Association of Obesity With Mortality Over 24 Years of Weight History

Xu, Hanfei; Cupples, L. Adrienne; Stokes, Andrew; Liu, Ching‐Ti

doi:10.1001/jamanetworkopen.2018.4587

Cited by 138 publications

(101 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…These interpretation maps are consistent with 7/14 prior knowledge that obese patients often have challenging image acquisition 31,32 , however it is surprising the degree of precision it brings to predicting height and weight. Prediction of these systemic phenotypes suggest that imaging based predictions of mortality and life expectancy could have high predictive value as systemic phenotypes of age, sex, and body mass index are highly correlated with cardiovascular outcomes and overall life expectancy [33][34][35] .…”

Section: Discussionmentioning

confidence: 99%

Deep Learning Interpretation of Echocardiograms

Ghorbani

Ouyang

Abid

et al. 2019

Preprint

View full text Add to dashboard Cite

Echocardiography uses ultrasound technology to capture high temporal and spatial resolution images of the heart and surrounding structures and is the most common imaging modality in cardiovascular medicine. Using convolutional neural networks on a large new dataset, we show that deep learning applied to echocardiography can identify local cardiac structures, estimate cardiac function, and predict systemic phenotypes that modify cardiovascular risk but not readily identifiable to human interpretation. Our deep learning model, EchoNet, accurately identified the presence of pacemaker leads (AUC = 0.89), enlarged left atrium (AUC = 0.85), normal left ventricular wall thickness (AUC = 0.75), left ventricular end systolic and diastolic volumes(R 2 = 0.73 and R 2 = 0.68), and ejection fraction (R 2 = 0.48) as well as predicted systemic phenotypes of age (R 2 = 0.46), sex (AUC = 0.88), weight (R 2 = 0.56), and height (R 2 = 0.33). Interpretation analysis validates that EchoNet shows appropriate attention to key cardiac structures when performing human-explainable tasks and highlight hypothesis-generating regions of interest when predicting systemic phenotypes difficult for human interpretation. Machine learning on echocardiography images can streamline repetitive tasks in the clinical workflow, standardize interpretation in areas with insufficient qualified cardiologists, and more consistently produce echocardiographic measurements.Related works Current literature have already shown that it is possible to identify standard echocardiogram views from unlabeled datasets. 5,6,24 Previous works have used convolutional neural networks (CNNs) trained on images and videos from echocardiography to perform segmentation to identify cardiac structures and derive cardiac function. In this study, we extend previous analyses to show that EchoNet, our deep learning model using echocardiography images, local cardiac structures and anatomy, estimate volumetric measurements and metrics of cardiac function, and predict systemic human phenotypes that modify cardiovascular risk. Additionally, we show the first application of interpretation frameworks to understand deep learning models from echocardiogram images. Human identifiable features, such as the presence of pacemaker and defibrillator leads, left ventricular hypertrophy, and abnormal left atrial chamber size identified by our convolutional neural network were validated using interpretation frameworks to highlight the most relevant regions of interest. To the best of our knowledge, we develop the first deep learning model that can directly predict age, sex, weight and height from echocardiogram images and use interpretation methods to understand how the model predicts these systemic phenotypes difficult for human interpreters. 2/14Lessons from model training and experiments EchoNet performance greatly improved with efforts to augment data size, homogenize input data, and with optimize model training with hyperparameter search. Our experience shows that increasing number of unique patien...

show abstract

Section: Discussionmentioning

confidence: 99%

Deep Learning Interpretation of Echocardiograms

Ghorbani

Ouyang

Abid

et al. 2019

Preprint

View full text Add to dashboard Cite

show abstract

“…Compared with those maintaining normal weight over 16 years, those with maximum BMI in the overweight, obesity class I (BMI 30‐35 kg/m 2 ), and obesity class II (BMI ≥ 35 kg/m 2 ) categories had significantly and progressively increased risk for all‐cause mortality, CVD mortality, stroke mortality, cancer mortality, and respiratory disease mortality. Xu et al investigated the association between maximum BMI and all‐cause mortality among participants from the original and offspring Framingham Heart Studies for a combined sample of 6,197 participants. A total of 77.3% had a history of overweight or obesity using the participant’s maximum BMI, while 66.5% had overweight or obesity based on baseline BMI.…”

Section: Question 1: What Is the Evidence Regarding The Importance Ofmentioning

confidence: 99%

Weight History in Clinical Practice: The State of the Science and Future Directions

Kushner

Batsis

Butsch

et al. 2019

Obesity

View full text Add to dashboard Cite

Eliciting a weight history can provide clinically important information to aid in treatment decision‐making. This view is consistent with the life course perspective of obesity and the aim of patient‐centered care, one of six domains of health care quality. However, thus far, the value and practicality of including a weight history in the clinical assessment and treatment of patients with obesity have not been systematically explored. For these reasons, the Clinical Committee of The Obesity Society established a task force to review and assess the available evidence to address five key questions. It is concluded that weight history is an essential component of the medical history for patients presenting with overweight or obesity, and there are strong and emerging data that demonstrate the importance of life stage, duration of exposure to obesity, maximum BMI, and group‐based trajectory modeling in predicting risk for increased morbidity and mortality. Consideration of these and other patient‐specific factors may improve risk stratification and clinical decision‐making for screening, counseling, and management. Recommendations are provided for the key elements that should be included in a weight history, and several needs for future clinical research are outlined.

show abstract

“…Clinical and population health guidelines define obesity as BMI ≥ 30 kg/m 2 . BMI has been used as a population-based tool to assess mortality and cardiovascular risk (7)(8)(9)(10)(11). However, BMI has limitations in clinical use that are well known.…”

Section: Introductionmentioning

confidence: 99%

Comparing the Ability of Two Comprehensive Clinical Staging Systems to Predict Mortality: EOSS and CMDS

Ejima

Xavier

Mehta

2020

Obesity

View full text Add to dashboard Cite

Objective Differences in discriminative and predictive ability for all‐cause mortality of two clinical staging systems, the Edmonton Obesity Staging System (EOSS) and Cardiometabolic Disease Staging (CMDS), were estimated. Methods Data for nonpregnant persons aged 40 to 75 years were extracted from the National Health and Nutrition Examination Survey. Predictive and discriminative ability was assessed using pseudo‐R2 and C‐statistics. Median years of life lost were also computed for each score. Results Differences in out‐of‐sample estimates of pseudo‐R2 and C‐statistics (EOSS model as reference) were 0.02 (95% CI: 0.01‐0.04) (Kent pseudo‐R2), 0.03 (0.01‐0.04) (Royston pseudo‐R2), and 0.02 (0.01‐0.02) (C‐statistics). The median years of life lost for EOSS scores 2 and 3 (low to high risk) for a reference person were 1.19 and 6.76 years. Those for CMDS scores 1, 2, 3, and 4 (low to high risk) were 1.53, 2.90, 3.91, and 8.51 years. Consistent results from the in‐sample estimates were observed. Conclusions CMDS had statistically significantly greater predictive and discriminative ability than EOSS for persons aged 40 to 75. While the clinical relevance of these differences is unknown, CMDS may have greater clinical utility given that it uses fewer items to risk stratify. The clinical relevance and utility need to be studied further.

show abstract

Association of Obesity With Mortality Over 24 Years of Weight History

Cited by 138 publications

References 47 publications

Deep Learning Interpretation of Echocardiograms

Deep Learning Interpretation of Echocardiograms

Weight History in Clinical Practice: The State of the Science and Future Directions

Comparing the Ability of Two Comprehensive Clinical Staging Systems to Predict Mortality: EOSS and CMDS

Contact Info

Product

Resources

About