Patient age did not predict the SEEG subsequent memory effect in FR1 and independent catFR1 patients, but low reliability limits inference

We tested whether patient age predicts the size of the encoding subsequent memory effect (SME) in stereo-electroencephalography (SEEG) during free recall. The SME was the recalled-minus-not-recalled difference in log10 power 0–1.6 s after word onset, averaged across bipolar pairs. The primary analysis used 51 FR1 patients; replication used 15 catFR1 patients not included in FR1, and a pooled analysis included 66 patients. In FR1, age was not associated with HFA (Spearman rho = -0.00, 95% CI -0.25 to 0.25), theta (-0.13, -0.39 to 0.16), or tilt (0.12, -0.16 to 0.39; smallest Holm-adjusted p = 1.0). The independent catFR1 and pooled analyses also found no age association (smallest Holm-adjusted p = 0.78 and 0.66). In the same FR1 patients, age was associated with recall rate (rho = -0.33, 95% CI -0.57 to -0.05, p = 0.019). SME split-half reliability was low (Spearman-Brown 0.06 HFA, 0.23 theta, 0.26 tilt), limiting detection of small between-patient effects. The results do not establish that age has no effect. The tested ages were 18–65 years, only session 0 was used, electrode placement varied clinically, and region-specific effects were not tested.

stereo-electroencephalography · subsequent memory effect · free recall · age · HFA · theta · split-half reliability