Estimating circadian rhythm across infancy: Development of intra- and interdaily stability, and relative amplitude

Previous studies have emphasized the connection of sleep quality, length, and rhythmicity during infancy with brain maturation. There is, however, a lack of consensus regarding variables and methodology in relation to the diverse age periods. We thus conducted a longitudinal study with 136 healthy infants at ages 3, 6, and 12 months, aiming to capture maturation of the sleep-wake rhythm. We estimated circadian rhythmicity derived from actimetric data (Intradaily Variability, Interdaily Stability, Relative Amplitude and Circadian Function Index). Results confirm significant effects across ages. Intradaily Variability decreased with age, reflecting continuously less fragmented sleep and sleep-wake transitions. Interdaily Stability and Relative Amplitude increased with age, reflecting the gradual approaching to synchronization with the day-night cycle. More globally, the Circadian Function Index, averaging those three variables to indicate the maturation of the sleep-wake rhythm, illustrated that infants’ circadian status improves significantly from 3 to 12 months. These findings highlight the first year of life as a period of great changes in sleep-wake patterns and offer new normative data regarding estimates of circadian rhythm in infancy. Such maturation curves are fundamental to detect atypical development of sleep rhythm. Furthermore, early detection allows intervention to improve negative development outcomes.


INTRODUCTION
Infants' sleep can be a source of sizable concern for parents, to such an extent that poor children's sleep has been reported to be the best predictor of parents' dissatisfaction (Sepa et al., 2004). Its variability and unpredictability can cause considerable struggle for parents and many believe that their infant's sleep is abnormal (Cook et al., 2020). Researchers have thus tried to grasp sleep characteristics to evaluate their impact on neural development. Two main sleep stages exist. The first one is characterized by rapid eye movements (REM sleep) and is usually associated with high-frequency brain activity. The second one does not contain rapid eye movements (non-REM sleep) and Article Cortica 2022. 1(2), 445-463 www.revue-cortica.net | ISSN : 2813445-463 www.revue-cortica.net | ISSN : -1940445-463 www.revue-cortica.net | ISSN : https://doi.org/10.26034/cortica.2022 Ó MARET, D. 2022 447 corresponds to a quieter state of rest and lower-frequency brain activity (Arditi-Babchuk et al., 2009). In infants, the sleep states are not as clearly distinguished, and two similar states have been outlined. Quiet sleep is comparable to non-REM sleep, while active sleep is described as an immature form of REM sleep, with a higher intensity and more phasic neuronal activity (Mirmiran et al., 2003).
Whereas adults sleep is mainly composed of non-REM sleep, term-born infants spend more than 50% of their sleep in active sleep (Heraghty et al., 2008;Tarullo et al., 2011).
However, by 6 months old, the proportion of active sleep is already reduced to 25%, approaching adult-like levels, while the proportion of quiet sleep increases (Arditi-Babchuk et al., 2009;Louis et al., 1997;Mirmiran et al., 2003).
Simultaneously, the human brain develops rapidly prenatally and during the first 3 months of life (Heraghty et al., 2008). Increased active sleep has thus been hypothesized to crucially support the development of learning, memory integration and consolidation that the infants experience across their first months and years (Bathory & Tomopoulos, 2017;Tarullo et al., 2011). By building a fundament of the normative patterns of sleep rhythm development, we can pinpoint perturbations in these maturational dynamics, and likely alleviate, with timely action, secondary problems of impaired daytime functioning, behavioral outcomes, central nervous system disorders or even mental diseases (Byars et al., 2012;Shimada et al., 1999;Tham et al., 2017).
Right after birth, newborns' sleep broadly lacks a sense of day-night periodicity. Only around 10-12 weeks a diurnal sleep rhythm starts to evolve, with a decrease in daytime sleep and gradual increase in nighttime sleep (Iwata et al., 2017). Slowly, the infants' sleepwake rhythm blends with the regular light-dark 24h cycle, supported by environmental cues, such as light, maternal feeding etc. (Jenni et al., 2006). The circadian rhythm is an endogenous rhythm close to a 24h cycle. This type of rhythmicity is not only intertwined with behavioral sleep-wake patterns, but also reflected in multiple processes of human physiology, including body temperature, blood pressure, hormone secretion, or digestive secretions (Minors & Waterhouse, 2013). It is mainly regulated by the suprachiasmatic nuclei (SCN) of the hypothalamus, which acts as a biological clock that also receives contextual information from environmental elements such as light. The circadian rhythm is thus largely independent of sleep pressure. Cortica 2022. 1(2), 445-463 www.revue-cortica.net | ISSN : 2813-1940https://doi.org/10.26034/cortica.2022 This supports sleeping at night, where daylight is lacking, and vice versa, staying awake during daytime (Borbély, 1982;Rivkees, 2003 Limitations of actimetry are a general lack of standardization among researchers in methodology, data processing and analysis (Berger et al., 2008), which can yet be addressed with precise methodology reporting (Schoch et al., 2021 (Sadeh, 2011), i.e. daily logs (Tikotzky & Sadeh, 2001). In this study, we will thus combine actimetry, an objective measure, with sleep diaries, a subjective measure.  In summary, the emergence of the circadian rhythm happens between 3 weeks and 6 months old, can depend on factors such as feeding mode, and the computational approach applied.

Research question
As infants age, their sleep becomes less fragmented but shorter, and matures into a circadian rhythm (Glickman, 2010;Iwata et al., 2017;Jenni et al., 2006). This can be shown We first hypothesized that the fragmentation of the rhythm decreases with age, and that IV gradually decreases from age 3 to 12 months.
Secondly, we investigated the stability of the In addition, they were also sent online questionnaires about the family background, general health, and demographics. As a compensation, they received small gifts (bodysuits, hats, etc).

RESULTS
Prior to the analysis, Kolmogorov-Smirnov normality tests were performed, to test the   Table 2). In other words, this data confirms the expected age-related decrease in IV, reflecting a decrease in rhythm fragmentation.

Figure 2
Circadian Function Index scores according to the infants' age To summarize, results confirm a general tendency across all variables towards a maturation of the sleep-wake rhythm, which is reflected by the IV decreasing with age and IS, RA and CFI increasing between 3 and 12 months old. Cortica 2022. 1(2), 445-463 www.revue-cortica.net | ISSN : 2813-1940https://doi.org/10.26034/cortica.2022  and light conditions need to be controlled for (Jenni et al., 2006). Relatedly, prior findings revealed that breastmilk might be a stronger zeitgeber than formula milk (Kikuchi et al., 2020), which in turn could impact infants sleep-wake rhythmicity. In our study, we only included infants who received breastmilk for more than 50% of their food intake at 3 months, the values we presented must thus be taken with caution, as they might vary depending on infants' food intake.  (Byars et al., 2012;Cook et al., 2020;Shimada et al., 1999;Tham et al., 2017). Normative data about the sleep-wake rhythm for this age period are essential for studying how sleep variables relate to neurological development.

Article
Circadian rhythm norms allow to identify evolution of infants' sleep-wake patterns in health and disease and support practitioners in detecting and diagnosing potential early risk factors before they affect the infants' development.