From a clinical point of view, typical AD is characterised by insidious and progressive onset and a cognitive profile where the initial and most prominent deficit is impairment in episodic memory, including learning and recall of recently learned information (McKhann et al., 2011). However, a burgeoning body of research has shown that changes in connected speech (i.e., spoken language used in continuous sequence) are amongst the earliest signs of cognitive decline in AD (Forbes-McKay & Venneri, 2005; Ahmed et al., 2013; Mueller et al., 2021) and that these changes progress through the successive clinical stages of the disease (e.g., Ahmed et al., 2013; Forbes-McKay et al., 2013; Mueller et al., 2018; Mueller et al., 2021). This knowledge reveals the potential value of connected speech analysis in the early identification and monitoring of AD. Moreover, unpacking the linguistic and cognitive mechanisms underlying the breakdown of connected speech in AD is critical for the development of new communication therapies to assist patients, families. and health and care professionals during the dementia journey. Indeed, support to overcome communication difficulties is among the core recommendation included in the NICE Dementia Care Pathway (National Collaborating Centre for Mental Health, 2018). Supporting communication is one of the core principles underlying “Living Well” and “Supporting Well”, key pillars of the NHS England Dementia Wellbeing Pathway and the UK National Dementia strategy (Department of Health, 2009).
Successful production of connected speech involves simultaneous use and coordination of several linguistic (e.g., morpho-syntactic, lexico-semantic, phonetic, phonological, pragmatic) and cognitive processes (e.g., memory, attention, executive function, speed of processing). Therefore, it provides an opportunity to identify specific levels of linguistic deficits from spoken output. Recent literature reviews on characteristics of connected speech in AD point to a pattern of deficit in several linguistic levels including speech rate, syntactic structure and complexity, lexical content, semantic content and efficiency, as well as spontaneity and fluency of speech (Boschi et al., 2017; Filiou et al., 2020; Mueller et al., 2018; Slegers et al., 2018). Specifically, the key features that distinguish AD from healthy control participants are: reduced speech rate and spontaneity including increased repetitions and revisions; simplified syntax and sentence structures including shorter and grammatically simpler sentences; word finding difficulties and increased use of pronouns (i.e., over production of he, she, they, it, rather than use of specific nouns); inflectional errors in nouns and verbs (e.g., difficulty producing verb tense-play, plays, played, playing); and reduced semantic content and uninformative speech with low idea density and efficiency. The progress in the field has been encouraging, however, a significant drawback remains with regard to the diversity of languages studied, and how fragmentation of linguistic features differs across different languages (Beveridge & Bak, 2011).
Our understanding of linguistic breakdown in AD is limited as most studies have been conducted in English speaking participants, and handful of other European languages (Boschi et al., 2017; Filiou et al., 2020; Petti et al., 2020). This is far from capturing the structural and typological diversity of languages spoken in the world. Cross-linguistic research in language impairments has shown that impairments are determined by the structure of the language system (Paradis, 2001). Research indicates that features of language impairment and specific diagnostic markers in AD depend on the structure of the language itself (Bose et al., 2021; Kavé & Levy, 2003; see Auclair-Ouellet, 2015). That is, there are distinct differences in how language impairment in AD manifests itself in English compared to other languages. To illustrate, our own research has shown that Bengali-speaking AD patients produce fewer pronouns, in direct contrast with the overuse of pronouns by English-speaking AD patients (Bose et al., 2021). Furthermore, these patients did not show difficulty producing verb tenses, which is among the common diagnostic features in English-speaking patients (Ahmed et al., 2013). Similarly, Kavé & Levy (2003) reported that Hebrew-speaking AD patients produced a similar proportion of inflected words compared to controls in Cookie Theft picture description, a difference that is typically found in English-speaking AD patients (Ahmed et al., 2013; Sajjadi et al., 2012). These are not idiosyncratic findings. Rather, it highlights that the features that are impaired in AD depend on the nature of the language itself.
Several reviews on connected speech in AD has focused on findings from English-speaking patients, and none have specifically focused on linguistic diversity of AD populations. We will therefore conduct a scoping review (Munn et al., 2018) to map the literature regarding connected speech in AD across non-English languages. The specific purposes of this scoping review are: 1) to identify the breath and extent of connected speech literature in non-English speakers with AD; 2) to determine their methodological characteristics; 3) to identify impaired linguistic features currently described across different linguistic levels, and 4) to identify language-specific features.