Ciprian M. Crainiceanu

Reduced Kidney Function as a Risk Factor for Incident Heart Failure: The Atherosclerosis Risk in Communities (ARIC) Study

Reduced kidney function is a risk factor for cardiovascular morbidity and mortality, and both heart failure (HF) and kidney failure incidences are increasing. This study therefore sought to determine the effect of decreased kidney function on HF incidence in a population-based study of middle-aged adults. From 1987 through 2002, 14,857 participants of the Atherosclerosis Risk in Communities (ARIC) study who were free of prevalent HF at baseline were followed for incident HF hospitalization or death (International Classification of Diseases, Ninth Revision/10th Revision 428/I50). Estimated GFR (eGFR) was calculated using the abbreviated Modification of Diet in Renal Disease (MDRD) Study equation, and kidney function was categorized as normal (eGFR > or =90 ml/min per 1.73 m(2); n = 7143), mildly reduced (eGFR 60 to 89 ml/min per 1.73 m(2); n = 7311), and moderately/severely reduced (eGFR <60 ml/min per 1.73 m(2); n = 403). Cox proportional hazards models were used to control for demographic and cardiovascular risk factors; analyses were stratified by the presence of coronary heart disease at baseline. During a mean follow-up of 13.2 yr, 1193 participants developed HF. The incidence of HF was three-fold higher for individuals with eGFR <60 ml/min per 1.73 m(2) compared to the reference group with eGFR > or =90 ml/min per 1.73 m(2) (18 versus 6 per 1000 person-years). The overall adjusted relative hazard of developing HF was 1.94 (1.49 to 2.53) for individuals with eGFR <60 ml/min per 1.73 m(2) compared to the reference group and was significantly increased for individuals with and without prevalent coronary heart disease at baseline. A substantially greater decline in kidney function occurred in individuals concomitant with HF hospitalization/death compared to those who did not develop HF. In summary, middle-aged adults with moderately/severely reduced kidney function are at high risk for developing HF.

Optical coherence tomography segmentation reveals ganglion cell layer pathology after optic neuritis

Post-mortem ganglion cell dropout has been observed in multiple sclerosis; however, longitudinal in vivo assessment of retinal neuronal layers following acute optic neuritis remains largely unexplored. Peripapillary retinal nerve fibre layer thickness, measured by optical coherence tomography, has been proposed as an outcome measure in studies of neuroprotective agents in multiple sclerosis, yet potential swelling during the acute stages of optic neuritis may confound baseline measurements. The objective of this study was to ascertain whether patients with multiple sclerosis or neuromyelitis optica develop retinal neuronal layer pathology following acute optic neuritis, and to systematically characterize such changes in vivo over time. Spectral domain optical coherence tomography imaging, including automated retinal layer segmentation, was performed serially in 20 participants during the acute phase of optic neuritis, and again 3 and 6 months later. Imaging was performed cross-sectionally in 98 multiple sclerosis participants, 22 neuromyelitis optica participants and 72 healthy controls. Neuronal thinning was observed in the ganglion cell layer of eyes affected by acute optic neuritis 3 and 6 months after onset (P < 0.001). Baseline ganglion cell layer thicknesses did not demonstrate swelling when compared with contralateral unaffected eyes, whereas peripapillary retinal nerve fibre layer oedema was observed in affected eyes (P = 0.008) and subsequently thinned over the course of this study. Ganglion cell layer thickness was lower in both participants with multiple sclerosis and participants with neuromyelitis optica, with and without a history of optic neuritis, when compared with healthy controls (P < 0.001) and correlated with visual function. Of all patient groups investigated, those with neuromyelitis optica and a history of optic neuritis exhibited the greatest reduction in ganglion cell layer thickness. Results from our in vivo longitudinal study demonstrate retinal neuronal layer thinning following acute optic neuritis, corroborating the hypothesis that axonal injury may cause neuronal pathology in multiple sclerosis. Further, these data provide evidence of subclinical disease activity, in both participants with multiple sclerosis and with neuromyelitis optica without a history of optic neuritis, a disease in which subclinical disease activity has not been widely appreciated. No pathology was seen in the inner or outer nuclear layers of eyes with optic neuritis, suggesting that retrograde degeneration after optic neuritis may not extend into the deeper retinal layers. The subsequent thinning of the ganglion cell layer following acute optic neuritis, in the absence of evidence of baseline swelling, suggests the potential utility of quantitative optical coherence tomography retinal layer segmentation to monitor neuroprotective effects of novel agents in therapeutic trials.

Cadmium Exposure and Hypertension in the 1999–2004 National Health and Nutrition Examination Survey (NHANES)

Introduction Cadmium induces hypertension in animal models. Epidemiologic studies of cadmium exposure and hypertension, however, have been inconsistent. Objective We aimed to investigate the association of blood and urine cadmium with blood pressure levels and with the prevalence of hypertension in U.S. adults who participated in the 1999–2004 National Health and Nutrition Examination Survey (NHANES). Methods We studied participants ≥ 20 years of age with determinations of cadmium in blood (n = 10,991) and urine (n = 3,496). Blood and urine cadmium were measured by atomic absorption spectrometry and inductively coupled plasma–mass spectrometry, respectively. Systolic and diastolic blood pressure levels were measured using a standardized protocol. Results The geometric means of blood and urine cadmium were 3.77 nmol/L and 2.46 nmol/L, respectively. After multivariable adjustment, the average differences in systolic and diastolic blood pressure comparing participants in the 90th vs. 10th percentile of the blood cadmium distribution were 1.36 mmHg [95% confidence interval (CI), −0.28 to 3.00] and 1.68 mmHg (95% CI, 0.57–2.78), respectively. The corresponding differences were 2.35 mmHg and 3.27 mmHg among never smokers, 1.69 mmHg and 1.55 mmHg among former smokers, and 0.02 mmHg and 0.69 mmHg among current smokers. No association was observed for urine cadmium with blood pressure levels, or for blood and urine cadmium with the prevalence of hypertension. Conclusions Cadmium levels in blood, but not in urine, were associated with a modest elevation in blood pressure levels. The association was stronger among never smokers, intermediate among former smokers, and small or null among current smokers. Our findings add to the concern of renal and cardiovascular cadmium toxicity at chronic low levels of exposure in the general population.

MULTILEVEL FUNCTIONAL PRINCIPAL COMPONENT ANALYSIS

The Sleep Heart Health Study (SHHS) is a comprehensive landmark study of sleep and its impacts on health outcomes. A primary metric of the SHHS is the in-home polysomnogram, which includes two electroencephalographic (EEG) channels for each subject, at two visits. The volume and importance of this data presents enormous challenges for analysis. To address these challenges, we introduce multilevel functional principal component analysis (MFPCA), a novel statistical methodology designed to extract core intra- and inter-subject geometric components of multilevel functional data. Though motivated by the SHHS, the proposed methodology is generally applicable, with potential relevance to many modern scientific studies of hierarchical or longitudinal functional outcomes. Notably, using MFPCA, we identify and quantify associations between EEG activity during sleep and adverse cardiovascular outcomes.

Penalized Functional Regression

We develop fast fitting methods for generalized functional linear models. The functional predictor is projected onto a large number of smooth eigenvectors and the coefficient function is estimated using penalized spline regression; confidence intervals based on the mixed model framework are obtained. Our method can be applied to many functional data designs including functions measured with and without error, sparsely or densely sampled. The methods also extend to the case of multiple functional predictors or functional predictors with a natural multilevel structure. The approach can be implemented using standard mixed effects software and is computationally fast. The methodology is motivated by a study of white-matter demyelination via diffusion tensor imaging (DTI). The aim of this study is to analyze differences between various cerebral white-matter tract property measurements of multiple sclerosis (MS) patients and controls. While the statistical developments proposed here were motivated by the DTI study, the methodology is designed and presented in generality and is applicable to many other areas of scientific research. An online appendix provides R implementations of all simulations.

Relationships between retinal axonal and neuronal measures and global central nervous system pathology in multiple sclerosis.

OBJECTIVE To determine the relationships between conventional and segmentation-derived optical coherence tomography (OCT) retinal layer thickness measures with intracranial volume (a surrogate of head size) and brain substructure volumes in multiple sclerosis (MS). DESIGN Cross-sectional study. SETTING Johns Hopkins University, Baltimore, Maryland. PARTICIPANTS A total of 84 patients with MS and 24 healthy control subjects. MAIN OUTCOME MEASURES High-definition spectral-domain OCT conventional and automated segmentation-derived discrete retinal layer thicknesses and 3-T magnetic resonance imaging brain substructure volumes. RESULTS Peripapillary retinal nerve fiber layer as well as composite ganglion cell layer+inner plexiform layer thicknesses in the eyes of patients with MS without a history of optic neuritis were associated with cortical gray matter (P=.01 and P=.04, respectively) and caudate (P=.04 and P=.03, respectively) volumes. Inner nuclear layer thickness, also in eyes without a history of optic neuritis, was associated with fluid-attenuated inversion recovery lesion volume (P=.007) and inversely associated with normal-appearing white matter volume (P=.005) in relapsing-remitting MS. As intracranial volume was found to be related with several of the OCT measures in patients with MS and healthy control subjects and is already known to be associated with brain substructure volumes, all OCT-brain substructure relationships were adjusted for intracranial volume. CONCLUSIONS Retinal measures reflect global central nervous system pathology in multiple sclerosis, with thicknesses of discrete retinal layers each appearing to be associated with distinct central nervous system processes. Moreover, OCT measures appear to correlate with intracranial volume in patients with MS and healthy control subjects, an important unexpected factor unaccounted for in prior studies examining the relationships between peripapillary retinal nerve fiber layer thickness and brain substructure volumes.

Cadmium exposure and all-cause and cardiovascular mortality in the U.S. general population.

Background: Urine cadmium concentrations were associated with all-cause and cardiovascular mortality in men in the 1988–1994 U.S. National Health and Nutrition Examination Survey (NHANES) population. Since 1988, cadmium exposure has decreased substantially in the United States. The associations between blood and urine cadmium and cardiovascular disease (CVD) mortality at more recent levels of exposure are unknown. Objectives: We evaluated the prospective association of blood and urine cadmium concentrations with all-cause and CVD mortality in the 1999–2004 U.S. population. Methods: We followed 8,989 participants who were ≥ 20 years of age for an average of 4.8 years. Hazard ratios for mortality end points comparing the 80th to the 20th percentiles of cadmium distributions were estimated using Cox regression. Results: The multivariable adjusted hazard ratios [95% confidence intervals (CIs)] for blood and urine cadmium were 1.50 (95% CI: 1.07, 2.10) and 1.52 (95% CI: 1.00, 2.29), respectively, for all-cause mortality, 1.69 (95% CI: 1.03, 2.77) and 1.74 (95% CI: 1.07, 2.83) for CVD mortality, 1.98 (95% CI: 1.11, 3.54) and 2.53 (95% CI: 1.54, 4.16) for heart disease mortality, and 1.73 (95% CI: 0.88, 3.40) and 2.09 (95% CI: 1.06, 4.13) for coronary heart disease mortality. The population attributable risks associated with the 80th percentile of the blood (0.80 μg/L) and urine (0.57 μg/g) cadmium distributions were 7.0 and 8.8%, respectively, for all-cause mortality and 7.5 and 9.2%, respectively, for CVD mortality Conclusions: We found strongly suggestive evidence that cadmium, at substantially low levels of exposure, remains an important determinant of all-cause and CVD mortality in a representative sample of U.S. adults. Efforts to further reduce cadmium exposure in the population could contribute to a substantial decrease in CVD disease burden.

Fast Adaptive Penalized Splines

This article proposes a numerically simple method for locally adaptive smoothing. The heterogeneous regression function is modeled as a penalized spline with a varying smoothing parameter modeled as another penalized spline. This is formulated as a hierarchical mixed model, with spline coefficients following zero mean normal distribution with a smooth variance structure. The major contribution of this article is to use the Laplace approximation of the marginal likelihood for estimation. This method is numerically simple and fast. The idea is extended to spatial and non-normal response smoothing.

Longitudinal functional principal component analysis

We introduce models for the analysis of functional data observed at multiple time points. The dynamic behavior of functional data is decomposed into a time-dependent population average, baseline (or static) subject-specific variability, longitudinal (or dynamic) subject-specific variability, subject-visit-specific variability and measurement error. The model can be viewed as the functional analog of the classical longitudinal mixed effects model where random effects are replaced by random processes. Methods have wide applicability and are computationally feasible for moderate and large data sets. Computational feasibility is assured by using principal component bases for the functional processes. The methodology is motivated by and applied to a diffusion tensor imaging (DTI) study designed to analyze differences and changes in brain connectivity in healthy volunteers and multiple sclerosis (MS) patients. An R implementation is provided.87.

Restricted Likelihood Ratio Testing for Zero Variance Components in Linear Mixed Models

The goal of our article is to provide a transparent, robust, and computationally feasible statistical platform for restricted likelihood ratio testing (RLRT) for zero variance components in linear mixed models. This problem is nonstandard because under the null hypothesis the parameter is on the boundary of the parameter space. Our proposed approach is different from the asymptotic results of Stram and Lee who assumed that the outcome vector can be partitioned into many independent subvectors. Thus, our methodology applies to a wider class of mixed models, which includes models with a moderate number of clusters or nonparametric smoothing components. We propose two approximations to the finite sample null distribution of the RLRT statistic. Both approximations converge weakly to the asymptotic distribution obtained by Stram and Lee when their assumptions hold. When their assumptions do not hold, we show in extensive simulation studies that both approximations outperform the Stram and Lee approximation and the parametric bootstrap. We also identify and address numerical problems associated with standard mixed model software. Our methods are motivated by and applied to a large longitudinal study on air pollution health effects in a highly susceptible cohort. Relevant software is posted as an online supplement.