Use of a non-invasive accelerometric method for diagnosing gastroesophageal reflux in premature infants.

OBJECTIVE: To evaluate the clinical usefulness of a non-invasive accelerometric device to diagnose GER in preterm babies. STUDY DESIGN: An accelerometer was taped over the sub-xiphoid process in 110 preterm (GA 29.6 ± 3.3 wk) infants (133 studies). Low frequency, sub-audible signals were captured via digital recording (sampling rate 200 Hz), then re-sampled (rate = 60 Hz) to create a spectrogram (focused range 0-30 Hz). Mean amplitude in the focused range was calculated. RESULTS: Of 85 studies with simultaneous pH-metry and accelerometry, 18 had concurrent positive and 23 had concurrent negative scores, 42 had negative pH scores when accelerometry was positive (≥1 µV), consistent with non-acid reflux. Eleven infants at high risk of aspiration received surgical interventions. All but 1 had negative pH scores while 10/11 had positive accelerometry. CONCLUSIONS: The non-invasiveness of this accelerometric technique allows for GER screening and for repeated testing to assess efficacy of interventions.

The Effect of Severe Intraventricular Hemorrhage on the Biorhythms of Feeding in Premature Infants.

Background: Suck-swallow rhythmicity and the integration of breathing into infant feeding are developmentally regulated. Neurological injury and breathing abnormalities can both impact feeding in preterm infants. Objective: To determine the effects of neurologic injury independent of effects of disordered breathing on feeding biorhythms in premature infants. Methods: Low-risk preterm infants (LRP), infants with Grade 3-4 Intraventricular Hemorrhage (IVH), those with bronchopulmonary dysplasia (BPD), and those with both BPD and IVH (BPD+IVH) were identified. Forty-seven infants, 32-42 weeks Postmenstrual Age (PMA) were evaluated on one or more occasions (131 studies). Of these, 39 infants (81 studies) were performed at >35 weeks PMA. Coefficient of variation (COV) (=standard deviation of the inter-event (e.g., suck-suck, swallow-breath, etc.) interval divided by the mean of the interval) was used to quantify rhythmic stability. Results: To adjust for PMA, only those infants >35-42 weeks were compared. Suck-suck COV was significantly lower (more rhythmically stable) in the LRP group [COV = 0.274 ± 0.051 (S.D.)] compared to all other groups (BPD = 0.325 ± 0.066; IVH = 0.342 ± 0.072; BPD + IVH = 0.314 ± 0.069; all p < 0.05). Similarly, suck-swallow COV was significantly lower in LRP babies (0.360 ± 0.066) compared to the BPD group (0.475 ± 0.113) and the IVH cohort (0.428 ± 0.075) (p < 0.05). The BPD+IVH group (0.424 ± 0.109), while higher, was not quite statistically significant. Conclusions: Severe IVH negatively impacts suck-suck and suck-swallow rhythms. The independent effect of neurological injury in the form of IVH on feeding rhythms suggests that quantitative analysis of feeding may reflect and predict neurological sequelae.

Variability of swallow-associated sounds in adults and infants.

We previously used cervical auscultation (CA) to describe the stability of swallow-associated sounds of infant feeding. To date, no similar studies have been performed in adults. The objectives of this study were to identify the initial discrete sounds (IDS) of adult swallows and compare the stability of IDS signals in infants to that of adults. We performed CA with a microphone and accelerometer fixed simultaneously to the neck of 20 healthy adults. Each participant consumed a liquid, puree, and solid. The microphone and accelerometer collected signals of similar duration. The variance index (VI), an assessment of the stability of the IDS, was compared in adults and a group of low-risk preterm infants. The VI of adults swallowing liquid (29.1 [24.1, 36.6] {25%, 75%}) did not differ from that of preterm infants older than 36 weeks PMA (36.3 [33.4, 41.9]), but was lower than the VI of infants younger than 36 weeks PMA (49.0 [46.4, 51.1]; p < 0.05). This is the first real-time comparison of microphones and accelerometers for CA. The stability of IDS of low-risk preterm infants approaches that of normal adults as the infants age. Because successful feeding in infants is often used as a surrogate for normal development, the stability of swallow-associated sounds deserves more investigation as a potential marker for neurologic well-being.

A New, Noninvasive Accelerometric Method for Diagnosing Acid andNon-Acid Gastroesophageal Reflux in Premature Infants: Proof of Concept.

Non-acid reflux is common in premature neonates. Current methods of diagnosing gastroesophageal reflux (GER) such as pH probes, multichannel impedance monitoring, X-rays, or endoscopy are either invasive or unable to diagnose non-acid reflux. Passage of a naso-esophageal tube is uncomfortable. Imaging studies are of short duration and may miss reflux entirely. Herein, we present proof of concept of a noninvasive accelerometric device that detects acid and non-acid reflux in premature infants. An accelerometer was taped over the subxiphoid process in patients suspected of having GER who were already scheduled for pH probe or multichannel impedance monitoring. The largest cohort was preterm infants, but term infants and toddlers were also studied. Low-frequency subaudible signals were obtained on a digital recorder (sampling rate 200 Hz) signals. Fast Fourier transforms graphically displayed the frequency and amplitude of signals. Data were then resampled at a rate of 60 Hz to create a spectrogram with a focused range of 0-30 Hz representing reflux-associated events. Proof of concept was attained through successful comparison with results from concurrent pH probes, multichannel impedance recordings, and ultrasound studies. We have thus validated accelerometry as a noninvasive method for assessing both acid and non-acid GER. The noninvasiveness of this diagnostic modality allows for repeated testing to assess the efficacy of anti-reflux medications, even when patients remain on antacids. This technology allows for more rational management of patients with GER and represents a major advance in the diagnosis and treatment of GER.

Respiratory patterns and strategies during feeding in preterm infants.

Because patterns of integration of respiration into rhythmic suck-swallow efforts are highly variable, we examined the vagaries of respiratory efforts as they evolve from the first tentative attempts at integration through more complex rhythmic interactions, with a focus on several strategies in which breathing and suck-swallow are coordinated. Thirty-four preterm infants (18 males, 16 females) of 26 to 33 weeks gestational age, 32 to 40 weeks postmenstrual age (PMA), and 2 to 12 weeks postnatal age were studied weekly from initiation of bottle feeding (using breast milk or preterm formula, both fed from a bottle) until discharge, with simultaneous digital recordings of pharyngeal pressure, nasal thermistor airflow, and thoraco-abdominal strain-gauge readings. Exceptional patterns of feeding-adapted variations of respiration were noted, including breathing during swallow, alternating blocks of suck-swallow and respiration efforts, narial airflow without thoracic movement, modulation of respiratory phase relationship against swallow rhythm, and paired rhythms with swallow:breath ratios of more than 1:1. Some of these strategies were developmentally regulated. Alternating blocks of suck-swallow and respiratory efforts were only seen in the earliest (PMA 32-33 wks) studies. In contrast, coordination and phase relationships of suck-swallow and breathing stabilized over time, as did the percentage of synchronized narial and thoracic respiratory efforts, which increased significantly after 36 weeks PMA compared with synchronization at 32 to 33.9 and 34 to 35.9 weeks PMA (p<0.05). There was also a significant positive correlation between percentage synchronization and PMA (r=0.58; p<0.001). The strategies and patterns noted here further clarify the developmentally regulated coordination of suck, swallow, and respiration into mature infant feeding, and may be predictive of those infants with short- and long-term feeding or developmental difficulties.

Maturational changes in the rhythms, patterning, and coordination of respiration and swallow during feeding in preterm and term infants.

To study the coordination of respiration and swallow rhythms we assessed feeding episodes in 20 preterm infants (gestational age range at birth 26-33wks; postmenstrual age [PMA] range when studied 32-40wks) and 16 term infants studied on days 1 to 4 (PMA range 37-41wks) and at 1 month (PMA range 41-45wks). A pharyngeal pressure transducer documented swallows and a thoracoabdominal strain gauge recorded respiratory efforts. Coefficients of variation (COVs) of breath-breath (BR-BR) and swallow-breath (SW-BR) intervals during swallow runs, percentage of apneic swallows (at least three swallows without interposed breaths), and phase of respiration relative to swallowing efforts were analyzed. Percentage of apneic swallows decreased with increasing PMA (16.6% [SE 4.7] in preterm infants 35wks; 1.5% [SE 0.4] in term infants; p<0.001). Term infants had significantly lower BR-BR COV (0.405 [SE 0.016]) than preterm infants (0.641 [SE 0.052] at >35wks' PMA; 0.693 [SE 0.059] at

Abnormalities in the coordination of respiration and swallow in preterm infants with bronchopulmonary dysplasia.

Individual rhythms of suck, swallow, and respiration are disrupted in preterm infants with bronchopulmonary dysplasia (BPD). Integration of respiration into suck-swallow efforts is critical for establishing coordinated suckle feeding. This study quantitatively assessed the coordination of respiration and swallow in infants with and without BPD. Thirty-four preterm infants of 26 to 33 weeks' gestational age were included: 14 participants with BPD (eight males, six females)and 20 comparison participants without BDP (10 males, 10 females). Participants were studied at postmenstrual age 32 to 40 weeks and postnatal age 2 to 12 weeks using digital recordings of pharyngeal pressure, nasal thermistor flow, and thoraco-abdominal plethysmography. The coefficients of variation (COV; standard deviation/mean) of the swallow-breath (SW-BR) and breath-breath (BR-BR) intervals during swallow runs, the percentage of 'apneic swallows' (runs of >or=3 swallows without interposed breaths), and phase relationships of respiration and swallow were used to quantify rhythmic coordination and integration of respiration into feeding episodes. Apneic swallows were significantly increased after 35 weeks in infants with BPD (mean 13.4% [SE 2.4]) compared with non-BDP infants (6.7% [SE 1.8]; p<0.05), as were SW-BR phase relationships involving apnea. The BPD cohort also had significantly higher SW-BR COV and BR-BR COV than non-BPD infants, indicating less rhythmic coordination of swallowing and respiration during feeding. Results emphasize the need for frequent rests and closer monitoring when feeding infants with respiratory compromise. Quantitative assessment of the underlying rhythms involved in feeding may be predictive of longer-term feeding and neurological problems.

Cervical accelerometry in preterm infants with and without bronchopulmonary dysplasia.

Cervical accelerometry with digital signal processing (DSP) can identify signals that are consistently associated with swallowing during feeding of infants. It is shown that these signals, called initial discrete sounds (IDSs), become more uniform with advancing postmenstrual age (PMA) in healthy preterm infants. The objectives of this study were to determine if cervical accelerometry and DSP at a higher sample rate than previously used (22 kHz versus 16 kHz), in conjunction with a DSP software package that allows manipulation of the sound files, would improve the accuracy of the previously developed variance index (VI) method of analyzing accelerometric information. The modified VI method was then used to measure developmental differences in the IDS morphology of infants with and without bronchopulmonary dysplasia (BPD). VIs were calculated for 24 feeding studies of infants of between 32 and 39 weeks PMA: 12 studies on healthy preterm infants (n = 10: three males, seven females; mean gestational age [GA] 28.6 weeks, SD 0.4; mean birthweight [BW] 1080 g, SD 82; PMA mean 35.2 weeks, SD 0.6) and 12 studies on infants with BPD (n = 7: five males, two females; GA 27.1 weeks, SD 0.4; BW 911 g, SD 71; PMA 36.2 weeks, SD 0.6). There was a significant inverse correlation between VI and PMA for the healthy preterm group (r = 0.66, m = -2.31/week,p = 0.02). There was no significant correlation between VI and PMA for the BPD cohort. The VI of infants with BPD was significantly different from that of infants without BPD (p < 0.007, multiple regression analysis, interaction PMA x Group). Additionally, using our modified VI technique, 100% of swallows were found to have the expected IDS signals.

Coordination of suck-swallow-respiration in infants born to mothers with drug-abuse problems.

To delineate quantitatively differences in suck-swallow-respiration coordination during feeding in term infants exposed to drugs in utero, 16 control infants (10 females, six males; mean birthweight 3209g, standard error of the mean [SEM] 103g) and 15 drug-exposed infants (seven females, eight males; seven cocaine, eight opiates; neonatal abstinence [Finnegan] scores less than 10; mean birthweight 3001g, SEM 93g) were studied in the first three days of life and again at one month of age. Coefficients of variation (COVs) of suck-suck, swallow-swallow, and breath-breath intervals were used as measures of stability of individual rhythms. COVs of suck-swallow and swallow-breath intervals were used to determine stability of multiply integrated rhythms. In the first three days of life, a significantly higher percentage of 'apneic swallows' (runs of three or more swallows not associated with breathing movements, divided by total run-swallows) was noted during feeding in the drug-exposed infants (mean 5.3% SEM 1.7%) compared with controls (mean 0.9%, SEM 0.4%;p<0.02), who also had less breath-breath rhythmic stability (higher COV) and shorter swallow-breath intervals. These differences were no longer evident at one month of age. Opiate-exposed infants were generally more affected than the cocaine-exposed infants. Swallowing was less rhythmic in the drug-exposed cohort, especially in opiate-exposed infants, but this difference also disappeared by one month. In the first days of life, drug-exposed infants were less efficient feeders, ingesting less volume per run-swallow than controls (mean 0.31ml, SEM 0.05ml, versus 0.45ml, SEM 0.05ml;p<0.05). The decreased efficiency appeared to be compensated for by a slightly faster swallow rate (mean 53 swallows per minute SEM 2 versus 47 swallows per min, SEM 1, in controls;p<0.005). By one month of age these differences between groups had disappeared as well. Intrauterine drug exposure, even in infants with relatively mild symptoms of withdrawal, may adversely impact development of brainstem respiratory and swallow centers, thus affecting, albeit transiently, the underlying biorhythms of feeding.

Integration of suck and swallow rhythms during feeding in preterm infants with and without bronchopulmonary dysplasia.

The aim of this study was to define quantitative measures for assessing the integration and maturation of suck and swallow rhythms in preterm infants as they relate to each other. Fourteen preterm infants (eight males, six females; gestational age range 26 to 32 weeks) with bronchopulmonary dysplasia (BPD) and an age-matched cohort of 20 infants (10 males, 10 females; gestational age range 26 to 33 weeks) without BPD were studied weekly from time of initiation of oral feeding using simultaneous recordings of nipple and pharyngeal pressure. The integration of suck and swallow rhythms was quantified by using the coefficient of variation (COV) of the suck-swallow dyad interval. Infants without BPD had a significant correlation between increasing postmenstrual age (PMA) and decreasing COV of the dyadic interval (increasing stabilization; r=0.45; p=0.008). In the non-BPD cohort 35 weeks or less PMA, the mean dyadic COV was 0.42 (SD 0.12) versus 0.34 (SD 0.09) in those more than 35 weeks PMA (p=0.039). In contrast, dyadic stability in infants with BPD was not correlated with PMA. Infants with BPD of more than 35 weeks PMA had less dyadic stability (0.45, SD 0.10) than did age-matched controls (p<0.001). Dyadic stability was also correlated with feeding efficiency in the non-BPD group (r=0.46;p=0.007) but not in the BPD cohort. Therefore, ontogeny of rhythmic suckle feeding can be described quantitatively in preterm infants, allowing comparison with at-risk populations. Infants with BPD do not follow predicted maturational patterns of suck-swallow rhythmic integration.

Cervical accelerometry in preterm infants.

The objective of this study was to develop a method to use digital signal processing (DSP) technology to describe quantitatively and statistically swallow-associated sounds in preterm infants and to use this method to analyze changes as infants mature. Twelve recordings of accelerometric and physiological data on bottle-feeding preterm infants between 32 and 39 weeks' postmenstrual age (PMA) were analyzed. Cervical auscultation was performed using an accelerometer attached over the larynx. Acoustic data were recorded and graphically displayed using DSP software. Initial discrete sounds (IDSs) were identified and used to construct an average waveform from which a 'variance index' (VI) was calculated for each infant. The shape of the IDS waveforms became progressively more uniform with advancing PMA, as indicated by a significant inverse correlation between VI and PMA (r=0.739; p=0.006). DSP technology facilitated the development of a new method to quantitatively analyze feeding in preterm infants. This method provides an elegant tool to track maturation of infant feeding and assessing feeding readiness. This technique makes the interpretation of cervical auscultation data less subjective by replacing the verbal description of the sounds of feeding with quantitative numeric values. It is anticipated that this method can be automated to facilitate further the analysis of cervical accelerometry data.

Changes in rhythmic suckle feeding patterns in term infants in the first month of life.

To quantify parameters of rhythmic suckle feeding in healthy term infants and to assess developmental changes during the first month of life, we recorded pharyngeal and nipple pressure in 16 infants at 1 to 4 days of age and again at 1 month. Over the first month of life in term infants, sucks and swallows become more rapid and increasingly organized into runs. Suck rate increased from 55/minute in the immediate postnatal period to 70/minute by the end of the first month (p<0.001). The percentage of sucks in runs of > or =3 increased from 72.7% (SD 12.8) to 87.9% (SD 9.1; p=0.001). Average length of suck runs also increased over the first month. Swallow rate increased slightly by the end of the first month, from about 46 to 50/minute (p=0.019), as did percentage of swallows in runs (76.8%, SD 14.9 versus 54.6%, SD 19.2; p=0.002). Efficiency of feeding, as measured by volume of nutrient per suck (0.17, SD 0.08 versus 0.30, SD 0.11 cc/suck; p=0.008) and per swallow (0.23, SD 0.11 versus 0.44, SD 0.19 cc/swallow; p=0.002), almost doubled over the first month. The rhythmic stability of swallow-swallow, suck-suck, and suck-swallow dyadic interval, quantified using the coefficient of variation of the interval, was similar at the two age points, indicating that rhythmic stability of suck and swallow, individually and interactively, appears to be established by term. Percentage of sucks and swallows in 1:1 ratios (dyads), decreased from 78.8% (SD 20.1) shortly after birth to 57.5% (SD 25.8) at 1 month of age (p=0.002), demonstrating that the predominant 1:1 ratio of suck to swallow is more variable at 1 month, with the addition of ratios of 2:1, 3:1, and so on, and suggesting that infants gain the ability to adjust feeding patterns to improve efficiency. Knowledge of normal development in term infants provides a gold standard against which rhythmic patterns in preterm and other high-risk infants can be measured, and may allow earlier identification of infants at risk of neurodevelopmental delay and feeding disorders.