Premature infants frequently experience difficulty in sucking, thereby delaying the transition from tube to full oral feeding [24–26]. According to the NOMAS cluster system, sucking difficulty can be classified as arrhythmical sucking pattern, inability to sustain sucking pattern, incoordination of SSR sucking pattern, and dysfunctional sucking pattern (Table 1). Although the dysfunctional sucking pattern is known to be associated with poor developmental outcome, the relationship with developmental outcome is not well known in other patterns. The results of this study suggest that incoordination of SSR sucking pattern could be more associated with the development in the cognitive domain both at 8-12 and 18-24 months than arrhythmical sucking or inability to sustain pattern.
Most of the preterm infants with an immature sucking pattern can successfully bottle-feed as the SSR matures, and the sucking pattern of term infants is characterized by the rhythmic alternation of suction and expression/compression . In the arrhythmical or inability to sustain pattern group, it could be assumed that SSR coordination has already been formed in preterm period, but sucking difficulty has occurred due to lack of oromotor strength and endurance , and this group showed a faster development than the group with less developed SSR coordination.
Previous studies have attempted to predict developmental delay through neonatal oromotor function. In a study of 27 premature infants without brain lesions, the risk of developmental delay increased when the premature infants exhibited a disorganized sucking pattern at 37 weeks PMA . However, this study only examined the presence of disorganized sucking pattern and did not distinguish between specific patterns among disorganized sucking pattern.
The results of this study are compatible with those of the study by Nieuwenhuis et al. , who classified the disorganized sucking patterns in the NOMAS into two categories: disorganized due to arrhythmic sucking and disorganized due to lack of coordination of SSR. They concluded that uncoordinated sucking patterns, but not arrhythmic sucking patterns, were associated with abnormal fidgety movement at 14 weeks post-term . However, in the previous study, they did not discern dysfunctional sucking pattern from incoordination sucking pattern in the NOMAS. Additionally, the study did not distinguish the inability to sustain sucking pattern separately.
Regarding brain ultrasound finding, GMH-IVH grades 3 and 4 are widely considered predictors of developmental delay, whereas the implications of GMH-IVH grades 1 and 2 remains controversial [27,28]. In concert with previous studies, GMH-IVH grades 3 and 4 were identified as predictors of the cognition composite score in a multiple linear analysis in the present study. In the present study, BPD was also analyzed as a statistically significant predictor of developmental delay. Mizuno and colleagues reported that infants with BPD demonstrated not only poorer feeding coordination, but also poorer feeding endurance and performance , which might have affected development .
In the NOMAS, stress signals, including nasal flaring, head turning, and extraneous movement, are regarded as the symptoms of incoordination of SSR (Table 1), as described by Palmer et al. [16,18]. However, the relationship between incoordination of SSR and stress signals was suggested through clinical observation and not by direct measurement of SSR. To demonstrate the incoordination of SSR, recordings of intraoral pressure (rhythmic alternation of suction and expression/compression) [11,24], pharyngeal pressure [24,25], nasal thermistor flow [26,30], and thoracoabdominal plethysmography  have been used, but these methods are not widely used, particularly in the clinical settings, due to their complexity and invasiveness. However, in terms of research, the relationship between SSR incoordination and clinically observed stress symptoms could be investigated by directly measuring SSR. There might be certain stress signals that are more relevant to SSR incoordination than other signals.
There are a few limitations to this study. First, we did not include more objective signs that could be assessed in premature infants with sucking difficulty, including episodes of desaturation, apnea, and bradycardia. Those measurable signs might complement NOMAS, which is composed of observational findings. Second, because this study was conducted retrospectively, we only evaluated the sucking pattern before 40 weeks PMA since most premature infants are discharged from the NICU before term age. In our study, clusters 2 and 3 (i.e., arrhythmia and inability to sustain sucking without stress signals) were less relevant to neurodevelopmental outcomes. However, if those symptoms persist during the post-term period, the clinical relevance could be changed. For example, Wolthuis-Stigter et al.  reported that the inability to sustain sucking at 46 weeks significantly increased the odds of abnormal neurodevelopmental outcomes at 2 years of age. Therefore, for future studies, a prospective study design would be necessary for improving our understanding of the clinical significance of sucking difficulties that are observed in premature infants.