Neurostten��e Letters 88 (1988) 271 274 271 Elsevier Scientific Pubhshers Ireland Ltd NSL 05341 Increased geomagnetic activity and the occurrence of bereavement hallucinations: evidence for melatomn-medmted microseizuring in the temporal lobe? M i c h a e l A P e r s m g e r Behavtoral Neuro~ttente Laborator~ Department o[ Pv~tholog~ Laurenttan Unt~erstt~ Sudhurx Ont ( ( anada ) (Recevmd 28 October 1987, Rewsed version received 10 February 1988 Accepted 19 Februar) 1988) Ke~ ~tords Temporal lobe, Hallucination Death Bereavement apparition Geomagnetlc,u-tlxlt) Eplde- mlology Melatonm This study was designed to test the hypothesis that common bereavement apparitions are hallucinatory e,cpenences evoked by transient electrical instability within the (glucocortlcold) sensitized meslobasal tem- poral lobes All first hand reports of'postmortem apparition' experiences were collected from a pubh~hcd data base The days on which the experiences occurred d~splayed s~gmficantty greater (mean lncreasc- 10 gamma) geomagnetic activity compared to the days before or afterv, ards These results suggest that bereavement apparitions are Sltuat~on-specLfic hallucinations evoked by m~croselzures ~lthm scnsltlv~-d temporal lobe structures, the occurrence of these mlcroselzures might be facilitated by suppression m mela- tomn levels that could accompany sudden increases m geomagnetic act~wtv Deep temporal lobe structures, such as the amygdala and htppocampus, are the most electrically unstable structures of the human brain [3] H t p p o c a m p a l lablhty can be slgmficantly sensitized by glucocortlcolds [8, 19] Rajaram and Mltra [18] recently reported that episodes of temporal lobe epilepsy were correlated wxth enhanced geo- magnetic activity, a decrease m seizure threshold was assocmted with changes of only about 20 g a m m a [7] In this context traditional correlations between psychiatric ad- mtsslons and increased geomagnetic actwtty [2] become more relevant because of the intimate involvement of temporal lobe factors m sch~zophremc and affect~ve d~s- orders [23] That electrical stimulation o f mes~obasal structures can evoke meaningful but hal- lucinatory experiences is well d o c u m e n t e d [3, 6, 9, 21] One hypothesis [13, 15] is that a c o m m o n form of bereavement p h e n o m e n a often labelled ' p o s t m o r t e m apparitions' ts due to mlcroselzures within the meslobasal temporal lobes (hlppocampus and (orre~pondente M A Persmger, Behavioral Neurosclence Laboratory Department ot Ps~cholog~ Laur- entlan Umvers~ty Sudbury, Ont, Canada P3E 2C6 0"~04-394I)'88 $ 03 50 O 1988 Elsevier Sclent111c Publishers Ireland Ltd

272 amygdala) that have been sensitized by elevated circulating glucocorticolds, they are a correlate of grief Because of the endogenous bursting characteristics of hIppocam- pal (and hmblc) neurons [1], alterations In the endogenous antlconvulsant effects of circulating melatonln by subtle extremely low frequency magnetic fields [5] produced by geomagnetic activity might evoke mlcroselzures that release intense situation-spe- cific (grief-relevant) imagery There is ample correlative evidence that mystical bereavement experiences are tem- poral lobe phenomena They are dominated by visual ('apparition') and then au&- tory ('messages') components [13, 14], the primary and secondary modahties that project into the human temporal lobe [24] A large proportion ofapparmons involves details that are memory fragments, particularly from the childhood of the expe- rlencer [13, 14] Most apparitions occur within the first 3 days of grieving (during peak circulating glucocortlcoid periods), between 02 00 and 04 00 h local time and are frequently associated with 'suddenly awakening' or dreams, these periods are associated with pontine-geniculo-occipltal (PGO) spikes and normal eplsodm (REM) activity that can be considered microseizures [21] Recent evidence suggests that the normal brain contains electrically active mlcrofocl within cortical columns [4] and that temporal lobe lability is strongly coupled with a propensity to report mystmal (psi) experiences I16, 17] To test this hypothesis all of the postmortem experiences that contained the day, month and year of occurrence were obtained from FATE magazine [17], a total of 203 cases that were pubhshed over a 37 year period met this criterion The cases were collected in 3 separate efforts (rephcatmns) from pubhcatmn years 1963-1969 (n = 75), 1970-1985 (n = 65) and 1949-1962 (n = 63) The aa (average antipodal) index [12] was used as the measure of global geomagnetic activity It was selected because dady lndmes were avadable for the years 1867 to present and many of the cases oc- curred before the beginning of more standard geomagnetic indices The average aa indices for each of the 3 days before, the day of and for each of the 3 days after the days of the experience were obtained If the hour was given, adjustments were made for the discrepancy between Greenwmh time and local time by using appropriate half-day [12, 17] values The mean aa values for the months and the years In which the experience occurred were also obtained for comparison only The results of the analysis can be seen m Fig l Multiple analysis of variance (Manova, SPSSX software, VAX computer) demonstrated no significant difference between rephcatmns (F2200=0 10, P 0 0 5 ) but a highly significant difference between geomagnetic actiwty on different days (/761200 = 6 07, P 0 001), the interac- tion was not ssgmficant (F12,1200-----0 98, P 0 05) A posterlon correlated t-test analy- sis demonstrated that the major source of the difference was due to increased geo- magnetm activity on the days of the experiences compared to either or all of the days before or afterwards These results support the hypothesis that sudden increases in geomagnetm activity can facd~tate the experience of bereavement halluclnatmns The average increase was about l0 aa umts, whmh is equivalent (in this range) to about 10-20 gamma, it might be recalled that an increase of 19 gamma was associated with a significant reductmn