Multivariate semiparametric spatial methods for imaging data.

Univariate semiparametric methods are often used in modeling nonlinear age trajectories for imaging data, which may result in efficiency loss and lower power for identifying important age-related effects that exist in the data. As observed in multiple neuroimaging studies, age trajectories show similar nonlinear patterns for the left and right corresponding regions and for the different parts of a big organ such as the corpus callosum. To incorporate the spatial similarity information without assuming spatial smoothness, we propose a multivariate semiparametric regression model with a spatial similarity penalty, which constrains the variation of the age trajectories among similar regions. The proposed method is applicable to both cross-sectional and longitudinal region-level imaging data. We show the asymptotic rates for the bias and covariance functions of the proposed estimator and its asymptotic normality. Our simulation studies demonstrate that by borrowing information from similar regions, the proposed spatial similarity method improves the efficiency remarkably. We apply the proposed method to two neuroimaging data examples. The results reveal that accounting for the spatial similarity leads to more accurate estimators and better functional clustering results for visualizing brain atrophy pattern.Functional clustering; Longitudinal magnetic resonance imaging (MRI); Penalized B-splines; Region of interest (ROI); Spatial penalty.

Penalized nonlinear mixed effects model to identify biomarkers that predict disease progression.

Precise modeling of disease progression in neurodegenerative disorders may enable early intervention before clinical manifestation of a disease, which is crucial since early intervention at the premanifest stage is expected to be more effective. Neuroimaging biomarkers are indicative of the underlying disease pathology and may be used to predict future disease occurrence at the premanifest stage. As observed in many pivotal studies, longitudinal measurements of clinical outcomes, such as motor or cognitive symptoms, often present nonlinear sigmoid shapes over time, where the inflection points of the trajectories mark a meaningful time in disease progression. Therefore, to identify neuroimaging biomarkers predicting disease progression, we propose a nonlinear mixed effects model based on a sigmoid function to predict longitudinal clinical outcomes, and associate a linear combination of neuroimaging biomarkers with subject-specific inflection points. Based on an expectation-maximization (EM) algorithm, we propose a method that can fit a nonlinear model with many potentially correlated biomarkers for random inflection points while achieving computational stability. Variable selection is introduced in the algorithm in order to identify important biomarkers of disease progression and to reduce prediction variability. We apply the proposed method to the data from the Predictors of Huntington's Disease study to select brain subcortical regional volumes predictive of the inflection points of the motor and cognitive function trajectories. Our results reveal that brain atrophy in the striatum and expansion of the ventricular system are highly predictive of the inflection points. Furthermore, these inflection points may precede clinically defined disease onset by as early as a decade and thus may be useful biomarkers as early signs of Huntington's Disease onset.

Pointwise influence matrices for functional-response regression.

We extend the notion of an influence or hat matrix to regression with functional responses and scalar predictors. For responses depending linearly on a set of predictors, our definition is shown to reduce to the conventional influence matrix for linear models. The pointwise degrees of freedom, the trace of the pointwise influence matrix, are shown to have an adaptivity property that motivates a two-step bivariate smoother for modeling nonlinear dependence on a single predictor. This procedure adapts to varying complexity of the nonlinear model at different locations along the function, and thereby achieves better performance than competing tensor product smoothers in an analysis of the development of white matter microstructure in the brain.

A semiparametric joint model for terminal trend of quality of life and survival in palliative care research.

Palliative medicine is an interdisciplinary specialty focusing on improving quality of life (QOL) for patients with serious illness and their families. Palliative care programs are available or under development at over 80% of large US hospitals (300+ beds). Palliative care clinical trials present unique analytic challenges relative to evaluating the palliative care treatment efficacy which is to improve patients' diminishing QOL as disease progresses towards end of life (EOL). A unique feature of palliative care clinical trials is that patients will experience decreasing QOL during the trial despite potentially beneficial treatment. Often longitudinal QOL and survival data are highly correlated which, in the face of censoring, makes it challenging to properly analyze and interpret terminal QOL trend. To address these issues, we propose a novel semiparametric statistical approach to jointly model the terminal trend of QOL and survival data. There are two sub-models in our approach: a semiparametric mixed effects model for longitudinal QOL and a Cox model for survival. We use regression splines method to estimate the nonparametric curves and AIC to select knots. We assess the model performance through simulation to establish a novel modeling approach that could be used in future palliative care research trials. Application of our approach in a recently completed palliative care clinical trial is also presented.

Age exacerbates HIV-associated white matter abnormalities.

Both HIV disease and advanced age have been associated with alterations to cerebral white matter, as measured with white matter hyperintensities (WMH) on fluid-attenuated inversion recovery (FLAIR) magnetic resonance imaging (MRI), and more recently with diffusion tensor imaging (DTI). This study investigates the combined effects of age and HIV serostatus on WMH and DTI measures, as well as the relationships between these white matter measures, in 88 HIV seropositive (HIV+) and 49 seronegative (HIV-) individuals aged 23-79 years. A whole-brain volumetric measure of WMH was quantified from FLAIR images using a semi-automated process, while fractional anisotropy (FA) was calculated for 15 regions of a whole-brain white matter skeleton generated using tract-based spatial statistics (TBSS). An age by HIV interaction was found indicating a significant association between WMH and older age in HIV+ participants only. Similarly, significant age by HIV interactions were found indicating stronger associations between older age and decreased FA in the posterior limbs of the internal capsules, cerebral peduncles, and anterior corona radiata in HIV+ vs. HIV- participants. The interactive effects of HIV and age were stronger with respect to whole-brain WMH than for any of the FA measures. Among HIV+ participants, greater WMH and lower anterior corona radiata FA were associated with active hepatitis C virus infection, a history of AIDS, and higher current CD4 cell count. Results indicate that age exacerbates HIV-associated abnormalities of whole-brain WMH and fronto-subcortical white matter integrity.

Edgewise and subgraph-level tests for brain networks.

Resting-state functional magnetic resonance image is a useful technique for investigating brain functional connectivity at rest. In this work, we develop flexible regression models and methods for determining differences in resting-state functional connectivity as a function of age, gender, drug intervention, or neuropsychiatric disorders. We propose two complementary methods for identifying changes of edges and subgraphs. (i) For detecting changes of edges, we select the optimal model at each edge and then conduct contrast tests to identify the effects of the important variables while controlling the familywise error rate. (ii) We adopt the network-based statistics method to improve power by incorporating the graph topological structure. Both methods have wide applications for low signal-to-noise ratio data. We propose stability criteria for the choice of threshold in the network-based statistics procedure and utilize efficient massive parallel procedure to speed up the estimation and inference procedure. Results from our simulation studies show that the thresholds chosen by the proposed stability criteria outperform the Bonferroni threshold. To demonstrate applicability, we use both methods in the context of the Oxytocin and Aging Study to determine effects of age, gender, and drug treatment on resting-state functional connectivity, as well as in the context of the Autism Brain Imaging Data Exchange Study to determine effects of autism spectrum disorder on functional connectivity at rest. Copyright © 2016 John Wiley & Sons, Ltd.

Quantile rank maps: a new tool for understanding individual brain development.

We propose a novel method for neurodevelopmental brain mapping that displays how an individual's values for a quantity of interest compare with age-specific norms. By estimating smoothly age-varying distributions at a set of brain regions of interest, we derive age-dependent region-wise quantile ranks for a given individual, which can be presented in the form of a brain map. Such quantile rank maps could potentially be used for clinical screening. Bootstrap-based confidence intervals are proposed for the quantile rank estimates. We also propose a recalibrated Kolmogorov-Smirnov test for detecting group differences in the age-varying distribution. This test is shown to be more robust to model misspecification than a linear regression-based test. The proposed methods are applied to brain imaging data from the Nathan Kline Institute Rockland Sample and from the Autism Brain Imaging Data Exchange (ABIDE) sample.

Optimally weighted L(2) distance for functional data.

Many techniques of functional data analysis require choosing a measure of distance between functions, with the most common choice being L2 distance. In this article we show that using a weighted L2 distance, with a judiciously chosen weight function, can improve the performance of various statistical methods for functional data, including k-medoids clustering, nonparametric classification, and permutation testing. Assuming a quadratically penalized (e.g., spline) basis representation for the functional data, we consider three nontrivial weight functions: design density weights, inverse-variance weights, and a new weight function that minimizes the coefficient of variation of the resulting squared distance by means of an efficient iterative procedure. The benefits of weighting, in particular with the proposed weight function, are demonstrated both in simulation studies and in applications to the Berkeley growth data and a functional magnetic resonance imaging data set.

A note on a nonparametric regression test through penalized splines.

We examine a test of a nonparametric regression function based on penalized spline smoothing. We show that, similarly to a penalized spline estimator, the asymptotic power of the penalized spline test falls into a small- K or a large-K scenarios characterized by the number of knots K and the smoothing parameter. However, the optimal rate of K and the smoothing parameter maximizing power for testing is different from the optimal rate minimizing the mean squared error for estimation. Our investigation reveals that compared to estimation, some under-smoothing may be desirable for the testing problems. Furthermore, we compare the proposed test with the likelihood ratio test (LRT). We show that when the true function is more complicated, containing multiple modes, the test proposed here may have greater power than LRT. Finally, we investigate the properties of the test through simulations and apply it to two data examples.

On testing an unspecified function through a linear mixed effects model with multiple variance components.

We examine a generalized F-test of a nonparametric function through penalized splines and a linear mixed effects model representation. With a mixed effects model representation of penalized splines, we imbed the test of an unspecified function into a test of some fixed effects and a variance component in a linear mixed effects model with nuisance variance components under the null. The procedure can be used to test a nonparametric function or varying-coefficient with clustered data, compare two spline functions, test the significance of an unspecified function in an additive model with multiple components, and test a row or a column effect in a two-way analysis of variance model. Through a spectral decomposition of the residual sum of squares, we provide a fast algorithm for computing the null distribution of the test, which significantly improves the computational efficiency over bootstrap. The spectral representation reveals a connection between the likelihood ratio test (LRT) in a multiple variance components model and a single component model. We examine our methods through simulations, where we show that the power of the generalized F-test may be higher than the LRT, depending on the hypothesis of interest and the true model under the alternative. We apply these methods to compute the genome-wide critical value and p-value of a genetic association test in a genome-wide association study (GWAS), where the usual bootstrap is computationally intensive (up to 10(8) simulations) and asymptotic approximation may be unreliable and conservative.

Chronic lesion activity and disability progression in secondary progressive multiple sclerosis.

Objective: Slowly expanding lesions (SELs), a subgroup of chronic white matter lesions that gradually expand over time, have been shown to predict disability accumulation in primary progressive multiple sclerosis (MS) disease. However, the relationships between SELs, acute lesion activity (ALA), overall chronic lesion activity (CLA) and disability progression are not well understood. In this study, we examined the ASCEND phase III clinical trial, which compared natalizumab with placebo in secondary progressive MS (SPMS). Methods: Patients with complete imaging datasets between baseline and week 108 (N=600) were analysed for SEL prevalence (the number and volume of SELs), disability progression, ALA (assessed by gadolinium-enhancing lesions and new T2-hyperintense lesions) and CLA (assessed by T1-hypointense lesion volume increase within baseline T2-non-enhancing lesions identified as SELs and non-SELs). Results: CLA in both SELs and non-SELs was greater in patients with SPMS with confirmed disability progression than in those with no progression. In the complete absence of ALA at baseline and on study, SEL prevalence was significantly lower, while CLA within non-SELs remained associated with disability progression. Natalizumab decreased SEL prevalence and CLA in SELs and non-SELs compared with placebo. Conclusions: This study shows that CLA in patients with SPMS is decreased but persists in the absence of ALA and is associated with disability progression, highlighting the need for therapeutics targeting all mechanisms of CLA, including smouldering inflammation and neurodegeneration. Trial registration number: NCT01416181.

A penalized spline approach to functional mixed effects model analysis.

In this article, we propose penalized spline (P-spline)-based methods for functional mixed effects models with varying coefficients. We decompose longitudinal outcomes as a sum of several terms: a population mean function, covariates with time-varying coefficients, functional subject-specific random effects, and residual measurement error processes. Using P-splines, we propose nonparametric estimation of the population mean function, varying coefficient, random subject-specific curves, and the associated covariance function that represents between-subject variation and the variance function of the residual measurement errors which represents within-subject variation. Proposed methods offer flexible estimation of both the population- and subject-level curves. In addition, decomposing variability of the outcomes as a between- and within-subject source is useful in identifying the dominant variance component therefore optimally model a covariance function. We use a likelihood-based method to select multiple smoothing parameters. Furthermore, we study the asymptotics of the baseline P-spline estimator with longitudinal data. We conduct simulation studies to investigate performance of the proposed methods. The benefit of the between- and within-subject covariance decomposition is illustrated through an analysis of Berkeley growth data, where we identified clearly distinct patterns of the between- and within-subject covariance functions of children's heights. We also apply the proposed methods to estimate the effect of antihypertensive treatment from the Framingham Heart Study data.

Prediction-based structured variable selection through the receiver operating characteristic curves.

In many clinical settings, a commonly encountered problem is to assess accuracy of a screening test for early detection of a disease. In these applications, predictive performance of the test is of interest. Variable selection may be useful in designing a medical test. An example is a research study conducted to design a new screening test by selecting variables from an existing screener with a hierarchical structure among variables: there are several root questions followed by their stem questions. The stem questions will only be asked after a subject has answered the root question. It is therefore unreasonable to select a model that only contains stem variables but not its root variable. In this work, we propose methods to perform variable selection with structured variables when predictive accuracy of a diagnostic test is the main concern of the analysis. We take a linear combination of individual variables to form a combined test. We then maximize a direct summary measure of the predictive performance of the test, the area under a receiver operating characteristic curve (AUC of an ROC), subject to a penalty function to control for overfitting. Since maximizing empirical AUC of the ROC of a combined test is a complicated nonconvex problem (Pepe, Cai, and Longton, 2006, Biometrics62, 221-229), we explore the connection between the empirical AUC and a support vector machine (SVM). We cast the problem of maximizing predictive performance of a combined test as a penalized SVM problem and apply a reparametrization to impose the hierarchical structure among variables. We also describe a penalized logistic regression variable selection procedure for structured variables and compare it with the ROC-based approaches. We use simulation studies based on real data to examine performance of the proposed methods. Finally we apply developed methods to design a structured screener to be used in primary care clinics to refer potentially psychotic patients for further specialty diagnostics and treatment.

Assessment of a disease screener by hierarchical all-subset selection using area under the receiver operating characteristic curves.

In many clinical settings, a commonly encountered problem is to assess the accuracy of a screening test for early detection of a disease. In this article, we develop hierarchical all-subset variable selection methods to assess and improve a psychosis screening test designed to detect psychotic patients in primary care clinics. We select items from an existing screener to achieve best prediction accuracy based on a gold standard psychosis status diagnosis. The existing screener has a hierarchical structure: the questions fall into five domains, and there is a root question followed by several stem questions in each domain. The statistical question lies in how to implement the hierarchical structure in the screening items when performing variable selection such that when a stem question is selected in the screener, its root question should also be selected. We develop an all-subset variable selection procedure that takes into account the hierarchical structure in a questionnaire. By enforcing a hierarchical rule, we reduce the dimensionality of the search space, thereby allowing for fast all-subset selection, which is usually computationally prohibitive. To focus on prediction performance of a selected model, we use area under the ROC curve as the criterion to rank all admissible models. We compare the procedure to a logistic regression-based approach and a stepwise regression that ignores the hierarchical structure. We use the procedure to construct a psychosis screening test to be used at a primary care clinic that will optimally screen low-income, Latino psychotic patients for further specialty referral.

Pre-meal anxiety and food intake in anorexia nervosa.

Anorexia nervosa (AN) is a serious mental illness characterized by reduced caloric intake that often persists after acute weight restoration. This preliminary study assesses the relationship between pre-meal anxiety and food intake in recently weight-restored individuals with AN. We hypothesized that pre-meal anxiety is inversely related to caloric intake in AN. Caloric intake and pre-meal anxiety were measured in three laboratory-based assessments (yogurt snack, multi-item lunch, macaroni and cheese lunch). Anxiety was measured by Spielberger State-Trait Anxiety Inventory (STAI-S) administered prior to the meal. Acutely weight-restored patients with AN were compared with healthy controls (HCs). Associations between anxiety and intake were analyzed first within each meal type separately and then using a model to combine the sample. In the multi-item lunch and the macaroni and cheese lunch, AN ate significantly less than HC (p=0.01, p<0.001). Pre-meal anxiety was significantly correlated with intake among AN, but not HC. In the yogurt snack, there was no significant association between anxiety and intake among patients or controls, and the groups did not differ in caloric intake. The association between pre-meal anxiety and intake among weight-restored individuals with AN suggests a potential target for relapse prevention treatment.

Omega-6: Omega-3 PUFA Ratio, Pain, Functioning, and Distress in Adults With Knee Pain.

OBJECTIVES: Osteoarthritis (OA) is associated with inflammation, chronic pain, functional limitations, and psychosocial distress. High omega-3 (n-3) polyunsaturated fatty acids (PUFAs) are associated with lower levels of inflammatory mediators, anti-nociception, and adaptive cognitive/emotional functioning. High omega-6 (n-6) PUFAs are associated with inflammation, nociception, and psychological distress. While findings related to n-3 supplementation in knee OA are mixed, consideration of the n-6:n-3 ratio and additional outcome measures may provide improved understanding of the potential relevance of these fatty acids in OA. On the basis of recommended and typical ranges of the n-6:n-3 ratio, we hypothesized that in adults with knee pain, those with a high n-6:n-3 ratio would have greater pain/functional limitations, experimental pain sensitivity, and psychosocial distress compared with those with a low n-6:n-3 ratio. MATERIALS AND METHODS: A cross-sectional investigation of clinical and experimental pain and physical and psychosocial functioning was completed in 167 adults ages 45 to 85 meeting knee OA screening criteria. Blood samples were collected and the plasma n-6:n-3 PUFA ratio determined. Quartile splits were computed and low (n=42) and high (n=41) ratio groups were compared. RESULTS: The high ratio group reported greater pain and functional limitations, (all Ps<0.04), mechanical temporal summation (hand and knee, P<0.05), and perceived stress (P=0.008) but not depressive symptoms. DISCUSSION: In adults with knee pain, a high n-6:n-3 ratio is associated with greater clinical pain/functional limitations, experimental pain sensitivity, and psychosocial distress compared with a low ratio group. Findings support consideration of the n-6:n-3 PUFA ratio and additional clinical endpoints in future research efforts.

Frontal Gamma-Aminobutyric Acid Concentrations Are Associated With Cognitive Performance in Older Adults.

BACKGROUND: Gamma-aminobutyric acid (GABA), the brain's principal inhibitory neurotransmitter, has been associated with perceptual and attentional functioning. Recent application of magnetic resonance spectroscopy (MRS) provides in vivo evidence for decreasing GABA concentrations during adulthood. It is unclear, however, how age-related decrements in cerebral GABA concentrations contribute to cognitive decline, or whether previously reported declines in cerebral GABA concentrations persist during healthy aging. We hypothesized that participants with higher GABA concentrations in the frontal cortex would exhibit superior cognitive function and that previously reported age-related decreases in cortical GABA concentrations continue into old age. METHODS: We measured GABA concentrations in frontal and posterior midline cerebral regions using a Mescher-Garwood point-resolved spectroscopy (MEGA-PRESS) 1H-MRS approach in 94 older adults without history or clinical evidence of mild cognitive impairment or dementia (mean age, 73 years). We administered the Montreal Cognitive Assessment to assess cognitive functioning. RESULTS: Greater frontal GABA concentrations were associated with superior cognitive performance. This relation remained significant after controlling for age, years of education, and brain atrophy. GABA concentrations in both frontal and posterior regions decreased as a function of age. CONCLUSIONS: These novel findings from a large, healthy, older population indicate that cognitive function is sensitive to cerebral GABA concentrations in the frontal cortex, and GABA concentration in frontal and posterior regions continue to decline in later age. These effects suggest that proton MRS may provide a clinically useful method for the assessment of normal and abnormal age-related cognitive changes and the associated physiological contributors.

Accelerated aging in adults with knee osteoarthritis pain: consideration for frequency, intensity, time, and total pain sites.

INTRODUCTION: Individuals with osteoarthritis (OA) show increased morbidity and mortality. Telomere length, a measure of cellular aging, predicts increased morbidity and mortality. Telomeres shorten with persisting biological and psychosocial stress. Living with chronic OA pain is stressful. Previous research exploring telomere length in people with OA has produced inconsistent results. Considering pain severity may clarify the relationship between OA and telomeres. OBJECTIVES: We hypothesized that individuals with high OA chronic pain severity would have shorter telomeres than those with no or low chronic pain severity. METHODS: One hundred thirty-six adults, ages 45 to 85 years old, with and without symptomatic knee OA were included in the analysis. Peripheral blood leukocyte telomere length was measured, and demographic, clinical, and functional data were collected. Participants were categorized into 5 pain severity groups based on an additive index of frequency, intensity, time or duration, and total number of pain sites (FITT). Covariates included age, sex, race or ethnicity, study site, and knee pain status. RESULTS: The no or low chronic pain severity group had significantly longer telomeres compared with the high pain severity group, P = 0.025. A significant chronic pain severity dose response emerged for telomere length, P = 0.034. The FITT chronic pain severity index was highly correlated with the clinical and functional OA pain measures. However, individual clinical and functional measures were not associated with telomere length. CONCLUSION: Results demonstrate accelerated cellular aging with high knee OA chronic pain severity and provide evidence for the potential utility of the FITT chronic pain severity index in capturing the biological burden of chronic pain.

Oxytocin's effect on resting-state functional connectivity varies by age and sex.

The neuropeptide oxytocin plays a role in social cognition and affective processing. The neural processes underlying these effects are not well understood. Modulation of connectivity strength between subcortical and cortical regions has been suggested as one possible mechanism. The current study investigated effects of intranasal oxytocin administration on resting-state functional connectivity between amygdala and medial prefrontal cortex (mPFC), as two regions involved in social-cognitive and affective processing. Going beyond previous work that largely examined young male participants, our study comprised young and older men and women to identify age and sex variations in oxytocin's central processes. This approach was based on known hormonal differences among these groups and emerging evidence of sex differences in oxytocin's effects on amygdala reactivity and age-by-sex-modulated effects of oxytocin in affective processing. In a double-blind design, 79 participants were randomly assigned to self-administer either intranasal oxytocin or placebo before undergoing resting-state functional magnetic resonance imaging. Using a targeted region-to-region approach, resting-state functional connectivity strength between bilateral amygdala and mPFC was examined. Participants in the oxytocin compared to the placebo group and men compared to women had overall greater amygdala-mPFC connectivity strength at rest. These main effects were qualified by a significant three-way interaction: while oxytocin compared to placebo administration increased resting-state amygdala-mPFC connectivity for young women, oxytocin did not significantly influence connectivity in the other age-by-sex subgroups. This study provides novel evidence of age-by-sex differences in how oxytocin modulates resting-state brain connectivity, furthering our understanding of how oxytocin affects brain networks at rest.

Investigating the Burden of Chronic Pain: An Inflammatory and Metabolic Composite.

Background. Chronic pain is associated with increased morbidity and mortality, predominated by cardiovascular disease and cancer. Investigating related risk factor measures may elucidate the biological burden of chronic pain. Objectives. We hypothesized that chronic pain severity would be positively associated with the risk factor composite. Methods. Data from 12,982 participants in the 6th Tromsø study were analyzed. Questionnaires included demographics, health behaviors, medical comorbidities, and chronic pain symptoms. The risk factor composite was comprised of body mass index, fibrinogen, C-reactive protein, and triglycerides. Chronic pain severity was characterized by frequency, intensity, time/duration, and total number of pain sites. Results. Individuals with chronic pain had a greater risk factor composite than individuals without chronic pain controlling for covariates and after excluding inflammation-related health conditions (p < 0.001). A significant "dose-response" relationship was demonstrated with pain severity (p < 0.001). In individuals with chronic pain, the risk factor composite varied by health behavior, exercise, lower levels and smoking, and higher levels. Discussion. The risk factor composite was higher in individuals with chronic pain, greater with increasing pain severity, and influenced by health behaviors. Conclusions. Identification of a biological composite sensitive to pain severity and adaptive/maladaptive behaviors would have significant clinical and research utility.