Published ahead of print on June 19, 2008, doi:10.1164/rccm.200803-432OC
© 2008 American Thoracic Society doi: 10.1164/rccm.200803-432OC
Prenatal Cigarette Smoke Exposure Attenuates Recovery from Hypoxemic Challenge in Preterm Infants1 Department of Pediatrics and Institute of Maternal and Child Health, Faculty of Medicine, University of Calgary, Calgary, Alberta, Canada Correspondence and requests for reprints should be addressed to Shabih U. Hasan, M.D., Department of Pediatrics, Health Sciences Center, 3330 Hospital Drive NW, Calgary, AB, T2N 4N1 Canada. E-mail: hasans{at}ucalgary.ca
Rationale: The effects of prenatal cigarette smoke (CS) exposure and hypoxemia on cardiorespiratory control have been investigated in full-term infants. However, few data are available in preterm infants, who form a particularly vulnerable population, with developmentally immature cardiorespiratory control. Objectives: To investigate the effects of prenatal CS exposure on the duration and recovery of breathing pauses and oxygen saturation levels under baseline and hypoxemic conditions in preterm infants. Methods: The study was performed on 22 (12 born to smoking and 10 to nonsmoking mothers) spontaneously breathing preterm infants between 28 and 36 weeks' gestation. Cardiorespiratory variables were recorded under baseline normoxemic and hypoxemic conditions. Measurements and Main Results: Breathing pauses, pause indices, time to recovery, percent pause recovery, oxygen saturation (SpO2), periods of wakefulness, and cardiorespiratory rates were compared between the two groups. Spontaneous recovery of breathing pauses (P = 0.03) and SpO2 levels (P = 0.017) were attenuated in CS-exposed infants as compared with the control group during the hypoxemic and posthypoxemic periods, respectively. The episodes of wakefulness during the hypoxemic challenge were similar between the two groups. Furthermore, CS-exposed infants showed a greater increase in heart rate (P < 0.001) during the hypoxemic challenge when compared with control infants. Conclusions: We provide evidence of how prenatal CS exposure and hypoxemic episodes affect the duration and recovery of breathing pauses in preterm infants. These observations could help explain why these infants are at a particularly high risk for sudden infant death syndrome.
Key Words: apnea • hypoxemia • SIDS • smoking • oxygen saturation
Cigarette smoke (CS) exposure during pregnancy increases the risk of several devastating effects on the feto-placental unit, including spontaneous abortions, placental abruption, preterm birth, and low birth weight in a dose-dependent fashion (1). Furthermore, since the advocacy of the supine sleep position, pre- and postnatal exposure to CS is currently the principal and leading independent risk factor for the occurrence of sudden infant death syndrome (SIDS) (2–4). In addition, preterm birth has also been identified as a major risk factor for SIDS as compared with the infants born at term gestation (5, 6). Therefore, preterm infants who have been prenatally exposed to CS are at a particularly high risk of succumbing to SIDS (5). Despite the recent marked reduction in SIDS rates, SIDS remains the leading cause of infant deaths beyond the neonatal period (7, 8). Apneas occur frequently in preterm infants, and the incidence and severity of the apnea of prematurity are inversely related to the gestational age (9, 10). However, there is no evidence that suggests a relationship between apnea of prematurity and SIDS (11). Prenatal CS exposure has been shown to increase the frequency and duration of obstructive apneas in a dose-dependent manner in term infants (12). In addition, Sawnani and colleagues found that preterm infants exposed to prenatal CS have an increased apnea index for obstructive events during active sleep and a decreased arousal index with a specific decrease in percentage of arousals after respiratory events (13). However, the effects of hypoxia on respiratory control were not investigated. Previous studies performed in term infants born to smoking mothers were unable to detect differences in cardiorespiratory and behavioral responses under normoxic conditions. However, exposure to hypoxia uncovered blunted ventilatory responses (14–17). Furthermore, studies investigating the effects of prenatal nicotine exposure on hypoxic ventilatory responses in animals and term infants have provided contradictory results (15, 18–22). Evidence suggests that cardiorespiratory control abnormalities may play a role in SIDS (6, 23, 24) and that prenatal CS exposure affects respiratory control in term infants (12, 15). Although preterm infants exposed prenatally to CS form a particularly vulnerable population, few data are available on cardiorespiratory control in this group of infants. Furthermore, currently no data are available on the effects of prenatal CS exposure on respiratory control during hypoxemia, one of the factors postulated as an important trigger for SIDS (25, 26). The specific aims of the current study were to investigate the effects of prenatal CS exposure on the duration and recovery of breathing pauses under baseline and hypoxemic conditions in preterm infants. We hypothesize that preterm infants prenatally exposed to CS have an increased number of breathing pauses and impaired cardiorespiratory responses during the baseline and hypoxemic periods.
Study Population We studied 22 (12 born to smoking and 10 to nonsmoking or control mothers) spontaneously breathing neonates between 28 and 36 weeks' gestation. This study was designed to investigate healthy preterm infants without cardiorespiratory dysfunction, including the need for mechanical ventilation and without major cardiovascular and/or neurologic disorders. Therefore, neonates with a history of maternal recreational drug use or alcohol abuse during pregnancy, congenital malformations, craniofacial malformations, neuromuscular diseases, congenital heart disease, patent ductus arteriosus, proven or presumed septicemia within 3 days of study and/or meningitis, hypoxic–ischemic encephalopathy, necrotizing enterocolitis (any Bell's stage), or any intraventricular hemorrhage (Papile classification grade I–IV) were excluded from the study. Patent ductus arteriosus was excluded on both echocardiographic and/or clinical criteria. The infants requiring assisted ventilation, including continuous positive airway pressure to treat lung disease or apneas, were also excluded in both CS and control groups. As per the guidelines of the American College of Obstetricians and Gynecologists, hypoxic–ischemic encephalopathy was defined as an Apgar score of 3 or less at 5 minutes after birth, an umbilical arterial cord pH of less than 7.0 in association with metabolic acidosis (base deficit = 12 mmol/L), neonatal sequelae, and organ dysfunction within 72 hours of birth. Infants with intrauterine growth restrictions, defined as a birth weight 2 SD below the expected weight for gestation, were also excluded from the study.
Study Protocol Maternal and neonatal demographics were obtained from a thorough review of the medical records. Neonatal hemoglobin, hematocrit, length, and head circumference were recorded within the week preceding the study. The infant weight was obtained from the daily neonatal care flow sheet. All studies were performed using the Sandman-Elite neonatal sleep polygraphic diagnostic system, version 6.0 (Puritan Bennett, Kanata, ON, Canada). The standard infant montage was used and the following variables were recorded simultaneously: ECG, pulse oximetry and pulse wave form (Nellcor NPB-295; Puritan Bennett), nasal airflow (Ultima Dual Airflow Pressure Sensor 0580D; Braebon Medical Corp., Carp, ON, Canada), and chest and abdominal movements. Periods of behavioral wakefulness were recorded throughout the study. Infants were studied at the bedside in the neonatal level II units at the Rockyview General Hospital and the Foothills Medical Center in Calgary, which are staffed by a citywide regional neonatal team. The infants were not moved to the Sleep Laboratory to ensure maintenance of their usual environment. During the study, infants received routine care, including feeding, assessments, and medication administration. During the hypoxemic challenge, one of the authors (S.U.H.), a certified neonatologist, was always present at the bedside during the hypoxemic and posthypoxemic periods to closely monitor the infants. Full neonatal resuscitation equipment was readily available, if needed. The hypoxemic challenge was discontinued if the SpO2 levels fell below 80% for 1 minute or more of consecutive recording time. However, at times, the values fell below 80% during breathing pauses. This study consisted of two consecutive phases (Figure 1): phase I (baseline) and phase II (hypoxemic challenge). Phase I was recorded in basal inspired oxygen concentration over a 3-hour mean period of valid recording time. Infants were connected to the recording system immediately before their morning feed to minimize any disturbance to the infant and reduce the confounding effects of metabolic–ventilatory interactions. After the feed, infants were placed in their incubator, and once settled, the recording was started. Phase II, the hypoxemic challenge, which was given only once, immediately followed the baseline study. The hypoxemic challenge was divided into three 5-minute periods: prehypoxemia, hypoxemia, and posthypoxemia. The pre- and posthypoxemic periods were recorded under basal inspired oxygen conditions. During the hypoxemic period, medical grade hypoxic gas mixtures of 15% or 12% O2 with a balance of nitrogen (Praxair, Calgary, AB, Canada) were delivered through the Braebon 0582Ox-INF (Braebon, Kanata, ON, Canada) divided cannula at a flow rate of approximately 0.5 L/minute or less. The decision to use 12% versus 15% O2 depended on the baseline SpO2 values. Infants with a baseline SpO2 level of 95% or greater received 12% O2, whereas infants with a baseline SpO2 less than 95% received 15% O2. The inspired O2 concentration was decreased by 6% for infants receiving supplemental oxygen in accordance with institutional ethics guidelines.
The biologic variables obtained from the baseline study and hypoxemic challenge were evaluated for the following: breathing pauses, pause index, time to recovery, percent pause recovery, and respiratory and heart rates. Breathing pauses were defined as cessation of nasal airflow accompanied by either asynchrony of chest and abdominal movements or no respiratory effort from the chest and abdomen. Pauses lasting for 3 seconds or longer were identified and classified by duration (3–5 s, 5.1–10 s, 10.1–15 s, and >15 s). Pause index, expressed as number of pauses per hour, was calculated for each category. Breathing pauses were defined as recovered when the SpO2 values, after the breathing pause, reached the initial prepause values. If the SpO2 values did not reach prepause levels before the onset of a subsequent pause, the initial breathing pause was classified as "not recovered." The time to recovery was calculated for each breathing pause from the time the SpO2 reached the nadir after the breathing pause until the time at which the SpO2 returned to its prepause value. The number of pauses from which the neonate did and did not recover was recorded and the percent pause recovery was determined. The hypoxemic challenge was administered while the infants were asleep. Furthermore, we investigated whether termination of breathing pauses during the hypoxemic challenge was associated with a change in behavioral state. Wakefulness was defined using behavioral criteria (27).
Statistical Analysis
Demographic Data Twenty-two infants met the criteria for enrollment into the study. There were 12 neonates in the CS-exposed group and 10 neonates in the control group. The maternal and neonatal demographic data are presented in Tables 1 and 2. The birth weight and the gestational ages were not statistically different between the CS and control groups. Similarly, there was no difference in postconceptional age or infant weight at the time of study. The number of infants receiving caffeine and supplemental oxygen therapy were similar between the CS and control groups. During the hypoxemic challenge, nine neonates in the CS-exposed group and six neonates in the control groups received the hypoxic gas mixture, whereas three CS-exposed infants and four control infants required a reduction in FIO2 (P = 0.65).
Cardiorespiratory Variables during Baseline (Phase I) The pause indices of various durations were similar between the prenatal CS-exposed and control group (Figure 2). In addition, the respiratory rate was not different between the CS and control groups (56.5 ± 4.4 and 63.1 ± 3.9 breaths/min, respectively). Furthermore, the time to recovery and the percent pause recovery were similar between the CS and control groups during the baseline phase (22.4 ± 1.4 and 23.1 ± 3.1, respectively; P = 0.86).
Cardiorespiratory Variables during Hypoxemic Challenge (Phase II) Respiratory rates during the prehypoxemic and hypoxemic periods were similar between the two groups. During the posthypoxemic period, both CS and control infants showed an increase in respiratory rate as compared with the hypoxemic period (P = 0.004; Figure 3A). During the hypoxemic challenge, there was a similar increase only in the pause index for pauses of 3 to 5 seconds for both groups (P = 0.007). The infants in the CS-exposed group showed an increase in heart rate during the hypoxemic period as compared with prehypoxemic values (Figure 3B; P < 0.001). In contrast, no difference in heart rates was observed in control infants between the prehypoxemic and hypoxemic periods.
The SpO2 values for both the CS-exposed and control groups decreased in a similar fashion during the hypoxemic period. However, in contrast to the infants in the control group, prenatal CS-exposed infants showed an attenuated SpO2 recovery during the posthypoxemic period as compared with prehypoxemic values (P = 0.017; 95% confidence interval [CI] –14.8 to –6.9; Figure 4A). Furthermore, the infants in the CS-exposed group showed a decreased percent pause recovery during the hypoxemic period as compared with the control infants (P = 0.03; Figure 4B). Subanalysis of the four groups comprising CS and control infants receiving hypoxic gas mixtures or lowered FIO2 to induce hypoxemia did not show any intergroup differences for any of the recorded cardiorespiratory variables.
Behavioral Arousal during Hypoxemic Challenge (Phase II) Episodes of wakefulness were observed in all infants in the CS group and in all but one infant in the control group during hypoxemic challenge. No significant difference was observed in the number of breathing pauses associated with wakefulness between the two groups (Fisher's exact test; P = 1.0).
Our current study contributes to the understanding of cardiorespiratory control during hypoxemia in preterm infants exposed to prenatal CS. We have shown that, compared with infants of nonsmoking mothers, spontaneous recovery of breathing pauses and oxygen saturation values are adversely affected in CS-exposed infants during the hypoxemic and posthypoxemic periods, respectively. Furthermore, CS-exposed infants showed a greater increase in heart rate during the hypoxemic challenge when compared with control infants. Finally, we did not observe any difference in the recorded cardiorespiratory variables in preterm infants of smoking versus nonsmoking mothers under baseline conditions. These observations may provide insights as to why preterm infants exposed prenatally to CS are particularly vulnerable to succumbing to SIDS. Studies indicate that a combination of hypoxia and hypercarbia may be acute precursors to SIDS (28). Infants at the greatest risk for SIDS have been shown to have both attenuated arousal and ventilatory responses to hypoxia (14, 16) and/or hypercarbia (29), although some studies have found divergent results (22). Attenuation of both the spontaneous recovery of breathing pauses and oxygen saturation levels in relation to hypoxic and asphyxial challenges has not previously been reported in CS-exposed preterm infants. A number of studies have investigated the ventilatory responses during hypoxia in animal models (18–20). A recent study has provided evidence that prenatal CS exposure under hypoxemic and/or thermal stress can lead to severe adverse effects on respiratory control (30). The results from animal studies that use nicotine as a surrogate for cigarette smoking need to be interpreted with caution because the effects of nicotine are dependent on the target cell, dose, route, and duration of exposure, and these factors may help explain the divergent results of these animal studies. The relevance and validity of animal work in relation to CS and nicotine exposure have been discussed in detail previously (31, 32). In term infants, investigation into the effects of prenatal CS exposure on ventilatory responses has produced inconsistent results (15, 16, 21, 22), which may be due in part to vastly different experimental protocols, including the use of sedation (15), increasing postnatal ages, and different combinations of gas mixtures (22), as well as by the extent of pre- and postnatal CS exposure. Our baseline results further corroborate the finding of other investigators (14, 22), who also did not observe a difference in the resting respiratory rate or heart rate between CS-exposed and control infants. Furthermore, in the present study, no difference was seen in the pause index under baseline conditions between the study groups. Although the effects of prenatal CS exposure on the number and duration of apneas have previously been reported in term infants (12, 33), only limited information from a single study is available for preterm infants (13). It is, however, difficult to make meaningful comparisons with the present study because, in addition to using hypoxemic challenge, we evaluated neonates at a much younger postconceptional age than the infants in the previous study (13). The precise mechanisms through which prenatal CS exposure decreased the spontaneous recovery of breathing pauses and oxygen saturation levels during hypoxemia in the current study remain unknown. However, two distinct possibilities exist: effects on neural control of breathing and altered pulmonary development. Evidence suggests that nicotine interacts with highly selective endogenous neuronal nicotinic acetylcholine receptors (34–36) and may affect the development of areas in the central nervous system essential for respiratory control (37). Second, alterations in lung development (38) and lung mechanics (39–41) may also contribute to suboptimal gas exchange, leading to attenuated hypoxemic recovery observed in the CS-exposed group. The trend of the respiratory rate to increase more dramatically in the control group as compared with the CS group during the posthypoxemic period versus the prehypoxemic period lends support to the neural mechanism for reduced pause recovery in the CS-exposed infants (42). Studies using auditory stimuli have shown that infants of smokers are less arousable than those born to nonsmoking mothers (43–45). Lewis and Bosque provided evidence that fetal exposure to CS impairs chemoreceptor control of hypoxic awakening responses, although the ventilatory responses to hypoxia and hypercapnia were similar in both groups (14). In contrast, Campbell and coworkers observed no difference in waking responses to hypercarbic hypoxic challenge between the infants of smoking and nonsmoking mothers (22). In summary, arousal responses are an important although not critical mechanism for the termination of breathing pause (46). In the current study, we found that prenatal CS-exposed infants had an increase in heart rate during the hypoxemic period, whereas no significant increase in heart rate was observed in the control group. Animal studies have found that prenatal nicotine exposure decreases the heart rate response to hypoxia (18, 19) and the magnitude of the diminished heart rate responsiveness is both dose and age dependent (47). In human infants exposed prenatally to CS, increasing number of cigarettes smoked has been correlated with smaller heart rate increases during hypoxia (48). However, the results of these studies may not be applicable to our preterm population in which parasympathetic–sympathetic interaction is less well developed (49). It has been shown that the autonomic nervous system requires 37 weeks of in utero development to reach a threshold value (50), and preterm infants have a deficit in the degree of parasympathetic maturation (49, 51). In our preterm population, the hypoxemic challenge would have elicited a sympathetic response, heavily dependent on medullary regulatory sites (52). In contrast, mature and older infants have more regulation by rostral brain structures, which are considered "protective" for cardiovascular challenges (52). In addition, prenatal CS exposure has been shown to decrease the number of surviving Purkinje cells in the cerebellum in rat offspring, which may further enhance the sympathetic outflow (53). In the current study, hypoxemia was induced either via reduction in the inspired O2 concentrations in infants receiving supplemental O2 or by exposure to hypoxic gas mixture to the infants in room air. The difference for inducing hypoxemia could potentially affect the cardiorespiratory responses. Alveolar hypoxia could result in increased pulmonary vascular resistance causing ventilation–perfusion mismatch. However, in our infants, alveolar hypoxia was administered only for 5 minutes and prolonged alveolar/arterial hypoxemia or repeated hypoxic/hypoxemic episodes are usually the triggering events to increase pulmonary vascular resistance (54). Alveolar hypoxia could also affect the hypoxic sensing neuroepithelial bodies, although their precise function in this regard has not been clearly elucidated (55). Systemic arterial hypoxemia in infants with chronic lung disease (infants on supplemental oxygen) may not elicit a similar ventilatory response due to the underlying lung disease as compared with the infants with relatively normal lungs. However, in our study, no difference was observed in the cardiorespiratory variables during hypoxemia between the infants receiving the hypoxic gas mixture and those requiring a reduction in inspired oxygen levels. Because preterm infants continue to have significant cardiorespiratory events after discharge from the hospital (56), our study may help identify the infants at risk for attenuated recovery from hypoxemic episodes while at home. Furthermore, it might help distinguish the infants who will arouse in response to hypoxemia. The infants identified to be at risk can subsequently be further investigated and/or monitored at home. Our study has certain methodologic limitations that deserve comment. Because we did not recruit infants until 2 to 3 weeks after birth, we were unable to measure urine cotinine concentrations. However, studies have shown strong correlation between urine cotinine concentrations and self-reported smoking (57). In the current study, EEG-determined sleep stages were not taken into consideration while evaluating the infants and it is possible that attenuated recovery of breathing pauses and oxygen saturation levels during the hypoxemic challenge might, at least in part, have resulted from differing sleep stages and/or diminished EEG-based arousability as has been previously reported in term infants (16, 17, 45). Although infants were studied over a wide range of gestational ages (28–36 wk), which may have an effect on certain cardiorespiratory variables, there was no difference in the gestational age either at birth or at the time of study between the two groups. No previous studies have investigated the effects of prenatal CS exposure and hypoxemic challenge on cardiorespiratory control in preterm infants and thus the incidence and prevalence of breathing pauses and the ability to recover from breathing pauses in this group of infants are not known, limiting our ability to determine exact sample size. Thus, our sample size represents one of convenience. Furthermore, it was extremely challenging to obtain parental consent for exposure to a hypoxemic gas mixture, especially from the parents of the CS-exposed infants. Due to the limited availability of parental consent for hypoxia, other studies have also used small numbers of infants. In conclusion, our study provides data on the effects of prenatal CS exposure on the recovery of breathing pauses and oxygen saturation levels during hypoxemia in preterm infants. The attenuation of these responses in CS-exposed preterm infants signifies the deleterious effects of prenatal CS exposure on respiratory control in preterm infants and may provide new insights into the physiologic mechanisms through which prenatal CS exposure places these infants at a particularly high risk for SIDS. Furthermore, this study may help further increase public awareness on the adverse effects of prenatal CS exposure.
The authors thank the parents of the infants who graciously agreed to participate in the study. They also acknowledge the Canadian Institute of Health Research, the Alberta Heritage Foundation for Medical Research, and the SIDS Calgary Society for their financial support of this project; Dr. Tak Fung, Associate Professor, Faculty of Sciences; and Gisela Engels, Department of Information Technologies, University of Calgary, for assistance with the statistical analysis; and Ms. Linda Brigan for her assistance with the manuscript formatting.
Supported by the Canadian Institutes of Health Research, Alberta Heritage Foundation for Medical Research, and SIDS Calgary Society. Originally Published in Press as DOI: 10.1164/rccm.200803-432OC on June 19, 2008 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form March 19, 2008; accepted in final form June 16, 2008
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
2 test as appropriate. Respiratory rate, percent pause recovery, and pause index during the baseline study were compared between the two groups using the unpaired t test. Analysis of variance (ANOVA) for repeated measures was used to compare the pause index, breathing rate, heart rate, SpO2, and percent pause recovery between the two groups during the hypoxemic challenge. When a statistically significant condition effect (within-subject effect) was detected by the ANOVA, paired t tests were performed (with Bonferroni's correction) to identify where the differences were. Furthermore, a subanalysis was performed on four groups (CS, control, those who received hypoxic gas mixtures, and those received lower inspired oxygen to induce hypoxemia) using ANOVA for repeated measures followed by Student-Newman-Keuls post hoc test. Onset of wakefulness during the hypoxemic challenge was analyzed using Fisher's exact test. All results are presented as mean ± SEM, and a P value of 0.05 or less was considered to be statistically significant. 
P = 0.02). No change in heart rate was seen in the control (open triangles) group. No significant difference was seen in the heart rate between the two groups.
P = 0.03 between the groups).