Skip to main content
eScholarship
Open Access Publications from the University of California

Scholarly Works (16 results)

  • Thesis
  • Peer Reviewed
UC Irvine Electronic Theses and Dissertations (2023)

In a longitudinal cohort study, a group of subjects is chosen based on certain characteristicsand then followed at routine intervals over time. At each visit, some measurements are recorded and the goal is oftentimes to model how they develop over time. During the followup, the collection of the data can be stopped by a terminal event. In this dissertation, we study several statistical methods for modeling longitudinal data while adjusting for the terminal event. In Chapter II, we propose a nonparametric bivariate time-varying coefficient model for longitudinal measurements with the occurrence of a terminal event that is subject to right censoring. The time-varying coefficients capture the longitudinal trajectories of covariate effects along with both the followup time and the residual lifetime, in contrast to the existing work that either models longitudinal measures as a function of only the forward time or the backward time, or poses strong parametric assumptions. We consider a kernel smoothing method for estimating regression coefficients in our model and use cross-validation for bandwidth selection, applying undersmoothing in the final analysis to eliminate the asymptotic bias of the kernel estimator. We show that the kernel estimates follow a finite-dimensional normal distribution asymptotically under mild regularity conditions, and provide an easily computed sandwich covariance matrix estimator. In Chapter III, we study the lifetime Medicare spending patterns of patients with end-stage renal disease (ESRD) stratified by waitlisting and kidney transplant. In addition to the terminal event, a non-terminal event that happened in the follow-up could also have an impact on the longitudinal measures. To study the heterogeneous Medicare cost trajectories across groups stratified by waitlisting and transplant, we proposed two models: a semiparametric regression model with both fixed and bivariate time-varying coefficients to compare unwaitlisted and waitlisted groups, and a bivariate time-varying coefficient model with different starting times (time since the first ESRD service and time since the kidney transplant) to compare untransplanted and transplanted groups. We use sandwich variance estimators to construct confidence intervals and validate inference procedures through simulations. Our analysis of the Medicare claims data reveals that waitlisting is associated with a lower daily medical cost at the beginning of ESRD service among waitlisted patients which gradually increases over time. Averaging over lifespan, however, there is no difference between waitlisted and unwaitlisted groups. A kidney transplant, on the other hand, reduces the medical cost significantly after an initial spike. In Chapter IV, we study how the onset of a non-terminal events is associated with the terminal event. Existing methods under the framework of multi-state model or the semi-competing risks model rely on certain semiparametric assumptions for modeling the joint distribution of these two events, thus may subject to model misspecification and lack clear interpretations of their association. Moreover, they assume the independence between the censoring time C and both event times (S, T), which can be easily violated in realistic situations. We propose to estimate the onset of non-terminal event conditional on the terminal event time using a generalized Beran’s estimator. We consider a left truncation time L in addition to the right censoring time C and only need a relaxed assumption of the independence between (L,C) and the non-terminal event time S conditional on the terminal event time T. Such estimates also enjoy more meaningful interpretations.

    Cover page: Statistical Methods for Cohort Studies with Terminal Events
    • Thesis
    • Peer Reviewed
    UC Irvine Electronic Theses and Dissertations (2024)

    This dissertation presents novel neural network-based methods for estimating conditional distribution functions across diverse settings. We introduce a robust approach for survival analysis with right-censored and time-varying covariates, estimating the conditional hazard function and deriving the conditional survival function. This method outperforms traditional Cox proportional hazards models and recent neural network models in both simulated and real-world data scenarios. Expanding the methodology to uncensored data, we propose a unified approach for estimating conditional distribution functions for continuous responses with multiple covariates. Our method shows improved robustness, accuracy, and computational efficiency compared to kernel-based and mean regression neural network methods. Further, we address limitations of traditional models by using DeepONet to capture complex, long-term effects of covariates on hazard functions, demonstrating enhanced flexibility and adaptability. We also develop a deep operator neural network framework for functional data analysis, where covariates are functions, paving the way for advanced analysis of complex data structures.

      Cover page: Conditional Distribution Function Estimation Using Neural NetworksCreative Commons 'BY-NC-ND' version 4.0 license
      • Article
      • Peer Reviewed
      UC Irvine Previously Published Works (2019)
      Cross-ratio is an important local measure of the strength of dependence among correlated failure times. If a covariate is available, it may be of scientific interest to understand how the cross-ratio varies with the covariate as well as time components. Motivated by the Tremin study, where the dependence between age at a marker event reflecting early lengthening of menstrual cycles and age at menopause may be affected by age at menarche, we propose a proportional cross-ratio model through a baseline cross-ratio function and a multiplicative covariate effect. Assuming a parametric model for the baseline cross-ratio, we generalize the pseudo-partial likelihood approach of Hu et al. (Biometrika 98:341-354, 2011) to the joint estimation of the baseline cross-ratio and the covariate effect. We show that the proposed parameter estimator is consistent and asymptotically normal. The performance of the proposed technique in finite samples is examined using simulation studies. In addition, the proposed method is applied to the Tremin study for the dependence between age at a marker event and age at menopause adjusting for age at menarche. The method is also applied to the Australian twin data for the estimation of zygosity effect on cross-ratio for age at appendicitis between twin pairs.
        Cover page: Proportional cross-ratio model
        • Article
        • Peer Reviewed
        UC Irvine Previously Published Works (2023)
        Modeling and drawing inference on the joint associations between single-nucleotide polymorphisms and a disease has sparked interest in genome-wide associations studies. In the motivating Boston Lung Cancer Survival Cohort (BLCSC) data, the presence of a large number of single nucleotide polymorphisms of interest, though smaller than the sample size, challenges inference on their joint associations with the disease outcome. In similar settings, we find that neither the debiased lasso approach (van de Geer et al., 2014), which assumes sparsity on the inverse information matrix, nor the standard maximum likelihood method can yield confidence intervals with satisfactory coverage probabilities for generalized linear models. Under this large n, diverging p scenario, we propose an alternative debiased lasso approach by directly inverting the Hessian matrix without imposing the matrix sparsity assumption, which further reduces bias compared to the original debiased lasso and ensures valid confidence intervals with nominal coverage probabilities. We establish the asymptotic distributions of any linear combinations of the parameter estimates, which lays the theoretical ground for drawing inference. Simulations show that the proposed refined debiased estimating method performs well in removing bias and yields honest confidence interval coverage. We use the proposed method to analyze the aforementioned BLCSC data, a large-scale hospital-based epidemiology cohort study investigating the joint effects of genetic variants on lung cancer risks.
          Cover page: Debiased lasso for generalized linear models with a diverging number of covariates.
          • Article
          • Peer Reviewed
          UC Irvine Previously Published Works (2023)
          For statistical inference on regression models with a diverging number of covariates, the existing literature typically makes sparsity assumptions on the inverse of the Fisher information matrix. Such assumptions, however, are often violated under Cox proportion hazards models, leading to biased estimates with under-coverage confidence intervals. We propose a modified debiased lasso method, which solves a series of quadratic programming problems to approximate the inverse information matrix without posing sparse matrix assumptions. We establish asymptotic results for the estimated regression coefficients when the dimension of covariates diverges with the sample size. As demonstrated by extensive simulations, our proposed method provides consistent estimates and confidence intervals with nominal coverage probabilities. The utility of the method is further demonstrated by assessing the effects of genetic markers on patients' overall survival with the Boston Lung Cancer Survival Cohort, a large-scale epidemiology study investigating mechanisms underlying the lung cancer.
            Cover page: Statistical inference for Cox proportional hazards models with a diverging number of covariates
            • Article
            • Peer Reviewed
            UC Irvine Previously Published Works (2014)

            Objective

            This study aims to examine whether blood pressure reductions differ by estrogen use among overweight glucose-intolerant women.

            Methods

            We conducted a secondary analysis of Diabetes Prevention Program postmenopausal participants who used oral estrogen with or without progestogen at baseline and 1-year follow-up (n = 324) versus those who did not use oral estrogen with or without progestogen at either time point (n = 382). Changes in systolic blood pressure (SBP) and diastolic blood pressure (DBP) were examined by randomization arm (intensive lifestyle change [ILS], metformin 850 mg twice daily, or placebo). Associations between changes in blood pressure and changes in sex hormone-binding globulin, estradiol, testosterone, and dehydroepiandrosterone were also examined.

            Results

            Estrogen users and nonusers had similar prevalences of baseline hypertension (33% vs 34%, P = 0.82) and use of blood pressure medications at baseline (P = 0.25) and on follow-up (P = 0.10). Estrogen users and nonusers randomized to ILS had similar decreases in SBP (-3.3 vs -4.7 mm Hg, P = 0.45) and DBP (-3.1 vs -4.7 mm Hg, P = 0.16). Among estrogen users, women randomized to ILS had significant declines in SBP (P = 0.016) and DBP (P = 0.009) versus placebo. Among nonusers, women randomized to ILS had significant declines in DBP (P = 0.001) versus placebo, but declines in SBP were not significant (P = 0.11). Metformin was not associated with blood pressure reductions versus placebo regardless of estrogen therapy. Blood pressure changes were not associated with changes in sex hormones regardless of estrogen therapy.

            Conclusions

            Among overweight women with dysglycemia, the magnitude of blood pressure reductions after ILS is unrelated to postmenopausal estrogen use.
              Cover page: Does hormone therapy affect blood pressure changes in the Diabetes Prevention Program?
              • Article
              • Peer Reviewed
              UC San Diego Previously Published Works (2022)
              Background: Understanding medication use patterns for patients with COVID-19 will provide needed insight into the evolution of COVID-19 treatment over the course of the SARS-CoV-2 pandemic and aid clinical management considerations. Objectives: To systematically determine most frequently used medications among COVID-19 patients overall and by hospitalization status. Secondary objective was use measurement of medications considered potential therapeutic options. Methods: Retrospective cohort study was performed using data from the University of California COVID Research Data Set (UC CORDS) patients between March 10, 2020, and December 31, 2020. Main outcomes were percentages of patients prescribed medications, overall, by age group, and by comorbidity based on hospitalization status for COVID-19 patients. Use percentage by month of COVID-19 diagnosis was measured. Cumulative count of potential therapeutic options was measured over time. Results: Dataset included 22 896 unique patients with COVID-19 (mean [SD] age, 42.4 [20.4] years; 12 154 [53%] women). Most frequently used medications in patients overall were acetaminophen (21.2%), albuterol (14.9%), ondansetron (13.9%), and enoxaparin (10.8%). Dexamethasone use increased from fewer than 50 total hospitalized patients through April who had received the medication, to more than 500 patients by mid-August. Cumulative count of enoxaparin users was the largest throughout the study period. Conclusion and Relevance: In this retrospective cohort study, across age and comorbidity groups, predominant utilization was for supportive care therapy. Dexamethasone and remdesivir experienced large increases in use. Conversely, hydroxychloroquine and azithromycin use markedly dropped. Medication utilization rapidly shifted toward more evidence-concordant treatment of patients with COVID-19 as rigorous study findings emerged.
                Cover page: Examination of Medication Use Patterns by Age Group, Comorbidity, and Month in COVID-19 Positive Patients in a Large Statewide Health System During the Pandemic in 2020.
                • Article
                • Peer Reviewed
                UC San Diego Previously Published Works (2015)

                Objective

                Persons with diabetes have accelerated muscle loss. The association of fasting and postchallenge glucose levels per se to grip strength, a clinical marker of poor physical function, and potential sex differences in this relationship has not been previously described.

                Design

                Longitudinal cohort.

                Setting

                USA.

                Participants

                Participants were community-dwelling older adults (mean age 71.3 years) without self-reported diabetes and/or use of diabetes medication with glucose measured at baseline (1992-1996).

                Measurements

                Fasting plasma glucose (FPG) was measured in 1019 women and 636 men. Two-hour glucose (2HG) levels after a 75 g oral glucose tolerance test were also available (women, n=870; men, n=559). Dominant hand grip strength was assessed using a hand-held dynamometer at 3.0±1.6 visits over a median 7.0 years. Mixed linear models examined the association of baseline glucose levels with grip strength, accounting for repeated visits, and adjusting for covariates.

                Results

                Sex-specific FPG quartiles were associated with unadjusted differences in grip strength among women (p=0.03) but not men (p=0.50). However, in men, adjusting for age, education, height, weight, peripheral neuropathy, physical activity, and comorbidities, each SD (SD=17 mg/dL) higher FPG was associated with persistently lower grip strength (-0.44±0.22 kg, p=0.049); 2HG (SD=50 mg/dL) was unrelated to grip strength (-0.39±0.25 kg, p=0.13). In women, neither FPG (SD=16 mg/dL) nor 2HG (SD=45 mg/dL) was associated with grip strength (0.02±0.12 kg, p=0.90; and -0.20±0.14 kg, p=0.14; respectively) after adjustment. The rate of change in grip strength did not differ across FPG or 2HG quartiles in either sex.

                Conclusions

                In age-adjusted analyses, elevated fasting glucose levels are associated with persistently lower grip strength in older men, but not women. Future studies are needed to elucidate reasons for these sex differences and may provide further insight into accelerated loss of muscle function as a complication of diabetes in older adults.
                  Cover page: Sex differences in the association of fasting and postchallenge glucose levels with grip strength among older adults: the Rancho Bernardo Study
                  • Article
                  • Peer Reviewed
                  UC San Diego Previously Published Works (2015)

                  Objectives

                  To examine the impact of a weight loss intervention upon follicle stimulating hormone (FSH) levels in postmenopause.

                  Methods

                  Participants were postmenopausal, overweight, glucose-intolerant women not using exogenous estrogen (n = 382) in the Diabetes Prevention Program. Women were randomized to intensive lifestyle change (ILS) with the goals of weight reduction of at least 7% of initial weight and 150 min per week of moderate-intensity exercise, metformin 850 mg twice a day, or placebo administered twice a day.

                  Results

                  Randomization to ILS led to small increases in FSH between baseline and 1-year follow-up vs. placebo (2.3 IU/l vs. -0.81 IU/l, P < 0.01). Increases in FSH were correlated with decreases in weight (r = -0.165, P < 0.01) and estradiol (E2) (r = -0.464, P < 0.0001) after adjustment for age, race/ethnicity, and randomization arm. Changes in FSH were still significantly associated with changes in weight even after adjustment for E2 levels. Metformin users had reductions in weight but non-significant changes in FSH and E2 levels vs. placebo.

                  Conclusions

                  Weight loss leads to small increases in FSH among overweight, postmenopausal women, potentially through pathways mediated by endogenous estrogen as well as other pathways.
                    Cover page: Weight loss decreases follicle stimulating hormone in overweight postmenopausal women
                    • Article
                    • Peer Reviewed
                    UC Irvine Previously Published Works (2017)

                    Objective

                    The menopausal transition (MT) is a critical period associated with physiologic changes that influence women's long-term health and longevity. Information is, however, limited regarding factors that influence age at the onset of the MT and its duration (ie, time from MT onset to the final menstrual period).

                    Methods

                    We analyzed data for 1,145 women from four sites of the Study of Women's Health Across the Nation who participated in the menstrual calendar substudy, had the start of the MT identified, and had no missing covariate information. Participants included from four racial/ethnic groups: African American, white, Chinese, and Japanese. Women completed daily menstrual calendars from 1996 to 2006 and questions on hormone therapy use monthly. Baseline measures included education, economic strain, and menstrual cycle characteristics. Annual measures included height, weight, and smoking status. Cox proportional hazards models were used to analyze the data.

                    Results

                    The adjusted median duration of the MT ranged from 4.37 years among the oldest age-at-onset quartile to 8.57 years among the youngest age-at-onset quartile (P < 0.001). Cigarette smoking was associated with an earlier onset (P < 0.001) and a shorter duration (P < 0.001). African American women had a longer duration (P = 0.012) than white women. Body mass index was associated with a later onset of the MT (P = 0.001) but not its duration.

                    Conclusions

                    The duration of the MT was largely influenced by the age at which it began: earlier onset was associated with a longer transition. This finding provides a strong rationale for developing improved markers of the onset of the early MT.
                      Cover page: Duration of the menopausal transition is longer in women with young age at onset
                      Top