Elsevier

Neuropsychologia

Volume 42, Issue 9, 2004, Pages 1212-1222
Neuropsychologia

Verbal fluency performance in dementia of the Alzheimer’s type: a meta-analysis

https://doi.org/10.1016/j.neuropsychologia.2004.02.001Get rights and content

Abstract

A meta-analysis of 153 studies with 15,990 participants was conducted to compare the magnitude of deficits upon tests of phonemic and semantic fluency for patients with dementia of the Alzheimer’s type (DAT) relative to healthy controls. As has been found for patients with focal temporal cortical lesions (but not for patients with focal frontal cortical lesions), DAT patients were significantly more impaired on tests of semantic relative to phonemic fluency (r=0.73 and 0.57, respectively). Thus, since phonemic and semantic fluency are considered to impose comparable demands upon executive control processes such as effortful retrieval, but the latter is relatively more dependent upon the integrity of semantic memory, these results suggest that the semantic memory deficit in DAT reflects a degradation of the semantic store. Also supporting this conclusion, confrontation naming, a measure of semantic memory that imposes only minimal demands upon effortful retrieval, was significantly more impaired than phonemic fluency (r=0.60 versus 0.55, respectively). However, since semantic fluency was also significantly more impaired than confrontation naming (r=0.73 versus 0.61), deficits in semantic memory and effortful retrieval may be additive. Semantic, but not phonemic fluency, was significantly more impaired than measures of verbal intelligence and psychomotor speed. Thus, the semantic memory deficit in DAT qualifies as a differential deficit, but executive dysfunction as indexed by phonemic fluency does not constitute an additional isolated feature of the disorder. Dementia severity was not significantly related to the relative magnitude of deficits upon phonemic and semantic fluency.

Introduction

Episodic amnesia is widely regarded as the earliest and most severe deficit associated with dementia of the Alzheimer’s type (DAT, see, e.g.; Hodges, Salmon, & Butters, 1990). Although there is a great deal of evidence that DAT is also characterised by a prominent deficit in semantic memory (Hodges, Salmon, & Butters, 1992), there remains widespread disagreement as to what this reflects. The most prominent position is that the deficit is attributable to degradation in the structure or content of semantic knowledge. Whilst some argue that semantic memory is abnormally organised but intact (Grober, Buschke, Kawas, & Fuld, 1985), others suggest that representations are actually lost (Chertkow & Bub, 1990, Hodges et al., 1992). In either case, it is presumed that impoverishment of conceptual knowledge will result, as defining attributes become inaccessible, and the associational strength between related concepts weakened.

However, an alternative position that has been suggested is that the deficit in semantic memory reflects a non-linguistic deficit in controlled cognitive processing that disturbs access to semantic knowledge from a relatively preserved semantic store (Balota & Ferraro, 1996; Bayles, Tomoeda, Kaszniak, & Trosset, 1991; Grande, McGlinchey-Berroth, Milberg, & D’Esposito, 1996; Nebes, 1989, Nebes, 1994; Nebes, Martin, & Horn, 1984; Ober & Shenaut, 1988). Since executive processes are responsible for the more supervisory, or strategic aspects of cognition (Crawford & Henry, in press, Stuss & Benson, 1986), this latter perspective suggests that the semantic memory deficit is attributable to a breakdown in the executive control mechanisms responsible for effortful retrieval. In the present study we will address this issue by applying meta-analytic techniques to the verbal fluency literature in DAT. As will be discussed, comparison of the relative magnitude of deficits on phonemic and semantic fluency may be used to draw inferences regarding the prominence of effortful retrieval deficits and semantic store degradation, and, when in the context of performance on other cognitive measures, whether these deficits qualify as differential deficits (Henry & Crawford, in press).

Tests of verbal fluency are amongst the most widely employed measures used to assess cognitive functioning following neurological damage, and involve associative exploration and retrieval of words based on phonemic or semantic criteria (phonemic and semantic fluency, respectively), usually conducted in the setting of a time constraint. Thus, whilst for phonemic fluency participants are asked to generate as many words as possible beginning with a specified letter (e.g. F), for semantic fluency search is constrained by a specified category (e.g. animals). These measures are considered to impose comparable demands upon executive or supervisory processes because both require efficient organisation of verbal retrieval and recall, as well as self-monitoring aspects of cognition (the participant must keep track of responses already given), effortful self-initiation, and inhibition of responses when appropriate (Crawford & Henry, in press, Rosser & Hodges, 1994). However, whilst phonemic fluency requires the creation of search strategies based primarily on lexical representations, tests of semantic fluency require searching for semantic extensions of a target superordinate, and thus depend intrinsically upon the integrity of semantic associations within the lexicon (Rohrer, Salmon, Wixted & Paulsen, 1999). Deficits on tests of semantic fluency may therefore reflect problems with semantic memory, and not executive dysfunction.

In a recent meta-analytic review that included 31 studies and 1791 participants, Henry and Crawford (in press) investigated the relative magnitude of cognitive deficits upon tests of phonemic and semantic fluency for patients with focal cortical lesions. The pattern of results strongly suggested that whilst the two types of fluency impose comparable demands upon executive functioning, semantic fluency is relatively more dependent upon the integrity of semantic memory. Henry and Crawford (in press) found that focal frontal lobe injuries were associated with equivalent phonemic and semantic fluency deficits (r=0.52 and 0.54, respectively). Whilst it is important not to treat the terms ‘frontal’ and ‘executive’ as interchangeable, there is nevertheless a great deal of evidence that frontal structures are particularly implicated in executive functioning (see; Stuss & Benson, 1986), and thus comparable impairment upon tests of phonemic and semantic fluency for patients with focal frontal cortical injuries suggests that the two measures impose equivalent demands upon executive control processes such as effortful retrieval. However, semantic fluency was more impaired following focal temporal damage (r=0.61), and this deficit was substantially larger than the corresponding phonemic fluency deficit (r=0.44). Since there is a great deal of evidence that temporal structures are the neural substrates particularly responsible for semantic memory this was presumed to reflect the greater reliance of semantic fluency upon the integrity of semantic memory. Comparison of the relative magnitude of deficits on phonemic and semantic fluency may therefore be used to draw inferences regarding the prominence of effortful retrieval deficits and semantic store degradation, respectively (see also; Rohrer et al., 1999).

Thus, of particular importance in the debate regarding whether the semantic memory deficit in DAT reflects a degradation of the semantic store or retrieval difficulties, has been the pattern of disproportionately impaired semantic relative to phonemic fluency often reported (Butters, Granholm, Salmon, Grant, & Wolfe, 1987; Monsch et al., 1992, Monsch et al., 1997, Rohrer et al., 1999). Whilst most studies of DAT have found both phonemic and semantic fluency to be significantly impaired, a deficit for the latter is more consistently reported, and is typically the more prominent of the two. Monsch et al. (1997) for instance, found that whilst both semantic and phonemic fluency were significantly impaired, when these values were converted to standard scores the deficit for the former was nearly twice as large.

However, the opposite deficit pattern, or equivalent performance on both fluency measures has also, although less frequently, been reported (Nebes et al., 1984; Ober, Dronkers, Koss, Delis, & Friedland, 1986; Suhr & Jones, 1998) suggesting that the semantic memory deficit associated with DAT may instead reflect a disruption of the executive control processes responsible for effortful retrieval, and not a degradation of the semantic store (Henry & Crawford, in press). Indeed, there is considerable evidence that executive deficits are also an important feature of the cognitive deterioration in DAT (Amieva, Phillips, Della Sala, & Henry, in press) and that these deficits typically occur early in the disease, and may be the first non-memory deficits to occur (Perry, Watson, & Hodges, 2000).

To the present authors’ knowledge, the current paper is the first to apply meta-analytic techniques to compare performance upon phonemic and semantic fluency in DAT. The results of this study will be particularly useful because they emphasise the magnitude of effects. Although researchers are strongly encouraged to report effect sizes for their individual studies (American Psychological Association, 2001), this is rarely done in practice, yet is far more informative than simply reporting whether a particular effect is significant or not. Moreover, because using meta-analysis it is possible to integrate effects across studies that differ in both the participants sampled and methodology employed, the effects can be considered to be very reliable, robust estimates of the corresponding parameters of interest. Thus, an effect’s generalisability can be subjected to a level of scrutiny not possible in a single study, and with a level of objectivity and methodological consistency that is difficult to achieve in non-quantitative reviews (Stanley, 2001).

The first aim was to derive effect size estimates for phonemic and semantic fluency for patients with DAT relative to healthy controls. Although there is a trend for semantic fluency to be more impaired than phonemic fluency, inconsistencies have been noted. Thus, it will be extremely informative to derive reliable estimates of the magnitude of the effects associated with each of these measures. As has been discussed, comparison of the relative magnitude of each will permit an assessment of whether the semantic memory deficit associated with DAT predominantly reflects a disrupted/degraded semantic store, or difficulties with effortful retrieval.

The second aim was to estimate effect sizes for other cognitive measures in order to provide comparison standards, and thus assess to what extent fluency deficits in DAT qualify as differential deficits. Of particular importance was to address the possibility that phonemic and semantic fluency deficits reflect a more general deficit in verbal intelligence (see; Miller, 1984). Thus, the pattern of deficits across fluency versus verbal intelligence as measured by the WAIS (Wechsler, 1981) verbal and vocabulary scales (VIQ) will be compared. Vocabulary has been permitted to contribute to the construct of VIQ because it is the scale that correlates most highly with the Total score on the VIQ scale; the average correlation across nine age-ranges was 0.85 (Wechsler, 1981). It will also be investigated whether deficits on semantic and phonemic fluency are in excess of deficits on the Trail Making Test Part A (TMT A; Reitan, 1990), a widely used measure of psychomotor speed; this will address the possibility that deficits on tests of verbal fluency reflect generalised slowing (i.e. a reduction in cognitive speed). Performance on tests of phonemic and semantic fluency will also be compared with scores on the Boston Naming Test (BNT; Kaplan, Goodglass, & Weintraub, 1983), a measure of semantic memory that imposes only minimal demands upon effortful retrieval, and Wechsler and Stone’s (1974) Immediate Logical Memory (ILM) test of episodic memory.

Finally, a third issue relates to whether the relative prominence of retrieval deficits and semantic store degradation is comparable at each stage of the disease. Lack of consistency across studies in terms of the relative prominence of deficits upon tests of phonemic and semantic fluency has been attributed to differences in dementia severity between patient groups (Butters et al., 1987; Margolin, Pate, Friedrich, & Elia, 1990). It has, for instance, been suggested that whilst executive deficits are present in the early stages of the disorder, they become more pronounced as the illness progresses (Duke & Kaszniak, 2000), which would predict that deficits on the two measures should become more comparable with increasing disease severity. To address this possibility, correlational analyses will be conducted to test whether the difference in the relative magnitude of semantic and phonemic fluency deficits is related to dementia severity, as measured by the mini mental state exam (MMSE; Folstein, Folstein, & McHugh, 1975).

Section snippets

Sample of studies

A computer-based search involving the Web of Science, Psych Lit CD-ROM, and Science Direct databases was undertaken, using the following terms; ‘letter fluency’, ‘FAS’, ‘semantic fluency’, ‘category fluency’, ‘controlled oral word association’, ‘COWA(T)’, ‘word fluency’ ‘verbal fluency’, ‘oral fluency’, ‘phonemic fluency’, ‘executive test’ and ‘frontal test’. In addition, a manual search of issues of Neuropsychologia, The Journal of the International Neuropsychological Society, Neuropsychology,

Participant characteristics

153 research articles published between 1983 and 2002 contributed to the present study (a complete list of the papers included in this meta-analysis, can be found by following the link at http://www.abdn.ac.uk/∼psy299/dept/). In total data from 8356 DAT patients and 7634 controls were included in these research articles. Patients and controls were closely matched for age (M=71.60, S.D.=4.77 versus M=69.66, S.D.=7.26, respectively) and gender (43.46% versus 42.95% male, respectively), but

Storage degradation versus retrieval slowing

For both phonemic and semantic fluency the presence of DAT was associated with mean deficits that were significantly different from zero (P<0.001), and large in practical importance according to Cohen’s (1977) criteria. However, for studies that assessed both measures, the semantic fluency deficit was significantly larger. As discussed earlier, it is thought that, relative to phonemic fluency, semantic fluency imposes greater demands upon the integrity of the semantic network, but that the two

References (50)

  • C Back-Madruga et al.

    Functional ability in executive variant Alzheimer’s disease and typical Alzheimer’s disease

    Clinical Neuropsychologist

    (2002)
  • D.A Balota et al.

    Lexical, sublexical, and implicit memory processes in healthy young and healthy older adults and in individuals with dementia of the Alzheimer type

    Neuropsychology

    (1996)
  • K.A Bayles et al.

    Alzheimer’s disease’s effects on semantic memory: loss of structure or impaired processing?

    Journal of Cognitive Neuroscience

    (1991)
  • H Braak et al.

    Evolution of neuropathology of Alzheimer’s disease

    Acta Neurologica Scandinavica

    (1996)
  • N Butters et al.

    Episodic and semantic memory: a comparison of amnesic and demented patients

    Journal of Clinical and Experimental Neuropsychology

    (1987)
  • A.S Chan et al.

    An assessment of the semantic network in patients with Alzheimer’s disease

    Journal of Cognitive Neuroscience

    (1993)
  • H Chertkow et al.

    Semantic memory loss in dementia of the Alzheimer’s type: what do various measures measure?

    Brain

    (1990)
  • R.F Coen et al.

    Letter and category fluency in Alzheimer’s disease: A prognostic indicator of progression?

    Dementia

    (1996)
  • Cohen, J. (1977). Statistical power analysis for the behavioral sciences (Revised ed.). New York: Academic...
  • Craik, F. I. M. (1986). A functional account of age differences in memory. In F. Klix & H. Hagendorf (Eds.), Human...
  • Crawford, J. R., & Henry, J. D. (in press). Assessment of executive deficits. In P.W. Halligan & N. Wade (Eds.), The...
  • L.M Duke et al.

    Executive control functions in degenerative dementias: A comparative review

    Neuropsychology Review

    (2000)
  • B Giffard et al.

    The hyperpriming phenomenon in normal aging: A consequence of cognitive slowing?

    Neuropsychology

    (2003)
  • B Giffard et al.

    The dynamic time course of semantic memory impairment in Alzheimer’s disease: Clues from hyperpriming and hypopriming effects

    Brain

    (2002)
  • T Gomez-Isla et al.

    Profound loss of layer II entorhinal cortex neurons in very early Alzheimer’s disease

    Journal of Neuroscience

    (1996)
  • Cited by (585)

    View all citing articles on Scopus
    View full text