scholarly journals Two Postestimation Commands for Assessing Confounding Effects in Epidemiological Studies

2007 ◽  
Vol 7 (2) ◽  
pp. 183-196 ◽  
Author(s):  
Zhiqiang Wang
Keyword(s):  
Logistic Regression ◽  
Bayesian Information Criterion ◽  
Proportional Hazards ◽  
Conditional Logistic Regression ◽  
Information Criterion ◽  
Epidemiological Studies ◽  
Cox Proportional Hazards ◽  
P Value ◽  
Effect Estimate ◽  
Cox Proportional Hazards Models

Confounding is a major issue in observational epidemiological studies. This paper describes two postestimation commands for assessing confounding effects. One command (confall) displays and plots all possible effect estimates against one of p-value, Akaike information criterion, or Bayesian information criterion. This computing-intensive procedure allows researchers to inspect the variability of the effect estimates from various possible models. Another command (chest) uses a stepwise approach to identify variables that have substantially changed the effect estimate. Both commands can be used after most common estimation commands in epidemiological studies, such as logistic regression, conditional logistic regression, Poisson regression, linear regression, and Cox proportional hazards models.

The Longitudinal Association of Vision Impairment with Transitions to Cognitive Impairment and Dementia: Findings from The Aging, Demographics and Memory Study

2021 ◽  
Author(s):  
Joshua R Ehrlich ◽  
Bonnielin K Swenor ◽  
Yunshu Zhou ◽  
Kenneth M Langa
Keyword(s):  
Visual Acuity ◽  
Logistic Regression ◽  
Cognitive Impairment ◽  
Proportional Hazards ◽  
Population Based ◽  
Cox Proportional Hazards ◽  
Vision Impairment ◽  
Logistic Regression Models ◽  
Cox Proportional Hazards Models ◽  
Longitudinal Association

Abstract Background Vision impairment (VI) is associated with incident cognitive decline and dementia. However, it is not known whether VI is associated only with the transition to cognitive impairment, or whether it is also associated with later transitions to dementia. Methods We used data from the population-based Aging, Demographics and Memory Study (ADAMS) to investigate the association of visual acuity impairment (VI; defined as binocular presenting visual acuity <20/40) with transitions from cognitively normal (CN) to cognitive impairment no dementia (CIND) and from CIND to dementia. Multivariable Cox proportional hazards models and logistic regression were used to model the association of VI with cognitive transitions, adjusted for covariates. Results There were 351 participants included in this study (weighted percentages: 45% male, 64% age 70-79 years) with a mean follow-up time of 4.1 years. In a multivariable model, the hazard of dementia was elevated among those with VI (HR=1.63, 95%CI=1.04-2.58). Participants with VI had a greater hazard of transitioning from CN to CIND (HR=1.86, 95%CI=1.09-3.18). However, among those with CIND and VI a similar percentage transitioned to dementia (48%) and remained CIND (52%); there was no significant association between VI and transitioning from CIND to dementia (HR=0.94, 95%CI=0.56-1.55). Using logistic regression models, the same associations between VI and cognitive transitions were identified. Conclusions Poor vision is associated with the development of CIND. The association of VI and dementia appears to be due to the higher risk of dementia among individuals with CIND. Findings may inform the design of future interventional studies.

The impact of sorafenib on the treatment and survival of advanced hepatocellular carcinoma (HCC): Analysis of the National Cancer Database (NCDB) from 2004 to 2014.

2019 ◽  
Vol 37 (15_suppl) ◽  
pp. e15682-e15682
Author(s):  
Aman Opneja ◽  
Gino Cioffi ◽  
Asrar Alahmadi ◽  
Nirav Patil ◽  
David Lawrence Bajor ◽  
...  
Keyword(s):  
Proportional Hazards ◽  
Single Agent ◽  
Cox Proportional Hazards ◽  
P Value ◽  
Advanced Hepatocellular Carcinoma ◽  
Advanced Hcc ◽  
Cox Proportional Hazards Models ◽  
Before And After ◽  
Time Frames ◽  
The Impact

e15682 Background: HCC is a common cause of mortality in the U.S. among men and women (5thand 7th, respectively) with overall five-year survival of ~18%. Sorafenib was the only FDA approved therapy for advanced HCC from 2007 until 2018. This study analyzes trends in the treatment and survival of advanced HCC before and after sorafenib approval. Methods: Adult patients ( > 18 years) with diagnosis of HCC treated with only chemotherapy from 2004 – 2014 were identified in NCDB database. Comparisons were made between 3 time frames: 2004 – 2007 (pre-sorafenib), 2008 – 2011 (early sorafenib) and 2012 – 2014 (late sorafenib). Patients treated with single or multi-agent chemotherapy were analyzed. Cox proportional hazards models were used for univariate and multivariable analyses. Kaplan-Meier method was used for survival analysis. Results: The NCDB contained 33,136 patients with HCC diagnosed between 2004 – 2014 and treated with chemotherapy alone. Patients were generally men (77.4%), over the age of 50 years (92.4%), with an elevated AFP at diagnosis (64.4%), and had limited co-morbidities (76.0%, Charlson/Deyo score of 0-1). The T-stages were T1 (26.3%), T2 (20.5%), T3 (25.6%), and T4 (16.2%). The number and proportion of patients treated with single agent chemotherapy increased significantly during the study period: 2,733 (45.3%) pre-sorafenib, 9,723 (72.7%) early sorafenib, and 13,502 (86.1%) late sorafenib. The proportion of all HCC patients in the NCDB receiving only chemotherapy increased from 17.2% to 26.4% to 28.3% across the 3 time frames. The survival of patients with advanced HCC treated only with chemotherapy improved significantly in the early and late sorafenib cohorts compared to the pre-sorafenib cohort (10.3 months (95% CI: 9.8-10.6) vs. 12.3 months (12.0-12.7) vs. 15.5 months (15.1-15.9), p-value < 0.001). Age > 70 years, male sex, higher Charlson/Deyo score ( > 1), elevated AFP at diagnosis, and higher T-stage were associated with worse survival (p value < 0.001). Conclusions: The approval of sorafenib has dramatically increased the use of chemotherapy for the treatment of advanced HCC and has resulted in a significant survival advantage.

scholarly journals Late-life depression and mortality: influence of gender and antidepressant use

2008 ◽  
Vol 192 (1) ◽  
pp. 12-18 ◽  
Author(s):  
Joanne Ryan ◽  
Isabelle Carriere ◽  
Karen Ritchie ◽  
Robert Stewart ◽  
Gwladys Toulemonde ◽  
...  
Keyword(s):  
Depressive Symptoms ◽  
Proportional Hazards ◽  
Depression Scale ◽  
Severe Depression ◽  
Epidemiological Studies ◽  
Cox Proportional Hazards ◽  
Community Dwelling ◽  
Late Life Depression ◽  
Cox Proportional Hazards Models ◽  
Antidepressant Use

BackgroundDepression may increase the risk of mortality among certain subgroups of older people, but the part played by antidepressants in this association has not been thoroughly explored.AimsTo identify the characteristics of older populations who are most at risk of dying, as a function of depressive symptoms, gender and antidepressant use.MethodAdjusted Cox proportional hazards models were used to determine the association between depression and/or antidepressant use and 4-year survival of 7363 community-dwelling elderly people. Major depressive disorder was evaluated using a standardised psychiatric examination based on DSM-IV criteria and depressive symptoms were assessed using the Center for Epidemiological Studies Depression scale.ResultsDepressed men using antidepressants had the greatest risk of dying, with increasing depression severity corresponding to a higher hazard risk. Among women, only severe depression in the absence of treatment was significantly associated with mortality.ConclusionsThe association between depression and mortality is gender-dependent and varies according to symptom load and antidepressant use.

scholarly journals Insomnia in the context of short sleep increases suicide risk

SLEEP ◽  
2020 ◽  
Author(s):  
Hedström Anna Karin ◽  
Ola Hössjer ◽  
Rino Bellocco ◽  
Weimin Ye ◽  
Lagerros Ylva Trolle ◽  
...  
Keyword(s):  
Sleep Duration ◽  
Suicide Risk ◽  
Proportional Hazards ◽  
Cox Proportional Hazards ◽  
Total Effect ◽  
P Value ◽  
Short Sleep ◽  
Cox Proportional Hazards Models ◽  
The Relationship ◽  
Insomnia Symptoms

Abstract Study Objectives The relationship between insomnia and suicide risk is not completely understood. We aimed to investigate the influence of insomnia on suicide risk, taking both sleep duration and depression into consideration. Methods The present study is based on a Swedish prospective cohort study of 38,786 participants with a mean follow-up time of 19.2 years. Cox proportional hazards models with attained age as time-scale were used to estimate hazard ratios (HRs) of death by suicide with 95% confidence intervals (CI) for participants categorized by frequency of insomnia symptoms. Causal mediation analysis was performed to assess to what extent the relationship between insomnia and suicide risk is mediated by depression. Results Insomnia was only associated with suicide risk among short sleepers, whereas no significant association was observed among those who slept 7 h/night or more. The total effect of insomnia in the context of short sleep on suicide risk, expressed on the HR scale, was 2.85 (95% CI 1.42–5.74). The direct effect was 2.25 (95% CI 1.12–4.54) and the indirect effect, mediated by depression, was 1.27 (95% CI 1.05–1.53). Of the total effect, 32% was mediated by depression. The association between insomnia and suicide risk became more pronounced with decreasing depressive symptoms (p value for trend &lt;0.05). Conclusions Insomnia in the context of short sleep increases suicide risk, both directly and indirectly by affecting the risk of depression. Abnormalities of sleep duration and insomnia symptoms should be evaluated when assessing suicide risk.

FC 080VARIABILITY IN SERUM PHOSPHATE ASSESSED BY DIRECTIONAL CHANGE IS ASSOCIATED WITH INCREASED MORTALITY

2021 ◽  
Vol 36 (Supplement_1) ◽  
Author(s):  
Xiaoling (Janice) Ye ◽  
Karlien Ter Meulen ◽  
Len A Usvyat ◽  
Frank Van Der Sande ◽  
Constantijn Konings ◽  
...  
Keyword(s):  
Proportional Hazards ◽  
Smoothing Splines ◽  
Cox Proportional Hazards ◽  
Serum Phosphate ◽  
P Value ◽  
Cox Proportional Hazards Models ◽  
Increased Risk ◽  
All Cause Mortality ◽  
Directional Change ◽  
Wide Variability

Abstract Background and Aims Prior studies showed that there is a wide variability between serial pre-dialysis measurements of serum phosphate (P). Serum P vary can be due to changes in nutritional intake, underlying bone disorders, medication use or inflammation. Various variability markers have been investigated to study the association between P variability and its association with outcomes, however, the directional trends have not been studied in depth. We aimed to study directional changes and investigated its association with outcomes. Method All adult incident HD patients treated in Fresenius Medical Care North America (FMCNA) clinics between 01/2010 and 10/2018 were included in this retrospective cohort study. Serum P levels were averaged from month 1 to 6 after the initiation of dialysis (baseline). Baseline absolute and directional range (DR) of serum P were calculated. DR of P was calculated as: P min/max (t2) – P max/min (t1), with P (t1) and P (t2) represents the timepoint when either the min P value or max P value was measured, whichever comes first, and with t2 happened after t1. It is positive when the minimum antedates the maximum, otherwise negative. All-cause mortality was recorded between months 7 and 18. Cox proportional hazards models with spline terms were applied to explore the association between absolute and DR of P and all-cause mortality. Additionally, tensor product smoothing splines were computed to study the interactions of P with absolute P and DR of P and their joint associations with outcomes, respectably. Results We studied 353,142 patients. The average age was 62.7 years, 58% were male, 64% were diabetic. Baseline P was 4.98 mg/dL, median absolute range was 2.40 mg/dL, median DR was 1.1 mg/dL. Across different levels of P, both higher levels of absolute range and DR of P were associated with higher risk of mortality (Figure 1, Figure 2). The associations even seemed stronger in patients with lower levels of serum P and with negative DR (Figure 1). Conclusion Lower levels of serum P are independently associated with an increased risk of all-cause mortality. Whereas both a positive and negative DR of P are in general associated with increased mortality, the effects of an increase are most predominant in patients with higher levels of serum P, whereas a negative directional range are most predominant in patients with low serum P. This could be explained by the fact that patients with lower levels of P are generally malnourished or inflamed, where a further reduction indicates nutritional deterioration.

scholarly journals THE HEALTHY AGING INDEX AND ITS ASSOCIATION WITH MORTALITY IN OLDER MEXICAN AND EUROPEAN AMERICANS

2019 ◽  
Vol 3 (Supplement_1) ◽  
pp. S960-S960 ◽  
Author(s):  
Vinutha Ganapathy ◽  
Mini E Jacob ◽  
Meghan I Short ◽  
Mitzi M Gonzales ◽  
Claudia Satizabal Barrera ◽  
...  
Keyword(s):  
Interaction Effect ◽  
Mexican Americans ◽  
Healthy Aging ◽  
Proportional Hazards ◽  
Cox Proportional Hazards ◽  
P Value ◽  
Multiple Imputations ◽  
Aging Index ◽  
European Americans ◽  
Cox Proportional Hazards Models

Abstract Mexican Americans (MA) have higher morbidity compared to European Americans (EA); however, the mortality rate remains lower (Hispanic Paradox). The Healthy Aging Index (HAI) captures clinical and subclinical morbidity in older adults and is useful in examining ethnic differences in mortality for a given disease burden. We assessed the association between baseline HAI and all-cause mortality over 12 years of follow-up among older MAs (n=394) and EAs (n=355) in the San Antonio Longitudinal Study of Aging (SALSA) and examined differences between ethnic groups. HAI incorporates non-invasive measures (systolic blood pressure, forced vital capacity, creatinine, fasting plasma glucose (FPG), Mini-Mental State Exam). Missing baseline data for HAI components and covariates were imputed using multiple imputations. Proteinuria was used instead of creatinine due to non-availability. Scores of 0, 1, 2 were given from lowest to highest tertile HAI categories; diagnosis of diabetes, hypertension, and renal failure were included in the highest tertiles. Cox proportional hazards models estimated the association between HAI and mortality, adjusting for confounders. After adjusting for age, gender, education, income, BMI, smoking and ethnicity, HAI was independently associated with mortality (HR 1.25 (1.16-1.35), p-value &lt;0.0001). We found no interaction effect between HAI and ethnicity on mortality ((p-value for interaction = 0.78). In the SALSA sample, HAI is a predictor of mortality after adjusting for confounders in both MAs and EAs. The absence of a significant HAI*ethnicity interaction effect further demonstrated that HAI works equally well as a predictor of mortality in both MAs and EAs.

Indigeneity, Prisoner Visitation and Reincarceration in Australia: The Association Between Visits in Prison and Reincarceration for Indigenous and Non-Indigenous People

2020 ◽  
Vol 60 (4) ◽  
pp. 1056-1079
Author(s):  
Nicole Ryan ◽  
Jeff Ackerman ◽  
Justin Ready ◽  
Stuart A Kinner
Keyword(s):  
Logistic Regression ◽  
Criminal Justice ◽  
Indigenous People ◽  
Administrative Data ◽  
Proportional Hazards ◽  
Cox Proportional Hazards ◽  
Travel Distance ◽  
Proportional Hazards Models ◽  
Cox Proportional Hazards Models ◽  
Criminal Justice Systems

Abstract Prison visitation is believed to protect against reincarceration post-release. This research explored differences in prisoner visitation between Indigenous and non-Indigenous people, the predictors of visitation and the effects of visitation on the risk of reincarceration. Descriptive analyses, logistic regression and Cox proportional hazards models were conducted using survey and linked administrative data for 1,238 Australian prisoners. We found that Indigenous people compared to non-Indigenous people were less likely to be visited, and travel distance decreased the likelihood of visitation for both groups. Moreover, visitation protected against reincarceration for non-Indigenous people only. Potential explanations for why visitation was not found to protect against reincarceration for Indigenous people are discussed as are potential implications for criminal justice systems that house Indigenous offenders.

scholarly journals Trajectories of Depression and Their Associations with Mortality in Older People in Korea over 12 Years

2021 ◽  
Vol 21 (2) ◽  
pp. 45-55
Author(s):  
Sinhyung Lee ◽  
Hyeon Ho Choi ◽  
Seung Hee Kim ◽  
Jungun Lee
Keyword(s):  
Depressive Symptoms ◽  
Proportional Hazards ◽  
Depression Scale ◽  
Epidemiological Studies ◽  
Cox Proportional Hazards ◽  
Men And Women ◽  
Women Subjects ◽  
Cox Proportional Hazards Models ◽  
Hazard Ratios ◽  
Korean Elderly

Background: As the course of depression and depressive symptoms over a lifetime varies between individuals, we used trajectory models based on the Korean Longitudinal Study of Aging to repeatedly measure symptoms over a follow-up period of 12 years to reveal the association with mortality.Methods: Three thousand five hundred sixty-one (1,483 men and 2,078 women) subjects aged over 65 years were included. Using the 10-item Center for Epidemiological Studies Depression Scale (CES-D 10), a trajectory model was classified into different depressive symptom groups by gender. Cox proportional hazards models were used to investigate the association between depression trajectory and all-cause mortality.Results: We identified four trajectories of depressive symptoms in both men and women characterized by low CES-D 10 scores throughout the study: Low trajectory (LT), Moderate high trajectory (MHT), High, but increasing trajectory (HIT), and High, but decreasing trajectory (HDT). The adjusted hazard ratios of the HIT, HDT, and MHT compared with LT in men were 2.12 (95% confidence interval [CI], 1.43-3.16), 1.52 (95% CI, 0.96-2.40), and 1.58 (95% CI, 1.10-2.26), respectively. In women, ratios of each group were 1.62 (95% CI, 1.25-2.10), 0.84 (95% CI, 0.61-1.16), and 1.20 (95% CI, 0.99-1.46).Conclusions: Highly depressive symptoms that increased over the 12 years period were associated with the highest risk of mortality in the Korean elderly population. The trajectory group with remitting depressive symptoms (HDT), despite having a similar baseline level of mood symptoms as the high increasing group (HIT) experienced a lower mortality risk in both men and women.

Sign in / Sign up

Export Citation Format

Share Document