Time-dependent Cox Proportional Hazards Models Research Articles

HEART FAILURE OUTCOMES AND PREDICTORS OF SUBCLINICAL ATRIAL FIBRILLATION PROGRESSION

Long-term monitoring with implanted cardiac devices can detect asymptomatic, subclinical atrial fibrillation (SCAF). We hypothesized that progression from shorter to longer SCAF episodes would be associated with a higher risk of congestive heart failure (CHF), one of the leading causes of AF-related morbidity and mortality. Subjects in the Asymptomatic Atrial Fibrillation and Stroke Evaluation in Pacemaker Patients and the Atrial Fibrillation Reduction Atrial Pacing Trial (ASSERT) were ≥65 years, had a history of hypertension, no known prior clinical AF, and had an implanted pacemaker or defibrillator. Among participants whose longest episode of SCAF was between >6 minutes and 24 hours during the first 12 months after enrollment (N=415), we evaluated the relationship between subsequent development of SCAF >24 hours or clinical AF and CHF hospitalization using time-dependent Cox proportional-hazards models. Over a mean follow-up of 2 years, 65 patients (15.7%) progressed to SCAF >24 hours or clinical AF (incidence rate 8.8%/year). The rate of CHF hospitalization among patients with versus without SCAF progression was 8.9 and 2.5%/year, respectively. SCAF progression was independently associated with CHF hospitalization in multivariable models (HR 4.58, 95%-CI 1.64 – 12.8, p=0.004). Similar results were observed after exclusion of patients with prior history of CHF (7.06, 95%-CI 1.82-27.3, p=0.005); when examining progression in patients with SCAF between >6 minutes and 12 hours in the first year (HR 7.73, 95%-CI 2.52-23.8, p < 0.001); and when defining SCAF progression as >24 hours only (HR 3.68, 95%-CI 1.27-10.7, p=0.016). Older age (HR 1.59, 95%-CI 1.05-2.39 per 10 year increment), increasing body mass index (HR 1.83, 95%-CI 1.14-2.94 per 10 kg/m2 increment) and longer SCAF duration within the first year of enrollment (HR 1.13, 95%-CI 1.09-1.17 per 1 hour increment) were independent predictors of SCAF progression. In this prospective cohort of older, hypertensive patients with an implanted pacemaker or defibrillator, SCAF progression was a strong and independent predictor of CHF hospitalization.

Androgen Deprivation Therapy and the Risk of Anemia in Men with Prostate Cancer.

The use of androgen deprivation therapy in prostate cancer may be associated with an increased risk of anemia, but the evidence remains limited. This study aimed to determine if androgen deprivation is associated with increased risk of anemia in patients newly diagnosed with prostate cancer. This was a population-based cohort study using the United Kingdom Clinical Practice Research Datalink linked to the Hospital Episode Statistics repository. The cohort consisted of 10,364 men newly diagnosed with nonmetastatic prostate cancer between 1 April 1998 and 30 September 2015. We used time-dependent Cox proportional hazards models to estimate adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for anemia (hemoglobin <130 g/L) associated with current and past use of androgen deprivation therapy, compared with nonuse. There were 3,651 incident anemia events during 31,574 person-years of follow-up (rate: 11.6/100 person-years). Current androgen deprivation therapy use was associated with a nearly three-fold increased hazard of anemia, compared with nonuse (23.5 vs. 5.9 per 100 person-years, respectively; HR: 2.90, 95% CI: 2.67, 3.16). The HR was elevated in the first 6 months of use (HR: 2.20, 95% CI: 1.95, 2.48) and continued to be elevated with longer durations of use. Past androgen deprivation therapy use was associated with a lower estimate (HR: 1.27, 95% CI: 1.12, 1.43), which returned closer to the null ≥25 months after treatment discontinuation (HR: 0.95, 95% CI: 0.79, 1.15). The use of androgen deprivation therapy is associated with increased risk of anemia, which reverses upon treatment discontinuation.

HIV Infection Increases the Risk of Incident Psoriasis: A Nationwide Population-Based Cohort Study in Taiwan

HIV can cause an imbalance of T lymphocytes, which may contribute to the onset of psoriasis. However, the association of HIV with incident psoriasis has not been extensively studied. The aim of this nationwide population-based cohort study was to determine the association of HIV with incident psoriasis. Since January 1, 2000, we identified adult people living with HIV/AIDS (PLWHA) from the Taiwan National Health Insurance Research Database. A control cohort without HIV infection, matched for age and sex, was selected for comparison. All patients were followed until December 31, 2012, and observed for the occurrence of psoriasis. The time-dependent Cox proportional hazards model was used to determine the association of HIV with incident psoriasis, while considering death as a competing risk event. Of the 102,070 patients (20,294 PLWHA and 81,776 matched controls), 248 (0.24%) had incident psoriasis during a mean follow-up period of 5.53 years, including 81 (0.40%) PLWHA and 171 (0.21%) controls. After adjusting for age, sex, and comorbidities, HIV infection was found to be an independent risk factor for incident psoriasis (adjusted hazard ratio, 1.80; 95% confidence interval: 1.38 to 2.36). The population of PLWHA is living longer; clinicians need to be aware of their higher risk of psoriasis.

Effect of complete or partial proteinuria recovery on long-term outcomes of lupus nephritis.

We aimed to evaluate the effect of complete recovery (CR), partial recovery (PR), and no recovery (NR) of proteinuria at 2 years from the diagnosis of LN on long-term renal and extra-renal outcomes. Patients with LN and proteinuria attending the Lupus Center from 1970 to 2015 were analyzed. At 2 years from diagnosis of LN, patients were divided into three groups (CR, PR, and NR), and long-term outcomes were studied up to 15 years or last visit available. CR was defined as resolution of proteinuria, PR as a reduction ≥50% in baseline proteinuria, and NR as a reduction <50% compared to baseline. Long-term outcomes examined included renal outcomes [low eGFR, ESRD, and composite renal (low eGFR, ESRD, and dialysis/transplant)], cardiovascular outcomes, damage, and death. Kaplan-Meier plots, time-independent and time-dependent Cox proportional hazards models were applied to examine the effect of CR, PR, or NR on long-term outcomes. Of 277 patients, 71.8% achieved CR, 18.4% PR, and 9.8% NR at 2 years. CR compared to NR and CR compared to PR were protective against low eGFR and composite renal outcome in time-independent and time-dependent analyses. CR compared to PR protected against damage in the time-independent analysis. Overall, the comparison of CR and PR favored CR for long-term renal outcomes. CR at 2 years from diagnosis of LN protected against renal outcomes (low eGFR, ESRD/dialysis, and transplant). CR is more favorable compared to PR and clinicians should aim for CR to improve long-term outcomes in LN.

The eukaryotic gut virome in hematopoietic stem cell transplantation: new clues in enteric graft-versus-host disease.

Much attention has been focused on the role of the bacterial microbiome in human health, but the virome is understudied. Although previously investigated in individuals with inflammatory bowel diseases or solid-organ transplants, virome dynamics in allogeneic hematopoietic stem cell transplantation (HSCT) and enteric graft-versus-host disease (GVHD) remain unexplored. Here we characterize the longitudinal gut virome in 44 recipients of HSCT using metagenomics. A viral 'bloom' was identified, and significant increases were demonstrated in the overall proportion of vertebrate viral sequences following transplantation (P = 0.02). Increases in both the rates of detection (P < 0.0001) and number of sequences (P = 0.047) of persistent DNA viruses (anelloviruses, herpesviruses, papillomaviruses and polyomaviruses) over time were observed in individuals with enteric GVHD relative to those without, a finding accompanied by a reduced phage richness (P = 0.01). Picobirnaviruses were detected in 18 individuals (40.9%), more frequently before or within a week after transplant than at later time points (P = 0.008). In a time-dependent Cox proportional-hazards model, picobirnaviruses were predictive of the occurrence of severe enteric GVHD (hazard ratio, 2.66; 95% confidence interval (CI) = 1.46-4.86; P = 0.001), and correlated with higher fecal levels of two GVHD severity markers, calprotectin and α1-antitrypsin. These results reveal a progressive expansion of vertebrate viral infections over time following HSCT, and they suggest an unexpected association of picobirnaviruses with early post-transplant GVHD.

Incidence of Pressure Ulcers During Home and Institutional Care Among Long-Term Care Insurance Beneficiaries With Dementia Using the Korean Elderly Cohort

ObjectiveTo assess whether type of long-term care service is a risk factor of the incidence of pressure ulcers among older adults with dementia who are receiving long-term care insurance (LTCI). MethodsData from LTCI beneficiaries (benefit level 1 or 2) with dementia, aged 60 and older (n = 7841), in the Korean Elderly Cohort data set from 2008 to 2013 were used. Type of long-term care service was categorized into home or institutional care using the LTCI Claims Database. The National Health Insurance Claims Database was used to identify the incidence of pressure ulcers as the outcome variable in a survival analysis using the time-dependent Cox proportional hazard model. ResultsOf the 7841 participants, 98 (1.2%) exhibited pressure ulcers. Compared with beneficiaries receiving home care, those receiving institutional care had a higher adjusted hazard ratio for pressure ulcers (hazard ratio 6.48, 95% confidence interval 3.48–10.86). These associations were particularly strong among beneficiaries without pressure ulcers during the mandatory assessment for benefit eligibility and who were partially ambulatory. DiscussionBeneficiaries receiving institutional care were more likely to have pressure ulcers than were those receiving home care. The government must monitor the quality of institutional long-term care services and encourage service providers to improve such care.

Metformin use and survival after non-small cell lung cancer: A cohort study in the US Military health system.

Research suggests that metformin may be associated with improved survival in cancer patients with type II diabetes. This study assessed whether metformin use after non-small cell lung cancer (NSCLC) diagnosis is associated with overall survival among type II diabetic patients with NSCLC in the U.S. military health system (MHS). The study included 636 diabetic patients with histologically confirmed NSCLC diagnosed between 2002 and 2007, identified from the linked database from the Department of Defense's Central Cancer Registry (CCR) and the Military Health System Data Repository (MDR). Time-dependent multivariate Cox proportional hazards models were used to assess the association between metformin use and overall survival during follow-up. Among the 636 patients, 411 died during the follow-up. The median follow-up time was 14.6 months. Increased post-diagnosis cumulative use (per 1 year of use) conferred a significant reduction in mortality (adjusted hazard ratio (HR) = 0.76; 95% CI = 0.65-0.88). Further analysis by duration of use revealed that compared to non-users, the lowest risk reduction occurred among patients with the longest duration of use (i.e. use for more than 2 years) (HR = 0.19; 95% CI = 0.09-0.40). Finally, the reduced mortality was particularly observed only among patients who also used metformin before lung cancer diagnosis and among patients at early stage of diagnosis. Prolonged duration of metformin use in the study population was associated with improved survival, especially among early stage patients. Future research with a larger number of patients is warranted.

Metformin and the incidence of viral associated cancers in patients with type 2 diabetes.

Limited studies have associated metformin with a reduced risk of viral associated cancers, however these had a number of methodological shortcomings. This study investigated whether the use of metformin is associated with a reduced risk of viral associated cancers in patients with type 2 diabetes. A cohort of 137,754 patients newly-prescribed non-insulin antidiabetic drugs between January 1, 1988 and March 31, 2016 was identified from the UK Clinical Practice Research Datalink and followed until a first-ever diagnosis of a viral associated cancer, death from any cause, end of registration with the practice, or March 31, 2016. Time-varying use of metformin was compared with use of other antidiabetic drugs, with exposures lagged by one year for latency purposes. Time-dependent Cox proportional hazards models were used to estimate adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) of incident viral associated cancer with use of metformin overall, by cumulative duration of use and viral etiology. Overall, there were 424 viral associated cancers during 759,810 person-years of follow-up (crude rate of 5.6 per 10,000 person-years). Metformin was not associated with a decreased rate of viral associated cancer (HR: 0.93, 95% CI: 0.65-1.32). There was no evidence of a duration-response relationship in terms of cumulative duration of use (p trend = 0.69), including with use of metformin for more than 10 years (HR 1.02, 95% CI: 0.52-1.99), or by viral etiology. In this large population-based cohort study, the use of metformin was not associated with a reduced risk of viral associated cancer.

Clinical Calculator for Early Mortality in Metastatic Colorectal Cancer: An Analysis of Patients From 28 Clinical Trials in the Aide et Recherche en Cancérologie Digestive Database.

Purpose Factors contributing to early mortality after initiation of treatment of metastatic colorectal cancer are poorly understood. Materials and Methods Data from 22,654 patients enrolled in 28 randomized phase III trials contained in the ARCAD (Aide et Recherche en Cancérologie Digestive) database were pooled. Multivariable logistic regression models for 30-, 60-, and 90-day mortality were constructed, including clinically and statistically significant patient and disease factors and interaction terms. A calculator (nomogram) for 90-day mortality was developed and validated internally using bootstrapping methods and externally using a 10% random holdout sample from each trial. The impact of early progression on the likelihood of survival to 90 days was examined with time-dependent Cox proportional hazards models. Results Mortality rates were 1.4% at 30 days, 3.4% at 60 days, and 5.5% at 90 days. Among baseline factors, advanced age, lower body mass index, poorer performance status, increased number of metastatic sites, BRAF mutant status, and several laboratory parameters were associated with increased likelihood of early mortality. A multivariable model for 90-day mortality showed strong internal discrimination (C-index, 0.77) and good calibration across risk groups as well as accurate predictions in the external validation set, both overall and within patient subgroups. Conclusion A validated clinical nomogram has been developed to quantify the risk of early death for individual patients during initial treatment of metastatic colorectal cancer. This tool may be used for patient eligibility assessment or risk stratification in future clinical trials and to identify patients requiring more or less aggressive therapy and additional supportive measures during and after treatment.

Dietary glycemic and insulin scores and colorectal cancer survival by tumor molecular biomarkers.

Accumulating evidence suggests that post-diagnostic insulin levels may influence colorectal cancer (CRC) survival. Yet, no previous study has examined CRC survival in relation to a post-diagnostic diet rich in foods that increase post-prandial insulin levels. We hypothesized that glycemic and insulin scores (index or load; derived from food frequency questionnaire data) may be associated with survival from specific CRC subtypes sensitive to the insulin signaling pathway. We prospectively followed 1,160 CRC patients from the Nurses' Health Study (1980-2012) and Health Professionals Follow-Up Study (1986-2012), resulting in 266 CRC deaths in 10,235 person-years. CRC subtypes were defined by seven tumor biomarkers (KRAS, BRAF, PIK3CA mutations, and IRS1, IRS2, FASN and CTNNB1 expression) implicated in the insulin signaling pathway. For overall CRC and each subtype, hazard ratio (HR) and 95% confidence interval (95% CI) for an increase of one standard deviation in each of glycemic and insulin scores were estimated using time-dependent Cox proportional hazards model. We found that insulin scores, but not glycemic scores, were positively associated with CRC mortality (HR = 1.19, 95% CI = 1.02-1.38 for index; HR = 1.23, 95% CI = 1.04-1.47 for load). The significant positive associations appeared more pronounced among PIK3CA wild-type cases and FASN-negative cases, with HR ranging from 1.36 to 1.60 across insulin scores. However, we did not observe statistically significant interactions of insulin scores with PIK3CA, FASN, or any other tumor marker (p interaction > 0.12). While additional studies are needed for definitive evidence, a high-insulinogenic diet after CRC diagnosis may contribute to worse CRC survival.

Prognostic impact of peritonitis in hemodialysis patients: A national-wide longitudinal study in Taiwan.

BackgroundPeritonitis has been independently associated with increased morbidity and mortality in peritoneal dialysis patients. However, there are few reports on peritonitis in hemodialysis patients. We aim at investigating both the risk profiles and prognostic impact of peritonitis in hemodialysis patients.MethodsThis nation-wide longitudinal study uses claims data obtained from the Taiwan National Health Insurance Research Database. A total of 80,733 incident hemodialysis patients of age ≥ 20 years without a history of peritonitis were identified between January 1, 1998 and December 31, 2009. Predictors of peritonitis events were estimated using Cox proportional hazard models. Time-dependent Cox proportional hazard models were used to estimate hazard ratio for mortality attributed to peritonitis exposure.ResultsOf 80,733 incident hemodialysis patients over a 13-year study period, peritonitis was diagnosed in 935 (1.16%), yielding an incidence rate of 2.91 per 1000 person-years. Female gender, liver cirrhosis and polycystic kidney disease were three of the most significant factors for peritonitis in both non-diabetic and diabetic hemodialysis patients. The cumulative survival rate of patients with peritonitis was 38.8% at 1 year and 10.1% at 5 years. A time-dependent Cox multivariate analysis showed that peritonitis had significantly increased hazard ratio for all cause mortality. Additionally, the risk of mortality remained significantly higher for non-diabetic hemodialysis patients that experienced peritonitis.ConclusionsThe risk of peritonitis in hemodialysis patients is higher in female gender, liver cirrhosis and polycystic kidney disease. Although peritonitis is a rare condition, it is associated with significantly poorer outcome in hemodialysis patients.

Soft drink consumption is associated with increased incidence of the metabolic syndrome only in women.

Prospective studies on the association between soft drink consumption and incident risk of the metabolic syndrome (MetS) have not been carried out in Asians. We explored the sex-specific association between soft drink consumption and incident risk of the MetS in Korean adults during 10 years of follow-up. A total of 5797 subjects who were free of the MetS at baseline were studied. Soft drink consumption was assessed using a semi-quantitative FFQ. Time-dependent Cox proportional hazard model was used to examine hazard ratios (HR) of incidence of the MetS and its components in relation to soft drink consumption. In women, the multivariable-adjusted HR for developing the MetS was 1·8-fold higher in frequent consumers of soft drinks (≥4 servings/week) compared with rare consumers (95 % CI 1·23, 2·64). The adjusted HR for elevated blood pressure increased by 2-fold (95 % CI 1·24, 3·14) and for hypertriacylglycerolaemia by 1·9-fold (95 % CI 1·19, 2·88) in frequent consumers of soft drinks compared with rare consumers. However, in men, there was no association between soft drink consumption and incident risk of the MetS or its components. Frequent soft drink consumption was associated with increased risk of developing the MetS and its components only in middle-aged Korean women, suggesting sex differences for the risk of the MetS related to diet.

Balancing the risk of spontaneous ischemic and major bleeding events in acute coronary syndromes

In the PLATelet inhibition and patient Outcomes (PLATO) trial, we compared spontaneous ischemic events (myocardial infarction or stroke) with spontaneous major bleeding events (PLATO major, Thrombolysis In Myocardial Infarction [TIMI] major, Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries [GUSTO] severe) with respect to risk of mortality using time-dependent Cox proportional hazards models. The comparison was performed using ratio of hazard ratios for mortality increase after ischemic vs bleeding events. A total of 822 patients (4.4%) had ≥1 spontaneous ischemic event; 485 patients (2.6%), ≥1 spontaneous PLATO major bleed, 282 (1.5%), ≥1 spontaneous TIMI major bleed; and 207 (1.1%), ≥1 spontaneous severe GUSTO bleed. In patients who had both events, bleeding occurred first in most patients. Regardless of classification, major bleeding events were associated with increased short- and long-term mortality that were not significantly different from the increase associated with spontaneous ischemic events: ratio of hazard ratios (95% CIs) for short- and long-term mortality after spontaneous ischemic vs bleeding events: 1.46 (0.98-2.19) and 0.92 (0.52-1.62) (PLATO major); 1.26 (0.80-1.96) and 1.19 (0.58-2.24) (TIMI major), 0.72 (0.47-1.10) and 0.83 (0.38-1.79) (GUSTO severe) (all P>0.05) CONCLUSIONS: In patients with ACS on dual antiplatelet therapy, spontaneous major bleeding events seem "prognostically equivalent" to spontaneous ischemic complications. This result allows quantitative comparisons between both actual and predicted bleeding and ischemic risks. Our findings help to better define net clinical benefit of antithrombotic treatments and more accurately estimate mortality after ischemic and bleeding events in patients with ACS.

Influenza vaccination reduces hemorrhagic stroke risk in patients with atrial fibrillation: A population-based cohort study

PurposeThe risk of hemorrhagic stroke in patients with atrial fibrillation (AF) is low but the consequences of its occurrence are extremely severe. In this study, we investigated the association of influenza vaccination with the risk of hemorrhagic stroke to develop an efficient strategy for reducing this risk in patients with AF. MethodsIn this study, data were retrieved from the Taiwan National Health Insurance Research Database. The study cohort comprised all patients who received a diagnosis of AF (n=14,454) before January 1, 2005 (index date) and were followed until December 31, 2012. Propensity scores were calculated using a logistic regression model to determine the effects of vaccination by accounting for covariates that predict receiving the intervention (vaccine). A time-dependent Cox proportional hazard model was used to calculate the hazard ratios (HRs) for hemorrhagic stroke in vaccinated and unvaccinated patients with AF. ResultsThe study population comprised 6570 patients who did (2547 [38.77%]) and did not receive (4023 [61.23%]) influenza vaccination. The adjusted HRs (aHRs) for hemorrhagic stroke were lower in the vaccinated patients than in the unvaccinated patients (influenza season, noninfluenza season, and all seasons: aHRs=0.97 [0.59–1.60], 0.51 [0.30–0.87], and 0.72 [0.50–1.03], respectively). ConclusionsInfluenza vaccination exerts dose–response and synergistic protective effects against hemorrhagic stroke in patients with AF who have a high risk of hemorrhagic stroke (i.e., male sex, age≥75years, Charlson comorbidity index ≥3, and hypertension) and reduces the incidence of hemorrhagic stroke.

Androgen deprivation therapy for prostate cancer and the risk of hospitalisation for community-acquired pneumonia

BackgroundAndrogens have been shown to influence both the immune system and lung tissue, raising the hypothesis that androgen deprivation therapy (ADT) for prostate cancer may increase the risk of pneumonia....

Drug survival on tumour necrosis factor inhibitors in patients with rheumatoid arthritis in Finland

Objective: A systematic review found that an average of 27% of rheumatoid arthritis (RA) patients using tumour necrosis factor (TNF) inhibitors discontinue their treatment within 1 year. The aim of this study was to assess drug survival on TNF inhibitors among patients with RA.Methods: Patients were identified from the National Register for Biologic Treatment in Finland (ROB-FIN), which is a longitudinal cohort study established to monitor the effectiveness and safety of biologic drugs in rheumatic diseases. Inclusion was limited to TNF-inhibitor treatments started as the patient’s first, second, or third biologic treatment between 2004 and 2014. Follow-up was truncated at 36 months. The results of a time-dependent Cox proportional hazards model were reported as adjusted hazard ratios (HRs) with 95% confidence intervals (CIs).Results: Of the 4200 TNF-inhibitor treatment periods identified from ROB-FIN, 3443 periods from 2687 patients met the inclusion criteria. Twenty-seven per cent of the patients discontinued their treatment within 12 months. Infliximab (HR 1.8, 95% CI 1.3–2.5) and certolizumab pegol (HR 1.7, 95% CI 1.2–2.3) had lower drug survival compared to golimumab. A similar trend was seen with adalimumab (HR 1.2, 95% CI 0.90–1.7) and etanercept (HR 1.2, 95% CI 0.87–1.6). Concomitant use of methotrexate (MTX) was associated with improved drug survival (HR 0.76, 95% CI 0.64–0.90) in comparison with TNF-inhibitor monotherapy.Conclusions: Golimumab was better in terms of drug survival than infliximab or certolizumab pegol and at least as good as adalimumab and etanercept. Concomitant use of MTX improved drug survival on TNF inhibitors.

Androgen Deprivation Therapy and the Risk of Dementia in Patients With Prostate Cancer.

Purpose Recent observational studies have associated the use of androgen deprivation therapy (ADT) with an increased risk of dementia and Alzheimer's disease, but these studies had limitations. The objective of this study was to determine whether the use of ADT is associated with an increased risk of dementia, including Alzheimer's disease, in patients with prostate cancer. Patients and Methods Using the United Kingdom's Clinical Practice Research Datalink, we assembled a cohort of 30,903 men newly diagnosed with nonmetastatic prostate cancer between April 1, 1988 and April 30, 2015, and observed them until April 30, 2016. Time-dependent Cox proportional hazards models were used to estimate adjusted hazard ratios with 95% CIs of dementia associated with the use of ADT compared with nonuse. ADT exposure was lagged by 1 year to account for delays associated with the diagnosis of dementia and to minimize reverse causality. Secondary analyses assessed whether the risk varied with cumulative duration of use and by ADT type. Results During a mean (standard deviation) follow-up of 4.3 (3.6) years, 799 patients were newly diagnosed with dementia (incidence, 6.0; 95% CI, 5.6 to 6.4) per 1,000 person-years. Compared with nonuse, ADT use was not associated with an increased risk of dementia (incidence, 7.4 v 4.4 per 1,000 person-years, respectively; adjusted hazard ratio, 1.02; 95% CI, 0.87 to 1.19). In secondary analyses, cumulative duration of use ( P for heterogeneity = .78) and no single type of ADT were associated with an increased risk of dementia. Conclusion In this population-based study, the use of ADT was not associated with an increased risk of dementia. Additional studies in different settings are needed to confirm these findings.

Abstract 12116: Long-term Effects of Cumulative Fine Particulate Matter Exposures on Hemorrhagic Stroke Event: Nationwide Cohort Study

Introduction: Epidemiological studies have revealed adverse associations between ambient fine particulate matter (PM 2.5 ; aerodynamic diameter &lt; 2.5 μm) and cardiovascular diseases. Recently, there has been increased interest in investigating the association between long-term PM 2.5 exposure and hemorrhagic stroke (HS) and identifying the most susceptible subpopulations to the PM 2.5 exposure, if any. Hypothesis: We assessed the hypothesis that long-term cumulative averages of PM 2.5 exposure affect HS and its effect might be modified by risk factors. Methods: This study was based on a nationwide representative cohort of 1,025,340 participants, or 2% of the National Health Insurance Service enrollees in Korea (2002-2013). We analyzed the effects of PM 2.5 on time to first hospital diagnosis for HS for participants living in Seoul. A time-dependent Cox proportional hazards model was used with adjustments for age, sex, household income, insurance type, body mass index, smoking status, exercise, medical history, and family history. Annual cumulative mean PM 2.5 concentrations for 25 districts were used as the time-dependent variables. Subgroup analyses were performed to evaluate potential effect modification. Results: Cumulative PM 2.5 exposure was associated with HS events with a hazard ratio of 1.04 (95% confidence interval [CI]: 1.01-1.07) per 1 μg/m 3 elevation in PM 2.5 . Its adverse effects were modified by risk factors with greater effects observed among women (1.05 [1.01-1.09]) and subgroups with hypertension (1.06 [1.03-1.10]), without diabetes (1.05 [1.02-1.09]), or without family history of stroke (1.04 [1.01-1.07]). Conclusions: To our best knowledge, this is the first study to investigate the effect of cumulative averages of PM 2.5 exposure on time to first hospital diagnosis for HS. Cumulative PM 2.5 exposure might increase the risk of HS. Participants with hypertension, without diabetes or family history of stroke, or women were more susceptible to PM 2.5 exposure. Our findings show the first time that long-term cumulative PM 2.5 exposure could be crucial to cerebrovascular event and have public health implications.

Influence of Donor and Recipient CYP3A4, CYP3A5, and ABCB1 Genotypes on Clinical Outcomes and Nephrotoxicity in Liver Transplant Recipients

This study investigated the influence of the CYP3A4*22, CYP3A5*3, and ABCB1 exons 12, 21, and 26 polymorphisms in donors and recipients on clinical outcomes and renal function in 170 liver transplant patients on cyclosporin A (CsA) or tacrolimus (Tac). Allelic discrimination assays were used for genotyping. Multivariate time-dependent Cox proportional hazard models, multiple linear regression using the generalized estimating equation and linear mixed-effect models were used for statistical analysis. Expression of CYP3A5 by either or both the donor and the recipient was significantly associated with lower Tac, but not CsA, dose-normalized trough levels. In the whole population, graft loss was only significantly associated with longer exposure to high calcineurin inhibitor (CNI) concentrations (hazard ratio, 6.93; 95% confidence interval, 2.13-22.55), P = 0.00129), whereas in the Tac subgroup, the risk of graft loss was significantly higher in recipient CYP3A5*1 expressers (hazard ratio, 3.39; 95% confidence interval, 1.52-7.58; P = 0.0028). Renal function was significantly associated with: (1) baseline modification of diet in renal disease (β = 0.51 ± 0.05; P < 0.0001); (2) duration of patient follow-up (per visit, β = -0.98 ± 0.22; P < 0.0001); and (3) CNI exposure (per quantile increase, β = -2.42 ± 0.59; P < 0.0001). No genetic factor was associated with patient survival, acute rejection, liver function test results, recurrence of viral or other initial liver disease, or renal function. This study confirms the effect of CYP3A5*3 on tacrolimus dose requirement in liver transplantation and shows unexpected associations between the type of, and exposure to, CNI and either chronic rejection or graft loss. None of the genetic polymorphisms studied had a noticeable impact on renal function degradation at 10 years.

Serum Potassium Levels and Mortality in Hemodialysis Patients: A Retrospective Cohort Study

Background: Hyperkalemia is common in patients receiving maintenance hemodialysis. However, few studies have examined the association between serum potassium level and mortality. Methods: This study used annual cohorts of hemodialysis patients during 2007-2010. To determine hyperkalemia prevalence, monthly hyperkalemia was defined as serum potassium level ≥5.5 mEq/l; prevalence was calculated as a ratio of hyperkalemia episodes to follow-up time, reported separately by long and short interdialytic interval. To determine the impact of hyperkalemia on mortality, patients in the 2010 cohort were followed from first potassium measurement until death or a censoring event; hyperkalemia was defined, sequentially, by potassium levels 5.5-6.0 mEq/l at 0.1 mEq/l intervals. Time-dependent Cox proportional hazards modeling was used to estimate the association between hyperkalemia and mortality. Results: The 4 annual cohorts ranged from 28,774 to 36,888 patients. Mean age was approximately 63 years, about 56% were men, 51% were white and 44% had end-stage renal disease caused by diabetes. Hyperkalemia prevalence was consistently estimated at 16.3-16.8 events per 100 patient-months. Prevalence on the day after the long interdialytic interval was 2.0-2.4 times as high as on the day after the short interval. Hyperkalemia, when defined as serum potassium ≥5.7 mEq/l, was associated with all-cause mortality (adjusted hazards ratio (AHR) 1.13, 95% CI 1.01-1.28, p = 0.037, vs. <5.7 mEq/l) after adjustment. AHRs increased progressively as the hyperkalemia threshold increased, reaching 1.37 (95% CI 1.16-1.62, p < 0.0001) for ≥6.0 mEq/l. Conclusions: The long interdialytic interval was associated with increased likelihood of hyperkalemia. Hyperkalemia was associated with all-cause mortality beginning at serum potassium ≥5.7 mEq/l; mortality risk estimates increased ordinally through ≥6.0 mEq/l, suggesting a threshold at which serum potassium becomes substantially more dangerous.