Development and Resolution of Nasal Fricatives in a Child with Repaired Bilateral Cleft Lip and Palate: A Case Report

1Craniofacial Center, Division of Craniofacial and Surgical Care, Adams School of Dentistry, University of North Carolina at Chapel Hill

2Craniofacial Center, Division of Craniofacial and Surgical Care, Adams School of Dentistry, University of North Carolina at Chapel Hill

3Craniofacial Center, Division of Craniofacial and Surgical Care, Adams School of Dentistry, University of North Carolina at Chapel Hill

This case report describes the development, characteristics, and resolution of anterior nasal fricatives (ANFs) – a learned maladaptive articulation error – in a young girl with repaired bilateral cleft lip and palate.

The girl was observed every two months from 12 to 24 months of age with follow-ups at 36, 48, and 67 months of age.

At 12 months of age, the girl nasalized /b/ inconsistently and had mild conductive hearing loss. At 18 months of age, she exhibited audible nasal air emission on some plosives and used ANFs to replace /s/ and /z/, often with a nasal grimace. At 24 months of age, the child continued to experience mild conductive hearing loss, obligatory nasal air emission, and ANFs for /s/ and /z/. At 36 months of age, pressure-flow testing documented significant velopharyngeal (VP) dysfunction. The girl then used ANFs for /f/ and /s/, phonetically marked by different oral stops. At 48 months of age, although VP impairment continued, speech therapy largely eliminated ANFs. By 67 months of age, VP closure was nearly normal.

Multiple factors including VP dysfunction, audible nasal air emission, and conductive hearing loss contributed to the development of ANFs. Clinical and etiological implications are discussed.

Children with repaired cleft palate can exhibit audible nasal air escape that differs in both auditory-perceptual characteristics and underlying causes. Clinicians have long recognized two distinct types of obligatory nasal air escape that is caused by velopharyngeal (VP) inadequacy – audible nasal air emission and nasal turbulence (also referred to as nasal rustle) (Peterson-Falzone, Hardin-Jones, and Karnell, 2001; Zajac and Vallino, 2017). Audible nasal air emission occurs when there is an open VP port with air becoming turbulent as it passes through and exits the nasal cavity. The sound of audible nasal air emission can be simulated by forcibly exhaling through the nose, creating a hissing-like noise (Peterson-Falzone et al, 2001). Zajac and Preisser (2016) showed that this type of nasal air escape is characterized acoustically by aperiodic noise. Nasal turbulence, in contrast, is a perceptually more salient type of noise that is characterized by varying degrees of periodic noise (Zajac and Preisser, 2016). Peterson-Falzone et al. (2001) suggest that the distinct sound of nasal turbulence is due to either displacement of mucous in the VP port and/or tissue vibration. Importantly, Kummer, Curtis, Wiggs, Lee, and Strife (1992) showed that audible nasal air emission is typically associated with relatively large VP gaps while nasal turbulence is associated with relatively small VP gaps. Either kind of obligatory nasal air escape can accompany any or all of the high-pressure oral consonants and typically requires physical management. Table 1 summarizes the terms used in this report to describe obligatory nasal air escape.

Terms used by the authors to describe obligatory audible nasal air escape.

Some children with and without palatal anomalies exhibit audible nasal air escape that is learned as a maladaptive articulation error, called nasal fricatives. Harding and Grunwell (1998) described active nasal fricatives as articulations produced with oral stops – either bilabial, lingual-alveolar, or lingual-velar – that directs all airflow through the nose. Frication noise occurs as air passes through and exits the nose. Some children also exhibit a nasal grimace as a component of an active nasal fricative. Nasal fricatives can replace any or all of the oral fricatives and affricates (Peterson-Falzone and Graham, 1990). Harding and Grunwell (1998) suggested that these articulations develop within a phonological framework to establish phonemic contrasts. They further suggested that some children mark contrasts among active nasal fricatives by using different places for oral stopping, similar to nasal consonants. That is, a nasal fricative used to replace /f/ might be produced with a bilabial gesture while a nasal fricative to replace /s/ might be produced with a lingual-alveolar gesture, thereby maintaining a place contrast. Harding and Grunwell (1998) also cite Harding (1993), however, who observed that some speakers typically use a lingual-velar placement of stopping for all nasal fricatives. Perceptually, active nasal fricatives can sound similar to audible nasal air emission that may occur in the presence of VP dysfunction as an obligatory symptom. Indeed, because the oral stopping gesture of an active nasal fricative may be difficult to perceive, these articulations may sound similar (or even identical) to obligatory audible nasal air emission.

Harding and Grunwell (1998) further noted that, at times, active nasal fricatives are produced with a “noisier, turbulent quality” that they referred to as “nasal turbulence” (p. 335). This description is similar to that of the posterior nasal fricative (PNF) described by Trost (1981). The PNF is produced with a partially closed VP port that results in audible frication noise and associated nasal air escape. Trost (1981) noted that radiographic imaging showed “a blurring of movement or a velar flutter” (p. 197). Trost (1981) further noted that PNFs occur in speakers without cleft palate and with normal production of other consonants. In these cases, she referred to the phenomenon as “phoneme-specific velopharyngeal inadequacy” (p. 197). As indicated by Peterson-Falzone et al. (2001), the use of PNFs without nasal emission on other consonants is often called “phoneme-specific nasal emission” (p. 210).1

Zajac (2015) used oral-nasal audio recordings obtained with the Nasometer (model 6200, Pentax Medical, Montvale, NJ) to describe two perceptually distinct types of active nasal fricatives that he called anterior and posterior. The ANF was produced with oral stopping and relatively diffuse nasal spectral energy that was quite similar to normal /s/ production. That is, spectral energy of the ANF was aperiodic and concentrated well above 4-5 kHz (see Zajac, 2015, Figure 2, p. 19). ANFs are consistent with the less noisy, less “turbulent” active nasal fricative described by Harding and Grunwell (1998). The PNF was also produced with an oral stopping gesture. Zajac (2015) further suggested that the distinctive nasal “turbulence” of the PNF was due to VP tissue flutter as identified by Trost (1981). This was evidenced by quasi-periodic components in the nasal spectra of children who produced PNFs but not ANFs (see Zajac, 2015, Figure 4, p. 22).2 Two points need to be emphasized at this time. First, Zajac (2015) used the terms anterior and posterior to refer to the primary location of the fricative noise as either at the exit of the nose (anterior) or the VP port (posterior). Second, as shown by Zajac (2015), although PNFs are often described as “turbulent”, these productions are primarily characterized by periodic, not aperiodic (or turbulent), spectral energy. Likewise, Zajac and Preisser (2016) have shown that obligatory nasal turbulence (or rustle) is primarily characterized by similar periodic noise components.3 Table 2 summarizes the terms used in this report to describe learned nasal fricatives.

Terms used by the authors to describe nasal air escape as part of maladaptive articulations.

Because clinicians encounter relatively few children who produce active nasal fricatives, a limited number of studies have investigated this misarticulation. Peterson-Falzone and Graham (1990) reported the speech characteristics of 36 children, three to 16 years of age, who produced phoneme-specific nasal emission. Nineteen of the children had no known structural anomalies of the VP mechanism. Peterson-Falzone and Graham (1990) described five patterns of phonemes affected in two or more children. All of the children, however, exhibited “posterior nasal frication on /s/ in some syllable position or particular phonetic context” (p. 134). The investigators further reported that a majority of the children had “highly suspicious” early histories of otitis media (p. 138), suggesting conductive hearing loss as a factor in the development of phoneme-specific nasal emission.

Morgan and Zajac (2016) described the clinical characteristics of 21 children, three to 15 years of age, who produced active nasal fricatives. Eleven of the children had repaired cleft palate (with or without cleft lip), three had unrepaired submucous cleft palate, three had hemifacial microsomia with ear atresia, and four had no known structural anomalies. Morgan and Zajac (2016) reported that PNFs were produced by 18 (86%) of the children and ANFs were produced by three (14%) of the children. This was determined using oral-nasal audio recordings as described by Zajac (2015). All children used nasal fricatives to replace at least the /s/ phoneme, consistent with Peterson-Falzone and Graham (1990). Hearing information was available for 17 of the children and all but two had positive histories for hearing loss and/or surgical insertion of ventilation tubes. Lastly, pressure-flow information was available for 18 of the children. Sixteen (89%) of these children had adequate (but not necessarily complete) VP closure during production of /p/ as evidenced by estimated areas of 5 mm2 or less.

It should be emphasized that the above studies were retrospective and, therefore, lacked important information on early hearing and speech development of many of the children. This is an inherent limitation of studies that investigate low-frequency behaviors. In contrast, a case report with well-documented longitudinal information has potential to significantly contribute to our understanding and treatment of children who produce active nasal fricatives, especially those with palatal anomalies who may also exhibit co-occurring obligatory symptoms. Zajac (2019), for example, recently described a girl with repaired cleft palate and conductive hearing loss who at 18 months of age exhibited obligatory nasal turbulence characterized by periodic noise during production of plosives. At 3 years of age, she produced PNFs for /s/ targets. Spectral analysis showed similar periodic noise in both her obligatory nasal turbulence during production of plosives and in her PNFs during production of /s/. This spectral evidence suggests that the girl may have learned to prefer nasal frication, perhaps as a result of enhanced auditory and vibrotactile feedback provided by nasal turbulence experienced during production of early plosives. Clearly, case reports can provide significant longitudinal information.

The purpose of the present case report is to describe the development, characteristics, and resolution of ANFs in a child with repaired bilateral cleft lip and palate (BCLP). The child was followed longitudinally from 12 to 67 months of age. Similar to the case reported by Zajac (2019), the child also experienced early conductive hearing loss and VP dysfunction. The child’s VP dysfunction, however, was characterized by obligatory nasal air emission, not turbulence, during production of plosives. At 36 months of age, she was using ANFs to replace both /f/ and /s/, marked by different locations of oral stopping. Pressure-flow testing also documented significant VP dysfunction, consistent with audible nasal air emission on plosives. At 48 months of age, although VP dysfunction continued, speech therapy largely eliminated the use of ANFs.

This case report presents observations of a Caucasian girl born with complete BCLP. The child was a participant in an ongoing study on the development of stop consonants in children with repaired cleft palate. She was followed longitudinally from 12 to 48 months of age as part of the study. She had no known syndromes, sensorineural hearing loss, or developmental delays. The child’s lip and alveolar clefts were severe (i.e., wide) with significant collapse of the nostrils. Because of this, she underwent presurgical nasal alveolar molding (NAM) starting at one month of age until her lip surgery at 4 months, 25 days. NAM therapy consists of an intraoral appliance designed to bring the cleft maxillary segments together (Grayson, Santiago, Brecht, and Cutting, 1991). The appliance also has extensions designed to lift the collapsed nostrils. The child reportedly experienced an initial ear infection at 6 months of age with recurring infections until she underwent myringotomy with insertion of ventilation tubes at the time of her palate surgery (two-flap palatoplasty) at 8 months, 19 days. Surgical notes from the myringotomy indicated thick bilateral middle-ear fluid that required suctioning. Because the child had cleft lip and palate, she qualified for early intervention services provided by the state. This started at 11 months of age and consisted of play therapy with a focus on communication and motor skills.

As part of the longitudinal study, the child was seen for speech and language assessments every two months between 12 and 24 months of age. She was also seen for follow-up assessments at 36 and 48 months of age. Signed, informed consent was obtained from the parent for the longitudinal study. The girl was also seen for routine team evaluation at 67 months of age. For this case report, the assessments and observations at 12, 18, 24, 36, 48, and 67 months of age are presented. The types of assessment are described first followed by specific findings at each age.

This was done by the first author at every visit to rule-out the occurrence of oronasal fistulae and/or surgical dehiscence of the palate.

Bilateral tympanograms were obtained at every visit using a Maico easyTymp automatic impedance unit (Maico Diagnostics).

Sound-field audiometry at the frequencies of 500, 1000, 2000, and 4000 Hz was done by a certified audiologist at 12, 24, and 36 months of age. Because sound-field testing evaluates both ears simultaneously, results reflect hearing sensitivity of the better ear. At 48 months of age, audiometry was done using headphones to obtain right and left hearing thresholds.

The Communication and Symbolic Behavior Scales Developmental Profile (CSBS DP) (Wetherby & Prizant, 2002) was completed at 12, 18, and 24 months of age. The CSBS DP takes approximately 30 minutes and consists of six semi-structured communication opportunities that involve the child, parent, and examiner. It assesses social communication; use of gestures, sounds, and words; and receptive understanding and object use of children up to 24 months of age. It also provides a total language score. The CSBS DP was video recorded. The Goldman-Fristoe Test of Articulation 3 (GFTA3) (Goldman & Fristoe, 2015) was administered at 36 and 48 months of age. The GFTA3 was video recorded.

Nasal ram pressure (NRP) monitoring was completed at 18 and 24 months of age to assess VP function. NRP detects the presence of air pressure at the exit of the nares during speech and breathing (Thom, Hoit, Hixon, & Smith, 2006). A pediatric-sized nasal cannula was looped over the child’s ears and inserted into the nostrils. The end of the cannula was connected to a bidirectional air pressure transducer to detect NRP. The child was shown pictures of common objects to elicit production of plosive consonants. Plosives produced with positive NRP were considered to have some degree of VP opening (see Eshghi et al., 2017, for details of the NRP procedure). It is important to note that, unlike pressure-flow testing described next, NRP does not provide area estimates of VP opening. NRP findings, therefore, are reported as a percentage of plosives produced with some degree of VP opening (i.e., positive NRP).

Pressure-flow testing and posterior rhinomanometry were completed at 36, 48, and 67 months of age to determine VP area during production of the syllables /pi/ and /mi/ and nasal area during breathing, respectively (Zajac, 2005). Nasal airflow was obtained using a small mask coupled to a calibrated and heated pneumotachograph. Oral air pressure was obtained with a catheter held behind the lips that was connected to a calibrated differential pressure transducer. Nasal air pressure was obtained by tapping mask pressure with a catheter connected to a second differential pressure transducer. VP and nasal areas were estimated using pressure-flow data and PERCI-SARS software (Microtronics, Inc., Chapel Hill, NC) (see Zajac, 2005, for details of pressure-flow and posterior rhinomanometry). VP function was described using the palatal closure efficiency (PaCE) index (Zajac, Powell, and Perrotta, 2021). PaCE is a speaker-centered metric that shows an individual’s ability to achieve VP closure during an oral plosive relative to his/her VP opening during a nasal consonant. It is calculated using the equation: PaCE% = (100 – (plosive area ÷ nasal consonant area) x 100)). As way of example, if VP areas are 5 and 10 mm2 for /p/ and /m/, respectively, then PaCE will be 50%, meaning that the speaker was able to achieve only 50% VP closure during /p/ relative to his/her opening during /m/.

The microphone headset of the Nasometer (model 6200, Pentax Medical, Montvale, NJ) was used to obtain oral and nasal audio recordings at 36 and 48 months of age. The unfiltered audio outputs of the Nasometer were input into two channels of the Computerized Speech Lab (CSL, model 4500, Kay Pentax) at a sampling rate of 44.1 kHz and low-pass filtered at 17.6 kHz. It should be noted that the Nasometer was used only as a recording device to confirm the presence and characteristics of ANFs. That is, we examined the oral signal for evidence of stop gaps (i.e., silent intervals) and the nasal signal for evidence of frication of suspected ANFs. Because the software component of the Nasometer was bypassed, nasalance data (i.e., ratios of nasal to oral-plus-nasal acoustic energy) were not obtained.

An oral examination revealed an intact hard and soft palate with no oronasal fistulae. The girl’s nostrils were small and constricted with a short columella. The mother commented that the child had “difficulty breathing through the nose since birth”.

Tympanometry revealed bilateral large volume type B tracings consistent with patent ventilation tubes. Sound-field testing with pure tones showed thresholds of 20, 20, 25, and 25 dB HL at the frequencies of 500, 1000, 2000, and 4000 Hz, respectively, in the better ear.

The mother reported that the child was babbling /m/, /n/, /d/, and /j/ consonants. During administration of the CSBS DP, the child was observed to use /m/ and /b/ in both babble and word approximations. At times, /b/ targets were nasalized (i.e., /b/ in bye sounded somewhat like /m/). The child obtained a total standard score for language of 100 on the CSBS DP that placed her in the 50th percentile for age.

The mother reported that the child had begun to receive speech therapy as part of early intervention. Speech therapy focused on establishing age appropriate consonants and increasing vocabulary. She continued to receive play therapy with a focus on communication, turn-taking, and motor skills.

An oral examination revealed an intact hard and soft palate. Tympanometry revealed bilateral large volume type B tracings consistent with patent ventilation tubes. Parent report indicated that the child was using the following consonants in word attempts: /m/, /n/, /w/, /b/, /d/, /t/, /j/, /h/, /f/, and /s/. The mother further indicated that the child at times “pushed air through the nose” for /s/. Consonants observed during administration of the CSBS DP included /m/, /n/, /b/, /d/, /t/, /s/, /z/, /l/, /g/, and /j/. Based on both live and video observation of the CSBS DP, the child produced ANFs with a grimace for /s/ and /z/ targets (e.g., yes, nose, toes). Perceptually, these were characterized by a nasal grimace with audible, aperiodic nasal air emission. The girl’s production of /z/ in the words nose and toes occurred with clear lip closure, consistent with the use of an active nasal fricative. There was no lip closure, however, associated with /s/ in yes. In addition, the child grimaced at times during production of plosives. For example, she grimaced on both the /d/ and /g/ in the word doggie with audible, aperiodic nasal air emission accompanying the /d/. During approximately 15 minutes of NRP monitoring, the child produced four /b/ plosives. Examination of the NRP recording indicated that the VP port was open to some degree during two (50%) of the plosives.

The mother reported that the child had experienced two ear infections during the previous two months. She was seen by an ENT physician who reported that the child’s right tube was extruded in the ear canal; the left tube could not be clearly seen due to earwax. The child was scheduled to undergo repeat myringotomies and insertion of ventilation tubes in approximately two weeks. The mother also reported that speech and play therapy had continued since the previous visit.

An oral examination revealed an intact hard and soft palate. Tympanometry revealed a small volume type B tracing on the right and a large volume type B tracing on the left. These findings were consistent with the recent ENT report. Sound-field testing with pure tones showed thresholds of 25, 25, 20, and 25 dB HL at the frequencies of 500, 1000, 2000, and 4000 Hz, respectively, in the better ear. The child obtained a total standard score for language of 134 on the CSBS DP that placed her in the 99th percentile for age. Parent report indicated that the child was producing the following consonants in word attempts: /m/, /n/, /w/, /b/, /d/, /t/, /g/, /k/, /j/, /h/, /f/, and /s/. The mother continued to report a description of ANFs for /s/ targets and now also /f/ targets. All of the consonants reported by the parent were also observed during administration of the CSBS DP, along with /v/, /z/, and /l/. Similar to the assessment at 18 months, the child was observed to use ANFs with a grimace for /s/ and /z/ targets (e.g., bubbles), but not for /f/ as reported by the mother. Also, similar to the previous visit, the child exhibited audible nasal air emission with a grimace at times during production of some plosives (e.g., the initial /p/ in puppy). During approximately 15 minutes of NRP monitoring, the child produced 23 plosives (13 /b/, 7 /p/, 2 /d/, 1 /k/). Examination of the NRP recording showed that 19 (83%) of the plosives were produced with some degree of VP opening; only four of the plosives (all /b/) showed VP closure.

The mother reported that the child had new ear ventilation tubes placed soon after the previous visit. Surgical notes indicated that the original right tube was extruded and the middle ear had effusion. A myringotomy was done and a new tube was inserted in that ear. The original left tube was also extruded but a perforation remained in the eardrum. A new tube was inserted in that ear. The mother also reported that the child was receiving speech therapy once per week at a university clinic for articulation that targeted correction of nasal fricatives. Although details of the therapy were not available, the mother reported that the child used the See Scape device (PRO-ED, Inc., Austin, TX) for visual feedback of nasal airflow during therapy and at home during practice. The See Scape consists of a nasal olive and flexible tube attached to a rigid cylinder with a float inside. The device is intended to provide visual feedback of nasal airflow by placing the olive in one of the nostrils. The girl also received occupational therapy for a sensory processing disorder.

An oral exam revealed that the girl had an intact hard and soft palate. The tonsils were large. Tympanometry revealed bilateral large volume type B tracings consistent with patent ventilation tubes. Sound-field testing with pure tones showed thresholds of 20 dB HL at each of the frequencies of 500, 1000, and 2000 Hz (4000 Hz was not tested) in the better ear.

During administration of the GFTA3, the child was observed to exhibit a nasal grimace during production of both ANFs and some plosives, similar to her previous assessment at 24 months of age. Production of ANFs, however, were limited to /f/ and /s/ targets, phonetically marked by bilabial and alveolar stops, respectively (further described below). Including the nasal grimace as an error, she obtained a raw score of 39, a standard score of 96, and a percentile of 39 on the GFTA3. Nasal grimacing accounted for approximately 50% of the 39 errors. She also exhibited one instance of stopping (e.g., b/v), one devoicing error, and one instance of a labio-lingual production of /t/.

Posterior rhinomanometry indicated a nasal area of 9.6 mm2 during inspiration. This area was likely reduced for age given that nasal area for a 4-year-old is approximately 20 mm2 and nasal area is expected to increase only 2-3 mm2 per year with growth (Drake, Davis, and Warren, 1993). Pressure-flow testing during speech indicated a mean VP area of 4.9 mm2 during multiple productions of /p/ in the syllable /pi/. This area falls in the upper range of adequate (i.e., 0-4.9 mm2) as suggested by Warren and colleagues (e.g., Warren et al., 1989). However, the girl’s mean VP area during production of /m/ in the syllable /mi/ was only 8.2 mm2. Thus, the girl’s PaCE index was 40%, indicating significant VP inadequacy. Mean oral air pressure during production of /p/ was unexpectedly high at 9-11 cm H2O. This may have occurred due to the child’s reduced nasal area and/or excessive vocal effort. Consistent with the latter, the child was judged to have mild vocal hoarseness. The child was also judged to have mixed, mild hyper- and hyponasal resonance. These perceptual judgments were consistent with the clinical and instrumental findings.

We explored the spectral characteristics of the girl’s obligatory nasal air emission and ANFs by recording her using the oral-nasal microphone headset of the Nasometer. Figure 1 displays the oral signal, nasal signal, and nasal spectrogram of the child saying /pi/ five times, with obligatory nasal air emission. The nasal spectrogram shows aperiodic, relatively high-frequency energy during each production of /p/. The two vertical cursors in the figure show the stop phase of the second /p/. Playback of this segment revealed a hissing-type noise, similar to the other stops. Spectral moment analysis was done using TF32 software (Milenkovic, 2000). A 20 ms analysis window was centered between the two cursors shown in Figure 1. This analysis indicated a mean first spectral moment (average concentration of energy) of 9.7 kHz, consistent with a hissing-type noise. We need to emphasize two points. First, the spectral energy of the audible nasal air emission during /p/ is quite similar to the spectral energy expected for /s/ (Daniloff, Wilcox, & Stephens, 1980; Hixon, Weismer, & Hoit, 2008). With every production of /p/, therefore, the girl generated acoustic information that mimicked a fricative. Second, because of the girl’s conductive hearing loss, the nasal frication noise associated with /p/ was likely most salient to her when the oral cavity was closed due to bone conduction. Thus, because of enhanced feedback during the stop phase of /p/, the girl may have learned to associate production of a fricative with oral stopping. This point is discussed later in the section on etiologic considerations.

Oral acoustic signal (top), nasal acoustic signal (middle), and spectrogram of the nasal signal (bottom) of the child producing /pi/ /pi/ /pi/ /pi/ /pi/ with audible nasal air emission.

Figure 2 displays the oral signal, nasal signal, and nasal spectrogram of the child using ANFs during three productions of the word sissy. Playback of the oral signal indicated that the child produced alveolar stops, both /d/ and /t/ inconsistently, for each intended /s/ target, consistent with an active nasal fricative. The oral stops are clearly seen as silent intervals with release bursts for each /s/ target. Playback of the nasal signal indicated fricative /s/-like noise, similar to the noise in Figure 1. This is seen in the nasal spectrogram as aperiodic energy above 3 kHz, with the most intense energy above 5 kHz. Spectral moment analysis indicated a mean first spectral moment of 8.7 kHz for the six frication segments displayed in the nasal spectrogram. As indicated above for /p/, the spectral characteristics of the girl’s ANFs for /s/ are also quite similar to those expected for normally produced /s/. Daniloff et al. (1980), for example, reported first and second spectral peaks of 6.5 and 10.0 kHz, respectively, for children during production of /s/.

Oral acoustic signal (top), nasal acoustic signal (middle), and spectrogram of the nasal signal (bottom) of the child producing “sissy sissy sissy” with anterior nasal fricatives.

Figure 3 displays the oral signal, nasal signal, and nasal spectrogram of the child repeating the syllable /fi/ six times. Playback of the oral signal indicated that the child produced bilabial plosives /p/ for the intended /f/ targets, again confirming an active nasal fricative. As seen in the oral signal, the third, fourth, and fifth /p/ targets were frictionalized. This may have occurred due to restricted movement of the child’s repaired upper lip that prevented complete closure at times. Alternatively, it is also possible that because the child was receiving speech therapy that targeted ANFs, she may have inconsistently attempted oral production of /f/. Regardless, similar to Figure 2, the nasal spectrogram shows frication-like noise with a mean first spectral moment of 8.9 kHz for the six frication segments. This value is similar to, but slightly higher, than the first spectral moment of the ANFs produced for /s/ in Figure 2.

Oral acoustic signal (top), nasal acoustic signal (middle), and spectrogram of the nasal signal (bottom) of the child for producing /fi fi fi fi fi fi/ with anterior nasal fricatives. The diacritic x indicates frictionalization of the oral stop.

We also determined the spectral characteristics of the girl’s production of /ʃ/ targets from the GFTA3 that were produced normally and recorded with a single head-mounted microphone. For six /ʃ/ targets (3 in word initial positions, 3 in word final positions), the mean first spectral moment was 6.4 kHz. This mean is more than 2 kHz lower than the spectral mean for the ANF /s/ and is consistent with an expected /s/- /ʃ/ spectral distinction, at least for adult speakers (Haley, Seeingler, Mandulak, & Zajac, 2010). The significance of this finding is discussed later.

The mother reported that the child was still receiving speech therapy once per week at a university clinic. The treating SLP, however, wanted to discharge the child due to the resolution of ANFs and age appropriate articulation. The mother reported no new surgeries relative to ear ventilation tubes.

Tympanometry revealed a small volume type B tracing in the right ear and a large volume type B tracing in the left ear. The latter was consistent with a patent ventilation tube. Pure-tone testing with headphones showed thresholds of 25, 25, 20, and 20 dB HL for the frequencies of 500, 1000, 2000, and 4000 Hz, respectively, in the right ear. Thresholds were 20 dB HL at each of the frequencies in the left ear. The girl obtained a standard score of 116 on the Expressive Vocabulary Test-2 (Williams, 2007) which placed her in the 86th percentile for age.

During administration of the GFTA3, the child did not produce any ANFs. She exhibited audible nasal air escape with a grimace, however, at times during production of plosives. Her raw score was 10, the standard score was 107, and percentile was 68. Of the child’s 10 errors, approximately 50% were due to nasal grimacing. She also exhibited an f/θ substitution and a labio-dental production of /m/. The child’s resonance was judged to be within acceptable limits overall, with mixed, mild hyper- and hyponasal resonance.

Posterior rhinomanometry during breathing indicated a total nasal area of 8.4 mm2 during inspiration, slightly less than the area obtained at 36 months of age. This may have occurred due to transient nasal congestion at the time of testing. Pressure-flow testing during speech indicated a mean VP orifice area of 5.8 mm2 during multiple productions of /p/ in the syllable /pi/. VP area during production of /m/ in the syllable /mi/ averaged 12.3 mm2. The PaCE index was 53%. Although PaCE was improved (i.e., better VP closure) compared to 36 months of age, it still reflected significant VP dysfunction.4 Mean oral air pressure was 5 to 6 cm H2O during production of /pi/, reduced from 36 months of age but grossly within normal limits (Zajac, 2000).

Oral-nasal audio recordings largely confirmed the resolution of ANFs. During multiple production of /s/ targets, the girl exhibited oral frication for all /s/ segments. The nasal audio signal, however, showed evidence of obligatory nasal air emission as would be expected based on the pressure-flow findings. The girl showed inconsistent oral stopping on some but not all /f/ targets, suggesting that ANFs were not completely resolved for this phoneme. Similar to the assessment at 36 months of age, she produced bilabial oral stops for some /f/ targets.

The child was seen as part of her regular team evaluation at the UNC-CH Craniofacial Center. The mother reported that a recent ENT evaluation indicated a left tympanic membrane perforation that would require surgery. The child was not receiving speech therapy. Mother reported hearing nasal fricatives rarely – typically when the child was excited and talking fast – and she easily self-corrected when reminded to make the sound through the mouth.

An oral exam revealed an intact palate. There was active elevation of the soft palate along with formation of a small Passavant’s ridge during phonation. The child’s resonance was within acceptable limits, mildly hyponasal. There was no audible nasal air emission during production of plosives. Articulation was within acceptable limits; no ANFs were observed.

Posterior rhinomanometry during breathing indicated a total nasal area of 12.2 mm2 during inspiration, approximately 30% reduced for age compared to children without clefts (Drake et al., 1993). Pressure-flow testing during speech indicated mean VP areas of 0.6 and 7.6 mm2 during multiple productions of /pi/ and /mi/, respectively. This resulted in a PaCE index of 92%, indicating nearly complete VP closure. The girl’s oral air pressure during production of /pi/ averaged 8.8 cm H2O, within normal limits for age (Zajac, 2000).

Table 3 summarizes the above observations of the child regarding middle ear function, hearing, VP status, and articulation relative to the use of ANFs from 12 to 67 months of age.

Middle ear function (MEF), hearing, velopharyngeal (VP) status, and articulation relative to the use of ANFs of the child from 12 to 67 months of age.

ND = no data obtained

The purpose of this case report was to describe the development, characteristics, and resolution of ANFs in a young girl with repaired BCLP. At 12 months of age, VP function was suspect as indicated by inconsistent nasalization of the oral plosive /b/. At 18 months of age, the girl exhibited audible nasal air emission on some plosives and produced ANFs to replace /s/ and /z/. NRP monitoring at this time confirmed some degree of VP dysfunction. The girl experienced two ear infections between 18 and 24 months of age. She continued to exhibit VP dysfunction and ANFs at 24 months of age. At 36 months of age, oral-nasal audio recordings revealed the use of different oral stops to contrast ANFs used for /f/ and /s/. Pressure-flow testing showed a PaCE index of 40%, indicating significant VP dysfunction. At 48 months of age, although VP impairment continued as evidenced by a PaCE index of 53%, speech therapy largely eliminated the use of ANFs. At 67 months of age, ANFs were rare and VP closure was nearly complete as evidenced by a PaCE index of 92%.

The development of ANFs by the girl in the present case parallels the development of PNFs in the child described by Zajac (2019). In both cases, the children experienced early conductive hearing loss and VP dysfunction during production of plosives. In both cases, hearing was reduced at higher frequencies associated with spectral characteristics of /s/. In the case described by Zajac (2019), the girl exhibited obligatory nasal turbulence characterized by periodic noise during production of plosives. In the present case, however, the girl exhibited obligatory nasal air emission characterized by aperiodic noise during production of plosives. Zajac (2015) speculated that the type of obligatory nasal air escape that a child experiences, periodic or aperiodic, may predispose the child to developing PNFs or ANFs, respectively. This point is discussed later relative to etiologic considerations.

This case report highlights several important implications for clinicians. To begin with, the differential diagnosis of active nasal fricatives, either ANFs or PNFs, versus obligatory nasal air emission or nasal turbulence is essential for making appropriate management decisions for a child with repaired cleft palate. Confirmation of active nasal fricatives, however, can be difficult even for experienced clinicians in some young children when co-occurring obligatory symptoms are present. A frequently used clinical technique is to occlude the nose of the child when producing suspected nasal fricatives. Under this condition, all expiratory airflow will stop if there is oral occlusion, confirming an active nasal fricative. Many 18- and even 24-month-olds, however, may not be cooperative for this. Although better cooperation would be expected with older children, nose occlusion is not fool proof. In our experience, manually occluding the nose of some children who produce active nasal fricatives may stimulate oral airflow, especially if the child is receiving speech therapy. In the present case, we were confident that the child was producing ANFs at 18 months of age based on video recordings that showed clear lip closure during production of some fricative targets. At 36 months of age, we used the microphones of the Nasometer in conjunction with CSL to obtain oral-nasal audio recordings to further confirm and describe ANFs. While many clinicians, especially those in medical settings, may have access to the Nasometer, some may not have the CSL. Newer versions of the Nasometer (e.g., model 4550), however, will automatically save the separate oral and nasal acoustic signals which can be examined for evidence of oral occlusion. Given that the misdiagnosis of active nasal fricatives as obligatory nasal air escape can lead to unnecessary surgery, and, the misdiagnosis of obligatory nasal air escape as active nasal fricatives can lead to unsuccessful (and frustrating) speech therapy, clinicians are obligated to use all available instrumentation to ensure accurate differential diagnosis.5

This case report further emphasizes the critical point that, once active nasal fricatives are confirmed in a young child with repaired cleft palate, speech therapy needs to be initiated prior to consideration of surgical management. To be sure, when the child was 36 months of age, she exhibited obligatory symptoms of VP dysfunction such as audible nasal air emission during production of plosives, at times with a nasal grimace, and mild hypernasality. Pressure-flow testing showed a PaCE index of only 40%. These perceptual and aerodynamic findings suggested the need for secondary surgical management. Indeed, the mother of the child was counseled that speech surgery would likely be needed in the future. As this report illustrates, however, not only did speech therapy eliminate active nasal fricatives, but the girl’s overall VP function improved to the point that, at 67 months of age, there was no consideration of surgery. We need to make plain that we are not attributing the improvement in VP function to speech therapy per se. Undoubtedly, it is likely that the girl’s overall motoric, phonologic, and communicative competency increased as a result of both speech therapy and maturation. This can be seen by noting that as the resolution of ANFs occurred over time, there was a concomitant improvement of VP function as reflected by PaCE indices that increased from 40% at 36 months of age, to 53% at 48 months of age, and to 92% at 67 months of age. The development of a Passavant’s ridge at 67 months of age also likely contributed to improved VP function. Regardless of the reasons, this case report clearly reinforces the need for speech therapy to address learned behaviors such as active nasal fricatives before consideration of surgery when obligatory symptoms coexist in children with repaired cleft palate.

Lastly, school-based and/or private practice clinicians who may rarely encounter children who produce nasal fricatives need to seek collaboration with an experienced craniofacial SLP. This is critical not only relative to accurate differential diagnosis but also to receive guidance regarding appropriate management techniques, especially when obligatory symptoms of VP dysfunction may also be present.

The cause of active nasal fricatives, either anterior or posterior, remains elusive. Peterson-Falzone et al. (2001) stated that the etiology is “certainly” heterogeneous given the pattern of the phenomenon and the fact that it occurs in both patients with and without structural anomalies (p. 210). Referring to children with repaired cleft palate, they speculated that “either a previous history of true physiological velopharyngeal inadequacy or hearing loss could easily be a basis for mislearning of how to produce a normal sibilant or affricate or other fricatives” (p. 210). They further acknowledged that the cause was not readily apparent in children without cleft palate.

Zajac (2015) proposed an etiology that was common to children both with and without cleft palate. He hypothesized that early conductive hearing loss combined with VP dysfunction during production of plosives triggered the development of active nasal fricatives. Both conditions are likely to occur in children with repaired cleft palate and may occur in some children without cleft palate. Children who have large and/or irregular adenoids, for example, may experience nasal air escape during production of plosives due to incomplete VP closure (see Zajac and Vallino, 2017, Figure 8-6, p. 211). These children are also likely to experience early bouts of otitis media with effusion due to eustachian tube dysfunction. Thus, in the presence of conductive hearing loss and degraded auditory feedback, a child who produces plosives with nasal air escape is likely to experience enhanced feedback from bone conduction and/or vibrotactile clues that may lead to the mislearning of fricatives. Zajac (2015) further speculated that the occurrence of obligatory nasal air emission characterized by aperiodic noise might lead to ANFs while the occurrence of obligatory nasal turbulence characterized by periodic noise might lead to PNFs. As evidenced in the present case, obligatory nasal air emission with aperiodic noise is likely to occur in children who have reduced (i.e., obstructed) nasal area that causes airflow to become turbulent as it exits the nose.

The observations presented in this case report are consistent with a history of obligatory nasal air emission characterized by aperiodic noise during production of plosives along with conductive hearing loss leading to the development of ANFs. To be sure, at 12 months of age, the girl had mild conductive hearing loss and suspected VP dysfunction. Between 18 and 24 months of age, the girl experienced two ear infections that likely further reduced hearing acuity during a critical period of phonological development. At 24 months of age, she was exhibiting audible nasal air emission during /p/ and used ANFs for /s/ and /z/. An audiogram documented mild conductive hearing loss, especially at higher frequencies that are important for auditory feedback of oral fricatives. Because of the girl’s conductive hearing loss, she may have learned to prefer nasal frication as it would be more salient due to bone conduction than oral frication. In addition, because the spectral and perceptual characteristics of the child’s ANFs were so similar to normal /s/ production, it is likely that listeners reinforced, albeit unintentionally, the child’s use of ANFs for /s/.

We need to note that 25 dB was considered a slight hearing loss based on categories recommended by DeBonis and Donohue (2008). For young children, they considered thresholds of −10 to 15 dB HL to be normal and 16 to 25 dB HL to represent slight hearing loss (p. 39). They emphasized that even a slight hearing loss may have significant effects on communication. In support of this, Shriberg et al. (2000) reported that children with average hearing thresholds greater than 20 dB at 12 to 18 months had a significantly higher risk of clinical or subclinical speech delay at 36 months compared to children with less than 20 dB average thresholds at 12 to 18 months. It should be further emphasized that the child in this case study may have experienced even more severe hearing loss at other times, especially during periods of active middle ear effusion and/or infections, such as those reported by the mother between 18 to 24 months of age.

Regardless of etiology, Harding and Grunwell (1998) commented that a child may attempt to incorporate nasal fricatives into his/her developing phonological system. This was evidenced by the girl in the present case who, at 36 months of age, phonetically marked her intended productions of /f/ and /s/ with bilabial and lingual-alveolar stops, respectively, thereby maintaining a place contrast. Of interest, the girl did not produce ANFs for /ʃ/ targets. This sibilant was produced orally and with a reduced first spectral moment as compared to /f/ and /s/ targets produced as ANFs. Thus, the girl appeared to develop motoric and phonologic strategies that maintained articulatory and acoustic-perceptual contrasts, although deviant, among her early developing fricatives and sibilants.

Lastly, we need to note that many clinicians consider active nasal fricatives as compensatory to VP dysfunction. Although there was clearly VP dysfunction in the present case, we avoided the term “compensatory” for two reasons. First, most sources attribute the development of compensatory articulations such as glottal and pharyngeal stops to the inability of the child to generate adequate air pressure in the oral cavity (e.g., Hoch, Golding-Kushner, Siegel-Sadewitz, & Shprintzen, 1986; McWilliams & Cohn, 2002; Peterson-Falzone et al., 2001; Warren, 1986). Although the girl in the present case had VP dysfunction, she was quite capable of generating adequate levels of oral air pressure for plosives – perhaps due to reduced nasal area and the use of a nasal grimace. Second, and more important, the use of the term “compensatory” may lead some clinicians to believe that a speaker has VP dysfunction, and thus, therapy will not be successful until after surgery. As this case report has shown, this thinking should not be encouraged.

Supported by NIDCR grant number 1R01DE022566.

Conflict of Interest

The authors have no conflicts of interest.

1At a recent American Cleft Palate-Craniofacial Association virtual speech-pathology forum, David Fitsimmons commented that the term “phoneme-specific nasal emission” may mislead some clinicians into believing that this is an obligatory symptom. We agree and suggest that this term be avoided in favor of describing the type of nasal fricative that a child may use as a misarticulation.

2Zajac (2015) did not use imaging to identify the source of the periodic vibration. While tissue flutter is the most likely source, it is also possible that vibration and/or displacement of mucous in the VP port contributed to the spectral pattern.

3Nasal airflow can become turbulent and, thus, audible after passing through either the posterior VP port and/or the anterior nasal cavity. In addition to some turbulent (i.e., aperiodic) noise, both PNFs and obligatory nasal turbulence typically have a perceptually prominent periodic quality that can best be described as a flutter or raspberry-like sound.

4It should be noted that the effects of transient nasal congestion would affect the production of both /pi/ and /mi/, likely reducing obtained VP areas. Because PaCE is essentially a ratio of the two VP areas, the overall interpretation of the index remains valid even in the presence of some nasal congestion.

5Pentax Medical and Glottal Enterprises both market systems that are capable of recording oral and nasal airflow and acoustic signals. Although we do not have (or endorse) these systems, both appear capable of facilitating differential diagnosis of active nasal fricatives.