Region-mean MTL HFA showed no subsequent memory effect for once-presented words in RAM RepFR1 (12 SEEG subjects, wide CI); parahippocampal pairs showed a positive effect for these words (10 subjects)

We analyzed archived medial temporal lobe (MTL) high-frequency activity features for 26 subjects from RAM RepFR1, session 0, in OpenNeuro ds005411 v1.0.1. A preregistered analysis tested subsequent memory effects (SMEs) for words shown once (1x), which the prior analysis of repeated words did not test. The MTL-mean 1x SME was +0.003 z (95% CI -0.091 to +0.094; 12 subjects; p = 0.94); 14 subjects had too few recalled 1x words for inclusion, so power was low and the interval does not exclude effects up to 0.09 z. The first-presentation repeated-word SME minus the 1x SME was -0.014 z (CI -0.141 to +0.113; p = 0.83). Entorhinal pairs showed no 1x SME (-0.046 z; n = 8), while parahippocampal pairs showed a positive 1x SME (+0.101 z; CI +0.042 to +0.169; n = 10; Holm p = 0.035). An exploratory, unregistered region analysis found a parahippocampal SME for repeated words at their second presentation (+0.063 z; CI +0.021 to +0.103; n = 21; p = 0.008), but not at the first presentation. These results indicate that the MTL-mean features did not show a reliable SME even for 1x words, so the earlier null result is not specific to repetition. They do not show that the measure is insensitive to memory: averaging across regions can hide a parahippocampal SME. The region samples are small, the repeated-word region result is exploratory, and both require replication.

subsequent memory effect · high-frequency activity · medial temporal lobe · parahippocampal · entorhinal · intracranial EEG · word recall · RAM RepFR1