Salivary cytokine cluster moderates the association between caregivers perceived stress and emotional functioning in youth

https://doi.org/10.1016/j.bbi.2021.02.025Get rights and content

Highlights

  • Psychiatric disorders associate with early life stress and systemic inflammation.

  • Approach was established to identify children at risk of psychiatric problems.

  • Hierarchical clustering technique was used to cluster child salivary cytokine levels.

  • Clusters moderated caregivers perceived stress on child’s emotionality or behavior.

Abstract

Some individuals exposed to early life stress show evidence of enhanced systemic inflammation and are at greater risk for psychopathology. In the current study, caregivers and their offspring (0–17 years) were recruited at a pediatric clinic visit at the University of California, San Francisco (UCSF). Mothers and seven-year-old children from the Growing Up in Singapore Towards healthy Outcomes (GUSTO) prospective birth cohort were used as a replication cohort. Caregivers perceived stress was measured to determine potential intergenerational effects on the children’s functioning and inflammation levels. Children’s emotional functioning in the UCSF cohort was evaluated using the Pediatric Quality of Life (PedsQL) inventory. Child emotional and behavioral functioning was measured using the Child Behavior Checklist (CBCL) in GUSTO. Saliva was collected from the children and salivary levels of IL-6, IL-1β, IL-8 and TNF-α were measured using an electrochemiluminescent cytokine multiplex panel. Child IL-6, IL-1β, IL-8 cytokine levels were clustered into low, average, and high cytokine cluster groups using hierarchical cluster analysis. We did not find that salivary cytokine clusters were significantly associated with children’s emotional or behavioral function. However, cytokine clusters did significantly moderate the association between increased caregiver perceived stress and reduced child emotional functioning (UCSF cohort) and increased Attention-Deficit-Hyperactivity (ADH) problems (GUSTO cohort, uncorrected Cohen’s F2 = 0.02). Using a cytokine clustering technique may be useful in identifying those children exposed to increased caregiver perceived stress that are at risk of emotional and attention deficit hyperactivity problems.

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)

  • M.F. O'Connor et al.

    To assess, to control, to exclude: effects of biobehavioral factors on circulating inflammatory markers

    Brain Behav Immun.

    (2009)
  • M.F. O'Connor et al.

    When grief heats up: pro-inflammatory cytokines predict regional brain activation

    Neuroimage.

    (2009)
  • E.J. Ramos et al.

    Is obesity an inflammatory disease?

    Surgery.

    (2003)
  • J.L. Riis et al.

    Maternal distress and child neuroendocrine and immune regulation

    Soc Sci. Med.

    (2016)
  • D.C. Slavish et al.

    Salivary markers of inflammation in response to acute stress

    Brain Behav Immun.

    (2015)
  • J.M. Wolf et al.

    Parent psychological states predict changes in inflammatory markers in children with asthma and healthy children

    Brain Behav Immun.

    (2008)
  • T.M. Achenbach et al.

    DSM-oriented and empirically based approaches to constructing scales from the same item pools

    J. Clin. Child Adolesc. Psychol.

    (2003)
  • Achenbach, T.M., Rescorla, L.A, 2001. Manual for the ASEBA School-Age Forms and Profiles. University of Vermont,...
  • M. Aebi et al.

    Accuracy of the DSM-oriented attention problem scale of the child behavior checklist in diagnosing attention-deficit hyperactivity disorder

    J. Atten. Disord.

    (2010)
  • E. Alisic et al.

    Rates of post-traumatic stress disorder in trauma-exposed children and adolescents: meta-analysis

    Br. J. Psychiatry.

    (2014)
  • Baskind, M.J., Taveras, E.M., Gerber, M.W., Fiechtner, L., Horan, C., Sharifi, M., 2019. Parent-Perceived Stress and...
  • P.J. Barnes

    Anti-inflammatory actions of glucocorticoids: molecular mechanisms

    Clin. Sci. (Lond)

    (1998)
  • D. Baumeister et al.

    Childhood trauma and adulthood inflammation: a meta-analysis of peripheral C-reactive protein, interleukin-6 and tumour necrosis factor-α

    Mol. Psychiatry.

    (2016)
  • J. Baxter et al.

    Attention deficit disorder and the internalizing dimension in males, ages 9–0 through 11–11

    Int. J. Neurosci.

    (2004)
  • T.M. Beckie

    A systematic review of allostatic load, health, and health disparities

    Biol. Res. Nurs.

    (2012)
  • M.E. Beutel et al.

    Childhood adversities and distress - the role of resilience in a representative sample

    PLoS ONE

    (2017)
  • Biederman, J., Faraone, S.V., 2005. Attention-deficit hyperactivity disorder. Lancet; 366 (9481), 237–248....
  • D. Boerrigter et al.

    Using blood cytokine measures to define high inflammatory biotype of schizophrenia and schizoaffective disorder

    J. Neuroinflammation.

    (2017)
  • K.S. Chan et al.

    The PedsQL: reliability and validity of the short-form generic core scales and Asthma Module

    Med. Care.

    (2005)
  • S. Chandelkar et al.

    Resilience, Emotional Stability and Perceived Stress among 10th graders

    Indian J. Mental Health.

    (2019)
  • Chen, T.J., Ji, C.Y., Wang, S.S., Lichtenstein, P., Larsson, H., Chang, Z., 2016. Genetic and environmental influences...
  • Christensen, D.S., Dich, N., Flensborg-Madsen, T., Garde, E., Hansen, Å., Mortensen, E.L., 2019. Objective and...
  • E. Chen et al.

    Maternal warmth buffers the effects of low early-life socioeconomic status on pro-inflammatory signaling in adulthood

    Mol. Psychiatry

    (2011)
  • D. Cicchetti et al.

    Child maltreatment, inflammation, and internalizing symptoms: Investigating the roles of C-reactive protein, gene variation, and neuroendocrine regulation

    Dev. Psychopathol.

    (2015)
  • M.S. Clark et al.

    Environmental stress, psychological stress and allostatic load

    Psychol. Health Med.

    (2007)
  • L. Cohen et al.

    A Modular Cytokine Analysis Method Reveals Novel Associations With Clinical Phenotypes and Identifies Sets of Co-signaling Cytokines Across Influenza Natural Infection Cohorts and Healthy Controls

    Front. Immunol.

    (2019)
  • S. Cohen et al.

    A global measure of perceived stress

    J. Health Soc. Behav.

    (1983)
  • S. Cohen et al.

    Who’s Stressed? Distributions of Psychological Stress in the United States in Probability Samples from 1983, 2006, and 2009

    J. Appl. Soc. Psychol.

    (2012)
  • E.M. Condon et al.

    Associations between maternal experiences of discrimination and biomarkers of toxic stress in school-aged children

    Matern. Child Health J.

    (2019)
  • A. Danese et al.

    Elevated inflammation levels in depressed adults with a history of childhood maltreatment

    Arch. Gen. Psychiatry.

    (2008)
  • A. Danese et al.

    Childhood maltreatment predicts adult inflammation in a life-course study

    Proc. Natl. Acad. Sci. U S A.

    (2007)
  • David, J., Measelle, J., Ostlund, B., Ablow, J., 2017. Association between early life adversity and inflammation during...
  • Cited by (0)

    View full text