Left with the voices or hearing right? Lateralization of auditory verbal hallucinations in schizophrenia

In an important contribution to the literature on the neural basis of hallucinations, Ait Bentaleb et al1 describe, in a patient with schizophrenia, more metabolic activity in the left primary auditory cortex and the right middle temporal gyrus during auditory hallucinations than while listening to external speech. The authors conclude this finding indicates that the hypotheses of defective internal monitoring and of aberrant auditory activation are not mutually exclusive. In addition, however, their finding raises the question of whether hallucinations might be associated with aberrant lateralization of language function. The finding of decreased cerebral asymmetry in schizophrenia has been replicated with several techniques.2 In addition, functional imaging studies have reported decreased lateralization of language-related activation in patients with schizophrenia compared with healthy controls.3 It could be hypothesized that inner speech, originating from right cerebral homologues of the language areas, is perceived as auditory hallucinations. Self-produced language activity normally leads to inhibition of language perception areas.4 When this inhibitory mechanism is failing, verbal thoughts may not be recognized as originating from the self and may erroneously be attributed to an external source. Indeed, inhibition of language perception might be more prone to failure when language activity is derived from an unusual site (i.e., from contralateral homologue areas in the right hemisphere). This hypothesis can be tested by reviewing studies meta-analytically that report functional activation in patients with schizophrenia while they are experiencing hallucinations. The following inclusion criteria were used: bilateral measurement of functional activity, right-handed patients with a diagnosis of schizophrenia, hallucination-related activity measured either in a block design or with event-related functional magnetic resonance imaging protocols. Because we are interested in laterality of the verbal component of auditory hallucinations, our analysis was restricted to activity in areas that are known to be involved in language (i.e., Brodmann areas 21, 22, 38, 39, 40, 41, 42, 44, 45 and 52). Significant hallucination-related activity was statistically integrated and compared among areas in the left and the right hemisphere using the logaritmic risk ratio method. To be less dependent on statistical thresholds of individual studies, data were also analyzed with a vote-counting method, which compares the number of language-related areas that are significantly activated in the right and left hemisphere. Five studies met our inclusion criteria (Table 1), but only 3 could be included in the risk ratio analysis (Lennox et al,6 Dierks et al8 and Shergill et al9) because one study5 provided insufficient data and the other7 did not report activity in language-related areas. The resulting mean, weighted risk ratio (left/right) was 3.42 (95% confidence interval 2.89–4.81; κ = 3; total n = 11), implying stronger activity of left hemisphere language areas. All 5 studies were included in the vote-counting analysis, which also demonstrated that more left hemisphere (n = 14) than right hemisphere (n = 7) language-related brain areas were significantly activated during auditory verbal hallucinations. Table 1 Using 2 different methods, our results showed that language-related areas in the left hemisphere were significantly more activated than the right-sided homotope regions. Thus, the hypothesis that auditory verbal hallucinations arise from a right hemisphere source of language production (inner speech) and are subsequently perceived as originating from an external source is not supported by the available evidence. Instead, these findings call for a more thorough investigation of auditory verbal imagery and language perception in relation to hallucinations.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.