Longitudinal changes in grey and white matter during adolescence

This paper on the longitudinal changes in grey and white matter during adolescence is the "first study to examine within-subject longitudinal changes in diffusion data in healthy adolescents".  24 teenagers, between 13 and 18 years old (at time one), returned 2.5 years after their first scan and the study investigates changes in their grey and white matter volumes and the fractional anisotropy (FA) of two major white matter tracts.

The results corroborate previous cross-sectional studies which have found increases in white matter volume and FA and decreases in grey matter volume.

Regions of grey (in blue–light blue) and white (in red–yellow) matter where volume shows significant (corrected p < 0.05) decreases and increases, respectively, with age over time.

White matter tracts (in red–yellow) showing significant (corrected p < 0.05) age-related FA increases over time in our group of adolescents.

The main reason I have this paper saved on my desk though, is for the statistical method they use to analyze longitudinal data (because I'm going to have to use them very soon!)  The authors created a design matrix which had regressors for each subject individually, to remove any within subject variance and then a regressor of interest which was their ages.  They orthogonalized that regressor to the subject regressors which (I believe) means that they were looking at the effects of age after the individual subject variance is accounted for, which is, after all, their question of interest :)

I also really liked this figure (below), for two reasons: One, I find it really interesting how little FA changes when you look across the sample.  It's only when you look at the pairs of data that you see an effect of age.  That's great evidence for the power of longitudinal studies: a cross sectional study wouldn't have been as sensitive to these changes. Two, its a change in parallel diffusivity that is driving this change in FA.  Not in perpendicular diffusivity.  I'm now really interested in seeing whether there is a qualitative difference between younger and older subjects.  Maybe myelination is driving a change in perpendicular diffusivity for children and reorganization, decreased tortuosity (curvy-ness) or increased axonal diameter is driving a change in parallel diffusivity in adolescence.  Who knows?  Stay tuned ;)

Scatterplots, with longitudinal data points connected by a line, of mean values of FA, parallel and perpendicular diffusivity from white matter regions showing significant relationship with age over time in adolescents. Orange lines indicate the three left-handed subjects.

Oh, and PS: the paper outlines very clearly how to deliniate the corticospinal tract and the arcuate fasciculus.  That'll be really helpful when I come to draw those tracts because there are so many different ways to do it that it's nice to be able to directly compare with previously published work :)

Longitudinal changes in grey and white matter during adolescence.

Giorgio A, Watkins KE, Chadwick M, James S, Winmill L, Douaud G, De Stefano N, Matthews PM, Smith SM, Johansen-Berg H, James AC.

Neuroimage. 2010 Jan 1;49(1):94-103. Epub 2009 Aug 11.
PMID: 19679191