OUP user menu

Impoverishment of spontaneous language and the prediction of Alzheimer's disease

Annalena Venneri, Katrina E. Forbes-Mckay, Michael F. Shanks
DOI: http://dx.doi.org/10.1093/brain/awh419 E27 First published online: 23 March 2005

The article by Garrard and colleagues (2005) reported the results of the systematic linguistic analysis of Iris Murdoch's final novel published in the year before the diagnosis of Alzheimer's disease was made. Garrard and colleagues observed signs of deterioration in her writing, especially in semantic skills and sophistication of vocabulary. Subtle alterations of language were also detected by Brian Butterworth in his analysis of Ronald Reagan's speeches prior to his re-election to the USA presidency in 1984. This was 10 years before the diagnosis of Alzheimer's disease was made in his case (The Sunday Times, November 4, 1984). Garrard and colleagues referred to the study of Croisile and colleagues, who had reported impairments of spontaneous oral and written language in their heterogeneous group of Alzheimer's disease patients (Croisile et al., 1996). There is additional evidence, however, from prospective laboratory-based studies of spontaneous language in patients with minimal and mild Alzheimer's disease (Forbes et al., 2002, 2003, 2004). Simplification of grammatical structure, ineffective communication of information, failure to identify pictorial themes and loss of vocabulary acquired late in life all indicated that impairments at the level of the semantic system were present very early in the course of the disease. There was no evidence of a similar decline in normal elderly controls (Fig. 1).

Fig. 1

Mean scores achieved by elderly controls and Alzheimer's disease patients for some aspects of written language in a spontaneous writing task. (Adapted from Forbes et al., 2004.)

In contrast to other cognitive skills (e.g. memory, attention and speed of information processing), therefore, semantic abilities seem relatively resistant to the physiological decline which accompanies the process of normal ageing. The findings by Garrard and colleagues are supportive of this earlier work, but have added impact since these premonitory language deficits were demonstrable even in the case of a talented and accomplished writer like Iris Murdoch. The presence of subtle changes in the semantic aspects of spontaneous speech and writing might, therefore, represent a cognitive marker of pathological cognitive decline and help the early and differential diagnosis from normal ageing. It is not always possible, however, to look at samples of spontaneous writing over a long period before referral (Snowdon et al., 1996), since people differ in their writing and record-keeping habits and the analysis of retrospective material will be limited by the lack of standardized procedures. Contemporary assessment of language may be equally or even more effective if age- and education-based norms were available. Our findings appear to indicate that detailed analysis of the semantic aspects of language using laboratory tests can detect subtle changes at a very early stage with accuracy and high discriminatory power. There are evident advantages in the detection of preclinical Alzheimer's disease and in the screening of at-risk groups such as patients with mild cognitive impairment and those homozygous for apolipoprotein ε4.