There is presently evidence to suggest that even in the absence of overt dementia, sentence comprehension and the speed of lexical activation may be compromised in patients with Parkinson's disease (PD). It has also been suggested that the neuromodulatory influence of dopamine may modulate access to lexical-semantic information within semantic networks. Nonetheless, reports of compromised lexical activation in PD have been inconsistent, and the specific neuromodulatory influence of dopamine depletion on lexical access also remains unclear. In addition, whilst emerging research has suggested that slowed lexical access may be linked to the sentence comprehension difficulties typically evident in a subset of patients with PD, studies in this area are currently limited.
Therefore, the first aim of this thesis was to profile lexical-semantic function and sentence comprehension in PD patients and healthy control participants (matched to the PD group in age and education). Secondly, this thesis aimed to use a battery of online semantic priming experiments to investigate lexical activation across time in PD, and how dopamine depletion may influence this process. Finally, this thesis also aimed to delineate whether slowed lexical access contributes to sentence processing deficits in PD.
Twenty PD patients and 23 healthy controls participated in the experiments. Firstly, an offline assessment of language function confirmed the presence of noncanonical sentence processing difficulties in the PD group. Subsequently, a self-paced reading task was implemented to investigate the locus of the sentence processing deficit in PD. The results of this experiment indicated that only the PD group was delayed in their sensitivity to the critical processing region of noncanonical sentences. These findings may potentially be explained by the presence of slowed lexical access in the PD group.
In order to further investigate possible changes to the speed of lexical access in PD, an online list priming paradigm was used to chart the time course of automatic lexical activation. Analyses revealed that semantic priming effects were delayed in the PD group, relative to controls. More importantly, however, following division of the PD patients into those classified as 'good' and 'poor' comprehenders of noncanonical sentences, a significantly longer delay in lexical activation was observed for the 'poor' comprehenders. Hence, these results supported the notion that delayed lexical access in PD contributes to these patients' sentence comprehension difficulties, which may be related to dopamine depletion in the disease.
Using a multi-priming paradigm, whereby two prime words are presented prior to the target word, potential changes to other aspects of lexical activation in PD were then further investigated. Whilst semantic priming effects were evident for all conditions in the control group, an absence of priming effects across an intervening unrelated prime word was observed for patients with PD. The result was consistent with a reduced signal-to-noise ratio (SNR) within semantic networks in PD, which decreases the salience of related prime words and increases the salience of unrelated prime words. Further, such alterations to the SNR appeared to occur independently of any changes in the time course of lexical activation. Subsequent division of patients into 'good' and 'poor' comprehenders of noncanonical sentences illustrated delayed lexical activation only for the 'poor' comprehenders.
Further investigations of semantic priming in PD patients whilst 'off' levodopa medication were conducted using both the list priming and multi-priming tasks. Dopamine depletion in PD during medication withdrawal led to both delays in lexical activation as well as disruptions to semantic priming across an intervening unrelated word. These results confirmed the neuromodulatory contribution of dopamine depletion to the changes in lexical activation observed in the earlier studies. Specifically, dopamine appeared to exert a dual ncuromodulatory influence on lexical access, modulating both the temporal availability and the saliency of lexical-semantic information within semantic networks.
In the final study of this thesis, a cross-modal priming task was used to specifically investigate the presence or absence of trace reactivation in the two participant groups. Surprisingly, a significant correlation between age and trace reactivation was observed for the control participants, such that trace reactivation was only observed for the younger control participants. In contrast, trace reactivation was evident for neither the ·poor' nor 'good' PD comprehenders. Limitations in task design represent one possible explanation for this result.
To summarise, the results reported in this thesis suggest that lexical activation is compromised by changes to dopaminergic function in PD. More specifically, the onset of automatic lexical activation appears to be delayed in PD, with longer delays evident in PD patients with poor comprehension of complex noncanonical sentences. Further, a reduced SNR within semantic networks in PD appears to modulate the saliency of both related and unrelated prime words, increasing the susceptibility of lexical activation to disruption. Overall, these results illustrate that dopamine may exert a dual neuromodulatory influence on lexical access. Moreover, the results suggest that this neuromodulatory influence may be at least partially mediated through frontal-striatal systems.