A streptococcus infection: an observational study

BackgroundGroup A streptococci (GAS) are known to cause serious invasive infections, but little is known about outcomes when patients with these infections are admitted to intensive care. We wanted to describe critically ill patients with severe sepsis or septic shock due to invasive GAS (iGAS) and compare them with other patients with severe sepsis or septic shock.MethodsAdult patients admitted to a general intensive care unit (ICU) in Sweden (2007–2019) were screened for severe sepsis or septic shock according to Sepsis 2 definition. Individuals with iGAS infection were identified. The outcome variables were mortality, days alive and free of vasopressors and invasive mechanical ventilation, maximum acute kidney injury score for creatinine, use of continuous renal replacement therapy and maximum Sequential Organ Failure Assessment score during the ICU stay. Age, Simplified Acute Physiology Score (SAPS 3) and iGAS were used as independent, explanatory variables in regression analysis. Cox regression was used for survival analyses.ResultsiGAS was identified in 53 of 1021 (5.2%) patients. Patients with iGAS presented a lower median SAPS 3 score (62 [56–72]) vs 71 [61–81]), p < 0.001), had a higher frequency of cardiovascular cause of admission to the ICU (38 [72%] vs 145 [15%], p < 0.001) and had a higher median creatinine score (173 [100–311] vs 133 [86–208] μmol/L, p < 0.019). Of the GAS isolates, 50% were serotyped emm1/T1 and this group showed signs of more pronounced circulatory and renal failure than patients with non-emm1/T1 (p = 0.036 and p = 0.007, respectively). After correction for severity of illness (SAPS 3) and age, iGAS infection was associated with lower mortality risk (95% confidence interval (CI) of hazard ratio (HR) 0.204–0.746, p < 0.001). Morbidity analyses demonstrated that iGAS patients were more likely to develop renal failure.ConclusionCritically ill patients with iGAS infection had a lower mortality risk but a higher degree of renal failure compared to similarly ill sepsis patients. emm1/T1 was found to be the most dominant serotype, and patients with emm1/T1 demonstrated more circulatory and renal failure than patients with other serotypes of iGAS.

Variable selection for distribution-free models for longitudinal zero-inflated count responses.

Zero-inflated count outcomes arise quite often in research and practice. Parametric models such as the zero-inflated Poisson and zero-inflated negative binomial are widely used to model such responses. Like most parametric models, they are quite sensitive to departures from assumed distributions. Recently, new approaches have been proposed to provide distribution-free, or semi-parametric, alternatives. These methods extend the generalized estimating equations to provide robust inference for population mixtures defined by zero-inflated count outcomes. In this paper, we propose methods to extend smoothly clipped absolute deviation (SCAD)-based variable selection methods to these new models. Variable selection has been gaining popularity in modern clinical research studies, as determining differential treatment effects of interventions for different subgroups has become the norm, rather the exception, in the era of patent-centered outcome research. Such moderation analysis in general creates many explanatory variables in regression analysis, and the advantages of SCAD-based methods over their traditional counterparts render them a great choice for addressing this important and timely issues in clinical research. We illustrate the proposed approach with both simulated and real study data. Copyright © 2016 John Wiley & Sons, Ltd.

Is awareness of public charging associated with consumer interest in plug-in electric vehicles?

Policymakers often seek to increase the visibility of plug-in electric vehicle (PEV) chargers in public locations in effort to build familiarity and interest in PEVs. However, it is not clear if the visibility of public charging stations actually has an impact on PEV demand. The purposes of the present study are to (1) assess the current levels of visibility for public PEV charging infrastructure within Canada and (2) identify whether or not a statistically significant relationship exists between consumer awareness of public charging infrastructure and interest in purchasing a PEV. We use data collected from a sample of 1739 Canadian new-vehicle buyers in 2013. About 18% of Canadian respondents have seen at least one public charger, while the proportion is highest in British Columbia (31%). We find a significant bivariate relationship between public charger awareness and PEV interest. However, when controlling for multiple explanatory variables in regression analyses, the relationship is weak or non-existent. While perceived existence of at least one charger exhibits no significant relationship with PEV interest, perceived existence of multiple chargers can have a weak but significant relationship. Thus, public charger awareness is not a strong predictor of PEV interest; other variables are more important, such as the availability of level 1 (110/120-volt) charging at home.

The use of cognitive ability measures as explanatory variables in regression analysis.

Cognitive ability measures are often taken as explanatory variables in regression analysis, e.g., as a factor affecting a market outcome such as an individual’s wage, or a decision such as an individual’s education acquisition. Cognitive ability is a latent construct; its true value is unobserved. Nonetheless, researchers often assume that a test score, constructed via standard psychometric practice from individuals’ responses to test items, can be safely used in regression analysis. We examine problems that can arise, and suggest that an alternative approach, a “mixed effects structural equations” (MESE) model, may be more appropriate in many circumstances.

Borderline personality disorder as a female phenotypic expression of psychopathy?

Evidence suggests that the combination of the interpersonal-affective (F1) and impulsive-antisocial (F2) features of psychopathy may be associated with borderline personality disorder (BPD), specifically among women (e.g., Coid, 1993; Hicks, Vaidyanathan, & Patrick, 2010). However, empirical research explicitly examining gendered relationships between BPD and psychopathy factors is lacking. To further inform this area of research, we investigated the hypothesis that the interplay between the two psychopathy factors is associated with BPD among women across two studies. Study 1 consisted of a college sample of 318 adults (51% women), and Study 2 consisted of a large sample of 488 female prisoners. The interpersonal-affective (F1) and impulsive-antisocial psychopathy (F2) scores, measured with self-report and clinician-rated indices, respectively, were entered as explanatory variables in regression analyses to investigate their unique contributions to BPD traits. Across two independent samples, results indicated that the interaction of high F1 and F2 psychopathy scores was associated with BPD in women. This association was found to be specific to women in Study 1. These results suggest that BPD and psychopathy, at least as they are measured by current instruments, overlap in women and, accordingly, may reflect gender-differentiated phenotypic expressions of similar dispositional vulnerabilities.

Gender differences in emotional risk for self- and other-directed violence among externalizing adults.

Women and men generally differ in how frequently they engage in other- and self-directed physical violence and may show distinct emotional risk factors for engagement in these high-impact behaviors. To inform this area, we investigated gender differences in the relationship of emotional tendencies (i.e., anger, hostility, and anhedonic depression) that may represent risk for other-directed violence (i.e., physical fighting, attacking others unprovoked) and self-directed violence (i.e., self-injury, suicide attempts). The ethnically diverse sample consisted of 372 adults (252 men and 120 women age 18-55) with a history of criminal convictions. Facets of emotional risk assessed with the Aggression Questionnaire (Buss & Warren, 2000) and Mood and Anxiety Symptom Questionnaire (Watson et al., 1995) were entered simultaneously as explanatory variables in regression analyses to investigate their unique contributions to other- and self-directed physical violence in men and women. Analyses revealed that anhedonic depressive tendencies negatively predicted other-directed violence and positively predicted self-directed violence in men and women, consistent with a model of depression in which aggression is turned inward (Henriksson et al., 1993). Gender differences, however, emerged for the differential contributions of anger and hostility to other- and self-directed violence. Trait anger (i.e., difficulty controlling one's temper) was associated with other-directed violence selectively in men, whereas trait hostility (i.e., suspiciousness and alienation) was associated with self- and other-directed violence among women. The divergent findings for trait anger and hostility underscore the need to examine gender-specific risk factors for physical violence to avoid excluding potentially useful clinical features of these mental health outcomes.

Explaining the differential distribution of Clean Development Mechanism projects across host countries

The Clean Development Mechanism (CDM) of the Kyoto Protocol represents an opportunity to involve all developing countries in the effort to reduce greenhouse gas emissions while also promoting sustainable development. To date, however, the majority of CDM projects have gone to emerging markets such as China, India, Brazil, and Mexico, while very few least developed countries have hosted projects. This paper investigates the differential distribution of CDM activities across countries. We develop a conceptual model for project profitability, which helps to identify potential country-level determinants of CDM activity. These potential determinants are employed as explanatory variables in regression analysis to explain the actual distribution of projects. Human capital and greenhouse gas emission levels influenced which countries have hosted projects and the amount of certified emission reductions (CER) created. Countries that offered growing markets for CDM co-products, such as electricity, were more likely to be CDM hosts, while economies with higher carbon intensity levels had greater CER production. These findings work against the least developed countries and help to explain their lack of CDM activity.

Predicting psychiatric admission rates.

To determine the numbers of actual and expected psychiatric admissions for the residents of the district health authorities of England and to develop a model to indicate which social, health status, and service provision factors best explain the variation of the actual from the expected psychiatric admissions; to use this model to predict psychiatric admission for district health authorities as an aid to resource allocation. The actual psychiatric admission for district health authority residents were extracted from data of the 1986 Mental Health Enquiry. Expected admissions were calculated using the age, sex, and marital status structure of each district health authority and the national psychiatric admission rates related to age, sex, and marital status. Standardised psychiatric admission ratios were calculated as the ratios of the numbers of actual to expected psychiatric admissions. A wide range of social, health status, and service provision data were used as the explanatory variables in regression analyses to determine which combination of factors best explained the variation between districts of standardised psychiatric admission ratios. The 168,652 psychiatric admissions recorded for the 1986 Mental Health Enquiry, after exclusion of mental handicap and psychogeriatric admissions. The actual number of psychiatric admissions varied from 79% above to 54% below the expected number of admissions from age, sex, and marital status for the districts of England. The most powerful variables to explain this variation were the rate of notification of drug misusers, standardised mortality ratios, and levels of illegitimacy in each district. A complex model was developed which could be used to predict district psychiatric admissions as an aid to resource allocation. A simpler model was also developed (which was less powerful than the more complex model) based on the underprivileged area score. One advantage of this model was that it could be used at the level of electoral wards as well as district health authorities.