Is visual activation associated with changes in cerebral high-energy phosphate levels?

Phosphorus magnetic resonance spectroscopy (31 P MRS) has been employed before to assess phosphocreatine (PCr) and other high-energy phosphates in the visual cortex during visual stimulation with inconsistent results. We performed functional 31 P MRS imaging in the visual cortex and control regions during a visual stimulation paradigm at an unprecedented sensitivity, exploiting a dedicated RF coil design at a 7 T MR system. Visual stimulation in a 3 min 24 s on-off paradigm in eight young healthy adults generated a clear BOLD effect with traditional 1 H functional MRI in the visual cortex (average z score 9.9 ± 0.2). However, no significant event-related changes in any of the 31 P metabolite concentrations, linewidths (7.9 ± 1.8 vs 7.8 ± 1.9 Hz) or tissue pH (7.07 ± 0.13 vs 7.06 ± 0.07) were detectable. Overall, our study of 31 P MRSI in 15 cm3 voxels had a detection threshold for changes in PCr, Pi and γ-ATP between stimulation and rest of 5, 17 and 10%, respectively. In individual subjects, the mean coefficients of variance for PCr and Pi levels of control voxels were 6 ± 3 and 19 ± 8% (three time point average of 3 min 24 s). Altogether this indicates that energy supply for neuronal activation at this temporal resolution does not drain global PCr resources.