Cumulative effects of excess high-normal alanine aminotransferase levels in relation to new-onset metabolic dysfunction-associated fatty liver disease in China

BACKGROUND

Within the normal range, elevated alanine aminotransferase (ALT) levels are associated with an increased risk of metabolic dysfunction-associated fatty liver disease (MAFLD).

AIM

To investigate the associations between repeated high-normal ALT measurements and the risk of new-onset MAFLD prospectively.

METHODS

A cohort of 3553 participants followed for four consecutive health examinations over 4 years was selected. The incidence rate, cumulative times, and equally and unequally weighted cumulative effects of excess high-normal ALT levels (ehALT) were measured. Cox proportional hazards regression was used to analyse the association between the cumulative effects of ehALT and the risk of new-onset MAFLD.

RESULTS

A total of 83.13% of participants with MAFLD had normal ALT levels. The incidence rate of MAFLD showed a linear increasing trend in the cumulative ehALT group. Compared with those in the low-normal ALT group, the multivariate adjusted hazard ratios of the equally and unequally weighted cumulative effects of ehALT were 1.651 [95% confidence interval (CI): 1.199-2.273] and 1.535 (95%CI: 1.119-2.106) in the third quartile and 1.616 (95%CI: 1.162-2.246) and 1.580 (95%CI: 1.155-2.162) in the fourth quartile, respectively.

CONCLUSION

Most participants with MAFLD had normal ALT levels. Long-term high-normal ALT levels were associated with a cumulative increased risk of new-onset MAFLD.

Clinical consequences of a genetic predisposition toward higher benign prostate-specific antigen levels

BACKGROUND

Prostate-specific antigen (PSA) levels are influenced by genetic variation unrelated to prostate cancer risk. Whether a genetic predisposition to a higher PSA level predisposes to a diagnostic work-up for prostate cancer is not known.

METHODS

Participants were 3110 men of African and European ancestries ages 45-70, without prostate cancer and with a baseline PSA < 4 ng/mL, undergoing routine clinical PSA screening. The exposure was a polygenic score (PGS) comprising 111 single nucleotide polymorphisms associated with PSA level, but not prostate cancer. We tested whether the PGS was associated with a: 1) PSA value > 4 ng/mL, 2) International Classification of Diseases (ICD) code for an elevated PSA, 3) encounter with a urologist, or 4) prostate biopsy. Multivariable Cox proportional hazards models were adjusted for age and genetic principal components. Analyses were stratified by age (45-59 years, and 60-70 years old). Association estimates are per standard deviation change in the PGS.

FINDINGS

The median age was 56.6 years, and 2118 (68%) participants were 45-59 years. The median (IQR) baseline PSA level was 1.0 (0.6-1.7) ng/mL. Among men ages 45-59, the PGS was associated with a PSA > 4 (hazard ratio [HR] = 1.35 [95% CI, 1.17-1.57], p = 4.5 × 10-5), an ICD code for elevated PSA (HR = 1.30 [1.12-1.52], p = 8.0 × 10-4), a urological evaluation (HR = 1.34 [1.14-1.57], p = 4.8 × 10-4), and undergoing a prostate biopsy (HR = 1.35 [1.11-1.64], p = 0.002). Among men ages 60-70, association effect sizes were smaller and not significant.

INTERPRETATION

A predisposition toward higher PSA levels was associated with clinical evaluations of an elevated PSA among men ages 45-59 years.

FUNDING

National Institutes of Health (NIH).

Effectiveness of 13-valent pneumococcal conjugate vaccine on all-cause pneumonia in children under 5 years in Shanghai, China: An observational study

BACKGROUND

Streptococcus pneumoniae (Spn) is a common respiratory pathogen and the main cause of bacterial pneumonia, meningitis, and bacteremia, acute otitis media. Imported 13-valent pneumococcal conjugate vaccine (PCV13) was licensed in China and introduced in Shanghai in 2017. We aim to describe PCV13 vaccination trends and pneumonia incidence of children under 5 from 2017 to 2020, then estimate the effectiveness of PCV13 against community-acquired pneumonia (CAP) in children under 5 in Shanghai, China.

METHODS

By calculating propensity scores with logistic regression, a comparison group was formed by frequency matching one unvaccinated child to one vaccinated child. For matching, we used the nearest-neighbor matching algorithm and exact matching, and then created distinct matched analysis sets for two cohorts. A Kaplan-Meier analysis was conducted to measure the cumulative incidence of all-cause pneumonia in both groups and used the log-rank test to assess the differences between the two cumulative incidence curves. Cox proportional hazards regression was used to compare the adjusted hazard ratios (HR) of differences in all-cause pneumonia between the two groups.

RESULTS

Children received three or more doses PCV13 accounted for 85.7% of all vaccinated children. The incidence of pneumonia in Shanghai's Songjiang district decreased rapidly from 2017, when PCV13 vaccination presented an overall increasing trend. The estimated vaccine effectiveness against visits for all-cause pneumonia was 19% (95% CI: 3 to 32) after the first dose in children vaccinated with at least one dose of PCV13. The protective effectiveness of PCV13 was found to be higher for hospitalized pneumonia (30%, 95% CI: 5% to 49%) than for outpatient pneumonia (19%, 95% CI: 4% to 32%).

CONCLUSIONS

PCV13 vaccination among children aged 0-5 years substantially reduced the incidence of all-cause pneumonia. Direct immunization of children under 5 years is an effective strategy to combat outpatient pneumonia, and hospitalized pneumonia.

A Longitudinal Study of Sleep Habits and Leukemia Incidence Among Postmenopausal Women

We sought to assess the relationship between sleep duration, sleep disturbance, and leukemia incidence among postmenopausal women. This study included 130,343 postmenopausal women aged 50-79 years who were enrolled in the Women's Health Initiative (WHI) during 1993-1998. Information on self-reported typical sleep duration and sleep disturbance was obtained by questionnaire at baseline, and sleep disturbance level was defined according to the Women's Health Initiative Insomnia Rating Scale (WHIIRS). WHIIRS scores of 0-4, 5-8, and 9-20 comprised 37.0%, 32.6%, and 30.4% of all women, respectively. After an average of 16.4 years (2,135,109 cumulative person-years) of follow-up, 930 of the participants were identified as having incident leukemia. Compared with women with the lowest level of sleep disturbance (WHIIRS score 0-4), women with higher sleep disturbance levels (WHIIRS scores of 5-8 and 9-20) had 22% (95% confidence interval (CI): 1.04, 1.43) and 18% (95% CI: 1.00, 1.40) excess risks of leukemia, respectively, after multivariable adjustment. A significant dose-response trend was found for the association between sleep disturbance and leukemia risk (P for trend = 0.048). In addition, women with the highest level of sleep disturbance had a higher risk of myeloid leukemia (for WHIIRS score 9-20 vs. WHIIRS score 0-4, hazard ratio = 1.39, CI: 1.05, 1.83). Higher sleep disturbance level was associated with increased risk of leukemia, especially for myeloid leukemia among postmenopausal women.

Survival Analysis, Kaplan-Meier Curves, and Cox Regression: Basic Concepts

Survival analysis is used to analyze data from patients who are followed for different periods of time and in whom the outcome of interest, a dichotomous event, may or may not have occurred at the time the study is halted; data from all patients are used in the analysis, including data from patients who dropped out, regardless of the duration of follow-up. This article discusses basic concepts in survival analysis, explains technical terms such as censoring, and provides reasons why ordinary methods of analysis cannot be applied to such data. The Kaplan-Meier survival curve is described, as is the Cox proportional hazards regression and the hazard ratio. Supplementary information includes a data file, graphs with explanations, and additional discussions; these are provided to enhance the reader's experience and understanding.

Utilizing shared frailty with the Cox proportional hazards regression: Post discharge survival analysis of CHF patients

Understanding patients' survival probability as well as the factors affecting it constitute a significant concern for researchers and practitioners, in particular for patients with severe chronic illnesses such as congestive heart failure (CHF). CHF is a clinical syndrome characterized by comorbidities and adverse medical events. Risk stratification to identify patients most likely to die shortly after hospital discharge can improve the quality of care by better allocating organizational resources and personalized interventions. Probability assessment improves clinical decision-making, contributes to personalized care, and saves costs. Although one of the most informative indices is the time to an adverse event for each patient, commonly analyzed using survival analysis methods, these are often challenging to implement due to the complexity of the medical data. Numerous studies have used the Cox proportional hazards (PH) regression method to generate the survival distribution pattern and factors affecting survival. This model, although advantageous for survival analysis, assumes the homogeneity of the hazard ratio across patients and independence of the observations in terms of survival time. These assumptions are often violated in real-world data, especially when the dataset is composed of readmission data for chronically ill patients, since these recurring observations are inherently dependent. This study ran the Cox PH regression on a feature set selected by machine learning algorithms from a rich hospital dataset. The event modeled here was patient mortality within 90 days post-hospital discharge. The sample was composed of medical records of patients hospitalized in the Israeli Sheba Medical Center more than once, with CHF as the primary diagnosis. We modeled the survival of CHF patients using the Cox PH regression with and without the shared frailty correction that addresses the shortcomings of the Cox Model. The results of the two models of the Cox PH regression - with and without the shared frailty correction were compared. The results demonstrate that the shared frailty correction, which was statistically significant in our analysis, improved the performance of the basic Cox PH model. While this is the main contribution, we also show that this model outperforms two commonly used measures (ADHERE and EFFECT) for predicting early mortality of CHF patients. Thus, the results illustrate how applying advanced analytics can outperform traditional methods. An additional contribution is the feature set selected using machine-learning methods that is different from those used in the extant literature.

Lower Expression of CFTR Is Associated with Higher Mortality in a Meta-Analysis of Individuals with Colorectal Cancer

Individuals with cystic fibrosis (CF), caused by biallelic germline mutations in the cystic fibrosis transmembrane conductance regulator (CFTR), have higher risk and earlier onset of colorectal cancer (CRC). A subset of CRC patients in the non-CF population expresses low levels of tumoral CFTR mRNA which may also cause decreased CFTR activity. To determine the consequences of reduced CFTR expression in this population, we investigated association of tumoral CFTR expression with overall and disease-specific mortality in CRC patients. CFTR mRNA expression, clinical factors and survival data from 1177 CRC patients reported in the Cancer Genome Atlas (TCGA) and Gene Expression Omnibus studies GSE39582 and GSE17538 were included. Log-transformed and z-normalized [mean = 0, standard deviation (SD) = 1] CFTR expression values were modeled as quartiles or dichotomized at the median. Univariate and multivariable Cox proportional hazards regression models were used to estimate hazard ratios (HR) and 95% confidence intervals (CI) for overall and disease-specific mortality in individual studies and meta-analyses. Analyses of each of the three individual datasets showed a robust association of decreased CFTR expression with increased mortality. In meta-analyses adjusted for stage at diagnosis, age and sex, CFTR expression was inversely associated with risk of overall death [pooled HR (95% CI): 0.70 (0.57-0.86)] and disease-specific death [pooled HR (95% CI): 0.68 (0.47-0.99)]. Associations did not differ by stage at diagnosis, age, or sex. Meta-analysis of overall death stratified by microsatellite instable (MSI) versus microsatellite stable (MSS) status indicated potential interaction between MSI/MSS status and CFTR expression, (p-interaction: 0.06). The findings from these three datasets support the hypothesis that low CFTR expression is associated with increased CRC mortality.

Exploration of Influencing Factors for Postoperative Recurrence in Patients with Madelung's Disease on the Basis of Multivariate Stepwise Cox Regression Analysis

Purpose

Madelung's disease (MD) is a rare condition of massive deposits of fat accumulations between superficial and deep fascia at typical locations. There is an absence of systematic studies related to MD in the Chinese cohort. Thus, the objective of the study was to investigate the clinical features of the MD cases in our institution and to explore the clinical variables associated with postoperative recurrence.

Materials and Methods

We retrospectively analyzed the clinical information of 21 individuals with MD from 2013 to 2021 enrolled in our institution. The paired t-test and χ ² test were, respectively, used to determine the difference between continuous and classified variables. The univariate Kaplan-Meier analysis by log-rank and multivariate stepwise Cox regression analysis were used to explore variables possibly associated with postoperative recurrence in MD individuals.

Results

In the current study, 90.48% of the studied patients were male with a mean age of 48.76 years old. About 61.90% exhibited type I MD. MD patients who experienced postoperative recurrence had a higher age, BMI, incidence of chronic complications, and prevalence of alcoholism than the other MD patients without recurrence (P < 0.05). The univariate Kaplan-Meier analysis by log-rank identified that age, BMI, alcoholism, and comorbidities were influencing factors related with postoperative recurrence (P < 0.05).

Conclusion

Demographic characteristics of the 21 studied Chinese cases with MD were generally in accordance with previously published data of other foreign populations. The factors possibly influencing the postoperative recurrence for patients with MD were age, BMI, alcoholism, and a combination of comorbidities. This is the first time that a summarization of clinical characteristics and postoperative recurrence variables of Chinese patients with MD has been reported.

Multi-threshold proportional hazards model and subgroup identification

We propose a novel two-stage procedure for change point detection and parameter estimation in a multi-threshold proportional hazards model. In the first stage, we estimate the number of thresholds by formulating the threshold detection problem as a variable selection problem and applying the penalized partial likelihood approach. In the second stage, the change point locations are refined by a grid search and the standard inference for segment regression can then follow. The proposed model and estimation procedure could lend support to subgroup identification and personalized treatment recommendation in medical research. We establish the consistency of the threshold estimators and regression coefficient estimators under technical conditions. The finite sample performance of the method is demonstrated via simulation studies and two cancer data examples.

Analysis of composite endpoints with component-wise censoring in the presence of differential visit schedules

Composite endpoints are very common in clinical research, such as recurrence-free survival in oncology research, defined as the earliest of either death or disease recurrence. Because of the way data are collected in such studies, component-wise censoring is common, where, for example, recurrence is an interval-censored event and death is a right-censored event. However, a common way to analyze such component-wise censored composite endpoints is to treat them as right-censored, with the date at which the non-fatal event was detected serving as the date the event occurred. This approach is known to introduce upward bias when the Kaplan-Meier estimator is applied, but has more complex impact on semi-parametric regression approaches. In this article we compare the performance of the Cox model estimators for right-censored data and the Cox model estimators for interval-censored data in the context of component-wise censored data where the visit process differs across levels of a covariate of interest, a common scenario in observational data. We additionally examine estimators of the cause-specific hazard when applied to the individual components of such component-wise censored composite endpoints. We found that when visit schedules differed according to levels of a covariate of interest, the Cox model estimators for right-censored data and the estimators for cause-specific hazards were increasingly biased as the frequency of visits decreased. The Cox model estimator for interval-censored data with censoring at the last disease-free date is recommended for use in the presence of differential visit schedules.

Relationship between serum sodium levels and all-cause mortality in congestive heart failure patients: A retrospective cohort study based on the Mimic-III database

BACKGROUND

The relationship between serum sodium levels and mortality in congestive heart failure (CHF) patients has not been well-studied previously. The non-linear correlation between serum sodium levels and mortality in patients with heart failure is currently controversial, and the relationship between different serum sodium levels and mortality is disputed. The goal of this study is to look into the relationship between serum sodium levels and all-cause mortality in people with CHF after controlling for other factors.

METHODS

The publicly accessible Mimic III database was the source of data for our study. We use the ICU Admission Scoring System to collect demographic data, laboratory findings, comorbidities, vital signs, and scoring information for each patient. Cox proportional risk analysis, smooth curve fitting, and the Kaplan-Meier survival curve were used to assess the relationship between baseline sodium levels and all-cause mortality in CHF patients.

RESULTS

The segmentation regression model discovered a turning point value of serum sodium levels (137.5 mmol/L) between serum sodium levels and all-cause mortality. According to the results of the fully adjusted Cox proportional hazard model, lower serum sodium levels (<137.5 mmol/L) were associated with an increased risk of 30, 90, 365-day, and 4-year all-cause deaths. The HRs and 95th confidence intervals were 0.96 (0.94, 0.99), 0.96 (0.94, 0.99), 0.96 (0.94, 0.98), and 0.96 (0.95, 0.98), respectively; the higher serum sodium levels (≥137.5 mmol/L) were related to an associated multiplied risk of 30, 90, 365-day, and 4-year all-cause deaths; the HRs and 95th confidence intervals were 1.02 (1.00, 1.05), 1.02 (1.00, 1.04), 1.02 (1.00, 1.03), and 1.02 (1.00, 1.03), respectively.

CONCLUSION

Serum sodium levels were u-shaped about all-cause mortality. In individuals with CHF, serum sodium levels are linked to an elevated risk of short-, medium-, and long-term all-cause mortality.

Association between nonalcoholic fatty liver disease and incident diabetes mellitus among Japanese: a retrospective cohort study using propensity score matching

BACKGROUND

Previous studies have demonstrated that nonalcoholic fatty liver disease (NAFLD) is a significant risk factor for diabetes mellitus (DM). However, these studies did not completely determine the relationship between NAFLD and DM due to unbalanced confounding factors. The propensity score (PS) is the conditional probability of having a particular exposure, given a set of baseline measured covariates. Propensity score matching (PSM) analysis could minimise the effects of potential confounders. Thus, this study aimed to use PSM analysis to explore the association between NAFLD and DM in a large Japanese cohort.

METHODS

This retrospective PSM cohort study was performed on 14,280 Japanese participants without DM at baseline in Murakami Memorial Hospital between 2004 and 2015. The independent variable was NAFLD at baseline, and the outcome was the incidence of DM during follow-up. One-to-one PSM revealed 1671 participants with and without NAFLD. A doubly robust estimation method was applied to verify the correlation between NAFLD and DM.

RESULTS

The risk of developing DM in participants with NAFLD increased by 98% according to the PSM analysis (HR = 1.98, 95% confidence interval [CI]: 1.41-2.80, P < 0.0001). The risk of developing DM in the NAFLD participants was 2.33 times that of the non-NAFLD participants in the PSM cohort after adjusting for the demographic and laboratory biochemical variables (HR = 2.33, 95% CI: 1.63-3.32, P < 0.0001). The participants with NAFLD had a 95% increased risk of DM after adjusting for PS (HR = 1.95, 95% CI: 1.39-2.75, P = 0.0001). All potential confounding variables were not significantly associated with NAFLD and DM after PSM in the subgroup analysis. In the sensitivity analysis, the participants with NAFLD had a 2.17-fold higher risk of developing DM in the original cohort (HR = 2.17, 95% CI: 1.63-2.88, P < 0.0001) and were 2.27-fold more likely to develop DM in the weighted cohort (HR = 2.27, 95% CI: 1.91-2.69, P < 0.00001).

CONCLUSIONS

NAFLD was an independent risk factor for the development of DM. The risk of developing DM in the NAFLD participants was 2.33 times that of the non-NAFLD participants in the PSM cohort after adjusting for the demographic and laboratory biochemical variables. The participants with NAFLD had a 95% increased risk of DM after adjusting for PS.

Making apples from oranges: Comparing noncollapsible effect estimators and their standard errors after adjustment for different covariate sets

We revisit the well-known but often misunderstood issue of (non)collapsibility of effect measures in regression models for binary and time-to-event outcomes. We describe an existing simple but largely ignored procedure for marginalizing estimates of conditional odds ratios and propose a similar procedure for marginalizing estimates of conditional hazard ratios (allowing for right censoring), demonstrating its performance in simulation studies and in a reanalysis of data from a small randomized trial in primary biliary cirrhosis patients. In addition, we aim to provide an educational summary of issues surrounding (non)collapsibility from a causal inference perspective and to promote the idea that the words conditional and adjusted (likewise marginal and unadjusted) should not be used interchangeably.

Survival analysis-part 2: Cox proportional hazards model

Learning objectives: 1. To understand the log-rank test and limitations of the log-rank test in comparing survival between groups. 2. To understand the fundamental concepts of the proportional hazards assumption. 3. To understand basic steps in the development of the Cox proportional hazards model and reported hazard ratios. 4. To understand how results of a Cox model run using STATA© (a commonly used proprietary statistical software) can be understood and interpreted.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12055-020-01108-7.

Analysis of genomic and transcriptomic variations as prognostic signature for lung adenocarcinoma

Background

Lung cancer is the leading cause of the largest number of deaths worldwide and lung adenocarcinoma is the most common form of lung cancer. In order to understand the molecular basis of lung adenocarcinoma, integrative analysis have been performed by using genomics, transcriptomics, epigenomics, proteomics and clinical data. Besides, molecular prognostic signatures have been generated for lung adenocarcinoma by using gene expression levels in tumor samples. However, we need signatures including different types of molecular data, even cohort or patient-based biomarkers which are the candidates of molecular targeting.

Results

We built an R pipeline to carry out an integrated meta-analysis of the genomic alterations including single-nucleotide variations and the copy number variations, transcriptomics variations through RNA-seq and clinical data of patients with lung adenocarcinoma in The Cancer Genome Atlas project. We integrated significant genes including single-nucleotide variations or the copy number variations, differentially expressed genes and those in active subnetworks to construct a prognosis signature. Cox proportional hazards model with Lasso penalty and LOOCV was used to identify best gene signature among different gene categories.

We determined a 12-gene signature (BCHE, CCNA1, CYP24A1, DEPTOR, MASP2, MGLL, MYO1A, PODXL2, RAPGEF3, SGK2, TNNI2, ZBTB16) for prognostic risk prediction based on overall survival time of the patients with lung adenocarcinoma. The patients in both training and test data were clustered into high-risk and low-risk groups by using risk scores of the patients calculated based on selected gene signature. The overall survival probability of these risk groups was highly significantly different for both training and test datasets.

Conclusions

This 12-gene signature could predict the prognostic risk of the patients with lung adenocarcinoma in TCGA and they are potential predictors for the survival-based risk clustering of the patients with lung adenocarcinoma. These genes can be used to cluster patients based on molecular nature and the best candidates of drugs for the patient clusters can be proposed. These genes also have a high potential for targeted cancer therapy of patients with lung adenocarcinoma.

Associations between body mass index and mortality in acute-on-chronic liver failure patients

INTRODUCTION AND OBJECTIVES

The association between the level of body mass index (BMI) and the mortality of patients with critical liver disease remains unclear. This study aimed to examine the association between BMI and hospital mortality of patients with acute-on-chronic liver failure (ACLF).

METHODS

Clinical data from 146 ACLF patients were collected and analyzed. BMI was categorized into three groups: lower BMI (<18.5kg/m²), normal BMI (18.5-24.9kg/m²), and overweight (25.0-32.0kg/m²). BMI and laboratory parameters were measured one day before, or on the day of the start of the treatment. Values of BMI and laboratory parameters were compared between survivors and non-survivors, and then hospital mortality rates were compared among patients with different BMI levels.

RESULTS

The prognosis of ACLF patients was significantly correlated with international normalized ratio (INR), albumin and BMI. The ACLF patients with low albumin level and high INR values tend to have a high mortality rate. Also, survival time was significantly shorter in the ACLF patients with lower BMI, while patients with normal and overweight values had longer survival time.

CONCLUSIONS

A graded association between BMI and hospital mortality with a strong significant trend was found in ACLF patients in China.

Comparison of Cox proportional hazards regression and generalized Cox regression models applied in dementia risk prediction

INTRODUCTION

The frequently used Cox regression applies two critical assumptions, which might not hold for all predictors. In this study, the results from a Cox regression model (CM) and a generalized Cox regression model (GCM) are compared.

METHODS

Data are from the Survey of Health, Ageing and Retirement in Europe (SHARE), which includes approximately 140,000 individuals aged 50 or older followed over seven waves. CMs and GCMs are used to estimate dementia risk. The results are internally and externally validated.

RESULTS

None of the predictors included in the analyses fulfilled the assumptions of Cox regression. Both models predict dementia moderately well (10-year risk: 0.737; 95% confidence interval [CI]: 0.699, 0.773; CM and 0.746; 95% CI: 0.710, 0.785; GCM).

DISCUSSION

The GCM performs significantly better than the CM when comparing pseudo-R2 and the log-likelihood. GCMs enable researcher to test the assumptions used by Cox regression independently and relax these assumptions if necessary.

Smoking, air pollution, and lung cancer risk in the Nurses' Health Study cohort: time-dependent confounding and effect modification

The proportional hazards (PH) model is commonly used in epidemiology despite the stringent assumption of proportionality of hazards over time. We previously showed, using detailed simulation data, that the impact of a modest risk factor cannot be estimated reliably using the PH model in the presence of confounding by a strong, time-dependent risk factor. Here, we examine the same and related issues using a real dataset. Among 97,303 women in the prospective Nurses' Health Study cohort from 1994 through 2010, we used PH regression to investigate how effect estimates for cigarette smoking are affected by increasingly detailed specification of time-dependent exposure characteristics. We also examined how effect estimates for fine particulate matter (PM2.5), a modest risk factor, are affected by finer control for time-dependent confounding by smoking. The objective of this analysis is not to present a credible estimate of the impact of PM2.5 on lung cancer risk, but to show that estimates based on the PH model are inherently unreliable. The best-fitting model for cigarette smoking and lung cancer included pack-years, duration, time since cessation, and an age-by-pack-years interaction, indicating that the hazard ratio (HR) for pack-years was significantly modified by age. In the fully adjusted best-fitting model for smoking including pack-years, the HR per 10-µg/m3 increase in PM2.5 was 1.06 (95% confidence interval (CI) = 0.90, 1.25); the HR for PM2.5 in the full cohort ranged between 1.02 and 1.10 in models with other smoking adjustments, indicating a residual confounding effect of smoking. The HR for PM2.5 was statistically significant only among former smokers when adjusting for smoking pack-years (HR = 1.35, 95% CI = 1.00, 1.82 in the best-fitting smoking model), but not in models adjusting for smoking duration and average packs (pack-years divided by duration). The association between cumulative smoking and lung cancer is modified by age, and improved model fit is obtained by including multiple time-varying components of smoking history. The association with PM2.5 is residually confounded by smoking and modified by smoking status. These findings underscore limitations of the PH model and emphasize the advantages of directly estimating hazard functions to characterize time-varying exposure and risk. The hazard function, not the relative hazard, is the fundamental measure of risk in a population. As a consequence, the use of time-dependent PH models does not address crucial issues introduced by temporal factors in epidemiological data.

Prognostic value of differentially methylated gene profiles in bladder cancer

DNA methylation can regulate gene expression and is pivotal in the occurrence and development of bladder cancer. In this study, we analyzed whole-genome DNA methylation on the basis of data from The Cancer Genome Atlas to select epigenetic biomarkers predictive of survival and further understand the molecular mechanisms underlying methylation patterns in bladder cancer. We identified 540 differentially methylated genes between tumor and normal tissues, including a number of independent prognostic factors based on univariate analysis. Genes (MIR6732, SOWAHC, SERPINI1, OR10W1, OR7G3, AIM1, and ZFAND5) were integrated to establish a risk model for prognostic assessment based on multivariate Cox analysis. The methylation of SOWAHC was negatively correlated with its messenger RNA expression, and together these were significantly correlated with prognosis. This study took advantage of high-throughput data mining to provide new bioinformatics evidence and ideas for further study into the pathogenesis and prognosis of bladder cancer.

Exploration of methylation-driven genes for monitoring and prognosis of patients with lung adenocarcinoma

Background

As one of the most common malignant tumors in humans, lung cancer has experienced a gradual increase in morbidity and mortality. This study examined prognosis-related methylation-driven genes specific to lung adenocarcinoma (LUAD) to provide a basis for prognosis prediction and personalized targeted therapy for LUAD patients.

Methods

The methylation and survival time data from LUAD patients in the TCGA database were downloaded. The MethylMix algorithm was used to identify the differential methylation status of LUAD and adjacent tissues based on the β-mixture model to obtain disease-related methylation-driven genes. A COX regression model was then used to screen for LUAD prognosis-related methylation-driven genes, and a linear risk model based on five methylation-driven gene expression profiles was constructed. A methylation and gene expression combined survival analysis was performed to further explore the prognostic value of 5 genes independently.

Results

There were 118 differentially expressed methylation-driven genes in the LUAD tissues and adjacent tissues. Five of the genes, CCDC181, PLAU, S1PR1, ELF3, and KLHDC9, were used to construct a prognostic risk model. Overall, the survival time was significantly lower in the high-risk group compared with that in the low-risk group (P < 0.05). In addition, the methylation and gene expression combined survival analysis found that the combined expression levels of the genes CCDC181, PLAU, and S1PR1 as well as KLHDC9 alone can be used as independent prognostic markers or drug targets.

Conclusion

Our findings provide an important bioinformatic basis and relevant theoretical basis for guiding subsequent LUAD early diagnosis and prognosis assessments.

Electronic supplementary material

The online version of this article (10.1186/s12935-018-0691-z) contains supplementary material, which is available to authorized users.

Development and Validation of a Prognostic Tool for Identifying Residents at Increased Risk of Death in Long-Term Care Facilities

BACKGROUND

To promote better care at the end stage of life in long-term care facilities, a culturally appropriate tool for identifying residents at the end of life is crucial.

OBJECTIVE

This study aimed to develop and validate a prognostic tool, the increased risk of death (IRD) scale, based on the minimum data set (MDS).

DESIGN

A retrospective study using data between 2005 and 2013 from six nursing homes in Hong Kong.

SETTING/SUBJECTS

A total of 2380 individuals were randomly divided into two equal-sized subsamples: Sample 1 was used for the development of the IRD scale and Sample 2 for validation.

MEASUREMENTS

The measures were MDS 2.0 items and mortality data from the discharge tracking forms. The nine items in the IRD scale (decline in cognitive status, decline in activities of daily living, cancer, renal failure, congestive heart failure, emphysema/chronic obstructive pulmonary disease, edema, shortness of breath, and loss of weight), were selected based on bivariate Cox proportional hazards regression.

RESULTS

The IRD scale was a strong predictor of mortality in both Sample 1 (HR_sample1 = 1.50, 95% confidence interval [CI]: 1.37-1.65) and Sample 2 (HR_sample2 = 1.31, 1.19-1.43), after adjusting for covariates. Hazard ratios (HRs) for residents who had an IRD score of 3 or above for Sample 1 and Sample 2 were 3.32 (2.12-5.21) and 2.00 (1.30-3.09), respectively.

CONCLUSIONS

The IRD scale is a promising tool for identifying nursing home residents at increased risk of death. We recommend the tool to be incorporated into the care protocol of long-term care facilities in Hong Kong.

Physical frailty and mortality risk in Japanese older adults

AIM

The association between frailty and increased mortality risk is unknown. The present study assessed the longitudinal relationship between frailty and mortality risk in Japanese community-dwelling older adults.

METHODS

Participants included 841 randomly chosen community-dwelling Japanese individuals, including 175 older adults aged 65-88 years with incomplete data at the baseline examination (July 2006-July 2008). Participants were followed from baseline to 31 December 2015 (mean 7.9 years). Frailty was diagnosed according to frailty criteria, including unintentional weight loss (shrinking), exhaustion, low activity, low grip strength and low gait speed. Information on deaths was obtained from a population dynamics survey. The relationship between frailty and mortality was assessed using Kaplan-Meier survival curves and Cox proportional hazards regression. The Cox proportional hazards model was used to control for potential confounders, including age at baseline, body fat, education, the Mini-Mental State Examination score, the Center for Epidemiologic Studies Depression Scale score, total physical activity, total caloric intake, alcohol intake, current smoking, household income and the number of current diseases.

RESULTS

The fully adjusted hazard ratio for all-cause mortality in the frailty group was 2.63 (95% confidence interval, 1.28-5.39; P for trend <0.01). The age- and sex-adjusted hazard ratio for mortality of cancer in the frailty group was 3.33 (95% confidence interval, 1.15-9.62; P for trend <0.05).

CONCLUSION

Complications of frailty, which include shrinking, exhaustion, low activity, weakness, and slowness, appear to be significant risks for mortality in Japanese older adults. Geriatr Gerontol Int 2018; 18: 1085-1092.

Estimation of causal effect measures with the R-package stdReg

Measures of causal effects play a central role in epidemiology. A wide range of measures exist, which are designed to give relevant answers to substantive epidemiological research questions. However, due to mathematical convenience and software limitations most studies only report odds ratios for binary outcomes and hazard ratios for time-to-event outcomes. In this paper we show how logistic regression models and Cox proportional hazards regression models can be used to estimate a wide range of causal effect measures, with the R-package stdReg. For illustration we focus on the attributable fraction, the number needed to treat and the relative excess risk due to interaction. We use two publicly available data sets, so that the reader can easily replicate and elaborate on the analyses. The first dataset includes information on 487 births among 188 women, and the second dataset includes information on 2982 women diagnosed with primary breast cancer.

Risk Factors for Culling, Sales and Deaths in New Zealand Dairy Goat Herds, 2000-2009

The aim of this study was to identify risk factors for culling, sales and deaths in intensively managed dairy goat herds in New Zealand. A data set provided by the New Zealand Dairy Goat Cooperative (n = 13,197 does) was analyzed using a Cox proportional hazard model. The outcome of interest was length of productive life (LPL), defined as the number of days from the date of second kidding to the date of removal from the herd or the date on which follow-up was terminated, whichever occurred first. Milk solids yield in the first lactation (MSL1) as a predictor of LPL was parameterized in the model as a penalized spline term. To account for MSL1 violating the proportional hazards assumption of the Cox model, LPL was divided into two intervals: T1 (less than or equal to 730 days from the date of second kidding) and T2 (greater than 730 days from the date of second kidding). MSL1 was then included in the model as a time-dependent covariate. A frailty term was included in the model to account for unmeasured, herd-level effects on LPL. During T1, the daily hazard of removal for does that produced 80 kg milk solids in the first lactation was 0.84 (95% CI 0.58–1.23) times the daily hazard of removal for does that produced 30 kg milk solids in the first lactation. During T2, the daily hazard of removal for does that produced 80 kg milk solids in the first lactation was 1.44 (95% CI 0.79–2.65) times the daily hazard of removal for does that produced 30 kg milk solids in the first lactation. We conclude that involuntary losses may be avoided if high MSL1 yielding does are preferentially managed from 2 years beyond the date of second kidding.

Associations of objectively measured moderate-to-vigorous-intensity physical activity and sedentary time with all-cause mortality in a population of adults at high risk of type 2 diabetes mellitus

The relationships of physical activity and sedentary time with all-cause mortality in those at high risk of type 2 diabetes mellitus (T2DM) are unexplored. To address this gap in knowledge, we examined the associations of objectively measured moderate-to-vigorous-intensity physical activity (MVPA) and sedentary time with all-cause mortality in a population of adults at high risk of T2DM. In 2010–2011, 712 adults (Leicestershire, U.K.), identified as being at high risk of T2DM, consented to be followed up for mortality. MVPA and sedentary time were assessed by accelerometer; those with valid data (≥ 10 hours of wear-time/day with ≥ 4 days of data) were included. Cox proportional hazards regression models, adjusted for potential confounders, were used to investigate the independent associations of MVPA and sedentary time with all-cause mortality. 683 participants (250 females (36.6%)) were included and during a mean follow-up period of 5.7 years, 26 deaths were registered. Every 10% increase in MVPA time/day was associated with a 5% lower risk of all-cause mortality [Hazard Ratio (HR): 0.95 (95% Confidence Interval (95% CI): 0.91, 0.98); p = 0.004]; indicating that for the average adult in this cohort undertaking approximately 27.5 minutes of MVPA/day, this benefit would be associated with only 2.75 additional minutes of MVPA/day. Conversely, sedentary time showed no association with all-cause mortality [HR (every 10-minute increase in sedentary time/day): 0.99 (95% CI: 0.95, 1.03); p = 0.589]. These data support the importance of MVPA in adults at high risk of T2DM. The association between sedentary time and mortality in this population needs further investigation.

•

Objectively measured MVPA time was strongly associated with all-cause mortality.

•

Objectively measured sedentary time was not associated with all-cause mortality.

•

These data support the importance of MVPA in adults at high risk of T2DM.

Random Survival Forest in practice: a method for modelling complex metabolomics data in time to event analysis

BACKGROUND

The application of metabolomics in prospective cohort studies is statistically challenging. Given the importance of appropriate statistical methods for selection of disease-associated metabolites in highly correlated complex data, we combined random survival forest (RSF) with an automated backward elimination procedure that addresses such issues.

METHODS

Our RSF approach was illustrated with data from the European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam study, with concentrations of 127 serum metabolites as exposure variables and time to development of type 2 diabetes mellitus (T2D) as outcome variable. Out of this data set, Cox regression with a stepwise selection method was recently published. Replication of methodical comparison (RSF and Cox regression) was conducted in two independent cohorts. Finally, the R-code for implementing the metabolite selection procedure into the RSF-syntax is provided.

RESULTS

The application of the RSF approach in EPIC-Potsdam resulted in the identification of 16 incident T2D-associated metabolites which slightly improved prediction of T2D when used in addition to traditional T2D risk factors and also when used together with classical biomarkers. The identified metabolites partly agreed with previous findings using Cox regression, though RSF selected a higher number of highly correlated metabolites.

CONCLUSIONS

The RSF method appeared to be a promising approach for identification of disease-associated variables in complex data with time to event as outcome. The demonstrated RSF approach provides comparable findings as the generally used Cox regression, but also addresses the problem of multicollinearity and is suitable for high-dimensional data.

ASSOCIATION OF TOTAL AND DIFFERENTIAL WHITE BLOOD CELL COUNTS TO DEVELOPMENT OF TYPE 2 DIABETES IN MEXICAN AMERICANS IN CAMERON COUNTY HISPANIC COHORT

OBJECTIVE

To evaluate the relationship between total and differential White Blood Cell (WBC) counts with time to transition to type 2 diabetes in Mexican Americans using prospective data from the Cameron County Hispanic Cohort (CCHC).

RESULTS

Multivariable Cox proportional hazards regression models revealed that obese Mexican-American cohort participants whose total WBC or granulocyte count increased over time had 1.39 and 1.35 times higher risk respectively of transition to type 2 diabetes when compared to overweight participants. The granulocyte or total WBC count in participants with BMI≥35 were significant risk factors for transition to type 2 diabetes.

CONCLUSIONS

Increased total WBC and WBC differential counts, particularly lymphocytes and granulocytes, are associated with risk of transition to type 2 diabetes in obese Mexican Americans, after adjusting for other potential confounders. Screening and monitoring the WBC counts, including lymphocytes and granulocytes can help with monitoring potential transition to type 2 diabetes.

Patient-specific meta-analysis for risk assessment using multivariate proportional hazards regression

We propose a method for assessing an individual patient's risk of a future clinical event using clinical trial or cohort data and Cox proportional hazards regression, combining the information from several studies using meta-analysis techniques. The method combines patient-specific estimates of the log cumulative hazard across studies, weighting by the relative precision of the estimates, using either fixed- or random-effects meta-analysis calculations. Risk assessment can be done for any future patient using a few key summary statistics determined once and for all from each study. Generalizations of the method to logistic regression and linear models are immediate. We evaluate the methods using simulation studies and illustrate their application using real data.

Extending the Peters-Belson approach for assessing disparities to right censored time-to-event outcomes

The Peters-Belson (PB) method was developed for quantifying and testing disparities between groups in an outcome by using linear regression to compute group-specific observed and expected outcomes. It has since been extended to generalized linear models for binary and other outcomes and to analyses with probability-based sample weighting. In this work, we extend the PB approach to right-censored survival analysis, including stratification if needed. The extension uses the theory and methods of expected survival on the basis of Cox regression in a reference population. Within the PB framework, among the groups to be compared, one group is chosen as the reference group, and outcomes in that group are modeled as a function of available predictors. By using this fitted model's estimated parameters, and the predictor values for a comparator group, the comparator group's expected outcomes are then calculated and compared, formally with testing and informally with graphics, with their observed outcomes. We derive the extension, show how we applied it in a study of incontinence in nursing home elderly, and discuss issues in implementing it. We used the 'survival' package in the R system to do computations.

Bias Correction Methods for Misclassified Covariates in the Cox Model: comparison offive correction methods by simulation and data analysis

Measurement error/misclassification is commonplace in research when variable(s) can notbe measured accurately. A number of statistical methods have been developed to tackle this problemin a variety of settings and contexts. However, relatively few methods are available to handlemisclassified categorical exposure variable(s) in the Cox proportional hazards regression model. Inthis paper, we aim to review and compare different methods to handle this problem - naïvemethods, regression calibration, pooled estimation, multiple imputation, corrected score estimation,and MC-SIMEX - by simulation. These methods are also applied to a life course study with recalleddata and historical records. In practice, the issue of measurement error/misclassification should beaccounted for in design and analysis, whenever possible. Also, in the analysis, it could be moreideal to implement more than one correction method for estimation and inference, with properunderstanding of underlying assumptions.