Oral feeding problems often occur in premature infants¹. The prevalence of feeding problems in pre-mature infants is reported to be 10.5-42% in a variable range according to definition^2,3. Some contributors of premature infant feeding problems include: immature swallow-related aspiration, gastroesophageal reflux disease (GERD), suck-swallow-breathe incoordination, neurologic immaturity, abnormal tone and posturing, oral aversion, decreased sucking, and impairment of other organ systems including cardiac, pulmonary, or renal^1,4.

Videofluoroscopic swallow study (VFSS) is the gold standard method for detecting subglottic aspiration. In adults, due to the risk of aspiration pneumonia, clinicians tend to restrict oral feeding in patients with VFSS- confirmed aspiration. However, prolonged exte-rnal stimulation via orogastric or nasogastric tubes, can disrupt the oral phase in preterm infants^4-6. Moreover, direct oral feeding is essential for the development of swallowing and feeding abilities in infants. A previous study reported that children who exhibited penetra-tion or aspiration on VFSS had a higher risk of deve-loping pneumonia than did those with normal VFSS⁷. On the contrary, another study reported that pneu-monia was significantly associated with underlying diseases such as asthma, Down syndrome, GERD, or prior lower respiratory tract infection, rather than the presence of aspiration⁸. Hence, there is no consensus regarding the association of aspiration on VFSS and aspiration pneumonia incidence in preterm infants. Furthermore, no study has reported the relationship between aspiration on VFSS and the occurrence of aspiration pneumonia in preterm infants only. It is unclear whether oral feeding should be restricted in the presence of aspiration on VFSS in preterm infants, unlike in adults. Therefore, this study aimed to inve-stigate the risk of aspiration pneumonia in preterm infants on oral feeding after aspiration had been con-firmed on VFSS.

1. Subjects

This study was conducted in a tertiary-care hospital. We retrospectively reviewed the medical records of preterm infants with dysphagia from December 2017 to May 2021. Inclusion criteria were: 1) preterm infants born less than 37 weeks, 2) subglottic aspiration confirmed on VFSS, and 3) initiation of partial oral feeding or full oral feeding after the VFSS. The sym-ptoms of dysphagia that were the reason for the re-ferral included poor sucking, feeding refusal, feeding desaturation, choking, or coughing with feeding. Among 50 preterm infants who had undergone VFSS, subglo-ttic aspiration was identified in 37 patients. Of the 37 infants with confirmed aspiration, two were lost to follow- up and nine were maintained on non-oral feeding due to massive aspiration. Finally, we analyzed 26 patients with aspiration confirmed on VFSS but who had initia-ted oral feeding.(Fig. 1)

Figure 1. Flowchart of patient inclusion.

2. Methods

Clinical characteristics included sex, gestational age (GA) at birth, and corrected age when the VFSS was performed. VFSS was conducted after confirming that the test could be safely performed in consultation with the attending physician. Before performing the VFSS examination, a pediatrician evaluated the preterm infa-nts for any airway disease, and those with congenital airway malformations, such as laryngeal cleft and tracheal agenesis, were excluded from the study. VFSS was conducted as described by Logemann⁹ with some modifications. All VFSS procedures were performed with the infants seated upright, with the head and neck in the neutral position and the chest and abdomen in the true lateral plane. To maintain a corrected posi-tion during the test, they were held from behind by their caregivers, who wore radiation-protective gear. Image of the swallowing process were taken in the lateral view using fluoroscopic equipment. Infants were bottle-fed with formula milk mixed with barium sulfate (Solotop sol 140; Taejoon Pharm, Seoul, Korea). One physiatrist fed the infants, while another phy-siatrist examined the conditions of the patients and conducted tests. The total VFSS time was 1 min 30 s to 2 min, and at least 20 swallows were observed. The VFSS was continued despite the observation of aspi-ration unless massive aspiration was observed.

The VFSSs were retrospectively reviewed by a phy-siatrist. In VFSS, the frequency and amount of aspira-tion, nasal penetration, suck-swallow ratio, and sucking power were checked. Aspiration was defined when milk passed through the vocal cords and entered the subglottic space. The suck-swallow ratio was calcu-lated by counting the number of suck cycles per one pharyngeal swallow. Sucking power was evaluated as good or weak. The amount of aspiration was also quan-titatively measured. For each aspiration, the percentage value of the aspirated amount to the swallowed amount was calculated, and the average value per patient was measured. The measurements were made using an image analysis software (ImageJ; National Institutes of Health).(Fig. 2)

Figure 2. Videofluoroscopic swallow study findings. (A) Amount of diluted barium swallowed. (B) Tracing of the diluted barium swallowed in dotted line. (C) Amount of diluted barium aspirated (arrow). (D) Tracing of the diluted barium aspirated in dotted line (arrow).

The characteristics of the subjects are described in Table 1. The mean GA was 33±3.69 weeks. The mean corrected age at the time of the VFSS was 33.15± 36.03 days. Of the 26 infants, 11 infants were normal- birth weight (≥2,500 g), 10 infants were low-birth weight (<2,500 g), and 5 infants were very low-birth weight (<1,500 g). And among 26 patients, 15 infants had respiratory distress syndrome, 4 infants had aspi-ration pneumonia, and 1 infant each had bronchitis, seizure, meconium aspiration syndrome, necrotizing enterocolitis, torticollis, intracerebral hemorrhage, and developmental delay. Fig. 3 shows the frequency of aspiration (%) of the 26 preterm infants, and Fig. 4 shows the mean amount (%) of aspiration of the 26 preterm infants. On average, the frequency of aspiration (percentage of the number of aspirations to the total number of swallowing) was 10.22±3.62 (%). The mean amount of aspiration (mean percentage of aspirated amount to the swallowed amount) was 2.28±1.02 (%). Nasal penetration was observed in 12 patients. The average suck-swallow ratio was 2.10 (SD:±0.93): 1 in this study population. Generally, in preterm infants, the suck-swallow ratio is considered functional up to 2-3 sucks per swallow¹⁰. Of the total 26 preterm infants, 16 exhibited good sucking power, and 10 exhibited weak sucking power. Twenty patients initiated full oral feeding after VFSS, and six initiated partial oral feeding, and subsequently progressed to full oral feeding. There was no case of complications in all preterm infants who per-formed the test.

Table 1 . Subject characteristics.

	Preterm infants (n=26)
Sex (male:female) (%)	18:8 (69.2:30.8%)
Gestational age (weeks)	33±3.69
Corrected age at VFSS (days)	33.15±36.03
Daysfrom birth to VFSS date (days)	72.65±51.58
Frequency of aspiration (%)	10.22±3.62
Mean amount of aspiration (%)	2.28±1.02
Nasal penetration (%)	12 (46.1%)
Suck-swallow ratio	2.10±0.93:1
Sucking power (good:weak) (%)	16:10 (61.5:38.5%)
Oral feeding after VFSS (full:partial) (%)	20:6 (76.9:23.1%)

Values are presented as number (%) or mean±standard deviation.

Figure 3. Frequency of aspiration (%) (percentage of the number of aspirations to the total number of swallowing).

Figure 4. Mean amount of aspiration (%) (mean percentage of aspirated amount to the swallowed amount).

One preterm male infant (subject 1) developed aspiration pneumonia. He was born at 27+6 weeks and was diagnosed with respiratory distress syndrome. He was born at very low-birth-weight (1,170 g) and had weak sucking power. He was on orogastric tube feeding until the VFSS exam (66 days from birth). The percentage of swallows comprising aspirations was 12.5%. The mean percentage of the aspirated amount to the swallowed amount was 6.64%. Full oral feeding was initiated immediately after the VFSS, however, 10 days later, the patient developed aspiration pneumonia.

To our knowledge, this was the first study to investigate the risk of aspiration pneumonia in pre-term infants who had initiated oral feeding, despite aspiration confirmed on VFSS. Moreover, no study has measured the degree of aspiration on VFSS in pre-term infants. Therefore, we also analyzed the fre-quency and the mean amount of aspiration. Of the 26 preterm infants who were analyzed the aspiration amount and frequency, most infants tolerated oral fee-ding. There was no occurrence of aspiration pneumonia when partial oral feeding was initiated and progre-ssed to full oral feeding. On the other hand, the pat-ient who developed aspiration pneumonia had started full oral feeding immediately after the VFSS. The VFSS result of that case revealed aspiration as not frequent compared to other infants.(Fig. 3); however, the amount aspirated was higher than those observed in the rem-aining infants.(Fig. 4) However, as only one patient developed aspiration pneumonia, statistical analysis was not performed.

In this study, the mean corrected age at the time of the VFSS was 33.15±36.03 days, which was after the full-term period. But, in case of the patient who developed aspiration pneumonia, oral feeding was performed after VFSS before full-term period, which was earlier than other preterm births. It is considered that caution is necessary when the timing of starting oral feeding after VFSS is relatively early.

When analyzed 9 preterm infants who were reco-mmended for non-oral feeding in this study, the aspi-ration frequency was 1 per 2-3 swallowing, and at least 1 time, the aspirated amount was 50% or more of the swallowed amount at VFSS. In these cases, the clinician judged that the aspiration was massive, so non-oral feeding was recommended. In this study, ana-lysis was performed on cases other than these subjec-tively judged massive aspiration cases.

Dysphagia can occur at different stages of swallowing in preterm infants. In the oral phase, immature or absent oral reflexes result in weak or disorganized sucking, immature patterns of biting or chewing, and drooling due to poor bolus formation and propu-lsion^5,6. Consequently, sucking skills improve over time in preterm infants^11,12. In the present study, weak sucking power was observed among the following: seven out of nine preterm infants (77.8%) with GA less than 34 weeks; three out of 17 preterm infants (17.6%) with GA more than 34 weeks and less than 37 weeks. Higher degree of prematurity is associated with higher incidence of oral phase dysfunction, which increases the risk of aspiration6. In the pharyngeal phase, the swallowing mechanism is linked to respiratory effici-ency, and successful feeding requires a coordinated response between swallowing and breathing¹³. Sucking- swallowing coordination usually develops by 32 to 34 weeks of GA and sucking-swallowing-breathing coordi-nation, by 37 weeks¹⁴. An incoordination of sucking- swallowing-breathing during feeding can compromise the airway safety, resulting in food particles entering the nasopharynx (nasopharyngeal reflux) or the airway, leading to laryngeal penetration (food particles entering the larynx but above the vocal cords) or tracheal aspiration (food particles entering the trachea)⁴. The pharyngeal swallow reflex and pharyngo-lower esopha-geal sphincter reflex are immature and underdeve-loped in preterm infants compared to those in term infants¹³. Furthermore, the presence of retrograde esophageal peristalsis and non-peristaltic esophageal motility in preterm infants contribute to abnormal esophageal clearance in preterm infants^6,15.

One study reported that prolonged nasogastric tube feeding for more than 2 weeks was independently associated with increased eating difficulties¹⁶. The oral phase of swallowing can be altered in preterm infants who undergo prolonged external stimulation via oral or nasal endotracheal tubes, non-invasive respiratory support interfaces, orogastric or nasogastric tubes, tapes, or securing devices⁴. Moreover, other problems including feeding tube dependency¹⁷, non- specific feeding difficulties, and oral aversion^3,16,18 develop. Therefore, timely and proper oral feeding is important in preterm infants when considering deve-lopmental milestones.

There is no consensus on whether partial and full oral feeding should be initiated in preterm infants with aspiration confirmed on VFSS. In our study, six of 26 preterm infants initiated partial feeding and then progressed to full oral feeding, and none of these infants developed aspiration pneumonia. However, the decision to start with partial or full oral feeding was subjective. Future studies are needed to establish more objective standards.

However, this study had several limitations. First, the relationship between the risk of aspiration pneu-monia and the aspiration frequency or amount was not statistically analyzed due to the low incidence of aspiration pneumonia. Therefore, we could not con-clude whether these values can be considered signifi-cant. Although silent aspiration can occur even in normal newborns, normal values for aspiration fre-quency or aspiration amount have not been establi-shed, so it was not possible to determine the signifi-cance of the values confirmed in this study. Second, a clinician recommended non-oral feeding for nine of 37 preterm infants who exhibited massive aspiration on VFSS, which could have contributed to a selection bias and limitation. Third, GERD is another risk factor for retrograde aspiration that can lead to aspiration pneumonia; however, not all the preterm infants were evaluated for GERD via VFSS. Despite these limita-tions, this study suggests another direction for the presence of aspiration in VFSS and dietary restriction in preterm infants.

In this study, most infants with aspiration confirmed on VFSS tolerated oral feeding provided that massive aspiration was not present, especially when partially initiated oral feeding. The presence of aspiration did not necessarily lead to aspiration pneumonia in premature infants on oral feeding. However, this study has limited implications because preterm infants with confirmed massive aspiration were excluded from this analysis. But, since appropriate oral feeding is critical to develop oral function in the infancy period, partial oral feeding may be initiated despite confirmed mild aspiration on VFSS in consideration of the clinical situation such as birth-weight, gestational age, under-lying disease, and sucking power. A larger-scale study is needed in the future to quantify the risk of aspi-ration pneumonia in association with the amount and frequency of aspiration.

The authors have no conflicts of interest relevant to this article.