Feasibility of Spectral Analysis as a Tool in Nursing Research to Quantify Patterns of Respiration in Premature Infants

Abstract

Background. Respiratory difficulties are a common concern in preterm infants, and they can lead to long-term health problems. Few studies have investigated the use of spectral analysis as a biomarker to quantify respiration patterns in preterm infants. Objective. To evaluate the feasibility of using spectral analysis of heart rate variability as a biomarker for the quantification of respiratory patterns in very-low-birth-weight preterm infants compared to direct observation. Methods. In a comparative, small-scale feasibility study, 18 preterm infants born during their 27th to 28th gestational week (weighing <1500 grams) participated by convenience. Respiratory patterns (regular or irregular; shallow or deep) were directly observed on the 28th week during playback of speech recording. Heart rate variability was simultaneously measured using spectral analysis of heart periods, from which the mean values influenced by respiratory sinus arrhythmia (frequencies of 0.30–1.0 Hz) were compared to each observed respiratory pattern. The magnitudes of respiratory sinus arrhythmia and the area under the curve were determined. Results. The magnitude of respiratory sinus arrhythmia (frequencies of 0.30–1.0 Hz) in infants observed to be displaying irregular shallow respiration was greater than that in infants with regular deep respiration. Further, there was a shift from lower frequencies ($\text{frequency}\text{peak}=0.30\text{Hz}$) to higher frequencies ($\text{peak}=0.70\text{Hz}$). Conclusion. In contrast with direct observation, spectral analysis allowed for the quantification of respiratory patterns in a vulnerable population of preterm infants of interest to the nursing scientific and practice community. Future directions include applying this biomarker to evaluate both developmental and pathological trends in the respiratory patterns of preterm infants.