Professional Documents
Culture Documents
Neuropsychologia
journal homepage: www.elsevier.com/locate/neuropsychologia
a r t i c l e i n f o a b s t r a c t
Article history: Temporal lobe epilepsy (TLE) has been associated with the phenomenon of accelerated long-term forget-
Received 19 January 2011 ting (ALF), in which memories are retained normally over short delays but are then lost at an accelerated
Received in revised form 5 April 2011 rate over days or weeks. The causes of ALF, and whether it represents a consolidation deficit distinct from
Accepted 14 April 2011
the one associated with forgetting over short delays, remain unclear. In addition, methodological issues
Available online 20 April 2011
have made results of some previous studies difficult to interpret. This study used improved methodology
to investigate the role of seizure activity in ALF. Forgetting was assessed in participants with TLE (who
Keywords:
have involvement of temporal lobe structures) and idiopathic generalised epilepsy (IGE; in which seizures
Temporal lobe epilepsy
Idiopathic generalised epilepsy
occur in the absence of identified structural pathology in the temporal lobes). Learning of novel stimuli
Memory was matched between patients with TLE, patients with IGE and healthy controls matched for age and IQ.
Accelerated long-term forgetting Results indicated that the TLE group showed accelerated forgetting between 30-min and three-weeks,
but not between 40-s and 30-min. In contrast, rates of forgetting did not differ between patients with
IGE and controls. We conclude that (1) ALF can be demonstrated in TLE in the absence of methodological
confounds; (2) ALF is unlikely to be related to the experience of epilepsy that does not involve the tem-
poral lobes; (3) neither seizures during the three-week delay nor polytherapy was associated with ALF.
0028-3932/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.neuropsychologia.2011.04.018
2418 N. Muhlert et al. / Neuropsychologia 49 (2011) 2417–2426
More recent studies have attempted to investigate the possi- known that investigation of forgetting rates in groups with memory
bility that ALF is related directly to TLE by investigating forgetting impairment involves serious methodological problems (see Isaac &
in patients with TLE without the atypical aetiologies of the single- Mayes, 1999a, 1999b). Difficulties arise when comparing rates of
case patients. Three studies (Blake, Wroe, Breen, & McCarthy, 2000; forgetting across groups that are performing at different levels as
Mameniskiene, Jatuzis, Kaubrys, & Budrys, 2006; Martin et al., it is known that forgetting curves are not linear and so comparing
1991) have reported ALF in groups with TLE and Jansari, Davis, performances which may be at different points on the forgetting
McGibbon, Firminger, and Kapur (2010) reported ALF in a case curve may result in scaling effects (e.g. Loftus, 1985). This difficulty
study of a patient with TLE. In addition, three studies (Butler et al., affects those studies in which performance of patients at the initial
2007; Manes, Graham, Zeman, Calcagno, & Hodges, 2005; Muhlert, delay is significantly poorer than that of controls (e.g. Bell, 2006;
Milton, Butler, Kapur, & Zeman, 2010) have reported ALF in groups Bell et al., 2005; Mameniskiene et al., 2006). One way of addressing
with transient epileptic amnesia, a phenomenon related to TLE this difficulty is to employ a matching procedure. However, the pro-
and associated with recurrent attacks of transient amnesia in the cedure of learning to criterion employed by Martin et al. (1991) and
context of ongoing difficulties with autobiographical and remote Blake et al. (2000) results in a ceiling effect at the short delays which
memory. However findings of ALF in groups with TLE have fre- may mask group differences. In these cases overlearning (Krueger,
quently not been replicated (Bell, 2006; Bell, Fine, Dow, Seidenberg, 1929) may occur, whereby repeated study-test trials beyond 100%
& Hermann, 2005; Giovagnoli, Cassaza, & Avanzini, 1995; Howard recall can result in forgetting rates being underestimated over ini-
et al., 2010). This implies either that (1) ALF is not necessarily char- tial delays (e.g. Driskell, Willis, & Cooper, 1992). Bell (2006) has
acteristic of TLE and may depend on individual epilepsy variables attributed the positive findings of ALF in the Martin et al. (1991)
or that (2) the phenomenon of ALF relates to methodological issues. and Blake et al. (2000) studies to these issues, although it should
In relation to the first possibility, studies have investigated fac- be noted that neither of these studies reported any group differ-
tors relating to structural pathology, antiepileptic medication and ences in a number of trials to reach criterion across patient and
seizure variables. control groups. Nevertheless, it remains to be established whether
Structural lesions have been reported in a minority of patients ALF occurs in TLE in the absence of scaling effects, ceiling effects
with ALF. Three of the 9 patients showed ALF in Blake et al. (2000); and overlearning.
11 of the 70 patients in Mameniskiene et al. (2006); 3 of the 10 To address these methodological difficulties, we matched learn-
patients in Manes et al. (2005); and none of the 24 patients in Butler ing by giving poor learners extended exposure to stimuli during
et al. (2007) had structural lesions. However, none of these stud- learning. Titrating presentations in this manner have been used
ies included detailed structural brain analysis such as volumetric previously to explore forgetting rates in groups with organic
measures. In one study that did include volumetry (Butler et al., amnesia syndromes (Isaac & Mayes, 1999a, 1999b). Using these
2009), subtle hippocampal atrophy was reported in patients with techniques this study had the following three aims. The first aim
transient epileptic amnesia, a subgroup of whom showed ALF. This was to compare forgetting over delays of up to three-weeks in
atrophy correlated negatively with retention over 30-min but not TLE, IGE and control groups to determine whether the presence
over one week (Butler et al., 2009). This indicates that hippocam- of generalised seizure activity in the absence of specific tempo-
pal function is important for retention over short delays. Whether ral lobe involvement is associated with abnormal forgetting. The
it underlies ALF remains uncertain. second aim was to conduct individual analyses of forgetting rates
The effects of antiepileptic medication on ALF have been within the three groups of participants to investigate the possibil-
reported in a single case, in which ALF was demonstrated both ity of heterogeneity and possible bases for this. The third aim was
before and after the administration of antiepileptic medication to investigate the relationship between rate of forgetting and per-
(Jansari et al., 2010). ALF has also been reported in patients with ceived memory, quality of life, mood and antiepileptic medication.
transient epileptic amnesia in whom attacks usually resolve even
2. Methods
on relatively low doses of antiepileptic medication, doses which
would be unlikely to lead to memory problems (e.g. Butler et al., The study was approved by the South Sheffield Research Ethics Committee. All
2007; Muhlert et al., 2010). participants gave informed consent to participate.
The relationship between seizures during the study period
and retention is variable. Patients with TLE who experience 2.1. Participants
seizures during retention periods have shown both greater for- Fourteen patients with TLE (ten females and four males), and fourteen with IGE
getting (Mameniskiene et al., 2006) and similar forgetting (Bergin, (eight females and six males) were recruited from the local neurology department.
Thompson, Fish, & Shorvon, 1995; Blake et al., 2000) to those who Fifteen healthy control participants (seven males and eight females) with no neu-
were seizure-free. ALF has also been reported in patients who have rological conditions or previous head injury were also studied. All participants met
the following inclusion criteria: (1) English as a native language; (2) age between
been seizure-free during, and for a minimum of four-months prior
16 and 75 years; (3) Wechsler Abbreviated Scale of Intelligence (WASI; Wechsler,
to, testing (Butler et al., 2007; Muhlert et al., 2010). 1999) Full Scale IQ greater than 80. Patients were allocated to the TLE or IGE groups
The role of seizures in ALF can also be investigated by examining by experienced epileptologists on the basis of all available clinical data. For further
forgetting in patients with extra-temporal forms of epilepsy. Idio- clinical details see Table 1.
pathic generalised epilepsies (IGE) are characterised by bilateral Patients in the TLE group reported a mean of 3.8 seizures per month. Two were
not on medication, six were on monotherapy, four were on two antiepileptic med-
seizure onset on electroencephalography without structural brain ication, and two were on three antiepileptic medication (see Table 1).
lesions (Commission of ILAE, 1989). Evidence for ALF in IGE would In the IGE group, four patients had a diagnosis of juvenile myoclonic epilepsy,
suggest that epilepsy variables other than lesions are sufficient to one of generalised tonic–clonic seizures on awakening and the remainder of unclas-
cause ALF. This may include seizure-related activity, antiepileptic sified IGE. These patients reported a mean of 5.4 seizures per month. One was not
on medication, eight were on monotherapy, three were on two antiepileptic medi-
medication, mood or other psychosocial disturbances.
cation, one was on three antiepileptic medication, and one was on four antiepileptic
In the only study of ALF in IGE, children with IGE showed medication (see Table 1).
impaired one-week retention on story recall, but normal reten- A one-way analysis of variance (ANOVA) indicated a significant group difference
tion on story recognition and a visual recall task (Davidson, Dorris, for age (F(2,42) = 5.1, p = .010). Post hoc Tukey’s test showed that the TLE group was
O’Regan, & Zuber, 2007). However, these authors attributed their older than both IGE (t(26) = −3.0, p = .016) and control groups (t(27) = 2.7, p = .032),
but that there was no significant difference between IGE and control groups (p > .1).
findings to inefficient learning rather than ALF. To account for this difference, the TLE group was compared to a sub-group of age-
The second possibility outlined above is that the presence or matched controls (termed CON group), using the age of the youngest patient in the
absence of ALF is determined by methodological issues. It is well TLE group (21 years) as a lower limit for selection. The CON group included ten
N. Muhlert et al. / Neuropsychologia 49 (2011) 2417–2426 2419
Table 1
Patient characteristics. MTS, mesial temporal sclerosis; SPS, simple partial seizure; CPS, complex partial seizure; GTCS, generalised tonic–clonic seizure; AEDs, antiepileptic
medication; LEV, levetiracetam; CRB, carbamazepine; PRG, pregabalin; PHY, phenytoin; LMT, lamotrigine; OXC, oxcarbazepine; SVP, sodium valproate; CLB, clobazam; TPR,
topiramate.
MRI EEG Age Seizure types experienced Age of onset Duration AEDs
TLE
1 Left MTS – 65 SPS, CPS 16 49 LEV
2 Right glioma – 51 SPS, GTCS 50 1
3 Left MTS Left 47 SPS, CPS, GTCS 3 44 CRB, PRG, PHY
4 Normal – 43 CPS 28 15 CRB
5 Right MTS Right 45 CPS, GTCS 25 25 LMT, OXC
6 Left MTS Left 55 SPS, CPS 7 48 CRB, PRG
7 Right MTS – 56 SPS, GTCS 10 46 CRB, PHY, LEV
8 Normal Bilaterala 52 SPS 36 16 PHY
9 Normal – 21 SPS, CPS 13 8
10 Right MTS Bilaterala 42 CPS 28 13 LEV, PRG
11 Right stroke – 37 SPS, GTCS 34 4 LEV
12 Normal Right 48 SPS, CPS, GTCS 19 29 PRG, LMT
13 Right epidermoid cyst – 55 CPS, GTCS 44 11 CRB
14 Left MTS – 32 CPS, GTCS 3 29 CRB
IGE
1 – – 56 GTCS, absence 11 45 LEV, SVP, LMT, CLB
2 – Spike and wave 23 GTCS 10 13 LMT, SVP
3 – – 22 GTCS 13 9 SVP
4 – Slow wave, spikes 27 GTCS, absence 7 20 LMT, TPR, SVP
5 – – 52 GTCS, absence 24 28 LEV
6 – – 35 GTCS 15 20
7 – – 20 GTCS 10 10 SVP, LMT
8 – Poly-spike and wave rhythmic slow-wave 21 GTCS, absence 14 7 LMT
9 – – 25 GTCS, absence 18 7 LEV
10 – Spike and wave 18 GTCS 17 1 LMT
11 – Poly-spike and wave 43 GTCS, absence 22 21 LMT
12 – – 18 GTCS 11 7 LEV, SVP
13 – – 58 GTCS 17 41 SVP
14 – Poly-spike and wave 24 GTCS 20 4 LMT
a
These patients had evidence of seizure onset in left and right temporal lobes on different occasions.
Table 2
Demographics, epilepsy characteristics, and standardised neuropsychological test scores for TLE, IGE, control, and CON groups. T1, independent samples t-tests comparing TLE
to CON groups on the relevant measures (df = 22). T2, independent samples t-tests comparing IGE to control groups on the relevant measures (df = 27). Figures in parentheses
show standard deviations. WASI, Wechsler Abbreviated Scale of Intelligence; WTAR, Wechsler Test of Adult Reading; HADS, Hospital Anxiety and Depression Scale; AMIPB,
Adult Memory and Information Processing Battery; WRMT, Warrington Recognition Memory Test.
N 14 10 14 15
Gender (M/F) 4/10 3/7 6/8 7/8
Age 46.4 (11.0) 40.0 (13.3) 1.3 31.6 (14.6) 33.3 (15.4) −0.3
Age of onset 22.1 (14.1) – 14.9 (5.0) –
Duration of epilepsy 24.3 (16.9) – 16.6 (13.5) –
WASI
Full scale 112.0 (13.5) 112.7 (10.8) −0.2 113.1 (15.5) 117.5 (12.2) −0.8
Verbal 111.2 (10.3) 111.0 (8.8) 0.3 114.4 (16.4) 116.1 (12.6) −0.3
Performance 106.1 (15.9) 111.6 (12.8) −0.6 108.8 (13.4) 114.8 (11.5) −1.3
WTAR 104.0 (6.7) 107.8 (3.4) −1.6 101.3 (11.2) 107.3 (5.4) −1.6
HADS
Anxiety 7.9 (3.6) 6.5 (3.0) 1.0 7.4 (5.2) 5.7 (3.2) 1.0
Depression 3.1 (2.3) 3.4 (2.6) −0.3 3.6 (2.9) 3.3 (2.4) 0.3
Test of Everyday Attention
Elevator Counting 6.9 (0.3) 7.0 (0) N/Aa 6.7 (0.7) 7.0 (0) N/Aa
Elevator Counting with Distraction 6.9 (3.1) 8.2 (2.9) 1.0 7.5 (2.7) 8.3 (2.7) 0.7
AMIPB list
Learning trials 52.9 (10.6) 60.7 (3.5) −2.1* 59.1 (9.0) 60.6 (3.7) −1.3
Post-interference recall 10.6 (2.8) 12.9 (1.5) −2.6* 12.6 (2.3) 12.6 (1.4) −1.9
Delayed recall 10.6 (2.8) 13.5 (1.7) −2.4* 10. 9 (7.8) 13.3 (1.7) −1.5
AMIPB figure
Copy 78.3 (2.8) 79.4 (0.8) N/A 78.6 (2.1) 79.3 (0.9) N/A
Imm recall 53.4 (17.2) 61.7 (16.6) −1.2 64.7 (13.2) 66.5 (15.2) −0.3
Delayed recall 50.1 (14.6) 60.1 (16.0) −1.6 65.7 (12.4) 59.1 (16.7) 0.9
WRMT
Words 46.1 (4.9) 49.4 (1.1) −2.1* 48.3 (2.1) 49.3 (1.1) −1.5
Faces 39.3 (4.8) 45.3 (3.2) −3.4** 42.1 (4.1) 45.1 (3.2) −1.9
a
Ceiling effects in controls on the Elevator Counting subtest prevented reliable statistical analysis. According to normalised scoring procedure one TLE patient and two
IGE patients performed in possibly abnormal range, and one IGE performed in abnormal range on this test.
*
p < .05.
**
p < .01.
2420 N. Muhlert et al. / Neuropsychologia 49 (2011) 2417–2426
Fig. 1. (a) Procedure for presentation of scenes. For each scene foreground objects were presented individually in front of the background for 2 s. Following presentation of
each of the five objects, the entire scene was presented for 10 s. (b) Recall sheet for beach scene.
participants and was matched to the TLE group for age and IQ. The IGE group was The position of the correct answer was varied pseudorandomly between questions,
matched to the controls for age and IQ (see Table 2). appearing equally often in each position.
and a 30-min filled delay. During the 30-min delay, participants were administered Table 3
subtests from the Test of Everyday Attention. After 30-min delayed testing, patients Percentage of TLE, IGE and control participants given one, two or three presentations
completed the QoLIE-31 and were given a diary to record any seizures between of the experimental stimuli.
the second and third sessions. The participants were not informed that memory for
Number of presentations of stimuli Frequency
the stimuli would be subsequently assessed, so as to minimise rehearsal during the
delay. TLE IGE Control CON
In the third session which took place three-weeks later, retention of the exper-
imental tests was assessed. In addition, seizure diaries were reviewed, participants 1 29% 71% 73% 70%
completed the Words and Faces subtests of the RMT, the HADS, and the Long-Term 2 64% 29% 20% 20%
Memory Questionnaire. 3 7% 0% 7% 1%
15
4 TLE
TLE
10 IGE IGE
Control Control
CON 2 CON
5
0 0
40 second 30 Minute 3 Week 40 second 30 Minute 3 Week
b Story Recall
b Spatial Discrimination Memory (Sp) 80
60
TLE
6 TLE IGE
IGE 40
Control
4 Control
CON
CON
20
2
0
0
40 second 30 Minute 3 Week
40 second 30 Minute 3 Week
c Story Recognition
c Corrected Descriptive Recall (Des) 12
Mean Recognition (Max = 12)
Mean descriptive score (Des)
100
9
TLE
75 IGE
TLE
6
IGE Control
50 Control CON
CON 3
25
0
0 40 second 30 Minute 3 Week
40 second 30 Minute 3 Week
Fig. 3. Mean performance of TLE, IGE, control and CON groups on the visual scenes
Fig. 2. Mean performance of TLE, IGE, control and CON groups on the visual scenes recall test. (a) Performances on item recall. (b) Performances on spatial recall. (c)
recall test. (a) Performances on item recall. (b) Performances on spatial recall. (c) Performances on descriptive recall. Error bars show 95% confidence intervals.
Performances on descriptive recall. Error bars show 95% confidence intervals. Sp:
spatial score. Des: descriptive score. Both spatial score and descriptive score are
contingent upon successful recall of items (see Appendix A).
3.3. Effects of seizures
100
60
IGE
Control
TLE
40
20
0
30 Min 3 Week 30 Min 3 Week 30 Min 3 Week 30 Min 3 Week 30 Min 3 Week
Item recall Spatial recall Descriptive recall Story Recall Story Recognition
Fig. 4. Percentage of participants in each group impaired between 40-s and 30-min delays (30 min) and between 30-min and three-week delays (3 week) on each of the
experimental tests. Participants performing worse than 1.96 SDs from the mean of the controls were considered impaired.
trace = 0.43, F(10,42) = 1.2, p = .332, 2p = .21) or patients with IGE (visual scenes item and descriptive recall, and story recognition)
alone (Pillai’s trace = 0.63, F(10,42) = 1.9, p = .071, 2p = .31), indicat- between 40 s and 30 min delays (zMEM), and between 30 min and
ing that experiencing seizures during the delay did not significantly three-week delays (zALF), respectively. This summarised forgetting
affect retention. across the different tests and reduced the number of variables used
in analyses.
3.4. Effects of medication Pearson product–moment correlations in patients showed a sig-
nificant correlation between zMEM and perceived memory (r = .4,
To assess the effects of medication on memory, patients on p = .014) but there were no significant correlations with QoLIE-
mono-therapy (six TLE and eight IGE) were compared to those on 31 full score (r(28) = .2, p = .598), or total HADS score (r(28) = −.1,
poly-therapy (six TLE and five IGE). The two patients with TLE who p = .843). Similarly, zALF did not correlate with perceived memory
were not taking antiepileptic medication at the time of testing were (r(28) = .2, p = .155), QoLIE-31 full score (r(28) = .2, p = .232), or total
omitted from these analyses. HADS score (r(28) = −.1, p = .626).
A MANOVA comparing performances across the five experimen-
tal tests found no delay by group (monotherapy and polytherapy) 4. Discussion
interaction over the three delays (Pillai’s trace = 0.11, F(10,86) = .5,
p = .859, 2p = .06). Similar MANOVAs found no delay by group inter- This study explored forgetting rates in TLE, IGE and control
action in the TLE group alone (Pillai’s trace = 0.65, F(10,34) = 1.7, groups after matching for learning using titrated presentations.
p = .130, 2p = .33) or in the IGE group alone (Pillai’s trace = 0.54, Once matched for learning, patients with TLE showed normal reten-
F(10,38) = 1.4, p = .217, 2p = .27). tion on all tests over 30-min. Over three weeks, they showed
significantly greater forgetting on visual scenes item recall, descrip-
3.5. Analysis of performance of individuals in the study tive recall and story recognition, compared with age-matched
controls. They exhibited normal retention on spatial recall and story
The percentage of TLE, IGE and control participants with recall tests. Patients with IGE showed no evidence of increased for-
impaired retention (defined as forgetting greater than 1.96 stan- getting at either 30 min or three-week delays. Seizures during the
dard deviations from the mean of the control group) over 30 min delay were not associated with worse retention in either group, nor
and three-week delays can be seen in Fig. 4. was polytherapy, although it should be noted that these analyses
Whereas a similar number of IGE and control participants were based on small sample sizes. Forgetting was not significantly
showed impaired retention over the three-week delay, a greater related to perceived memory or quality of life.
percentage of TLE participants showed impaired retention over this The IGE group was not impaired on either standardised or exper-
time (see Fig. 4). Both left and right TLE patients were impaired on imental tests. These findings are in line with the view that memory
visual scenes and story tests. A greater number of patients with deficits are rare in IGE (e.g. Hommet, Sauerwein, De Toffol, &
mesial temporal sclerosis showed impaired retention than non- Lassande, 2006). They also indicate that when learning is matched
mesial temporal sclerosis TLE patients (see Table 5). in the absence of overlearning, patients with IGE show normal
retention. Our findings support those of Davidson et al. (2007)
3.6. Perceived memory and quality of life and also lend credence to their hypothesis that impaired reten-
tion on story recall in their study was related to impaired learning
All of the patients and ten of the controls (7 CON) completed efficiency.
the perceived memory questionnaire. Self-ratings of memory The TLE group showed significantly worse retention than
were lower in the TLE (mean = 5.7, SD = 0.9) than CON group controls over the three-week delay on an omnibus assessment.
(mean = 7.8, SD = 1.3; t(19) = −4.2, p < .001, r = .69), and lower in the Analysis of results from the individual tests indicated ALF in three of
IGE (mean = 6.5, SD = 1.4) than control group (mean = 7.7, SD = 1.1; the five measures that were included in this analysis (visual scenes
t(22) = −2.2, p = .041, r = .42). recognition was not included because of a ceiling effect). These
As with Butler et al. (2009), composite measures of reten- were visual scenes item recall, visual scenes descriptive recall and
tion over 30-min (termed ‘zMEM’) and three-week delays (termed story recognition. On one of the two tests on which ALF was not
‘zALF’) were calculated. Retention scores were summed on the found (spatial recall), performance of the groups was not success-
three experimental tests on which patients with TLE showed ALF fully matched at the short delay and so the results could have been
2424 N. Muhlert et al. / Neuropsychologia 49 (2011) 2417–2426
subject to a scaling effect. On the other (story recall) four patients patients with and without mesial temporal sclerosis including
with TLE performed at floor level at the three-week delay and so measures of diffuse temporal lobe pathology will be important
floor effects may have masked differences in retention on this test. to establish the relationship between mesial temporal sclerosis
Furthermore, although group analysis indicated that retention on and ALF.
spatial recall and story recall tests was not impaired in patients The individual level analyses were not consistent with the
with TLE, analyses at the individual level showed greater evidence material-specific memory model. According to this model, right
of impaired three-week retention in individuals with TLE compared temporal lobe damage is associated with impaired visuo-spatial
to the controls on these tests. This indicates that, contrary to the memory (e.g. Milner, 1965), whereas left temporal lobe damage is
hypothesis of Bell (2006), ALF can occur in TLE in the absence of associated with impaired verbal memory (Delaney, Rosen, Mattson,
overlearning. & Novelly, 1980; Mungas, Ehlers, Walton, & McCutchen, 1985). This
Individual analyses may help to explain the discrepancy model can explain some findings in studies of ALF, as left but not
between studies. Bell (2006) found no significant difference right TLE has been associated with greater forgetting of a story
between percentage of participants with TLE and percentage of (Blake et al., 2000), but other studies have found no effect of lat-
control participants who showed ALF over two weeks. In con- erality on ALF (Bell, 2006; Bell et al., 2005; Giovagnoli et al., 1995;
trast, our individual analyses indicated that a greater percentage Martin et al., 1991). Although numbers were small, individual anal-
of participants with TLE showed impaired retention over three yses revealed ALF on both the visual scenes and story recognition
weeks in comparison with controls. This inconsistency may reflect tests regardless of the lateralisation of TLE. The absence of a later-
the fact that in Bell’s study, even at the short delay, patients ality effect on visual scenes recall tests may reflect a strong verbal
performed worse than controls and so results may have been component to this task. Indeed Dulay et al. (2002) report that the
influenced by a scaling effect. However, our individual anal- task’s homologue, the Family Pictures test from the WMS-III, may
ysis revealed heterogeneity in forgetting rates within the TLE involve verbalisation of information. Retention on story recogni-
group, as approximately half of these patients (four to eight tion was also impaired in both left and right TLE, arguing against
patients across the tests) showed impaired retention. It is also a strong effect of laterality of seizure onset on the experimental
possible, therefore, that rather than methodological differences results.
alone, Bell’s finding may relate to inclusion of a sample with Most studies of ALF have found similar patterns of forgetting in
a number of patients who did not show ALF. Although the patients who experience seizures during a retention period and
results of these individual analyses should be made tentatively those who remain seizure-free (Bergin et al., 1995; Blake et al.,
they could give valuable insights into the potential causes of 2000), and ALF is reported in patients who have been seizure-
ALF. free for six months or more (Butler et al., 2007). Our results are
In the individual analysis, impaired retention between 30 min in line with these studies: patients experiencing seizures showed
and three-weeks was more frequently identified in patients with similar forgetting to those who remained seizure-free. This find-
mesial temporal sclerosis (from 4 out of 7 patients on story recall ing held when assessing the TLE group, the IGE group, or both.
to 7 out of 7 patients on visual scenes spatial recall) than patients These findings are however inconsistent with those of a large group
without mesial temporal sclerosis (from 1 out of 7 patients on visual study by Mameniskiene et al. (2006), who report greater forget-
scenes spatial recall to 2 out of 7 patients on visual scenes item ting in patients experiencing seizures. They also found significant
recall), across the tests. These results could be consistent with those relationships between the quantity and type of seizures; patients
of Butler et al. (2009) who reported significant hippocampal atro- experiencing more than four seizures forgot more material than
phy in patients with transient epileptic amnesia who showed ALF. those experiencing fewer than four, and seizures affecting con-
Although they found no direct relationship between hippocampal sciousness (complex partial or secondarily generalised seizures)
volume and ALF, their findings and the present results suggest that were associated with more forgetting than seizures not affecting
ALF may in some way be associated with hippocampal dysfunction. consciousness (simple partial seizures). It should be noted, how-
It is possible that ALF is not directly related to hippocampal/medial ever, that this study did not differentiate participants showing ALF
temporal sclerosis but that it relates to a third underlying variable, from those who forgot across both short and long delays. It is pos-
such as diffuse temporal lobe pathology, which has been reported sible, therefore that the effects of seizures on ALF are reflective
to be associated with hippocampal sclerosis (e.g. Yogarajah et al., of a correlate of more severe epilepsy than seizures per se. Our
2008). This explanation could also account for the fact that ALF results suggest that overt seizures are unlikely to play a major
has not been associated with hippocampal damage in cases in role in ALF. It is not possible to exclude the possibility that sub-
whom the primary aetiology has not been temporal lobe epilepsy. clinical epileptiform activity may disrupt consolidation processes,
Further research comparing forgetting rates in larger samples of however.
Table 5
Number of patients with left and right TLE, with and without mesial temporal sclerosis (MTS), who showed impaired retention between 40 s and 30 min delays, and between
30 min and three-week delays.
At present the cause of ALF is unclear, although the temporal Appendix A. Correction for spatial and descriptive memory
lobe is clearly implicated. The timescale over which ALF occurs is discrimination scores
also unclear. This is mainly due to methodological difficulties in
assessing ALF over multiple delays when participants may antic- Spatial information was scored using a corrected measure
ipate testing and rehearse material which may then obliterate because the possible spatial score varied as a function of the num-
the effect (e.g. Jansari et al., 2010). This difficulty has resulted ber of items recalled (see Hunkin, Parkin, & Longmore, 1994). For
in the vast majority of studies including assessment at only one example, recall of one item with correct spatial location would lead
long delay. Nevertheless ALF has been reported at delays as short to a higher score (100%) than recall of six items with correct spa-
as 24 h (Jansari et al., 2010; Martin et al., 1991; Muhlert et al., tial location for five of those items (80%). To compensate for the
2010; O’Connor et al., 1997). This raises the possibility that sleep possibility of attaining extreme scores (0% or 100%), Hunkin et al.’s
disruption may play a role as it is well known that sleep plays discrimination score (Sp) was used. This is calculated as:
an important role in memory consolidation (e.g. Stickgold, 2005).
r−x
Recently Westerberg et al. (2010) reported that in a group of peo- Sp =
SD
ple with amnestic mild cognitive impairment, poorer recognition of
items learned 24 h previously was associated with reports of poorer where r is the correct spatial responses; x = n·p; n is the number of
quality sleep. In relation to epilepsy it has been reported that people items recalled; p is the probability of recalling spatial information
with simple and complex partial seizures have a higher incidence by chance; SD is the square root of (n·p·q); and q = (1 − p).
of sleep complaints than healthy controls (De Weerd et al., 2004). Descriptive recall was also measured using Des. This was calcu-
In addition Machline Carrion, Lahorgue Nunes, Longo Martinez, lated as
Wetters Portuguez, and da Costa (2010) reported a trend for poorer d/2
sleep quality to be associated with nocturnal seizures in a group of Des = × 100
i
patients with TLE pre-surgery. Subjective sleep quality improved
post-surgery along with seizure reduction. These studies raise the where d is the descriptive recall raw score (halved as two points
possibility that sleep disruption may play a role in ALF. Whether were available for each item); i is the item recall raw score.
this can be attributed to structural damage or nocturnal seizure
activity remains unclear. References
We found that patients’ subjectively perceived memory cor-
related with their retention over 30-min on the tests that were Alvarez, P., & Squire, L. R. (1994). Memory consolidation and the medial temporal
lobe: A simple network model. Proceedings of the National Academy of Sciences
sensitive to ALF. In contrast, perceived memory did not signif-
USA, 91, 7041–7045.
icantly correlate with retention over three weeks. Blake et al. Bell, B. (2006). WMS-III logical memory performance after a two week delay in
(2000) also found no significant relationship between ALF and temporal lobe epilepsy and control groups. Journal of Clinical & Experimental
perceived memory. However their findings were not replicated Neuropsychology, 28, 1435–1443.
Bell, B., Fine, J., Dow, C., Seidenberg, M., & Hermann, B. (2005). Temporal lobe epilepsy
in Butler et al. (2009), who found that a composite measure of and the selective reminding test: The conventional 30-minute delay suffices.
ALF predicted self-ratings of memory on the everyday memory Psychological Assessment, 17, 103–109.
questionnaire in patients with transient epileptic amnesia. It may Bergin, P. S., Thompson, P. J., Fish, D. R., & Shorvon, S. D. (1995). The effect of seizures
on memory for recently learned material. Neurology, 45, 236–240.
be that the negative findings between ALF and subjectively per- Blake, R., Wroe, S., Breen, E., & McCarthy, R. (2000). Accelerated forgetting in patients
ceived memory in patients with TLE (but not transient epileptic with epilepsy: Evidence for an impairment in memory consolidation. Brain, 123,
amnesia) relate to the small sample sizes used in these analyses. 472–483.
Butler, C. R., Bhaduri, A., Acosta-Cabronero, J., Nestor, P. J., Kapur, N., Graham, K.
Subjectively perceived memory in TLE may also relate to forgetting S., et al. (2009). Transient epileptic amnesia: Regional brain atrophy and its
over shorter, 30 min delays, as shown in this study. Alternatively, relationship to memory deficits. Brain, 132(2), 357–368.
reports of poor subjective memory in TLE may relate to mood Butler, C. R., Graham, K. S., Hodges, J. R., Kapur, N., Wardlaw, J. M., & Zeman, A. Z.
(2007). The syndrome of transient epileptic amnesia. Annals of Neurology, 61,
and adjustment related issues (see Hall, Isaac, & Harris, 2009 for
587–598.
a review). Commission of the ILAE. (1989). Proposal for revised classification of epilepsies and
Quality of life and mood were also not correlated with mea- epileptic syndromes. Epilepsia, 30, 389–399.
Coughlan, A. K., & Hollows, S. E. (1985). The adult memory and information processing
sures of retention over 30-min or three-week delays. This suggests
battery (AMIPB).
that ALF may have no greater impact on patients’ mood and Davidson, M., Dorris, L., O’Regan, M., & Zuber, S. M. (2007). Memory consolidation and
quality of life than other epilepsy-related issues (e.g. seizure- accelerated forgetting in children with idiopathic generalised epilepsy. Epilepsy
worry). It should be noted, however, that this result was based & Behavior, 11(3), 394–400.
De Renzi, E., & Lucchelli, P. (1993). Dense retrograde amnesia, intact learning capa-
on a small sample. Further research in larger samples may bility and abnormal forgetting rate: A consolidation deficit? Cortex, 29, 449–466.
help to fully establish the influence of forgetting on quality of De Weerd, A., De Haas, S., Otte, A., Kasreleijn-Nolst-Trenite, D., Van Erp, G., Cohen,
life. A., et al. (2004). Subjective sleep disturbance in patients with partial epilepsy: A
questionnaire-based study on prevalence and impact on quality of life. Epilepsia,
45, 1397–1404.
Delaney, R. C., Rosen, A. J., Mattson, R. H., & Novelly, R. A. (1980). Memory function
5. Conclusion in focal epilepsy: A comparison of non-surgical, unilateral temporal lobe and
frontal lobe samples. Cortex, 16, 103–117.
Driskell, J. E., Willis, R. P., & Cooper, C. (1992). Effect of overlearning on retention.
This study demonstrated evidence for ALF in TLE but not IGE Journal of Applied Psychology, 77(5), 615–622.
in the absence of overlearning and ceiling effects. Forgetting did Dulay, M. F., Schefft, B. K., Testa, S. M., Fargo, J. D., Privitera, M., & Yeh, H. S.
not appear to be related to the experience of overt seizures (2002). What does the family pictures subtest of the Wechsler Memory Scale-III
measure? Insight gained from patients evaluated for epilepsy surgery. Clinical
between learning and testing. This suggests that ALF was not Neuropsychology, 16(4), 452–462.
caused by a history of seizures, overt seizures during retention Giovagnoli, A. R., & Avanzini, G. (1999). Learning and memory impairment in patients
periods, antiepileptic medication, or mood disturbance. ALF was with temporal lobe epilepsy: Relation to the presence, type and location of brain
lesion. Epilepsia, 40(7), 904–911.
not correlated with subjectively perceived memory or quality of Giovagnoli, A. R., Casazza, M., & Avanzini, G. (1995). Visual learning on a selective
life. Preliminary findings suggest an association between forget- reminding procedure and delayed recall in patients with temporal lobe epilepsy.
ting over three-weeks in TLE and the presence of mesial temporal Epilepsia, 36, 704–711.
Giovagnoli, A. R., Cassaza, M., Broggi, G., & Avanzini, G. (1996). Verbal learning and
sclerosis which could form the focus of future research into this
forgetting in patients with temporal lobe epilepsy. European Journal of Neurology,
phenomenon. 3, 345–353.
2426 N. Muhlert et al. / Neuropsychologia 49 (2011) 2417–2426
Hall, K. E., Isaac, C. L., & Harris, P. (2009). Memory complaints in epilepsy: An accurate Martin, R. C., Loring, D., Meador, K., Lee, G., Thrash, N., & Arena, J. (1991). Impaired
reflection of memory impairment or an indicator of poor adjustment? A review long-term retention despite normal auditory learning in patients with temporal
of the literature. Clinical Psychology Review, 29(4), 354–367. lobe dysfunction. Neuropsychology, 5, 3–12.
Hendriks, M. P. H., van Kampen, A., Aldenkamp, A. P., van der Vlugt, H., Alpherts, W. Mayes, A. R., Isaac, C. L., Holdstock, J. S., Cariga, P., Gummer, A., & Roberts, N. (2003).
C. J., & Vemeulen, J. (2003). Recognition memory of serially or simultaneously Long-term amnesia: A review and detailed illustrative case study. Cortex, 39,
presented words or figures, of epilepsy patients with or without mesial temporal 567–603.
sclerosis. Epilepsy Research, 57(2), 137–144. Milner, B. (1965). Visually guided maze learning in man: Effects of bilateral hip-
Hommet, C., Sauerwein, H. C., De Toffol, B., & Lassande, M. (2006). Idiopathic epileptic pocampal, bilateral frontal and unilateral cerebral lesions. Neuropsychologia, 3,
syndromes and cognition. Neuroscience & Biobehavioral Review, 30, 85–96. 317–338.
Howard, C. E., Andres, P., Broks, P., Noad, R., Sadler, M., Coker, D., et al. (2010). Muhlert, N., Milton, F., Butler, C., Kapur, N., & Zeman, A. Z. (2010). Accelerated
Memory, metamemory and their dissociation in temporal lobe epilepsy. Neu- long-term forgetting of real-life events in transient epileptic amnesia. Neuropsy-
ropsychologia, 48(4), 921–932. chologia, 48(11), 3235–3244.
Hunkin, N. M., Parkin, A. J., & Longmore, B. E. (1994). Aetiological variation in the Mungas, D., Ehlers, C., Walton, N., & McCutchen, C. B. (1985). Verbal learning differ-
amnesic syndrome: Comparisons using the list discrimination task. Neuropsy- ences in epileptic patients with left and right temporal lobe foci. Epilepsia, 26,
chologia, 32(7), 819–825. 340–345.
Isaac, C. L., & Mayes, A. R. (1999a). Rate of forgetting in amnesia. I: Recall and recogni- O’Connor, M., Sieggreen, M. A., Ahern, G., Schomer, D., & Mesulam, M. (1997). Accel-
tion of prose. Journal of Experimental Psychology: Learning, Memory & Cognition, erated forgetting in association with temporal lobe epilepsy and paraneoplastic
25(4), 942–962. encephalitis. Brain & Cognition, 35, 71–84.
Isaac, C. L., & Mayes, A. R. (1999b). Rate of forgetting in amnesia. II: Recall and recog- Psychology Software Tools. (2002). E-prime. Pittsburgh: Psychology Software Tools,
nition of word lists at different levels of organisation. Journal of Experimental Inc.
Psychology: Learning, Memory & Cognition, 25(4), 963–977. Robertson, I. H., Ward, T., Ridgeway, V., & Nimmo-Smith, I. (1994). The test of everyday
Jansari, A. S., Davis, K., McGibbon, T., Firminger, S., & Kapur, N. (2010). When “long- attention (TEA). Suffolk, England: Thames Valley Test.
term memory” no longer means “forever”: Analysis of accelerated long-term Stevens, J. P. (2002). Applied multivariate statistics for the social sciences. Hillsdale, NJ:
forgetting in a patient with temporal lobe epilepsy. Neuropsychologia, 48(6), Lawrence Erlbaum.
1707–1715. Stickgold, R. (2005). Sleep dependent memory consolidation. Nature, 437,
Kapur, N., Millar, J., Colbourn, C., Abbott, P., Kennedy, P., & Docherty, T. (1997). Very 1272–1278.
long-term amnesia in association with temporal lobe epilepsy: Evidence for Vickrey, B. G., Perrine, K. R., Hays, R. D., Hermann, B. P., Cramer, J. A., Meador, K. J., &
multiple-stage consolidation processes. Brain & Cognition, 35(1), 58–70. Devinsky, O. (1993). Quality of life in epilepsy QoLIE-31 scoring manual.
Kapur, N., Scholey, K., Moore, E., Barker, S., Brice, J., & Thompson, S. (1996). Long-term Warrington, E. K. (1984). Recognition memory test. Windsor, UK: NFER
retention deficits in two cases of disproportionate retrograde amnesia. Journal Nelson.
of Cognitive Neuroscience, 8, 416–434. (1999). Wechsler Abbreviated Scale of Intelligence (WASI). San Antonio: The Psycho-
Krueger, W. C. F. (1929). The effect of overlearning on retention. Journal of Experi- logical Corporation, a Harcourt Assessment Company.
mental Psychology, 12, 71–78. (1997). Wechsler Memory Scale (WMS-III). San Antonio: The Psychological Corpora-
Loftus, G. R. (1985). Evaluating forgetting curves. Journal of Experimental Psychology: tion, a Harcourt Assessment Company.
Learning, Memory & Cognition, 11(2), 397–405. (2001). Wechsler Test of Adult Reading Manual. San Antonio: The Psychological Cor-
Lucchelli, F., & Spinnler, H. (1998). Ephemeral new traces and evaporated remote poration, a Harcourt Assessment Company.
engrams: A form of neocortical temporal lobe amnesia? A preliminary case Westerberg, C. E., Lundgren, E. M., Florczak, S. M., Mesulam, M. M., Weintraub, S., Zee,
report. Neurocase, 4, 447–459. P. C., et al. (2010). Sleep influences the severity of memory disruption in amnes-
Machline Carrion, M. J., Lahorgue Nunes, M., Longo Martinez, J. V., Wetters Portuguez, tic mild cognitive impairment. Alzheimer’s Disease and Associated Disorders, 24,
M., & da Costa, J. C. (2010). Evaluation of sleep quality in patients with refractory 325–333.
seizures who undergo epilepsy surgery. Epilepsy & Behavior, 17, 120–123. Yogarajah, M., Powell, H. W. R., Parker, G. J. M., Alexander, D. C., Thompson, P. J.,
Mameniskiene, R., Jatuzis, D., Kaubrys, G., & Budrys, V. (2006). The decay of memory Symms, M. R., et al. (2008). Tractography of the parahippocampal gyrus and
between delayed and long-term recall in patients with temporal lobe epilepsy. material specific memory impairment in unilateral temporal lobe epilepsy. Neu-
Epilepsy & Behavior, 8, 278–288. roimage, 40(4), 1755–1764.
Manes, F., Graham, K., Zeman, A. Z., Calcagno, M. L., & Hodges, J. R. (2005). Auto- Zigmond, A. S., & Snaith, R. P. (1983). The hospital anxiety and depression scale. Acta
biographical amnesia and accelerated forgetting in transient epileptic amnesia. Psychiatrica Scandinavica, 67, 361–370.
Journal of Neurology, Neurosurgery & Psychiatry, 76, 1387–1391.