Distributional regression modeling via generalized additive models for location, scale, and shape: An overview through a data set from learning analytics

Distributional moderation analysis: Unpacking moderation effects in intervention research

Moderation and subgroup analyses are well-established statistical tools to evaluate whether intervention effects vary across subpopulations defined by participants' demographic and contextual factors. Moderation effects themselves, however, can be subject to heterogeneity and can manifest in various outcome parameters that go beyond group-specific averages (i.e., means) that are typically the focus of main and moderation effect analyses. The present study introduces distributional moderation analysis using the framework of inflated Generalized Additive Models for Location, Scale, and Shape (GAMLSS) that allows researchers to holistically characterize intervention effect modifiers through simultaneously modeling conditional mean-, variance-, skewness-, and kurtosis-based intervention effects, as well as moderated treatment effects located at the endpoints of the response scale (i.e., floor/ceiling effects). Data from a large-scale randomized controlled trial evaluating the effects of a teacher classroom management program on students' disruptive classroom behavior are used to provide a step-by-step guide for applying distributional moderation analysis in school-based intervention research. Although a traditional mean-focused analysis suggests that the intervention reduced students' average disruptive behavior only for students receiving special education, an evaluation of distributional treatment effects reveals a general decrease in the average disruptive behavior for at-risk students. In addition, distributional moderation analysis suggests that this average decrease is moderated by students' race and that the moderation effect of special education status initially seen in the traditional analysis is not located in the means, but in the chance to show no disruptive behavior patterns at all. Thus, we conclude that distributional moderation analysis constitutes a valuable complementary tool to provide a fine-grained characterization of treatment effect modifiers.

The cardiovascular exercise response in children with overweight or obesity measured by cardiovascular magnetic resonance imaging

BACKGROUND

Overweight and obesity are among the main causes of cardiovascular diseases. Exercise testing can aid in the early detection of subtle cardiac dysfunction not present in rest. We hypothesized that the cardiovascular response to exercise is impaired among children with overweight or obesity, characterized by the inability of the cardiovascular system to adapt to exercise by increasing cardiac volumes and blood pressure. We performed a cardiovascular stress test to investigate whether the cardiovascular exercise response is altered in children with overweight and obesity, as compared to children with a normal weight.

SUBJECTS

A subgroup of the Generation R population-based prospective cohort study, consisting of 41 children with overweight or obesity and 166 children with a normal weight with a mean age of 16 years, performed an isometric exercise.

METHODS

Continuous heart rate and blood pressure were measured during rest, exercise and recovery. Cardiovascular magnetic resonance (CMR) measurements were performed during rest and exercise.

RESULTS

Higher BMI was associated with a higher resting systolic and diastolic blood pressure (difference: 0.24 SDS (95% CI 0.10, 0.37) and 0.20 SDS (95% CI 0.06, 0.33)) and lower systolic and diastolic blood pressure increases from rest to peak exercise (-0.11 SDS (95% CI -0.20, -0.03) and -0.07 SDS (95% CI -0.07, -0.01)). BMI was also associated with a slower decrease in systolic and diastolic blood pressure during recovery (p values < 0.05). Higher childhood BMI was associated with lower BSA corrected left ventricular mass, end-diastolic volume and stroke volume (p values < 0.05). There were no associations of childhood BMI with the cardiac response to exercise measured by heart rate and CMR measurements.

CONCLUSION

Childhood BMI is, across the full range, associated with a blunted blood pressure response to static exercise but there were no differences in cardiac response to exercise. Our findings suggest that adiposity may especially affect the vascular exercise reaction without affecting cardiac response.

In vivo marker of brainstem myelin is associated to quantitative sleep parameters in healthy young men

The regional integrity of brain subcortical structures has been implicated in sleep-wake regulation, however, their associations with sleep parameters remain largely unexplored. Here, we assessed association between quantitative Magnetic Resonance Imaging (qMRI)-derived marker of the myelin content of the brainstem and the variability in the sleep electrophysiology in a large sample of 18-to-31 years healthy young men (N = 321; ~ 22 years). Separate Generalized Additive Model for Location, Scale and Shape (GAMLSS) revealed that sleep onset latency and slow wave energy were significantly associated with MTsat estimates in the brainstem (pcorrected ≤ 0.03), with overall higher MTsat value associated with values reflecting better sleep quality. The association changed with age, however (MTsat-by-age interaction-pcorrected ≤ 0.03), with higher MTsat value linked to better values in the two sleep metrics in the younger individuals of our sample aged ~ 18 to 20 years. Similar associations were detected across different parts of the brainstem (pcorrected ≤ 0.03), suggesting that the overall maturation and integrity of the brainstem was associated with both sleep metrics. Our results suggest that myelination of the brainstem nuclei essential to regulation of sleep is associated with inter-individual differences in sleep characteristics during early adulthood. They may have implications for sleep disorders or neurological diseases related to myelin.

Visual search and childhood vision impairment: A GAMLSS-oriented multiverse analysis approach

The aim of this report was to analyze reaction times and accuracy in children with a vision impairment performing a feature-based visual search task using a multiverse statistical approach. The search task consisted of set sizes 4, 16, and 24, consisting of distractors (circle) and a target (ellipse) that were presented randomly to school-aged individuals with or without a vision impairment. Interactions and main effects of key variables relating to reaction times and accuracy were analyzed via a novel statistical method blending GAMLSS (generalized additive models for location, scale, and shape) and distributional regression trees. Reaction times for the target-present and target-absent conditions were significantly slower in the vision impairment group with increasing set sizes (p < .001). Female participants were significantly slower than were males for set sizes 16 and 24 in the target-absent condition (p < .001), with male participants being significantly slower than females in the target-present condition (p < .001). Accuracy was only significantly worse (p = .03) for participants less than 14 years of age for the target-absent condition with set sizes 16 and 24. There was a positive association between binocular visual acuity and search time (p < .001). The application of GAMLSS with distributional regression trees to the analysis of visual search data may provide further insights into underlying factors affecting search performance in case-control studies where psychological or physical differences may influence visual search outcomes.