Assessing patient risk, benefit, and outcomes in drug development: A decade of ramucirumab clinical trials.

OBJECTIVE: This study aims to evaluate published clinical trials of ramucirumab to assess the risk/benefit profile and burden over time for patients. BACKGROUND: The burden of oncologic drug development on patients paired with increasing clinical trial failure rates emphasizes the need for reform of drug development. Identifying and addressing patterns of excess burden can guide policy, ensure evidence-based protections for trial participants, and improve medical decision-making. METHODS: On May 25, 2023 a literature search was performed on Pubmed/MEDLINE, Embase, Cochrane CENTRAL, and ClinicalTrials.gov for clinical trials using ramucirumab as monotherapy or in combination with other interventions for cancer treatment. Authors screened titles and abstracts for potential inclusion in a masked, duplicate fashion. Following data screening, data was extracted in a masked, duplicate fashion. Trials were classified as positive when meeting their primary endpoint and safety, negative or indeterminate. RESULTS: Ramucirumab was initially approved for gastric cancer but has since been tested in 20 cancers outside of its FDA approved indications. In our analysis of ramucirumab trials, there were a total of 10,936 participants and 10,303 adverse events reported. Gains in overall survival and progression-free survival for patients were 1.5 and 1.2 months, respectively. FDA-approved indications have reported more positive outcomes in comparison to off-label indications. CONCLUSION: We found that FDA-approved indications for ramucirumab had better efficacy outcomes than non-approved indications. However, a concerning number of adverse events were observed across all trials assessed. Participants in ramucirumab randomized controlled trials saw meager gains in overall survival when evaluated against a comparison group. Clinicians should carefully weigh the risks associated with ramucirumab therapy given its toxicity burden and poor survival gains.

On Smart Gaze Based Annotation of Histopathology Images for Training of Deep Convolutional Neural Networks.

Unavailability of large training datasets is a bottleneck that needs to be overcome to realize the true potential of deep learning in histopathology applications. Although slide digitization via whole slide imaging scanners has increased the speed of data acquisition, labeling of virtual slides requires a substantial time investment from pathologists. Eye gaze annotations have the potential to speed up the slide labeling process. This work explores the viability and timing comparisons of eye gaze labeling compared to conventional manual labeling for training object detectors. Challenges associated with gaze based labeling and methods to refine the coarse data annotations for subsequent object detection are also discussed. Results demonstrate that gaze tracking based labeling can save valuable pathologist time and delivers good performance when employed for training a deep object detector. Using the task of localization of Keratin Pearls in cases of oral squamous cell carcinoma as a test case, we compare the performance gap between deep object detectors trained using hand-labelled and gaze-labelled data. On average, compared to 'Bounding-box' based hand-labeling, gaze-labeling required 57.6% less time per label and compared to 'Freehand' labeling, gaze-labeling required on average 85% less time per label.

Pakistan Oral Cancer Collaborative: analyzing barriers and obstacles to oral cancer diagnosis, treatment, and prevention in Pakistan.

Oral cancer is a global health problem with increasing case numbers worldwide and no significant improvement in prognosis over the last few decades. It is one of the most common cancers and a leading cause of death in Pakistan, although the number reported is significantly underreported owing to the lack of a national cancer repository, and the true magnitude of this challenge is not known. Bilateral discussions and workshops funded by the Global Challenges Research Fund brought together a number of like-minded researchers and clinicians from the United Kingdom and Pakistan to analyze the status quo and plan the future course. This article reviews some of these discussions as well as barriers to oral cancer diagnosis in Pakistan and makes recommendations to investigate the magnitude and develop measures that may help tackle this devastating disease.

Automated feature detection in dental periapical radiographs by using deep learning.

OBJECTIVE: The aim of this study was to investigate automated feature detection, segmentation, and quantification of common findings in periapical radiographs (PRs) by using deep learning (DL)-based computer vision techniques. STUDY DESIGN: Caries, alveolar bone recession, and interradicular radiolucencies were labeled on 206 digital PRs by 3 specialists (2 oral pathologists and 1 endodontist). The PRs were divided into "Training and Validation" and "Test" data sets consisting of 176 and 30 PRs, respectively. Multiple transformations of image data were used as input to deep neural networks during training. Outcomes of existing and purpose-built DL architectures were compared to identify the most suitable architecture for automated analysis. RESULTS: The U-Net architecture and its variant significantly outperformed Xnet and SegNet in all metrics. The overall best performing architecture on the validation data set was "U-Net+Densenet121" (mean intersection over union [mIoU] = 0.501; Dice coefficient = 0.569). Performance of all architectures degraded on the "Test" data set; "U-Net" delivered the best performance (mIoU = 0.402; Dice coefficient = 0.453). Interradicular radiolucencies were the most difficult to segment. CONCLUSIONS: DL has potential for automated analysis of PRs but warrants further research. Among existing off-the-shelf architectures, U-Net and its variants delivered the best performance. Further performance gains can be obtained via purpose-built architectures and a larger multicentric cohort.

Brain tumor classification in MRI image using convolutional neural network.

Brain tumor is a severe cancer disease caused by uncontrollable and abnormal partitioning of cells. Recent progress in the field of deep learning has helped the health industry in Medical Imaging for Medical Diagnostic of many diseases. For Visual learning and Image Recognition, task CNN is the most prevalent and commonly used machine learning algorithm. Similarly, in our paper, we introduce the convolutional neural network (CNN) approach along with Data Augmentation and Image Processing to categorize brain MRI scan images into cancerous and non-cancerous. Using the transfer learning approach we compared the performance of our scratched CNN model with pre-trained VGG-16, ResNet-50, and Inception-v3 models. As the experiment is tested on a very small dataset but the experimental result shows that our model accuracy result is very effective and have very low complexity rate by achieving 100% accuracy, while VGG-16 achieved 96%, ResNet-50 achieved 89% and Inception-V3 achieved 75% accuracy. Our model requires very less computational power and has much better accuracy results as compared to other pre-trained models.

Prognostic Relevance of Cardiorespiratory Fitness as Assessed by Submaximal Exercise Testing for All-Cause Mortality: A UK Biobank Prospective Study.

OBJECTIVE: To investigate whether the inverse associations of cardiorespiratory fitness (CRF) with all-cause and cardiovascular mortality in the general population vary among individuals who are at different levels of pretest risk. PATIENTS AND METHODS: Cardiorespiratory fitness was assessed through submaximal bicycle tests in 58,892 participants aged 40 to 69 years who completed baseline questionnaires between January 1, 2006, and December 31, 2010, in the UK Biobank Prospective Study. Participants were categorized into risk categories, which determined allocation to an individualized bicycle protocol. The groups at minimal risk (category 1), small risk (category 2), and medium risk (category 3) were tested at 50%, 35% of the predicted maximal workload, and constant level, respectively. We investigated associations of CRF with mortality across different levels of pretest risk and determined whether CRF improves risk prediction. RESULTS: During a median follow-up of 5.8 years, 936 deaths occurred. Cardiorespiratory fitness was linearly associated with mortality risk. Comparing extreme fifths of CRF, the multivariable-adjusted hazard ratios (95% CIs) for mortality were 0.63 (0.52-0.77), 0.54 (0.36-0.82), 0.81 (0.46-1.43), and 0.58 (0.48-0.69) in categories 1, 2, and 3 and overall population, respectively. The addition of CRF to a 5-year mortality risk score containing established risk factors was associated with a C-index change (0.0012; P=.49), integrated discrimination improvement (0.0005; P<.001), net reclassification improvement (+0.0361; P=.005), and improved goodness of fit (likelihood ratio test, P<.001). Differences in 5-year survival were more pronounced across levels of age, smoking status, and sex. CONCLUSION: Cardiorespiratory fitness, assessed by submaximal exercise testing, improves mortality risk prediction beyond conventional risk factors and its prognostic relevance varies across cardiovascular risk levels.

Finnish sauna bathing does not increase or decrease the risk of cancer in men: A prospective cohort study.

INTRODUCTION: Evidence suggests that heat therapy can be used to prevent and treat cancer; anecdotal reports suggest passive heat therapies may increase cancer risk. Finnish sauna bathing has been linked to a reduced risk of chronic diseases, but its association with cancer risk is unknown. We aimed to assess the prospective association between frequency of sauna bathing and the risk of all-cause and site-specific cancers using the Kuopio Ischemic Heart Disease prospective cohort. METHODS: Baseline sauna bathing habits were assessed in 2173 men aged 42-61 years with no history of cancer. Hazard ratios (HRs) with 95% confidence intervals (CIs) for cancer were calculated using Cox proportional hazard models. We corrected for within-person variability in sauna bathing habits using data from repeat assessments taken 11 years apart. RESULTS: During a median follow-up of 24.3 years, 588 (27.1%) all-cause cancer cases were recorded. The age-adjusted regression dilution ratio of sauna bathing frequency was 0.69 (0.62-0.76). In multivariable-adjusted analyses, the HRs (95% CIs) of all-cause cancer were 0.92 (0.76-1.11) and 0.92 (0.66-1.27) for men who had 2-3 and ≥4 sauna sessions per week, respectively, compared with men who had ≤1 sauna session per week. The non-significant findings were consistent for prostate, gastrointestinal and lung cancers on multivariate adjustment. CONCLUSION: Frequent Finnish sauna bathing is not associated with the risk of cancer in a middle-aged male Caucasian population. Further studies are required to confirm or refute these findings, particularly in women and other age groups.

Cardiorespiratory Fitness and the Risk of Serious Ventricular Arrhythmias: A Prospective Cohort Study.

Cardiorespiratory fitness (CRF) is an established risk factor for cardiovascular disease outcomes. However, the relationship of CRF with risk of ventricular arrhythmias (VAs) is unknown. We aimed to assess the prospective association of CRF with the risk of serious VAs. Cardiorespiratory fitness, as measured by maximal oxygen uptake, was assessed using a respiratory gas exchange analyzer in 2299 middle-aged men in the Kuopio Ischemic Heart Disease prospective cohort. We corrected for within-person variability in CRF levels using data from repeated measurements 11 years apart. During median follow-up of 25.3 years (interquartile range, 18.7-27.2 years), 73 serious VAs were recorded. The age-adjusted regression dilution ratio of CRF was 0.58 (95% CI, 0.53-0.64). In analysis adjusted for age, the hazard ratio (HR) for serious VAs per 1-SD increase in CRF was 0.64 (95% CI, 0.49-0.84). The association persisted on additional adjustment for body mass index, systolic blood pressure, history of hypertension, prevalent coronary heart disease, smoking, history of diabetes, cholesterol level, alcohol consumption, and physical activity (HR, 0.67; 95% CI, 0.51-0.88). The corresponding adjusted HRs (95% CIs) were 0.29 (0.14-0.59) and 0.32 (0.15-0.65), respectively, comparing the top vs bottom tertiles. The associations were stronger on correction for regression dilution bias, remained consistent on exclusion of men with a history of coronary heart disease, and did not vary importantly in several relevant clinical subgroups. Cardiorespiratory fitness is inversely associated with future risk of serious VAs, independently of several cardiovascular disease risk factors. Further research is needed to assess the causal relevance of these findings.

Skin Lesion Classification Using GAN based Data Augmentation.

Early detection and frequent monitoring are critical for survival of skin cancer patients. Unfortunately, in practice a significant number of cases remain undetected until advanced stages, reducing the chances of survival. An appealing approach for early detection is to employ automated classification of dermoscopic images acquired via low-cost, smartphone-based hardware. By far, the most successful classification approaches on this task are based on deep learning. Unfortunately, most medical image classification tasks are unable to leverage the true potential of deep learning due to limited sizes of training datasets. Investigation of novel data generation techniques is thus an appealing option since it can enable us to augment our training data by a large number of synthetically generated examples. In this work, we investigate the possibility of obtaining realistic looking dermoscopic images via generative adversarial networks (GANs). These images are then employed to augment our existing training set in an effort to enhance the performance of a deep convolutional neural network on the skin lesion classification task. Results are compared with conventional data augmentation strategies and demonstrate that GAN based augmentation delivers significant performance gains.

Cancer Image Quantification With And Without, Expensive Whole Slide Imaging Scanners.

Automated analysis of digitized pathology images in tele-health applications can have a transformative impact on under-served communities in the developing world. However, the vast majority of existing image analysis algorithms are trained on slide images acquired via expensive Whole-Slide-Imaging (WSI) scanners. High scanner cost is a key bottleneck preventing large-scale adoption of digital pathology in developing countries. In this work, we investigate the viability of automated analysis of slide images captured from the eyepiece of a microscope via a smart phone. The mitosis detection application is considered as a use case.Results indicate performance degradation when using (lower-quality) smartphone images; as expected. However, the performance gap is not too wide (F1-score smartphone=0.65, F1-score WSI=0.70) demonstrating that smartphones could potentially be employed as image acquisition devices for digital pathology at locations where expensive scanners are not available.

Sleep Duration and Risk of Fatal Coronary Heart Disease, Sudden Cardiac Death, Cancer Death, and All-Cause Mortality.

BACKGROUND: Sleep duration has been shown to be associated with all-cause mortality; however, its relationship with cause-specific fatal events remains uncertain. We examined the relationship between sleep duration and risk of fatal coronary heart disease, sudden cardiac death, cancer-related death, and all-cause mortality. METHODS: Sleep duration was self-reported at baseline examinations performed between March 20, 1984, and December 5, 1989, in 2361 men aged 42-61 years from the Kuopio Ischemic Heart Disease study. Of these, 1734 (73.4%) men were free from coronary heart disease and cancer at baseline. RESULTS: A total of 802 all-cause deaths, 202 fatal coronary heart disease events, 141 sudden cardiac events, and 229 cancer-related deaths were reported during a median follow-up of 25.9 (interquartile range, 20.6-28.2) years. Multivariable adjusted hazard ratios comparing the top quartile (>10 hours) of sleep duration vs the bottom quartile (<8 hours) was 1.19 (95% confidence interval [CI], 1.01-1.43) for all-cause mortality, 1.27 (95% CI, 0.88-1.84) for fatal coronary heart disease, 1.20 (95% CI, 0.78-1.86) for sudden cardiac death, and 1.29 (95% CI, 0.92-1.80) for cancer death. No differences in association of sleep duration with outcomes were found in clinically relevant subgroups, including age, history of coronary heart disease, body mass index, physical activity, and C-reactive protein levels. CONCLUSIONS: Longer duration of sleep was associated with significantly increased all-cause mortality. The mechanistic link between these findings remains to be explored further.

Long-Term Change in Cardiorespiratory Fitness in Relation to Atrial Fibrillation and Heart Failure (from the Kuopio Ischemic Heart Disease Risk Factor Study).

The benefits of aerobic fitness in relation to all-cause and cardiovascular mortality is well established; however, the associations of long-term change in cardiorespiratory fitness (CRF) with incident heart failure (HF) and atrial fibrillation (AF) have not been studied before. The Kuopio Ischaemic Heart Disease Risk Factor Study is a prospective cohort comprising men aged 42 to 60 years from the city of Kuopio and its surroundings, with a baseline examination between 1984 and 1989 (V1), a re-examination at 11 years (V2), and up to 15 years of follow-up from V2. CRF, as assessed by VO2max, was measured at both visits using respiratory gas exchange during maximal exercise tolerance test. The difference (ΔVO2max) was estimated as VO2max (V2) - VO2max (V1). Participants with no missing data on both baseline and 11-year exercise test were included (n = 481). The mean ΔVO2max was -5.4 ml/min⋅kg (standard deviation 5.4). During a median follow-up of 14.3 years (interquartile range 13.3 to 15.1), 46 incident HF (9.6%) and 73 incident AF (15.2%) events were recorded. In a multivariate analysis adjusted for baseline age, baseline VO2max, systolic blood pressure, smoking, type 2 diabetes, and cardiovascular disease, per 1 ml/min⋅kg higher ΔVO2max was log linearly associated with incident HF with a 10% relative risk reduction of HF (hazard ratio 0.90, 95% confidence interval 0.83 to 0.97). No significant relation of ΔVO2max with incident AF was observed. In conclusion, overall long-term improvement in CRF is associated with reduced risk of HF, indicating the importance of maintaining good CRF over time.

Physical activity classification using time-frequency signatures of motion artifacts in multi-channel electrical impedance plethysmographs.

Physical activities are known to introduce motion artifacts in electrical impedance plethysmographic (EIP) sensors. Existing literature considers motion artifacts as a nuisance and generally discards the artifact containing portion of the sensor output. This paper examines the notion of exploiting motion artifacts for detecting the underlying physical activities which give rise to the artifacts in question. In particular, we investigate whether the artifact pattern associated with a physical activity is unique; and does it vary from one human-subject to another? Data was recorded from 19 adult human-subjects while conducting 5 distinct, artifact inducing, activities. A set of novel features based on the time-frequency signatures of the sensor outputs are then constructed. Our analysis demonstrates that these features enable high accuracy detection of the underlying physical activity. Using an SVM classifier we are able to differentiate between 5 distinct physical activities (coughing, reaching, walking, eating and rolling-on-bed) with an average accuracy of 85.46%. Classification is performed solely using features designed specifically to capture the time-frequency signatures of different physical activities. This enables us to measure both respiratory and motion information using only one type of sensor. This is in contrast to conventional approaches to physical activity monitoring; which rely on additional hardware such as accelerometers to capture activity information.

Cardiorespiratory fitness and nonfatalcardiovascular events: A population-based follow-up study.

BACKGROUND: To examine the prognostic value of cardiorespiratory fitness (CRF) with risk of first major nonfatal myocardial infarction (MI), stroke, and heart failure (HF) events. METHODS: Cardiorespiratory fitness, as measured by maximal oxygen uptake, was assessed at baseline in a prospective cohort of 2,089 men aged 42 to 61years. RESULTS: During a mean (SD) follow-up of 19.1(8.4) years, 522 nonfatal acute MI events, 198 acute all-cause nonfatal stroke events, and 221 nonfatal HF events were recorded. The hazard ratio per 1-metabolic-equivalent increase in CRF was 0.93 (95% CI 0.88-0.97) for nonfatal MI, 0.94 (95% CI0.87-1.01) for nonfatal stroke, and 0.84 (95% CI 0.78-0.91) for nonfatal HF events after adjustment for cardiovascular risk factors (age, systolic blood pressure, body mass index, history of cardiovascular disease, diabetes, smoking, alcohol use, serum creatinine, low-density lipoprotein levels, physical activity, and socioeconomic status). Further adjustment for left ventricular hypertrophy and resting heart rate did not attenuate these associations. Addition of CRF to conventional cardiovascular disease risk factors significantly improved both discrimination (C index) and category free net reclassification index (cf-NRI) for nonfatal MI (change in C index, 0.015 [95% CI 0.010-0.020] and change in cf-NRI 0.27, P<.01) and HF (change in C index 0.040 [95% CI 0.010-0.060] and change in cf-NRI 0.88, P<.01). CONCLUSION: In this Finnish population, there is a strong, inverse, and independent association between CRF and acute nonfatal MI and HF risk.

Long-term Change in Cardiorespiratory Fitness and All-Cause Mortality: A Population-Based Follow-up Study.

Few studies have investigated long-term changes in cardiorespiratory fitness (CRF), defined by indirect measures of CRF, and all-cause mortality. We aimed to investigate whether long-term change in CRF, as assessed by the gold standard method of respiratory gas exchange during exercise, is associated with all-cause mortality. A population-based sample of 579 men aged 42 to 60 years with no missing data at baseline examination (V1) and at reexamination at 11 years (V2) were included. Maximal oxygen uptake (VO2max) was measured at both visits using respiratory gas exchange during maximal exercise testing, and the difference (ΔVO2max) was calculated as VO2max (V2) - VO2max (V1). Deaths were ascertained annually using national death certificates during 15 years of follow-up after V2. The mean ΔVO2max was -5.2 mL/min*kg. During median follow-up of 13.3 years (interquartile range, 12.5-14.0 years), 123 deaths (21.2%) were recorded. In a multivariate analysis adjusted for baseline age, VO2max, systolic blood pressure, smoking status, low- and high-density lipoprotein cholesterol and triglyceride levels, C-reactive protein level, body mass index, alcohol consumption, physical activity, socioeconomic status, and history of type 2 diabetes mellitus and ischemic heart disease, a 1 mL/min*kg higher ΔVO2max was associated with a 9% relative risk reduction of all-cause mortality (hazard ratio, 0.91; 95% CI, 0.87-0.95). This study suggested that in this population, long-term CRF reduction was associated with an increased risk of mortality, emphasizing the importance of maintaining good CRF over the decades.

Is lipoprotein (a) protective of dementia?

Lipoprotein(a) [Lp(a)]-an established risk factor for vascular disease, has been suggested to be associated with risk of dementia, however no prospective evidence exists to support this finding. We aimed to assess the association of Lp(a) with dementia risk. Lp(a) concentration was assessed at baseline in a prospective cohort of 2532 men aged 42-61 years. During a median follow-up of 24.9 years, 228 new cases of dementia were recorded. Lp(a) was approximately log-linearly associated with dementia risk. In age-adjusted analysis, the hazard ratio for dementia in a comparison of extreme quartiles of Lp(a) levels was 0.68 (95 % CI: 0.47-0.99), which persisted after adjustment for several physical measures, history of coronary heart disease, smoking status, history of diabetes, serum lipids, alcohol consumption, and socio-economic status 0.68 (0.46-0.99). Lp(a) is protective of future dementia risk in a middle-aged male Caucasian population. Further research is needed replicate these findings.

A cross-sectional study of the prevalence and correlates of tobacco use in Chennai, Delhi, and Karachi: data from the CARRS study.

BACKGROUND: Tobacco burdens in India and Pakistan require continued efforts to quantify tobacco use and its impacts. We examined the prevalence and sociodemographic and health-related correlates of tobacco use in Delhi, Chennai (India), and Karachi (Pakistan). METHODS: Analysis of representative surveys of 11,260 participants (selected through multistage cluster random sampling; stratified by gender and age) in 2011 measured socio-demographics, tobacco use history, comorbid health conditions, and salivary cotinine. We used bivariate and multivariate regression analyses to examine factors associated with tobacco use. RESULTS: Overall, 51.8 % were females, and 61.6 % were below the age of 45 years. Lifetime (ever) tobacco use prevalence (standardized for world population) was 45.0 %, 41.3 %, and 42.5 % among males, and 7.6 %, 8.5 %, and 19.7 % among females in Chennai, Delhi, and Karachi, respectively. Past 6 month tobacco use prevalence (standardized for world population) was 38.6 %, 36.1 %, and 39.1 % among males, and 7.3 %, 7.1 %, and 18.6 % among females in Chennai, Delhi, and Karachi, respectively. In multivariable regression analyses, residing in Delhi or Karachi versus Chennai; older age; lower education; earning less income; lower BMI; were each associated with tobacco use in both sexes. In addition, semi-skilled occupation versus not working and alcohol use were associated with tobacco use in males, and having newly diagnosed dyslipidemia was associated with lower odds of tobacco use among females. Mean salivary cotinine levels were higher among tobacco users versus nonusers (235.4; CI: 187.0-283.8 vs. 29.7; CI: 4.2, 55.2, respectively). CONCLUSION: High prevalence of tobacco use in the South Asian region, particularly among men, highlights the urgency to address this serious public health problem. Our analyses suggest targeted prevention and cessation interventions focused on lower socioeconomic groups may be particularly important.

Cardiorespiratory fitness and atrial fibrillation: A population-based follow-up study.

BACKGROUND: Atrial fibrillation (AF) is the most common arrhythmia worldwide and has a complex association with physical fitness. The relationship of cardiorespiratory fitness (CRF) with the risk for AF has not been previously investigated in population-based studies. OBJECTIVE: The purpose of this study was to determine the relationship of CRF with incident AF. METHODS: CRF, as assessed by maximal oxygen uptake (VO2max) during exercise testing, was measured at baseline in 1950 middle-aged men (mean age 52.6 years, SD 5.1) from the Kuopio Ischaemic Heart Disease (KIHD) study. RESULTS: During average follow-up of 19.5 years, there were 305 incident AF cases (annual AF rate of 65.1/1000 person-years, 95% confidence interval [CI] 58.2-72.8). Overall, a nonlinear association was observed between CRF and incident AF. The rate of incident AF varied from 11.5 (95% CI 9.4-14.0) for the first quartile of CRF, to 9.1 (95% CI 7.4-11.2) for the second quartile, 5.7 (95% CI 4.4-7.4) for the third quartile, and 6.3 (95% CI 5.0-8.0) for the fourth quartile. Age-adjusted hazard ratio comparing top vs bottom fourth of usual CRF levels was 0.67 (95% CI 0.48-0.95), attenuated to 0.98 (95% CI 0.66-1.43) upon further adjustment for risk factors. These findings were comparable across age, body mass index, history of smoking, diabetes, and cardiovascular disease status at baseline. CONCLUSION: Improved fitness as indicated by higher levels of CRF is protective of AF within a certain range, beyond which the risk of AF rises again. These findings warrant further replication.