Metabolite T₁ relaxation times decrease across the adult lifespan

Relaxation correction is an integral step in quantifying brain metabolite concentrations measured by in vivo magnetic resonance spectroscopy (MRS). While most quantification routines assume constant T-1 relaxation across age, it is possible that aging alters T-1 relaxation rates, as is seen for T-2 relaxation. Here, we investigate the age dependence of metabolite T-1 relaxation times at 3 T in both gray- and white-matter-rich voxels using publicly available metabolite and metabolite-nulled (single inversion recovery TI = 600 ms) spectra acquired at 3 T using Point RESolved Spectroscopy (PRESS) localization. Data were acquired from voxels in the posterior cingulate cortex (PCC) and centrum semiovale (CSO) in 102 healthy volunteers across 5 decades of life (aged 20-69 years). All spectra were analyzed in Osprey v.2.4.0. To estimate T-1 relaxation times for total N-acetyl aspartate at 2.0 ppm (tNAA(2.0)) and total creatine at 3.0 ppm (tCr(3.0)), the ratio of modeled metabolite residual amplitudes in the metabolite-nulled spectrum to the full metabolite signal was calculated using the single-inversion-recovery signal equation. Correlations between T-1 and subject age were evaluated. Spearman correlations revealed that estimated T-1 relaxation times of tNAA(2.0) (r(s) = -0.27; p < 0.006) and tCr(3.0) (r(s) = -0.40; p < 0.001) decreased significantly with age in white-matter-rich CSO, and less steeply for tNAA(2.0) (r(s) = -0.228; p = 0.005) and (not significantly for) tCr(3.0) (r(s) = -0.13; p = 0.196) in graymatter-rich PCC. The analysis harnessed a large publicly available cross-sectional dataset to test an important hypothesis, that metabolite T-1 relaxation times change with age. This preliminary study stresses the importance of further work to measure age-normed metabolite T-1 relaxation times for accurate quantification of metabolite levels in studies of aging.