J Neurol Neurosurg Psychiatry 83:1145-1148 doi:10.1136/jnnp-2012-302437
  • Cognitive neurology
  • Research paper

Episodic memory and the medial temporal lobe: not all it seems. Evidence from the temporal variants of frontotemporal dementia

  1. Yolande Pijnenburg2
  1. 1Department of Neurology, Diakonessenhuis, Utrecht, the Netherlands
  2. 2Department of Neurology, Alzheimer Center, VU University Medical Center, Amsterdam, the Netherlands
  3. 3Department of Medical Psychology, VU University Medical Center, Amsterdam, the Netherlands
  4. 4Department of Neuroradiology, VU University Medical Center, Amsterdam, the Netherlands
  5. 5Department of Epidemiology and Biostatistics, VU University Medical Center, Amsterdam, the Netherlands
  1. Correspondence to C Machiel Pleizier, Department of Neurology, Diakonessenhuis, PO Box 80250, 3508 TG, Utrecht, The Netherlands; mpleizier{at}
  1. Contributors CMP and YP were responsible for the study concept, design, analysis and interpretation of data, wrote the article and did final revisions of the manuscript. AEvdV was involved in data management, analysis and interpretation of data. EK and PS were involved in the collection of clinical data. TK interpreted the neuropsychological data. FB was responsible for the collection of radiological data (MTA rating). WMvdF was involved in study design and interpretation of data. All authors were involved in editing the manuscript and approved the final version of the submitted article.

  • Received 4 February 2012
  • Revised 20 July 2012
  • Accepted 1 August 2012
  • Published Online First 29 August 2012


Background Disproportionate medial temporal lobe atrophy (MTA) is an early finding in Alzheimer's disease (AD). Episodic memory impairment in AD is associated with the degree of MTA. Episodic memory impairment and MTA are also found in semantic dementia (SD) and in right temporal lobe atrophy (RTLA), the temporal variants of frontotemporal dementia, but their relationship is unclear.

Objective To compare episodic memory impairment among patients with these temporal variants of frontotemporal dementia with that of patients with AD with the same degree of MTA.

Methods Episodic memory was tested with the visual association test, and semantic memory (SM) with animal fluency and the visual association naming test. MTA was measured using a visual rating scale. Each patient with SD or RTLA was matched for MTA with two patients with AD. Comparisons of episodic memory and SM were made for patients with SD versus matched patients with AD; patients with RTLA versus matched patients with AD and for SD, RTLA and all patients with AD.

Results 27 patients with SD and 11 with RTLA were matched with 54 and 22 patients with AD, respectively. Episodic memory was less impaired in patients with SD than in those with AD (8 versus 2; p<0.001) and in patients with RTLA than in those with AD (10 versus 4.5; p=0.009). Semantic memory was more affected in patients with SD than in those with AD, and the Mini Mental State Examination score was higher in patients with RTLA than in those with AD. Comparison of the three diagnostic groups showed that episodic memory was most impaired in AD, whereas SM was most impaired in SD.

Conclusion Since episodic memory impairment is more severe in AD than in SD and RTLA, despite a comparable degree of MTA, atrophy of the medial temporal lobe alone cannot account for episodic memory dysfunction.


Episodic memory impairment is an early finding in Alzheimer's disease (AD). In early AD structural MRI studies show atrophy predominantly in the medial temporal lobes with symmetrical distribution and particular involvement of the hippocampus and entorhinal cortex.1 ,2 It has been suggested that declining episodic memory with ageing is related to reduced hippocampal volume.3 Others found a direct correlation between loss of (episodic) memory function and medial temporal lobe atrophy (MTA) in subjects with AD, mild cognitive impairment or even in healthy elderly people.4 ,5

MTA is also found in semantic dementia (SD), one of the temporal variants of frontotemporal dementia (FTD), clinically characterised by profound impairment in semantic memory (SM).6 ,7 SD, however, is characterised by asymmetrical temporal lobe atrophy with more left than right anterior temporal lobe atrophy.8

Although it has previously been thought that episodic memory in SD or other variants of FTD is relatively preserved, this has recently been challenged.7 ,9–12 One study even found episodic memory impairment in 90% of patients with SD and in patients with the radiologically opposite predominant right temporal lobe atrophy (RTLA).10 Patients with FTD with right temporal atrophy differ clinically from patients with SD in that they display more behavioural and personality changes and their language is less affected. According to the international consensus diagnostic criteria most patients with RTLA are diagnosed either as SD or as the behavioural variant of FTD (bvFTD).8

It remains to be resolved if in SD and RTLA episodic memory impairment is directly related to the degree of MTA.

In this study we compared impairment in episodic memory and SM in patients with SD, RTLA and AD with the same degree of MTA. We hypothesised that despite similar degrees of medial temporal atrophy, episodic memory is less affected in SD and RTLA than in AD. Moreover, we hypothesised that SM impairment is highest in SD compared with AD and RTLA.

Patients and methods


Twenty-seven patients with SD and 11 patients with RTLA were consecutively enrolled from the prospectively collected database at the Alzheimer's Center of the VU University Medical Center. All subjects had undergone a standard diagnostic procedure, including medical history, informant-based history, physical and neurological examination, neuropsychological testing, screening laboratory tests, EEG and MRI. Patients were diagnosed in a multidisciplinary consensus meeting. SD was diagnosed according the Neary criteria.8 Patients with a previous diagnosis of SD with severe and predominantly right rather than left anterior temporal atrophy were reclassified as having RTLA. Patients diagnosed as bvFTD with an announcement of ‘right temporal variant’, were also classified as RTLA.10 For each patient with SD and RTLA two patients with a diagnosis of probable AD according to the NINCDS–ADRDA criteria13 were matched for degree of MTA, age, sex and education level. The level of education was classified using the system of Verhage, ranging from 1 (not finished primary school) to 7 (university degree).14 The maximum time difference between MRI and neuropsychological assessment was 6 months; patients with longer intervals were not included. Moreover, patients with previous stroke or severe vascular white matter damage (Fazekas' scale 2 or 315) were not included.

Neuropsychological assessment

From all patients except one patient with SD Mini Mental State Examination (MMSE) scores were obtained. The neuropsychological test battery included the visual association test (VAT) as a test of episodic memory.16 In this test, patients are shown cue cards with an object as well as association cards with the previously seen object plus an interacting object and are asked to name them. Subsequently, the cue cards are shown without delay and patients are asked to recall the now missing interacting objects (score 0–12 for two trials). In addition, we used VAT picture naming (score 0–12) and category fluency (animals, 60 s) to assess SM. In the case of poor naming in patients with SD, memory nevertheless could be tested by, for example, describing or drawing the missing object. In 13 patients no data on VAT picture naming were available (10 with AD, two with SD and one with RTLA) and animal fluency was not tested in one patient with SD. The other episodic memory tests of the battery included the Rey auditory-verbal learning test and the Rey–Osterrieth complex figure copying test, but the first interferes with language (poor performance expected in patients with SD) and the latter was not routinely available for all patients. The Boston naming test, also included in the battery, was not assessed in all patients. The other tests (trail making test B, elements of the behaviour assessment of dysexecutive syndrome, Stroop effect and letter fluency) were used for assessment of executive function and in making a diagnosis, but are beyond the scope of this study.

MTA and matching

MTA was assessed using a visual rating scale, ranging from 0 (no atrophy) to 4 (severe volume loss) for each hemisphere.17 Since patients with AD have less asymmetrical MTA than both patients with SD and those with RTLA, matching for MTA was sometimes challenging. If an exact match was impossible because of severe right–left differences, defined as a difference in MTA score of two or more points, we chose to match the left MTA score of the patients with SD with the left MTA score of subjects with AD with a right MTA match that was as close as possible. Vice versa, in RTLA the right MTA score was selected if severe asymmetry made a proper match troublesome. We also calculated an MTA sum score to be sure that patients with SD or RTLA and those with AD had a roughly equal total atrophy score.

Statistical analysis

SPSS V.17.0, Windows, IBM, was used for statistical evaluation. The Mann–Whitney U test was used to confirm adequate matching for MTA, age and education level. A χ2 test was performed to test for differences in sex distribution. Since the assumption for normality was met only for category fluency, we chose to use Mann–Whitney U test to assess differences in performance on MMSE and all neuropsychological tests for SD versus AD and RTLA versus AD (each with their own set of matched patients with AD). A comparison between all patients of the three diagnostic groups was performed using the Kruskall–Wallis test followed by the Mann–Whitney U test.


Twenty-seven patients with SD and 11 patients with RTLA were matched with 54 and 22 subjects with AD, respectively. In patients with SD MTA scores left and right were more asymmetrical, which made matching more difficult with patients with AD in whom more symmetrical MTA was found. Patients with AD were older in both matched groups but there were no significant differences in age, sex, educational level or MTA scores between patients with AD and patients with either SD or RTLA (table 1).

Table 1

Demographic characteristics and MTA scores

Episodic memory was less impaired in both patients with SD compared with those with AD (p<0.001) and in patients with RTLA compared with those with AD (p=0.009, table 2).

Table 2

MMSE and neuropsychological test scores in SD compared with AD and RTLA compared with AD

Semantic memory as shown by both VAT picture naming and category fluency was more impaired in patients with SD than in those with AD (p<0.001 and p=0.001, respectively). There was no difference in SM performance between patients with RTLA and those with AD. Patients with RTLA had a higher median MMSE score than those with AD (26 vs 23.5; p=0.03).

When patients with SD, RTLA and all patients with AD were compared with each other the aforementioned differences were confirmed: episodic memory was impaired in all groups but to a lesser extent in patients with RTLA (p<0.001) and SD (p<0.01) than in those with AD. Although the RTLA–AD group had a less severe MTA sum score than the SD–AD group, performance on the VAT memory test and MMSE were not statistically significant different between patients with SD and RTLA.

Semantic memory impairment was more severe in patients with SD than in all those with AD (VAT picture naming p<0.001, animal fluency p<0.001), but also in comparison with patients with RTLA (VAT picture naming p<0.01, animal fluency p<0.001). Median MMSE scores were higher in patients with RTLA than in all patients with AD (26 vs 22.5; p<0.01). No difference in MMSE was found between patients with SD and RTLA or all patients with AD (figure 1).

Figure 1

Performance on Mini Mental State Examination (MMSE), episodic memory and semantic memory tests and comparisons between patients with semantic dementia (SD), right temporal lobe atrophy (RTLA) and all patients with Alzheimer's disease (AD). *p=0.001 (SD vs AD) and p<0.001 (RTLA vs AD); †p<0.01 (RTLA vs AD); ‡p<0.001 (SD vs RTLA and AD); §p=0.001 (SD vs RTLA) and p<0.001 (SD vs AD). VAT, visual association test.


In this study we demonstrated that despite similar MTA scores, both patients with SD and those with RTLA performed significantly better on a visual episodic memory task than patients with AD.

This finding is in favour of our hypothesis that episodic memory impairment in relation to MTA may differ between various neurodegenerative disorders owing to differential involvement of anatomical pathways besides the medial temporal lobe.

We chose to subdivide the patients with FTD into SD and RTLA because of, by definition, their opposite temporal atrophy patterns and therefore possibly different performance on cognitive tests. Unlike Chan et al, however, we included our patients with RTLA on a clinical basis, as they fulfilled the Neary criteria for bvFTD or SD. We therefore avoided including subjects with other underlying pathologies such as AD, solely based on their distribution of atrophy.10 This might also explain why in our study, episodic memory impairment was not a prominent feature in RTLA. Few studies that assessed cognitive functioning in SD differentiated between subjects with predominant left or right temporal atrophy. Episodic memory tests that assessed different modalities were found to be lateralised.11 The patients with mainly left temporal atrophy demonstrated poor scores on verbal episodic memory tests while those with right temporal atrophy showed poor performance on both verbal and visual episodic memory tasks, which did not even discriminate them from patients with AD. These findings were not supported by a study in which a visual episodic memory task was equally impaired in both groups.7 This is in accordance with our study in which no relationship between laterality of MTA and performance on a visual episodic memory task was found.

Studies that showed differences in episodic memory impairment in patients with AD and FTD had different starting points. As in this study, some assessed episodic memory function at an initial visit to the memory clinic or in ‘the early stage’, others matched for MMSE or dementia severity. Although various results were reported, most studies found more severe episodic memory impairment in AD than in FTD.18–20 MTA is widely used as a biomarker in AD and episodic memory impairment in AD is associated with MTA. Although the presence of MTA has been described in a substantial proportion of patients with FTD,21 ,22 its relationship with episodic memory impairment in these disorders is less clear. Considering the similar MTA in patients with SD/RTLA and those with AD in our study, it seems that episodic memory function depends not, or not solely, on the integrity of the medial temporal lobe. This is in line with a functional study in which patients with early AD and SD underwent positron emission tomography with [18F]fluoro-2-deoxy-D-glucose tracer with co-registered MRI.19 MTA in patients with SD was equivalent to, or even more severe, than in those with AD. Episodic memory impairment was only present in patients with AD and was associated with hypometabolism in a network beyond the medial temporal lobe (MTL) with involvement of the posterior cingulate cortex. Another study found obvious episodic memory impairment in patients with FTD and this was associated with grey matter volume loss in left posterior regions and especially in the left MTL.12

We also found SM above all compromised in patients with SD and to a much lesser extent and only concerning animal fluency in patients with RTLA and those with AD. We could confirm the different cognitive profile of patients with RTLA in comparison with those with SD as measured by a VAT picture naming subtest and animal fluency and to a lesser extent by VAT scores and MMSE, in which a trend of better achievement was seen in RTLA. In most studies SM impairment was found more in left predominant SD than in predominant right SD or RTLA,7 ,10 ,23 while another study found equal involvement.11 One study made a clear distinction between SD and RTLA based on the radiologically opposite atrophy pattern and different clinical profile,10 while others suggested that both right and left predominant SD are degenerative processes that ultimately lead to a multimodal semantic loss.23 ,24 In SD SM impairment seems radiologically obviously related with left (lateral) temporal atrophy and functionally with bilateral rostral temporal lobe hypometabolism.19 In a more recent study the authors confirmed the bilateral temporal lobe hypometabolism, especially in the anterior fusiform region. They found involvement in the left and right fusiform temporal lobes which was obvious when the tests were studied in patients with respectively more left than right (verbal SM) and more right than left (non-verbal SM) temporal atrophy.24 Involvement of different parts in this semantic network may explain partly the ultimately different clinical profile of patients with SD and RTLA, which therefore needs different management strategies.

Because of the lack of patient series on SD and RTLA, the large number of patients in our study is of particular note and the match for MTA with twice the number of subjects with AD enables an appropriate comparison of episodic memory and SM performance between these diagnostic groups. In most cases we could make an adequate match for both right and left MTA. Only the right MTA score within the SD group demonstrated a non-significant difference between patients with SD and those with AD.

Our study has some limitations. We did not match for dementia severity with MMSE or Clinical Dementia Rating test scores. Clinical Dementia Rating was not available in many cases and MMSE was not used for matching because of possible language interference in the patients with SD. On the other hand, our aim was not primarily to compare patients with equal severity of dementia but to study the clinical profile of patients with equal MTA. In any case, MMSE scores were in the same range between diagnostic groups with a tendency for patients with RTLA to perform better on the MMSE. Our study was based on previously collected data, and patients were recruited over several years. Despite the limited number of psychometric tests that had been performed in all patients, three appropriate general tests for both visual episodic memory and SM were applied, together with the MMSE. We determined MTA with a visual rating scale. The visual MTA score has been shown to be of value in daily clinical practice and correlates fairly well with volumetric methods.21 ,25 ,26 Although volumetric analysis might have been more precise in measuring atrophy of the hippocampus along the axis, this would not have resolved our difficulties in matching patients with SD and RTLA with those with AD because of their still varying degrees of asymmetry.

In conclusion, episodic memory is relatively preserved in SD and RTLA in comparison with AD and this difference is unrelated to differences in MTA. Semantic memory is more affected in SD than in RTLA or AD. The role of the hippocampus and the MTL in episodic memory has to be further elucidated for the different types of dementia. The use of functional imaging may further explain the role of different networks in episodic memory and SM.


  • Funding The Alzheimer's Center VUmc is supported by Alzheimer's Nederland and Stichting VUmc fonds. The clinical database structure was developed with funding from Stichting Dioraphte.

  • Competing interests None.

  • Ethics approval The study was approved by the ethical review board of VU University Medical Center.

  • Provenance and peer review Not commissioned; externally peer reviewed.


Visit the full archive of podcasts for JNNP here >>

Free sample
This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of JNNP.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.