Salivary cytokine cluster moderates the association between caregivers perceived stress and emotional functioning in youth
Introduction
Exposure to stress in childhood affects emotional and behavioral functioning and the risk for childhood and adult-onset psychiatric disorders, including major depression, anxiety, schizophrenia spectrum disorders and post-traumatic stress disorder (Green et al., 2010, Nemeroff, 2016, Krugers et al., 2017, Heim et al., 2008, Alisic et al., 2014, Varese et al., 2012, Faravelli et al., 2017). However, there is considerable inter-individual variability in how individuals respond to early life stress. While some individuals are resilient to early life stress, others show more susceptibility, developing stress-related health and psychiatric disorders (Werner, 1992, Werner, 1996, Beutel et al., 2017, Ebner and Singewald, 2017). Since individuals demonstrate differential sensitivity to the effects of early life stress, finding methods to identify these at-risk children early on is imperative to avail them of treatment and intervention options.
While it is well known that early life stress associates with numerous health and mental health problems, the role that objective versus perceived stress has on this link is still being elucidated (Pervanidou and Chrousos, 2018). Both objective measures of stress such as the frequency of adverse events and the subjective perception of stress are associated with biological markers of allostatic load (Evans, 2003, Upchurch et al., 2015). Allostatic load measures the physiological dysregulation in different biological systems including the hypothalamic–pituitary−adrenal (HPA) axis, the immune system and metabolism that are associated with the adaptation to chronic stress. It is this biological “wear and tear” that is theorized to form the basis for stress-related health and mental health disorders (Beckie, 2012). Perceived stress levels show a greater association with allostatic load biomarkers than objective stress measures, especially with regards to neuroendocrine activity (Clark et al., 2007, Hawkley et al., 2011, Glei et al., 2007) though an exception exists (Christensen et al., 2019). Perceived stress has also been shown to predict physical health better than questionnaires that measure the frequency of adverse life events (Cohen et al., 1983).
Not all individuals will respond the same way to the presentation of adverse or traumatic events in their lives. Some individuals who perceive adverse events as more stressful may be at greater risk for adverse consequences than individuals that perceive the same experiences as less stressful or impactful and thus the subjective appraisal of objective events is key in determining how individuals will respond to these events (Lazarus and Stress, 1984). Part of the link between perceived stress levels and stress-related health and mental health outcomes could potentially be mediated through personality factors. When controlling for the Big Five personality traits though, the association between perceived stress and allostatic load was shown to still have a weakly modest but significant association (Christensen et al., 2019). Further, measures of resilience and emotional stability are negatively correlated with perceived stress levels (Lauridsen et al., 2017, Chandelkar and Shetty, 2019). Meanwhile, higher perceived stress scores are found in patients with mental health problems including psychotic-like experiences and depression compared to healthy controls (Turley et al., 2019, Klein et al., 2016). Higher perceived stress scores are also found in individuals that experienced early life stress compared to control subjects and adult perceived stress scores are positively correlated with scores on the Child Trauma Questionnaire (Han et al., 2016). Perceived stress may then be a more powerful indicator of distress in an individual than objective measures of adverse and traumatic events (Cohen et al., 1983).
Chronic stress increases the risk for psychopathology and the mechanism for this association could be due in part to the influence of chronic stress on glucocorticoid receptor sensitivity and receptor numbers. Prolonged exposure to cortisol in the face of prolonged HPA axis activation can downregulate glucocorticoid receptor numbers and sensitivity in different tissues including on immune cells such as monocytes and macrophages. Glucocorticoid receptor desensitization associates with an enhanced pro-inflammatory cytokine profile in immune cells leading to a state of low-grade inflammation associated with different disease states including some mental health disorders. A strong association has been found between early life stress and systemic inflammation later in life (Danese et al., 2007).
Interestingly, parent perceived stress has been shown to predict the inflammatory profile of their children with and without asthma over a six-month period. Further, the association between the parents perceived stress and the child’s inflammatory profile was not explained by the child’s own psychological state. The parent’s perceived stress accounted for 7.2 to 7.6% of the variance in the child’s inflammatory profile. Further, the association between the parent’s perceived stress and the child’s inflammatory profile was greater than between the parent’s depression scores and the child’s inflammatory profile. Children may be more responsive to the perceived stress of their parents than by their parents’ mood and the child’s experience of the parents perceived stress appears to get under the child’s skin becoming biologically embedded (Wolf et al., 2008). Parents perceived stress levels have also been shown to associate with other aspects of their children’s biology. For example, there is a positive correlation between parents perceived stress levels and children’s body mass index in families with low income (Baskind et al., 2019).
Recently, salivary cytokines have been considered as potential biomarkers that could be used to predict the risk for the development of psychological dysfunction in response to stress (Riis et al., 2016). Previous studies show that cytokines can be clustered based on their profile levels in blood and that these clusters are associated with clinical phenotypes including those related to psychiatric illnesses such as major depression, schizophrenia, and schizoaffective disorder (Lu et al., 2013, Fillman et al., 2016, Boerrigter et al., 2017). Adult patients with depression have higher peripheral cytokine levels, especially those patients that experienced early life stress. Clustering cytokines by their profile levels in peripheral plasma was able to differentiate depressed patients with a history of early life stress from healthy controls (Lu et al., 2013).
Given the findings in adults, it is likely that cytokines might reflect the effects of early life stress on health outcomes in children and might serve to identify those most affected by early life stress (Hantsoo et al., 2019). Since the effect sizes for the association between individual cytokines and early life stress in childhood are quite small though (Kuhlman et al., 2019, Cohen et al., 2019), a statistical technique that clusters children together based on similar pro-inflammatory profiles may increase the signal to noise ratio and identify children most at risk of psychopathology.
We measured cytokines with pro-inflammatory activity including IL-6, IL-1β, TNF-α and IL-8 in children’s saliva in the UCSF and GUSTO cohorts, as these salivary cytokines have been shown to reliably increase in the presence of a social-cognitive psychological stressor (reviewed in Slavish et al., 2015, Shields et al., 2016). Collection of saliva is a less intrusive and costly sampling approach for large intervention studies based in community settings and increases the likelihood of subject consent. Despite a lack of clear overlap between all cytokine levels between different compartments like blood and saliva, some cytokines do correlate well between serum and blood (Byrne et al., 2013) and salivary cytokines are known to be responsive to different types of stressors (reviewed in Slavish et al., 2015). Several cytokines show higher levels in saliva than in blood (Byrne et al., 2013). Salivary cytokines also correlate with regional brain activation (O'Connor et al., 2009a) and cytokines produced from oral inflammation can lead to fever indicating a link between oral inflammation and brain activation (Navarro et al., 2006). Oral inflammation due to periodontal disease is sometimes a confound in the measurement of salivary cytokine levels but this confound is minimized in children. Salivary cytokines may be potentially useful biomarkers to assess the effects of early life stress in laboratory and naturalistic settings (reviewed in Slavish et al., 2015).
The perceived stress scale (PSS) was used to measure the degree to which the children’s caregivers in the UCSF and GUSTO cohorts found their lives to be uncontrollable, unpredictable and overloaded over the last 8 weeks (Cohen et al., 1983). The PSS is considered a useful tool to determine the role of perceived stress levels in the development of health and behavioral disorders in the subjects and we are examining how this simple tool could be used to possibly predict intergenerational effects of the caregivers perceived stress levels on the children’s emotional and behavioral functioning and inflammatory cluster profiles in real-time.
The UCSF cohort was recruited from caregivers and children visiting a large urban safety-net hospital serving a mostly low income and ethnically and racially diverse patient population. Unlike many community samples that are biased towards recruiting mostly individuals at low risk of experiencing life stress, this sample offers a unique opportunity to determine how life stress as perceived by the children’s caregivers associates with children’s cytokine levels, and how these interactions can predict risk for problems in children’s emotional and behavioral functioning. We then explored similar relationships in a cross-cultural replication using the GUSTO cohort, a cohort comprising an ethnically dissimilar sample compared to the UCSF cohort. We hypothesized that higher salivary cytokine profiles in the children and higher perceived caregiver stress would be associated with greater problems in the children’s emotional (UCSF and GUSTO) and behavioral functioning (GUSTO). We used different versions of the Pediatric Quality of Life Inventory (PedsQL™) to measure emotional functioning in UCSF children aged 0–17 years and the Child Behavior Checklist (CBCL) to measure emotional and behavioral functioning in GUSTO children aged 7 years. The PedsQL has also been shown to have value in predicting the most common behavioral disorder in childhood, ADHD (Varni and Burwinkle, 2006). We used clustering analyses to determine if children’s salivary cytokine levels can moderate the association between the caregiver’s perceived stress and emotional and behavioral problems in childhood. We hypothesized that children with higher salivary pro-inflammatory cytokine levels would exhibit greater emotional and behavioral problems when exposed to environments their caregivers perceived as more stressful.
Section snippets
University of California San Francisco (UCSF) cohort
Data were collected from 622 child-caregiver dyads. Only mothers, fathers and stepfathers who presented to the clinic as the child’s primary caregiver were included in the current study’s analyses (607 subjects). Any caregivers who presented to the hospital as the children’s extended relatives or family friends were excluded from the study. Of the 622 child-caregiver dyads that were screened for the study, 3 subjects were excluded due to the presenting caregiver at the interview being an adult
UCSF cohort
The children in the UCSF cohort ranged in age between 0 and 215 months. The UCSF cohort’s mean age is 80.23 months with a standard deviation of 59 and a median of 70. Associations between the participants demographic variables and the children’s emotional functioning scores were also assessed in the UCSF cohort.
Descriptives statistics were obtained for cytokine levels. Based on the skew (IL1β 5.74, IL6 17.52, IL8 2.49, TNF-α 4.15) and kurtosis (IL1β 41.62, IL6 340.34, IL8 6.69, TNF-α 24.23) for
UCSF cohort
Baseline characteristics for the total sample and subjects used in the cytokine clusters in the UCSF cohort are described in Table 1. Supplementary Table 1 lists the exposure and outcome variables used in the UCSF cohort. Of the original 622 caregiver-child dyads screened for this study, 15 dyads were excluded because the caregiver was not the child’s mother, father or stepfather. There was a significant difference in the excluded childrens ages compared to the children that remained in the
Discussion
Our study is the first to explore if clustering pro-inflammatory cytokines by their profile levels in saliva can predict the emotional function of children aged 0–17 in response to caregiver perceived stress. While the salivary cytokine clusters were not significantly associated with the emotional function outcomes, the cytokine clusters did significantly moderate the association between the caregivers’ perceived stress scale scores and the children’s emotional functioning in one sample and
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
Funding for this research was supported by The JPB Foundation through a grant to The JPB Research Network on Toxic Stress: A Project of the Center on the Developing Child at Harvard University provided by the JPB Foundation to MJM and PPS. Several research assistants from the Social Interventions Research and Evaluation Network, Center for Health and Community, University of California, San Francisco were involved in the collection of data for the UCSF cohort. These research assistants include
References (85)
- et al.
Acute phase protein and cytokine levels in serum and saliva: a comparison of detectable levels and correlations in a depressed and healthy adolescent sample
Brain Behav. Immun.
(2013) - et al.
Neuroinflammation as a risk factor for attention deficit hyperactivity disorder
Pharmacol. Biochem. Behav.
(2019) - et al.
Individual differences in stress susceptibility and stress inhibitory mechanisms
Curr. Opinion Behav. Sci.
(2017) - et al.
The link between childhood trauma and depression: insights from HPA axis studies in humans
Psychoneuroendocrinology.
(2008) - et al.
Early life adversity: Lasting consequences for emotional learning
Neurobiol Stress.
(2017) - et al.
Elevated specific peripheral cytokines found in major depressive disorder patients with childhood trauma exposure: a cytokine antibody array analysis
Compr Psychiatry.
(2013) - et al.
Child abuse is related to inflammation in mid-life women: role of obesity
Brain Behav Immun.
(2014) - et al.
Clustering of depression and inflammation in adolescents previously exposed to childhood adversity
Biol Psychiatry.
(2012) - et al.
Childhood Adversity and Current Stress are related to Pro- and Anti-inflammatory Cytokines in Major Depression
J Affect Disord.
(2019) Paradise Lost: The Neurobiological and Clinical Consequences of Child Abuse and Neglect
Neuron
(2016)