Autoregressive Correlation Structure Research Articles

Descent of the fetal head (station) during the first stage of labor

High station at specific points in the first stage of labor, such as a floating head on admission, or at 4-cm dilation or when arrest of dilation occurs, is associated with higher rates of failure to deliver vaginally. Therefore it could be useful to know if station is within an expected range at a given dilation during first stage. Arrest of descent disorders have been defined thus far on criteria applicable in the second stage. Statistical modeling is an attractive methodology to characterize the relationship between station and dilation because the resulting mathematical expressions could be used as a reference for comparison in the future. In addition, they can be used to produce a finely graded assessment of descent using numerical terms such as percentile rankings. A 2-step approach to potentially improving the assessment of station could be to develop a statistical model that describes the general relationship between station and dilation in the first stage of uncomplicated births and then determine if such a model would have identified births with complications related to poor labor progress. Given the complex nature of labor data, especially the imprecision of dilation and station measurement, it is not immediately evident that such a model is identifiable or what its precision would be. We sought to characterize in mathematical terms the relationship of station to dilation during the first stage of labor for nulliparous and multiparous women with spontaneous vaginal births. This retrospective cohort study included 28,121 exams from 5555 women with singleton cephalic presentations at ≥37 weeks' gestation with electronic fetal monitoring tracings, who delivered vaginally without instrumentation and had 5-minute Apgar scores >6 at 2 academic community referral hospitals in 2012 through 2013. Women with a previous cesarean birth were excluded. We used longitudinal statistical techniques suitable to biological data that were irregularly sampled with repeated measures over time. A linear relationship was observed between station and dilation. For both nulliparous and multiparous women the final model was a linear regression with random effects for intercept and slope and a first-order autoregressive correlation structure. The 5th-95th range of station at any given dilation spanned about 3-4 cm. Our results demonstrate a general trend of increasing descent of the presenting part as dilation advances during the first stage of labor in women who delivered vaginally without instrumentation. We propose that the mathematical expressions describing this relationship may be valuable in the assessment of first-stage labor progression.

Do symptom-specific stages of change predict eating disorder treatment outcome?

Interview methods to assess stages of change (SOC) in eating disorders (ED) indicate that SOC are positively correlated with symptom improvement over time. However, interviews require significant time and staff training and global measures of SOC do not capture varying levels of motivation across ED symptoms. This study used a self-report, ED symptom-specific SOC measure to determine prevalence of stages across symptoms and identify if SOC predict treatment outcome. Participants [N = 182; age 13-58 years; 92% Caucasian; 96% female; average BMI 21.7 (SD = 5.9); 50% ED not otherwise specified (EDNOS), 30.8% bulimia nervosa (BN), 19.2% anorexia nervosa (AN)] seeking ED treatment at a diverse-milieu multi-disciplinary facility in the United States completed stages of change, behavioral (ED symptom use and frequency) and psychological (ED concerns, anxiety, depression) measures at intake assessment and at 3, 6 and 12 months thereafter. Descriptive summaries were generated using ANOVA or Kruskal-Wallis (continuous) and χ (2) (categorical) tests. Repeated measures linear regression models with autoregressive correlation structure predicted treatment outcome. At intake assessment, 53.3% of AN, 34.0% of BN and 18.1% of EDNOS patients were in Preparation/Action. Readiness to change specific symptoms was highest for binge-eating (57.8%) and vomiting (56.5%). Frequency of fasting and restricting behaviors, and scores on all eating disorder and psychological measures improved over time regardless of SOC at intake assessment. Symptom-specific SOC did not predict reductions in ED symptom frequency. Overall SOC predicted neither improvement in Eating Disorder Examination Questionnaire (EDE-Q) scores nor reduction in depression or trait anxiety; however, higher overall SOC predicted lower state anxiety across follow-up. Readiness to change ED behaviors varies considerably. Most patients reduced eating disorder behaviors and increased psychological functioning regardless of stages of change, indicating the benefits of treatment and effectiveness of treatment-as-usual for overall psychiatric improvement.

General Linear Models in a Missing Outcome Environment of Clinical Trials Incorporating with Splines for Time-Invariant Continuous Adjustment

Missing data is a common occurrence in longitudinal studies of health care research. Although many studies have shown the potential usefulness of current missing analyses, e.g., (1) Complete Case (CC) analysis; (2) imputation methods such as Last Observation Carried Forward (LOCF), multiple imputations, Expectation-Maximization algorithm approach; and (3) methods using all available data such as linear mixed model and generalized estimation equations approach, the CC analysis or LOCF imputation method have been popular due to their simplicity of execution regardless of some critical drawbacks. The proposed approach employs the generalized least squares method using all available data without deletion or imputations for missing outcomes, producing the best linear unbiased estimate. A simulation study was conducted to compare the proposed approach to commonly used missing analyses under each missing data mechanism and showed the validity of the proposed approach, especially with the first order autoregressive correlation structure. B-spline is applied to the proposed model to manage non-linear relationships between outcome and continuous covariate. Application to a cell therapy clinical trial is presented.

Crash frequency modeling using negative binomial models: An application of generalized estimating equation to longitudinal data

The prediction of crash frequency models can be improved when several years of crash data are utilized, instead of three to five years of data most commonly used in research. Crash data, however, generates multiple observations over the years that can be correlated. This temporal correlation affects the estimated coefficients and their variances in commonly used crash frequency models (such as negative binomial (NB), Poisson models, and their generalized forms). Despite the obvious temporal correlation of crashes, research analyses of such correlation have been limited and the consequences of this omission are not completely known. The objective of this study is to explore the effects of temporal correlation in crash frequency models at the highway segment level.In this paper, a negative binomial model has been developed using a generalized estimating equation (GEE) procedure that incorporates the temporal correlations amongst yearly crash counts. The longitudinal model employs an autoregressive correlation structure to compare to the more traditional NB model, which uses a Maximum Likelihood Estimation (MLE) method that cannot accommodate temporal correlations. The GEE model with temporal correlation was found to be superior compared to the MLE model, as it does not underestimate the variance in the coefficient estimates, and it provides more accurate and less biased estimates. Furthermore, an autoregressive correlation structure was found to be an appropriate structure for longitudinal type of data used in this study. Ten years (2002–2011) of Missouri Interstate highway crash data have been utilized in this paper. The crash data comprises of traffic characteristics and geometric properties of highway segments.

Abstract P303: Economic Crisis in Western Europe and Ischemic Heart Disease Mortality

Introduction: The effects of economic crisis, like the current recession affecting Europe since 2008, on health have been widely debated. Our objective was to evaluate the effects of economic changes on mortality trends from ischemic heart disease (IHD) in Europe for the period 1980-2010. Hypothesis: We hypothesize that economic recession is associated with either increases in IHD mortality rates or halting of previous declining trends. Methods: We gathered economic (World Bank) and vital statistics data (WHO Mortality Database) for all Western European countries with more than 5 million inhabitants that had complete mortality data for the period 1980-2010. We defined a period of economic recession as an annual GDP (Gross Domestic Product) growth below 0%. We also computed the cumulative years lived through a given recession period (set to 0 if GDP growth rises above 0% in a given year). We computed age-adjusted (2013 European Standard Population) mortality rates for IHD (ICD9 Codes: 410-414, ICD10 Codes: I20-I25) for the study period and computed yearly changes. We modelled the association between recession and IHD mortality using Generalized Estimating Equations accounting for the correlation among data within country and time using first order autoregressive within-group correlation structures. Changes in ICD coding were introduced as a covariate to account for the shift from ICD-9 to ICD-10. We tested for associations using 0 to 10 year lags. Sensitivity analyses were conducted using unemployment as an alternative indicator of economic recession. Results: Twelve Western European countries were included (Austria, Finland, France, Germany, Greece, Ireland, Netherlands, Norway, Spain, Sweden, Switzerland and the United Kingdom). Economic recession was associated with a 2-year lagged 1.49% increase in IHD mortality in men (p=0.016) and 1.74% increase in women (p=0.019). Each additional year of recession was associated with a 2-year lagged 1.21% increase in IHD mortality in men (p=0.003) and 1.17% increase in women (p=0.016). Sensitivity analyses using unemployment as a second measure of economic recession showed analogous results with strongest associations seen in 0-year lags. GDP and Unemployment were highly correlated with two-year lags between GDP and Unemployment changes, confirming the robustness of these associations. Conclusions: This study showed a deleterious effect of economic recession on ischemic heart disease mortality rates in Europe for the period 1980-2010. Further studies should look into the specific mechanisms of this association and into the possible interaction with policy responses to economic crisis. 1

Role of personality traits in cocaine craving throughout an outpatient psychosocial treatment program

Cocaine dependence is a major international public health concern. Its chronically relapsing nature is possibly related to craving intensity, which can be influenced by diverse biological and psychological aspects. This study aimed to evaluate the role of different personality traits in craving measured throughout a psychosocial treatment program. The sample comprised 66 cocaine-dependent outpatients who were enrolled in an individual and manualized cognitive-behavioral therapy program. The influence of personality traits on craving intensity, frequency, and duration was analyzed using a generalized estimating equations model with an autoregressive correlation structure. Craving varied during treatment. The personality traits of novelty seeking, reward dependence, and harm avoidance interacted with craving intensity, and the personality trait of persistence interacted with craving duration throughout the treatment period. Furthermore, there were significant interactions between drug use and craving intensity, and between different routes of administration and craving intensity. Participants who used cocaine/crack while in treatment and concurrent users of crack (i.e., freebase) cocaine and powder cocaine also had a higher craving intensity. The extent of craving variation can depend on certain personality styles. This study shows that craving is influenced by personality traits, and this may presumably change clinical expression involved in disease.

Kronecker Product Linear Exponent AR(1) Correlation Structures for Multivariate Repeated Measures

Longitudinal imaging studies have moved to the forefront of medical research due to their ability to characterize spatio-temporal features of biological structures across the lifespan. Credible models of the correlations in longitudinal imaging require two or more pattern components. Valid inference requires enough flexibility of the correlation model to allow reasonable fidelity to the true pattern. On the other hand, the existence of computable estimates demands a parsimonious parameterization of the correlation structure. For many one-dimensional spatial or temporal arrays, the linear exponent autoregressive (LEAR) correlation structure meets these two opposing goals in one model. The LEAR structure is a flexible two-parameter correlation model that applies to situations in which the within-subject correlation decreases exponentially in time or space. It allows for an attenuation or acceleration of the exponential decay rate imposed by the commonly used continuous-time AR(1) structure. We propose the Kronecker product LEAR correlation structure for multivariate repeated measures data in which the correlation between measurements for a given subject is induced by two factors (e.g., spatial and temporal dependence). Excellent analytic and numerical properties make the Kronecker product LEAR model a valuable addition to the suite of parsimonious correlation structures for multivariate repeated measures data. Longitudinal medical imaging data of caudate morphology in schizophrenia illustrates the appeal of the Kronecker product LEAR correlation structure.

Comparison of multiplicative heterogeneous variance adjustment models for genetic evaluations

Two heterogeneous variance adjustment methods and two variance models were compared in a simulation study. The method used for heterogeneous variance adjustment in the Nordic test-day model, which is a multiplicative method based on Meuwissen (J. Dairy Sci., 79, 1996, 310), was compared with a restricted multiplicative method where the fixed effects were not scaled. Both methods were tested with two different variance models, one with a herd-year and the other with a herd-year-month random effect. The simulation study was built on two field data sets from Swedish Red dairy cattle herds. For both data sets, 200 herds with test-day observations over a 12-year period were sampled. For one data set, herds were sampled randomly, while for the other, each herd was required to have at least 10 first-calving cows per year. The simulations supported the applicability of both methods and models, but the multiplicative mixed model was more sensitive in the case of small strata sizes. Estimation of variance components for the variance models resulted in different parameter estimates, depending on the applied heterogeneous variance adjustment method and variance model combination. Our analyses showed that the assumption of a first-order autoregressive correlation structure between random-effect levels is reasonable when within-herd heterogeneity is modelled by year classes, but less appropriate for within-herd heterogeneity by month classes. Of the studied alternatives, the multiplicative method and a variance model with a random herd-year effect were found most suitable for the Nordic test-day model for dairy cattle evaluation.

Complications, outcomes, and need for fusion after minimally invasive posterior cervical foraminotomy and microdiscectomy

Background contextPosterior cervical foraminotomy (PCF) with or without microdiscectomy (posterior cervical discectomy [PCD]) is a frequently used surgical technique for cervical radiculopathy secondary to foraminal stenosis or a laterally located herniated disc. Currently, these procedures are being performed with increasing frequency using advanced minimally invasive techniques. Although the safety and efficacy of minimally invasive PCF/PCD (MI-PCF/PCD) have been established, reports on long-term outcome and need for secondary surgical intervention at the index or adjacent level are lacking. PurposeTo determine the rates of complications, long-term outcomes, and need for secondary surgical intervention at the index or adjacent level after MI-PCF and microdiscectomy. Study designRetrospective analysis of a prospective cohort. Patient sampleSeventy patients treated with MI-PCF and/or MI-PCD for cervical radiculopathy. Outcome measuresVisual Analog Scale for neck/arm (VASN/A) pain and Neck Disability Index (NDI). MethodsNinety-seven patients underwent MI-PCF with or without MI-PCD between 2002 and 2011. Adequate prospective follow-up was available for 70 patients (95 cervical levels). The primary outcome assessed was need for secondary surgical intervention at the index or adjacent level. The secondary outcomes assessed included complications and improvements in NDI and VASN/A scores. All complications were reviewed. Mixed-model analyses of variance with random subject effects and autoregressive first-order correlation structures were used to test for differences among NDI, VASA, and VASN measurements made over time while accounting for the correlation among repeated observations within a patient. All statistical hypothesis tests were conducted at the 5% level of significance. ResultsPatients were followed for a mean of 32.1 months. Of 70 patients operated, there were 3 (4.3%) complications (1 cerebrospinal fluid leak, 1 postoperative wound hematoma, and 1 radiculitis), none of which required a secondary operative intervention. Five patients required an anterior cervical discectomy and fusion (eight total levels fused) on average 44.4 months after the index surgery. Of those, five (5.3%) were at the index level and three (2.1%) were at adjacent levels. Neck Disability Index scores improved significantly (p<.0001) immediately postoperatively and continued to decrease gradually with time. Visual Analog Scale for neck/arm scores improved significantly (p<.0001) from baseline immediately postoperatively but tended to plateau with time. ConclusionsMinimally invasive PCF with or without MI-PCD is an excellent alternative for cervical radiculopathy secondary to foraminal stenosis or a laterally located herniated disc. There is a low rate (1.1% per index level per year) of future index site fusion and a very low rate (0.9% per adjacent level per year) of adjacent-level disease requiring surgery.

Effect of carryover and presampling procedures on the results of real-time PCR used for diagnosis of bovine intramammary infections with Streptococcus agalactiae at routine milk recordings

The use of PCR tests as diagnostics for intramammary infections (IMI) based on composite milk samples collected in a non-sterile manner at milk recordings is increasing. Carryover of sample material between cows and non-aseptic PCR sampling may be incriminated for misclassification of IMI with Streptococcus agalactiae (S. agalactiae) in dairy herds with conventional milking parlours. Misclassification may result in unnecessary costs for treatment and culling. The objectives of this study were to (1) determine the effect of carryover on PCR-positivity for S. agalactiae at different PCR cycle threshold (Ct) cut-offs by estimating the between-cow correlation while accounting for the milking order, and (2) evaluate the effect of aseptic presampling procedures (PSP) on PCR-positivity at the different Ct-value cut-offs.The study was conducted in four herds with conventional milking parlours at routine milk recordings. Following the farmers’ routine pre-milking preparation, 411 of 794 cows were randomly selected for the PSP treatment. These procedures included removing the first streams of milk and 70% alcohol teat disinfection. Composite milk samples were then collected from all cows and tested using PCR. Data on milking order were used to estimate the correlation between consecutively milked cows in each milking unit. Factors associated with the PCR-positivity for S. agalactiae were analyzed using generalized estimating equations assuming a binomially-distributed outcome with a logit link function. Presampling procedures were only significant using cut-off 37. A first-order autoregressive correlation structure provided the best correlation between consecutively milked cows. The correlation was 13%, 11%, 9% at cut-offs <40, 37, and 34, respectively. PSP did not reduce the odds of cows being PCR-positive for S. agalactiae.In conclusion, carryover and non-aseptic sampling affected the PCR results and should therefore be considered when samples from routine milk recordings are used. In relative terms, higher cut-offs resulted in higher between-cow correlation, but the absolute amount of carryover may not be affected although this was not tested.

Pathogen-specific effects on milk yield in repeated clinical mastitis episodes in Holstein dairy cows

The objective of this study was to estimate the effects of clinical mastitis (CM) cases due to different pathogens on milk yield in Holstein cows. The first 3 CM cases in a cow’s lactation were modeled. Eight categories of pathogens were included: Streptococcus spp.; Staphylococcus aureus; coagulase-negative staphylococci (CNS); Escherichia coli; Klebsiella spp.; cases with CM signs but no bacterial growth (above the level detectable by our microbiological procedures) observed in the culture sample, and cases with contamination (≥3 pathogens in the sample); other pathogens that may be treated with antibiotics (included Citrobacter, Corynebacterium bovis, Enterobacter, Enterococcus, Pasteurella, Pseudomonas; “other treatable”); and other pathogens not successfully treated with antibiotics (Trueperella pyogenes, Mycoplasma, Prototheca, yeasts; “other not treatable”). Data from 38,276 lactations in cows from 5 New York State dairy herds, collected from 2003–2004 until 2011, were analyzed. Mixed models with an autoregressive correlation structure (to account for correlation among the repeated measures of milk yield within a lactation) were estimated. Primiparous (lactation 1) and multiparous (lactations 2 and 3) cows were analyzed separately, as the shapes of their lactation curves differed. Primiparas were followed for up to 48wk of lactation and multiparas for up to 44 wk. Fixed effects included parity, calving season, week of lactation, CM (type, case number, and timing of CM in relation to milk production cycle), and other diseases (milk fever, retained placenta, metritis, ketosis, displaced abomasum). Herd was modeled as a random effect. Clinical mastitis was more common in multiparas than in primiparas. In primiparas, Streptococcus spp. occurred most frequently as the first case. In multiparas, E. coli was most common as the first case. In subsequent cases, CM cases with no specific growth or contamination were most common in both parity groups. The hazard of CM increased with case number. Mastitic cows were generally higher producers before the CM episode than their nonmastitic herdmates. Milk loss varied with pathogen and case number. In primiparas, the greatest losses were associated with E. coli and “other not treatable” organisms. In multiparas, the greatest losses were associated with Klebsiella spp. and “other not treatable” organisms. Milk loss was not associated with occurrence of CNS. The findings may help farmers to make optimal management decisions for their cows.

Vitamin D Level and Its Correlation With Hemoglobin In Pediatric Sickle Cell Disease Patients

Introduction It has been well characterized that people living with sickle cell disease are more likely than the general population to be vitamin D deficient. New literature suggests there is an association between vitamin D deficiency and increased anemia in patients with chronic anemia. To date, this has been examined only in patients with end stage renal disease, heart failure, and anemia of other chronic diseases, but not in sickle cell disease, a chronic inflammatory disease with an elevation in inflammatory markers. Vitamin D has been shown to increase erythroid precursor proliferation by increasing erythropoietin sensitivity. It is also thought that vitamin D has pleiotropic effects, which may have influence on the bone marrow. Additionally, vitamin D has been shown to suppress pro-inflammatory cytokines, which causes a decrease in the anemia of chronic inflammation. Severe vitamin D deficiency also causes hyperparathyroidism, which leads to an inhibition of endogenous production of erythropoietin. Our hypothesis implies supplementation of vitamin D will improve anemia of vitamin D deficient sickle cell patients. Methods We utilized laboratory and demographic data from the Bronx Lebanon Hospital Center's patient electronic medical record between November 2009 and July 2013. Our sample included a population of 50 sickle cell disease patients aged 0 to 21 who were vitamin D deficient (serum 25-hydroxyvitamin D (25-OHD) &lt;30ng/ml). Sickle cell disease types included SS (64%), SC (32%), Sβ+ thalassemia and Sβ0 thalassemia (4%). Using a linear mixed model with an autoregressive correlation structure to account for variance within the subject, we examined the association between time dependent 25-OHD level (ng/ml) and hemoglobin in (g/dl), as well as between 25-OHD and reticulocyte percentage over time in days. We began at the time of first supplementation of 25-OHD to the most recent blood levels in each patient. Results After adjusting for mean corpuscular volume (MCV), hemoglobin, sickle cell type and hydroxyurea (Table 1), there was a linear increase in reticulocyte percentage over time associated with increasing 25-OHD levels (β = 0.000060, SE = 0.000030, p =0.050). However, after controlling for MCV, gender, race, sickle cell type, and hydroxyurea (Table 2), we found trend of a negative association between 25-OHD and hemoglobin levels, which was statistically significant (β = -0.000017, SE = 0.000007, p =0.017). Conclusion Our results did not support our original hypothesis, which stated that supplementation of vitamin D in vitamin D deficient sickle cell patients would lead to an improvement in anemia. However, a trend was noted of a decrease in hemoglobin with increasing 25-OHD. In turn, there was a statistically significant increase in reticulocyte percentage with increasing 25-OHD. This is suggestive that there is increased hemolysis with increased erythropoiesis caused by increasing 25-OHD. This is seen even when controlling for patients taking hydroxyurea. Further study is required to elucidate this possible correlation. It is important to examine this relationship further, as supplementing vitamin D in vitamin D deficient sickle cell disease patients is becoming standard of care for improvement in bone density. It is unclear what potential implications this may have on increased production of sickle cells from the bone marrow, including a possible increase in hemolysis and a worsened anemia. Disclosures: Seelaboyina: Bronx Lebanon Hospital: Employment.

A default prior distribution for contingency tables with dependent factor levels

A default prior distribution is proposed for the Bayesian analysis of contingency tables. The prior is specified to allow for dependence between levels of the factors. Different dependence structures are considered, including conditional autoregressive and distance correlation structures. To demonstrate the prior distribution, a dataset is considered which involves estimating the number of injecting drug users in the eleven National Health Service board regions of Scotland using an incomplete contingency table where the dependence structure relates to geographical regions.

Assessment of Heterogeneity of Efficacy of a Three-Dose Regimen of a Type III Secreted Protein Vaccine for Reducing STEC O157 in Feces of Feedlot Cattle

Preharvest control of Escherichia coli O157:H7 (STEC O157) may prevent human illness by reducing the presence of STEC O157 throughout the beef production chain. Immunization of cattle with a type III secreted protein vaccine inhibits colonization of cattle with STEC O157 and reduces the probability of fecal shedding and hide contamination. Our objectives were to perform a meta-analysis to estimate efficacy of a three-dose regimen of TTSP vaccine at reducing the presence of STEC O157 in the feces of feedlot cattle and to test factors that might modify vaccine efficacy. Pen-level data (n=184 pens, 1462 cattle) from four randomized controlled vaccine trials conducted from 2002 to 2008 at the University of Nebraska-Lincoln were analyzed. Factors explaining a culture-positive fecal sample were tested in generalized estimating equations logistic regression and log-binomial models. An autoregressive correlation structure was defined to account for clustering of repeated test-periods within block. Clustering or potential confounding by study was accounted for by treating study as a fixed effect. STEC O157 was detected from 661 of 5451 postvaccination fecal samples. The probability to detect STEC O157 postvaccination was 8.4% and 15.8% in vaccinated and unvaccinated cattle, respectively. Interactions between vaccination and (1) study; (2) prevalence of control pens within each time-place cluster; and (3) days from vaccination were not significant or fit poorly with observed data. Adjusting for study, cattle in pens receiving three doses of vaccine were less likely to shed STEC O157 (odds ratio=0.46, p<0.0001). Model-adjusted vaccine efficacy was 48% (95% confidence interval: 0.37-0.57). We concluded that a three-dose regimen type III secreted protein vaccine was efficacious at reducing the probability of detecting STEC O157 in the feces of cattle and that vaccine efficacy was not modified by study or level of prevalence observed in control pens.

Fitting Logistic Growth Curve with Nonlinear Mixed-effects Models

A nonlinear mixed-effects modeling approach was used to model individual tree diameter increment based on Logistic growth function for dahurian larch (Larix gmelinii Rupr.) plantations in northeastern China. The study involved the estimation of fixed and random parameters, as well as procedures for determining random effects variance-covariance matrices. Results showed that the mixed-effects model provided better model fitting than the fixed-effects model. The logistic model with three random parameters b₁, b₂, b₃ was considered the best mixed model. Time series correlation structures included Autoregressive correlation structure AR (1) and AR (2), Moving Average correlation structure MA (1) and MA (2) and Autoregressive-Moving Average correlation structure [ARMA (1, 1)] and ARMA (2, 2) were incorporated into the best mixed model. The mixed model with MA (2) correlation structure showed lower AIC and BIC values and significantly improved model performance (LRT = 545.6, p<0.0001). Techniques for calibrating the diameter growth model for a particular tree of interest were also explored. The results indicated that the mixed-effects model provided better diameter predictions than the models using only fixed-effects parameters.

Nonrespiratory symptoms before loss of asthma control in children

Nonrespiratory symptoms before loss of asthma control in children

Repeated measures and longitudinal data analysis using higher‐order entropies

This paper implements the generalized maximum entropy (GME) method in longitudinal data setup to investigate the regression para meters and correlation among the repeated measurements. We derive the GME system using Shannon classical entropy as well as some higher‐order entropies assuming an autoregressive correlation structure. This method is illustrated using two simulated examples to study the effect of changing the support range and compare the performance of the GME approach with the classical estimation methods.

Impact of diabetes on long-term mortality following multivessel percutaneous interventions: An insight into optimal statistical analysis

BackgroundSeveral studies have demonstrated better long-term outcomes with drug eluting stents (DES) as compared to bare metal stents (BMS) among diabetics with coronary artery disease (CAD). A significant heterogeneity exists with respect to the optimal statistical strategy to analyze stent related data. MethodsWe used our percutaneous intervention (PCI) registry to identify all diabetics with CAD, who underwent PCI on two or more vessel territories between 2003 and 2009. Long-term mortality was assessed using the social security death index. Six different analytical strategies were applied. ResultsA total of 1568 DES and 336 BMS interventions were encountered in 756 diabetics. Considerable differences were observed in the results between the methods applied. Generalized estimating equation (GEE) approach with an autoregressive correlation structure (GEE) was a robust method to account for the cluster structure, since the measurements taken through time on the same person were assumed to be highly correlated, if they were spaced more closely in time. Diabetics undergoing PCI with BMS had a significantly higher long-term mortality as compared to the patients undergoing DES-PCI [Hazard ratio (95% CI): 1.47 (1.04–2.09)]. ConclusionThere is a great potential for erroneous interpretation of PCI data due to complex spatial and temporal clustering. Use of GEE with autoregressive correlation matrix and robust variance is most optimal to account for the clustered nature of the PCI related data. Using GEE, we observed that there is a 47% (4%–119%) higher hazard for mortality among diabetics undergoing BMS-PCI as compared to diabetics undergoing DES-PCI.

Analyzing temporally correlated dolphin sightings data using generalized estimating equations

AbstractMany of the statistical techniques commonly used in ecology assume independence among responses. However, there are many marine mammal survey techniques, such as those involving time series or subgroups, which result in correlations within the data. Generalized estimating equations (GEEs) take such correlations into account and are an extension of generalized linear models. This study demonstrates the application of GEEs by modeling temporal variation in bottlenose dolphin presence from sightings data. Since dolphins could remain in the study area for several hours resulting in temporal autocorrelation, an autoregressive correlation structure was used within the GEE, each cluster representing hours within a day of survey effort. The results of the GEE model showed that there was significant diel, tidal, and interannual variation in the presence of dolphins. Dolphins were most likely to be seen in the early morning and during the summer months. Dolphin presence generally peaked during low tide, but this varied among years. There was a significantly lower probability of dolphins being present in 2003 than 2004, but not between 2004 and the other years (1991, 1992, and 2002). GEE‐model fitting packages are now readily available, making this a valuable, versatile tool for marine mammal biologists.

LUNG INFLAMMATION AMONG CHILDREN WITH ASTHMA EXPOSED TO INDUSTRIAL AND TRAFFIC POLLUTION

Background and Aims: Children with asthma are susceptible to the effects of ambient air pollution, including increased symptoms and lung inflammation. Industrial emissions are an important source of air pollution in many areas. Oil refinery emissions are of particular concern as they comprise a complex mixture of organic and inorganic pollutants. We examined associations between fractional exhaled nitric oxide (FeNO), an indicator of airway inflammation, and exposure to oil refinery and traffic emissions on the health of children with asthma in Montreal, Canada. Methods: We recruited 69 children (age 8-13 years) with a physician diagnosis of asthma from schools and an asthma clinic. Subjects participated in the panel study for 10 consecutive days between October 2009 and April 2010. We measured personal exposures to sulphur dioxide (SO2) using Ogawa passive samplers (Ogawa &amp; Company, Pompano Beach, FL, USA) and fine particulate matter (PM2.5) using Harvard Personal Environmental Monitors (HPEM, BGI, MA, USA) and continuously using the personal DataRAM (pDR-1200, MIE Inc, Bedford, MA). Filters were also analyzed for metals associated with oil refinery emissions. We recorded online FeNO daily using the NIOX MINO monitor (Aerocrine, Solna, Sweden) and collected participants’ reports on health, medication use, and activities. Linear mixed-effects regression models with autoregressive correlation structure were used to estimate the association between FeNO and pollutant exposure. Results: The geometric mean of FeNO was 20.9 ppb (geometric standard deviation: 2.2). Mean (SD) personal exposure to PM2.5 was 9.5 (13.4) µg/m3, while for SO2 it was 0.81 (3.21) ppb. Preliminary models indicate that an increase of 10 µg/m3 in previous 8-hour personal exposure to PM2.5 was associated with a 1% (95% CI: 0.1-2.0%) increase in FeNO, adjusted for corticosteroid use, age and sex. Conclusions: Preliminary results indicate an association between personal exposures to PM2.5 and increased airway inflammation in children with asthma.