Cortisol levels of young children in full-day childcare centers: relations with age and temperament

Abstract

Cortisol levels of 70 children, aged 39–106 months, were sampled at home and at their full-day childcare centers at two times of day, mid-morning and mid-afternoon. Parents and teachers completed questionnaires assessing child temperament (negative affectivity, surgency or extroversion, and effortful control) and aggressive behavior. The results replicated a previous study showing increases in cortisol levels over the day at childcare for preschool-aged children, while home levels followed the expected circadian decrease in cortisol from morning to afternoon for most children regardless of age. At childcare, 3- and 4-year olds were more likely to show elevations in cortisol by mid-afternoon than were older children. Controlling statistically for age, shyness for boys, and poor self-control and aggression for both sexes were associated with increases in cortisol over the day at childcare. The results suggest that younger children and those with more immature social skills may frequently experience elevations in cortisol as the day progresses in group care contexts.

Introduction

Influenced by a social and political interest, research on the impact of non-parental care during early childhood has intensified in the last decade (NICHD, 1996, NICHD, 1997, Lamb, 1997, Scarr, 1997). Much of this work has focused on behavioral issues such as the relations between the timing, quality, and quantity of non-parental childcare and parent–child attachment relationships and children’s cognitive and social development. Despite evidence that group care contexts are associated with increased respiratory illness rates in young children (Collet et al., 1991), only scant attention has been paid to the physiological responses of children in childcare that might influence immune competence (Legendre and Korintus, 1996, Gunnar et al., 1997, Tout et al., 1998).

Although most studies of behavioral outcomes have revealed largely neutral or positive effects of early child care experiences (NICHD, 1996, NICHD, 1997, Lamb, 1997, Scarr, 1997) group care may be challenging for young children. The length of the day, the need to sustain interactions with many other children, and the need to organize and reorganize security seeking behaviors around multiple adults are examples of the challenges facing young children in these settings. While these challenges may stimulate the development of social skills, they may also tax emotional resources and coping competencies and result in intermittent stimulation of stress-sensitive physiological systems, including the hypothalamic-pituitary-adrenocortical (HPA) system.

The HPA system is sensitive to physical and psychological stressors. Increased activity of the HPA axis during the states of threatened homeostasis is important for mobilizing energy stores, influencing the activity of the immune system, and coordinating behavioral responses to threat (Johnson et al., 1992). Activity of the HPA axis can be estimated using measures of its end products, glucocorticoids. The primary glucocorticoid in humans is cortisol and it can be assessed in small samples of free-flowing saliva. Salivary and plasma estimates of cortisol are highly correlated (Kirschbaum and Hellhammer, 1994). The activity of the HPA axis is of special interest because of its close links to the amygdala and hippocampus, which are both limbic system structures involved in emotions and memory (de Kloet, 1991, Jacobson and Sapolsky, 1991, Eichbaum et al., 1992, McEwen, 1994, Nelson and Carver, 1998). In addition, elevations in glucocorticoids, if frequent or sustained, may interfere with optimal functioning of the immune system (Coe, 1993, Coe, 1994, Black, 1994).

Studies examining salivary cortisol in children during periods of group care have shown that morning levels are either similar to or slightly lower than levels obtained when children are at home (Lundberg et al., 1993, Gunnar et al., 1997). In contrast to concerns that group care might be stressful for young children, these data suggest the opposite. It was surprising, therefore, to find that cortisol levels for 81% of children rose from morning to afternoon in a study of preschoolers in full-day, center-based care (Tout et al., 1998). This was reverse of the expected circadian decrease in cortisol from morning to afternoon. Typically, cortisol levels are high early in morning and decrease over the day in children, as in adults (Shimada et al., 1995, Knutsson et al., 1997). This daily pattern is established in infancy (Price et al., 1983, Stansbury and Gunnar, 1994, Mantagos et al., 1998). Levels consistent with a decrease over the day have been observed in preschoolers attending half-day, nursery school programs (Gunnar et al., 1997). Like the nursery school study, the study by Tout and her colleagues involved multiple days (approximately 20) of cortisol sampling. The rise in cortisol over the day found by Tout and colleagues was, thus, based on information averaged over multiple days of sampling morning and afternoon levels in the same children. Unlike the nursery school study, however, the study by Tout and colleagues (1998), did not involve home cortisol levels as comparison. Thus, it was unknown whether the cortisol increase over the day reflected something peculiar to the childcare context or a pattern particular to these children that would also be observed if they were assessed at home. In addition, because Tout and colleagues did not have home levels of cortisol, they could not tell whether the cortisol increase reflected suppressed morning levels returning to levels similar to home levels by afternoon, or whether the results reflected a stress response of the HPA axis with afternoon levels that exceeded home baseline levels.

Tout et al. (1998) did find that temperament and social behavior were associated with the rise in cortisol over the day. Specifically, for boys in particular, increases in cortisol from morning to afternoon were associated with teacher reports of shy/anxious behavior. Other studies have also shown relations between behavior and cortisol levels for children in group care settings. In addition to shy/anxious behavior, poor attention regulation and poor self- or inhibitory control competence has been shown to correspond to higher cortisol levels (Gunnar et al., 1997). Similarly, angry/aggressive behavior has been found to relate to frequent ‘spiking’ or elevations of corticoids (Montagner et al., 1978, Gunnar et al., 1997). Thus, negative emotionality and/or aggressive, poorly regulated behavior may increase the likelihood of cortisol elevations among children coping with the complex social nature of group childcare.

In the study of Tout et al. (1998), it was also unclear whether this rising pattern of cortisol in childcare would persist among school-aged children, or was peculiar to the preschool years. However, based on evidence that poor self-control and low social skills are related to increased levels of cortisol for children in group care settings, we suspected that younger children with their generally less-developed social skills would be more likely than older children to show this pattern. Thus, the present study was designed to address the following questions: (a) Could we replicate the rise in cortisol over the day in full-day, center-based childcare found by Tout et al. (1998)? (b) Would data on home levels of cortisol show that afternoon levels at childcare were elevated over home baselines, or not? (c) Would the rise in cortisol be observed among older, school-aged children, or was it particular to the preschool-aged child? (d) Would children whose teachers or parents described them as shy/anxious or aggressive/poorly controlled show larger increases in cortisol as the day progressed in childcare? (e) And would younger children be described more in these terms than older children, suggesting that immature social skills were involved in creating conditions that elevated cortisol in full-day childcare?