Disentangling the Effect of Brain Size from Sex Differentiated Aging Trajectories
Poster No:
1196
Submission Type:
Abstract Submission
Authors:
Aliza Brzezinski-Rittner1,2, Mahsa Dadar1,2, Yashar Zeighami1,2
Institutions:
1McGill University, Montreal, Quebec, 2Douglas Mental Health University Health Centre, Montreal, Quebec
First Author:
Aliza Brzezinski-Rittner
McGill University|Douglas Mental Health University Health Centre
Montreal, Quebec|Montreal, Quebec
McGill University|Douglas Mental Health University Health Centre
Montreal, Quebec|Montreal, Quebec
Co-Author(s):
Mahsa Dadar
McGill University|Douglas Mental Health University Health Centre
Montreal, Quebec|Montreal, Quebec
McGill University|Douglas Mental Health University Health Centre
Montreal, Quebec|Montreal, Quebec
Yashar Zeighami
McGill University|Douglas Mental Health University Health Centre
Montreal, Quebec|Montreal, Quebec
McGill University|Douglas Mental Health University Health Centre
Montreal, Quebec|Montreal, Quebec
Introduction:
The aging process leads to loss of cortical and subcortical gray and white matter volumes, and these changes are differentiated between females and males (Bethlehem et al., 2022). Although there are general aging trajectories, different biological and environmental factors lead to distinct outcomes, and sex is one of these factors. However, many volumetric sex differences found in the brain are induced by differences in brain size and disappear once adjusting for intracranial volume (Sanchis-Segura et al., 2020). In this work, we seek to disentangle the effects of sex from those derived by allometric differences in regional aging trajectories as estimated by deformation based morphometry (DBM), and volumetric data from the UK Biobank (UKBB) dataset.
Methods:
Average cortical and subcortical DBM values (Zeighami et al., 2015) were extracted based on Schaeffer 1000 (Schaefer et al., 2018) and Xiao (Xiao et al., 2019)atlases, respectively. Precomputed FreeSurfer-based volumetric information was provided by the UKBB based on the DKT atlas (Desikan et al., 2006). Following visual quality control, data from 35,752 participants was included. We then created an age and total intracranial volume (TIV) matched sample of males and females (with a threshold of maximum one month of age difference at acquisition time and 0.2% difference in TIV) of 13,200 participants (referred to as matched sample), another sample of the same size matched just by age (age matched sample), and a non-matched sample of the same size (non matched sample). We modeled the sex differentiated aging trajectories for all the regions in each of the samples. We examined how the model estimates for different statistical contrasts change between these matching strategies and which coefficients were significant before and after multiple comparisons corrections for the different samples. We tested two different linear models to model aging trajectories. Model one tested the linear interaction between age and sex (ROI ~ 1 + age * sex), while the second model also included a quadratic interaction (ROI ~ 1 + (age + age2) * sex). A series of paired t-tests were used to assess the significance of the differences in model estimates for each sample.
Results:
For the DBM data, overall, model one fit the regional sex differentiated aging trajectories best according to the Akaike information criterion (AIC). We found that after multiple comparisons correction, the proportion of regions in which sex showed a significant effect in their aging trajectories was smaller in the matched sample than in the others (8.1% for the non matched sample, 3.2% for the age match sample and only 0.8% for the matched sample for the interaction between sex and age) (Fig. 1). Interestingly, only one subcortical region showed a significant effect of sex in the matched sample, namely the right nucleus accumbens. Model estimates were significantly different between the samples (Fig. 2, A).
For the Freesurfer volumetric data, overall, model two fit the regional sex differentiated aging trajectories best according to the AIC. Model estimates were significantly different among samples (Fig. 2, B).
For both DBM and the volumetric models, the linear effect of age was negative for all samples, showing a general trend of loss of gray matter tissue. The main effect of sex was very pronounced in the non-matched sample and even in the age-matched sample, however, after matching for age and TIV, this effect diminished and in some cases it disappeared altogether. Furthermore, in the case of the volumetric data, there was a sex differentiated effect of the quadratic term for age, however, in the matched sample, this trend diminished greatly, while the main effect of the quadratic term of age became stronger (Fig. 2).
For the Freesurfer volumetric data, overall, model two fit the regional sex differentiated aging trajectories best according to the AIC. Model estimates were significantly different among samples (Fig. 2, B).
For both DBM and the volumetric models, the linear effect of age was negative for all samples, showing a general trend of loss of gray matter tissue. The main effect of sex was very pronounced in the non-matched sample and even in the age-matched sample, however, after matching for age and TIV, this effect diminished and in some cases it disappeared altogether. Furthermore, in the case of the volumetric data, there was a sex differentiated effect of the quadratic term for age, however, in the matched sample, this trend diminished greatly, while the main effect of the quadratic term of age became stronger (Fig. 2).
Conclusions:
Our results suggest that in most cases, the regional differences that can be found between females and males in aging trajectories can be attributed to the brain size differences between the two populations.
Lifespan Development:
Aging 1
Modeling and Analysis Methods:
Univariate Modeling 2
Other Methods
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
Cortical Anatomy and Brain Mapping
Keywords:
Aging
Cortex
Modeling
MRI
Sexual Dimorphism
Other - Brain Size
1|2Indicates the priority used for review
Provide references using author date format
- Desikan, R. S., et. al. (2006). An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. NeuroImage, 31(3), 968–980.
- Sanchis-Segura, C., et. al. (2020). Effects of different intracranial volume correction methods on univariate sex differences in grey matter volume and multivariate sex prediction. Scientific Reports, 10(1), Article 1.
- Schaefer, A., et. al. (2018). Local-Global Parcellation of the Human Cerebral Cortex from Intrinsic Functional Connectivity MRI. Cerebral Cortex, 28(9), 3095–3114.
- Xiao, Y., et. al. (2019). An accurate registration of the BigBrain dataset with the MNI PD25 and ICBM152 atlases. Scientific Data, 6(1), Article 1.
- Zeighami, Y., et. al. (2015). Network structure of brain atrophy in de novo Parkinson’s disease. eLife, 4, e08440.