Clinical manifestations

Presentation and course

Aprosodia presents as poor comprehension or expression of affective or nonaffective prosody. Aprosodia may manifest acutely, such as following a cerebrovascular accident, or, more rarely, as primary progressive aprosodia (96). Importantly, any pathological process affecting the neural substrates of prosody can result in dysprosodic features. Gorelick and Ross suggested that the affective aprosodias have a similar recovery pattern as the aphasias (99). For example, these authors documented a case of global aprosodia that quickly evolved into motor aprosodia. However, this pattern of recovery has not been observed consistently. Please refer to the Prognosis and Management sections for further information regarding the recovery and treatment of aprosodia.

Ross proposed that the subtypes of affective aprosodias mirror those of the aphasias (215; 219). This theory has received support from clinical-anatomical correlates (99; 217). However, a systematic relationship between the type of aprosodia and lesion location has not always been found among individuals with right-hemisphere damage (281). Baum and Pell provide a review of this issue (15). The proposed variants of aprosodia are motor, sensory, global, conduction, anomic, transcortical motor, transcortical sensory, and mixed. Note that the following aprosodic subtypes are all affective in nature.

Motor aprosodia. Motor aprosodia refers to the impairment of spontaneous expression of prosody and gesturing with relatively spared comprehension of prosodic speech. These patients appear to have flat affect. Gorelick and Ross provided evidence for the speculation that lesions resulting in motor aprosodia are analogous to lesions producing motor Broca’s aphasia (ie, the right frontal operculum) (99). Motor aprosodia has also been seen in patients with brainstem and right-sided basal ganglia lesions (63; 116).

Sensory aprosodia. Sensory aprosodia is characterized by spared spontaneous prosodic expression with poor repetition and comprehension of prosodic speech (288). This disorder has been observed following infarction of the temporoparietal region of the right hemisphere (99) as well as lesions to the right thalamus and posterior limb of the internal capsule (285).

Global aprosodia. Global aprosodia presents as impairments in the comprehension, repetition, and expression of prosody, and there are few accounts of it in the literature. Global aprosodia has been seen with perisylvian frontoparietal infarction (99). A somewhat similar symptom profile has been documented in a patient with frontotemporal dementia with a behavioral variant showing atrophy and hypometabolism in the temporal and frontal lobes, more remarkable on the right side, using MRI and PET scans, respectively (69).

Conduction aprosodia. Conduction aprosodia was first reported by Gorelick and Ross (215; 99; 288). Conduction aprosodia is the inability to repeat prosodic utterances in the setting of preserved spontaneous expression and comprehension of prosodic speech. Similar to conduction aphasia, conduction aprosodia is observed in patients with extensive lesions involving the temporal and parietal lobes. This is an extremely rare disorder and is mentioned primarily for conceptual completeness.

Anomic aprosodia. Anomic aprosodia refers to poor comprehension of emotional gesturing (eg, limb movements and facial expression) with sparing in other areas of prosodic communication (288).

Transcortical motor aprosodia. Transcortical motor aprosodia is characterized by impaired production of affective prosody, with spared imitation and comprehension. Patients with lesions to the superior lateral surface of the right frontal lobe have been documented to present with transcortical motor aprosodia (259).

Transcortical sensory aprosodia. Transcortical sensory aprosodia presents as normal spontaneous prosody and imitation but involves poor comprehension of prosodic speech. It has been observed following damage to the right posterior neostriatum and posterior limb of the internal capsule (99).

Mixed aprosodia. Mixed aprosodia is marked by poor spontaneous prosody and prosodic comprehension but good repetition. Heilman and colleagues suggest that mixed aprosodia would follow damage to the right superior-medial frontal region (109).

Of note, aprosodias are often simplified in the literature as either “expressive” or “receptive” (287).

Expressive aprosodia. Expressive aprosodia encompasses aprosodic subtypes involving deficits in prosody production (eg, motor and transcortical motor aprosodias).

A systematic review of right-hemisphere lesion locations associated with affective aprosodia showed that damage to frontal right-hemisphere regions, especially the inferior frontal lobe and insula, were more consistently related to expressive than receptive deficits (76).

Receptive aprosodia. Receptive aprosodia, in turn, encompasses aprosodic subtypes involving deficits in prosody comprehension (eg, sensory and transcortical sensory aprosodias). Sheppard and colleagues showed that the right posterior superior temporal gyrus in the right hemisphere ventral stream is important for affective speech recognition (245). In a second paper, Sheppard and colleagues characterized subtypes and neural correlates of receptive aprosodia in acute right-hemisphere stroke (246). Impaired receptive emotional prosody was predicted by greater subcortical damage, particularly to the caudate, and it was concluded that receptive aprosodia can result from impairments at different processing stages (eg, at the level of acoustic analysis or analysis of the abstract representation of acoustic characteristics that convey emotion). Durfee and colleagues also revealed that damage to posterior right-hemisphere regions, especially in the temporal lobe, basal ganglia, thalamus, and amygdala, were more consistently related to receptive than expressive affective aprosodia (76).

Cases of right-hemisphere damage have also revealed that associated aprosodia depends on recovery stage (76). That is, whether the infarct is in the acute, subacute, or chronic stage may influence the type and extent of prosodic deficits. For example, patients who exhibit aprosodia in the acute stage may not show the same deficits in the subacute or chronic stages despite no change in lesion (70; 245). Because affective aprosodias are often observed following damage to the right hemisphere, they are likely to be accompanied by neglect disorders (110), anosognosia (255), agraphia (221), and left-sided motor or sensory weakness (116). Dysarthria (281) and amusia (64) have also been observed in patients with aprosodia. A systematic review examining the co-occurrence of cognitive and communication disorders with aprosodia revealed that receptive linguistic aprosodia may be related to neglect disorders and amusia, whereas receptive affective aprosodia was related to interpersonal communication deficits and may be related to the ability to process affective semantics and facial expressions but not to unilateral neglect (247). This suggests that different types of aprosodia (receptive vs. expressive and affective vs. nonaffective) may be accompanied by different concomitant cognitive, communicative, or motoric disorders.

Prognosis and complications

Prognosis. The prognosis for affective aprosodia likely depends on the observed deficit profile (288; 149) and etiology (eg, focal brain damage or psychiatric disorder). At least some degree of recovery can occur (222), but decline has also been observed (173). There is some evidence, primarily from case reports, to suggest that the existence of a progressive neurodegenerative disease can affect affective prosody. Ghacibeh and Heilman described a case study of a woman with progressive expressive affective aprosodia accompanied by amusia and loss of automatic speech, yet she retained receptive affective prosody as well as nonaffective (ie, syntactic) expressive and receptive prosody (96). Based on the association between clinical findings and MRI findings showing right frontal cortical atrophy, the authors proposed that this woman exhibited the right-hemisphere equivalent of primary progressive aphasia. Graff-Radford and colleagues reported another case of progressive affective prosody in a 64-year-old woman whose neurologic and neuropsychological examinations were otherwise unremarkable (100). A PET scan revealed focal hypometabolism in the right superior medial frontal lobe, and an MRI revealed mild generalized cerebral volume loss without specific lobar distribution. Most recently, a case study was reported in which a 58-year-old man exhibited significant impairment in the expression of emotional prosody along with prosopalgia, aka “trigeminal neuralgia” (14). The condition’s etiology was not clear, and the man reportedly did not meet criteria for any specific neurodegenerative disease but had “striking focal atrophy” in the perisylvian regions (right more than left). The existence of these progressive neurodegenerative affective aprosodia syndromes appears to be quite rare, and reports are limited to just a few case studies at this point in time. Research on the presence of impaired prosody in dementia has revealed mixed findings both within and across different forms of dementia (ie, dementia due to Alzheimer disease, Parkinson disease, Huntington disease, frontotemporal dementia, or vascular dementia), as well as in mild cognitive impairment. A review of this literature concluded that the variable research findings are a result of small sample sizes, inconsistent inclusion criteria (eg, degree of cognitive impairment present), and perhaps, most importantly, heterogeneous assessment methods used across studies (162). Although longitudinal studies of prosody in neurodegenerative diseases have not been reported, given the general progressive decline in dementia, it is reasonable to conclude that prosodic abilities would likely follow a trajectory of decline similar to that seen for the cognitive and behavioral abilities in these syndromes.

How well intervention strategies target the underlying mechanism of etiology is an important factor in prognosis (89; 77). In progressive neurodegenerative diseases, such as those cited above, such cases are rare and there has been no follow-up to inform prognosis. However, given the progressive course of the underlying disease, it is likely that the outcome will be poor. Further, there have been no reports of interventions for any of these disorders. In cases of aprosodia arising from neurologic insult, both the rate and degree of recovery likely depend on lesion site and size as well as the amount of time since the initial insult (70; 173; 245). Ross and Monnot note that, as with aphasia, affective prosodic deficits are more severe when both cortical and subcortical lesions are present, and recovery is more limited than when only cortical lesions are present (223). Full recovery has been reported. For example, 40% to 50% of patients with right-hemisphere stroke who show symptoms of aprosodia within 10 days after stroke may not show any prosodic deficits in the chronic stage (70; 245). Following phenytoin treatment, a patient with a 1-month history of expressive aprosodia and amusia (associated with right temporo-occipital seizure) returned to baseline function, though, due to a lack of experimental control, it is unclear whether recovery was directly related to the pharmacologic intervention or to spontaneous recovery (17). In one case of receptive aprosodia following right hemisphere ischemic stroke, a patient was verbally presented with several readings of the same sentence every day (eg, “I never said he took my money.”), each reading with a different word vocally stressed. Just as young children sequentially master parts of speech (eg, nouns, then verbs, and then pronouns), the patients recovered their ability to perceive inflection for nouns, verbs, and pronouns in the same ordinal sequence (59). In aprosodia occurring from multiple sclerosis, the greatest deficits in affective prosody have been shown to occur in individuals with long disease duration and severe physical disability (18). However, a study by Kraemer and colleagues has shown that deficits in affective prosody can occur in individuals with only a short disease duration and low physical disability, suggesting that in multiple sclerosis, affective aprosodia can begin early and possibly progress over the disease duration (138; 122). Meta-analysis of prosody in autism spectrum disorder, Williams syndrome, and Down syndrome suggests that research on expressive and receptive emotional prosody in autism spectrum disorder demonstrated mixed results, with some studies showing impairment and others demonstrating no significant differences with age-matched peers; Williams syndrome literature indicates that receptive prosody in patients with Williams syndrome is less accurate than in typically developing peers but exceeds the accuracy of IQ-matched peers; Down syndrome literature suggests global prosodic deficits in patients with Down syndrome but that receptive abilities for emotional prosody tend to be better than expressive abilities (153). In aprosodia that follows from autism spectrum disorder, a 6-step form of imitative treatment, previously utilized to treat aprosodia in adults with acquired deficits, was effective in an adolescent (04). A systematic review of prosody interventions for autism spectrum disorder found moderate-to-large improvements when evidence-based treatments were used to directly target prosody (117). However, the duration of these treatment effects was not examined and, thus, remains unclear. Additional interventions are discussed in the Management section of this article. Other research suggests that premorbid IQ is inversely correlated with aprosodia, as has been shown in a sample of patients with multiple sclerosis, suggesting that cognitive reserve may protect against aprosodic decline (94).

Prosodic recovery can be complex. Changes in overall functional profiles have been reported, and prognosis for prosodic recovery may depend on which hemisphere is primarily damaged. Although it has been suggested that both cerebral hemispheres are necessary for prosodic perception, right hemisphere damage has been shown to result in a greater decline, particularly in affective prosodic processing (284; 98; 32). A meta-analysis by Witteman and colleagues showed that both linguistic and affective prosodic perception decline similarly following left-hemisphere damage, whereas right-hemisphere damage primarily impairs affective prosodic perception. Impaired affective prosodic perception is greater in right as opposed to left hemisphere damage (284). A systematic review by Blake and colleagues did not find a reliable difference between right- and left-hemisphere damage for comprehension and production of linguistic prosody nor production of emotional prosody; however, they did find that right-hemisphere damage was more likely to result in impaired comprehension of emotional prosody than left-hemisphere damage (Blake and colleagues 2021). Sheppard and colleagues examined predictors of both impairment and recovery of the ability to identify emotional prosody in individuals with acute (within 48 hours) right-hemisphere stroke (245). They found that greater damage to the right posterior superior temporal gyrus (independent of total lesion volume) and older age were predictive of greater impairment and that damage to the right amygdala was associated with impaired fear recognition. Follow-up assessment of nine patients (of 23 who participated in the acute assessment) at 6 to 12 months revealed that two patients improved over time, three patients remained impaired, and three remained unimpaired at follow-up; one patient who was not impaired acutely was impaired at follow-up. This latter finding could not be attributed to expansion of the initial infarct or to a new infarct. These findings were consistent with prior research implicating the right hemisphere in the recognition of emotional prosody. Nonetheless, patients with left-hemisphere stroke sustain impaired expressive prosody compared to healthy controls, regardless of concurrent aphasia (149). Nakhutina and colleagues observed that patients with left-hemisphere-lesions tended to improve in affective prosody, whereas right-hemisphere-damaged individuals significantly declined (173). In a sample of 14 patients with right-hemisphere stroke, Gorelick and Ross observed sufficient recovery of affective-prosodic repetition and changes in functional abilities to warrant reclassification of the aprosodia subtype (99). Although longitudinal studies of emotional prosody deficits following stroke are lacking, Sihvonen and colleagues reported some data that inform recovery (249). Using an affective prosody task using one-word utterances, 24 of 39 patients in their study were classified as aprosodic (based on performance falling two standard deviations below that of healthy controls) in the subacute timeframe, whereas 16 patients were aprosodic at the 3-month assessment. They also found that affective prosodic performance was correlated with music perception (r = 0.66 and r = 0.62 at subacute and 3-month assessments) and that damage to the right inferior fronto-occipital fasciculus was associated with prosodic and music perception deficits.

Research on the prognosis of affective aprosodia is limited, in part, due to the range of possible etiologic causes as described above. However, affective aprosodia is less likely to be formally assessed than other aspects of communicative impairment (eg, aphasia). Indeed, a survey found that, although speech-language pathologists are more likely to diagnose aprosodia than other rehabilitation professionals, most speech-language pathologists use informal observation to assess prosody rather than objective measures (206). In their survey, Ramsey and Blake found that only 50% of clinicians screened for aprosodia, and they point out that, although observation may be sufficient to detect expressive aprosodia, it is unlikely to detect receptive aprosodia (206). Alba-Ferrara and colleagues discussed the value of using intractable epilepsy to study affective aprosodia, which could further inform prognosis based on functional reorganization of the brain following surgical treatment of seizures (05). Ultimately, increased awareness of affective aprosodia and more longitudinal research are necessary to better understand its prognosis.

To address the lack of available research on recovery following acquired prosodic deficits, Hillis proposed a study to assess recovery of affective prosodic deficits following stroke that would address several of the limitations in the current research literature (NIH Project Number 5R01DC015466-05). The goals of the research were “to integrate (1) detailed longitudinal analysis of the impaired perceptual, cognitive, and motor processes underlying prosody in each patient at four time points over the first year after stroke, (2) detailed longitudinal analysis of the structural and functional lesions (eg, in right ventral and dorsal streams) and functional connectivity between critical regions in each patient at the same four time points over the first year after stroke, and (3) analyses of variables (such as timing and doses of antidepressants) and co-morbidities (such as depression) that influence recovery.” Further efforts to study the impact of right hemisphere damage are underway through the development of a shared database of resources called the RHDBank (161) as well as through the creation of a Right Hemisphere Damage Working Group (RHDWG), part of the Evidence-Based Clinical Research Committee of the Academy of Neurologic Communication Disorders and Sciences. Sheppard and colleagues reported that this latter group “was established to identify and distinguish core, co-occurring deficit patterns of cognitive-communication deficits following right-hemisphere stroke reported in the literature by producing systematic reviews and meta-analyses to describe gaps in the extant literature and to form recommendations for future research” (247). Thus, although at the time of this update, minimal longitudinal research has been published to inform the prognosis of emotional prosody deficits, future research appears promising in this area.

Complications. Individuals with affective aprosodia may experience difficulties in several domains, including social and occupational functioning, due to their emotional communication deficits (258; 111). Aprosodia can result in difficulty understanding another person’s emotional state and generating an appropriate response; this can impair interpersonal relationships and, in particular, be mistaken for apathy or a lack of empathy (287). For instance, in patients with right hemisphere strokes, the ability to decode the meaning of prosody and facial expressions has been associated with decreased marital satisfaction (34), and caregivers report aprosodia as an important problem (113). Caregivers of other patient populations with emotion recognition deficits have reported increased burden and depression (156). For those returning to work, residual deficits in affective prosody can disrupt interpersonal efficacy with colleagues and supervisors (289). Aprosodia can also affect interactions in medical contexts. Wymer and colleagues note that "Aprosodia may also interfere in medical diagnosis and treatment; patients may be unable to accurately convey their mood state, and difficulty with prosodic comprehension may limit the benefit of didactic instruction and psychotherapies" (288; 238). Deficits in affective prosody early in a disease, for example, multiple sclerosis, can be especially challenging. Good communication and stability in interpersonal relationships can be critical for individuals coping with the early stages of a chronic disease. Again, poor communication and misunderstandings can disrupt interpersonal relationships and can challenge effective therapy (138; 56). Furthermore, in some cases, aprosodia may be accompanied by facial expressivity deficits (for example, comorbidity may be present in certain neuropsychiatric disorders) (39; 43). For individuals living with multiple compromised communication channels, social functioning is likely to be particularly challenging. Social anxiety has been associated with a tendency to misinterpret social information, particularly neutral faces and voices, more threateningly (191). In addition, other symptoms that may co-occur with aprosodia due to a nondominant hemisphere lesion include hemineglect, other cognitive symptoms (eg, decreased alertness and attention), severe personality and emotional changes (eg, increased irritability and apathy), and psychotic symptoms (35). Moreover, other communication deficits that might co-occur with aprosodia following right-hemisphere damage include dysarthria, impairments in discourse comprehension and production, and impaired pragmatics (31; 140), which could exacerbate the disruptive effects of aprosodia on interpersonal relationships and quality of life. In patients with Parkinson disease, speech prosody impairment, together with symptoms of REM sleep behavior disorder, predict cognitive decline (209).

Based on a systematic review, receptive emotional aprosodia, in particular, as opposed to expressive emotional aprosodia, is associated with deficits in emotional facial recognition, interpersonal interactions, and emotional semantics (247). Receptive linguistic aprosodia is associated with amusia and hemispatial neglect.

In sum, aprosodia can substantially negatively impact an individual’s recovery, well-being, and overall quality of life by interfering with social and emotional functioning, and concomitant cognitive-communication impairments may exacerbate this effect.

Clinical vignette

The following vignette is a fictitious example of an aprosodia case study designed to present a “snapshot” of how a bilingual and multicultural patient with aprosodia might present, emphasizing the impact of aprosodia on social interactions, mood, and quality of life.

The patient is a 52-year-old, right-handed, native English- and Spanish-speaking woman born in Santiago, Chile. She was raised as a native bilingual as both languages were fluently spoken in the home by her parents, and her family often interacted with individuals in the community who spoke both languages. The patient also understands some Mapudungun because her maternal grandparents identified as Mapuche and resided with the family during her childhood. The Mapuche group represents an indigenous people primarily located in the countries of Argentina and Chile. When the patient was 15 years of age, her parents relocated to the United States, where she attended an international high school. She received her bachelor’s degree in biochemistry and spent her junior year in a study-abroad program in Spain. The patient is married with two children, and the family communicates primarily in spoken English, although both children are proficient in Spanish. She has been employed for the past 18 years as a pharmacy assistant. Medical history is significant for type 2 diabetes, hypertension, and hyperlipidemia. Family history is notable for stroke (paternal grandfather). Her current medications include metformin, furosemide, and rosuvastatin.

On February 15, 2020, the patient awoke with “loss of feeling” in her left arm and face and with speech difficulties. On attempting to stand, she collapsed and lost consciousness. A family member transported the patient to the emergency department of a local hospital, where a cerebral angiogram revealed an occlusion in the M1 distribution of the right middle cerebral artery. Despite tissue plasminogen activator administration, symptoms persisted, and mechanical thrombectomy was performed. The patient was transferred to an acute rehabilitation hospital for comprehensive rehabilitation therapies. Following 2 weeks of inpatient rehabilitation, she was discharged to her home. The patient continued to complete outpatient physical, occupational, and speech therapies, but she was unable to return to work.

At the 6-month follow-up, the family reported changes in mood, social interactions, and behavior (ie, apathy, nervousness, and agitation). The patient was referred for neurologic evaluation, which was conducted in spoken English. Her score on the Mini-Mental Status Exam (85) was 25/30, which was lower than expected. Brief self-report screening measures revealed symptoms of mild depression and moderate anxiety. Brain MRI with FLAIR contrast revealed scattered hyperintensities within the regions of the right frontal lobe, insula, and internal capsule. In light of these behavioral and cognitive findings, the patient was referred for neuropsychological assessment to an American-born bilingual neuropsychologist with linguistic fluency in both of the patient’s primary languages and familiarity with South America's cultures. She underwent a comprehensive evaluation consisting of a clinical interview and collateral interviews with her wife and the older son, testing attention, memory, processing speed, executive functioning, language, and visuospatial abilities and assessing social cognition, mood, anxiety, and apathy. All tests were conducted in English, but additional speech and language measures were administered in Spanish during a subsequent session.

During the clinical interview, the patient’s family reported ongoing reduced affect; mood changes, including apathy, irritability, and nervousness; indifference to family and home life; and poorer quality of social interactions with them. They described several incidents in which the patient had become very irritated with the family, stating that they were “ridiculing” her and “se estaban burlando de ella [were making fun of her]” when the conversations actually involved the typical joking manner the entire family used with each other throughout the day. The patient became visibly agitated when questioned as to her emotional and cognitive functioning during the interview and denied any difficulties other than a slight weakness of the left hand. Formal testing revealed deficits in visuospatial processing and nonverbal executive functioning as well as elevated apathy, mild depression, and moderate anxiety. Assessment of social cognition was largely within expected ranges with the exception of difficulty identifying emotional tone in both spoken English and Spanish. On observation, the patient’s speech was at expectation with regards to articulation, rate, pitch, rhythm, and volume.

Given the evaluation findings and family report, difficulties with the patient’s affective prosody recognition were suspected. The patient was referred to an ongoing research study examining affective prosodic comprehension and production. As part of this study, she completed several facial and prosodic perception tasks and was video-recorded while being asked to imitate and produce sad, happy, fearful, angry, and neutral sounds and sentences. These productions were analyzed on a 5-point Likert scale by a number of native English-speaking adult judges, blind to the possible diagnosis. Results indicated intact rate, pause time, pitch, and loudness in affective and nonaffective free speech as well as normal accuracy on affective facial perception tasks. However, the patient was only about 20% accurate in her ability to identify prosodic speech and to differentiate among five emotion categories (ie, happiness, sad, anger, fear, and neutral) and in her ability to repeat affective prosodic speech. Based on the research norms, her ability to produce spontaneous affective prosody was slightly below the lower end of the average range. Given her language background, she also was administered the linguistic and emotional prosody tasks in Spanish from the Protocolo MEC: Protocolo para la Evaluación de la Comunicación de Montréal (83). The performance profile on these latter tasks was similar to the research-based tasks presented in English, except that her comprehension and repetition performance on the Protocolo MEC emotional prosody tasks were approximately 10% and 16% accurate, respectively, compared to 20% in English. Based on these findings, the patient was diagnosed with impairment in the comprehension and repetition of affective prosodic speech. This profile most closely aligns with sensory aprosodia, as per Gorelick and Ross’s categorization system (99). Although the slight reduction in spontaneous affective prosody is not typical for this subtype, it may have been an artifact of testing in English given that her performance on the Protocolo MEC was typical for her age. Based on these findings, several treatment options were suggested to the patient and her family. Emotion-focused cognitive behavioral therapy, which is designed to target the development of emotion comprehension and emotion regulation skills, was the primary recommendation (260; 196). The various activities of ECBT facilitate the patient’s understanding of emotional experiences and the ability to process emotional context and emotional cues when listening to others. As part of ECBT, emotionally provocative vignettes are used in which the patient listens to a story presented orally and considers which emotions the protagonist is most likely to be feeling (eg, happy or sad) in the particular situation; the patient is required to explain why the protagonist may be feeling that way and how the protagonist might know that they are feeling that way (ie, emotion understanding). There is almost no literature directly examining ECBT with adults from South American Spanish-speaking countries, but several studies include small subsets of self-identified Latinx participants. Although the patient primarily uses fluent spoken English in her day-to-day interactions at work and in the local community, it was believed that she would benefit from ECBT with a bilingual therapist.

Additional recommendations included (1) music therapy to aid the patient’s understanding of auditory nuances, such as rhythm and semantics, which, in turn, might improve comprehension of emotional responses (269; 11; 234) and (2) additional speech therapy techniques specifically targeting both inferencing (171; 30) and figurative language comprehension (155; 154). To target inferencing, the patient is presented with pictures or stories and guided to describe what is happening and to make inferences or identify the gist. To target metaphor comprehension, the patient undergoes structured semantic training whereby they are presented with visuals to explicitly identify the semantic components of a metaphor and are guided to make associations between concepts and ultimately interpret the metaphor (154).

The patient subsequently participated in 4 months of ECBT. She also completed 3 months of speech therapy focused on making inferences and figurative language comprehension. At a 3-month follow-up, she showed a 20% increased ability to repeat prosodic speech in English and a marked increase in the ability to recognize orally-presented (recorded or live), happy-toned speech. Although most of the work was conducted in spoken English, the bilingual therapist was able to provide some intervention and further assessment in Spanish. On re-administration of the Protocolo MEC at the end of treatment, the patient was demonstrating a similar level of improvement as with English-based prosodic speech. Her anxiety resolved, but she continued to experience mild symptoms of depression. Further, her spouse reported persistent strain on their marriage due primarily to continued misinterpretation of social language cues on the part of the patient. Her wife also noticed that the patient’s ability to codeswitch between languages was more laborious and that she experienced some continued mild difficulties tracking conversations that included both languages. Thus, the patient and her wife will subsequently participate in family therapy to help them build compensatory communication skills and provide the spouse with supportive strategies.

For published case studies describing individuals with other types of aprosodia, also see Calvo and colleagues (53), Dara and colleagues (69), Deonna and colleagues (73), Stringer and Hodnett (259), Confavreux and colleagues (64), Cohen and colleagues (63), Heilman and colleagues (109), Cercy and Kuluva (58), and Russell and colleagues (228).

Biological basis

Etiology and pathogenesis

Aprosodia can occur with cortical damage to either the left or right hemisphere (48; 237; 185; 223; 35; 219; 32) or with subcortical damage (54; 76). Several published reviews address affective prosodic deficits in brain-damaged individuals (37; 106; 108; 38; 292; 76; 256). Because damage to brain structures associated with prosodic processing can result from a number of factors, the etiology of aprosodia is varied. Vascular insult, however, is thought to be the most common cause of aprosodia (62). Although the research literature has associated primarily right-hemisphere lesion with affective aprosodia, Leung and colleagues have found that those with left-hemisphere damage due to stroke also can have altered affective prosody, both in production and interpretation (149). Gorelick and Ross, for example, found that the incidence of aprosodia in cases of right-hemisphere stroke was equal to that of aphasia in left-hemisphere stroke (12 instances of aprosodia among 14 individuals with right-sided stroke; 12 instances of aphasia in 15 individuals with left-sided stroke) (99). It has also been demonstrated that sensory aprosodia can serve as an indicator of acute infarction in the inferior division of the right middle cerebral artery (70).

Although much of the aprosodia literature focuses on stroke, tumor, or other focal injuries, affective prosodic difficulties have also been well documented in more diffuse neurologic conditions. Prosodic deficits have been reported in congenital and developmental neurologic disorders, such as autism spectrum disorder (188; 101), Friedrich ataxia (143), Down syndrome (144), and Williams syndrome (194). They have also been reported in neurodegenerative diseases, such as Huntington disease (253), multiple sclerosis (138), Parkinson disease (67; 253; 291; 51; 12), Alzheimer disease (262), frontotemporal dementia (96), multi-infarct dementia (189), and primary progressive aphasia (213) and in psychiatric disorders, such as depression (92; 236; 179; 278), schizophrenia (87; 43; 225; 79; 150; 123; 47; 195), bipolar disorder (172), and posttraumatic stress disorder (86). Of note, affective prosodic deficits have been associated with structural anomalies, such as agenesis of the corpus callosum (181). Further, deficits or difficulties with affective prosody have also been observed in seizure disorders (17; 58; 05) and in neurotypical aging adults (177; 84). Of note, event-related potential studies have found irregular early waveforms in schizophrenia (195), major depressive disorder (179), and Williams syndrome (194).

Concerning substance use, alcohol consumption has been linked with affective prosodic deficits. Direct alcohol exposure and prenatal exposure, with and without resultant fetal alcohol syndrome or fetal alcohol effects, have been reported to be associated with affective prosody comprehension deficits (167; 166; 252; 49). A study comparing 17 patients with severe alcohol use disorder, 16 with Korsakoff syndrome, and 19 healthy controls found significant deficits in decoding affective prosody, particularly for negative emotions, for both the severe alcohol use disorder and Korsakoff syndrome groups when compared to healthy controls. There were no significant differences between the severe alcohol use disorder and the Korsakoff syndrome participants (49).

Because multiple perceptual, cognitive, and expressive processes contribute to the reception and expression of prosody, the pathology underlying aprosodia varies as a function of the location, severity, and extent of cortical and subcortical damage among affected individuals. As such, prosodic dysfunction can vary greatly among patients, manifesting as a disorder with widespread behavioral changes or presenting as a relatively isolated deficit (248; 245). Consequently, extensive neuropsychological testing is needed to characterize the extent and type of functional impairment (ie, the extent to which expression vs. comprehension is affected and the degree to which affective or nonaffective prosody is compromised). The development of diagnostic tools to identify the exact acoustic, emotional, or cognitive breakdown in the comprehension or production of prosody could also improve characterizing the impairment. For example, Wright and colleagues proposed a cognitive architecture of receptive emotional prosody that breaks down the comprehension of emotional prosody into stages for which they developed tasks to evaluate performance at each stage (286). Sheppard and colleagues updated this model, which describes three main stages: acoustic analysis, analysis of abstract representation of acoustic characteristics that convey emotion (ARACCE), and semantic representation access (246). Ongoing development of these types of models and tasks will allow for the development of more precise diagnostic tools and better characterization of impairments.

Neuroanatomically, affective aprosodia has been strongly linked to right-hemisphere dysfunction (215; 107; 106; 108; 33; 183; 212; 220). Lesion and imaging studies have reported specific areas within the right hemisphere that are necessary for processing affective prosody; these include the anterior portion of the right temporal lobe (54; 246) and the mid- and anterior portions of the right superior temporal lobe (282; 82). An MRI study on the impact of lesion load on affective prosody after right-hemisphere stroke found that the greatest impact on affective prosody was caused by infarct in the right inferior frontal gyrus, supramarginal gyrus, or associated white matter tracts, predominantly right inferior fronto-occipital fasciculus, superior longitudinal fasciculus, and uncinate fasciculus (180). It has been suggested that an emotional prosody network within the right hemisphere involves the supramarginal gyrus, superior temporal gyrus (245), inferior frontal gyrus, and middle temporal gyrus (145). Although affective processing is strongly thought of as a right-hemisphere function (prosody included), affective aprosodia has been reported following damage both to the left and right hemispheres (54; 185; 149). Further, additional research has shown that the neuronal areas involved in affective prosody are not restricted to only the temporal lobes. Affective aprosodia has occurred following damage to the right and left basal ganglia (54; 217; 48; 218; 03; 271; 219; 246), left, right, and midline white matter regions (57), and right frontal regions (108; 223; 282). A group of 12 patients with right-hemisphere stroke evidenced greater receptive prosody and emotional facial expression recognition impairments as a function of greater damage to the caudate nucleus of the basal ganglia (246). Another recent lesion-symptom mapping study showed that affective receptive aprosodia was associated with damage to the right basal ganglia (249). Of note, Van Lancker and Sidtis found that regardless of lesion side, when relying on acoustic cues, deficits were observed among both left- and right-hemisphere-damaged groups (270). For a more comprehensive representation of affective prosodic deficits in clinical populations, several reviews are available for further reading (37; 106; 38; 292; 220).

As illustrated above, there is considerable conflicting evidence regarding the pathophysiology of affective aprosodia. Similarly, the theoretical underpinnings of affective aprosodia remain in dispute. Ross and colleagues argue that affective prosody is a lateralized and dominant function of the right hemisphere and that it is functionally organized in a manner similar to propositional language (ie, words and syntax) within the left hemisphere. Accordingly, as seen in the aphasic syndromes (95), Ross posits that the neurofunctional organization of the right hemisphere is analogous to the language functions of the left hemisphere (215). This theory suggests that anterior lesions of the right hemisphere would be associated with prosodic production deficits whereas comprehension deficits would be associated with posterior lesions (222; 215; 223). A systematic review of literature reporting receptive and expressive affective aprosodia associated with right hemisphere lesions found that damage to right frontal areas was more often associated with deficits in expression of affective prosody and to right temporal and subcortical regions more often associated with deficits in recognition of affective prosody (76), validating earlier observations by Ross and colleagues (215; 223). Durfee and colleagues also noted that, in line with these observations, expressive aprosodia may result from damage to the dorsal stream and receptive aprosodia from damage to the ventral stream (76), resembling the dorsal and ventral streams of language expression and comprehension in the left hemisphere (112; 88). In contrast, Van Lancker and Sidtis theorize that affective prosody is not lateralized to a single hemisphere; rather, prosodic processing comprises multiple skills and functions distributed throughout the brain (270; 163).

This notion that "affective prosody is functionally and anatomically organized in the right hemisphere in a manner analogous to propositional language in the left hemisphere" (223) frequently appears in the literature (283; 109) and merits further comment. It is observations consistent with this locationist view that have led to aprosodia classifications analogous to traditional aphasia classifications. However, it should be noted that findings of affective prosodic deficits following certain types of left-hemisphere damage might be reconcilable with the view that affective prosody is a dominant and lateralized function of the right hemisphere. Ross and Monnot note that if the neurologic basis of affective prosody resides in the right hemisphere, prosodic processes must necessarily be connected with the left hemisphere if they are to modify speech (223). Therefore, according to Ross and Monnot, left-hemisphere lesions, especially those involving the deep white matter that connects the hemispheres, may disrupt the "integration of affective prosody [processed in the right hemisphere] with the propositional aspects of language [processed in the left hemisphere]” (223; 219). Commenting on the debate surrounding affective prosody lateralization, Pell states that there is a “growing consensus that the right hemisphere’s participation in emotional prosody constitutes a relative rather than an absolute dominance in this processing domain” (185). Given the current body of research, it is reasonable to suggest that the processing of affective prosody occurs in the right hemisphere. This information is then transferred to the left hemisphere via deep white matter pathways of the right and left hemispheres (including the corpus callosum) where it is then integrated within the linguistic areas of the left hemisphere. Any lesion that damages the affective processing area in the right hemisphere or the network of fibers connecting the right to the left hemisphere language areas could result in an affective prosody deficit. Also, approaching prosody from an integration perspective, Sidtis and Van Lancker Sidtis argue that both affective and nonaffective prosodic behaviors result from a widespread neuroanatomical network that includes recruitment of several neuronal areas involved in perception, motor behavior, and organization (248). This view is well supported across the literature (283; 109; 185; 223; 282) and is also consistent with more contemporary models of language processing in the left hemisphere that view language processing as supported by several interconnected networks and pathways rather than separate regions devoted to discrete language functions (112; 199). Such a large neuronal network allows for a variety of neuroanatomical injuries to result in the varying presentations and severity seen in the aprosodias (248). This view exposes the limitations of the traditional aprosodia and aphasia classification systems, as many patient symptom profiles do not neatly fit within these traditional classifications. A more constructionist approach to understanding the functional organization of prosody and aprosodia could result in more precise diagnoses and, subsequently, more personalized interventions.

In regard to nonaffective prosodic processing, it does not appear to be localized exclusively to either the left or right hemisphere, and nonaffective aprosodia can result from both left- and right-hemisphere damage as well as from subcortical damage (279; 16; 93; 282; 219). For example, nonaffective prosody has been shown to be affected in Broca’s aphasia resulting from left-hemisphere stroke (02). Geigenberger and Zeigler argue that the right hemisphere is central to nonaffective prosody, whereas the left hemisphere interfaces with those elements necessary for comprehension of linguistic material (93), an argument that to some extent parallels Ross and Monnot’s model of affective aprosodia (223). However, work in right-hemisphere-damaged participants by both Emmorey and Behrens indicates that linguistic prosody at the word level, specifically emphatic and lexical stress, is not a function exclusively or even primarily of the right hemisphere, as right-hemisphere-damaged participants were either mildly impaired or had no impairment in these areas (80; 19). Similarly, a systematic review and meta-analysis of literature synthesizing work that compared prosodic abilities of participants with right-hemisphere damage to participants with no brain damage concluded that right-hemisphere damage had no reliable effect on word- and phrase-level linguistic prosody or production of “speech act” prosody (ie, grammatical prosody to distinguish speech acts, such as interrogatives from imperatives) (256). In contrast, they found that right-hemisphere damage appeared to affect discrimination and identification of speech-act prosody and emotional prosody. A meta-analysis of 21 studies examined the effects of right and left hemisphere damage on the production and comprehension of linguistic and emotional prosody (268). The meta-analysis revealed that there was not strong evidence for lateralization of linguistic prosodic production or comprehension nor for emotional prosodic production. Some evidence suggests that comprehension of emotional prosody is more impaired following right compared to left hemisphere damage. However, this was not statistically significant (268). The authors highlight the methodological heterogeneity across studies and the small number of empirical studies that met quality review criteria and provided sufficient information for inclusion in the meta-analysis as limiting factors (268). In their lesion-symptom mapping study, including both right- and left-hemisphere stroke survivors, comprehension of linguistic prosody was associated with the right inferior frontal gyrus, Rolandic operculum, insula, basal ganglia, and limbic system (249).

Some lateralization models of prosody focus less on the affective or nonaffective distinction and instead emphasize hemispheric differences in processing the physical or acoustic subcomponents of prosody. The “functional” or “task-dependent” theory, posits that the lateralization of prosody occurs on a continuum, such that the more affective the cue is, the more it will lateralize to the right hemisphere, whereas the more nonaffective the cue, the more it will lateralize to the left hemisphere (Van Lancker 1980; 219). Alternately, other researchers have proposed an “acoustic” or “cue dependent” model, which posits lateralization based on left hemisphere specialization for timing or temporal aspects of prosody and right hemisphere specialization for pitch or spectral aspects (27; 68; 210; 270). In support of this model, production and comprehension of temporal prosodic stimuli, such as duration of pauses, word stress timing, and phrase boundaries (243), are generally considered left-hemisphere functions. Similarly, we would also assume left hemisphere recruitment while listening to linguistically complex prosodic material (eg, speech with high articulatory demands, like a “tongue twister”) (192). However, evidence is mixed regarding tone localization, suggesting a more complex functional division of labor. Investigations of native speakers of tonal languages (eg, Mandarin and Thai) illustrate this complexity. In tonal languages, pitch variations convey the affective component of language and are also essential in distinguishing word meaning. If all aspects of prosodic pitch processing were lateralized to the right hemisphere, right-hemisphere damage would be expected to affect both affective and nonaffective aspects of language in native speakers of tonal languages. Yet, relative sparing of propositional language has been reported in a group of right-hemisphere-damaged native Mandarin speakers, whereas affective prosody was severely compromised (120). Conversely, among speakers of tonal languages, left-hemisphere damage seems to result in deficits in identifying word meaning (90) and in the production of tonal phonemes (178). Furthermore, Gandour and colleagues noted right-hemisphere activation for intonational tasks (eg, pitch contours) in native English speakers; however, left- and bilateral-hemisphere activation was seen for the same tasks in native Chinese speakers. This suggests that left-hemisphere involvement in intonation may vary as a function of how important tonal qualities are to word meaning (91). Moreover, Pihan and colleagues observed bilateral activation in pitch perception tasks in healthy adults (193). Work by Ross and colleagues has also challenged the acoustic model; on the basis of observed reductions in pitch-intonation ability in individuals with both right and left hemisphere damage, they conclude that the modulation of prosodic stress is distributed across both hemispheres and “fairly resistant to focal brain damage” (219). In addition to the functional anatomic relationships described above, a sample of more specific findings is provided in the table below. Studies are grouped by hemisphere, intrahemispheric anatomy (where applicable), and research findings. For further review of the neural mechanisms underlying both affective and nonaffective prosody, see Baum and Pell (15). Sidtis and Van Lancker Sidtis also provide a critical review of lateralization models of nonaffective prosody and a discussion of the role of subcortical regions in both affective and nonaffective prosody (248). A meta-analysis of neuroimaging studies on linguistic and affective prosody (20) found mixed support for hemispheric lateralization for both types of prosody, finding aprosodic involvement with areas in both hemispheres. The authors found that affective prosody was lateralized to the superior temporal sulcus in the right hemisphere (RH), middle temporal gyrus (RH), inferior frontal gyrus pars orbitalis in the left hemisphere (LH), and thalamus (LH). Linguistic prosody was lateralized to the ventral section of the inferior frontal gyrus pars opercularis (RH and LH), middle frontal gyrus (RH and LH), angular gyrus (RH), claustrum (RH), and superior temporal gyrus (LH).

Table 1. A Summary of Research Studies on Affective and Nonaffective Prosody

Affective Prosody
Hemisphere	Specific Region	Experimental Condition	Reference
Right	Right inferior frontal gyrus, supramarginal gyrus, right inferior fronto-occipital fasciculus, superior longitudinal fasciculus, and uncinate fasciculus	Affective prosody production	(180)
Right	Supramarginal gyrus, superior temporal gyrus, inferior frontal gyrus, and middle temporal gyrus	Literature review of right-hemisphere functions	(145)
Right	Supplementary motor areas	Affective prosody production	(109)
Right	Right frontal regions and right basal ganglia	A review of expressive and comprehension deficits	(212)
Right	---	Processing of pitch (receptive)	(244; 210)
Right	---	Processing of affective speech (evidence from comprehension or expression deficits)	(110; 267; 222; 215; 132; 39; 43; 33; 282; 137)
Right	Right superior-medial frontal cortex	Affective but not syntactic aprosodia	(109)
Right	Right frontal operculum	Acute deficits in spontaneous production of affective prosody	(223)
Right	Right frontal operculum	Affective repetition deficits	(215; 99; 54; 223)
Right	Right temporal operculum	Poor comprehension of affective prosody	(223; 137)
Right	Right anterior temporal pole	Deficits in processing pitch (receptive) and in repetition (expressive)	(287)
Right	Right thalamus damage	Deficits in affective prosodic comprehension	(285; 223)
Right	Right anterior putamen	Affective prosodic comprehension deficits	(223)
Right	Right middle cerebral artery (inferior divisional infarction)	Sensory aprosodia has been proposed as a specific marker of right middle cerebral artery infarction	(70)
Right	---	Comprehension and expression of emotions from prosodic and facial cues	(69)
Right	Posterior superior temporal gyrus, right amygdala	Comprehension of affective prosody	(245)
Right	Basal ganglia (caudate nucleus)	Comprehension of emotions from prosodic and facial cues	(246)
Right	Basal ganglia, thalamus, inferior frontal lobe pars opercularis, frontotemporal, arcuate fasciculus, temporoparietal, ventral stream white matter tracts (ie, inferior longitudinal fasciculus, inferior fronto-occipital fasciculus), angular and supramarginal gyri	Right-hemisphere loci for treatment response targeting receptive prosody following right-hemisphere stroke	(77)
Right	Frontotemporal	fMRI comparison of hemispheric activation for emotional prosody in a neurologically healthy sample	(242)
Right	Middle and inferior frontal lobe, insula, parietal lobe (esp. inferior), temporal lobe (esp. superior), anterior occipital lobe, amygdala, basal ganglia, thalamus, internal capsule	Systematic literature review of articles reporting right-hemisphere lesions associated with receptive aprosodia	(76)
Right	Frontal lobe (esp. inferior), insula, anterior and inferior parietal lobe, anterior and superior temporal lobe, occipital lobe (region unspecified), basal ganglia, inferior frontal occipital fasciculus, superior longitudinal fasciculus, uncinate fasciculus, internal capsule, corona radiata	Systematic literature review of articles reporting right-hemisphere lesions associated with expressive aprosodia	(76)
Right	Basal ganglia, putamen, globus pallidum, caudate	Identify the emotion in which a single word is spoken	(249)
Left	---	Affective prosody comprehension deficits	(235; 241; 264; 270; 48; 183; 03; 60; 139; 149; 140)
Left	---	Deficits in affective production of monosyllables	(149)
Left	Left "mid-white" matter (lesions below supplementary motor area and adjacent to anterior cingulate)	Deficits in spontaneous production and repetition of affective prosody	(223)
Left (by design*)	Inferior frontal gyrus orbitalis	Comprehension and production of both semantic meaning and affective prosody	(157; 21)
Bilateral: right > left	Mid and anterior right temporal lobe; bilateral inferior frontal cortex	Comprehension of affective prosody	(282; 82)
Bilateral	Right fronto-parietal operculum, bilateral frontal pole, and left frontal operculum	Affective prosody comprehension	(03)
Bilateral	Anterior cortical damage	Affective prosody deficits	(48)
Bilateral	Basal ganglia and right temporal lobe	Affective prosody deficits	(54)
---	Suprasylvian region	Expression of affective prosody	(216)
Bilateral: right greater than left	Right greater than left perisylvian region, right frontal	Deficits in affective prosody expression	(14)
Deep white matter regions	Deep white matter adjacent to corpus callosum	Deficits in spontaneous production and repetition of affective prosody	(217; 57)
White matter	Right sagittal stratum	Deficits in comprehension of attitudinal prosody, specifically sarcasm	(71)
White matter	Anterior four-fifths of corpus callosum	Production of affective prosody	(136)
Bilateral basal ganglia	Basal ganglia	Expression of affective prosody	(54; 217; 218; 219; 48; 03; 271; 212)
Bilateral	Right superior temporal sulcus and middle temporal gyrus, and left inferior frontal gyrus pars orbitalis and thalamus	Affective prosody	(20)
Bilateral	Pars orbitalis, amygdala, and anterior insula	Affective prosody	(242)
Nonaffective Prosody
Hemisphere	Specific Region	Experimental Condition	Reference
Right	---	Monotone speech characterized by poor articulation resulting from decreased acoustic energy	(132)
Right	---	Processing nonaffective speech--production of prosodic cues signaling phrase boundaries	(243)
Right	---	Processing of nonaffective speech--intonation and lexical tones	(91; 58)
Right	Inferior frontal gyrus, Rolandic operculum, insula, putamen, globus pallidum, caudate, amygdala, thalamus	Identifying word stress patterns	(249)
Left	---	Processing temporal aspects of prosody	(27; 210)
Left	---	Processing temporal linguistic aspects of prosody, eg, phase boundaries and pause durations	(192; 243)
Left	---	Processing pitch and word contours	(91; 277)
Left	---	Comprehending nonaffective prosody in single words	(80; 19)
Left	Broca’s area	Abnormal expression of prosodic patterns and timing	(02)
Bilateral	Basal ganglia	Expression of linguistic elements of prosody	(271)
Bilateral	Cortical and subcortical damage	Nonaffective aprosodia	(279; 16; 93; 282)
Bilateral	Bilateral middle frontal gyrus, claustrum, and inferior frontal gyrus pars opercularis, right angular gyrus, and left superior temporal gyrus	Linguistic prosody	(20)
* The left hemisphere was chosen by investigators as a region of interest.

Epidemiology

Information about the epidemiology of aprosodia is limited, and the prevalence of aprosodia has been difficult to quantify as a consequence of variability across populations. Gorelick and Ross suggest that affective aprosodia might be as common as the aphasias (99). Their study found that the incidence of aprosodia in cases of right-sided stroke (12 of 14 cases) was equal to that of aphasia in left-sided stroke (12 of 15 cases).

Other studies have estimated that 50% to 80% of adults may exhibit communicative deficits after right hemisphere damage (26; 65; 76) and that 30% of right hemisphere stroke cases display expressive aprosodia specifically (28; 29). Following right hemisphere stroke, individuals with affective aprosodia are also more likely to have subcortical structure damage (246). Specific training protocols have been developed to address affective aprosodia due to right-hemisphere stroke (77). Overall, vascular injury is thought to be the most common cause of aprosodia (62).

Aprosodia can co-occur with psychiatric disorders, such as schizophrenia (79), bipolar disorder (272; 06), and posttraumatic stress disorder; and with neurologic disorders or diseases, such as autism spectrum disorder (188; 251; 151; 274; 78), attention deficit disorder (176; 135; 169), multiple sclerosis (138), and amyotrophic lateral sclerosis (08; 22; 55). In psychiatric disorders that include psychosis, such as schizophrenia, deficits in prosody appear to occur in the early stages of psychosis. According to Bonfils and colleagues, deficits in affective prosody may be evident before a psychotic episode and could be a predictor of risk for psychosis (36). This supports the notion that aprosodia is a feature of psychosis (52).

Affective prosody deficits can also still be present when psychiatric conditions such as schizophrenia and bipolar disorder are symptomatically remittent (115), and there is evidence in posttraumatic stress disorder (174; 142) that long-term childhood trauma during the developmental periods of social cognition may result in affective prosody deficits. In bipolar I disorder, impairments in emotional prosody are still significantly present even during symptom remittance (114). A study by Mulcahy and colleagues highlights the possibility that increased introspective awareness (sensing of internal body signals) in autism spectrum disorder can mitigate deficits in emotional prosody (170). Relatedly, a study evaluating emotional prosody perception among college students with autism spectrum disorder found this social skill to be a relative strength (23). However, Icht and colleagues’ meta-analysis concluded that among individuals with autism spectrum disorder and low-support-needs, perceptual accuracy was only commensurate with typically developing peers when emotions were simple (eg, happy or sad) but was significantly worse for more complex affect (eg, envy or boredom) (121).

In addition to psychiatric and neurologic disorders, prosodic deficits have been examined across cultures, and the factors of sex and age have also been investigated. Yudiarto explored the occurrence of aprosodia among native Indonesian speakers. In a study of 15 Indonesian patients after right hemisphere cerebrovascular accident, 14 displayed prosodic deficits, manifesting most commonly as motor aprosodia (290).

Bozikas and colleagues and Vaskinn and associates found deficits in prosody perception most commonly in men with schizophrenia (46; 273), whereas other studies report no effect of sex (182; 224).

Several studies have illuminated age differences, suggesting that older adults show greater deficits in processing affective prosody (177; 164). Interestingly, one study found that in older adults (60 to 85 years of age), “a female perception benefit was less pervasive” (240). Additionally, men with co-occurring schizophrenia show greater deficits in processing affective prosody with age (205; 115).

Prevention

In general, aprosodia is associated with other neurologic and psychiatric disorders. Therefore, following suggestions for reducing the risk of these conditions would likely reduce the likelihood of developing aprosodia. For instance, because stroke is believed to be the most common cause of aprosodia (62), following guidelines for improved cardiovascular health may reduce the risk of aprosodia. Because both alcohol users and children subjected to prenatal alcohol exposure have demonstrated affective prosody comprehension deficits (167; 166; 252; 74; 49), limited alcohol consumption may reduce the risk of acquiring these deficits within these populations. Deficits in affective prosody have also been shown to be present not only in heavy, long-term alcohol users but also in binge drinkers (119). When neurosurgical intervention (eg, resection of lesion site in temporal lobe epilepsy) involves brain regions associated with affective prosody (105), the inclusion of prosodic language and affect assessment into language-mapping protocols may reduce the risk of postsurgical aprosodia.

Differential diagnosis

Associated or underlying disorders

Many disorders are marked by aprosodia or have features that resemble aprosodia. Furthermore, differential diagnosis pertaining to aprosodia is often focused on distinguishing psychiatrically-based versus neurologically-based aprosodia. Although constricted affect, including diminished prosody, is often interpreted by clinicians and lay public alike as psychiatric or psychological in etiology, these symptoms can also be attributes of a neurologic condition (219). Flat and affectless speech can be symptomatic of psychiatric disorders, including depression, schizophrenia, and posttraumatic stress disorder (86; 142; 190), or neurologic disorders, such as Parkinson disease (291), or the behavioral variant of frontotemporal dementia (69; 175; 276; 239). Notably, however, individuals with a pure aprosodia lack clinically significant psychiatric symptoms (eg, psychosis). Thus, formal psychological and psychiatric evaluations can differentiate between psychiatrically-based versus neurologically-based aprosodia. A psychological and psychiatric evaluation can also help determine if an underlying mood disturbance is present with a behavioral profile that may also include flat affect and emotional communication deficits (219). Additionally, a comprehensive review of the onset and progression of the prosodic disturbance can provide insight as to its etiology. For example, if arising from stroke or other acquired brain injury, aprosodia is likely to have a sudden onset that coincides with the insult, whereas individuals with psychiatric illness are more likely to have a longstanding history of aprosodia. In contrast, individuals with brain lesions or tumors that progress slowly may have aprosodic symptoms that are delayed or absent, as the brain has been able to reorganize during the incremental progression (219).

Distinguishing aprosodia from dysarthria can be particularly difficult. Although prosodic abnormalities may also be present in dysarthria, aprosodic symptoms rarely occur in isolation (214); rather, they are present within the context of other dysarthric symptoms, such as slurred speech due to weakness or reduced coordination of articulatory musculature, changes in resonance related to velopharyngeal dysfunction (ie, abnormal function of the velopharyngeal valve, which is comprised of the velum/soft palate and/or the pharyngeal walls), and reduced laryngeal control (265). Differential consideration should also be given to the possibility of an organic voice disorder unrelated to a neurologic or psychiatric syndrome. For example, temporary or permanent damage to the superior laryngeal nerve, which controls the muscle responsible for higher-pitched phonation, typically results in a limited pitch range and may affect both affective and nonaffective prosodic variation (211). Such disorders are familiar to laryngologists and speech-language pathologists specializing in voice, which may be helpful in making a differential diagnosis. Finally, other conditions can result in problems with prosodic perception and expression. For example, one study found that children with cochlear implants performed significantly worse on measures of perception and production of intonation than children with typical hearing (165; 81). Functional and structural neuroimaging may also assist in the differential diagnosis of aprosodia, though these resources are often limited in the clinical setting.

Diagnostic workup

There are no laboratory tests to definitively diagnose aprosodia. This disorder is most often diagnosed by a neurologist, neuropsychologist, or speech-language pathologist based on clinical judgment of the patient's spoken language. According to Wymer and colleagues, Ross designed the first formal assessment procedure for diagnosing affective aprosodia (215; 288). Completed at bedside, this assessment examined the patient's ability to produce and understand emotionally laden speech as well as correctly repeat emotionally laden sentences after presented by the examiner. It also measured the patient’s ability to recognize affective gestures.

Researchers have used additional assessment measures of affective aprosodia in experimental studies, including:

• Aprosodia Battery (217) • Battery of Emotional Expression and Comprehension (54) • Florida Affect Battery (33; 45) • Montreal Protocol for the Evaluation of Communication (128) • New York Emotion Battery (44; 42) • Test for Rating Emotions in Speech (24)

Contact restriction in response to the Sars-Cov-2/COVID-19 pandemic has led to the adaptation of prosodic assessment measures that accommodate testing via remote platforms (ie, video conferencing), with preliminary evidence to suggest their comparable validity and reliability (23).

For a comprehensive review of experimental and clinical measures developed to assess affective aprosodia and other emotional processing functions in neurologic populations, see Borod, Tabert, Santschi, and Strauss (42). See also the article by Kalathottukaren and colleagues for a review of behavioral assessments of nonaffective and affective prosodic abilities in adults and children (129). Although perceptual, rater-based, qualitative assessment is typically used to evaluate affective aprosodia, quantitative evaluation using acoustic measures can also be employed (13; 184; 275; 231; 24; 14). To evaluate nonaffective aprosodia, quantitative acoustic measures are typically used (184; 229; 213). Findings suggest that using simpler verbalizations (eg, those comprised of monosyllabic words) may have greater utility in evaluating aprosodia as producing these utterances is easier for patients with aphasia. Thus, aphasia severity is less likely to confound prosodic ability. Additionally, subjective and acoustic measures had greater sensitivity in distinguishing between the stroke group and healthy controls for the parts of the testing battery that used these simpler verbalizations (149).

Of note, socioeconomic or cultural factors may influence the presentation of affective prosodic difficulties. Some evidence suggests that affective prosody varies by socioeconomic level, with lower socioeconomic status associated with poorer performance, regardless of neurologic status (07). This raises a question about the utility of formal, non-culturally adjusted prosodic testing. Age may also affect the presentation of affective prosody. Specifically, older patients with brain damage seem to have more difficulty comprehending affective prosody than younger patients (224).

Although a phenomenologically based diagnosis, aprosodia should be considered a “focal” neurologic sign, prompting a neurologic work-up. Most of the pathologic processes that produce lesions causing aprosodia (eg, stroke) can easily be demonstrated with conventional brain scanning. When possible, diagnosis should be made on the basis of formal neurologic and neuropsychological testing in conjunction with clinical judgment and neuroimaging data. Notably, a survey of common practices in the evaluation of social cognition following brain injury suggests that there is a paucity of formal assessment of aprosodia used, even among clinical neuropsychologists and speech-language pathologists (131). As previously mentioned, a survey of speech-language pathologists, who are the most likely rehabilitation professionals to diagnose aprosodia, revealed that only 50% of respondents said they screened for aprosodia as part of their speech and language evaluations and that most speech-language pathologists use informal observation rather than standardized assessments, which may not be adequate to fully characterize prosodic deficits, especially for receptive aprosodia (206). A survey by Hawthorne and Fischer of speech-language pathologists indicated they do not assess or treat aprosodia, even when it is suspected, because they lack training and clinical tools, not because they think it is unimportant (104). As part of a series of systematic reviews conducted by the Academy of Neurological Communication Disorders and Sciences Right Hemisphere Damage Writing Group, authors stressed that, based on their findings, after right-hemisphere damage, clinicians should assess (1) both receptive and expressive emotional prosody, (2) comprehension of linguistic prosody as it pertains to distinguishing speech acts (eg, imperatives from interrogatives), and (3) pragmatic performance as it relates to emotional prosody performance (32; 76; 247; 256). The authors also noted limitations of the current literature and, therefore, made recommendations for improving research practices, such as better statistical practices, participant descriptions, and use of valid and reliable stimuli; they also advocated for making data accessible. Improving these research practices will allow for more robust findings, which will aid in clinical implementation.

Management

There is a relative paucity of empirical work that investigates the effects of treatment interventions for aprosodia. Research has been limited to case studies and small-sample clinical trials (30; 146). Furthermore, the aforementioned limited or omitted assessment of aprosodia by treatment providers (typically speech-language pathologists) suggests that it may not become a target for formal or informal intervention (206). Nevertheless, the extant literature suggests that there are, indeed, a few behavioral interventions that hold promise for the treatment of expressive aprosodia. In addition, a few nonbehavioral interventions (eg, pharmacotherapy) were not developed to treat aprosodia, per se, but have demonstrated potential to ameliorate prosody-related deficits. Below, different types of interventions and treatment targets (eg, prosody expression and recognition as well as nonaffective and affective prosody) are discussed (146; 103; 102). We refer the reader to Raskin and colleagues (207) for a review of treatment approaches (eg, pharmacotherapy, cognitive therapy, or behavior therapy) that are commonly used to address deficits in emotional processing, including aprosodia. We also refer the reader to Benedetti and colleagues (25) for a scoping review of evaluation tools and treatment approaches for aprosodia in adults with acquired brain injury, which can be very useful for speech therapists.

Behavioral treatment. Treatment interventions for expressive aprosodia are limited (228). The lack of clinical research for guiding the treatment of aprosodia (in comparison to the treatment of other communication disorders, such as aphasia, dysarthria, and apraxia of speech) may be because prosodic impairments occur across numerous clinical conditions, and prosodic deficits may be perceived as less limiting than other communication deficits (102). Aprosodia also appears to be a lower priority treatment target in acute care settings (206), likely further contributing to the limited research exploration into formal interventions. This, combined with the apparent lack of clinical training and tools to assess and treat aprosodia (104), results in fewer evidence-based interventions and poor implementation of existing interventions.

Imitative treatment is a well-known behavioral approach to the remediation of expressive aprosodia wherein a therapist models a prosodic tone, and the patient is instructed to imitate it (148; 146). A second approach is pitch biofeedback, in which a visual representation of a target pitch is presented, the patient imitates the target pitch and views the visual representation of it, then compares the target visual representation to the visual representation of their attempt, and, ideally, makes changes accordingly to improve their performance. Stringer reported a case study of a 36-year-old woman with motor aprosodia who participated in imitative treatment or expression modeling and pitch biofeedback (258). In this case, expression modeling entailed having the patient imitate paired and unpaired vocal and facial expressions and, then, receive verbal feedback from the clinician or make self-corrections using a mirror. The patient remained impaired relative to healthy individuals, but her vocal affect did, indeed, improve significantly with treatment, and such gains were seen at 2 months post-treatment. Silverman suggested that the use of an Augmentative and Alternative Communication system and a number of behavioral methods (eg, respiratory monitoring through visual feedback) might be effective in treating the deficits of affective prosody (250). It should be noted that in cases of affective aprosodia, treatment can be complicated by the presence of anosognosia (255; 288), as the lack of insight into deficits may interfere with self-monitoring and recognition of errors.

Another approach to the treatment of expressive aprosodia is cognitive-linguistic treatment. In cognitive-linguistic treatment, the patient reads descriptions of the prosodic features commonly used to communicate various emotions paired with pictures of facial expressions matching the emotion and is then asked to produce a particular tone of voice without a model. In a single-subject design study with three participants, “modest to substantial” treatment effects were shown for both a cognitive-linguistic intervention and imitative intervention; however, the treatment effect did not generalize to an untreated emotion (148; 146).

Leon and colleagues recommended that combined cognitive-linguistic and imitative interventions are more effective than either approach used individually (147). Both approaches involve a 6-step treatment process in which cues are faded out across the steps, although they noted that the results were preliminary.

Hargrove concluded that the four treatments reviewed (imitative approach, Samuelsson’s explicit intervention approach, the Lee Silverman Voice Treatment [LSVT] program, and targeting lexical stress) are evidence-based interventions for prosodic impairments (102). The LSVT approach had the strongest evidence because it was based on a prospective, randomized controlled design. To improve access to evidence-based interventions for prosodic deficits, Hargrove created the “Clinical Prosody” blog (clinicalprosody.wordpress.com), which permits clinicians to access information for their clinical practice and promote evidence-based practice in treating prosodic impairments.

Although aprosodia therapy initially focused on acquired brain injury (eg, stroke or traumatic brain injury), prosodic deficits in children with autism spectrum disorder have gained more attention in recent years. Akbari and Davis found that the 6-step imitative approach for expressive affective aprosodia in an adolescent with autism spectrum disorder improved both acoustic (duration and intensity) and perceptual (utterance and unstressed syllable duration) speech characteristics following a 10-week intervention (04).

Evidence supporting remediation for prosody recognition is limited but growing. In one study, individuals with acquired brain injuries (mostly traumatic brain injury) participated in a short treatment (three 2-hour sessions) to improve emotion prosody recognition (159). Treatment involved discussion of the seven basic emotions and categorization of emotion words into groups as well as exercises that involved prosody discrimination and production. Prosody recognition improved in six of the ten treated patients, though the effects were short-lived and not maintained at follow-up. Of note, given links between prosodic expression and recognition, it is possible that improving one may benefit the other (187). A study of patients in acute recovery following right-hemisphere stroke used an explicit acoustic-prosodic-emotion training paradigm to ameliorate receptive aprosodia (77). In the study, Durfee and colleagues used a scaffolded approach in which participants were first taught to discern acoustic features of pure tones (eg, pitch and volume: high or low), then to discern acoustic features of affective speech without an attempt to decode emotion (eg, pitch and volume: high or low), and, then, to assign acoustic features that they associated with a given emotion as presented in text (eg, sad – pitch, volume: high or low); following training, receptive prosody was significantly more accurate than at baseline.

Prosodic recognition deficits in schizophrenia have received increasing attention over the past few years, and it is now generally accepted that social cognitive deficits, including emotion perception, are just as important as cognitive deficits and psychotic symptoms in determining functioning. Lado-Codesido and colleagues assessed the efficacy of an online prosodic self-training intervention, “VOICES,” in individuals with schizophrenia and schizoaffective disorder in a randomized, single-blind, multicenter clinical trial (141). The treatment group completed eight 30-minute sessions of the VOICES computer program during which they heard a phrase and chose the emotion conveyed with feedback, which gradually increased in difficulty. Compared to a group of individuals given a conventional rehabilitation program, the treatment group showed improvement in prosodic test scores for the Reading Mind in the Voices-Spanish Version (RMV-SV), an adaptation of the Reading Mind in the Voices-Test Revised (RMV-TR) tool. The online nature of this program makes it easily available to a variety of populations, although, to our knowledge, it has not been utilized in patients with prosodic deficits resulting from acquired brain injury.

Another important aspect of behavioral treatment for aprosodia involves introducing patients and caregivers to compensatory strategies to maximize use of preserved functions, which is critically important because recoveries may be somewhat limited for patients with aprosodia (288). For instance, individuals with impairment in prosodic affective expression can be encouraged to employ facial expressivity to improve communication, to the extent possible, when conversing with others (118). Furthermore, patients with poor emotional prosody recognition, but preserved comprehension of emotional gesturing may be encouraged to engage in in-person communication where affective gestural information is also available (288). In addition, patients with aprosodia may be encouraged to compensate for their deficits by relying on their linguistic capacity. Patients can learn to communicate by placing emphasis on their word choice (ie, semantic material) rather than on prosody (288). Evidence for this approach comes, for instance, from a patient who compensated for the inability to communicate affect via prosody by providing explicit linguistic statements about her mood (eg, “I am angry and mean it”) (222).

Other behavior-related treatments. The Lee Silverman Voice Treatment (LSVT LOUD ®) (201; 203; 204; 202; 233) is aimed to improve vocal and speech characteristics in individuals with Parkinson disease. LSVT LOUD ® improves vocal volume, pitch range, and variability (200), as well as facial emotional expressivity (254; 75), in patients with Parkinson disease, and, thus, may be useful to other patient populations with aprosodia. Furthermore, music therapy can improve prosody (specifically, inflection, pitch, and intensity in speech) in patients with nonfluent aphasia (133). Similarly, Kim and Jo suggested that an accent-based music speech protocol can benefit nonaffective prosody in patients with stroke and mixed dysarthria (134).

Treatments of childhood apraxia of speech, where difficulties in affective prosody are a main symptom, include Rapid Syllable Transition Training (ReST) and Dynamic Temporal and Tactile Cueing (DTTC) interventions (158), Samuelsson’s explicit intervention approach (102), and ultrasound biofeedback (197; 198; 61). The ReST intervention focus is on appropriate lexical stress and prosodic control using both strong-weak (stress placed on first syllable of word) and weak-strong (stress placed on second syllable of word) words, and was found to significantly improve children’s production of lexical stress from pre-treatment to post-treatment (twelve 1-hour treatments over 3 weeks) (158). In the DTTC treatment, children are instructed to imitate words and phrases with different intonation patterns as modeled by the therapist (257). Although this treatment appears to directly target aprosodia, no studies have demonstrated the effectiveness of DTTC on affective prosody (257; 158). Samuelsson’s explicit intervention approach was created for children 4 to 6 years of age and targets production of meaningful prosody in speech, including vowel duration, word accents, early and late word stress placement, phrase stress, and intonation (232). This treatment improved affective expressive prosody in a single-subject experimental study, but further research must be conducted to assess generalizability. Ultrasound biofeedback has been used as an adjunct to traditional speech therapy interventions aimed at improving speech sound errors (197). It works by providing visual information to patients about their tongue movements while speaking through real-time ultrasound images of their tongue. The speech-language pathologist assists the patient in identifying what is wrong with the tongue placement and how to fix it to achieve the desired sound. Multiple studies have shown that this is an effective treatment for improving speech sound accuracy in children with childhood apraxia of speech (197; 198; 61), and applications to the treatment of aprosodia should be considered.

Nonbehavioral interventions with implications for the treatment of aprosodia. Pharmacological intervention has demonstrated somewhat variable efficacy for reduction of deficits in prosody in certain neuropsychiatric populations. Raymer and colleagues treated a patient with crossed-nonfluent aphasia with bromocriptine, a dopamine agonist, and observed substantial improvements in verbal fluency but did not find a significant improvement in emotional prosody or gesturing (208). Improved prosody has been observed following pharmacological intervention for seizures. Specifically, carbamazepine treatment for epilepsy has been shown to result in resolved seizure-related deficits in nonaffective prosody (73), and deficits in affective prosody resulting from seizures have been observed to resolve with phenytoin treatment (17), though questions remain about the level of experimental control exercised as these are both descriptive case studies. Further, treatment with L-Dopa has been associated with improvements in pitch variation in patients with Parkinson disease (97). One study found that the combination of LSVT with levodopa led to greater improvements in prosody than LSVT without levodopa in patients with Parkinson disease (09). Lastly, electroconvulsive therapy has been associated with improved affective prosody in a group of depressed individuals (152). Naturally, improvements seen in prosody associated with pharmacological interventions appear to be related to the treatment of the underlying disease process that is causing the prosodic deficits. Thus, the potential for the use of pharmacological interventions to improve aprosodia depends on the etiology of aprosodia and the effectiveness of treatments to address the underlying neurologic pathology.

Although subthalamic nucleus deep brain stimulation (STN DBS) might benefit prosodic processing in some individuals with Parkinson disease (50), much of the literature indicates that it instead worsens prosody (280; 261; 127; 130; 266; 01). However, subthalamic nucleus deep brain stimulation may be optimized to minimize consequent prosodic deficits, including lower frequency, lower pulse width, and higher voltage settings than the standard clinical settings used for deep brain stimulation (01).

Summary. Several behavioral interventions may ameliorate aprosodia. In addition, certain psychopharmacological interventions and electroconvulsive therapy have reduced communication deficits and improved prosody. Specific treatment strategies should be tailored based on the clinical presentation and associated etiologies because specific treatment outcomes can differ among populations and according to individual differences. As noted by Hargrove and others, treatment of aprosodia has received far less attention than other speech-related deficits (102; 104). Treatments for aprosodia deserve continued research, particularly in light of the aforementioned studies and the overwhelming evidence, in general, that affective difficulties improve with direct remediation techniques (263; 171; 40; 77). Moreover, translational research and expanded investigation of aprosodia’s impact on functional measures of life-participation and quality-of-life are strongly indicated to facilitate clinical practice and training of diagnostic and intervention techniques--particularly when considered in the context of responding to prolonged social isolation brought on by a global pandemic.

Special considerations

Pregnancy

Fetal alcohol syndrome, fetal alcohol effects, and in-utero alcohol exposure (not resulting in fetal alcohol syndrome or fetal alcohol effects) have been associated with affective prosody comprehension deficits (167; 166).

Acknowledgments

This work was supported, in part, by Professional Staff Congress – City University of New York Research Awards 67383-00-45, 68246-00-46, and 60285-00-48 and by NIH R01 DC01150.

Deena Schwen Blackett PhD is a postdoctoral fellow, and her time is funded by a grant from the National Institute on Deafness and Other Communication Disorders (T32DC014435).