One-year follow-up of disease burden and medication changes in patients with myasthenia gravis: From the MG Patient Registry.

INTRODUCTION/AIMS: We studied the progression of myasthenia gravis (MG) disease burden and medication adjustment among MG Patient Registry participants. METHODS: Participants diagnosed with MG (age ≥18 years), registered between July 1, 2013 and July 31, 2018 and completing both 6- and 12-month follow-up surveys, were included in this investigation. Participants were grouped into high-burden (Myasthenia Gravis Activity of Daily Living scale [MG-ADL] score ≥6) and low-burden (MG-ADL <6) groups based on MG-ADL scores at enrollment. Demographics and disease history were compared between groups. MG-ADL score change and medication changes (escalation, no change, de-escalation) between enrollment and 12-month follow-up were compared between groups. Minimal symptom expression (MSE, MG-ADL <2) at 12 months was compared between groups. Logistic regression analysis was performed to study factors associated with MSE at 12 months. RESULTS: In total, 520 participants (56% female) were included in high-burden (n = 248) and low-burden (n = 272) groups. Those in the high-burden group were more likely to be younger, female, and have shorter disease duration. At 12 months, MSE was achieved in 6% of the high-burden group and newly achieved (42 of 201, 21%) or maintained (52 of 71, 73%) in the low-burden group. In the multivariable analysis, being in the high-burden group and use of pyridostigmine were associated with less likelihood of MSE, whereas MG-ADL score improvement (>2 or >20%) at 6 months significantly increased the likelihood of achieving MSE at 12 months (P = .0004). DISCUSSION: In both groups, but more so in the high-burden group, patients infrequently achieved MSE after 1 year of MG treatment. Baseline low disease burden, improvement at 6 months and no pyridostigmine use were associated with a higher likelihood of MSE at 12 months.

Incorporating spatial structure into inclusion probabilities for Bayesian variable selection in generalized linear models with the spike-and-slab elastic net.

Spike-and-slab priors model predictors as arising from a mixture of distributions: those that should (slab) or should not (spike) remain in the model. The spike-and-slab lasso (SSL) is a mixture of double exponentials, extending the single lasso penalty by imposing different penalties on parameters based on their inclusion probabilities. The SSL was extended to Generalized Linear Models (GLM) for application in genetics/genomics, and can handle many highly correlated predictors of a scalar outcome, but does not incorporate these relationships into variable selection. When images/spatial data are used to model a scalar outcome, relevant parameters tend to cluster spatially, and model performance may benefit from incorporating spatial structure into variable selection. We propose to incorporate spatial information by assigning intrinsic autoregressive priors to the logit prior probabilities of inclusion, which results in more similar shrinkage penalties among spatially adjacent parameters. Using MCMC to fit Bayesian models can be computationally prohibitive for large-scale data, but we fit the model by adapting a computationally efficient coordinate-descent-based EM algorithm. A simulation study and an application to Alzheimer's Disease imaging data show that incorporating spatial information can improve model fitness.

Bayesian compositional models for ordinal response.

Ordinal response is commonly found in medicine, biology, and other fields. In many situations, the predictors for this ordinal response are compositional, which means that the sum of predictors for each sample is fixed. Examples of compositional data include the relative abundance of species in microbiome data and the relative frequency of nutrition concentrations. Moreover, the predictors that are strongly correlated tend to have similar influence on the response outcome. Conventional cumulative logistic regression models for ordinal responses ignore the fixed-sum constraint on predictors and their associated interrelationships, and thus are not appropriate for analyzing compositional predictors.To solve this problem, we proposed Bayesian Compositional Models for Ordinal Response to analyze the relationship between compositional data and an ordinal response with a structured regularized horseshoe prior for the compositional coefficients and a soft sum-to-zero restriction on coefficients through the prior distribution. The method was implemented with R package rstan using efficient Hamiltonian Monte Carlo algorithm. We performed simulations to compare the proposed approach and existing methods for ordinal responses. Results revealed that our proposed method outperformed the existing methods in terms of parameter estimation and prediction. We also applied the proposed method to a microbiome study HMP2Data, to find microorganisms linked to ordinal inflammatory bowel disease levels. To make this work reproducible, the code and data used in this paper are available at https://github.com/Li-Zhang28/BCO.

All-Cause Mortality in Reproductive-Aged Females by State: An Analysis of the Effects of Abortion Legislation.

OBJECTIVE: To estimate the association between state-level abortion legislation and all-cause mortality among all females of reproductive age and maternal, fetal, and infant mortality. METHODS: We conducted a retrospective cohort study using the Centers for Disease Control and Prevention's WONDER (Wide-ranging ONline Data for Epidemiologic Research) database. Generalized estimating equations were used to estimate the association between supportive, moderate, and restrictive state abortion regulations and all-cause mortality in reproductive-aged females. Secondary outcomes included maternal, fetal, and infant mortality. The association of the number and type of laws on mortality were estimated. RESULTS: Moderate and supportive states were not associated with a significant decrease in all-cause mortality compared with restrictive states. Maternal mortality (per 100,000 live births) was significantly lower in moderate (-5.79, 95% CI -9.88 to -1.70) compared with restrictive states, but not supportive states (-2.51, 95% CI -6.75 to 1.72). Infant mortality (per 1,000 live births) was significantly lower in both moderate (-0.56, 95% CI -1.09 to -0.04) and supportive (-1.10, 95% CI -1.56 to -0.64) states. Fetal mortality was lower in moderate states (-0.69, 95% CI -1.18 to -0.20) but not in supportive states (-0.64, 95% CI -1.14 to 0.13). Each additional abortion regulation was associated with an increase in maternal mortality (1.09/100,000 live births, 95% CI 0.36-1.82) and infant mortality (0.20/1,000 live births, 95% CI 0.12-0.26). CONCLUSION: Moderate state abortion legislation was associated with lower rates of maternal, fetal, and infant mortality but not lower all-cause mortality in reproductive-aged females compared with restrictive laws. An increasing number of laws restricting abortion was associated with increased maternal and infant mortality.

Breastfeeding initiation and duration among people with mild chronic hypertension: a secondary analysis of the Chronic Hypertension and Pregnancy trial.

BACKGROUND: Increased duration of breastfeeding improves maternal cardiovascular health and may be especially beneficial in high-risk populations, such as those with chronic hypertension. Others have shown that individuals with hypertension are less likely to breastfeed, and there has been limited research aimed at supporting breastfeeding goals in this population. The impact of perinatal blood pressure control on breastfeeding outcomes among people with chronic hypertension is unknown. OBJECTIVE: This study aimed to evaluate whether breastfeeding initiation and short-term duration assessed at the postpartum clinic visit differed according to perinatal blood pressure treatment strategy (targeting blood pressure <140/90 mm Hg vs reserving antihypertensive treatment for blood pressure ≥160/105 mm Hg). STUDY DESIGN: We performed a secondary analysis of the Chronic Hypertension and Pregnancy trial. This was an open-label, multicenter, randomized trial where pregnant participants with mild chronic hypertension were randomized to receive antihypertensive medications with goal blood pressure <140/90 mm Hg (active treatment) or deferred treatment until blood pressure ≥160/105 mm Hg (control). The primary outcome was initiation and duration of breastfeeding, assessed at the postpartum clinic visit. We performed bivariate analyses and log-binomial and cumulative logit regression models, adjusting models for variables that were unbalanced in bivariate analyses. We performed additional analyses to explore the relationship between breastfeeding duration and blood pressure measurements at the postpartum visit. RESULTS: Of the 2408 participants from the Chronic Hypertension and Pregnancy trial, 1444 (60%) attended the postpartum study visit and provided breastfeeding information. Participants in the active treatment group had different body mass index class distribution and earlier gestational age at enrollment, and (by design) were more often discharged on antihypertensives. Breastfeeding outcomes did not differ significantly by treatment group. In the active and control treatment groups, 563 (77.5%) and 561 (78.1%) initiated breastfeeding, and mean durations of breastfeeding were 6.5±2.3 and 6.3±2.1 weeks, respectively. The probability of ever breastfeeding (adjusted relative risk, 0.99; 95% confidence interval, 0.93-1.05), current breastfeeding at postpartum visit (adjusted relative risk, 1.01; 95% confidence interval, 0.94-1.10), and weeks of breastfeeding (adjusted odds ratio, 0.87; 95% confidence interval, 0.68-1.12) did not differ by treatment group. Increased duration (≥2 vs <2 weeks) of breastfeeding was associated with slightly lower blood pressure measurements at the postpartum visit, but these differences were not significant in adjusted models. CONCLUSION: In a secondary analysis of the cohort of Chronic Hypertension and Pregnancy trial participants who attended the postpartum study visit and provided breastfeeding information (60% of original trial participants), breastfeeding outcomes did not differ significantly by treatment group. This suggests that maintaining goal blood pressure <140/90 mm Hg throughout the perinatal period is associated with neither harm nor benefit for short-term breastfeeding goals. Further study is needed to understand long-term breastfeeding outcomes among individuals with chronic hypertension and how to support this population in achieving their breastfeeding goals.

The spike-and-slab elastic net as a classification tool in Alzheimer's disease.

Alzheimer's disease (AD) is the leading cause of dementia and has received considerable research attention, including using neuroimaging biomarkers to classify patients and/or predict disease progression. Generalized linear models, e.g., logistic regression, can be used as classifiers, but since the spatial measurements are correlated and often outnumber subjects, penalized and/or Bayesian models will be identifiable, while classical models often will not. Many useful models, e.g., the elastic net and spike-and-slab lasso, perform automatic variable selection, which removes extraneous predictors and reduces model variance, but neither model exploits spatial information in selecting variables. Spatial information can be incorporated into variable selection by placing intrinsic autoregressive priors on the logit probabilities of inclusion within a spike-and-slab elastic net framework. We demonstrate the ability of this framework to improve classification performance by using cortical thickness and tau-PET images from the Alzheimer's Disease Neuroimaging Initiative (ADNI) to classify subjects as cognitively normal or having dementia, and by using a simulation study to examine model performance using finer resolution images.

Measuring cognitive function by the SDMT across functional domains: Useful but not sufficient.

BACKGROUND: The Symbol Digit Modalities Test (SDMT) is a common screen of cognitive function for people with Multiple Sclerosis (pwMS) but growing acknowledgement that people with cognitive impairment are a heterogeneous population suggests that a single screen may provide limited information. OBJECTIVE: To assess the adequacy of the SDMT in capturing impairment across specific cognitive domains as measured by a multi-domain cognitive assessment battery (CAB, NeuroTrax). METHODS: 113 pwMS were assessed with SDMT and the CAB. Cognitive impairment in each CAB domain was defined as ≥1.5 SD below the normalized mean. Logistic regression models were fit for each CAB domain with domain-specific cognitive impairment as the outcome and SDMT as the predictor, and a classifier created by selecting cutpoints using the Youden Index. Model performance was assessed by predicting domain-specific cognitive impairment in an independent data set consisting of 81 pwMS. RESULTS: SDMT was a significant predictor of cognitive impairment in all outcomes considered (Odds Ratio: 0.885-0.950), but prediction metrics such as area under the receiver operating curve (AUC) were modest (0.623-0.778), and the alignment between observed/predicted impairment was less than optimal. CONCLUSION: The SDMT is not sufficient to differentiate between impaired and non-impaired pwMS across several cognitive domains.

Sickle cell trait and risk of cognitive impairment in African-Americans: The REGARDS cohort.

BACKGROUND: Sickle cell anemia may be associated with cognitive dysfunction, and some complications of sickle cell anemia might affect those with sickle cell trait (SCT), so we hypothesized that SCT is a risk factor for cognitive impairment. METHODS: The Reasons for Geographic and Racial Differences in Stroke (REGARDS) study enrolled a national cohort of 30,239 white and black Americans from 2003 to 7, who are followed every 6 months. Baseline and annual global cognitive function testing used the Six-Item Screener (SIS), a validated instrument (scores range 0-6; ≤ 4 indicates cognitive impairment). Participants with baseline cognitive impairment and whites were excluded. Logistic regression was used to calculate the association of SCT with incident cognitive impairment, adjusted for risk factors. Linear mixed models assessed multivariable-adjusted change in test scores on a biennially administered 3-test battery measuring learning, memory, and semantic and phonemic fluency. FINDINGS: Among 7743 participants followed for a median of 7·1 years, 85 of 583 participants with SCT (14·6%) developed incident cognitive impairment compared to 902 of 7160 (12·6%) without SCT. In univariate analysis, the odds ratio (OR) of incident cognitive impairment was 1·18 (95% CI: 0·93, 1·51) for those with SCT vs. those without. Adjustment did not impact the OR. There was no difference in change on 3-test battery scores by SCT status (all p > 0·11). INTERPRETATION: In this prospective cohort study of black Americans, SCT was not associated with incident cognitive impairment or decline in test scores of learning, memory and executive function. FUNDING: National Institutes of Health, American Society of Hematology.

Nasospheroids permit measurements of CFTR-dependent fluid transport.

Expansion of novel therapeutics to all patients with cystic fibrosis (CF) requires personalized CFTR modulator therapy. We have developed nasospheroids, a primary cell culture-based model derived from individual CF patients and healthy subjects by a minimally invasive nasal biopsy. Confocal microscopy was utilized to measure CFTR activity by analyzing changes in cross-sectional area over time that resulted from CFTR-mediated ion and fluid movement. Both the rate of change over time and AUC were calculated. Non-CF nasospheroids with active CFTR-mediated ion and fluid movement showed a reduction in cross-sectional area, whereas no changes were observed in CF spheroids. Non-CF spheroids treated with CFTR inhibitor lost responsiveness for CFTR activation. However, nasospheroids from F508del CF homozygotes that were treated with lumacaftor and ivacaftor showed a significant reduction in cross-sectional area, indicating pharmacologic rescue of CFTR function. This model employs a simple measurement of size corresponding to changes in CFTR activity and is applicable for detection of small changes in CFTR activity from individual patients in vitro. Advancements of this technique will provide a robust model for individualized prediction of CFTR modulator efficacy.