Light/Dark Cycle Promotes Weight Gain in Preterm Infants

Introduction Worldwide, Neonatal Intensive Care Units (NICU) are often maintained under 24-hour constant bright light (CBL) conditions, without significant variations in light intensity. As such, infants under these conditions are exposed to chaotic temporal signals, as opposed to predictable temporal signals, without cyclic variations in terms of sounds and light. Interestingly, the level of light to which premature infants are exposed represents an accessible environmental variable that can be easily controlled. Premature infants are generally maintained under CBL conditions for prolonged intervals, thus hampering their growth and development. CBL conditions can generate harmful effects to newborn eyes, and experimental studies have shown that newborn rats exposed to a CBL environment from the first day of birth develop damage in the cellular organization of the retina.

Previous studies have shown that a 12:12 hour light/darkness cycle (LDC) can promote an increase in infant sleep time, decrease feeding time and improve weight gain, compared with infants maintained in constant light or constant dark conditions. Further, evidence shows that infants exposed to a LDC since birth acquire circadian rhythmicity of activity/rest patterns as early as week 34 of life.

This study presents the results from a multicentric trial which sought to determine the benefits of implementing a LDC in the NICU of three different hospitals in Mexico. The primary endpoint was weight gain; secondary endpoints included early hospital discharge, salivary melatonin levels and the effect of reducing the intensity of nocturnal light in development of premature infants.

Methods Study Design and Participants This prospective, open-label, randomized multicentric clinical trial recruited patients in three different public-setting hospitals in Mexico between 2016 and 2020. Eligible patients were premature infants (gestational age <37 weeks) who were hospitalized in the low risk and high-risk neonatal units of participating institution, with a non-severe diagnosis for hospitalization, without concomitant illness, and classified as stable. Exclusion criteria included infants hospitalized with severe illness, congenital malformations, or important neurological diseases, while elimination criteria included infants initially classified as having a non-severe illness who progress to severe illness, infants who received intensive treatment for over a week due to medical complications (i.e. bacterial infections), as well as infants whose parents requested withdrawal from participating in the study. The parents or legal guardians of all included patients provided written informed consent to participate in this study. Participating institutions included Centro Medico Nacional "La Raza", Mexico City; Hospital General de Zona No.1 "Dr. Demetrio Mayoral Pardo" Oaxaca City, Oaxaca, Mexico (both hospitals from the public healthcare system "Mexican Institute for Social Security IMSS"), and General Hospital "Dr. Aurelio Valdivieso", Oaxaca City, Oaxaca, Mexico (from the public healthcare system of the Health Secretariat [SSA]). The study was approved by the scientific and bioethical committees of each participating institution (CMNR-No. 3504-35, HGZ 1 IMSS Oaxaca R-2001-2015-8, SSO-R. 023/2018 y UNAM-094/2014). The study was performed in accord with the declaration of Helsinki and the principles of good clinical practice. This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guidelines.

Sample Size The sample size was calculated for a two-sample comparison, given an 80% sensitivity and a 60% specificity and with a confidence interval of 90%. A total of 150 patients per group was estimated, for a total sample size of 300 subjects.

Procedures All study procedures were performed following a predefined study protocol and a uniform sample processing plan. Data was registered using a unified registry form in all participating sites, and all analyses were centrally performed at the main investigation site. All included patients provided information regarding family and perinatal history.

Each NICU from the participating institutions was evaluated in order to measure the illumination level using a luxmeter (HER-410-Steren), in 4 specific areas during 6 different times of day to ascertain homogeneous lighting conditions throughout the centers. Subjects randomized to the experimental group were allocated to alternating light/darkness conditions as follows: from 07:00 to 19:00 hours the subjects were kept under normal room light conditions; from 19:00 to 07:00 of the following day the conditions were modified by placing the patient under an acrylic cephalic helmet (length: 27 cm; width: 27 cm; height: 17.5 cm; opening: 17x12 cm). The helmet was covered with surgical cloth (green or blue) folded to 50x60cm rectangles, leaving the frontal part open in order to maintain an adequate air flow. This intervention exposed infants in the experimental group to light at 25 lux for 12 hours every day, while during daytime the cloth was removed in order for study subjects to be exposed to regular room lighting. The control group was kept under normal room light conditions (CBL) 24 hours a day (level of illumination was 275.82±14 lux during the day and 145.28±14 lux at night). All included subjects were under 24-hour surveillance for the period spanning the length of their hospital stay by the NICU personnel at each participating institution. Infants were kept in individual incubators regulated to maintain a body temperature ranging from 36.5 °C -37 °C.

Both study groups were offered feeding every 3 hours (as per NICU guidelines and vital signs were recorded every 3 hours (cardiac and respiratory rate, temperature). Every infant underwent an 8-12 hour fast after birth for intestinal resting, thereafter enteral feeding began at a dose of 12.5 ml/kg of breast milk, which was increased every 12 hours until reaching a maximum of 200 ml/kg/day. Every included infant was orally fed, in case of abdominal distension or absence of peristalsis underwent a 24-hour maximum fast. Once the infants reached 1600 g body weight, breast feeding was indicated in order to prepare for discharge. The total caloric content among different hospitals was not standardized in this present study.

Among 20 randomly selected infants (10 randomized to BCL and 10 randomized to LDC) saliva samples were collected two times a day (08:00, 23:00) every day during the first 20 days of NICU stay in order to determine biomarkers. The salivary concentration of melatonin was determined using a standardized ELISA kit, following instructions by the manufacturer (IBL International GmbH Flughafenstr. Hamburg, Germany).

Other procedures performed included capillary blood glucose, administration of medication as prescribed by their primary care physician, and sample acquisition for laboratory testing pertaining each condition presented by infants included in this trial. All included patients were monitored four weeks post hospital discharge to evaluate development and growth at home, as well as their circadian activity/rest patterns.

Healthy premature infants were considered for discharge when participants presented an adequate growth curve and met established growth criteria, including the ability to maintain a stable body temperature in an open crib, ability to perform all feedings without respiratory distress, and a constant weight gain >2000g.

Statistical Analysis Continuous variables were summarized as arithmetic means with standard deviations or medians and ranges according to the data distribution assessed with the Kolmogorov-Smirnov test. For descriptive purposes, categorical variables were summarized as frequencies and percentages. Data pertaining to body weight gain, feeding and melatonin concentration were classified per study group and evaluated as repeated measures according to post-natal days. Comparisons among study groups were performed using a two-way ANOVA, followed by a TUKEY post-test analysis. A p-value <0.5 was considered statistically significant. The initial body weight, gestational week, and length of hospital stay were compared using a one-way ANOVA, a p-value <0.5 was considered statistically significant. All analyses were performed using the STATISTICA program (V10; StatSoft, Inc. 1993), while figures were constructed using the program Sigma Plot (V14; Systat Software, San Jose, CA).