Abstract
Stress physiology in pediatric occupational therapy patients should be considered within the context of the family system; family-based interventions may be particularly helpful for reducing patient stress in this population.
Stress undoubtedly affects health and is a major public health risk for most Americans. Populations already experiencing chronic stress, such as families with children who are neurologically atypical, are at particular risk for developing stress-related disease (Sapolsky, 2004). Indeed, parents of children with developmental disabilities report more stress than parents of typically developing children, often related to their child’s behavior severity (Baker et al., 2002; Hayes & Watson, 2013).
Effects of stress are also observed on physiological, immunological, and emotional outcomes among parents of neurologically atypical children. For instance, stressed mothers of children with autism spectrum disorder (ASD) have reported poor health-related quality of life and increased immune-related physical health problems (Reed et al., 2016), and 30% of mothers of children with cerebral palsy (CP) have reported clinical depression (Manuel et al., 2003). Parents of children with attention deficit hyperactivity disorder (ADHD) and intellectual disabilities have also reported high parenting stress (Anastopoulos et al., 1992; Estes et al., 2009; Eisenhower et al., 2005).
Among neurologically atypical children, those with ASD are at increased risk of cellular oxidative stress (James et al., 2004) and respond to novel and social–evaluative stimuli with altered physiological stress (Corbett et al., 2006; Spratt et al., 2012; Taylor & Corbett, 2014). Medical record reviews reveal that as adults, children with ASD are more likely than their typically developing counterparts to experience stress-related disease, including stroke, Parkinson’s, immune conditions, gastrointestinal and sleep disorders, seizures, obesity, hypertension, and diabetes (Croen et al., 2015). In addition, pediatric patients with sensory processing disorder demonstrate low parasympathetic nervous system activity (Schaaf et al., 2010), a sign of decreased adaptability to stressors (Porges, 1995).
Given the strong association between stress and chronic disease development, patients with ASD and their parents are likely to be at particular risk for chronic disease and premature aging. Although stress in this population may be inevitable, the persistent nature of familial stress associated with ASD and other developmentally related conditions warrants concern and attention (Hayes & Watson, 2013; Zaidman-Zait et al., 2014).
Family systems theory suggests that family members are affected by each other’s stress. Moreover, each family member must be understood in the context of the family, and a person’s psychosocial and physical well-being are not independent from familial influence (Ackerman, 1959). Thus, a child’s stress likely affects parent stress, just as a parent’s stress likely affects child stress. Stress can therefore be conceptualized as a “contagion” within parent–child dyads (Cox et al., 2001). Given the nature of chronic stress in families of children with ASD (Zaidman-Zait et al., 2014), parents may have fewer resources from which to gain positive social support, further contributing to compounded stressful life experiences (Hartley et al., 2018). If stress is contagious, then so are the subsequent consequences, such as reduced immune responses to vaccines, increased production of proinflammatory cytokines, increased risk of infection, and increased telomere shortening (Sapolsky, 2004).
Evaluating how variables known to influence the perception of stress affect both parents and their children dyadically is invaluable to understanding health at the family level. At its core, attachment is one such variable that converges with a person’s perspective on and perception and experience of stress (Bottonari et al., 2007; Hankin et al., 2005). Attachment theory (Bowlby, 1958) posits that people whose physical and emotional needs are consistently met in infancy and childhood develop a secure working model of relationships and the world, thus forming a secure attachment style (i.e., characterized by low anxiety and low avoidance in interpersonal relationships). In contrast, people whose needs are never or inconsistently met or who experience dysfunctional family relationships develop an insecure working model of relationships and the world, thus forming an insecure attachment style (i.e., characterized by high anxiety and high avoidance in interpersonal relationships).
In families with children who are neurologically atypical, parents with an insecure attachment style may have increased negative consequences as a result of the chronic nature of associated stress. Successful coping for parents of children with ASD and developmental delays depends on social support, relationships with extended family, and marriage quality (Derguy et al., 2016; Siman-Tov & Kaniel, 2011), three factors compromised in parents with insecure attachment styles. In addition, insecure attachment styles, as opposed to secure styles, are related to abnormal hypothalamic–pituitary–adrenal (HPA) axis reactivity (Diamond & Fagundes, 2010), indicated by an abnormal stress response as measured by the stress hormone cortisol. If attachment style influences a person’s perception and experience of stress, then it is important to understand the implications of this perception and experience for other family members. This study focuses on the attachment style of the parent, examining its effect on the parent’s own perception of stress and on a physiologically based measurement of stress in the patient–parent dyad.
To our knowledge, it is unclear whether attachment style makes parents who have children who are neurologically atypical more susceptible to stress and how stress transmits within the parent–child dyad in the context of day-to-day activities, appointments, and treatment for these children. This study examined the relationship between the parent factor of attachment style and both parent and child stress in the context of a pediatric occupational therapy clinic visit. We used a biomarker, cortisol, to measure physiological stress in children who are neurologically atypical, because self-report by these children is potentially inaccurate and further complicated by characteristic communication difficulties. We also used cortisol level to measure stress in parents and examined the synchrony of cortisol level in the patient–parent dyads. The aim of this pilot study was to establish feasibility for understanding biological stress in patient–parent dyads in the context of pediatric occupational therapy visits.
Method
Participants
Ten parent and pediatric patient dyads were recruited from an occupational therapy clinic housed in a master’s-level occupational therapy department. Participants younger than age 4 yr or who could not cognitively understand what was being asked of them were excluded. To control variability within the sample and account for the diurnal pattern of cortisol, the study included participants who presented to the clinic between 2:00 p.m. and 5:00 p.m.; who had no history of glucocorticoid use; and who had not consumed caffeine within 3 hr, engaged in vigorous exercise that day, consumed alcohol that day, or consumed food within 2 hr of their study participation.
Nine of the 10 children had evidence of ASD symptoms reported in their medical records; however, most had not undergone diagnostic procedures sufficient to receive an ASD diagnosis. The documented diagnoses of the pediatric participants included sensory processing disorder (n = 9), ADHD (n = 4), dyspraxia (n = 1), ASD (n = 1), spastic hemiplegic CP (n = 1), spastic quadriplegic CP (n = 1), seizure disorder (n = 1), and nonautistic motor stereotypes (n = 1). Each child had two or more diagnoses noted in their chart. Children were between ages 4 and 11 yr (M = 6.70, SD = 2.16), and parents were between ages 33 and 49 yr (M = 39, SD = 6.29). On average, parents had been in a romantic relationship for 11.35 yr (SD = 4.97) and were experiencing 1.20 (SD = 0.92) stressors (e.g., financial difficulties, difficulties with work, relationship challenges) in addition to coping with their child’s conditions. Participants included 9 mother–child dyads and 1 father–child dyad. Seven of the 10 children were male. Nine of the 10 parents identified as White; we did not collect data on child race or ethnicity.
Instruments
Demographic Information
Parents completed a demographic questionnaire that asked questions about age, race, gender, biological relationship to the patient, history of glucocorticoid use, current medications, recent exercise, and recent (within 3 hr) consumption of food, tobacco, alcohol, and caffeine. Child age and diagnosis were obtained from patient charts.
Experiences in Close Relationships–Revised
The Experiences in Close Relationships–Revised (ECR–R; Fraley et al., 2011) measured the attachment style of the parent on the dimensions of anxiety and avoidance. High anxiety scores, high avoidance scores, or both are indicative of insecure attachment style, whereas a combination of low anxiety and low avoidance scores reflects secure attachment. Items assessing anxiety (e.g., “I worry about being alone”) and avoidance (e.g., “I get uncomfortable when a romantic partner wants to be very close”) are strongly intercorrelated (Cronbach’s αs = .91 and .94, respectively). This measure has 36 items, each assessed on a 7-point Likert scale ranging from 1 (disagree strongly) to 7 (agree strongly). The measure was given only to parents, in paper-and-pencil format.
Perceived Stress Scale
The Perceived Stress Scale (PSS; Cohen & Williamson, 1988) assesses the degree to which situations in a person’s life are perceived as stressful (Cronbach’s α = .84–.86). This measure includes 10 items and uses a 5-point Likert scale ranging from 0 (never) to 4 (very often). It was administered in paper-and-pencil format and was used to assess only parent stress.
Autism Parenting Stress Index
The Autism Parenting Stress Index (APSI; Silva & Schalock, 2012) measures parenting stress specific to ASD and comorbid symptoms (Cronbach’s α = .83). The measure was also validated with children with developmental delays (Silva & Schalock, 2012). Parenting stress associated with children with sensory processing difficulties is also captured by this measure, because items include tantrums, sleep problems, diet, and concern for their child’s future. The measure includes 13 items and uses a 5-point Likert scale ranging from 0 (not stressful) to 4 (so stressful sometimes we feel we cannot cope). The measure was given only to parents, in paper-and-pencil format. It samples aspects of children’s health (i.e., social development) and asks parents to rate how stressful that aspect is to the entire family.
Child and Adolescent Survey of Experiences
The Child and Adolescent Survey of Experiences, Parent Version (CASE; Allen et al., 2012), measures stressful life experiences relevant to children and adolescents. The CASE parallels a standard clinical interview of life events but is more sensitive to capturing these events (relative risk = 1.15); it has high test–retest reliability (r = .75; Allen et al., 2012). Higher scores on the CASE are associated with a greater negative impact of life events and anxiety on the child. For this study, the measure was given only to parents, in paper-and-pencil format. According to Allen et al. (2012), agreement between parent and child on events endorsed is adequate (Ms = 7.78 and 8.23, respectively).
Procedure
Study procedures were approved by the institutional review board at the University of Puget Sound (Tacoma, WA). Researchers approached pediatric patients and their parents between the hours of 2:00 p.m. and 5:00 p.m. within 10 min of the start of their clinic visit (for best practices, see Hanrahan et al., 2006). Ten of 11 patient–parent dyads approached agreed to participate in the study. Nine of the 10 dyads had mothers as the parent participant, and 1 dyad had the father as the parent participant. Parents provided consent, and children were required to give verbal assent.
Families were queried regarding exclusion criteria, specifically for the purposes of mitigating extraneous variability associated with accurately measuring cortisol. All enrolled families met the inclusion criteria. Children and parents provided a saliva sample through passive drool oral swab method for a duration of 2 min. We used SalivaBio Children’s Swab (to reduce choking hazard) for children and SalivaBio Oral Swab for adults (Salimetrics, Carlsbad, CA). Child salivary gland activation was elicited by asking them to discuss their favorite foods. The parent then completed the self-report measures.
Parents were directed to take a parent and a child saliva sample at home within 2 days of their clinic visit, using the same method as in the clinic. The samples were to be collected on a “normal, relaxing day” at the same time of day that samples were taken in the clinic. Although obtaining multiple samples in and out of the clinic is ideal, we aimed to decrease participant burden by requiring only two samples each from parent and child. In addition, the methodology for observing synchrony with one sample from each member of the dyad at a single time point has been documented elsewhere (Kornienko et al., 2013; Williams et al., 2013).
Participants were asked to abide by directives to abstain from eating, drinking caffeine or alcohol, smoking, and exercise for 2 hr before sample collection and to store the saliva sample in their home freezer until their next clinic visit. They were asked to bring the samples back to the clinic in a provided plastic bag with ice cubes. We compensated participants with a $5 gift card upon return of the home sample. Samples were stored at –20°C and then batch-shipped overnight on dry ice to Salimetrics. Cortisol levels were assayed from both saliva samples by Salimetrics using the enzyme-linked immunosorbent assay method. Data were analyzed using R Version 3.3.3 (R Foundation, Vienna, Austria).
Results
Factors Related to Parent Stress
Parent Self-Report
To examine the relationship between the parental attachment style dimensions of anxiety and avoidance and self-reported parenting stress, Pearson’s product–moment correlations were run between APSI and PSS scores and ECR–R Anxiety and Avoidance subscale scores (Table 1). No significant results were revealed; that is, no relationship was observed between either attachment style dimension and parenting stress or parent perceived stress.
Score Means and Standard Deviations and Pearson Correlations Between Measures
Note. Cortisol measured in micrograms per deciliter.
aMean adjusted for two outliers.
*p < .05. **p < .01.
Parent Cortisol Level at Home
Pearson’s product–moment correlations were calculated for parent cortisol level at home and ECR–R Anxiety and Avoidance subscale scores (see Table 1). No significant results and no patterns were revealed.
Parent Cortisol Level in the Clinic
Pearson’s product–moment correlations were calculated for parent cortisol level in the clinic and the ECR–R Anxiety and Avoidance subscale scores (see Table 1). A strong, positive relationship was revealed between parent cortisol level in the clinic and the Avoidance (r = .70, p = .01, two-tailed) and Anxiety (r = .67, p = .03, two-tailed) subscales.
Factors Related to Patient Stress
Parent-Reported Patient Stress
To examine the relationship between parent attachment anxiety and avoidance and parent-reported patient stress, as indicated on the CASE Negative Events subscale, we used Pearson’s product–moment correlations. These correlations were also calculated to examine the relationship between parent attachment anxiety and avoidance and positive patient affect, as indicated on the CASE Positive Events subscale (see Table 1). No results reached a conventional level of significance, but a moderate, nonsignificant negative relationship between parent attachment anxiety and positive child affect was observed.
Patient Cortisol Level at Home
Pearson’s product–moment correlations were calculated for patient cortisol level at home and parent ECR–R Anxiety and Avoidance subscale scores (see Table 1). These correlations revealed a strong, positive relationship between both parent attachment anxiety (r = .90, p < .01, two-tailed test) and avoidance (r = .86, p < .01, two-tailed test) and patient cortisol level at home.
Patient Cortisol Level in the Clinic
Pearson’s product–moment correlations were calculated for patient cortisol level in the clinic and parent ECR–R Anxiety and Avoidance subscale scores (see Table 1). These correlations revealed a strong, positive relationship between parent attachment anxiety and patient cortisol level in the clinic (r = .85, p < .01, two-tailed). A moderate, positive, nonsignificant relationship, but approaching significance (r = .59, p = .09), was found between parent attachment avoidance and patient cortisol level in the clinic.
Cortisol Level Synchrony Within the Dyad
Pearson’s product–moment correlations were calculated for the relationship between parent and patient cortisol level synchrony. We found a strong, positive relationship in the clinic (r =.70, p = .03) but no relationship at home.
Impact of the Clinic Visit on Stress Levels
To determine the degree to which a regularly scheduled visit to a community outpatient occupational therapy clinic was physiologically stressful for patients and parents, we conducted a paired-sample t test to detect differences between their home and clinic cortisol levels. Although we observed higher clinic values for 8 of 10 patients and 7 of 10 parents, there was no statistical difference between the home and clinic cortisol levels for parents or patients. See Table 1 for means and standard deviations for cortisol levels.
Discussion
This pilot study aimed to examine the relationship among a parent factor, attachment style, and stress in parents and their children who are neurologically atypical. We also considered stress in the context of an occupational therapy clinic visit versus at home. Finally, we used cortisol measures to bolster parent self-reported measures of stress and parents’ reports on their children. This study was exploratory in nature; we aimed to establish feasibility for examining stress in dyads in this population and guide future research directions. Results revealed that parent attachment anxiety and avoidance may have a meaningful impact on both parents’ and children’s physiological stress in this population. Increased parent attachment anxiety was related to higher child and parent cortisol levels in the clinic and higher child levels at home. Increased parent attachment avoidance was related to higher child cortisol levels at home, but not in the clinic.
Attachment style is a lens through which people perceive events, rendering them more or less susceptible to interpreting an event as stressful (Bowlby, 1958). Therefore, when a person scores higher in the dimensions of attachment anxiety or avoidance, we would expect their stress physiology to also increase. Indeed, we found that as parents’ attachment anxiety and avoidance increased, so did their cortisol levels in the clinic.
We were particularly interested in whether stress acted as a contagion within the family, a concept we tested by looking at synchrony between patient and parent cortisol levels in the clinic and at home. Our results indicated that as the parent’s clinic cortisol increased, so did the child’s clinic cortisol, but this relationship was not observed at home. In addition, Saxbe and colleagues (2017) found that HPA axis linkage occurred in children with ASD and their fathers, but not their mothers, in a stress paradigm done at home. We also did not observe synchrony within the dyads at home, with a primarily maternal sample.
We do not suggest that standard-of-care treatment of children who are neurologically atypical causes stress or is unwarranted; we simply observe that stress between the parent and the child was more closely related in the clinic than at home. The mechanism through which cortisol was related in our sample in the clinic is unclear and is worthy of further investigation. It could be that children’s cortisol level rises when they watch their parents become stressed. Alternatively, parents’ cortisol may spike when watching their children behaviorally react to new or less familiar environments. The directionality of the synchrony (i.e., who influences whom) is unknown and should be explored in future research. Whether this synchrony is adaptive remains unclear. In one study, higher environmental stress resulted in higher mother–child cortisol synchrony (Hibel et al., 2009), whereas other findings have indicated that greater parent sensitivity to and warmth toward their children in the context of stress increase cortisol synchrony (Sethre-Hofstad et al., 2002).
One of the main goals of this study was to establish feasibility for examining synchrony within dyads. Given the high enrollment rate and high rates of data completion, we suggest that future research may use such a protocol. Interestingly, we did not observe any relationship between a parent’s report of their child’s stress and the child’s actual physiological stress. These data suggest that future research should incorporate biobehavioral indicators to address patient stress and to accurately determine stress level because parent report of a child’s stress, although important, may not be reliable.
This study has several limitations, the first being the small sample size, which requires that the results be interpreted with caution. Although we used the term neurologically atypical to describe our current sample, we recognize the limitations of this term because typical is often on a spectrum. Only children who could verbally assent were enrolled in the study, limiting the generalizability of the results to nonverbal children. Parents may not have followed the proper procedure for sample collection at home even with guidance. We used only two samples of cortisol to determine stress synchrony at two distinct times, whereas determining cortisol levels at multiple time points over the course of several days and accounting for the diurnal pattern of cortisol is best practice and may have yielded different results. Requiring multiple cortisol samples may increase participant burden, however, and should be carefully considered before implementation.
Despite its limitations, this study contributes to the small but emerging literature on HPA axis linkage in this population. We also bolstered parent-report measures with physiological data to understand what happens to children who may not typically be able to report how stress manifests for them in the context of their family. Examining parents’ reporting on their children is crucial in this population because clinicians primarily rely on such reporting during practice. Finally, we demonstrated that this protocol is feasible and acceptable in this population.
Implications for Occupational Therapy Practice
The results of this study have the following implications for occupational therapy practice:
The strong, positive relationship between parent and child physiological stress in the clinic suggests that pediatric occupational therapy practitioners should pay attention to parent affect because it may be relevant to the child’s physiology.
If the parent seems overwhelmed, the child may experience a high stress response and may therefore benefit from a beginning activity that includes arousal reduction skills such as deep touch or rocking to reduce sympathetic nervous system activation (Parham & Mailloux, 2010).
Teaching arousal reduction skills may be particularly relevant outside of the clinic for children if their parents are experiencing a high degree of stress in other domains (e.g., relationships) or have low social support.
As recommended by the American Academy of Pediatrics, a family-based treatment approach may be helpful for addressing unwarranted or high physiological stress (Committee on Hospital Care and Institute for Patient- and Family-Centered Care, 2012).
Parents with high relationship stress or low social support may have insecure attachment characteristics (Bowlby, 1958). Therefore, referring them to psychotherapy may help them address inappropriate perceptions of stress and relationship challenges (Butler et al., 2006) that may physiologically affect their children.
Conclusion
We conclude that this protocol is feasible in pediatric occupational outpatient settings. On the basis of our preliminary data, parent and child cortisol may be more closely associated in clinic versus at home. Parent attachment anxiety and avoidance may be meaningfully associated with both parents’ and patients’ physiological stress in this population; however, further research is needed in this area. As recommended by the American Academy of Pediatrics, these data provide an emerging signal that family-centered care for pediatric occupational therapy patients may have implications for patient stress physiology.
Footnotes
Acknowledgments
We thank Renee Watling for her support in recruiting families for this study. In addition, we thank the families who participated. This research was funded by grants from the University of Puget Sound (Tacoma, WA) W. M. Keck Neuroscience Fellowship (Scott) and Psi Chi, the International Honor Society in Psychology (Scott).