Global aphasia is an acquired language disorder characterized by severe loss of comprehension with concomitant deficits in expressive abilities (Peach, 2001). Unlike other syndromes of aphasia, few distinctions are found between preserved and impaired components of these patients' language. The outlook for recovery from global aphasia tends to be bleak (Kertesz and McCabe, 1977). For this reason, the term global may be more prognostic than descriptive.

Several isolated areas of relatively preserved comprehension have been identified in global aphasia. These include recognition of specific word categories (Wapner and Gardner, 1979) and famous personal and geographic names (Yasuda and Ono, 1998). Globally aphasic subjects also show relatively better comprehension for personally relevant information (Van Lancker and Nicklay, 1992).

Globally aphasic patients are most severely impaired in their expressive abilities. The verbal output of many patients with global aphasia consists primarily of stereotypic recurring utterances or speech automatisms. Recurrent utterances have been described as being either nondictionary verbal forms (unrecognizable) consisting of consonant-vowel (CV) syllables (for example, do-do-do or ma-ma-ma) or dictionary forms (word or sentence) (Alajouanine, 1956). Blanken, Wallesch, and Papagno (1990) investigated the relationship between the nondictionary forms of recurrent utterances and comprehension disturbances in global aphasia. Although recurrent utterances are frequently associated with comprehension disturbances, the overall variability in language comprehension suggests that speech stereotypes cannot be used to infer the presence of severe comprehension deficits.

Patients with global aphasia give the impression of having more preserved communicative abilities than is actually the case because of their use of the suprasegmental aspects of speech. To investigate this, deBlesser and Poeck (1985) analyzed the spontaneous utterances of a group of globally aphasic subjects with output limited to CV recurrences. They concluded that the length and pitch of these utterances were stereotypical and that the prosody of these patients did not seem to reflect communicative intent. The contributions to conversation that are credited to these patients, therefore, may be the result of the communicative partners' need for informative communication rather than the patients' use of prosodic elements to convey intent.

The efficiency of communication following global aphasia depends on the type of question that is asked (Herrmann et al., 1989). Better performance is observed for responses to yes/no questions than for responses to interrogative pronoun questions and narrative requests. Patients with global aphasia mostly use gesture in their responses to yes/no questions. The other types of questioning require increased verbal output and thus create the need for more complex communicative responses.

Patients with global aphasia rarely take the initiative to communicate or expand on shared topics (Herrmann et al., 1989). Their most frequent communication strategies are those that enable them to secure comprehension (e.g., indicating comprehension problems, requesting support for establishing comprehension). Although these individuals rely most heavily on nonverbal strategies, the efficiency of their communication may approximate that of less impaired aphasic patients while imposing nearly as low a burden on the communication partner (Marshall, Freed, and Phillips, 1997).

Global aphasia results most commonly from a cerebrovascular event in the middle cerebral artery at a level inferior to the point of branching. The majority of lesions producing global aphasia are extensive and involve both prerolandic and postrolandic areas of the left hemisphere. These include Broca's (posterior frontal) and Wernicke's (superior temporal) areas and may extend to subcortical areas, including the basal ganglia, internal capsule, and thalamus (Murdoch et al., 1986). Occasionally, the lesion is confined to anterior, posterior, or deep cortical and subcortical regions (Mazzocchi and Vignolo, 1979). Global aphasia has also been described in patients with lesions restricted to subcortical regions, including the basal ganglia, internal capsule, periventricular white matter, temporal isthmus, and thalamus (Alexander, Naeser, and Palumbo, 1987).

Ferro (1992) investigated the influence of lesion site on recovery from global aphasia. The lesions in his subjects with global aphasia were grouped into five types with differing outcomes. Type 1 included patients with large pre-and postrolandic middle cerebral artery infarcts. These patients had a very poor prognosis. The remaining four groups were classified as follows: type 2, prerolandic; type 3, subcortical; type 4, parietal, and type 5, double frontal and parietal lesion. Patients in these latter groups demonstrated variable outcomes, improving generally to Broca's or transcortical aphasia. Complete recovery was observed in some patients with type 2 and 3 infarcts. In contrast, Basso and Farabola (1997) investigated recovery in three cases of aphasia based on the patients' lesion patterns. One patient had global aphasia from a large lesion involving both the anterior and posterior language areas, while two other patients had Broca's and Wernicke's aphasia from lesions restricted to either the anterior or posterior language areas, respectively. The patient with global aphasia recovered better than his two aphasic counterparts and had an outstanding outcome. Basso and Farrabola concluded that group recovery patterns based on aphasia severity and site of lesion may not be able to account for the improvement that is occasionally observed in individual patients.

Global aphasia has the lowest recovery rate of all the aphasias (Kertesz and McCabe, 1977). When assessing the language recovery that does occur, comprehension is found to improve more than expression (Prins, Snow, and Wagenaar, 1978). Differences have been reported in the temporal patterns of recovery depending on whether the subjects were receiving speech and language treatment. For globally aphasic patients not receiving treatment, improvement appears to be greatest during the first 6 months after onset (Pashek and Holland, 1988).

Globally aphasic patients receiving treatment demonstrate substantial improvements during the first 3–6 months but also continue to improve during the period between 6 and 12 months or more after onset. Sarno and Levita (1981) observed the most accelerated improvement between 6 and 12 months after stroke. Nicholas et al. (1993) found different patterns of recovery for language and nonlanguage skills during the first year. Substantial improvements in praxis and oral-gestural expression were noted only in the first 6 months after onset, while similar improvements in auditory and reading comprehension were observed only between 6 and 12 months after onset.

The majority of patients with global aphasia will not recover to less severe forms of the disorder. Some patients, however, will improve such that the condition evolves into other aphasia syndromes, including Broca's, transcortical motor, mixed nonfluent, conduction, anomic, and Wernicke's aphasias. Occasionally, patients make a complete recovery to normal language. One apparent explanation for the variability among these patients might be the greater instability of language scores (and therefore aphasia classifications) obtained during the first 4 weeks after stroke versus those obtained after the first month post onset. McDermott, Horner, and DeLong (1996) found greater magnitudes of change in scores and frequencies of aphasia type evolution in subjects tested during the first 30 days after onset than in subjects tested in the second 30 days after onset. Aphasia tends to be more severe during the acute stage, giving observers an initial impression of global aphasia. However, globally aphasic patients who do progress to some other form of aphasia may demonstrate changes that extend into the first months after onset (Pashek and Holland, 1988). In some cases, the global aphasia may not begin to evolve until after the first month has passed. The discrepancies in these studies, therefore, do not appear to be simply the result of the time at which the initial language observations were recorded. Apparently, evolution from global aphasia is the result of a complex interaction among a number of heretofore incompletely understood variables.

Several factors have been investigated for their prognostic significance with regard to global aphasia. The patient's age appears to have an impact on recovery: the younger the patient, the better the prognosis (Holland, Swindell, and Forbes, 1985). However, numerous exceptions to this trend have been described. Age may also relate to the type of aphasia at 1 year post stroke. In the study by Holland, Swindell, and Forbes (1985), younger globally aphasic patients evolved to a nonfluent Broca's aphasia while older patients evolved to increasingly severe fluent aphasias with advancing age. The oldest patients remained globally aphasic.

Absence of hemiparesis with global aphasia may be a positive indicator for recovery. Tranel et al. (1987) described globally aphasic patients with dual discrete lesions (anterior and posterior cerebral) that spared the primary motor area. Global aphasia improved significantly in this group within the first 10 months after onset. These conclusions are tempered by the results of Keyserlingk et al. (1997), who found that chronic globally aphasic patients with no history of hemiparesis did not fare any better with regard to language outcome than did their globally aphasic counterparts with hemiparesis from the time of onset.

The radiologic findings of patients with global aphasia have also been studied to determine whether lesion patterns found on computed tomography may provide prognostic information. Although the findings have been generally mixed, Naeser et al. (1990) were able to show significantly better recovery of auditory comprehension for a group of globally aphasic subjects whose damage did not include Wernicke's area (i.e., the lesions were limited to the subcortical temporal isthmus).

Finally, it appears that a lack of variability between auditory comprehension scores and other language scores may be viewed as a negative indicator (Mark, Thomas, and Berndt, 1992). The more performance differs among language tasks, the better the outlook. Within auditory comprehension scores, globally aphasic patients who produce yes/no responses to simple questions, regardless of their accuracy, seem to have a better outcome at 1 year post onset than those who cannot grasp the yes/no format.