Time-dependent Cox Proportional Hazards Models Research Articles

Dipeptidyl Peptidase 4 Inhibitors and the Risk of Bullous Pemphigoid Among Patients With Type 2 Diabetes.

There are uncertainties regarding the association between dipeptidyl peptidase 4 (DPP-4) inhibitors and bullous pemphigoid (BP), a potentially severe autoimmune skin disease. Thus, we conducted a population-based study to determine whether use of DPP-4 inhibitors, when compared with other second- to third-line antidiabetic drugs, is associated with an increased risk of BP in patients with type 2 diabetes. Using the U.K. Clinical Practice Research Datalink, we conducted a cohort study among 168,774 patients initiating antidiabetic drugs between January 2007 and March 2018. Using time-dependent Cox proportional hazards models, we estimated adjusted hazard ratios (HRs) with 95% CIs of incident BP associated with current use of DPP-4 inhibitors, compared with current use of other second- to third-line antidiabetic drugs. We also conducted a propensity score-matched analysis to assess the impact of residual confounding. During 711,311 person-years of follow-up, 150 patients were newly diagnosed with BP (crude incidence rate, 21.1 per 100,000 person-years). Current use of DPP-4 inhibitors was associated with an increased risk of BP (47.3 vs. 20.0 per 100,000 person-years; HR 2.21 [95% CI 1.45-3.38]). HRs gradually increased with longer durations of use, reaching a peak after 20 months (HR 3.60 [95% CI 2.11-6.16]). Similar results were obtained in the propensity score-matched analysis (HR 2.40 [95% CI 1.13-4.66]). In this large population-based study, use of DPP-4 inhibitors was associated with an at least doubling of the risk of BP in patients with type 2 diabetes, albeit the absolute risk was low.

Blood Pressure Change from Normal to 2017 ACC/AHA Defined Stage 1 Hypertension and Cardiovascular Risk.

The purpose of this study was to investigate the clinical significance of the 2017 American College of Cardiology (ACC)/American Heart Association (AHA) defined stage 1 hypertension (systolic blood pressure (SBP) 130–139 mmHg or diastolic blood pressure (DBP) 80–89 mmHg), and increase in BP from previously normal BP in Korean adults. We conducted a retrospective analysis of 60,866 participants from a nationally representative claims database. Study subjects had normal BP (SBP < 120 mmHg and DBP < 80 mmHg), no history of anti-hypertensive medication, and cardiovascular disease (CVD) in the first period (2002–2003). The BP change was defined according to the BP difference between the first and second period (2004–2005). We used time-dependent Cox proportional hazards models in order to evaluate the effect of BP elevation on mortality and CVD with a mean follow-up of 7.8 years. Compared to those who maintained normal BP during the second period, participants with BP elevation from normal BP to stage 1 hypertension had a higher risk for CVD (adjusted hazard ratio (aHR) 1.23; 95% confidence interval (CI), 1.08–1.40), and ischemic stroke (aHR 1.32; 95% CI, 1.06–1.64). BP elevation to 2017 ACC/AHA defined elevated BP (SBP 120–129 mmHg and DBP < 80 mmHg) was associated with an increased risk of CVD (aHR 1.26; 95% CI, 1.06–1.50), but stage 1 isolated diastolic hypertension (SBP < 130 and DBP 80–89 mmHg) was not significantly related with CVD risk (aHR 1.12; 95% CI, 0.95–1.31).

1665-P: Fracture Risk in Patients Using Sodium-Glucose Cotransporter 2 Inhibitors: A Real-World Study

There are concerns that sodium-glucose co-transporter 2 inhibitors (SGLT2Is) may increase the risk of fractures in patients with type 2 diabetes. To date, large cardiovascular outcome trials have generated conflicting findings, and few real-world studies have been conducted to assess this association. As fracture risk in this vulnerable population is associated with increased mortality, there is an urgent need to address this safety issue using real-world data. Thus, the objective of this population-based cohort study is to determine if use of SGLT2Is is associated with an increased fracture risk in patients with type 2 diabetes. Using the UK Clinical Practice Research Datalink, a large primary care database, we identified a cohort of patients newly treated with antihyperglycemic drugs between January 2013 and December 2017. Use of SGLT2Is was modeled as a time-varying exposure and compared with use of dipeptidyl peptidase-4 inhibitors (DPP-4Is). Time-dependent Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% CIs of incident fractures associated with use of SGLT2Is overall and according to specific SGLT2Is (dapagliflozin, empagliflozin, canagliflozin). We conducted several sensitivity analyses to address potential sources of bias. The cohort included 73,178 patients, which generated 153,179 person-years of observation. During follow-up, 1,973 patients were newly diagnosed with fractures (incidence rate: 12.9/1000 per year). Compared with DPP-4Is, SGLT2Is were not associated with an increased risk of fracture (HR: 0.97, 95% CI: 0.79-1.19). Similar findings were observed when stratifying on specific SGLT2Is, with HRs ranging between 0.67 and 1.42, with all CIs including the null value. The sensitivity analyses yielded highly consistent findings. The results of this large population-based study suggest that use of SGLT2Is is not associated with fracture incidence. This finding should provide reassurance on the safety of this class of drugs. Disclosure D. Abrahami: None. A. Douros: None. H. Yin: None. O. Yu: None. L. Azoulay: None. Funding Canadian Institutes of Health Research

Testosterone Replacement Therapy and the Risk of Prostate Cancer in Men With Late-Onset Hypogonadism.

The association between the use of testosterone replacement therapy (TRT) and prostate cancer remains uncertain. Thus, we investigated whether TRT is associated with an increased risk of prostate cancer in men with late-onset hypogonadism. We used the UK Clinical Practice Research Datalink to assemble a cohort of 12,779 men who were newly diagnosed with hypogonadism between January 1, 1995, and August 31, 2016, with follow-up until August 31, 2017. Exposure to TRT was treated as a time-varying variable and lagged by 1 year to account for cancer latency, with nonuse as the reference category. During 58,224 person-years of follow-up, a total of 215 patients were newly diagnosed with prostate cancer, generating an incidence rate of 3.7 per 1,000 person-years. In time-dependent Cox proportional hazards models, use of TRT was not associated with an overall increased risk of prostate cancer (hazard ratio = 0.97; 95% confidence interval: 0.71, 1.32) compared with nonuse. Results remained consistent in secondary and sensitivity analyses, as well as in a propensity score-matched cohort analysis that further assessed the impact of residual confounding. Overall, the use of TRT was not associated with an increased risk of prostate cancer in men with late-onset hypogonadism.

Association of changes in bone mineral parameters with mortality in haemodialysis patients: insights from the ARO cohort.

There is little information in haemodialysis (HD) patients on whether temporal changes in serum calcium, phosphate or intact parathyroid hormone (iPTH) are associated with mortality. We analysed associations of phosphate, total calcium and iPTH with all-cause and cardiovascular mortality in 8817 incident HD patients from the European second Analyzing Data, Recognizing Excellence and Optimizing Outcomes (AROii) cohort enrolled in 2007-09, which were prospectively followed for a median of 3 years, using time-dependent Cox proportional hazards models. We evaluated changes in risk over time depending on changes in phosphate, calcium or iPTH. The association of phosphate and iPTH with all-cause mortality was U-shaped, with the lowest risk ranges between 1.20 and 1.89 mmol/L for phosphate and between 239 and 710 ng/L for iPTH. For total calcium, the associations were J-shaped, with an increased risk for all-cause mortality at levels >2.36 mmol/L. Lowest risk ranges for cardiovascular mortality did not change markedly for all three parameters. If iPTH was below the lowest risk range at baseline (iPTH <239 ng/L), a subsequent increase in levels was associated with improved survival. For phosphate, an increase or decrease out of the lowest risk range was associated with increased mortality risk. For calcium, this was only the case when the values increased above the lowest risk range. In the AROii cohort, the ranges of bone mineral biomarkers associated with the lowest mortality ranges were largely consistent with the current Kidney Disease: Improving Global Outcomes chronic kidney disease-mineral and bone disorder guideline recommendations. Allowing a suppressed iPTH to increase was associated with a lower mortality, whereas shifts of phosphate or calcium outside the lowest risk range increased mortality.

Cardiovascular and Cerebrovascular Safety of Testosterone Replacement Therapy Among Aging Men with Low Testosterone Levels: A Cohort Study

PurposeWe assessed the risk of ischemic stroke, transient ischemic attack, and myocardial infarction associated with testosterone replacement therapy (TRT) among aging men with low testosterone levels. MethodsUsing the UK Clinical Practice Research Datalink, we formed a cohort of men aged 45 years or older with low testosterone levels and no evidence of hypogonadotropic or testicular disease, between 1995 and 2017. Hazard ratios (HRs) and 95% confidence intervals (CIs) of a composite of ischemic stroke/transient ischemic attack and myocardial infarction were estimated using time-dependent Cox proportional hazards models, comparing current use of TRT with nonuse. ResultsThe cohort included 15,401 men. During 71,541 person-years of follow-up, 850 patients experienced an ischemic stroke/transient ischemic attack/myocardial infarction (crude incidence rate 1.19 [95% confidence interval (CI), 1.11-1.27] per 100 persons per year). Compared with nonuse, current use of TRT was associated with an increased risk of the composite outcome (HR 1.21; 95% CI, 1.00-1.46). This risk was highest in the first 6 months to 2 years of continuous TRT use (HR 1.35; 95% CI, 1.01-1.79), as well as among men aged 45-59 years (HR 1.44; 95% CI, 1.07-1.92). ConclusionsTRT may increase the risk of cardiovascular events in aging men with low testosterone levels, particularly in the first 2 years of use. In the absence of identifiable causes of hypogonadism, TRT should be initiated with caution among aging men with low testosterone levels.

The association between platelet transfusions and mortality in patients with critical illness.

Platelet (PLT) transfusions are frequently administered in the setting of critical illness but their clinical impacts remain unknown. This study examined the association between PLT transfusions and death in a large intensive care unit (ICU) patient population. Using a transfusion registry spanning 2008 to 2015, this study assessed effect of in-ICU PLT transfusions on ICU and in-hospital mortality using a stratified, time-dependent Cox proportional hazards model adjusted for illness severity, thrombocytopenia, and other confounders. Patients with known malignancy were excluded. Exposure to PLT transfusions were analyzed dichotomously (ever or never transfused) and continuously (number of transfusions). Medical, general surgery, and cardiac surgery subgroups were analyzed separately. Overall 32,842 adult patients were admitted to ICU, and 4927 patients received PLT transfusion(s). Crude in-ICU and in-hospital mortality were higher for PLT-transfused patients compared to nontransfused patients (9.2% vs. 6.7% and 12.3% vs. 9.3%, respectively). After confounders were adjusted for, PLT transfusions (ever vs. never) were not associated with increased mortality in ICU (hazard ratio [HR], 0.78; 95% confidence interval [CI], 0.60-1.02; p = 0.06) or in hospital (HR, 0.89; 95% CI, 0.68-1.09; p = 0.41). Continuous exposure analysis also showed no association between PLT transfusions and death. PLT transfusions have a protective effect on in-hospital mortality in the subgroup of general surgery patients (HR, 0.71; 95% CI, 0.51-0.99; p = 0.04; ever or never analysis). Platelet transfusions were not associated with increased risk of death in critically ill patients. Further studies are required to identify subgroups for which PLT transfusions may be beneficial.

Abstract P345: Change in Cardiovascular Health Metrics Over Time, CVD Events and Total and Cause-Specific Mortality

Background: The impact of changes in cardiovascular health (CVH) on cardiovascular disease (CVD) and total and all-cause mortality has yet to be described. Methods: CVH was computed according to smoking, body mass index, total cholesterol, blood glucose and blood pressure, physical activity and diet. Change in CVH was defined as a point-to-point difference in each metric or the score. We used time-dependent Cox Proportional Hazard models to calculate hazard ratios for all-cause mortality and CVD events among 10,656 adult participants from the ARIC study, aged 44 to 66 years at baseline (1987-1989) and followed up until 2014. Hazard ratios for all-cause mortality and CVD event according to CVH change at the component metrics and an aggregate score level were calculated with Cox Proportional Hazard models with consistently low CVH considered as the reference group. Results: Overall, 17% of the sample improved their overall CVH, while the percentage which maintained a low CVH or decreased CVH was 29% and 21%, respectively. Higher levels of overall CVH over time were associated with a graded decrease in risk in CVD events and all-cause mortality. The hazard ratios for all-cause mortality among participant that decreased their overall CVH from favorable to low or moderate, that increased their CVH from low to moderate or favorable, and had consistently favorable CVH, as compared with the constantly low CVH group, were: 0.47 (95% confidence interval [CI], 0.39 to 0.57), 0.80 (95 CI%, 0.72 to 0.89), and 0.37 (95% CI, 0.30 to 0.46). The risk reductions were of a same magnitude for CVD events. In the adjusted Cox time dependent model, compared with low overall CVH, having moderate or favorable CVH was associated with a decreased risk in mortality: 24% (HR=0.76, 95% CI, 0.72 to 0.80) and 44% (HR=0.56, 95% CI, 0.51 to 0.62), respectively and a decreased risk in CVD 37% (HR=0.63, 95% CI, 0.59 to 0.66) and 56% (HR=0.44, 95% CI, 0.39 to 0.49), respectively. Conclusion: Earlier life stage favorable CVH was associated with lower CVD events and total and cause-specific mortality regardless of CVH change patterns over time. Furthermore, improving CVH is associated with lower CVD event risk and lower total and all-cause mortality. However, we observed an alarming low percentage of overall CVH improvement and a high percentage of maintaining low overall CVH. Understanding the mechanisms underlying CVH change patterns may help to tackle the low prevalence of moderate or optimal CVH.

Human Immunodeficiency Virus Increases the Risk of Incident Heart Failure.

Although the HIV can cause myocardial inflammation, the association of HIV infection with subsequent development of heart failure (HF) has not been extensively studied. This nationwide cohort study aimed to determine the risk of incident HF in people living with HIV/AIDS (PLWHA). We identified PLWHA using the Taiwan Centers for Disease Control and Prevention HIV Surveillance System. An age- and sex-matched control group without HIV infection was selected from the Taiwan National Health Insurance Research Database for comparison. All patients were followed up until December 2014 and were observed for a new diagnosis of HF. A time-dependent Cox proportional hazards model was used to determine the association of HIV and highly active antiretroviral therapy with incident HF, with death as a competing risk event. Of the 120,765 patients (24,153 PLWHA and 96,612 matched controls), 641 (0.53%) had incident HF during a mean follow-up period of 5.84 years, including 192 (0.79%) PLWHA and 449 (0.46%) controls. Time to diagnosis of incident HF was significantly shorter in PLWHA than in those without HIV infection (P < 0.001, the log-rank test). After adjusting for age, sex, and comorbidities, HIV infection was found to be an independent risk factor for incident HF (adjusted hazard ratio, 1.52; 95% confidence interval: 1.27 to 1.82). As the duration of highly active antiretroviral therapy increased, the risk of HF decreased (P = 0.014). HIV infection was an independent risk factor for incident HF. Clinicians need to be aware of the higher risk of HF in PLWHA.

Elevated Risk of Cancer After Solid Organ Transplant in Childhood: A Population-based Cohort Study.

Cancer risk is elevated among adult transplant recipients, but there is limited data regarding long-term cancer risk and mortality in pediatric recipients. We conducted a population-based retrospective cohort study in Ontario, Canada. We included pediatric recipients of solid organ transplants at the Hospital for Sick Children, Toronto, from 1991 to 2014, and compared rates of new cancers and cancer-specific mortality to nontransplanted Ontario children born in the same year. We constructed standard and time-dependent Cox proportional hazards models accounting for competing risk of death. A total of 951 recipients (kidney, n = 400; liver, n = 283; heart, n = 218; lung, n = 36; multiorgan/small bowel, n = 14) were compared with 5.3 million general population children. Mean (SD) age was 8.2 (6.4) years; 50% were male. Over a mean (SD) follow-up of 10.8 (7.1) years, cumulative incidence of cancer was 20% in recipients and 1.2% in the general population (incidence rate ratio, 32.9; 95% confidence interval [CI], 26.6-40.8). Risk was highest in the first year posttransplant (adjusted hazard ratio [aHR],176; 95% CI, 117-264), but remained elevated beyond 10 years (aHR, 10.8; 95% CI, 6.3-18.6). Lymphoproliferative disorders were predominant (77%); however, solid cancers (renal, sarcomas, genital, thyroid) were seen. Recipients of lung or multiorgan transplants were at highest risk. Cancer-specific mortality was also higher among recipients (HR, 93.1; 95% CI, 59.6-145.2). Childhood transplant recipients have a 30 times greater cancer incidence versus the general population. Further investigation is needed to guide screening strategies in this at-risk population.

Clinical events associated with cumulative corticosteroid use in patients with castration-resistant prostate cancer: An administrative claims analysis.

236 Background: Corticosteroids (CSs) are used in the treatment of castration-resistant prostate cancer (CRPC) to manage adverse events and tumor-related symptoms. This study evaluated the impact of cumulative CS use on the risk of developing specific clinical events in men with CRPC. Methods: Adult chemotherapy-naïve patients who initiated a CRPC treatment following surgical or medical castration were identified from the MarketScan database (2007–2016). Patients were grouped into 4 cohorts based on their cumulative CS dose since 1 year before initiating their CRPC treatment: no use (0 g); low use ( &lt; 0.5 g); medium use (0.5–2.0 g); and high use ( &gt; 2.0 g). Time-dependent Cox proportional hazards models adjusting for baseline characteristics were used to evaluate the association between CS use and the risk of events (fracture, infection, endocrine disorders, acute cardiovascular [CV] events, peptic ulcer, and mental health conditions), with no CS use as the reference group. Results: This study included 9425 patients: 6765 no CS use; 1660 low use; 655 medium use; and 345 high use. On average, the no-CS-use cohort was older and had a lower comorbidity burden than the other 3 cohorts. During the study period, cumulative CS exposure (across all categories) was associated with a significantly higher risk of developing an infection (comparing high-use vs. no-use cohort, adjusted hazard ratio [HR] = 2.55; p for trend &lt; 0.001), fracture (HR = 1.59; p for trend &lt; 0.001), and acute CV events (HR = 1.62; p for trend &lt; 0.001). Also, patients in the high-CS-use cohort were more likely to develop endocrine disorder (HR = 1.61; p for trend &lt; 0.001), peptic ulcer (HR = 1.91; p for trend &lt; 0.001), or mental health condition (HR = 1.28; p for trend = 0.014) in comparison with patients in the no-use cohort. Conclusions: Cumulative CS exposure was associated with a greater risk of clinical events such as infection, fracture, and acute CV events among men with CRPC. The risk was more pronounced with higher cumulative CS doses of &gt; 2.0 g. The study provides important data on the impacts of cumulative CS use on the risk of clinical events among CRPC patients who tend to be older and more prone to these events.

Clinical outcomes in patients with chronic hepatitis C after direct-acting antiviral treatment: a prospective cohort study

Although direct-acting antivirals have been used extensively to treat patients with chronic hepatitis C virus (HCV) infection, their clinical effectiveness has not been well reported. We compared the incidence of death, hepatocellular carcinoma, and decompensated cirrhosis between patients treated with direct-acting antivirals and those untreated, in the French ANRS CO22 Hepather cohort. We did a prospective study in adult patients with chronic HCV infection enrolled from 32 expert hepatology centres in France. We excluded patients with chronic hepatitis B, those with a history of decompensated cirrhosis, hepatocellular carcinoma, or liver transplantation, and patients who were treated with interferon-ribavirin with or without first-generation protease inhibitors. Co-primary study outcomes were incidence of all-cause mortality, hepatocellular carcinoma, and decompensated cirrhosis. The association between direct-acting antivirals and these outcomes was quantified using time-dependent Cox proportional hazards models. This study is registered with ClinicalTrials.gov, number NCT01953458. Between Aug 6, 2012, and Dec 31, 2015, 10 166 patients were eligible for the study. 9895 (97%) patients had available follow-up information and were included in analyses. Median follow-up was 33·4 months (IQR 24·0-40·7). Treatment with direct-acting antivirals was initiated during follow-up in 7344 patients, and 2551 patients remained untreated at the final follow-up visit. During follow-up, 218 patients died (129 treated, 89 untreated), 258 reported hepatocellular carcinoma (187 treated, 71 untreated), and 106 had decompensated cirrhosis (74 treated, 32 untreated). Exposure to direct-acting antivirals was associated with increased risk for hepatocellular carcinoma (unadjusted hazard ratio [HR] 2·77, 95% CI 2·07-3·71) and decompensated cirrhosis (3·83, 2·29-6·42). After adjustment for variables (age, sex, body-mass index, geographical origin, infection route, fibrosis score, HCV treatment-naive, HCV genotype, alcohol consumption, diabetes, arterial hypertension, biological variables, and model for end-stage liver disease score in patients with cirrhosis), exposure to direct-acting antivirals was associated with a decrease in all-cause mortality (adjusted HR 0·48, 95% CI 0·33-0·70) and hepatocellular carcinoma (0·66, 0·46-0·93), and was not associated with decompensated cirrhosis (1·14, 0·57-2·27). Treatment with direct-acting antivirals is associated with reduced risk for mortality and hepatocellular carcinoma and should be considered in all patients with chronic HCV infection. INSERM-ANRS (France Recherche Nord & Sud Sida-HIV Hépatites), ANR (Agence Nationale de la Recherche), DGS (Direction Générale de la Santé), MSD, Janssen, Gilead, AbbVie, Bristol-Myers Squibb, and Roche.

Use of dipeptidyl peptidase-4 inhibitors and the risk of arthralgia: Population-based cohort and nested case-control studies.

The objective of this study was to investigate the association between the administration of dipeptidyl peptidase-4 (DPP-4) inhibitors (cumulative duration, timing, and individual substance) and the risk of arthralgia by using a nationwide database with two methodological approaches including cohort and nested case-control study designs. Using Taiwan's National Health Insurance Research Database, we identified patients who were newly prescribed with DPP-4 inhibitors, thiazolidinediones (TZDs), or acarbose between 1 March 2009 and 31 December 2012. The exposure of studied drugs was categorized into five exclusive categories: DPP-4 inhibitor, TZD, acarbose, combined use, or non-use, and assessed in a time-varying manner. Time-dependent Cox proportional hazard models were used to estimate the association between DPP-4 inhibitors and the risk of arthralgia. Particularly, we tested the impact of different cumulative duration, timing, and individual substance of DPP-4 inhibitors use on risk of arthralgia. A corresponding nested case-control study using conditional logistic regression was conducted to verify this association. An increased risk of arthralgia was observed during the first year after initiating DPP-4 inhibitors (adjusted Hazard Ratio=1.35; 95% confidence interval [CI], 1.04-1.75) but the risk declined with cumulative use. This duration-response relation was not found in TZDs use and acarbose use. In the nested case-control study, there was a slightly increased risk of arthralgia (aOR=1.08; 95% CI, 1.04-1.12) associated with current DPP-4 inhibitor use. A relatively higher risk of arthralgia was associated with the initial administration of DPP-4 inhibitors, however, the risk declined among long-term users.

Association between thiopurine medication exposure and Alzheimer's disease among a cohort of patients with inflammatory bowel disease

IntroductionRas-related C3 botulinum toxin substrate 1 (Rac1), a member of the Rho-GTPase family of proteins, could be an Alzheimer's disease (AD) triggering co-factor due to its effect on both amyloid precursor protein (APP) and tau. Thiopurine medications, such as azathioprine and mercaptopurine, are immunosuppressants that suppress Rac1 activation. We hypothesize that due to their ability to suppress Rac1, thiopurines are associated with a lower risk of AD. MethodsTo explore the relationship between thiopurines and incident AD diagnosis, we conducted a national retrospective cohort study among U.S. Veterans with inflammatory bowel disease (IBD), including Crohn's disease (CD) or ulcerative colitis (UC), as well as a non-IBD control. We created propensity score-matched cohorts and estimated the hazard ratio via the time-dependent Cox proportional hazards model. ResultsThe study sample size was 66,312 patients and consisted of 24,057 IBD patients (4354 thiopurine exposed and 19,703 unexposed) and 42,255 patients without IBD or thiopurine exposure. Patients exposed to thiopurines have the lowest rate of AD, and our results demonstrate for each additional year of thiopurine exposure risk of AD is reduced by 8.3%% (adjusted HR = 0.917; 95% CI = [0.851–0.989]). DiscussionOur results support the preclinical findings implicating Rac1 in the AD disease process. A national cohort study demonstrated that Rac1 is associated with the AD process consistent with the preclinical evidence. Further exploration and evaluation of Rac1 inhibition are needed.

Use of Dipeptidyl Peptidase-4 Inhibitors and New-onset Rheumatoid Arthritis in Patients with Type 2 Diabetes.

Case reports have suggested a link between dipeptidyl peptidase-4 (DPP-4) inhibitors, antidiabetic drugs used as second- to third-line treatments, and incidence of rheumatoid arthritis. Because the DPP-4 enzyme is involved in several immunologic processes and possibly in the pathophysiology of rheumatoid arthritis, further research is warranted. This population-based study aimed to determine whether use of DPP-4 inhibitors is associated with incidence of rheumatoid arthritis. Using the United Kingdom Clinical Practice Research Datalink, we conducted a cohort study among 144,603 patients with type 2 diabetes initiating antidiabetic drugs between 2007 and 2016. We estimated hazard ratios (HRs) with 95% confidence intervals (CIs) for incident rheumatoid arthritis using time-dependent Cox proportional hazards models, comparing use of DPP-4 inhibitors with use of other antidiabetic drugs. We imposed a 6-month exposure lag period for latency and diagnostic delays. Secondary analyses included assessment of the duration-response relation and comparison with other second-line antidiabetic drugs, among others. During 567,169 person-years of follow-up, 464 patients were newly diagnosed with rheumatoid arthritis (crude incidence rate: 82 per 100,000/year). Compared with use of other antidiabetic drugs, use of DPP-4 inhibitors was not associated with an increased risk of rheumatoid arthritis (82 vs. 79 per 100,000/year; HR = 1.0; 95% CI = 0.8, 1.3), with no evidence of duration-response relation. The results did not change after using second-line antidiabetic drugs as the comparator group. In this large population-based study, use of DPP-4 inhibitors was not associated with an increased risk of incident rheumatoid arthritis.

Effects of anti-osteoporosis medications on total hip arthroplasty risks in osteoporotic patients with hip osteoarthritis in Taiwan: a nationwide cohort study.

This study aimed to assess the association between use of anti-osteoporosis medications (AOMs) and the risk of undergoing total hip arthroplasty (THA) in patients with hip osteoarthritis (OA). Using the 2008-2013 National Health Insurance Research Database, we identified patients who were first diagnosed as having hip OA. All identified patients were followed until THA, death of any cause, or December 31, 2013, whichever occurred first. All AOM exposures were divided into three categories: bisphosphonates use, non-bisphosphonates use, and no use of AOMs and assessed in a time-varying manners. The primary outcome was THA. The secondary outcome was the differences in the longitudinal utilization of NSAIDs between AOM users and non-users. Time-dependent Cox proportional hazards models were used to investigate the effect of AOM use on the risk of THA. We identified 35,870 patients who were first diagnosed as having hip OA and had no history of AOM use between 2009 and 2012. Among them, 3162 and 1667 patients had their first prescription of bisphosphonates and other non-bisphosphonates AOMs during the follow-up period. Mean age of bisphosphonates users, non-bisphosphonates users, and non-users was 75.62, 76.84, and 67.39years, respectively. Bisphosphonates or non-bisphosphonates users did not show significant change when compared to non-users in terms of risk of undergoing THA [adjusted hazard ratio (aHR) 0.972, 95% confidence interval (CI) 0.743-1.273; aHR 0.926, 95% CI 0.672-1.277]. Our results showed that the use of AOMs is not associated with decreased risk of THA in patients with hip OA.

Human Immunodeficiency Virus Infection Increases the Risk of Incident Uveitis Among People Living With HIV/AIDS.

Case reports indicated that HIV itself may be a direct cause of uveitis. However, the association of HIV with incident uveitis has not been extensively studied. This nationwide cohort study determined the association of HIV with incident uveitis. Since January 1, 2003, we identified adult people living with HIV/AIDS (PLWHA) from Taiwan Centers for Disease Control HIV Surveillance Database. A control cohort without HIV infection, matched for age and sex, was selected for comparison from the Taiwan National Health Insurance Research Database. The time-dependent Cox proportional hazards model was used to determine the associations of HIV and highly active antiretroviral therapy (HAART) with incident uveitis, while considering death as a competing risk event. Of the total 120,430 patients (24,086 PLWHA and 96,344 matched controls), 609 (0.51%) had incident uveitis, including 334 (1.39%) PLWHA and 265 (0.28%) controls. After adjusting for age, sex, and comorbidities, HIV infection was found to be an independent risk factor for incident uveitis [adjusted hazard ratio (AHR), 5.55; 95% confidence interval (CI): 4.67 to 6.59]. Within PLWHA, the risk of incident uveitis was significantly higher in those who received HAART (AHR, 2.46; 95% CI: 1.71 to 3.54). In addition, considering the short- and long-term effects of HAART on incident uveitis, HAART was found to associate with a higher risk of uveitis development within 1 year of treatment (AHR, 3.36; 95% CI: 2.41 to 4.69), but not after 1 year of HAART initiation (AHR, 1.14; 95% CI: 0.76 to 1.72). HIV infection is an independent risk factor for incident uveitis.

Glucagon-Like Peptide 1 Receptor Agonists and the Risk of Incident Diabetic Retinopathy.

Previous studies suggested that glucagon-like peptide 1 receptor agonists (GLP-1 RAs) may initially worsen and possibly increase the risk of diabetic retinopathy. However, data on this possible association remain limited. Thus, this population-based study aimed to determine whether use of GLP-1 RAs is associated with an increased risk of incident diabetic retinopathy. Using the U.K. Clinical Practice Research Datalink (CPRD), we conducted a cohort study among 77,115 patients with type 2 diabetes initiating antidiabetic drugs between January 2007 and September 2015. Adjusted hazard ratios (HRs) and 95% CIs of incident diabetic retinopathy were estimated using time-dependent Cox proportional hazards models, comparing use of GLP-1 RAs with current use of two or more oral antidiabetic drugs. In an ancillary analysis, new users of GLP-1 RAs were compared with new users of insulin. During 245,825 person-years of follow-up, 10,763 patients were newly diagnosed with diabetic retinopathy. Compared with current use of two or more oral antidiabetic drugs, use of GLP-1 RAs was not associated with an increased risk of incident diabetic retinopathy overall (HR 1.00, 95% CI 0.85-1.17). Compared with insulin, GLP-1 RAs were associated with a decreased risk of diabetic retinopathy (HR 0.67, 95% CI 0.51-0.90). The associations with diabetic retinopathy varied according to the type of comparator. When compared with use of two or more oral antidiabetic drugs, use of GLP-1 RAs was not associated with an increased risk of incident diabetic retinopathy. The apparent lower risk of diabetic retinopathy associated with GLP-1 RAs compared with insulin may be due to residual confounding.

Correction: Air Pollution and Newly Diagnostic Autism Spectrum Disorders: A Population-Based Cohort Study in Taiwan.

[This corrects the article DOI: 10.1371/journal.pone.0075510.].

Increased risk of all-cause mortality associated with domperidone use in Parkinson's patients: a population-based cohort study in the UK.

Domperidone is used to treat gastrointestinal symptoms in patients with Parkinson's disease (PD) and is linked to an increased risk of mortality. We sought to examine the risk of all-cause mortality associated with domperidone exposure in PD. We conducted a cohort study using data from the Clinical Practice Research Datalink database (1987-2011). The first recorded PD diagnosis defined index date. Time-dependent Cox proportional hazards models estimated hazard ratios (HRs) of all-cause mortality associated with domperidone use. PD patients were stratified by domperidone use (current/recent/past), with never used as the referent. Current domperidone users were stratified by daily dose, domperidone duration and other anti-Parkinson's medications. A secondary analysis compared PD patients to matched (1:1) non-PD patients. A total of 5114 PD patients were identified. Current use of domperidone among PD patients was associated with a two-fold increase in all-cause mortality (HRadj =2.00, 95% confidence interval [CI]: 1.64-2.45), as compared to patients never exposed to domperidone. All-cause mortality risk was highest in those starting domperidone in the previous month [HRadj =2.97, 95% CI: 2.06-4.27]. When compared to matched non-PD patients, PD was associated with a 43% increased risk of all-cause mortality, yet this increased to a 2.4-fold increased risk among PD patients currently using domperidone. Current use of domperidone was associated with a two-fold increased mortality risk in PD patients, as compared to PD patients that never used domperidone. The risk is highest in the first month of use and does not appear to be attributable to PD alone.