The Mini-Mental State Examination (MMSE) does allow some formal testing of language, but additional detailed testing is required to gain insight into the type of underlying dementia. Listening carefully to a patient's spontaneous speech is a key aspect of the language assessment ( Table 2 ). This can be done by asking open-ended questions such as, "Why have you come to see me?" and by asking the patient to describe a complex picture, such as the well known "cookie theft" picture from the Boston Diagnostic Aphasia Examination. In a pinch, a picture from a magazine or newspaper could be used. When listening to spontaneous speech, the clinician should pay attention to articulation (clarity or distortion of speech), fluency (rate of speech, phrase length), grammatical accuracy and variety, word finding (Are there pauses to search for words? Is a good range of vocabulary used?), and prosody or melodic line.
Naming ( Table 3 ) is tested by asking the patient to name objects around the room, body parts, or line drawings, such as those in the Boston Naming Test. Naming is typically impaired in dementia, and the types of errors can be informative. Phonemic or semantic errors can be observed and suggest nonfluent or fluent progressive aphasia, respectively. Patients who describe what a word means instead of providing the name are employing circumlocution. When patients fail to recognize what an object is, they are exhibiting visual agnosia caused by the loss of semantic knowledge for the object. Circumlocution and visual agnosia can be observed in both early Alzheimer's dementia and semantic dementia; initially, this occurs with infrequently encountered words or objects, but with disease progression, it happens even with (previously) familiar words or objects.
Repetition of words and sentences (see Table 3 ) should also be tested as this can differentiate the types of aphasia seen in dementia. It also provides localizing value: impaired repetition implies a lesion in the perisylvian area, while intact repetition in conjunction with aphasia implies a lesion in the extraperisylvian area and signifies a transcortical aphasic disorder (see Figure 1).
Comprehension is tested by asking the patient to follow a series of one-, two-, and three-step commands (see Table 3 ). Additional tests in comprehension can also include asking the definition of words to test the patient's semantic knowledge or asking for an interpretation of a story by providing a short scenario followed by a question. An example of story interpretation is, "A tiger and lion fought in the jungle. The tiger was eaten by the lion. Who won?"
Reading comprehension (see Table 3 ) can be tested by having the patient read commands and follow them–similar to one of the items in the MMSE. Reading aloud can be tested by having the patient read aloud either a standardized paragraph or one from a magazine. Reading of single words should be tested if paragraph reading is not successful.
Writing (see Table 3 ) can be tested by asking the patient to write a sentence. The sentence from the MMSE should suffice. The sentence should be examined for grammar, spelling errors, phrase length, and punctuation. In addition, writing to dictation (of regular and exception words) and copying should be tested if spontaneous writing is not successful.
Patients with a progressive aphasia often have insight into their difficulties and are frustrated by their problems with communication. These difficulties can be exacerbated when they are asked to converse or to perform tests quickly. Extra time must be provided to allow these patients to communicate.
Geriatrics and Aging. 2008;11(2):103-110. © 2008 1453987 Ontario, Ltd.
Cite this: Assessment of Language Function in Dementia - Medscape - Mar 01, 2008.