A philisophical question:

The underlying principal behind BOLD fMRI is that signal increases 4-6 s after an event reflects a transient increase in metabolic demand for a localized region of tissue ( or something like that).

One paradigm we use involving reward cues reliably elicits activation in a region (the nucleus accumbens) where preclinical literature shows afferents of interest (ventral tegmental DAergic neurons) contribute a predominantly *inhibitory* input on post-synaptic targets (spiny GABAergic neurons).

Under these circumstances, why do we always see "activation" in a location where in preclinical experiments, postsynpatic neuron populations are *inhibited*?

Is it because there are other accumbens neurons being excited (perhaps by other afferents)?

Is it because there was initial inhibition followed by some rebound excitatory activity that the BOLD timecourse does not resolve?

Or, is it because metabolic demands at a synapse are disproportionately driven by the *pre*-synaptic neurons (vesicle packaging/release/reuptake etc), making the postsynpaptic effect (inhibition vs excitation) less important? Under this scenario, BOLD increase would indicate "general synaptic activity."

Does anyone know of any literature that addresses this issue?

Jim