Synopsis:
Cross-sectional neuroimaging studies have indicated hemispheric asymmetry, gender differences and a decline in brain volume with advancing age. However, these studies often have a small sample size (≤ 500) and unbalanced groupings, resulting in contradictory findings that might not be reproducible. Precise estimates on a large cohort will facilitate the assessment and quantification of brain volume, resulting in improved diagnostics and establishing statistical norms. In this study we analyze 6739 individuals to draw normative conclusions about typical patterns of age-related brain changes. We provide whole-brain and regional volume aging-curves for males and females for raw and head-size adjusted data.
Summary:
Different ROIs decrease at different rates, which can be linear or nonlinear. Hemispheric asymmetry increases with aging. After 50 years, intracranial volume declines. In deep brain regions, genders show similar hemispheric asymmetry except for the putamen.
Introduction and Purpose:
Quantitative assessment of whole and regional volumes of the brain will improve our knowledge of brain atrophy with normal aging; this information may be helpful for distinguishing atrophy patterns associated with disease from age-related decline. Several studies have investigated the effects of age and gender on brain anatomy, however, the results are inconsistent and the topic continues to be the subject of active research 1-6. The accuracy of prior published brain-wide correlation studies have been questioned based on the limitations from the use of small sample sizes and the fact that such studies require thousands of participants for higher reproducibility and generalizability7. Therefore, in order to extract better correlations with age and gender on brain volume change, we performed a quantitative analysis of the brain volume on a large dataset of 6739 healthy individuals (age range: 30–80 years, 49% female), who underwent a T1-weighted scan as part of a preventative whole-body MRI (WB-MRI) screening program. Moreover, we evaluate the effects of raw and ICV(intracranial volume)-adjusted volumetric data on statistical analysis.
Materials and Methods:
A total of 6739 healthy participants, from 4 sites geographically spread across North America, were scanned on Siemens 1.5T MRI scanners as part of a preventative health screening program. 3D T1-weighted brain MRI (1mm isotropic), performed as part of the WB-MRI protocol, served as substrate in the present study to quantify the respective brain volumetrics by application of AI-solutions trained for this purpose. We used a deep learning framework - FastSurferCNN8 to segment the brain into 96 distinct regions. FastSurfer is trained over 134 participants aged 27-66. All MRI brain volumes were conformed to a standard slice orientation and resolution before feeding to the deep learning networks. To correct for the differences in the head size of the subjects, a separate deep learning model - nnUNet9- is trained on 60 scans to segment the ICV. All regional volumetric variables segmented by Fastsurfer were normalized for individual variations in ICV by dividing the raw values by ICV.
To analyze the volume loss and quantiles by age, after separating the female and male populations, we used a sliding-window method to integrate data around one year younger and older of the window's centering age. The analysis was performed per decade. The rate of decline in every decade is determined by measuring the slope of the robust regression line between targeted volume and age, controlling the effect of ICV. With the slope of the regression line and the corresponding mean of the targeted volume from the table, the reduction rate based on the percentage of ROI or total brain volume per decade can be calculated. The normative percentile of the healthy population at each age can be determined using the baseline mean, standard deviations, and involution rates. We compared the raw and ICV-corrected (by dividing a raw volume by ICV) volumetric values across genders and hemispheres using a two-tailed test.
Results and Discussion:
According to table 1, Intracranial volume increases throughout early adulthood and decreases during the fifth decade of life, and in each age group tested, males have a larger ICV, total and regional brain volume than women. However, women have greater ICV-corrected total and regional brain volumes, and the rate of volume loss is more noticeable in males than females in the majority of regions, and it is not linear. The brainstem and Cerebellum WM change minimally with age. Cerebral white matter volume increases up to the age-group of 40-50 in females, then decreases. Lateral ventricle shrinkage accelerates after 60 years in women and 50years in men and the relationship is non-linear.
From figure 1, we can compare the left and right ROIs before and after the normalization for males and females. For the majority of the ROIs, there is a small asymmetry between the left and right of ROIs but it is statically-significant (p-value <0.01). In the amygdala, caudate, and hippocampus, the right hemisphere has a larger volume than the left; The discrepancy with putamen is minimal and statistically insignificant in males, whereas in females, it is larger and statistically significant. Further, the right hemisphere has a lesser volume in the pallidum, thalamus, somatic and motor cortex. ICV-corrected ROIs display similar trends as before with reduced left-right discrepancy. Table 2 shows the reduction rate for different RoIs for male and female and left and right hemispheres. Table 3 shows a general trend of increasing asymmetry between the left and right hemispheres with age. The asymmetric analysis demonstrates a weak asymmetry between the thalamus, hippocampus, pallidum and the putamen. The caudate, and the primary motor cortex are moderately asymmetric. The amygdala and the primary sensory cortex are strongly asymmetric. According to table 4, for gray matter, the frontal lobes shrinks faster than all the others, while the parietal and temporal lobe shrinks at a relatively slower rate and the occipital lobe has the slowest reduction rate. These rates are nonlinear over decades.
Conclusion:
We investigated and contrasted the effects of age and gender in healthy brains across adulthood. Also, we reported volumetric differences between the left and right subcortical hemispheres. This large-scale investigation may explain contradictions and explain sex-specific distinctions in brain anatomy, cognition and behavior.
Acknowledgements:
We appreciate the assistance of the MRI Technologists, Backend, and Patient Care teams at Prenuvo with the data acquisition process.
References:
- Giorgio A, Watkins KE, Chadwick M, James S, Winmill L, Douaud G, de Stefano N, Matthews PM, Smith SM, Johansen-Berg H, James AC (2010) Longitudinal changes in grey and white matter during adolescence. Neuroimage 49(1):94–103
- Matsuda H (2013) Voxel-based morphometry of brain MRI in Normal aging and Alzheimer’s disease. Aging Dis 4(1):29–37
- Giorgio A, Santelli L, Tomassini V, Bosnell R, Smith S, de Stefano N, Johansen-Berg H (2010) Age-related changes in grey and white matter structure throughout adulthood. Neuroimage 51(3):943–951
- Smith CD, Chebrolu H, Wekstein DR, Schmitt FA, Markesbery WR (2007) Age and gender effects on human brain anatomy: a voxel-based morphometric study in healthy elderly. Neurobiol Aging 28(7):1075–1087
- Beheshti, I., Maikusa, N., & Matsuda, H. (2019). Effects of aging on brain volumes in healthy individuals across adulthood. Neurological Sciences, 40(6), 1191-1198.
- Ge Y, Grossman R, Babb J (2002) Age-related total gray matter and white matter changes in normal adult brain. Part I: volumetric MR imaging analysis. Am J Dent 23:1327–1333
- Overman, D. (2022). For Accuracy, Brain Studies of Complex Behavior Require Thousands of People. AXIS Imaging News.
- Henschel, L., Conjeti, S., Estrada, S., Diers, K., Fischl, B., & Reuter, M. (2020). Fastsurfer-a fast and accurate deep learning based neuroimaging pipeline. NeuroImage, 219, 117012.
- Isensee, F., Petersen, J., Kohl, S. A., Jäger, P. F., & Maier-Hein, K. H. (2019). nnu-net: Breaking the spell on successful medical image segmentation. arXiv preprint arXiv:1904.08128, 1(1-8), 2.
