Fast Bootstrap Confidence Intervals for Continuous Threshold Linear Regression

The global burden of heart failure attributable to interstitial lung diseases: insights from the global burden of disease study 2021

BACKGROUND

The burden of interstitial lung disease (ILD)-associated heart failure (HF) poses a significant challenge to the prognosis of ILD patients. This study aimed to characterize the disease burden and analyse future trends of ILD-associated HF, offering valuable insights to inform targeted prevention and control strategies.

METHODS

Data on the prevalence and years lived with disability (YLDs) of ILD-associated HF were retrieved from the Global Burden of Disease (GBD) database for the period 1990-2021. Trends in ILD-associated HF were evaluated using average annual percentage change (AAPC) and percentage change analyses. Future prevalence data were projected up to 2050 using predictive modelling.

RESULTS

Globally, the number of patients with ILD-associated HF increased from 20,229 in 1990 to 104,059 in 2021, with the prevalence rising from 0.53 to 1.41 per 100,000 population. Prevalence rates were disproportionately higher in older populations, with individuals over 95 years experiencing a 17.78-fold increase over the study period. Additionally, a positive correlation was observed between higher socioeconomic development index (SDI) levels and ILD-associated HF prevalence. Among 204 countries, 71.1% exhibited an increasing trend in prevalence. However, Bayesian age-period-cohort (BAPC) modelling predicts a declining trend over the next 28 years.

CONCLUSION

Over the past three decades, the global burden of ILD-associated HF has escalated, particularly among individuals aged over 65 and in regions with high SDI levels. These findings underscore the need for region-specific, personalized intervention strategies to mitigate disease progression and enhance the quality of life for ILD patients.

Trend analysis and age-period-cohort effects on morbidity and mortality of liver cancer from 2010 to 2020 in Guangzhou, China

Introduction

Liver cancer is one of the most common malignant gastrointestinal tumors worldwide. This study intends to provide insight into the epidemiological characteristics and development trends of liver cancer incidence and mortality from 2010 to 2020 in Guangzhou, China.

Methods

Data were collected from the Cancer Registry and Reporting Office of Guangzhou Center for Disease Control and Prevention. Cross-sectional study, Joinpoint regression (JPR) model, and Age-Period-Cohort (APC) model were conducted to analyze the age-standardized incidence rate (ASIR) and age-standardized mortality rate (ASMR) trend of liver cancer among the entire study period.

Results

The age-standardized incidence and mortality of liver cancer in Guangzhou showed an overall decreasing trend. The disparity in risk of morbidity and mortality between the two sexes for liver cancer is increasing. The cohort effect was the most significant among those born in 1965~1969, and the risk of liver cancer incidence and mortality in the total population increased and then decreased with the birth cohort. Compared with the birth cohort born in 1950~1954 (the reference cohort), the risk of liver cancer incidence and mortality in the males born in 1995~1999 decreased by 32% and 41%, respectively, while the risk in the females decreased by 31% and 32%, respectively.

Conclusions

The early detection, prevention, clinical diagnosis, and treatment of liver cancer in Guangzhou have made remarkable achievements in recent years. However, the risk of liver cancer in the elderly and the middle-aged males is still at a high level. Therefore, the publicity of knowledge related to the prevention and treatment of liver cancer among the relevant population groups should be actively carried out to enhance the rate of early diagnosis and treatment of liver cancer and to advocate a healthier lifestyle.

Modeling and estimating a threshold effect: An application to improving cardiac surgery practices

Estimating thresholds when a threshold effect exists has important applications in biomedical research. However, models/methods commonly used in the biomedical literature may lead to a biased estimate. For patients undergoing coronary artery bypass grafting (CABG), it is thought that exposure to low oxygen delivery (DO2) contributes to an increased risk of avoidable acute kidney injury. This research is motivated by estimating the threshold of nadir DO2 for CABG patients to help develop an evidence-based guideline for improving cardiac surgery practices. We review several models (sudden-jump model, broken-stick model, and the constrained broken-stick model) that can be adopted to estimate the threshold and discuss modeling assumptions, scientific plausibility, and implications in estimating the threshold. Under each model, various estimation methods are studied and compared. In particular, under a constrained broken-stick model, a modified two-step Newton-Raphson algorithm is introduced. Through comprehensive simulation studies and an application to data on CABG patients from the University of Michigan, we show that the constrained broken-stick model is flexible, more robust, and able to incorporate scientific knowledge to improve efficiency. The two-step Newton-Raphson algorithm has good computational performances relative to existing methods.

L-Lysine supplementation affects dietary protein quality and growth and serum amino acid concentrations in rats

Single amino acid (AA) supplementations in foods are increasing, however their potential nutritional and physiological impacts are not fully understood. This study examined the effects of L-lysine (Lys) supplementation on protein quality of diets, serum AA concentrations and associations between the ratio of supplemental Lys to dietary protein (X) with body weight gain (BWG) in Sprague-Dawley male rats. Rats were fed one of 10 diets containing either 7% or 20% casein and supplemented with 0% (Control), 1.5%, 3%, 6% Lys or 6% Lys + 3% L-arginine (Arg) (8 rats/diet group) for 1 week. Lys supplementation reduced the protein quality of the casein-based diets (p < 0.01). BWG was reduced by supplemental Lys when X > 0.18. Free Lys supplementation dose-dependently increased serum Lys levels (p < 0.01), while increased protein-bound Lys (1.4% vs 0.52%) had little effect on serum Lys (p > 0.05). In the 7% casein diets, ≥ 1.5% supplemental Lys reduced serum alanine, asparagine, glycine, isoleucine, leucine, serine, tyrosine, valine, carnitine, ornithine, and increased urea. Supplementation of ≥ 3% Lys additionally reduced tryptophan and increased histidine, methionine and α-aminoadipic acid (α-AAA) compared to the Control (p < 0.05). In the 20% casein diets, addition of ≥ 1.5% Lys reduced serum asparagine and threonine, and ≥ 3% Lys reduced leucine, proline, tryptophan, valine, and ornithine, and 6% Lys reduced carnitine, and increased histidine, methionine, and α-AAA. Overall, this study showed that free Lys supplementation in a Lys-sufficient diet reduced the protein quality of the diets and modified the serum concentrations of many amino acids. Excess free Lys intake adversely affected growth and utilization of nutrients due to AA imbalance or antagonism. Overall lower protein intake increases susceptibility to the adverse effects of Lys supplementation.

Estimation and inference for multikink expectile regression with longitudinal data

In this article, we investigate parameter estimation, kink points testing and statistical inference for a longitudinal multikink expectile regression model. The estimators for the kink locations and regression coefficients are obtained by using a bootstrap restarting iterative algorithm to avoid local minima. A backward selection procedure based on a modified BIC is applied to estimate the number of kink points. We theoretically demonstrate the number selection consistency of kink points and the asymptotic normality of all estimators. In particular, the estimators of kink locations are shown to achieve root-n consistency. A weighted cumulative sum type statistic is proposed to test the existence of kink effects at a given expectile, and its limiting distributions are derived under both the null and the local alternative hypotheses. The traditional Wald-type and cluster bootstrap confidence intervals for kink locations are also constructed. Simulation studies show that the proposed estimators and test have desirable finite sample performance in both homoscedastic and heteroscedastic errors. Two applications to the Nation Growth, Lung and Health Study and Capital Bike sharing dataset in Washington D.C. are also presented. The R codes for simulation studies and the real data are available at https://github.com/wangleink/MKER.

Fast Grid Search and Bootstrap-based Inference for Continuous Two-phase Polynomial Regression Models

Two-phase polynomial regression models (Robison, 1964; Fuller, 1969; Gallant and Fuller, 1973; Zhan et al., 1996) are widely used in ecology, public health, and other applied fields to model nonlinear relationships. These models are characterized by the presence of threshold parameters, across which the mean functions are allowed to change. That the threshold is a parameter of the model to be estimated from the data is an essential feature of two-phase models. It distinguishes them, and more generally, multi-phase models, from the spline models and has profound implications for both computation and inference for the models. Estimation of two-phase polynomial regression models is a non-convex, non-smooth optimization problem. Grid search provides high quality solutions to the estimation problem, but is very slow when done by brute force. Building upon our previous work on piecewise linear two-phase regression models estimation, we develop fast grid search algorithms for two-phase polynomial regression models and demonstrate their performance. Furthermore, we develop bootstrap-based pointwise and simultaneous confidence bands for mean functions. Monte Carlo studies are conducted to demonstrate the computational and statistical properties of the proposed methods. Three real datasets are used to help illustrate the application of two-phase models, with special attention on model choice.