The Effects of Kangaroo Care on the Sleep and Wake States... : Journal of Nursing Research (2024)

Introduction

The organization of sleep and wake states relates closely to brain maturation and reflects the central well-being of the nervous system (^{Holditch-Davis, 2010}). The development of deep sleep and quiet awake is important for preterm infants (^{Als, 2009}). In neonatal intensive care units (NICUs), many factors impinge on the development of deep sleep and quiet awake, including the number of manipulations (^{Franco et al., 2010}), bright lights and high noise levels, maternal separation (^{Morgan, Horn, & Bergman, 2011}), mechanical ventilation (^{Lehtonen & Martin, 2004}), the frequent experience of pain (^{Polkki & Korhonen, 2014}), and health risks such as infection (^{Franco et al., 2010}). These factors lead to states of active wakefulness, frequent crying, sleep interruptions and deprivation, and excessive drowsiness (^{Franco et al., 2010}), resulting in the impairment of sensory development, disorders of consciousness (^{Graven & Browne, 2008}), and mental–social development, among other problems (^{Holditch-Davis, 2010}). Infants in the NICU rarely reach the deep sleep and quietly awake states, which are required for healthy development. Accordingly, it is of great importance to develop interventions that aid in the improvement of deep sleep and quiet awake states for preterm infants (^{Hunt, 2008}). Kangaroo care (KC), first developed in Bogota, Columbia (^{Feldman & Eidelman, 2003}), is a safe and effective method to care for preterm infants. Studies have associated the use of KC with reduced mortality, length of hospital stay, and hypothermia; increased weight gain; breastfeeding; and improved maternal–infant bonding and mother satisfaction (^{Conde-Agudelo, Belizán, & Diaz-Rossello, 2011}) as well as improved maturation of the prefrontal cortex and better organization of sleep patterns (^{Feldman, Rosenthal, & Eidelman, 2014}). Few studies have investigated the effect of KC on the development of sleep and wake states in preterm infants. Those published to date have focused on comparing KC with incubator care (^{Cong, Cusson, Hussain, Zhang, & Kelly, 2012}; ^{Cong, Ludington-Hoe, McCain, & Fu, 2009}). It is unknown whether the effect of KC on the sleep and wake states of preterm infant is because of the presence of the mother or skin-to-skin contact. The current study set out to compare sleep and wake states in preterm infants receiving KC with those who were held in their mothers’ arms. In most NICUs in Iran, as in many other Middle and Far Eastern countries, preterm infants, if held at all, are held in their mothers’ clothed arms, that is, “in-arms-holding” (IAH). This study set out to investigate the hypothesis that KC generates better deep sleep and quiet awake states in preterm infants than IAH.

Methods

Design

This study is a randomized controlled trial that was conducted in the NICU of a mother–fetus research center at Valiasr Hospital in Tehran, Iran. Data were collected from October 2011 to February 2012.

Sample

The authors took into consideration type I error (α = .05), type II error (β = 0.2), the mean and standard deviation sleep state scores from a previous study (^{Ferber & Makhoul, 2004}), and a sampling formula to estimate the minimum number of subjects necessary for each group as 31. Seventy consecutively born and successfully recruited infants were enrolled in the research, with 35 infants in each group. Participants met the following criteria: gestational ages (GAs) of 32–37 weeks; birth weights greater than 1500 g; 5-minute Apgar score ≥ 6; feeding every 2 hours; breathing room air without mechanical support; free of illnesses and conditions such as congenital brain malformations, meningitis, seizures, encephalopathy, congenital abnormalities, intraventricular hemorrhage grades III and IV, herpes, sepsis, and heart or metabolic diseases; and free of medication interventions that necessitate the use of theophylline, phenobarbital, midazolam, or phentalene. In addition, the physical health status of the mothers of eligible infants were required to be sufficient to perform the intervention conditions under study; for example, mothers should be free of undue pain while sitting and holding infants.

The study was to be terminated for infants who developed tachycardia, bradycardia, tachypnea, bradypnea, arterial oxygen saturation of less than 88%, temperature decrease by more than 1° from the baseline temperature, and/or any need for chest tube, peripheral arterial line, or central catheter placement during the course of the intervention.

One of the researchers screened newly admitted infants for eligibility and recruited all eligible infants in their order of admission to the NICU. Of the 78 eligible infants, 72 parents agreed to participate. The most frequent reason for ineligibility/refusal was pain because of Cesarean section. The flow chart for this study is shown in Figure 1. Once successfully recruited, the recruiter randomly assigned the infant into one of the two groups (KC and IAH) by use of a simple random-numbers table, which was prepared by a researcher who did not participate in data collection. Group assignment status was blinded to infants’ mothers until the subject was allocated and to the data analyst until the conclusion of the study. Only the data collector was trained in how to use the tool (identify and record the behaviors of infants). Therefore, the data collector collected all of the data for the KC and IAH groups. Accordingly, the randomization was not kept concealed from data collector.

Setting

The NICU aspects experienced by all subjects included a steady room temperature of 25°C (77°F), many large windows that were shielded with dark curtains, the availability of a mothers’ room for rest, and an unlimited amount of time with their infants. The infants were monitored throughout the study for physiological parameters that included heart rate, breathing rate, and arterial oxygen saturation.

Data Collection

Instruments

Demographic questionnaire

Demographic data included gender, feeding type, delivery type, 5-minute Apgar score, maternal age, GA, postnatal age, and weight.

Sleep and wake states of the Newborn Individualized Developmental Care And Assessment Program

The state definitions of the sleep and wake state of the Newborn Individualized Developmental Care and Assessment Program (NIDCAP; ^{Als, 1999}) were used to assess the sleep and wake states of the preterm infants. This scale differentiates six levels, including deep sleep, light sleep, drowsy, quietly awake and/or alert, actively awake and aroused, and highly aroused and agitated/crying. Each state level is defined by various behavioral cues such as breathing patterns, facial expressions, body/eye/mouth movements, various sounds, and skin color. Six levels of sleep and wake states are divided into two categories: States A are the disorganized and diffuse analogs of the States B, which are robust, well-modulated states. For the purpose of this study, these two forms were scored as one level. For example, the researcher marked “deep sleep” for observations of the behavioral cues of deep sleep A and deep sleep B.

In this study, as in formal NIDCAP observations, NIDCAP state levels were recorded in 2-minute time epochs, with the occurrence of the predominant state for the 2-minute block assigned to the respective 2-minute time block. To be recognized as a distinct state, the behavioral configuration was required to persist for at least 2–3 seconds. Two observers, trained to interrater reliability, simultaneously and independently observed and recorded the sleep and wake behaviors of 10 nonstudy infants in the course of the data collection process. A Pearson product–moment test of correlation indicated a correlation coefficient between the two observers of 78% (p = .008). One of the second authors performed all of the study observations.

Preintervention phase

The initial study session was performed during the first interval between the afternoon feeding times. Light levels and peripheral sounds were beyond the control of the investigators. To mitigate these effects, all observations were performed in the afternoon when the study NICU was typically quieter and more muted in terms of light. Furthermore, the afternoon offered easier access to mothers and fewer interruptions from nursing interventions. Half an hour before the onset of this study, the investigator explained to the mothers that she would now begin to prepare the infant and the mother for the intervention.

Mothers. Before the onset of the observations, all mothers were encouraged to take the following steps: empty their bladder and bowels, abstain from using perfume, pump breast milk in case of fullness, satisfy hunger and thirst, and turn off mobile phones. All mothers were assisted into a seated position in a comfortable chair with a 45° soft backrest and with feet supported on a footrest to prevent fatigue. The mothers in the KC group were asked, before taking their seat, to remove their upper clothes and don an open-front gown.

Infants. Before commencing the observations, all infants, clad in only a diaper, were placed by a nurse in supine position in their incubator with the mattress at a 30° angle. All infants were observed for 20 minutes. States were recorded onto a scan sheet every 2 minutes (an electronic timer transmitted a brief ringtone to the researcher’s ears through earphones at the beginning of each 2-minute interval). The intervention was initiated immediately after the 20-minute observation period.

Intervention phase

Kangaroo Care Condition. Once the mother was seated, the KC group infants, clad in only a diaper, were carefully placed onto their mothers’ bare chest at a modified diagonal angle, with the infants’ head near their mothers’ heart, to ensure an open airway and facilitate direct mother–infant eye contact and transmission of the mother’s heartbeat through her skin. KC infants were helped to tuck their arms and legs against their body. A soft blanket was gently wrapped around the infant and mother for warmth.

In-Arms-Holding Condition. The IAH group infants, clad in only a diaper, were gently wrapped in a soft bedsheet and a blanket and carefully cradled in their mothers’ arms, with the infants’ head and back supported by their mothers’ left arm, to ensure an open airway and facilitate mother–infant eye contact and transmission of the mothers’ heartbeat through her clothing and the blanket. IAH infants were helped to tuck their arms and legs under their blankets. The observations were conducted in the same way as the KC group previously described.

The privacy of the infant and mother was ensured by pulling the bedside curtain around them in two groups. For all infants, intravenous medication infusions were continued during the interventions. In this study, infants were fed, diapers were changed, and other nursing interventions were performed every 2 hours. As these interventions disrupt the normal sleep and wake responses of infants, we performed the intervention during the intervals between feeding. Furthermore, the instrument (NIDCAP) requires observation of the infants 20 minutes before and 20 minutes after the intervention. Evidence indicates that 60 minutes is the minimum time that KC affects the states of sleep and wake in preterm infants, with more than 60 minutes of KC being almost intolerable for the mother and her infant (^{Kadivar & Noorbakhsh, 2009}). Thus, the intervention period lasted for 70 minutes, during which time the observer recorded the infants’ states.

Postintervention phase

Upon completion of the respective intervention observations, the diaper-clad infants were returned to their incubators, placed in supine position, and again observed for 20 minutes, while their states were recorded as described.

Outcome Measures

The primary outcome measures included the mean percentages of time spent in deep sleep and in the quiet awake/alert state as measured separately before, during, and after intervention for each of the two groups. The secondary outcomes included the mean percentages of time spent in each of the other states, again as measured separately before, during, and after intervention for each of the two groups.

Data Analysis

The SPSS-17 software program was used to conduct all analyses. Group differences at a level of p < .05 were considered as significant. The Fisher’s exact test, chi-square, and independent t test were used to compare individual characteristics between groups. After examining the data, percentages of time spent in the six states were distributed normally. Therefore, repeated measures analysis of variance (RM-ANOVA) was used to determine the effect of the intervention on the states measured. Furthermore, the data analysts were blinded to the group assignment of infants.

Ethical Consideration

This research was confirmed by the research council and the research ethics committee of the Tehran University of Medical Sciences and registered with the Iranian Registry of Clinical Trials (IRCT201109157565N1). Data collection was performed after explaining the research objectives and obtaining informed consent from the mothers of infants. The researchers bound themselves to observe the ethical issues that were outlined in the Declaration of Helsinki. All mothers were assured of their anonymity, of the confidentiality of their and their infants’ personal information, and of their right to refuse to participation or to withdraw from the study at any time. In addition, the necessary permissions were sought from the hospital authorities and the NICUs.

Results

The two groups were comparable in terms of all the measured individual characteristics (p > .05; Table 1). Furthermore, in the preintervention phase, the two groups were comparable in terms of the percentage of time spent in the six states of sleep and wake, respectively (p > .05).

The mean of deep sleep increased to 23.08 ± 4.22 in the KC group and to 5.31 ± 5.88 in the IAH group during the intervention. Therefore, deep sleep in both groups increased between the first and second assessments, but the magnitude of the increase was greater in the KC group, and the difference between the two groups was significant at the second assessment (p = .001; Table 2). The deep sleep at the second intervention was 0.45 ± 1.59 and 0.42 ± 1.42 in the KC and IAH groups, respectively, which was not significantly different (p = .93; Table 2). Moreover, RM-ANOVA showed a significant difference between the two groups in terms of the infants’ mean scores of deep sleep during the three study phases (F = 184.05, p = .001; Table 2).

The mean for quiet awake was not different between two groups at the preintervention phase. However, during the intervention, the quiet awake in the KC group increased more than in the IAH group (p = .004). Furthermore, a significant difference was found between the two groups in the postintervention phase (p = .001; Table 2). Moreover, RM-ANOVA showed a significant difference between the two groups in terms of the infants’ quiet awake during the three study phases (F = 21.64, p < .001).

In addition, results showed that the IAH group spent more time in light sleep, the drowsy state (p < .001), and the actively awake state (p = .02) than the KC group. No significant differences between the groups were found for crying (p > .05; Table 2).

In the IAH and KC groups, significant differences were found between before and during test stages and also between during and after test stages for deep sleep (p < .001). Furthermore, significant differences were identified between before and during test stages for quiet awake (p < .001). Moreover, in terms of quiet awake, the KC group exhibited a significant difference between before and after test stages (p = .005), whereas the IAH group exhibited a significant difference between during and after test stages (p < .001; Table 3).

Discussion

This study investigated the effects of KC and IAH on the sleep and wake states of preterm infants. As the results show, the two groups were comparable in terms of individual characteristics as well as of the time spent in sleep and wake states in the preintervention phase. This increases the confidence in the “during intervention” and postintervention findings.

Overall, the hypothesis that KC increases the two desirable states of deep sleep and quiet awake/alert in preterm infants as compared with IAH was supported. The infants in the KC group exhibited more deep sleep during the intervention than the IAH group. This finding is consistent with the results of ^{Cong et al. (2012)}, who examined the effects of KC during heel-stick blood sampling on the responses of twin preterm infants. They found that both infants during KC spent 80%–100% of their time in deep sleep. The current study, which examined the effects of 70 minutes of KC implementation, showed deep sleep results that were highly similar to those of earlier KC studies (^{Chwo et al., 2002}; ^{Scher & Loparo, 2009}) that investigated longer or more frequent KC implementation times. In contrast to the results of this study, ^{Lai et al. (2006)} reported no significant difference for preterm infants in terms of deep sleep state as measured on the first day of their study in one of two experimental conditions, namely, 1 hour of music exposure during KC as compared with being kept in the incubator. Their state recording interval of once every 10 minutes might not have been sensitive enough to identify possible group differences. Furthermore, the simultaneity of the two intervention modalities, music and KC, may have obscured the KC effects on the deep sleep state.

As already mentioned, deep sleep is recognized as an important state for the promotion of brain development because of its effect on increasing synaptogenesis (^{Peirano, Algarin, & Uauy, 2003)}. Deep sleep reduces tension and stress in infants and enhances awake availability. In addition, respirations are well regulated, which enhances the delivery of oxygen to tissues (^{Kadivar & Noorbakhsh, 2009}).

The potential underlying reasons for increased deep sleep during KC include the mother’s body warmth, which may promote increased infant tranquility and deep sleep (^{Chwo et al., 2002}; ^{Kadivar & Noorbakhsh, 2009}), as well as the odor of the mother’s body and breasts and the soothing sound of the mother’s heartbeat in combination with the rhythmic movements of her chest (^{Arnon et al., 2014}; ^{Kadivar & Noorbakhsh, 2009}; ^{Schlez et al., 2011}). In the current KC condition, the infants were placed vertically between the breasts of their mothers, which appears to create a more confined, containing, and cradling space that may be analogous to the internal space of the uterus and may protect infants from contact with excessive peripheral sensory stimulation, increase infants’ sense of security and comfort, and help infants move into deep sleep (^{Kadivar & Noorbakhsh, 2009}).

The current investigation showed that the KC group spent significantly less time in light sleep than the IAH group. This contrasts with ^{Cong et al. (2009)}, who found no significant differences in light sleep between the KC and control group infants, either before or during heel blood sampling. This discrepancy may be because of the very small number of subjects (two) in the latter study. However, the much larger studies by ^{Ferber and Makhoul (2004)} for term infants and by ^{Feldman and Eidelman (2003)} for preterm infants concur with the results of this study. A reduction in light sleep may be desirable for preterm infants, as the hypoxemia risk is significantly increased in this state. The existence of effortful motor activities and the poor or lacking Hering Breuer reflex (^{Verklan & Walden, 2010}) are other characteristics of light sleep.

Regarding the state of drowsiness, the infants in the KC group showed significantly less drowsiness than those in the IAH group. In the drowsy state, the infant experiences the benefits of neither sleep nor wakefulness and efforts to wake up or to go to sleep (^{Kadivar & Noorbakhsh, 2009}). Similar to the results of this study, ^{Chwo et al. (2002)} found that preterm infants of 34–36 weeks of GA who had received KC for three 1-hour between-feeding intervals showed less drowsiness than infants who were wrapped in a blanket and held by their mothers (control group; 2% vs. 15%, p = .0001). In another study by ^{Begum et al. (2008)}, 1-hour KC showed no effect on the drowsiness of the preterm study infants. However, these results are difficult to interpret as there was no control group and the sample sizes were small (n = 16). Furthermore, four infants had been given theophylline before the study, which are likely explanations for the inconsistent results that were found in these two studies.

The quiet awake state is the most critical state related to effective attention and interaction in preterm infants (^{Boxwell, 2000}). In the current investigation, KC was associated with a more significant increase in the quiet awake state. Even in the postintervention phase, the KC group had longer quiet awake periods than the IAH group. ^{Chwo et al. (2002)} also showed that infants who received KC had significantly more quiet awake and alert periods than infants who were held in their mothers’ arms for feeding (14% vs. 7.3%, p = .0001). However, ^{Begum et al. (2008)} found that none of their preterm study infants entered into the quiet awake/alert state during KC; the infants studied were mainly in sleep states. As already mentioned, this study lacked a control group and may not have sufficient statistical power because of the small sample size (n = 16).

Reducing the infants’ active/awake and crying states have been shown to result in better integrated sleep (^{Ferber & Makhoul, 2004}). In this study, the KC group showed significantly less time in the active awake state than the IAH group. Conversely, crying showed no significant difference between the two groups—perhaps because neither had yet developed sufficient energy to cry. Nevertheless, both groups exhibited significantly less crying during the intervention as compared with the preintervention phase. This may be attributable to the mother’s closeness in the KC group and in IAH that served to soothe and decrease this state. Other studies have shown that KC reduces the amount of crying (^{Cong et al., 2012}).

Limitations

The infants’ states were measured by direct observation because of the cultural conditions at the site, which made videotaping impractical. However, in the future, videotaping is suggested to increase the reliability of the observations. It is further recommended to observe subjects for more than only one experimental period. Moreover, to increase the generalizability of the results, this study should be replicated with medically less stable preterm populations, who would presumably benefit even more from the KC intervention.

Conclusion

The KC intervention significantly enhances the states of deep sleep and quiet awake/alert and significantly decreases the states of light sleep, drowsiness, and actively awake in comparison with the IAH intervention. These findings add to the previous evidence of maternal skin-to-skin contact as a nonpharmacologic intervention for the improvement of state organization in preterm infants. The results provide an important reference for Iranian nurses and the nurses of other countries where resistance remains to recommending KC interventions in the NICU environment.

Acknowledgments

This study was supported by the Tehran University of Medical Science. We thank Gloria McAnulty, PhD, Department of Psychiatry, Boston Children’s Hospital and Harvard Medical School, for her statistical advice and consultation. Furthermore, we thank Elaheh Amini, MD, Department of Mother and Fetus Research Center, Valiasr Hospital, Tehran University of Medical Science, for her scientific consultation services.

References