Nonword repetition in schizophrenia: Semantic long-term memory supports phonological loop


  • Field: Schizophrenia–Cognitive Psychology.

MANY STUDIES HAVE indicated an intact phonological loop (PL) as a subsystem of working memory (WM) in people with schizophrenia (PWS).1

Some WM tasks are more dependent on the PL than others, such as serial recall tasks.1 However, using a word serial recall test and a tone delay discrimination test, Bruce et al. showed one subgroup of PWS with a selective deficit in verbal memory.2

It is suggested that semantic long-term memory has an active role in the recall of short-term memory.3 There is a difference in semantic demand of verbal tasks such as forward digit span and nonverbal tasks, like the tone delay task. So it is assumed that the semantic aspect of mentioned tasks may play an important role in identifying any potential PL deficit.

The nonword repetition task (NRT) has been used in various studies to examine the PL in children with specific language impairment and other language impairments.4,5

To investigate PL by NRT, we enrolled nine chronic male PWS [mean age = 34.3y, standard deviation (SD) = 9.15] and ten age- and educationally-matched healthy men (mean age = 35.11, SD = 8.60). No speech or hearing impairment was reported in their recordings. All PWS were taking atypical antipsychotics for at least 6 months. Sixty-nine phonotactically similar monosyllabic nonwords in Azerbaijani language were used as task stimuli.

Using the Mann–Whitney test, a significant difference was found for the nonword repetition error numbers between both groups (P = 0.001). The PWS made more errors in repetition of the nonwords (mean = 48.88, SD = 13.57) than normal participants (mean = 19, SD = 11.01).

This finding is contrary to previous studies that used forward digit span tasks that indicated intact PL in PWS.1 But it is consistent with the findings of Wexler et al. and Stevens et al., which have shown dysfunctions in the left inferior frontal cortex in PWS who have specific performance deficits in verbal but not nonverbal WM.2,6

The NRT without semantic load revealed the deficits of PL in PWS.

As Green et al. have shown improvement of verbal WM and other cognitive functions in PWS who use risperidone, we suggest that further analysis of the effect of the drugs on the deficit is needed.7