Data Selection, Abstraction, and Synthesis
All three authors selected studies for inclusion and exclusion; the first author abstracted all data, the second and third authors arbitrated uncertainties and disagreements. We undertook a qualitative synthesis across studies because there was substantial heterogeneity with respect to research methodologies amongst the identified articles—ranging from randomized clinical trials, observational studies, and preclinical research on model organisms and in vitro studies. The substantial heterogeneity amongst these methodologies precluded a single metric of quality assessment.
Many studies utilized in vitro measures of receptor density and signal transduction, as differences in means before- and after-interventions. Briefly: assays for CB1/CB2 receptor density include autoradiography with tritiated ligands (usually [3H]CP55,940 or [3H]SR141716), Western blot or immunostaining with antibodies to CB1/CB2 proteins, and Northern blot with radio-labeled or fluorescent riboprobes for CB1/CB2 mRNA. Signal transduction studies measure cannabinoid-stimulated inhibition of adenylyl cyclase, cannabinoid-stimulated [35S]GTPγS binding, or electrophysiological assays of ex vivo brain slices. Electrophysiological studies include depolarization-induced suppression of excitation (DSE, via glutamatergic synapses), depolarization-induced suppression of inhibition (DSI, via GABAergic synapses), long-term depression of excitatory synaptic transmission (LTDE, via glutamatergic synapses), or long-term depression of inhibitory synaptic transmission (LTDI, via GABAergic synapsis).
Publication bias was addressed by asking investigators to contribute unpublished studies. Clinical interventions (intervention groups) with five or more studies are provided with an interpretive summary at the end of the section (e.g., the sections on NSAIDs, glucocorticoids, opiates, etc.).