A new look at the difference between the GEE and the GLMM when modeling longitudinal count responses

Life-course Accumulated Cannabis Use and Recent Cannabis-related Problems in the Washington Panel Survey

BACKGROUND

Previous studies have only investigated the short-term association of recent cannabis use with cannabis-related problems, without accounting for the onset, duration, and variations in frequency of use in the life-course.

METHODS

We obtained data from the Washington panel survey during 2014-2016. We constructed accumulated lifetime exposure to cannabis use, heavy drinking (5+ drinks on one occasion), and cigarette pack-years from age of onset based on a series of decades-based questions on cannabis use and heavy drinking, and tobacco use history. We used Generalized Estimating Equation with Poisson distribution to investigate the association between accumulated cannabis use and the past-6-month CUDIT score. We adjusted for accumulated heavy drinking and cigarette pack-years, substance co-use variables, demographics, and applied survey weights.

RESULTS

We found strong and statistically significant correlations for the lifetime measures across the four panel surveys, indicating that the life-course measures of cannabis use and heavy drinking were largely reliable. We found a statistically significant relationship between the lifetime accumulated exposure to cannabis and CUDIT. The results were robust to the inconsistencies in reported frequencies and onset age across panel surveys.

CONCLUSIONS

This study established the relationship between lifetime exposure to cannabis and cannabis-related problems in a representative sample of drinkers and marijuana users in Washington state. We have also provided test-retest validity and question details for the decades-based cannabis and heavy drinking measures to facilitate their use in future studies of cannabis and alcohol-related outcomes.

Associations between simultaneous use of alcohol and cannabis and cannabis-related problems in 2014-2016: evidence from the Washington panel survey

BACKGROUND

To address the research question of how simultaneous users of alcohol and cannabis differ from concurrent users in risk of cannabis use problems after the recreational marijuana legalization in Washington State.

METHODS

We used generalized estimating equations with a Poisson distribution to analyze the association between simultaneous use of alcohol and marijuana (SAM) and cannabis-related problems compared to concurrent use. The data is a longitudinal sample of drinkers and cannabis users (n = 257, 47% female) aged 18 years and older from Washington State in 2014-2016. We adjusted for survey weights to account for differential probability of selection and response rates. The primary outcome is the past-six-month CUDIT problem subscale (ranging from 0 to 28), which is the total score for seven CUDIT problem items, after excluding the three items that covered marijuana use frequency. Covariates include marijuana use frequency (daily/near daily use, regular use, or infrequent use), marijuana daily quantity, alcohol daily volume, panel survey cycle, medical marijuana recommendation, driving time to nearest marijuana outlet, age of marijuana use onset, and other demographics.

RESULTS

After adjusting for covariates, we found that compared to concurrent use, SAM was significantly positively associated with CUDIT problem subscale (IRR = 1.68, 95% CI: 1.25-2.27, p < 0.001); daily/near daily use of marijuana was strongly significantly associated with CUDIT problem subscale compared with infrequent use (IRR = 5.1, 2.71-9.57, p < 0.001) or regular use (IRR = 3.05, 1.91-4.85, p < 0.001). Secondary analyses using CUDIT total score as the outcome also showed a significant positive association with SAM compared to concurrent use (IRR = 1.17, 1.02-1.34, p < 0.05).

CONCLUSIONS

This study highlighted the importance of SAM, in addition to cannabis use frequency for predicting cannabis-related problems.

The Vermont Family Based Approach in Primary Care Pediatrics: Effects on Children's and Parents' Emotional and Behavioral Problems and Parents' Health-Related Quality of Life

This randomized controlled trial tested the Vermont Family Based Approach (VFBA) in primary care pediatrics. The VFBA is a model of healthcare delivery that shifts the focus from the individual to the family, emphasizes emotional and behavioral health, and uses evidence-based health promotion/prevention along with the treatment of emotional and behavioral problems. Participants were 81 families of 3-15-year-olds. For children, the VFBA was associated with greater reductions than the Control condition on the Child Behavior Checklist Emotionally Reactive, Withdrawn, Sleep Problems, Aggressive Behavior and Total Problems scales. For parents, the VFBA was associated with greater reductions than the Control condition on the Adult Self-Report Anxious/Depressed, Rule-Breaking Behavior, Internalizing Problems and Total Problems scales. The VFBA was also associated with greater improvement than the Control condition in the parents' health-related quality of life, as indicated by all scales of the Medical Outcomes Study Health Survey.

Discrepancies of recurrent birth intervals using longitudinal data analysis in Ethiopia: interim EDHS 2019

OBJECTIVE

This study aimed to determine whether the birth interval changes differently over time among women in Ethiopia and whether the change depends on women, children and household characteristics measured at the last visit.

METHODS

Longitudinal study design was implemented based on the data obtained from the 2019 Ethiopia Mini Demographic and Health Survey consisting of a total of 3630 mothers. Generalised estimating equation and generalised linear mixed model were employed to estimate the effect of the determinants given the correlation between birth intervals within a mother is under consideration.

RESULTS

The majority of women were Muslims (48.1%) and come from rural areas (82.2%). About 77.2% of women at first birth were below 20 years old. A significant correlation (p value <0.0001) between the first and second birth intervals of mothers was observed. The estimated birth interval of women from the poorest household was 0.877 (e^-0.1317) times the estimated birth intervals of women from the richest household. This indicates richest households were likely to have higher birth intervals as compared with the poorest households (95% CI e-0.1754=0.839 to e-0.088=0.916).

CONCLUSION

The birth intervals of over one-fifth of mothers were 1 year, less than the birth interval recommended by the WHO standard. It was also perceived that successive birth intervals are correlated. Mothers who have delivered female children had lower birth intervals than mothers who have delivered male children. As compared with the birth intervals of mothers from a household with higher economic status, the birth intervals of mothers from a household with lower economic status had lower birth intervals. In this study, significant effects of religion, contraceptive use, region, mothers' current age, education level and mothers' current marital status on birth intervals were also noted.

Statistical implications of endogeneity induced by residential segregation in small-area modelling of health inequities

Health inequities are assessed by health departments to identify social groups disproportionately burdened by disease and by academic researchers to understand how social, economic, and environmental inequities manifest as health inequities. To characterize inequities, group-specific small-area health data are often modeled using log-linear generalized linear models (GLM) or generalized linear mixed models (GLMM) with a random intercept. These approaches estimate the same marginal rate ratio comparing disease rates across groups under standard assumptions. Here we explore how residential segregation combined with social group differences in disease risk can lead to contradictory findings from the GLM and GLMM. We show that this occurs because small-area disease rate data collected under these conditions induce endogeneity in the GLMM due to correlation between the model's offset and random effect. This results in GLMM estimates that represent conditional rather than marginal associations. We refer to endogeneity arising from the offset, which to our knowledge has not been noted previously, as "offset endogeneity". We illustrate this phenomenon in simulated data and real premature mortality data, and we propose alternative modeling approaches to address it. We also introduce to a statistical audience the social epidemiologic terminology for framing health inequities, which enables responsible interpretation of results.

Association between pre-participation characteristics and risk of injury amongst pre-professional dancers

STUDY DESIGN

Prospective Cohort Study.

BACKGROUND

Few investigations utilize evidence-informed pre-participation evaluation, inclusive injury definitions, and prospective surveillance to identify risk factors for dance-related injuries.

OBJECTIVE

To evaluate pre-participation characteristics that may be associated with greater odds of dance-related musculoskeletal complaints in pre-professional dancers.

METHODS

Full-time pre-professional ballet [n = 85, 77 females, median (range) age 15-years (11-19)] and contemporary [n = 60, 58 females, 19-years (17-30)] dancers underwent pre-participation evaluation: baseline questionnaire, coping skills, body mass index, bone mineral density, ankle range-of-motion, active standing turnout, lumbopelvic control, and balance tests. Self-reported complaints (any physical problem making dance participation difficult, irrespective of medical attention or time-loss) were captured weekly via online questionnaires for one academic year. Self-reported musculoskeletal complaints were recorded weekly (yes/no). Potential risk factors were identified a-priori through systematic review. Associations between potential risk factors and musculoskeletal complaints were examined with generalized linear mixed method regression models.

RESULTS

Response rate was 99%, with 81% of dancers reporting at least one musculoskeletal complaint. Of 1521 complaints (19% first-time, 81% ongoing), the ankle (22%), knee (21%), and foot (12%) accounted for the majority. Injury history [odds ratio (OR) 7.37 (95% CI 3.41, 15.91)] and previous week's dance hours [OR 1.02 (1.01, 1.03)] were associated with dance-related musculoskeletal complaints.

CONCLUSIONS

Prevalence of musculoskeletal complaints amongst pre-professional dancers is high and associated with injury history and training volume. Further understanding of the relationship between training load and injury is needed, with particular consideration of the dynamic and recursive nature of dance injury etiology.

LEVEL OF EVIDENCE

Therapy / Prevention, Aetiology / Harm, level 2b.

Functional decline in the aphasic variant of Alzheimer's disease

INTRODUCTION

Primary progressive aphasia (PPA) is a clinical dementia syndrome associated with frontotemporal lobar degeneration (FTLD) or Alzheimer's disease (AD). Impairment in activities of daily living is essential for dementia diagnosis, yet less is known about the neuropathologic impact on functional decline in PPA, especially over time.

METHODS

Activities of Daily Living Questionnaire (ADLQ) ratings were compared by suspected underlying pathology between 17 PPA^Aβ+ and 11 PPA^Aβ- participants at 6-month intervals for 2 years using a linear mixed-effects model. A general linear model examined associations between functional decline and cortical thickness at baseline.

RESULTS

Groups did not differ in demographics or aphasia severity at baseline, yet overall and subdomain scores of the ADLQ were significantly worse for PPA^Aβ+ compared to PPA^Aβ- (P = .015) at each interval across 18 months.

DISCUSSION

Functional decline appears more pronounced and disrupts more aspects of life activities for individuals with non-semantic PPA with suspected AD versus non-AD neuropathology.

A longitudinal Bayesian mixed effects model with hurdle Conway-Maxwell-Poisson distribution

Dental caries (i.e., cavities) is one of the most common chronic childhood diseases and may continue to progress throughout a person's lifetime. The Iowa Fluoride Study (IFS) was designed to investigate the effects of various fluoride, dietary and nondietary factors on the progression of dental caries among a cohort of Iowa school children. We develop a mixed effects model to perform a comprehensive analysis of the longitudinal clustered data of IFS at ages 5, 9, 13, and 17. We combine a Bayesian hurdle framework with the Conway-Maxwell-Poisson regression model, which can account for both excessive zeros and various levels of dispersion. A hierarchical shrinkage prior distribution is used to share the temporal information for predictors in the fixed-effects model. The dependence among teeth of each individual child is modeled through a sparse covariance structure of the random effects across time. Moreover, we obtain the parameter estimates and credible intervals from a Gibbs sampler. Simulation studies are conducted to assess the accuracy and effectiveness of our statistical methodology. The results of this article provide novel tools to statistical practitioners and offer fresh insights to dental researchers on effects of various risk and protective factors on caries progression.

A comparison study on modeling of clustered and overdispersed count data for multiple comparisons

Data collected in various scientific fields are count data. One way to analyze such data is to compare the individual levels of the factor treatment using multiple comparisons. However, the measured individuals are often clustered - e.g. according to litter or rearing. This must be considered when estimating the parameters by a repeated measurement model. In addition, ignoring the overdispersion to which count data is prone leads to an increase of the type one error rate. We carry out simulation studies using several different data settings and compare different multiple contrast tests with parameter estimates from generalized estimation equations and generalized linear mixed models in order to observe coverage and rejection probabilities. We generate overdispersed, clustered count data in small samples as can be observed in many biological settings. We have found that the generalized estimation equations outperform generalized linear mixed models if the variance-sandwich estimator is correctly specified. Furthermore, generalized linear mixed models show problems with the convergence rate under certain data settings, but there are model implementations with lower implications exists. Finally, we use an example of genetic data to demonstrate the application of the multiple contrast test and the problems of ignoring strong overdispersion.

Distribution-free models for latent mixed population responses in a longitudinal setting with missing data

Many biomedical and psychosocial studies involve population mixtures, which consist of multiple latent subpopulations. Because group membership cannot be observed, standard methods do not apply when differential treatment effects need to be studied across subgroups. We consider a two-group mixture in which membership of latent subgroups is determined by structural zeroes of a zero-inflated count variable and propose a new approach to model treatment differences between latent subgroups in a longitudinal setting. It has also been incorporated with the inverse probability weighted method to address data missingness. As the approach builds on the distribution-free functional response models, it requires no parametric distribution model and thereby provides a robust inference. We illustrate the approach with both real and simulated data.

On performance of parametric and distribution-free models for zero-inflated and over-dispersed count responses

Zero-inflated Poisson (ZIP) and negative binomial (ZINB) models are widely used to model zero-inflated count responses. These models extend the Poisson and negative binomial (NB) to address excessive zeros in the count response. By adding a degenerate distribution centered at 0 and interpreting it as describing a non-risk group in the population, the ZIP (ZINB) models a two-component population mixture. As in applications of Poisson and NB, the key difference between ZIP and ZINB is the allowance for overdispersion by the ZINB in its NB component in modeling the count response for the at-risk group. Overdispersion arising in practice too often does not follow the NB, and applications of ZINB to such data yield invalid inference. If sources of overdispersion are known, other parametric models may be used to directly model the overdispersion. Such models too are subject to assumed distributions. Further, this approach may not be applicable if information about the sources of overdispersion is unavailable. In this paper, we propose a distribution-free alternative and compare its performance with these popular parametric models as well as a moment-based approach proposed by Yu et al. [Statistics in Medicine 2013; 32: 2390-2405]. Like the generalized estimating equations, the proposed approach requires no elaborate distribution assumptions. Compared with the approach of Yu et al., it is more robust to overdispersed zero-inflated responses. We illustrate our approach with both simulated and real study data.

Introduction to RNA-Seq and its applications to drug discovery and development

Preclinical Research RNA sequencing (RNA-Seq) is a novel high-throughput technology for comprehensive transcriptome study. It can measure the expression levels of thousands of genes simultaneously and provide insight into functional pathways and regulations in biological processes. In addition, RNA-Seq can provide copious information on alternative splicing, allele-specific expression, unannotated exons, and novel transcripts (gene or noncoding RNAs). This technology has revolutionized the way biologists examine transcriptomes and has been successfully applied in drug discovery and development, being able to identify drug-related genes, microRNAs, and fusion proteins. In this overview, we will review this technology including data analysis, and its recent applications in drug discovery and development.

Coping with time and space in modelling malaria incidence: a comparison of survival and count regression models

To study the effect of a mega hydropower dam in southwest Ethiopia on malaria incidence, we have set up a longitudinal study. To gain insight in temporal and spatial aspects, that is, in time (period  =  year-season combination) and location (village), we need models that account for these effects. The frailty model with periodwise constant baseline hazard (a constant value for each period) and a frailty term that models the clustering in villages provides an appropriate tool for the analysis of such incidence data. Count data can be obtained by aggregating for each period events at the village level. The mixed Poisson regression model can be used to model the count data. We show the similarities between the two models. The risk factor in both models is the distance to the dam, and we study the effect of the risk factor on malaria incidence. In the frailty model, each subject has its own risk factor, whereas in the Poisson regression model, we also need to average the risk factors of all subjects contributing to a particular count. The power loss caused by using village averaged distance instead of individual distance is studied and quantified. The loss in the malaria data example is rather small. In such a setting, it might be advantageous to use less labor-intensive sampling schemes than the weekly individual follow-up scheme used in this study; the proposed alternative sampling schemes might also avoid community fatigue, a typical problem in such research projects.

Distribution-free models for longitudinal count responses with overdispersion and structural zeros

Overdispersion and structural zeros are two major manifestations of departure from the Poisson assumption when modeling count responses using Poisson log-linear regression. As noted in a large body of literature, ignoring such departures could yield bias and lead to wrong conclusions. Different approaches have been developed to tackle these two major problems. In this paper, we review available methods for dealing with overdispersion and structural zeros within a longitudinal data setting and propose a distribution-free modeling approach to address the limitations of these methods by utilizing a new class of functional response models. We illustrate our approach with both simulated and real study data.