Search

Wearable devices enable the continuous monitoring of physical activity (PA) but generate complex functional data with poorly characterized errors. Most work on functional data views the data as smooth, latent curves obtained at discrete time intervals with some random noise with mean zero and constant variance. Viewing this noise as homoscedastic and independent ignores potential serial correlations. Our preliminary studies indicate that failing to account for these serial correlations can bias estimations. In dietary assessments, epidemiologists often use self-reported measures based on food frequency questionnaires that are prone to recall bias. With the increased availability of complex, high-dimensional functional, and scalar biomedical data potentially prone to measurement errors, it is necessary to adjust for biases induced by these errors to permit accurate analyses in various regression settings. However, there has been limited work to address measurement errors in functional and scalar covariates in the context of quantile regression. Therefore, we developed new statistical methods based on simulation extrapolation (SIMEX) and mixed effects regression with repeated measures to correct for measurement error biases in this context. We conducted simulation studies to establish the finite sample properties of our new methods. The methods are illustrated through application to a real data set.

Wearable devices permit the continuous monitoring of biological processes, such as blood glucose metabolism, and behavior, such as sleep quality and physical activity. The continuous monitoring often occurs in epochs of 60 seconds over multiple days, resulting in high dimensional longitudinal curves that are best described and analyzed as functional data. From this perspective, the functional data are smooth, latent functions obtained at discrete time intervals and prone to homoscedastic white noise. However, the assumption of homoscedastic errors might not be appropriate in this setting because the devices collect the data serially. While researchers have previously addressed measurement error in scalar covariates prone to errors, less work has been done on correcting measurement error in high dimensional longitudinal curves prone to heteroscedastic errors. We present two new methods for correcting measurement error in longitudinal functional curves prone to complex measurement error structures in multi-level generalized functional linear regression models. These methods are based on two-stage scalable regression calibration. We assume that the distribution of the scalar responses and the surrogate measures prone to heteroscedastic errors both belong in the exponential family and that the measurement errors follow Gaussian processes. In simulations and sensitivity analyses, we established some finite sample properties of these methods. In our simulations, both regression calibration methods for correcting measurement error performed better than estimators based on averaging the longitudinal functional data and using observations from a single day. We also applied the methods to assess the relationship between physical activity and type 2 diabetes in community dwelling adults in the United States who participated in the National Health and Nutrition Examination Survey.

While extensive work has been done to correct for biases due to measurement error in scalar-valued covariates prone to errors in generalized linear regression models, limited work has been done to address biases associated with functional covariates prone to errors or the combination of scalar and functional covariates prone to errors in these models. We propose Simulation Extrapolation (SIMEX) and Regression Calibration approaches to correct measurement errors associated with a mixture of functional and scalar covariates prone to classical measurement errors in generalized functional linear regression. The simulation extrapolation method is developed to handle the functional and scalar covariates prone to errors. We also develop methods based on regression calibration extended to our current measurement error settings. Extensive simulation studies are conducted to assess the finite sample performance of our developed methods. The methods are applied to the 2011-2014 cycles of the National Health and Examination Survey data to assess the relationship between physical activity and total caloric intake with type 2 diabetes among community-dwelling adults living in the United States. We treat the device-based measures of physical activity as error-prone functional covariates prone to complex arbitrary heteroscedastic errors, while the total caloric intake is considered a scalar-valued covariate prone to error. We also examine the characteristics of observed measurement errors in device-based physical activity by important demographic subgroups including age, sex, and race.

Physical activity (PA) is an important risk factor for many health outcomes. Wearable-devices such as accelerometers are increasingly used in biomedical studies to understand the associations between PA and health outcomes. Statistical analyses involving accelerometer data are challenging due to the following three characteristics: (i) high-dimensionality, (ii) temporal dependence, and (iii) measurement error. To address these challenges we treat accelerometer-based measures of physical activity as a single function-valued covariate prone to measurement error. Specifically, in order to determine the relationship between PA and a health outcome of interest, we propose a regression model with a functional covariate that accounts for measurement error. Using regression calibration, we develop a two-step estimation method for the model parameters and establish their consistency. A test is also proposed to test the significance of the estimated model parameters. Simulation studies are conducted to compare the proposed methods with existing alternative approaches under varying scenarios. Finally, the developed methods are used to assess the relationship between PA intensity and BMI obtained from the National Health and Nutrition Examination Survey data.

We treat the problem of testing for association between a functional variable belonging to Hilbert space and a scalar variable. Particularly, we propose a distribution-free test statistic based on Kendall's Tau which is one of the most popular methods to determine the association between two random variables. The distribution of the test statistic under the null hypothesis of independence is established using the theory of U-statistics taking values in a Hilbert space. We also consider the case when the functional data is sparsely observed, a situation that arises in many applications. Simulations show that the proposed method outperforms the alternatives under several conditions demonstrating the robustness of our approach. We provide data applications that further consolidate the utility of our method.

For all diseases, prevalence has been carefully studied. In the "classic" paradigm, the prevalence of different diseases has usually been studied separately. Accumulating evidences have shown that diseases can be "correlated". The joint analysis of prevalence of multiple diseases can provide important insights beyond individual-disease analysis, however, has not been well conducted. In this study, we take advantage of the uniquely valuable Taiwan National Health Insurance Research Database (NHIRD), and conduct a pan-disease analysis of period prevalence trend. The goal is to identify clusters within which diseases share similar period prevalence trends. For this purpose, a novel penalization pursuit approach is developed, which has an intuitive formulation and satisfactory properties. In data analysis, the period prevalence values are computed using records on close to 1 million subjects and 14 years of observation. For 405 diseases, 35 nontrivial clusters (with sizes larger than one) and 27 trivial clusters (with sizes one) are identified. The results differ significantly from those of the alternatives. A closer examination suggests that the clustering results have sound interpretations. This study is the first to conduct a pan-disease clustering analysis of prevalence trend using the uniquely valuable NHIRD data and can have important value in multiple aspects.

Measurement error is an important problem that has not been very well studied in the context of Functional Data Analysis. To the best of our knowledge, there are no existing methods that address the presence of functional measurement errors in generalized functional linear models. A framework is proposed for estimating the slope function in the presence of measurement error in the generalized functional linear model with a scalar response. This work extends the conditional-score method to the case when both the measurement error and the independent variables lie in an infinite dimensional space. Asymptotic results are obtained for the proposed estimate and its behavior is studied via simulations, when the response is continuous or binary. It's performance on real data is demonstrated through a simulation study based on the Canadian Weather data-set, where errors are introduced in the data-set and it is observed that the proposed estimate indeed performs better than a naive estimate that ignores the measurement error.