Source: Unknown

Source: Unknown

The issue of daycare is probably the only other topic to rival infant feeding as one of the most contentious in parenting discourse today.  The idea that daycare may be “bad” for a child has many parents seeing red, despite much research that decries the quantity and, most importantly, the quality of daycare being used in many (but not all) developed nations today (e.g., [1][2][3][4][5]).  One of the means we have to look at how children are adapting to daycare is to examine their cortisol levels during the transition to daycare, after the transition, and at home.  Cortisol can tell us about the degree of stress that is caused by the separation to the primary caregiver and for how long this type of stress continues.

What we seem to know so far is the following (with a caveat that although there is a fair bit of research that converges in this area, it is far from conclusive due in part to the many variations of care that exist):

  • Cortisol has been consistently found to increase over the course of the day in daycare (versus when a child is at home with a primary caregiver) (e.g., [6][7]) which raises a flag as to the effects on the developing brain (for a discussion of hyporesponsivity and why this increase matters, see here). Of note here is that the normal diurnal pattern is for cortisol to be highest in the morning and to gradually decrease as the day progresses so it is not just that this is muted in daycare, but we are often seeing the total opposite.
  • The rises in cortisol that have been shown to exist have been found to be at least partially dependent upon the quality of care provided. Specifically, the magnitude of increase is much higher in lower-quality daycare settings[8][9] and high-quality, home-based child care seems to be associated with little to no increase in cortisol levels[10].  This provides some evidence that the rises seen are indeed reflective of stress in the daycare environment and fit with the behavioural research cited above that shows significant differences in externalizing problems based on quality of care.
  • These rises are not necessarily temporary. In one study looking at young toddlers transition to daycare at 15 months, the increase that occurred during transition was still visible 15 months later[7].  This shouldn’t be too surprising as other research has not specifically looked at transition, but rather the cortisol levels of children who have been in daycare for a spell already.
  • Child temperament influences the degree of increase in cortisol with higher needs children showing greater increases even in higher quality daycare settings[10]. This is likely not too surprising given what we know of the need for what some researchers have called “optimum” care for these children (with “optimum” meaning very attachment-based parenting) (see here for a brief discussion).
  • Age influences this cortical effect, with younger children (3 and under) showing greater effects than older children[6][11], though what remains unclear is if this is due to developmental differences or the amount of time spend in daycare. That is, do older children show this effect because they are used to daycare already or is there something that occurs during development that leads to this change.

Enter some new research aimed at examining cortisol effects during a 10-week transition to a new daycare setting across a variety of ages to see if differences in cortisol levels in daycare by age previously have been due to development of the child or the time spent in daycare[12].  A total of 168 children in the United States aged 1.2 months to 8 years of age were included with a mean of 3.27 years of age, though 34 children were over age 5 and thus in school for part of the time and thus secondary analyses were done with these children removed.  Quality of daycare was not assessed, though the daycare was a center (which is often associated with lower-quality care), but was a university-based daycare (which may result in higher-quality care).  Cortisol samples were collected six days over the study time period of 10 weeks (day 1, weeks 2, 4, 6, 8, and 10), twice per day (morning and afternoon) and were collected during times that activities would not influence levels (e.g., eating or sleeping or playing outside).  For 129 children, parents collected saliva samples at home for 2 consecutive days prior to them entering the daycare.  Not surprisingly, there were many missing data points for cortisol levels given that a child who was away on any of the six time periods would have missing data; however, missing data analysis revealed no associations of missing data with most key variables with the exception of minority status in which the home morning sample was more likely to be missing (age was also a factor in that older children were more likely to be missing the center morning value but this is not surprising as older children were in school and thus not at the center for morning collection).

What was found?  In terms of the question of how does the transition to daycare influence the change in cortisol levels, in line with myriad other research, the researchers found that being in daycare resulted in an increase in cortisol levels compared to being at home.  Notably, though, over the 10-week period, the degree of increase also increased.  That is, the increase in cortisol from mid-morning to afternoon actually got worse over the 10-week transition to daycare.  Further analyses showed that this was due to a consistent decline in mid-morning cortisol with a stable-high afternoon cortisol level.  Finally, the effects remained the same for the infancy and toddler group when the school-age children were removed from the analysis (as previously mentioned they were removed as they often were missing morning data due to not actually being at the centre during those times).

There were some nuances though that are worth looking into.  In terms of the increase in change over time, this was not driven by a small group of children, but rather a higher percentage of children showed an increase in cortisol throughout the day as the weeks went by as opposed to some form of attenuation.  For example, in the infancy group, 39% showed a rise on day 1, 30% on the week 2 measurement, 38% on week 4, 54% on week 6, 60% on week 7, and 47% on week 8.  Though this may seem sporadic, there is a significant general linear trend and this is mirrored in all three age groups.

The question of interest, however, was how age influenced findings.  In this regard, there was a quadratic effect of age with the greatest increase in cortisol occurring in the preschool years (over infancy or school-age years).  There were also differences in the types of effects.  At home in infancy, the “normal” pattern was virtually no change in cortisol from mid-morning to afternoon (not surprising given that this diurnal pattern of a decline can take months to even remotely establish and further develops for years) whereas in daycare there was an increase from mid-morning to afternoon.  At home in the preschool years, there was a moderate decline from mid-morning to afternoon, but at daycare there was an increase (i.e., there were totally opposite patterns).  In school-age children, at home there was a steep decline from mid-morning to afternoon, but a relatively stable level in daycare.  Thus the daycare effect was similar for infants and toddlers, though the home effect differed, but the daycare effect differed for older children.

There are a few considerations that deserve discussion based on the pattern of findings.  First, it is worth noting that the failure to show home-like patterns at the end of 10 weeks does not support the idea that children fully adapt to the daycare environment.  Rather, there are continued struggles relative to a home environment.

Second, this finding of a failure to adapt coupled with the largest increase being in the preschool years suggests a large portion of this may be driven by peer interactions.  That is, toddlers and preschoolers are the group most likely to struggle with peer interactions and what is expected of them in this regard in a daycare environment (think sharing), causing the most difficulties over this time period (and based on other research, continuing for months[7]).  Infants don’t really engage with peers and school-age children have had far more experience in this realm and know what is expected of them, although infants do show the same rise in cortisol, suggesting they also experience stress, it is just not as magnified as with the toddler group.

Third, how should we interpret the drop in mid-morning cortisol levels?  There are three possibilities raised by the researchers:

  1. The results may suggest there may be some adaptation to the parental separation. The results found that there were decreases to the mid-morning levels of cortisol which could suggest that the early weeks were representative of stress upon leaving the parent which does seem to attenuate as the weeks progress.
  2. In line with other research[11][13], is that children (even infants) “anticipate” daycare days and have higher HPA axis activity overnight, resulting in periods of hypoactivation the following morning. This means that the anticipation of daycare causes stress and by the next morning, their brains have gone into a period of hypoactivation to compensate for the higher levels of adrenocortical activity overnight.  (Total aside: What’s interesting about this is that this idea of cortical “anticipation” might negate one of the criticisms of the research looking at cortisol levels during extinction sleep training[14].  Namely, the infant in that study showed high levels of cortisol just prior to the onset of the bedtime ritual on all days tested, even though mom was present with the child, and many speculated that this stress was due to the new environment, despite most research suggesting new environments with mom don’t result in this change.  My speculation had been the anticipatory nature of bedtime that had been problematic for the families leading them to the sleep clinic in the first place.  Anyway, just thought I’d share that tidbit here though it has no bearing on daycare.)
  3. The third possibility is that the sleep changes that often go along with transitions to daycares result in a shifting of the cortisol levels at the midmorning assessment if they have been waking earlier the longer they are in daycare. This presupposes that the patterns of sleep-wake times shifted significantly over the 10 weeks, something that was not measured in this particular study, but remains a possibility.

Most importantly, how does this fit with other research on daycare?  It seems to fit quite nicely in with what we already know: Namely, that daycare results in changes to cortisol levels, specifically changes in line with possible stress responses.  This adds to the body of research by demonstrating that these changes occur during transition regardless of age, though the magnitude and effect differ by age.  Perhaps most importantly, this study failed to assess quality of care which raises questions about the applicability to all daycares.

This is critical as other new research on daycare that was published recently was used to debunk the idea of any long-term externalizing problems[15].  However, this research is difficult to extrapolate or compare to the bulk of behavioural research cited earlier for one big reason: The daycare in question was in Norway, not the United States where the bulk of the behavioural research has been done.  Scandinavian countries are renowned for having excellent, high-quality daycare for all children, something that is lacking in the USA and elsewhere.  In line with this, many of the effects seen in daycare are eliminated or at least minimized when the daycare is deemed “high-quality” (and why I believe we need to push for high-quality daycare on a larger scale for families who need it), as mentioned previously.  This is important if we view the cortisol levels as being a flag for possible stress that would impact long-term development, particularly social development if the stress is, indeed, peer-related.

Interestingly, even in the Norwegian study, there were differences in levels of aggression based on time spent in daycare at 2 years of age, but rather these effects had faded by 4 years of age.  This could suggest that daycare is stressful for children, but that higher-quality daycare helps children adapt and cope with this stress in a manner that is conducive to longer-term well-being.  In lower-quality care, the stress is left to build up with little support from caregivers, leading to a higher risk of longer-term problems based on time spent in this lower-quality care and child temperament.

As I nearly always write when it comes to daycare, the take-home message here is that we need to fight for high-quality care and for families to have the options of staying home during these critical developmental windows.  Though I acknowledge not all families will want to or be able to take advantage of being at home, it should be an option for all.  (This is particularly important for parents of higher-needs children who seem to be most affected by their environment.)  For those that utilize daycare, being able to find and use high-quality daycare is essential for the well-being of our children.

_____________________________

[1] Loeb S, Bridges M, Bassoka D, Fuller B, Rumberger RW.  How much is too much?  The influence of preschhol centers on children’s social and cognitive development.  Economics of Education Review 2007; 26: 52-66.

[2] Magnuson K, Meyers M, Ruhm C, Waldfogel J.  Inequality in preschool education and school readiness.  American Educational Research Journal 2004; 41: 115-157.

[3] Vandell DL, Belsky J, Burchinal M, Steinberg L, Vandergrift N, NICHD ECCRN.  Do effects of early child care extend to age 15 years?  Results from the NICHD study of early child care and youth development.  Child Development 2010; 81: 737-756.

[4] Belsky J, Vandell DL, Burchinal M, Clarke-Stewart KA, McCartney K, Owen MT, NICHD ECCRN.  Are there long-term effects of early child care?  Child Development 2007; 78: 681-701.

[5] Pluess M, Belskey J.  Differential susceptibility to rearing experience: the case of childcare.  Journal of Child Psychology and Psychiatry 2009; 50: 396-404.

[6] Dettling AC, Gunnar MR, Donzella B.  Cortisol levels of young children in full-day childcare centers: relations with age and temperament.  Psychoneuroendocrinology 1999; 24: 519-36.

[7] Ahnert L, Gunnar MR, Lamb ME, Barthel M.  Transition to child care: Associations with infant-mother attachment, infant negative emotion, and cortisol elevations.  Child Development 2004; 75: 639-50.

[8] Legendre A.  Environmental features influencing toddlers’ bioemotional reactions in day care centers.  Environment and Behavior 2003; 35: 523-49.

[9] Lisonbee JA, Mize J, Payne AL, Granger DA.  Children’s cortisol and the quality of the teacher-child relationships in child care.  Child Development 2008; 79: 1818-32.

[10] Dettling AC, Parker SW, Lane S, Sebanc A, Gunnar MR.  Quality of care and temperament determine changes in cortisol concentrations over the day for young children in childcare.  Psychoneuroendocrinology 2000; 25: 819-36.

[11] Ouellet-Morin L, Tremblay RE, Boivin M, Meaney M, Kramer M, Cote SM.  Diurnal cortisol secretion at home and in child care: A prospective study of 2-year-old toddlers.  Journal of Child Psychology and Psychiatry 2010; 51: 295-303.

[12] Bernard K, Peloso E, Laurenceau JP, Zhang Z, Dozier M.  Examining change in cortisol patterns during the 10-week transition to a new child-care setting.  Child Development 2015; 86: 456-71.

[13] Gunnar MR, Tout K, de Haan M, Pierce S, Stansbury K.  Temperament, social competence, and adrenocortical activity in preschoolers.  Developmental Psychobiology 1997; 31: 65-85.

[14] Middlemiss W, Granger DA, Goldberg WA, Nathans L.  Asynchrony of mother-child hypothalamic-pituitary-adrenal axis activity following extinction of infant crying responses induced during the transition to sleep.  Early Human Development 2012; 88: 227-32.

[15] Dearing E, Zachrisson HD, Naerde A.  Age of entry into early childhood education and care as a predictor of aggression: faint and fading associations for young Norwegian children.  Psychological Science 2015; DOI: 10.1177/0956797615595011.