Gastro-esophageal reflux (GER) is a common condition among preterm infants, due to several physiological promoting factors. To limit the potentially harmful widespread of pharmacological treatment, a step-wise approach, which firstly applies conservative strategies, is currently considered the best choice to manage GER in the preterm population. Among the most common conservative strategies, postural measures seem to represent an effective measure to reduce GER features in symptomatic preterm babies, whereas feed thickening is almost ineffective. Non-nutritive sucking (NNS) is a care strategy largely applied in the Neonatal Intensive Care Unit (NICU) settings. The act of swallowing is both reported to trigger the onset of transient lower esophageal sphincter relaxations (TLESRs), thereby eliciting to GER episodes, and to promote the esophageal clearance of refluxate. Hence, a possible effect of NNS on GER features might be hypothesized.

This observational, prospective and explorative study primarily aims to explore the effect of NNS, applied by means of a pacifier, on acid and non-acid GER features, evaluated in symptomatic preterm infants (gestational age ≤32 weeks) undergoing a diagnostic combined pH and multiple intraluminal impedance (pH-MII). The secondary aim of this study is to evaluate, in the subgroup of patients with recurrent GER-related apneas, the effect of NNS on cardiorespiratory events, defined as bradycardias and total, central, obstructive, mixed apneas and detected by a simultaneous polysomnographic monitoring.

Due to several promoting factors related to the physiological immaturity of the gastro-intestinal tract, gastro-esophageal reflux (GER) is a frequent condition among preterm infants.

GER clinical presentation may vary within a wide range of symptoms, being vomiting, regurgitations and cardiorespiratory events, such as apnoeas or bradycardia, the most common in the preterm population. Particularly, apneic episodes are inversely related to the gestational age (GA) and might be associated with such pathological events as hypoxia, hypercapnia or pressure variability, which seem to negatively affect the long-term neurological development of preterm babies.

Combined pH and multiple intraluminal impedance (pH-MII) is currently considered the best choice for GER diagnosis in preterm infants. While pH monitoring is limited to the detection of acid reflux, combined pH-MII effectively identifies both acid and non-acid GER, which is predominant in preterm babies. Moreover, thanks to its multichannel probe, combined pH-MII allows to evaluate the height reached by the refluxate within the esophageal lumen and to distinguish retrograde fluid bolus movements (reflux) from the anterograde ones (swallows).

In the last decade, GER pharmacological treatment has increasingly widespread in NICUs. The provision of anti-histamine2 (H2) blockers and proton pump inhibitors (PPI) in the preterm population, however, might lead to such adverse effects as an increased incidence of necrotizing enterocolitis and infections. Hence, to avoid a useless and potentially harmful pharmacological overtreatment, a step-wise approach, firstly promoting conservative strategies, is currently considered the most advisable choice for GER management in symptomatic preterm infants.

Postural and dietetic interventions represent the most common conservative strategies and their efficacy in improving GER symptoms and features has been largely investigated.

As we assessed in a previous pH-MII study, both prone and left-side positions were more effective than the right-side or the supine ones to determine a significant improvement in both acid and non-acid GER features.

Feed thickening is widely employed to reducing vomiting and regurgitations in symptomatic term infants; with regard to the preterm population, however, little data is currently available. We have previously investigated the efficacy of thickened fortified human milk on preterm infants suffering from GER symptoms (frequent regurgitations, desaturations, bradycardia), observing no benefits. Furthermore, standard human milk fortification itself has been shown to increase GER frequency in preterm infants.

With regard to anti-regurgitation (AR) formulas, we have recently evaluated the effectiveness of a specific AR formula, thickened with amylopectin and specifically tailored for preterm infants' nutritional needs; the tested thickened formula effectively reduced the number of acid GERs, but no difference was observed in the frequency of non-acid GERs and in the mean esophageal acid exposure, which is known to be the main determinant for the development of gastro-esophageal reflux disease (GERD). Thus, according to these preliminary data, feed thickening seem to be almost ineffective in improving GER symptoms and features in preterm infants.

Non-nutritive sucking (NNS) is part of the routine neonatal care, being widely applied in NICU settings. NNS has been showed to positively influence gastric secretion, gut motility and to improve the digestive process. NICU hospitalization and the need for respiratory support, however, might delay the transition from tube feeding to full oral feeding, thereby increasing the length of hospital stay and negatively affecting the mother-neonate relationship. Over two decades ago, in their meta-analysis Schwartz et al. showed beneficial effects of NNS in reducing time to attain full oral feeding as well as the length of hospitalization. These results have been further confirmed in a recent review by Harding.

The role of NNS on GER, however, has not been yet clarified. It has been previously shown that the swallowing enhances the occurrence of lower oesophageal sphincter relaxations, which are known to represent a trigger stimulus for reflux episodes. On the other hand, the act of swallowing appears to promote the esophageal clearance of refluxate; hence, a possible influence of NNS on GER might be hypothesized. In a pH-metric study, performed by Orenstein et al. on infants younger than 6 months, NNS seemed to have a worsening effect on GER when applied in the prone posture, whereas in the seated position was associated with an ameliorative outcome; this controversial result, however, cannot be extended to the preterm population. To date, the efficacy of NNS by means of a pacifier on preterm infants symptomatic for GER has been preliminary evaluated only by Zhao et al., reporting an improvement of overall GER features (number of refluxes, reflux index, total time at pH<4) and a faster time of gastric emptying in the NNS group. However, this study was performed by pH monitoring, so that the effect of NNS on non-acid GER has not been evaluated yet.

Due to the buffering effect of milk feeding on gastric pH, and being the acidity of gastric juice GA-dependent, non-acid GERs represent a major percentage (70%) of the overall reflux episodes in the preterm population. Moreover, as we have recently observed, non-acid reflux is the main responsible for the occurrence of GER-related apneas. Hence, a beneficial effect on GER-related apneas and subsequent hypoxic episodes, potentially harmful for preterm infants, could be supplied by conservative strategies effective in decreasing non-acid GERs features.

The primary aim of this observational, prospective, exploratory study is to evaluate the effect of NNS on GER pH-MII features in preterm infants (GA ≤32 weeks) symptomatic for GER (i.e. suffering from frequent regurgitations, feeding difficulties, failure to thrive and/or post-prandial desaturations) for which a 24-hour pH-MII monitoring for GER diagnostic evaluation has been scheduled.

The secondary aim of this study is to assess, by means of a simultaneous non-invasive 6-hour polysomnographic recording, the effect of NNS on total, central, obstructive, mixed apneas and on cardiorespiratory events within the subgroup of patients with recurrent desaturations.

Overall thirty preterm infants (GA ≤32 weeks) will be enrolled if suffering from moderate/severe GER symptoms and if a diagnostic 24-hour pH-MII monitoring has been scheduled to evaluate GER features, according to our clinical practice.

During the 24-h pH-MII monitoring each infant will receive eight meals, four of which will be followed by NNS treatment with a pacifier. The sequence of NNS application will be consecutively alternated in study participants. For each evaluation, a minimum of 2 hours of NNS post-prandial treatment is required. The healthcare assistants (i.e. medical staff and nurses) will inspect the correct application of NNS, recording the effective time of pacifier usage and regularly checking the pacifier position in relation to the pH-MII probe and polysomnographic flow sensor, to avoid possible interferences. Due to their immaturity, the enrolled patients might show an ineffective sucking, potentially losing the pacifier. In this case, the healthcare assistants will offer the pacifier to the babies up to a maximum of four times for each postprandial period; measures to fix the pacifier will be avoided.

The pH-MII device in use for this study (Comfortec pH-MII Sandhill Scientific) is a nasogastric probe with seven rings, composed by six dipolar impedance channels and one antimony electrode responsible for pH detection and placed in the middle of the distal impedance dipole. Due to these characteristics, pH-MII is able to measure the esophageal height reached by refluxate and to identify the direction of esophageal bolus movements, thereby distinguish between a retrograde flow (reflux) and an anti-retrograde flow (swallow). Esophageal probe length will be determined using Strobel's formula: [baby length (cm) x 0.252 + 5)] - 13%.

Impedance monitoring detects a GER episode (MII-GER) if there is a sequential drop in impedance values to less than 50% of the baseline, starting distally and propagating retrogradely to at least the next two more proximal measuring segments (about 4.5 cm above the lower esophageal sphincter (LES)). A GER episode is then defined as acid or non-acid according to its pH: the MII episodes with a pH lower than 4 are defined as acid MII-GERs, whereas the MII episodes with a pH≥4 are defined as non-acid MII-GERs. GERs migration height is the maximum distance from the LES reached by the esophageal refluxate. An acid GER episode is recorded by pH monitoring as a pH drop to less than 4; to be detected by the pH sensor, the pH drop has to last at least 5 s. The total percent time of esophageal exposure to a pH <4 is named Reflux Index (RIpH).

Data from the pH-MII recording will be automatically acquired on a portable Sleuth Sandhill Scientific system and then stored in a personal computer containing a specific software (BioVIEW Analysis Sandhill Scientific, version 5.0.9). Data analysis will be performed by the aforementioned software and confirmed by direct visual evaluation.

In the subgroup of patients with recurrent desaturations, a simultaneous non-invasive 6-hour polysomnographic monitoring, comprehensive of two meals (one followed by NNS and the other not), will be performed to evaluate a possible effect of NNS on apneic episodes (total, central, obstructive, mixed) and cardiorespiratory events.

The polysomnographic device in use for this study (System Plus, Micromed, Mogliano Veneto, Italy) measures the oral/nasal airflow by a flow transducer placed below the nostrils, whereas abdominal movements are detected by a band placed around the abdomen which is connected to a respiratory inductance plethysmography. Each oral/nasal flow cessation ≥5 seconds will be considered an apneic episode; apneas are then classified as central (CA: absence of both oral/nasal flow and movements of the abdominal wall), obstructive (OA: abdominal movements with an inadequate oral/nasal flow) and mixed (MA: components of both CA and OA). Apneas included in periodic breathing patters will be ruled out from the analysis. Systemic oxygen saturation is recorded by means of a pulse oximeter sensor, placed on the infant's foot. Cardiorespiratory events will be characterized as hypoxic episodes, defined as the occurrence of SatO2 <80%, and/or bradycardia, defined as pulse rate <100 bpm. The polysomnographic recording is automatically acquired in a computer connected to the polysomnographic device. The number of total apneas, CA, OA, MA and cardiorespiratory events will be traced down and visually analyzed by the sub-investigators.

Statistical analysis will be performed by SPSS 20 (Statistical Package for the Social Sciences, SPSS Inc., Chicago, IL, USA) for Windows. Differences between NNS and control periods in terms of GER frequency and features, number of apneic and bradycardic episodes will be tested by Wilcoxon Signed Ranks Test. The level of significance is set at 0.05.

Inclusion Criteria:

• Preterm newborns (GA ≤ 32 weeks) able to sustain a feeding of at least 100 ml kg-1 day-1 of fortified human milk or standard preterm formula, affected by severe/moderate GER symptomatology (recurrent regurgitations, feeding difficulties, failure to thrive and/or post-prandial desaturations) which usual treatment includes 24h pH-MII diagnostic evaluation.
• Preterm newborns able to sustain pacifier usage for at least 2h, as scheduled in our study plan.
• Weight ≥ 1100 g at the time of enrollment
• Obtained written parental consent.

Exclusion Criteria:

• Newborns with serious chronic pathology.
• Ongoing pharmacological treatment which could interfere with gastro-enteral motion functions (pro-kinetics) and/or with acid gastric secretion (H2 antagonists; proton pump inhibitors).
• Major congenital malformations (e.g. congenital heart diseases, gastrointestinal abnormalities, malformation syndromes).
• Neonatal necrotizing enterocolitis (NEC).
• Ongoing infections.
• Patients with severe clinical conditions which can hinder his/her participation in this trial (e.g. patent ductus arteriosus, intra-ventricular hemorrhage, hemodynamic instability).
• Administration of experimental medication treatment during the previous two weeks.

Due to the exploratory nature of this study, at least 30 preterm infants admitted to the Neonatal Intensive Care Unit of Sant'Orsola-Malpighi Universitary Hospital and fulfilling the inclusion criteria will be enrolled.