Multiple Linear Mixed Models Research Articles

Iterative Usage of Fixed and Random Effect Models for Powerful and Efficient Genome-Wide Association Studies.

False positives in a Genome-Wide Association Study (GWAS) can be effectively controlled by a fixed effect and random effect Mixed Linear Model (MLM) that incorporates population structure and kinship among individuals to adjust association tests on markers; however, the adjustment also compromises true positives. The modified MLM method, Multiple Loci Linear Mixed Model (MLMM), incorporates multiple markers simultaneously as covariates in a stepwise MLM to partially remove the confounding between testing markers and kinship. To completely eliminate the confounding, we divided MLMM into two parts: Fixed Effect Model (FEM) and a Random Effect Model (REM) and use them iteratively. FEM contains testing markers, one at a time, and multiple associated markers as covariates to control false positives. To avoid model over-fitting problem in FEM, the associated markers are estimated in REM by using them to define kinship. The P values of testing markers and the associated markers are unified at each iteration. We named the new method as Fixed and random model Circulating Probability Unification (FarmCPU). Both real and simulated data analyses demonstrated that FarmCPU improves statistical power compared to current methods. Additional benefits include an efficient computing time that is linear to both number of individuals and number of markers. Now, a dataset with half million individuals and half million markers can be analyzed within three days.

Factors impacting unbound vancomycin concentrations in different patient populations.

The unbound drug hypothesis states that only unbound drug concentrations are active and available for clearance, and highly variable results regarding unbound vancomycin fractions have been reported in the literature. We have determined the unbound vancomycin fractions in four different patient groups by a liquid chromatography tandem mass spectrometry (LC-MS/MS) method and identified factors that modulate vancomycin binding. We have further developed and validated a prediction model to estimate unbound vancomycin concentrations. Vancomycin (unbound and total) concentrations were measured in 90 patients in four different hospital wards (hematology [n = 33 samples], intensive care unit [ICU] [n = 51], orthopedics [n = 44], and pediatrics [age range, 6 months to 14 years; n = 18]) by a validated LC-MS/MS method. Multiple linear mixed model analysis was performed to identify patient variables that were predictive of unbound vancomycin fractions and concentrations. The variables included in the model were patient age, ward, number of coadministered drugs with high protein binding, kidney function (estimated glomerular filtration rate [determined by Chronic Kidney Disease Epidemiology Collaboration formula]), alpha-1-acid glycoprotein, albumin, total bilirubin, IgA, IgM, urea, and total vancomycin concentrations. In the pediatric cohort, the median unbound vancomycin fraction was 81.3% (range, 61.9 to 95.9%), which was significantly higher (P < 0.01) than the unbound fraction found in the three adult patient cohorts (hematology, 60.6% [48.7 to 90.6%]; ICU, 61.7% [47.0 to 87.6%]; orthopedics, 56.4% [45.9 to 78.0%]). The strongest significant predictor of the unbound vancomycin concentration was the total drug concentration, completed by albumin in the pediatric cohort and albumin and IgA in the adult cohorts. Validation of our model was performed with data from 13 adult patients. A mean difference of 0.3 mg/liter (95% confidence interval [CI], -1.3 to 0.7 mg/liter; R(2) = 0.99 [95% CI, 0.95 to 0.99]) between measured and calculated unbound vancomycin concentrations demonstrated that the predictive performance of our model was favorable. Unbound vancomycin fractions vary significantly between pediatric and adult patients. We developed a formula to estimate the unbound fraction derived from total vancomycin, albumin, and IgA concentrations in adult patients.

Genome-wide association study on progression of carotid artery intima media thickness over 10 years in a Chinese cohort

BackgroundCarotid artery intima media thickness (IMT) in human is a marker of subclinical atherosclerosis with high heritability. Many genome-wide association studies (GWAS) were performed in European and American populations, yet discovery efforts have been limited in Asians. ObjectiveTo identify the genetic determinants of 10-year progression of IMT using GWA approach in a Chinese cohort. MethodsCardiovascular epidemiologic survey was carried out in 810 Chinese adults in 2012. 302,218 single-nucleotide polymorphisms (SNP) in whole genome were genotyped using gene chip and carotid IMT was measured. Most of these participants, had previous carotid IMT measurements in 2002 (n = 572), 2005 (n = 750), 2007 (n = 747), and 2010 (n = 671). General linear model (GLM) and multiple linear mixed-model (MLM) were used to assess the association between SNPs and carotid IMT. ResultsThe mean age (SD) of the sample was 61.3 (5.1) years; 33.6% were men. The adjusted GLM showed no SNP with significance association at genome-level (all p > 1 × 10−7). However, using MLM, after adjusting for age, sex, number of cigarettes smoked per day, systolic blood pressure, use of antihypertensive drugs in the past 2 weeks, serum cholesterol, body mass index, fasting glucose levels, use of insulin or hypoglycemic drugs, time of measuring IMT and its interaction with SNP, we identified two novel SNPs (rs36071027 in EBF1 gene on chromosome 5 and rs975809 close to PCDH15 gene on chromosome 10) that are significantly associated with carotid IMT at genome level (p < 1 × 10−7) and seven novel SNPs (rs2230307 in AGL gene on chromosome 1, rs12040273 in GALNT2 gene on chromosome 1, rs4536103 in NEUROG3 gene on chromosome 10, rs9855415 in LOC647323 gene on chromosome 3, rs2472647 in PCDHGA1 gene on chromosome 5, rs17433780 in GBP3 gene on chromosome 1, and rs7625806 in DLEC1 gene on chromosome 3) which are suggestive of significant association (p < 10−5). ConclusionThe study represents the first GWAS of association between SNPs and carotid IMT in an Asian population. We identified 2 novel loci associated with carotid IMT progression over 10 years.

Prenatal Ambient Air Pollution, Placental Mitochondrial DNA Content, and Birth Weight in the INMA (Spain) and ENVIRONAGE (Belgium) Birth Cohorts.

Background:Mitochondria are sensitive to environmental toxicants due to their lack of repair capacity. Changes in mitochondrial DNA (mtDNA) content may represent a biologically relevant intermediate outcome in mechanisms linking air pollution and fetal growth restriction.Objective:We investigated whether placental mtDNA content is a possible mediator of the association between prenatal nitrogen dioxide (NO2) exposure and birth weight.Methods:We used data from two independent European cohorts: INMA (n = 376; Spain) and ENVIRONAGE (n = 550; Belgium). Relative placental mtDNA content was determined as the ratio of two mitochondrial genes (MT-ND1 and MTF3212/R3319) to two control genes (RPLP0 and ACTB). Effect estimates for individual cohorts and the pooled data set were calculated using multiple linear regression and mixed models. We also performed a mediation analysis.Results:Pooled estimates indicated that a 10-μg/m3 increment in average NO2 exposure during pregnancy was associated with a 4.9% decrease in placental mtDNA content (95% CI: –9.3, –0.3%) and a 48-g decrease (95% CI: –87, –9 g) in birth weight. However, the association with birth weight was significant for INMA (–66 g; 95% CI: –111, –23 g) but not for ENVIRONAGE (–20 g; 95% CI: –101, 62 g). Placental mtDNA content was associated with significantly higher mean birth weight (pooled analysis, interquartile range increase: 140 g; 95% CI: 43, 237 g). Mediation analysis estimates, which were derived for the INMA cohort only, suggested that 10% (95% CI: 6.6, 13.0 g) of the association between prenatal NO2 and birth weight was mediated by changes in placental mtDNA content.Conclusion:Our results suggest that mtDNA content can be one of the potential mediators of the association between prenatal air pollution exposure and birth weight.Citation:Clemente DB, Casas M, Vilahur N, Begiristain H, Bustamante M, Carsin AE, Fernández MF, Fierens F, Gyselaers W, Iñiguez C, Janssen BG, Lefebvre W, Llop S, Olea N, Pedersen M, Pieters N, Santa Marina L, Souto A, Tardón A, Vanpoucke C, Vrijheid M, Sunyer J, Nawrot TS. 2016. Prenatal ambient air pollution, placental mitochondrial DNA content, and birth weight in the INMA (Spain) and ENVIRONAGE (Belgium) birth cohorts. Environ Health Perspect 124:659–665; http://dx.doi.org/10.1289/ehp.1408981

Influence of plant community composition on biomass production in planted grasslands.

United States energy policy mandates increased use of renewable fuels. Restoring grasslands could contribute to a portion of this requirement through biomass harvest for bioenergy use. We investigated which plant community characteristics are associated with differences in biomass yield from a range of realistic native prairie plantings (n = 11; i.e., conservation planting, restoration, and wildlife cover). Our primary goal was to understand whether patterns in plant community composition and the Floristic Quality Index (FQI) were related to productivity as evidenced by dormant season biomass yield. FQI is an objective measure of how closely a plant community represents that of a pre-European settlement community. Our research was conducted in planted fields of native tallgrass prairie species, and provided a gradient in floristic quality index, species richness, species diversity, and species evenness in south-central Wisconsin during 2008 and 2009. We used a network of 15 randomly located 1 m2 plots within each field to characterize the plant community and estimate biomass yield by clipping the plots at the end of each growing season. While plant community composition and diversity varied significantly by planting type, biomass yield did not vary significantly among planting types (ANOVA; P >0.05). Biomass yield was positively correlated with plant community evenness, richness, C4 grass cover, and floristic quality index, but negatively correlated with plant species diversity in our multi-season multiple linear mixed effects models. Concordantly, plots with biomass yield in the lowest quartile (biomass yield < 3500 kh/ha) had 8% lower plant community evenness and 9% lower FQI scores than those in the upper quartile (biomass yield > 5800 kh/ha). Our results suggest that promoting the establishment of fields with high species evenness and floristic quality may increase biomass yield, while simultaneously supporting biodiversity.

Identifying the environmental conditions favouring West Nile Virus outbreaks in Europe.

West Nile Virus (WNV) is a globally important mosquito borne virus, with significant implications for human and animal health. The emergence and spread of new lineages, and increased pathogenicity, is the cause of escalating public health concern. Pinpointing the environmental conditions that favour WNV circulation and transmission to humans is challenging, due both to the complexity of its biological cycle, and the under-diagnosis and reporting of epidemiological data. Here, we used remote sensing and GIS to enable collation of multiple types of environmental data over a continental spatial scale, in order to model annual West Nile Fever (WNF) incidence across Europe and neighbouring countries. Multi-model selection and inference were used to gain a consensus from multiple linear mixed models. Climate and landscape were key predictors of WNF outbreaks (specifically, high precipitation in late winter/early spring, high summer temperatures, summer drought, occurrence of irrigated croplands and highly fragmented forests). Identification of the environmental conditions associated with WNF outbreaks is key to enabling public health bodies to properly focus surveillance and mitigation of West Nile virus impact, but more work needs to be done to enable accurate predictions of WNF risk.

Do more detailed environmental covariates deliver more accurate soil maps?

In this study we evaluated whether investing in more spatially detailed environmental covariates improves the accuracy of digital soil maps. We used a case study from Southern Brazil to map clay content (CLAY), organic carbon content (SOC), and effective cation exchange capacity (ECEC) of the topsoil for a~2000ha area located on the edge of the plateau of the Paraná Sedimentary Basin. Five covariates, each with two levels of spatial detail were used: area-class soil maps, digital elevation models (DEM), geologic maps, land use maps, and satellite images. Thirty-two multiple linear regression models were calibrated for each soil property using all spatial detail combinations of the covariates. For each combination, stepwise regression was used to select predictor variables incorporated in the model. Model evaluation was done using the adjusted R-square of the regression. The baseline model, calibrated with the less detailed version of each covariate, and the best performing model were used to calibrate two linear mixed models for each soil property. Model parameters were estimated using restricted maximum likelihood. Spatial prediction was performed using the empirical best linear unbiased predictor. Validation of baseline and best performing linear multiple regression and linear mixed models was done using cross-validation. Results show that for CLAY the prediction accuracy did not considerably improve by using more detailed covariates. The amount of variance explained increased only ~2 percentage points (pp), less than that obtained by including the kriging step, which explained 4pp. On the other hand, prediction of SOC and ECEC improved by ~13pp when the baseline model was replaced by the best performing model. Overall, the increase in prediction performance was modest and may not outweigh the extra costs of using more detailed covariates. It may be more efficient to spend extra resources on collecting more soil observations, or increasing the detail of only those covariates that have the strongest improvement effect. In our case study, the latter would only work for SOC and ECEC, by investing in a more detailed land use map and possibly also a more detailed geologic map and DEM.

Relationship between exposure to fine particulates and ozone and reduced lung function in children

BackgroundA limited number of studies have reported an association between long-term exposure to ambient air pollutants and lung function growth among children, with inconclusive results. ObjectivesTo assess the relationship between air pollutant exposure and lung function growth, and to examine potential sex differences in the susceptibility of lung function growth to air pollution. MethodsWe conducted a two-year prospective cohort study of Taiwanese children aged 12 at baseline who were followed from October 1, 2007 to November 31, 2009. The study population comprised 2941 non-smoking children who completed pulmonary function tests at both baseline and follow-up surveys. We applied spatial modeling for individual-level exposure assessment to capture relevant exposures and also attempted to eliminate potential community-level confounding. The exposure parameters were annual averages and values calculated from 24-hourly PM2.5 and 8-hourly ozone (O3) concentrations, corresponding to the residential addresses over the study period. The effect estimates were presented as lung function growth deficits per interquartile range (IQR) for PM2.5 and O3. ResultsIn a multiple linear mixed effect model, adjusted for confounding, growth deficits in the forced vital capacity (FVC), forced expiration volume in 1s (FEV1), and forced expiratory flow between the 25th and 75th percentiles of the FVC were associated with increased exposure to PM2.5 and O3. For example, greater exposure to PM2.5 (IQR, 17.92μg/m3) was associated with an annual deficit in FVC growth of 75mL in boys and 61mL in girls (p for interaction <0.05). Similar associations were found for O3. ConclusionsThe study provides evidence that long-term exposure to PM2.5 and O3 may have a detrimental effect on the development of lung function in children. The estimated deficits were generally larger in boys, compared to girls.

Development of the Lémann Index to Assess Digestive Tract Damage in Patients With Crohn's Disease

There is a need for a scoring system that provides a comprehensive assessment of structural bowel damage, including stricturing lesions, penetrating lesions, and surgical resection, for measuring disease progression. We developed the Lémann Index and assessed its ability to measure cumulative structural bowel damage in patients with Crohn's disease (CD). We performed a prospective, multicenter, international, cross-sectional study of patients with CD evaluated at 24 centers in 15 countries. Inclusions were stratified based on CD location and duration. All patients underwent clinical examination and abdominal magnetic resonance imaging analyses. Upper endoscopy, colonoscopy, and pelvic magnetic resonance imaging analyses were performed according to suspected disease locations. The digestive tract was divided into 4 organs and subsequently into segments. For each segment, investigators collected information on previous operations, predefined strictures, and/or penetrating lesions of maximal severity (grades 1-3), and then provided damage evaluations ranging from 0.0 (no lesion) to 10.0 (complete resection). Overall level of organ damage was calculated from the average of segmental damage. Investigators provided a global damage evaluation (from 0.0 to 10.0) using calculated organ damage evaluations. Predicted organ indexes and Lémann Index were constructed using a multiple linear mixed model, showing the best fit with investigator organ and global damage evaluations, respectively. An internal cross-validation was performed using bootstrap methods. Data from 138 patients (24, 115, 92, and 59 with upper tract, small bowel, colon/rectum, and anus CD location, respectively) were analyzed. According to validation, the unbiased correlation coefficients between predicted indexes and investigator damage evaluations were 0.85, 0.98, 0.90, 0.82 for upper tract, small bowel, colon/rectum, anus, respectively, and 0.84 overall. In a cross-sectional study, we assessed the ability of the Lémann Index to measure cumulative structural bowel damage in patients with CD. Provided further successful validation and good sensitivity to change, the index should be used to evaluate progression of CD and efficacy of treatment.

Demographic characteristics and food choices of participants in the Special Diabetes Program for American Indians Diabetes Prevention Demonstration Project

Objective. American Indians and Alaska Natives (AI/ANs) suffer a disproportionate burden of diabetes. Identifying food choices of AI/ANs at risk of type 2 diabetes, living in both rural and urban settings, is critical to the development of culturally relevant, evidence-based education strategies designed to reduce morbidity and mortality in this population.Design. At baseline, 3135 AI/AN adults participating in the Special Diabetes Program for American Indians Diabetes Prevention Demonstration Project (SDPI-DP) completed a socio-demographic survey and a 27-item food frequency questionnaire (FFQ). The primary dietary behavior goal of SDPI-DP education sessions and lifestyle coaching is changes in food choices, i.e., increased fruits, vegetables and whole grains, decreased high sugar beverages, red meat, and processed foods. Subsequently, program assessment focuses on changes in food types. Foods were delineated using a ‘healthy’ and ‘unhealthy’ classification as defined by the educators advising participants. Urban and rural differences were examined using χ2 tests and two sample t-tests. Multiple linear regressions and linear mixed models were used to assess the association between socio-demographic factors and food choice.Results. Retired participants, those living in urban areas and with high income and education selected healthy foods most frequently. Young males, those with low income and education consumed unhealthy foods most frequently. Selection of unhealthy foods did not differ by urban and rural setting.Conclusions. The ubiquitous nature of unhealthy food choices makes them hard to avoid. Food choice differences by gender, age, income, and setting suggest that nutrition education should more effectively target and meets the needs of young AI/AN males.

A longitudinal study of gross motor coordination and weight status in children

This longitudinal study investigated the interrelationship between children's weight status and level of gross motor coordination over time, taking baseline physical activity (PA) into account as a possible mediator. Baseline measurements were collected in 2517 children (5-13 years, 52.8% boys), including (1) body height and weight to calculate body mass index (BMI) z-scores, (2) gross motor coordination using the Körperkoordinationstest für Kinder (KTK), (3) total PA estimated by a questionnaire. At follow-up, 754 participants (7-13 years, 50.8% boys) underwent anthropometric and KTK assessments again. Two hypothesized partial mediation models (i.e., KTK ↔ PA ↔ BMI z-score) were examined by multiple linear mixed models. A lower performance on the KTK at baseline significantly predicted an increase in BMI z-score (B = -0.003, P = 0.027). Conversely, a higher baseline BMI z-score also predicted a decrease in KTK performance (B = -1.792, P < 0.001). Since total PA at baseline was not significantly related to initial KTK performance (B = 1.628, P = 0.134) nor BMI z-score (B = 25.312, P = 0.130), its mediating effect was not further explored. Our results strongly suggest that children's weight status negatively influences future level of gross motor coordination, and vice versa. Prevention and intervention initiatives should consider this reciprocal causal relationship across developmental time.

Inflammation and echocardiographic parameters of ventricular hypertrophy in a cohort with preserved cardiac function

ObjectiveTo investigate the association between inflammation and selective echocardiographic parameters (EP) characteristic for ventricular hypertrophy in cross-sectional and longitudinal population-based analyses.MethodsBaseline (711 men, 659 women: 45–83 years) and 4-year follow-up...

Risk factors associated with increased mortality of farmed Pacific oysters in Ireland during 2011

The Pacific oyster, Crassostrea gigas, plays a significant role in the aquaculture industry in Ireland. Episodes of increased mortality in C. gigas have been described in many countries, and in Ireland since 2008. The cause of mortality events in C. gigas spat and larvae is suspected to be multifactorial, with ostreid herpesvirus 1 (OsHV-1, in particular OsHV-1μvar) considered a necessary, but not sufficient, cause. The objectives of the current study were to describe mortality events that occurred in C. gigas in Ireland during the summer of 2011 and to identify any associated environmental, husbandry and oyster endogenous factors. A prospective cohort study was conducted during 2010–2012, involving 80 study batches, located at 24 sites within 17 bays. All 17 bays had previously tested positive for OsHV-1μvar. All study farmers were initially surveyed to gather relevant data on each study batch, which was then tracked from placement in the bay to first grading. The outcome of interest was cumulative batch-level mortality (%). Environmental data at high and low mortality sites were compared, and a risk factor analysis, using a multiple linear regression mixed effects model, was conducted. Cumulative batch mortality ranged from 2% to 100% (median=16%, interquartile range: 10–34%). The final multivariable risk factor model indicated that batches imported from French hatcheries had significantly lower mortalities than non-French hatcheries; sites which tested negative for OsHV-1μvar during the study had significantly lower mortalities than sites which tested positive and mortalities increased with temperature until a peak was reached. There were several differences between the seed stocks from French and non-French hatcheries, including prior OsHV-1μvar exposure and ploidy. A range of risk factors relating to farm management were also considered, but were not found significant. The relative importance of prior OsHV-1μvar infection and ploidy will become clearer with ongoing selection towards OsHV-1μvar resistant oysters. Work is currently underway in Ireland to investigate these factors further, by tracking seed from various hatchery sources which were put to sea in 2012 under similar husbandry and environmental conditions.

Caffeine, alcohol and overall nutrient adequacy are associated with longitudinal cognitive performance among US adults

The relationships of caffeine, alcohol and nutrient adequacy with cognitive performance were examined using data from a prospective cohort study, the Baltimore Longitudinal Study of Aging (n=628–1,305;~2 visits/person). Outcomes included ten cognitive test scores, spanning various domains. Caffeine and alcohol intakes and a nutrient adequacy score(NAS) were estimated from 7‐d food diaries. Multiple linear mixed models (Time=years since baseline assessment), indicated that caffeine was associated with better baseline global cognition, overall and for subjects with baseline age(Agebase)≥70y. Caffeine was also related to better baseline performance on visual memory overall, but with a faster decline on that domain over time. Higher NAS was linked to slower decline on verbal memory, immediate recall(overall,men, Agebase&lt;70y) and attention(women, Agebase&lt;70y), better baseline performance on immediate and delayed recall(Agebase≥70y), but with faster decline on global cognition and visual memory(Agebase&lt;70y). For Agebase&lt;70y, alcohol was linked to faster decline on letter fluency and global cognition. Conversely, for Agebase≥70y, alcohol was related to better baseline attention and working memory. Models with Time=Age centered at 50y were consistent. In sum, mixed patterns of association indicate modification by baseline age and sex, requiring more longitudinal studies.

Vitamin D receptor and megalin gene polymorphisms are associated with central adiposity status and changes among US adults.

We examined longitudinal associations of vitamin D receptor (VDR) and megalin (LRP2; LDL receptor-related protein-2) gene polymorphisms with central adiposity. We used data from the Baltimore Longitudinal Study of Aging (BLSA), an ongoing prospective open cohort study. Study participants consisted of non-Hispanic white adults residing in Baltimore city, with one or more visits at age≥50 years, and complete data (n 609-617). Repeated assessments on waist circumference (WC) and waist:hip ratio (WHR) were available. Multiple linear mixed models were used to estimate mid-follow-up age central adiposity level and annual rate of change with cut-points set at the sex-specific 80th percentile. The four binary outcomes were: 'elevated central adiposity' (ECA-WC and ECA-WHR) and 'significant increase in central adiposity' (SICA-WC and SICA-WHR). SNP for VDR (four SNP: (1) rs11568820 (CdX-2:T/C); (2) rs1544410 (BsmI:G/A); (3) rs7975232 (ApaI:A/C); (4) rs731236 (TaqI:G/A)) and Megalin (three SNP: (1) rs3755166:G/A; (2) rs2075252:C/T; (3) rs4668123:C/T) genes were selected. SNP latent classes (SNPLC) and SNP haplotypes (SNPHAP) were created. Multiple logistic regression analyses indicated that, in men, higher ECA-WHR odds were associated with SNPLC Megalin2:rs3755166[-]/rs2075252[TT]/rs4668123[T-] (v. Megalin1:rs3755166[-]/rs2075252[CC]/rs4668123[-]) (OR 2·87; 95 % CI 1·15, 7·12; P=0·023) and that SNPLC Megalin3:rs3755166[-]/rs2075252[CT]/rs4668123[-] (v. Megalin1) was linked to lower SICA-WC odds (OR 0·48; 95 % CI 0·26, 0·88; P=0·019) (P>0·05 for sex×SNPLC). In women, VDR3 SNPHAP (GAA:bAT) was related to lower odds of ECA-WC (OR 0·37; 95 % CI 0·16, 0·87; P=0·023) (P<0·05 for sex×SNPHAP), VDR1 SNPHAP (GCA:baT) was associated with greater odds and VDR3 SNPHAP (GAA:bAT) with lower odds of SICA-WC (P>0·05 for sex×SNPHAP). Vitamin D-related gene polymorphisms were associated with central adiposity status and change. Future mechanistic studies are needed to confirm those polymorphisms' biological significance to central adiposity.

Efficient maximum likelihood estimation of multiple membership linear mixed models, with an application to educational value-added assessments

The generalized persistence (GP) model, developed in the context of estimating “value added” by individual teachers to their students’ current and future test scores, is one of the most flexible value-added models in the literature. Although developed in the educational setting, the GP model can potentially be applied to any structure where each sequential response of a lower-level unit may be associated with a different higher-level unit, and the effects of the higher-level units may persist over time. The flexibility of the GP model, however, and its multiple membership random effects structure lead to computational challenges that have limited the model’s availability. We develop an EM algorithm to compute maximum likelihood estimates efficiently for the GP model, making use of the sparse structure of the random effects and error covariance matrices. The algorithm is implemented in the package GPvam in R statistical software. We give examples of the computations and illustrate the gains in computational efficiency achieved by our estimation procedure.

Prediction of psilocybin response in healthy volunteers.

Responses to hallucinogenic drugs, such as psilocybin, are believed to be critically dependent on the user's personality, current mood state, drug pre-experiences, expectancies, and social and environmental variables. However, little is known about the order of importance of these variables and their effect sizes in comparison to drug dose. Hence, this study investigated the effects of 24 predictor variables, including age, sex, education, personality traits, drug pre-experience, mental state before drug intake, experimental setting, and drug dose on the acute response to psilocybin. The analysis was based on the pooled data of 23 controlled experimental studies involving 409 psilocybin administrations to 261 healthy volunteers. Multiple linear mixed effects models were fitted for each of 15 response variables. Although drug dose was clearly the most important predictor for all measured response variables, several non-pharmacological variables significantly contributed to the effects of psilocybin. Specifically, having a high score in the personality trait of Absorption, being in an emotionally excitable and active state immediately before drug intake, and having experienced few psychological problems in past weeks were most strongly associated with pleasant and mystical-type experiences, whereas high Emotional Excitability, low age, and an experimental setting involving positron emission tomography most strongly predicted unpleasant and/or anxious reactions to psilocybin. The results confirm that non-pharmacological variables play an important role in the effects of psilocybin.

The Effects of Climate Change on Harp Seals (Pagophilus groenlandicus)

Harp seals (Pagophilus groenlandicus) have evolved life history strategies to exploit seasonal sea ice as a breeding platform. As such, individuals are prepared to deal with fluctuations in the quantity and quality of ice in their breeding areas. It remains unclear, however, how shifts in climate may affect seal populations. The present study assesses the effects of climate change on harp seals through three linked analyses. First, we tested the effects of short-term climate variability on young-of-the year harp seal mortality using a linear regression of sea ice cover in the Gulf of St. Lawrence against stranding rates of dead harp seals in the region during 1992 to 2010. A similar regression of stranding rates and North Atlantic Oscillation (NAO) index values was also conducted. These analyses revealed negative correlations between both ice cover and NAO conditions and seal mortality, indicating that lighter ice cover and lower NAO values result in higher mortality. A retrospective cross-correlation analysis of NAO conditions and sea ice cover from 1978 to 2011 revealed that NAO-related changes in sea ice may have contributed to the depletion of seals on the east coast of Canada during 1950 to 1972, and to their recovery during 1973 to 2000. This historical retrospective also reveals opposite links between neonatal mortality in harp seals in the Northeast Atlantic and NAO phase. Finally, an assessment of the long-term trends in sea ice cover in the breeding regions of harp seals across the entire North Atlantic during 1979 through 2011 using multiple linear regression models and mixed effects linear regression models revealed that sea ice cover in all harp seal breeding regions has been declining by as much as 6 percent per decade over the time series of available satellite data.

Statistical methods for medical studies

Most textbooks for biostatistics only explain each individual statistical test with its mathematical formula. However, it is crucial to understand the relationships among the statistical methods and to properly integrate the individual methods to effectively apply them to real clinical research settings. The choice for valid statistical tests greatly depends on the dependency of the sample and the number of independent variables in the analyses as well as the measurement scale of dependent variables and independent variables. In this report, many statistical tests such as the two sample t-test, ANOVA, non-parametric tests, chi-square test, log-rank test, multiple linear regression, logistic regression, mixed model, and Cox regression model are addressed through hypothetical examples. The key for a successful analysis of a clinical experiment is to adopt suitable statistical tests. This study presents a guideline to clinical researchers for selecting valid and powerful statistical tests in their study design. The choice of suitable statistical tests increases the reliability of analytical results and therefore the possibility of accepting a researcher's clinical hypothesis. The proposed flowchart of appropriate tests of statistical inference will be of help to many clinical researchers to their study.

Effects of ambient air pollution on pulmonary function among schoolchildren

Literature has shown adverse effects of ambient air pollution exposure on various asthma related outcomes in childhood. However, the associated evidence on pulmonary function effects is still inconsistent. We conducted a population-based study comprised of seventh-grade children in 14 Taiwanese communities. Pulmonary function tests and questionnaires were completed on 3957 subjects. We evaluated the effects of ambient air pollution exposures based on the data collected in 2005-2007 by existing air monitoring stations. Multiple linear mixed effect models were fitted to estimate the relationship between community pollutant levels and pulmonary function indices. After adjustment for individual-level confounders, pulmonary function differed only slightly between communities with different levels of air pollution. We found greater effects of ambient air pollutants on pulmonary function for boys than for girls. Among boys, traffic-related pollutants CO, NOx, NO(2), and NO were generally associated with chronic adverse effects on FVC and FEV(1), and subchronic adverse effects mainly on maximal mid-expiratory flow (MMEF) and peak expiratory flow rate. Among girls, only NOx and NO(2) showed subchronic adverse effects on MMEF. Although effect estimates of SO(2), PM(10), and PM(2.5) were generally negative for boys, none achieved statistical significance. Our data suggests that ambient traffic-related pollution had chronic adverse effects on pulmonary function in schoolchildren, especially for boys.