No abstract

With age, the prevalence of diseases such as fatty liver disease, cirrhosis, and type two diabetes increases. Approaches to both predict abdominal age and identify risk factors for accelerated abdominal age may ultimately lead to advances that will delay the onset of these diseases. We build an abdominal age predictor by training convolutional neural networks to predict abdominal age (or “AbdAge”) from 45,552 liver magnetic resonance images [MRIs] and 36,784 pancreas MRIs (R-Squared = 73.3 ± 0.6; mean absolute error = 2.94 ± 0.03 years). Attention maps show that the prediction is driven by both liver and pancreas anatomical features, and surrounding organs and tissue. Abdominal aging is a complex trait, partially heritable (h_g2 = 26.3 ± 1.9%), and associated with 16 genetic loci (e.g. in PLEKHA1 and EFEMP1), biomarkers (e.g body impedance), clinical phenotypes (e.g, chest pain), diseases (e.g. hypertension), environmental (e.g smoking), and socioeconomic (e.g education, income) factors.

Heart disease is the first cause of death after age 65 and, with the world population aging, its prevalence is expected to starkly increase. We used deep learning to build a heart age predictor on 45,000 heart magnetic resonance videos [MRI] and electrocardiograms [ECG] from the UK Biobank cohort (age range 45-81 years). We predicted age with a root mean squared error [RMSE] of 2.81+/-0.02 years (R-Squared=85.6+/-0.2%) and found that accelerated heart aging is heritable at more than 35%. MRI-based anatomical features predicted age better than ECG-based electro-physiological features (RMSE=2.89+/-0.02 years vs. 6.09+/-.0.02 years), and heart anatomical and electrical aging are weakly correlated (Pearson correlation=.249+/-.002). Our attention maps highlighted the aorta, the mitral valve, and the interventricular septum as key anatomical features driving heart age prediction. We identified genetic (e.g titin gene) and non-genetic correlates (e.g smoking) of accelerated heart aging.

It is hypothesized that there are inter-individual differences in biological aging; however, differences in aging among (heart images vs. electrophysiology) and across (e.g., brain vs heart) physiological dimensions have not been systematically evaluated and compared. We analyzed 676,787 samples from 502,211 UK Biobank participants aged 37-82 years with deep learning approaches to build a total of 331 chronological age predictors on different data modalities such as videos (e.g. heart magnetic resonance imaging [MRI]), images (e.g. brain, liver and pancreas MRIs, full body X-rays, eye fundus and optical coherence tomography [OCT] images, carotid ultrasound images), time-series (e.g. electrocardiograms [ECGs], pulse wave analysis, wrist accelerometer data) and scalar data (e.g. blood biomarkers, anthropometric measures) to characterize the multiple dimensions of aging. We combined these age predictors into 11 main aging dimensions, 31 subdimensions and 84 sub-subdimensions ensemble models based on specific organ systems. Some of the organ systems were highly predictive of age. For example, heart features predict chronological age with a testing root mean squared error (RMSE) and standard error of 2.83+/-0.04 years, and musculoskeletal features predict age with a RMSE of 2.65+/-0.04 years. We defined "accelerated" agers as participants whose predicted age was greater than their chronological age and computed the correlation between these different definitions of accelerated aging. We found that most aging dimensions are largely uncorrelated (average correlation=.139+/-.090) but that dimensions that are biologically related tend to be more positively correlated. For example, we found that heart anatomical (from MRI) accelerated aging and heart electrical (from ECG) accelerated aging are correlated (average Pearson of .249+/-.005). Likewise, we found that most aging dimensions are genetically largely uncorrelated (average correlation=.104+/-.149). For example, heart anatomical and electrical accelerated aging are genetically .508+/-.089 correlated (r_g). We identified 9,697 SNPs in 3,318 genes associated with accelerated aging in at least one aging dimension and found an average GWAS-based heritability for accelerated aging of 26.1+/-7.42% (e.g. heart aging: 35.2+/-1.6%). Finally, we identified biomarkers, clinical phenotypes, diseases, family history, environmental variables and socioeconomic variables associated with accelerated aging in each aging dimension and computed the correlation between the different aging dimensions in terms of these associations. We found that environmental and socioeconomic variables are similarly associated with accelerated aging across aging dimensions (average correlations of respectively .639+/-.180 and .607+/-.309). We also built multivariate models to predict accelerated aging as a function of these same variables. For example, we predicted accelerated pulse wave analysis-measured arterial aging from diet with a R2 of 9.1%. Overall, most dimensions of aging are complex traits with both genetic and non-genetic correlates. These dimensions are weakly correlated with each other, highlighting the multidimensionality of the aging process and the need for multi-dimensional biological age predictors. Our results can be interactively explored on the following website: https://www.multidimensionality-of-aging.net/