0731 Salivary Cortisol and Sleep in Grandparents Raising Grandchildren and Non-Caregiving Grandparents

Abstract

Abstract Introduction Caregivers often exhibit unhealthy/dampened cortisol profiles and sleep problems due to high stress. Grandparent caregivers (GP-CG) may be particularly vulnerable with caregiving burden and age-related changes in sleep/health. Yet there is lack of empirical evidence examining baseline cortisol and reactivity in GP-CG and how they are associated with sleep. We examined differences between GP-CG and non-GP-CG in cortisol levels pre- and post-stress tasks and the moderating role of caregiving status in the relationship between cortisol and sleep. Methods 486 participants in the second wave of the 2009 MIDUS biomarkers project were characterized as GP-CG who cared for their grandchild for at least 1 year (n=63, Mage=61.80, 64% female) and non-GP-CG (n=423, Mage=62.98, 56.2% female). T-tests and multiple regressions (SPSS PROCESS) examined group differences and moderation by caregiving status in the associations between salivary cortisol (Salivette®; measured before and after Stroop and MATHTurner tasks) and sleep (work-day and non-work-day total sleep time-TST; frequency of Sleep Onset Latency problems-SOLfreq; 1-never to 5-always) over the past month. Analyses controlled for age, gender, cortisol collection time. Results GP-CG had lower baseline cortisol (M=10.35, SD=6.02) than non-GP-CG (M=12.15, SD=7.70; t(485)=2.03, p=.045). GP-CG had lower post-task cortisol reactivity (M=10.99, SD=5.83) than non-GP-CG (M=13.21, SD=7.73; t(485)=2.06, p=.040). Only non-GP-CG had a significant cortisol increase from baseline to post-task (t(1,422)=3.00, p=.001)). Caregiving status interacted with baseline cortisol (t(1, 484)=3.81, p=.04) in its association with SOLfreq (R2=.16, p=.03). Higher baseline cortisol was associated with less SOLfreq in GP-CG with no relationship in non-GP-CG. Additionally, caregiving status interacted with post-task cortisol (t(1,484)=7.79, p=.006) in its association with non-work-day TST (R2=.17, p=.04). Cortisol and TST had a positive relationship in GP-CG but a negative relationship in non-GP-CG. Conclusion GP-CG may have lower baseline cortisol and reactivity compared to non-GP-CG, consistent with previous literature examining individuals with prolonged stress compared to healthy controls. GP-CG’s reduced cortisol reactivity may be adaptive given that the positive relationship between cortisol and sleep in this group, contrary to the negative association found among healthy controls. Future research should utilize longitudinal data to determine the temporality of these relationships. Support (if any) NIA (P01AG020166, R37AG027343; McCrae, PI: R01AG061976)