Search

BackgroundThe ideal participants for Alzheimer's disease (AD) clinical trials would show cognitive decline in the absence of treatment (i.e., placebo arm) and would also respond to the therapeutic intervention.ObjectiveTo investigate if predictive models can be an effective tool for identifying and excluding people unlikely to show cognitive decline as an enrichment strategy in AD trials.MethodWe used data from the placebo arms of two phase 3, double-blind trials, EXPEDITION and EXPEDITION2. Patients had 18 months of follow-up. Based on the longitudinal data from the placebo arm, we classified participants into two groups: one showed cognitive decline (any negative slope) and the other showed no cognitive decline (slope is zero or positive) on the Alzheimer's Disease Assessment Scale-Cognitive subscale (ADAS-cog). We used baseline data for EXPEDITION to train regression-based classifiers and machine learning classifiers to estimate probability of cognitive decline. Models were applied to EXPEDITION2 data to assess predicted performance in an independent sample. Features used in predictive models included baseline demographics, apolipoprotein E ε4 genotype, neuropsychological scores, functional scores, and volumetric magnetic resonance imaging.ResultIn EXPEDITION, 46.3% of placebo-treated patients showed no cognitive decline and the proportion was similar in EXPEDITION2 (45.6%). Models had high sensitivity and modest specificity in both the training (EXPEDITION) and replication samples (EXPEDITION2) for detecting the stable group. Positive predictive value of models was higher than the base prevalence of cognitive decline, and negative predictive value of models were higher than the base rate of participants who had stable cognition.ConclusionExcluding persons with AD unlikely to decline from the active and placebo arms of clinical trials using predictive models may boost the power of AD trials through selective inclusion of participants expected to decline.

Infection by human papillomavirus (HPV) type 16, the most oncogenic HPV type, was found to be the least affected by HIV-status and CD4 count of any of the approximately 13 oncogenic HPV types. This relative independence from host immune status has been interpreted as evidence that HPV16 may have an innate ability to avoid the effects of immunosurveillance. However, the impact of immune status on other individual HPV types has not been carefully assessed. We studied type-specific HPV infection in a cohort of 2,470 HIV-positive (HIV[+]) and 895 HIV-negative (HIV[-]) women. Semi-annually collected cervicovaginal lavages were tested for >40 HPV types. HPV type-specific prevalence ratios (PRs), incidence and clearance hazard ratios (HRs), were calculated by contrasting HPV types detected in HIV[+] women with CD4

BackgroundAmyloid-β positivity (Aβ+) based on PET imaging is part of the enrollment criteria for many of the clinical trials of Alzheimer's disease (AD), particularly in trials for amyloid-targeted therapy. Predicting Aβ positivity prior to PET imaging can decrease unnecessary patient burden and costs of running these trials.ObjectiveThe aim of this study was to evaluate the performance of a machine learning model in estimating the individual risk of Aβ+ based on gold-standard of PET imaging.MethodsWe used data from an amnestic mild cognitive impairment (aMCI) subset of the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort to develop and validate the models. The predictors of Aβ status included demographic and ApoE4 status in all models plus a combination of neuropsychological tests (NP), MRI volumetrics, and cerebrospinal fluid (CSF) biomarkers.ResultsThe models that included NP and MRI measures separately showed an area under the receiver operating characteristics (AUC) of 0.74 and 0.72, respectively. However, using NP and MRI measures jointly in the model did not improve prediction. The models including CSF biomarkers significantly outperformed other models with AUCs between 0.89 to 0.92.ConclusionsPredictive models can be effectively used to identify persons with aMCI likely to be amyloid positive on a subsequent PET scan.

Elevated cancer incidence has been reported among World Trade Center (WTC)-exposed responders, with some incidence rate ratios (IRRs) varying over time. This study describes the influence that different reference populations have on relative cancer incidence and temporal trends. Participants from the WTC Combined Rescue/Recovery Cohort (n = 65,691) were observed between 1/1/2002 and 12/31/2015 using data obtained from 13 state cancer registries. Poisson regression was used to estimate IRRs controlling for age, race/ethnicity, and calendar year. IRRs and change-points were estimated using three reference populations (New York City (NYC), New York State (NYS), and a US population). IRRs for each cancer site varied in magnitude. Prostate and thyroid cancer IRRs were significantly greater in WTC-exposed responders, while colon and lung cancer IRRs were significantly lower compared with NYC, NYS, and US population reference groups. The range of IRRs varied by reference population. Mixed findings were observed for other cancers, as results were dependent on the reference group used. A significant change-point was found only for prostate cancer, and only when compared to a US population. Our findings suggest that reference population selection will influence the IRR, timing, and statistical significance of change-point estimation, varying with follow-up length. [ABSTRACT FROM AUTHOR]

BackgroundPredicting clinical course of cognitive decline can boost clinical trials' power and improve our clinical decision-making. Machine learning (ML) algorithms are specifically designed for the purpose of prediction; however. identifying optimal features or algorithms is still a challenge.ObjectiveTo investigate the accuracy of different ML methods and different features to classify cognitively normal (CN) individuals from Alzheimer's disease (AD) and to predict longitudinal outcome in participants with mild cognitive impairment (MCI).MethodsA total of 1,329 participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI) were included: 424 CN, 656 MCI, and 249 AD individuals. Four feature-sets at baseline (hippocampal volume and volume of 47 cortical and subcortical regions with and without demographics and APOE4) and six machine learning methods (decision trees, support vector machines, K-nearest neighbor, ensemble linear discriminant, boosted trees, and random forests) were used to classify participants with normal cognition from participants with AD. Subsequently the model with best classification performance was used for predicting clinical outcome of MCI participants.ResultsEnsemble linear discriminant models using demographics and all volumetric magnetic resonance imaging measures as feature-set showed the best performance in classification of CN versus AD participants (accuracy = 92.8%, sensitivity = 95.8%, and specificity = 88.3%). Prediction accuracy of future conversion from MCI to AD for this ensemble linear discriminant at 6, 12, 24, 36, and 48 months was 63.8% (sensitivity = 74.4, specificity = 63.1), 68.9% (sensitivity = 75.9, specificity = 67.8), 74.9% (sensitivity = 71.5, specificity = 76.3), 75.3%, (sensitivity = 65.2, specificity = 79.7), and 77.0% (sensitivity = 59.6, specificity = 86.1), respectively.ConclusionsMachine learning models trained for classification of CN versus AD can improve our prediction ability of MCI conversion to AD.

Background: Posttraumatic stress disorder (PTSD) symptomatology and poorer pulmonary function are highly prevalent psychiatric and medical conditions. In the present study, we tested for the individual, additive, and modifying associations of PTSD symptomatology and pulmonary function with cognitive performance. Methods: In this cross‐sectional study, a total of 1,401 World Trade Center (WTC) responders (mean age = 53, SD = 8 years, 92% males) participated in the study. Cogstate assessment measured cognitive performance. PTSD symptomatology was measured using the trauma‐specific version of the posttraumatic stress disorder checklist (PCL‐17) adapted for the WTC attacks. The 1‐second forced expiratory volume and forced vital capacity (FEV1/FVC) ratio was used to measure pulmonary function. Linear regressions with cognitive performance as the outcome were conducted to assess individual, additive, and moderating associations of PTSD symptomatology and pulmonary function. Results: Higher PTSD symptomatology and poorer pulmonary function were negatively associated with cognitive performance. A 10% increase on the FEV1/FVC ratio moderated the association between PTSD symptomatology and cognition, whereby its association with cognition was stronger when PTSD symptomatology was higher (est. = 0.01, 95%CI = 0.004, 0.01, p < 0.001). When stratified by responder type, these associations persisted in trained (est. = 0.01, 95%CI = 0.01, 0.02, p < 0.001), but not in non‐trained (est. = 0.004, 95% C.I. = −0.01, 0.02, p = 0.39) responders. Conclusions: In the presence of higher PTSD, better pulmonary functioning is associated with better cognitive performance. Early intervention efforts to mitigate preventable cognitive decline in high‐risk populations should be studied, especially since intervention in one modality may have an impact on others. [ABSTRACT FROM AUTHOR]

In the Baltimore Longitudinal Study of Aging (BLSA), we examined the temporal unfolding of declining performance on tests of episodic memory (Free Recall on the Free and Cued Selective Reminding Test), executive function (Category Fluency, Letter Fluency, and Trails), and Verbal Intelligence (Nelson, 1982; American Version of the Nelson Adult Reading Test [AMNART]) before the diagnosis of dementia in 92 subjects with incident Alzheimer's disease (AD) followed for up to 15 years before diagnosis. To examine the preclinical onset of cognitive decline, we aligned subjects at the time of initial AD diagnosis and examined the cognitive course preceding diagnosis. We found that declines in performance on tests of episodic memory accelerated 7 years before diagnosis. Declining performance on tests of executive function accelerated 2-3 years before diagnosis, and verbal intelligence declined in close proximity to diagnosis. This cognitive profile is compatible with pathologic data suggesting that structures which mediate memory are affected earlier than frontal structures during the preclinical onset of AD. It also supports the view that VIQ as estimated by the AMNART does not decline during the preclinical onset of AD.

Rationale: Low FEV1 is a biomarker of increased mortality. The association of normal lung function and mortality is not well described. Objectives: To evaluate the FEV1–mortality association among participants with normal lung function. Methods: A total of 10,999 Fire Department of the City of New York (FDNY) responders and 10,901 Third National Health and Nutrition Examination Survey (NHANES III) participants, aged 18–65 years with FEV1 ⩾80% predicted, were analyzed, with FEV1 percent predicted calculated using Global Lung Function Initiative Global race-neutral reference equations. Mortality data were obtained from linkages to the National Death Index. Cox proportional hazards models estimated the association between FEV1 and all-cause mortality, controlling for age, sex, race/ethnicity, smoking history, and, for FDNY, work assignment. Cohorts were followed for a maximum of 20.3 years. Measurements and Main Results: We observed 504 deaths (4.6%) of 10,999 for FDNY and 1,237 deaths (9.4% [weighted]) of 10,901 for NHANES III. Relative to FEV1 ⩾120% predicted, mortality was significantly higher for FEV1 100–109%, 90–99%, and 80–89% predicted in the FDNY cohort. In the NHANES III cohort, mortality was significantly higher for FEV1 90–99% and 80–89% predicted. Each 10% higher predicted FEV1 was associated with 15% (hazard ratio, 0.85; 95% confidence interval, 0.80–0.91) and 23% (hazard ratio, 0.77; 95% confidence interval, 0.71–0.84) lower mortality for FDNY and NHANES III, respectively. Conclusions: In both cohorts, higher FEV1 is associated with lower mortality, suggesting higher FEV1 is a biomarker of better health. These findings demonstrate that a single cross-sectional measurement of FEV1 is predictive of mortality over two decades, even when FEV1 is in the normal range. [ABSTRACT FROM AUTHOR]