 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* Department of Pediatrics, Yale University School of Medicine, New Haven, Connecticut
Yale Child Study Center, Yale University School of Medicine, New Haven, Connecticut
Division of Neonatology, Bridgeport Hospital, Bridgeport, Connecticut
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods. Premature infants who were born at <32 weeks’ postmenstrual age and were medically stable in neonatal intensive care unit rooms were randomly assigned between 32 and 34 weeks’ postmenstrual age to either continuous dim lighting (<25 lux; duration 24 days; control group; n = 29) or cycled lighting (239 ± 29 lux, 7:00 AM to 7:00 PM; <25 lux, 7:00 PM to 7:00 AM; duration: 25 days; experimental group; n = 33). Activity was continuously monitored from enrollment until approximately 1 month after discharge from the hospital. Weight and head circumference were also assessed up to 6 months after discharge from the hospital.
Results. Over the first 10 days at home, distinct day-night differences in activity were not seen in control subjects (D day-night: N 1.07 ± 0.02), but experimental group infants were more active during the day than at night (day-night: 1.25 ± 0.03). It was not until 21 to 30 days after discharge that day-night activity ratios in control infants matched those seen in experimental group infants shortly after discharge, yet even at this age, experimental group infants (day-night: 2.13 ± 0.19) were considerably more active during the day than at night as compared with control subjects (day-night: 1.43 ± 0.09).
Conclusion. Exposure of premature infants to low-intensity cycled lighting in the hospital nursery induces distinct patterns of rest-activity that are apparent within 1 week after discharge. In comparison, the appearance of distinct patterns of rest and activity are delayed in infants who are exposed to continuous dim lighting in the hospital. These observations show that day-night rhythms in activity patterns can be detected shortly after discharge to home in premature infants and that the circadian clock of developing infants is entrained by cycled lighting.
Key Words: circadian rhythm • premature • light • sleep • infant
Abbreviations: SCN, suprachiasmatic nuclei
Circadian rhythms are endogenously generated rhythms that have a period length of approximately 24 hours and synchronize physiologic processes with the 24-hour light-dark cycle and include the sleep-wake (rest-activity) cycle and daily rhythms in hormone production.1 It is well established that circadian rhythms are generated and regulated by a biological clock located in the suprachiasmatic nuclei (SCN) of the anterior hypothalamus.1
The oscillations of the SCN are not exactly 24 hours; thus, the circadian pacemaker needs to be readjusted on a daily basis so that endogenous circadian phase is synchronized (or entrained) with the external light-dark cycle.1,2 In the absence of such time cues, sleep-wake cycles and other hormonal rhythms will drift out of phase with the light-dark cycle.2 Light is the dominant signal that entrains the SCN.1,2 It was previously believed that high-intensity lighting was needed to regulate the human circadian clock,3 but it is now clear that moderate indoor lighting can entrain human adults.2,4
Little is known about how light regulates the circadian pacemaker during human development. Anatomic and functional studies in nonhuman primates thus have been used to provide insights into human circadian timing system development. On the basis of these studies, it seems that the SCN are present at the end of the first trimester of gestation and the SCN begin to have daily oscillations in utero.5 Day-night rhythms in heart rate and respiratory active and adrenal steroidogenesis can be detected in fetuses that are regulated by the mother.6,7 It also seems that 200 lux, which is equivalent to lighting in a moderately illuminated office, can entrain circadian phase5,8 and that SCN responsiveness to light begins at stages equivalent to 25 weeks’ postmenstrual age in humans.9
In clinical medicine, the importance of circadian biology is becoming increasingly appreciated, yet little consideration has been given to the importance of diurnal lighting cycles in the care of premature infants who are reared for extended periods in hospital nurseries. At present, we are unaware of systematic surveys that have assessed nursery lighting conditions either in the United States or in other countries. However, in many neonatal intensive care units in the United States, Canada, and Europe, both premature and term infants are not exposed to cycled lighting.10 Rather, infants are continuously exposed to dim lighting, which is achieved by using low-level room lighting and crib covers.10
The practice of nursery lighting has also changed in the past several decades, without a clear basis. Cycled lighting was often used in the hospital nurseries in the 1950s and 1960s, followed by continuous lighting when isolettes and neonatal intensive care units were introduced. In reaction to continuous bright light, continuous dim light was introduced in the 1980s and 1990s.
At present, it is not known how these different illumination practices influence the development of expressed rhythmicity in infants.10 To address this issue, we examined whether cycled lighting influences the development of expressed rhythmicity in patterns of rest and activity in premature infants. Patterns of rest and activity were selected for analysis, as activity patterns represent outputs of the circadian system and are among the earliest expressed circadian rhythms.5,8 Furthermore, is very difficult to detect day-night differences either in heart rate or in melatonin or cortisol production at <3 months of age.5 The studies reported here are parts of a comprehensive series of experiments examining the effects of lighting on infant and parent behavior.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Randomization
One individual randomly assigned infants to the control or experimental group using a table of random numbers. Two investigators enrolled the patients and supervised the lighting conditions. The investigators who analyzed and interpreted the data were blinded to the treatment groups.
Lighting Conditions
Infants were studied in the continuing-care rooms used to care for maturing, preterm infants in the Yale-New Haven Hospital (n = 42) or Bridgeport Hospital (n = 20) newborn special care units. In both groups, infants were fed every 3 to 4 hours. For infants who required an external heat source, the temperature in the isolettes was regulated to maintain a body temperature of 36.5°C to 37°C. When infants were mature enough to maintain their body temperature without a supplemental heat source, they were monitored to ensure that it was in the same range.
The different lighting conditions were achieved as follows. In all infant care rooms, background lighting was generally between 100 and 200 lux, with lighting provided by overhead fluorescent lighting. Control infants (dim lighting group) were kept in isolettes or cribs covered with quilts throughout the 24-hour day. This is currently the standard practice in the newborn special care units. Quilts were made of dark cotton material that reduced transmitted light. With the typical background lighting in the nurseries, the amount of transmitted light was 1 to 5 lux, resulting in continuous dim lighting.
Cycled lighting in the experimental group was achieved by removing the isolette or crib covers from 7:00 AM to 7:00 PM. During this period, an additional fluorescent light in the ceiling above the crib was turned on. From 7:00 PM to 7:00 AM, the isolettes or cribs were covered with quilts and the additional overhead lights were turned off.
No attempts were made to influence the lighting conditions to which the infants were exposed at home. Parents were instructed to illuminate their home as was their customary practice.
Light Monitoring
Background lighting in the nursery rooms was monitored using VWR light meters. In addition, an Actiwatch-L light sensor (Minimitter Co, Sun River, OR) was placed in each isolette near the head of the infant to monitor light exposure continuously during hospitalization. Thus, light measurements reflected the light reaching the infant. Actiwatch-L sensors were calibrated by the manufacturer to monitor lighting intensities <2000 lux. Light exposure data stored in Actiwatch-L devices were downloaded to a computer by placing the devices on a reader attached to an IBM computer. Data were then saved and analyzed using Actiware software (Minimitter). We did not monitor home lighting conditions.
Activity Assessment
We assessed activity in all enrolled infants. Activity was monitored continuously from the time of enrollment until 30 days after discharge from the hospital using Actiwatches (Minimitter) in 14 control and 15 experimental group infants. Activity was monitored continuously from the time of enrollment until 20 days after discharge from the hospital in 27 control and 24 experimental group infants. In 11 infants, activity data were not available through 20 days after discharge from the hospital, as a result of either premature removal of the watches or mechanical failure.
Actiwatches were placed on 1 ankle, and the position of the watch was changed from one ankle to the other at biweekly intervals. The manufacturer calibrated the Actiwatches so that collection of data and sensitivity between watches were consistent, with the number of activity counts recorded reflecting the degree and the speed of motion. Actiwatches record a digitally integrated measure of gross motor activity that is assessed by an internal accelerometer. Watch sensitivity was <0.01 Newtons.
Activity data were analyzed for successive 10-day blocks relative to the discharge date by 3 approaches. First, the absolute number of movements per day was determined. Second, for assessing whether there was diurnal variation in movement, day-night activity ratios were calculated from the absolute numbers of movements (not the duration of movements) from 7:00 AM to 7:00 PM and from 7:00 PM to 7:00 AM. This analysis was conducted for each infant to correct for potential differences between individuals and differences in Actiwatch sensitivity. Third, the circadian periods were calculated using 
2 periodogram analysis to assess periodicity, with cutoff set to detect period lengths between 12 and 36 hours.11 Analyses were conducted using Actiware software (Minimitter).
Statistical Analysis
Sample size was determined by power analysis using SamplePower (1.0; SPSS, Inc, Chicago, IL). For a 2-tailed 
level of .05, we calculated that 20 subjects in each group would be needed to observe a large effect size of 0.40 with a power of 0.76.
Multivariate parametric analysis was conducted using the repeated measures analysis of variance with Bonferroni posttest comparisons among all groups. For paired comparisons, t tests were conducted. Tests for homogeneity of variance revealed no significant differences in variances among the observations of day-night activity. Correlation coefficients were determined by least squares analysis. Statistical calculations were conducted using GraphPad Prism version 3.00 for Windows (San Diego, CA) and SamplePower (1.0; SPSS, Inc). Data are presented as mean ± standard error of the mean in text and in figures.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
In the control group, average lighting exposure from 7:00 AM to 7:00 PM was 28.5 ± 3 lux; from 7:00 PM to 7:00 AM, lighting exposure was 15 ± 5 lux. In the experimental group, the lighting exposure from 7:00 AM to 7:00 PM was 239 ± 29 lux (P < .0001 vs controls).
Actograms were generated for each infant from the time of enrollment until approximately 1 month after discharge from the hospital (Fig 1). Demonstrating the utility of Actiwatches for monitoring the activity of preterm infants, we were readily able to detect activity in all infants.
|
Next, activity data were examined to assess whether the distribution of activity varied between day (7:00 AM to 7:00 PM) and night (7:00 PM to 7:00 AM). When actograms of the control infants were inspected, day-night differences in rest and activity were difficult to observe in most infants over the 10 days preceding discharge from the hospital (Figs 1 and 2). Ratios of day-night activity were 1.07 ± 0.02 for control infants, indicating 7% more total activity during day than night. When actograms of experimental group infants were inspected over the 10 days before discharge, the day-night activity ratio was 1.25 ± 0.03, indicating 25% more activity during the day than night.
|
When we examined the infants with activity data for 20 days postdischarge, there was a significant change over time for infants in both groups for the day-night activity ratios, both before and after discharge (P < .001), and a significant main effect for lighting condition (P < .001). Across all assessment points, infants in the cycled lighting condition showed higher day-night activity ratios. There was also a time by lighting condition interaction (P = .005; Fig 2a). Post hoc contrasts showed that this interaction was accounted for by a greater increase in day-night phasic activity for the cycled lighting condition infants in the first 20 days after discharge (P = .002).
Similar findings were obtained when we examined the 19 experimental and 18 control group infants who had activity data for 1 month after discharge (Fig 2b). There was again a significant change over time for infants in both groups for the day-night activity (P < .001) and a significant main effect for lighting condition (P < .001). There was also a time by lighting condition interaction (P = .002). Post hoc contrasts showed that this interaction was accounted for by a greater increase in day-night phasic activity for the cycled lighting condition infants in the first 20 days after discharge (P = .008) and the first 30 days after discharge (P = .003).
We also observed that infants who were raised in either dim or cycled lighting showed higher day-night activity ratios over the first 10 days after discharge to home (Fig 3). Thus, it seems that patterns of in-hospital care influences expressed activity.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
When control and experimental groups were compared, they were similar in terms of birth weight and gestational age. The infants also had similar lengths of stays and the same weights at discharge. Thus, it is unlikely that physical or developmental variation among the groups contributed to the observed differences.
Although it is known that the primate circadian pacemaker begins oscillating during gestation, robust circadian rhythms in activity or humoral substances are not generally apparent until 2 months after birth.5 However, in human and nonhuman primate newborns, day-night differences in rest-activity patterns can be detected shortly after birth, with more absolute movement occurring during the day than the night.5,8,12 Thus, to assess whether there was diurnal variation in movement, we used day-night activity ratios to provide an index of expressed rhythmicity, along with period analysis.
When we examined patterns of rest and activity in the infants who were exposed to constant dim lighting in the nursery, we found that day-night differences in rest and activity patterns were generally not apparent for most control infants over the first 20 days at home, yet after 3 weeks at home, day-night differences in activity became apparent. In contrast, day-night differences in activity were seen over the first week after discharge in the infants who were exposed to cycled lighting before discharge. Supporting the notion that the infants who were exposed to cycled lighting had entrained circadian rhythmicity, the average period was 24 hours for the infants who were exposed to cycled lighting, whereas the period was 24.77 hours for the control infants, indicating that they were not entrained to the 24-hour day.
The appearance of day-night differences in activity shortly after discharge was also so strikingly apparent in infants in the experimental group that it was possible to identify the discharge date by visual examination of actograms. It was not possible to do this in the control infants. The appearance of day-night rhythmicity soon after discharge following in-hospital exposure to cycled lighting suggests that patterns of in-hospital care mask or disrupt expressed rhythmicity.
We also assessed whether day-night activity ratios or infant period after discharge was related to the number of days in cycled lighting. No such effects were observed. Thus, it is possible that the shortest period in cycled lighting was sufficient to influence the experimental group infants. This observation is consistent with findings in human adults suggesting that >10 days of cycled lighting exposure is sufficient to result in biological clock entrainment to new lighting cycles.2
Potential influences of cycled lighting on premature infants have been the subject of a few previous studies. In the Stanford Cycled Lighting Study,10,13 differences in circadian rhythms in temperature were not detected among infants who were exposed to either continuous dim lighting or cycled lighting before discharge.10,13 These infants were studied 1 and 3 months after discharge. Because we observed that infants in both groups manifest similar circadian phase by 30 days of age, treatment effects on the rhythm of core body temperature may no longer be distinct after 1 month of age.
Other investigators have suggested that exposing infants to light-dark cycles improves infant weight gain. Mann et al14 found that exposure to light-dark cycles before discharge resulted in better weight gain and more sleep over the 24-hour day than did chaotic lighting patterns. These effects were seen 6 weeks after discharge and not sooner.14 Because of this lag period, it has been suggested that the observed effects were not a direct result of cycled lighting exposure on the infant.10 Miller et al15 suggested that there was improved health of infants who were reared in cycled as compared with continuous dim lighting. More recently, it was suggested that exposing infants to light-dark cycles improves the in-hospital growth of infants if exposure occurs before 36 weeks of age.16 However, it seems that the infants who were not exposed to cycled lighting were more ill than the cycled lighting groups, as reflected by longer durations of hospitalization and ventilatory support.16 In comparison, studying infants closely matched at enrollment, we do not observe changes in growth in hospital.
Previous studies have suggested that day-night rhythmicity is not apparent in prematurely born infants until nearly 1 month after term-birth age equivalency is reached (>42 weeks’ postmenstrual age).10,13,17 These conclusions have been based on 24- to 48-hour assessments of rectal temperature and/or activity and sleep patterns, yet, using actigraphy to monitor rest-activity patterns for extended periods, we find that circadian phase can be detected in infants who are exposed to cycled lighting as early as a 37 weeks’ postmenstrual age. These observations demonstrate the clear advantage of using actigraphy, which provides an objective measure of activity, over the subjective nature of rest-activity assessment that is based on parental observations that are recorded in diaries.
In previous studies of nonhuman primate infants that were reared in constant conditions, it has been observed that day-night differences in rest and activity are apparent shortly after term birth.8 It was also found that day-night differences in activity could be detected several weeks before circadian rhythms in core body temperature can be observed using internal telemetry devices,8,18 possibly as a result of immature temperature regulation in the newborn.19 Thus, rest-activity patterns may provide the earliest index of developing circadian rhythmicity in infants, although other investigators regard the temperature rhythm as 1 of the earliest expressed circadian rhythms.13,20
Cycled lighting can potentially influence expressed rhythmicity by several means. First, cycled lighting can entrain the circadian clock to the phase of the solar light-dark cycle. Second, daytime hospital light may cause infants to be more active during the day, and these effects carry over to the home environment. Because it is difficult to assess day-night rhythms in temperature or humoral factors (melatonin, cortisol) in term infants who are <1 to 2 months of age,21,22 we do not have an independent marker of endogenous circadian phase to answer this question. Third, it is possible that the intervention influenced home lighting, which in turn influences the infant. In this study, we did not assess the lighting conditions at home. Thus, we do not know whether assignment to the cycled lighting group caused parents to illuminate homes differently than in the control group. Fourth, although the background lighting in the nursery was not influenced by the study, it is also possible that the different interventions influenced staff behavior, which indirectly influenced the infants. Nevertheless, irrespective of the mechanism, exposure of preterm infants to light-dark cycles in the nursery results in the early establishment of entrained rest-activity cycles.
Although continuous dim lighting is the current practice in most nurseries in the United States, the scientific basis for this practice is not clear.23 It has been suggested that nursery lighting may contribute to eye disease in premature infants,24 yet rigorous clinical studies have failed to show adverse effects of ambient nursery lighting on premature infants.25,26 Investigators who propose a Neonatal Individualized Developmental Care Assessment Program have also suggested that infants should be dark-reared.27 This approach overlooks that before birth, the infant is exposed to maternal time-of-day cues that synchronize the fetal clock with the external light-dark cycle.5 Rearing premature infants in the dark deprives infants of the time-of-day information that they would have received during full gestation. It is important to note that recent observations suggest that a dark-rearing approach does not improve developmental outcome or the patterns of sleep of premature infants.28 Furthermore, exposing premature infants to cycled lighting does not disrupt sleep organization.29
In summary, we find that by exposing premature infants to low-intensity cycled lighting while infants are maturing before discharge, we can induce rest-activity patterns that are apparent soon after the child leaves the hospital. These observations suggest that the circadian clock of developing infants is entrained by cycled lighting. Our studies also suggest that during the 2 weeks before discharge to home, cycled lighting is preferable to continuous dim lighting in the care of premature infants
|
ACKNOWLEDGMENTS |
|---|
Some of these data have been presented at the Society for Pediatric Research Meeting, May 2002,30 and referred to in a review article.31
Special thanks to Elaine Sherwonit, Sarah Conti, Patricia Gettner, Ireneusz Sielsk, and Carol Billingham for help with these studies.
|
FOOTNOTES |
|---|
Reprint requests to (S.A.R.) Yale Pediatrics, Yale Child Health Research Center, PO Box 208081, New Haven, CT 06520-8081. E-mail: Scott.Rivkees{at}yale.edu
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. M. Ludington-Hoe, M. W. Johnson, K. Morgan, T. Lewis, J. Gutman, P. D. Wilson, and M. S. Scher Neurophysiologic Assessment of Neonatal Sleep Organization: Preliminary Results of a Randomized, Controlled Trial of Skin Contact With Preterm Infants Pediatrics, May 1, 2006; 117(5): e909 - e923. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2004 American Academy of Pediatrics. All rights reserved. |
||
|
||
