Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Jun 4.
Published in final edited form as: Clin Nurs Res. 2020 Nov 20;30(5):680–689. doi: 10.1177/1054773820973274

Challenges in obtaining and assessing salivary cortisol and α-amylase in an over 60 population undergoing psychotherapeutic treatment for complicated grief: Lessons learned

Jesse M Bell 1, Tina M Mason 2, Harleah G Buck 3, Cindy S Tofthagen 4, Allyson R Duffy 5, Maureen W Groër 6, James P McHale 7, Kevin E Kip 8
PMCID: PMC8177751  NIHMSID: NIHMS1701870  PMID: 33218253

Abstract

Biomarkers may serve as objective measures in complicated grief (CG) potentially capturing responses to stress reduction treatment. This paper reports challenges in obtaining and assessing salivary cortisol and α-amylase (sAA) for a recent randomized clinical trial. Within-session changes in salivary cortisol and sAA for 54 older adults with CG who received Accelerated Resolution Therapy were compared with perceived stress measured by Subjective Units of Distress Scale. Bivariate correlations and multiple regressions examined changes in biomarkers. Protocols, study logs, and audit reports identified challenges. Challenges included obtaining unstimulated passive drool salivary samples and their analyses. Our sample of older females on multiple medications may have resulted in a perfect storm of moderating and intervening variables which affected the stress response. This paper contributes to the discussion on designing clinical trials for older adults which must account for physiologic changes, multimorbidity, and polypharmacy common in this population and makes recommendations moving forward.

Keywords: Accelerated Resolution Therapy, alpha amylase, biomarkers, complicated grief, cortisol, salivary

The death of a family member or friend is one of the most stressful life events an individual will experience (Dohrenwend, 2006; Simon et al., 2020). As many as 15% of bereaved individuals do not adjust to the death resulting in complicated grief (CG) (Center for Complicated Grief of the Columbia University School of Social Work, 2015; Shear et al., 2013). CG is unusually prolonged, severe grief which impairs daily functioning (Shear, 2015). As people age, the incidence of CG rises due to increases in peer deaths, including spouse/partners (Holland et al., 2013; Mason et al., 2020). Deleterious outcomes of CG include significant psychological stress, social isolation, loneliness, and anxiety, resulting in individuals neglecting their own physical and mental health (Allen et al., 2013; Marques et al., 2013; Sung et al., 2011; Utz et al., 2012).

Clinicians and researchers recognize the need for CG clinical trials (Bryant et al., 2014; Shear et al., 2016; Shear et al., 2014). However, there are challenges in building the evidence base for CG treatment. Certain challenges are conceptual, such as a lack of agreement on a specific definition and diagnostic criteria for this grief-related condition, with terms like prolonged grief disorder and persistent complex bereavement disorder used interchangeably with CG (Prigerson & Maciejewski, 2017; Simon et al., 2020). Similarly, CG is generally diagnosed based on clinical presentation or subjective measurement by self-report instruments such as the Inventory of Complicated Grief (Prigerson et al., 1995) or the Prolonged Grief Disorder-13 (Prigerson et al., 2009). It is known that cortisol levels and patterns are altered in persons with CG. An integrative review of a series of smaller CG studies (n=5 studies) found lower morning cortisol levels and a flattened slope at multiple times points throughout the day in those with CG (Mason & Duffy, 2019). It was hypothesized that this may indicate hypothalamic-pituitary-adrenal (HPA) axis dysregulation in CG but further examination in larger samples was needed. Given this very preliminary evidence base, it is possible that stress biomarkers such as cortisol and α-amylase (which vary in concentration based on stress levels) may serve as objective measures of stress in CG and potentially capture any response to stress reduction treatment (Beauchaine et al., 2008; Dhama et al., 2019; Steckl & Ray, 2018).

To address the lack of mechanistic data in mind-body interventions targeting CG, our team proposed to use less invasive objective measures of stress, salivary cortisol and α-amylase (sAA). These frequently used biomarkers serve as proxy measures of HPA axis (cortisol) and sympathetic nervous system (SNS) activation (Strahler et al., 2017). Cortisol is a major glucocorticoid produced by the adrenal cortex in response to stress (Richardson et al., 2015). Higher cortisol levels are powerful predictors of physiologic outcomes (including mortality) in CG, with gender variations (women > men) (Richardson et al., 2015). sAA, an important enzyme in the breakdown of starch, is elicited by SNS activation (Nater & Rohleder, 2009). While sAA has not been tested in CG it has shown abnormal variation in post-traumatic stress disorder which shares a similar symptomatology and high co-occurrence with CG suggesting that it might also be an appropriate stress biomarker (Thoma et al., 2012). However, few studies have evaluated stress biomarkers pre- and post-CG interventions (O’Connor, 2012; Mason & Duffy, 2019). This may be due to a second set of challenges which are more procedural than conceptual.

Our team had the needed expertise in CG, clinical trials in older adults, the specific intervention (Accelerated Resolution Therapy (ART)), and psychoneuroimmunology to conduct this study. Because of the well-known difficulties in obtaining saliva in older adults (Tanasiewicz et al., 2016; Wiener et al., 2010), we used a detailed, written saliva collection protocol which included current best practices, and research staff were well trained (and re-trained if needed) in the procedures. However, despite the strong team, protocol and staff training, both obtaining and then assessing these salivary stress markers proved daunting in this population. The purpose of this article is to report the challenges in obtaining and assessing salivary cortisol and sAA in older adults with CG for our recent randomized clinical trial, detailing both the challenges that were anticipated and those that were not.

Methods

Design

A total of 54 bereaved family caregivers were recruited from a large, multi-site hospice organization in the United States. This trial compared ART (Kip et al., 2012) versus a 4-week wait list control condition in relation to changes in CG symptoms. To be included participants had to be 60 years of age or older and experienced a death at least 12 months prior to enrollment. Additionally, they had to deny suicidal ideation, meet screening criteria for CG, and have elevated levels of trauma symptoms. Participants were excluded if they were receiving another form of psychotherapy or had a major psychiatric disorder, substance abuse, or cognitive impairment. The primary aim of the study was to compare pre-to-post ART symptom changes in magnitude and rate of change of CG, psychological trauma, and depression (Buck et al., 2020). The secondary aim of the study was to compare pre- and post-ART stress biomarkers, specifically salivary cortisol and sAA for older adults with CG to evaluate objective changes in the stress response in relation to treatment with ART. To accomplish this aim, pre- to post-ART session changes in the stress biomarkers were compared.

Procedure

Detailed procedures are available elsewhere (Buck et al., 2020) but briefly, ethical oversight was provided by a university Institutional Review Board (IRB # Pro00032358), the trial was registered with ClinicalTrials.gov (NCT03484338) and participants were randomly assigned to one of two groups after consenting – one group received ART once a week for the following four weeks; the other group waited 4 weeks (control condition) and then received ART. Each participant was seen by a trained (master’s level or above) ART therapist a minimum of one to a maximum of five times, leading to a total of 187 treatment sessions across participants. Pre- and post-session measures of salivary cortisol and sAA from these sessions were analyzed.

Measures

For this analysis, we correlated a measure of perceived stress (Subjective Units of Distress Scale (SUDS)) to stress biomarkers results (salivary cortisol and sAA). Participants completed the SUDS before and after each ART session. Stress biomarkers were either measured before and after ART session 1 and ART session 4 (cortisol) or before and after each ART session (sAA). The SUDS self-report rating scale (scale 0-10) is used to measure subjective feelings of disturbance or distress in an individual at a specific time point. Participants are asked to rate how they feel with regards to an issue(s) on a scale from 10 (unbearably bad) to 0 (total relief and serenity). The main purpose of this scale is to report progress to the individual (e.g. “last week you reported a ‘9’ and this week you are reporting a ‘7’). Salivary cortisol and sAA were measured by appropriate immunoassay kits by Salimetrics® using the protocols listed in the assay kit.

The study protocol called for saliva samples to be collected before and after treatment sessions with ART using the unstimulated passive drool method. Participants were asked to tilt their head forward while allowing the saliva to pool on the floor of the mouth, and then pass the saliva through the SalivaBio Collection Aid into a polypropylene vial. At consenting, participants were given written instructions to avoid eating, brushing teeth, or smoking for an hour prior to salivary collection, and to avoid alcohol intake for at least 12 hours before each collection. Participants were then verbally reminded of the instructions via telephone 24 hours prior to their treatment session. At the session participants were provided with water to rinse their mouth 10 minutes prior to the collection in order to avoid specimen compromise from sugar or acidic foods that lower sample pH or stimulate bacterial growth. Each specimen was labeled with the patient’s study ID number, date, and time of collection. To further avoid bacterial growth, all salivary samples were immediately put on ice and transported to the laboratory where they were frozen at or below −20°C within 8 hours of collection, where samples could be stored up to 6 months. At time of assay, the samples were thawed and centrifuged at 1500 x g (@3000 rpm) for 15 minutes.

Intervention

ART is a relatively new evidenced based alternative to traditional psychotherapy. ART has been shown to relieve stress symptoms in an average of approximately four, one-hour sessions (Kip et al., 2012). In a typical ART session, participants are asked to focus upon their traumatic event/experience (death of their family/friend in this case) while the therapist guides them through visualizations, having them direct attention to repetitive hand movements to perform lateral left-right eye movements. These eye movements are designed to help the brain reprocess the stimulus response of the event (Kip et al., 2012). The parent study hypothesized that ART would reduce CG and its associated psychological stress which was borne out in a small clinical trial (Buck et al., 2020) from which the data in the current study is derived.

Statistical Analysis

Participant demographics and clinical characteristics were assessed at the start of the trial using student t-test or Wilcoxon test (depending on distribution) for continuous variables such as salivary cortisol, sAA, and self-report test measurements; while categorical variables, such as sex, race, and medications were compared via Fisher’s exact test of proportions. To assess the association between the SUDS and each biological marker, bivariate correlations were examined. Similarly, to identify potential factors that may have influenced the biological analysis we used bivariate correlations with medications and demographic variables.

All analyses involved within session data. To capture the change in salivary cortisol, both Spearman’s correlations and forward stepwise multiple regression with the log transformed difference of the pre-first and final post-ART session cortisol value as the outcome variable were used. A constant was included to account for negative values (where cortisol levels increased). Similarly, for sAA analysis Spearman’s correlations and forward stepwise multiple regression were used with the log transformed difference of the pre- and final post-ART session sAA value as the outcome variable. A constant was again included to account for negative values (where sAA levels increased). Because participants assigned to the control condition crossed over to the intervention after the wait list period, they were subsequently included in the overall pool of ART participants for this analysis.

Results

Demographic Characteristics

A total of 65 participants were recruited, of whom, 54 (83.1%) were subsequently enrolled with 32 (59.3%) randomly assigned to receive the ART intervention immediately and the remaining 22 (41.7%) assigned to the control condition (4-week waitlist). All 32 participants assigned to immediately receive ART completed the assigned regimen. Four participants on the waitlist withdrew prior to treatment, the rest (n=18) subsequently were treated with ART after the waitlist period.

The average participant was a white female, approximately 67 years old, with some college/technical education or less, taking several prescription medications (reported by class) including antidepressants (47.6%), statins (28.6%), antianxiety (23.8%), antihypertensive (21.4%), and/or over the counter supplements (16.7%). See Table 1 for further demographic information. A total of 187 treatment sessions were conducted, with a mean of 3.7 (SD 0.8) sessions per study participant. The mean length of each ART session was 60.7 (SD 17.9) minutes. From clinical notes, participants worked on the following: traumatic events (95.3%), loss (92.3%), mixed grief and trauma (90.0%), simple grief (52.3%), and guilt (23.0%). Biological markers were only collected during ART participation, not during the control condition, so there was no “cross-over” with regard to these measurements.

Table 1.

Characteristics of Study Population by Sex

Characteristic Total (N=50) Female (N=42) Male (N=8) p-value
Age in years, mean, SD 68.0, 6.7 67.4, 6.6 71.1, 6.8 0.15
Age group by years, %
Less than 65 34.0 38.1 12.5 0.26
65 to 74 46.0 42.9 62.5
75 or older 20.0 19.0 25.0
Race, %
White 96.0 97.6 87.5 0.19
More than one race 4.0 2.4 12.5
Ethnicity, %
Not Hispanic 88.0 92.9 62.5 0.02
Hispanic 12.0 7.1 37.5
Marital status, %
Married/Partnered 12.2 14.6 0.0 0.37
Divorced 14.3 14.6 12.5
Widowed 67.3 63.4 87.5
Single/Never married 6.1 7.3 0.0
Formal education completed, %
Less than high school 20.4 14.6 50.0 0.01
Some college/technical 30.6 29.3 37.5
Associate degree 10.2 9.8 12.5
Bachelor’s degree 16.3 19.5 0.0
Graduate degree 22.4 26.8 0.0
No. times hospitalized since the death, n (%)
None 39(78.0) 32 (76.2) 7(87.5) 0.47
One time 2 (4.0) 2 (4.8) 0 (0.0)
Two times 2 (4.0) 1 (2.4) 1 (12.5)
3 or more times 6 (12.0) 6 (14.3) 0 (0.0)
Months since the death, mean, SD 24.6, 23.3 24.5, 23.1 25.5, 25.6 0.95
Medications by class, n (%)
Anti-anxiety medication 13(26.0) 10 (23.8) 3 (37.5) 0.30
Antibiotic/anti-viral medication 2 (4.0) 2 (4.8) 0 (0.0) 0.06
Anti-depressant medication 23 (46.0) 20 (47.6) 3 (37.5) 0.04
Anti-hypertensive medication 13 (26.0) 9 (21.4) 4 (50.0) 0.05
Anti-clotting medication 8 (16.0) 4 (9.5) 4 (50.0) 0.00
GERD 7 (14.0) 6 (14.3) 1 (12.5) 0.07
OTC supplement 7 (14.0) 7 (16.7) 0 (0.0) 0.04
Overactive bladder medication 3 (6.0) 3 (7.1) 0 (0.0) 0.05
Pain medication 9 (18.0) 7 (16.7) 2 (25.0) 0.05
Sleep medication 9 (18.0) 7 (16.7) 2 (25.0) 0.05
Statin medication 18 (36.0) 12 (28.6) 6 (75.0) 0.00
Steroid medication 2 (4.0) 2 (4.8) 0 (0.0) 0.06
Length of ART session in minutes, mean, SD 60.7, 17.9 59, 14.3 61.0, 18.6 0.77
Open in a new tab

Note. ART - Accelerated Resolution Therapy, GERD - gastroesophageal reflux disease, OTC - over the counter.

Challenge 1: Obtaining Saliva Samples

The first challenge, obtaining an unstimulated passive drool salivary sample, was evident from the beginning of the study. Despite extensive training the research assistants struggled with consistently implementing the protocol. This was in part because participants labored to produce sufficient saliva (a minimum of 2 ml was required) using passive drooling techniques. Retraining and reassignment of responsibility resulted in closer adherence. However, obtaining samples remained problematic. Participants reported disliking the process and when unaware that they were being observed, used more active suction methods resulting in more foam than saliva in the collection tube. The protocol was then amended to state that if there was difficulty producing saliva, paraffin wax would be offered to chew to stimulate salivary flow. Participants were cautioned not to swallow the wax or allow it to get into the saliva sample. Both amount of saliva per participant and participant’s comfort with the procedure subsequently improved with this protocol amendment.

Challenge 2: Assessing the Biomarkers

Change in Biomarkers Over Time.

The second challenge, assessing salivary cortisol and sAA in older adults with CG, only became evident as we began to analyze the data at the end of the study. At ART session #1, the mean (SD) within session change in salivary cortisol was positive, 0.11 (0.28) μg/dl. At session #4, the corresponding mean (SD) change was 0.01 (0.19) μg/dl. There was no discernable pattern with most participants showing little to modest variation in cortisol. Similarly, pre-ART session #1 to post-ART session #1, the mean (SD) within session change in sAA was −3.96 (76.6) μg/dl. At session #4, the corresponding mean (SD) within session change was −4.95 (72.6) μg/dl. Once again, there was wide variation (as evidenced by very large standard deviations) in magnitude and direction of within session changes in sAA over the 4 treatment sessions with ART. This wide variation suggests external factors may be contributing substantially to any acute change in levels of sAA.

Analysis of Subjective and Objective Stress Measures.

For the subjective measure of SUDS, the mean (SD) value at the beginning of the ART sessions was 8.0 (2.0) with a corresponding mean value of 2.7 (2.1) at the end of the ART sessions (Table 2). This represented a clinically meaningful mean within-session difference (reduction) of 5.3 (2.3). Spearman correlations between changes in the subjective measure of stress (SUDS) and objective measures of stress (salivary cortisol and sAA) by ART session minutes are listed in in Table 3. As seen, within (ART) session changes in the SUDS had little to no relationship with the within-session changes in salivary cortisol (r = −0.11, p=0.31) or sAA (r = −0.01, p=0.86). Given the minimal association between change in the subjective and objective measures of stress, associations with other variables were investigated.

Table 2.

Subjective and Objective Stress Measures before and after ART Sessions by Sex

Sex ART Session Number N SUDS Before ART SUDS After ART Cortisol Before ART* Cortisol After ART* sAA Before ART sAA After ART
male All Visits 26 7.15 2.58 0.30 0.23 129.73 154.74
female All Visits 152 8.13 2.66 0.25 0.19 145.60 147.70
male 1 7 7.86 4.14 0.40 0.28 157.84 179.35
female 1 42 8.19 3.46 0.27 0.15 151.15 152.21
male 2 7 6.57 1.86 124.59 159.53
female 2 39 8.18 2.36 152.20 142.36
male 3 6 8.17 2.33 131.17 151.10
female 3 34 8.21 2.82 142.44 159.56
male 4 6 6.00 1.83 0.19 0.18 101.49 124.09
female 4 37 7.95 2.00 0.23 0.22 135.24 137.32
Open in a new tab
*

Note. Cortisol measured pre-Visit 1 and post-Visit 4 only.

ART- Accelerated Resolution Therapy, SUDS- Subjective Units of Distress Scale, sAA- salivary alpha amylase

Table 3.

Correlation Coefficients between Change (Δ) in Biological Measures and ART Session Minutes

Δ in Cortisol Δ in sAA Δ in SUDS ART Session Minutes
Δ in Cortisol 1.0 −0.02 −0.11 −0.17
Δ in sAA 1.0 −0.01 0.01
Δ in SUDS 1.0 0.09
ART Session Minutes 1.0
Open in a new tab

Note: ART - Accelerated Resolution Therapy, sAA - salivary alpha amylase, SUDS - Subjective Units of Distress Scale

Our strategy for identifying potential factors that may have influenced the analysis was to examine bivariate correlations between medication and demographic variables and the SUDS, salivary cortisol, and sAA variables. For the SUDS, ART visit number (r=−.24), and anti-anxiety medications (r=−.23), both contributed to within-session decreases in self-reported distress, while overactive bladder medication (r=.26) appeared to relate to increased self-reported distress. For within-session changes in salivary cortisol, the following variables were weakly associated: antidepressant use (r=.23, p=.03)), minutes of ART session (r=.17, p=0.11), gastroesophageal reflux diseases (GERD) medications (r= −.19, p=.06), hormone replacement therapy (r= −.18, p=.07), ART visit number (r= −.21, p=.04), and Hispanic race (r= −.26, p=.012). For within-session changes in sAA, the following variables showed small-to-modest inverse correlations: over the counter medications (r= −.21, p=.005), hypothyroid medications (r= −.14, p<=05), anti-clotting (r= −.17, p=.02) and GERD medication (r= −.22, p=.005).

See Table 4 for characteristics of variables used in the following regressions. Multiple forward, stepwise regression (minimum step to enter at F < 0.15) on log transformed (base 10) values of change in salivary cortisol was used to account for the non-normal distribution of cortisol. After inclusion of pre-ART salivary cortisol value and change in SUDS, change in log-cortisol was regressed on the following medications: GERD, diabetes, anti-angina, hormone replace therapy (HRT), asthma, thyroid, statins, blood pressure, anti-anxiety, sleep, blood thinning, anti-coagulants, pain, steroids, overactive bladder, and anything over the counter (OTC), along with the variables: ART visit number, sex, age, race, income, and months since the death of their care recipient. As seen in Table 5, neither the pre-ART salivary cortisol value nor within-session change in SUDS (difference score) were associated with pre-to-post ART change in cortisol value. The variable most strongly associated with change in cortisol was use of antibiotic/anti-viral medication with a partial R2 value of 0.25 (F=26.05, p < 0.0001). Other variables associated with within-session change in cortisol value included use of over the counter supplement and ART visit number. Variables included in the model accounted for 39.6% of the observed variance in change in cortisol value.

Table 4.

Characteristics of non-Demographic Variables Entered into Regression Equations

Variables μ, SD Total (N=50) Female (N=42) Male (N=8) p-value
SUDS (pre-post) difference score 4.5, 2.5 4.7, 2.6 3.3, 1.7 0.13
Pre-ART Cortisol 0.3, 0.3 0.3, 0.3 0.4, 0.3 0.29
Pre-ART sAA 150.7, 84.6 151.5, 79.6 146.4, 113.4 0.88
Inventory of Complicated Grief 40.0, 9.3 40.5, 8.9 37.4, 11.4 0.39
Prolonged Grief Disorder Scale 39.2, 7.4 39.5, 7.5 38.0, 7.6 0.61
Charlson Comorbidity Index 0.7, 1.1 0.6, 1.0 1.4, 1.1 0.05
Open in a new tab
Table 5.

Factors Associated with Pre-to-Post ART Changes in Salivary Cortisol

Variable Number Factor β Partial R-Square Model R-Square Pr>F
* Pre-ART Cortisol 0.015 0.015 0.015 0.95
* SUDS difference score 0.032 0.015 0.015 0.29
3 Antibiotic/anti-viral medication −1.446 0.2539 0.2689 <.0001
4 ART Visit Number −0.090 0.0418 0.3107 0.04
5 OTC supplement −0.377 0.0401 0.3508 0.04
6 Overactive bladder medication −0.431 0.0264 0.3772 0.08
7 Pain medication −0.224 0.0187 0.3959 0.14
Open in a new tab
*

Note. Variable forced into the model prior to stepwise selection.

ART - Accelerated Resolution Therapy, SUDS - Subjective Units of Distress Scale, OTC - over the counter

As seen in Table 6, the pre-ART sAA value was strongly associated with pre-to-post ART change in sAA (F=24.4, p=0.0001) and explained 6.5% of the variation in this outcome. Change in SUDS rating was not associated with pre-to-post ART change in sAA. Consistent with the cortisol results, the variable most strongly associated with change in sAA value was use of antibiotic/anti-viral medication with a partial R2 value of 0.12 (F=22.2, p < 0.0001). Other variables associated with within-session change in sAA value are listed in Table 6 and include different medications, months since the death, and number of hospitalizations since the death. Variables included in the model accounted for 37.4% of the observed variance in change in sAA values.

Table 6.

Factors Associated with Pre-to-Post ART Changes in Salivary Alpha Amylase

Variable Number Factor β Partial R-Square Model R-Square Pr>F
* Pre-ART sAA 0.002 0.0645 0.0645 0.00
* SUDS difference score 0.024 0.0645 0.0645 0.24
3 Antibiotic/anti-viral medication −0.890 0.1206 0.185 <.0001
4 No. times hospitalized since the death −0.125 0.0512 0.2362 0.00
5 Months since the death 0.004 0.0227 0.259 0.03
6 OTC supplement −0.245 0.0185 0.2774 0.05
7 Anti-hypertensive medication −0.187 0.02 0.2974 0.04
8 Sleep medication −0.180 0.0122 0.3096 0.11
9 Annual household income 0.071 0.0132 0.3228 0.10
10 Prolonged Grief Disorder Scale total score 0.012 0.017 0.3398 0.06
11 Anti-Clotting Medication 0.304 0.0243 0.3641 0.02
12 Overactive bladder medication −0.266 0.0096 0.3737 0.14
Open in a new tab
*

Note. Variable forced into the model prior to stepwise selection.

ART - Accelerated Resolution Therapy, sAA - Salivary Alpha Amylase, SUDS - Subjective Units of Distress Scale, OTC - over the counter

Discussion

The purpose of this paper was to report the lessons that we learned when facing challenges in obtaining and assessing salivary cortisol and α-amylase in older adults with CG. Some of these challenges, such as difficulty in obtaining saliva samples were anticipated from the literature (Tanasiewicz et al., 2016). However, challenges in assessing the biomarkers and making meaning of our findings were not anticipated. Hence, this paper contributes to the discussion on designing clinical trials for older adults which must account for physiologic changes, multimorbidity, and polypharmacy common in this population (Herrera et al., 2010).

Lesson 1:

consider carefully the sampling method of the biomarker given the specific population being sampled. It was more than likely that challenges in obtaining saliva samples exacerbated the issues related to the analyses of the specific biomarkers being used by providing less than optimum amounts and types (active vs. passive techniques) of saliva. While we were well aware of the difficulties in obtaining saliva samples in older adults and quickly addressed them when they occurred in our study, we still question the quality of samples that were obtained. In the end, a blood draw might have been easier on both parties.

Lesson 2:

consider carefully which stress biomarker to use. Interpreting the data was complicated by some inherent limitations in the biomarkers that we used. Salivary cortisol is known to show large intra- and interindividual variability (Kudielka et al., 2009). In addition, cortisol increases with age with this effect greater in women than men (Strahler et al., 2017). Anti-depressant (Ruhé et al., 2015) and anti-anxiety medications (Manthey et al., 2010) affect salivary cortisol concentrations and patterns. Cortisol is particularly sensitive to medications such as GERD treatment, antibiotics, and pain medication (Granger et al., 2009; Kudielka et al., 2009; Sahu et al., 2014). Similarly, in sAA, while age is not typically associated directly with lower levels of sAA response, health conditions such as cancer, and medications such as anti-anxiety and GERD treatment can affect sAA response (Granger et al., 2009; Kudielka et al., 2009). Moreover, despite our best efforts, there were a number of patients who were not able to maintain consistency in appointment times, a problem which affects the overall diurnal variation of both cortisol and sAA. Considering each of these factors in combination, this sample of older women on multiple medications may have led to a perfect storm of moderating and intervening variables which affected both the stress response and its measurement in these participants.

Lesson 3:

consider carefully any potential confounding factors in the sampling frame which may impact the biomarker. Interpreting the data was also complicated by the disconnect between our subjective and objective stress measures. Significant reductions in the SUDS occurred from pre-to post ART, at the same time we did not observe expected changes in salivary cortisol or sAA. This suggests that there are factors apart from our intervention that can influence cortisol and sAA results in this population. There is a known disconnect between subjective and objective symptom measurement in other older adult populations (Andre et al., 2009; Riegel et al., 2018; Serra-Blasco et al., 2019) without commensurate examination of predisposing factors or plausible rationales for this disconnect. The lack of consistency between the subjective and objective measures of stress in this study may be due to the health status of our chosen population. Specifically, certain medications are known to impact stress biomarkers, including cortisol and sAA (Granger et al., 2009; Kudielka et al., 2009; Sahu et al., 2014). Our data provide additional evidence that this may be the case for older populations; in particular, this connection was revealed through our multivariable modeling and strong association between use of antibiotics and within-session change in cortisol and sAA.

Limitations

Limitations of this study include the absence of a control condition for the biological analysis, as all participants ultimately received the intervention. Additionally, there was insufficient racial diversity and disproportionately low male participation. Therefore, reliable estimation with regards to these groups is impossible. Finally, while we could account for some medications, it is impossible to adjust for all, for example (as reported in the Results section): generic OTC medications displayed a much higher correlation than would be expected (r=.21). While speculative at this point, this may in some measure reflect use of probiotics, which have been shown to affect stress hormone levels in animal models (Ait-Belgnaoui et al., 2012) and are used widely by older adults (Sarah et al., 2016).

Conclusions

The present study highlights the challenges that we faced while attempting to capture differences in results of an evidence-based psychotherapy (ART) in a new population (CG). We put forward issues that researchers should be aware in the older adult population when using salivary stress biomarkers. While we are not recommending against the use of salivary cortisol in older adults, we are suggesting using this method cautiously. Measuring stress, particularly in this population is important and depending on the research design there are other methods to consider (hair cortisol, for example) or possibly the inclusion of other biomarkers that can be collected and measured through blood. More research is needed in this area. For example, comparing salivary and hair cortisol in a large sample of older adults will provide valuable information regarding the reliability of both methods. Whereas some irregularity in results may be expected in any study, our results support the postulate that medications, particularly polypharmacy, common in older adults, may substantially influence objective measures of stress for research purposes. These challenges put a premium on considering health status and comorbidities of study populations in the selection of biomarkers for research purposes, particularly those sensitive to external influences. To continue to exclude older adults from clinical trials because it is a difficult population to assess is not an ethical option.

Acknowledgments

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Institute on Aging of the National Institutes of Health under award number [R21AG056584].

Footnotes

Declaration of Conflicting Interests: The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Dr. Kevin Kip is on the Board of Directors for the International Society of Accelerated Resolution Therapy but does not receive payment for this advising position. The rest of the authors have no conflict of interest to declare.

Contributor Information

Jesse M. Bell, University of South Florida, College of Public Health.

Tina M. Mason, Moffitt Cancer Center, Department of Nursing Research, Nurse Researcher and University of South Florida, College of Nursing.

Harleah G. Buck, University of South Florida, College of Nursing.

Cindy S. Tofthagen, Mayo Clinic, Division of Nursing Research.

Allyson R. Duffy, University of South Florida, College of Nursing.

Maureen W. Groër, University of South Florida, College of Nursing and Morsani College of Medicine.

James P. McHale, University of South Florida, Department of Psychology.

Kevin E. Kip, University of Pittsburgh Medical Center, Health Services Division.

References

  1. Ait-Belgnaoui A, Durand H, Cartier C, Chaumaz G, Eutamene H, Ferrier L, Houdeau E, Fioramonti J, Bueno L, & Theodorou V (2012). Prevention of gut leakiness by a probiotic treatment leads to attenuated HPA response to an acute psychological stress in rats. Psychoneuroendocrinology, 37(11), 1885–1895. 10.1016/j.psyneuen.2012.03.024 [DOI] [PubMed] [Google Scholar]
  2. Allen JY, Haley WE, Small BJ, Schonwetter RS, & McMillan SC (2013). Bereavement among hospice caregivers of cancer patients one year following loss: Predictors of grief, complicated grief, and symptoms of depression. Journal of Palliative Medicine, 16(7), 745–751. 10.1089/jpm.2012.0450 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Andre R, Vuyk H, Ahmed A, Graamans K, & Nolst Trenite G (2009). Correlation between subjective and objective evaluation of the nasal airway. A systematic review of the highest level of evidence. Clinical Otolaryngology, 34(6), 518–525. [DOI] [PubMed] [Google Scholar]
  4. Beauchaine TP, Neuhaus E, Brenner SL, & Gatzke-Kopp L (2008). Ten good reasons to consider biological processes in prevention and intervention research. Development and Psychopathology, 20(3), 745–774. 10.1017/s0954579408000369 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bryant RA, Kenny L, Joscelyne A, Rawson N, Maccallum F, Cahill C, Hopwood S, Aderka I, & Nickerson A (2014). Treating prolonged grief disorder: A randomized clinical trial. JAMA Psychiatry, 71(12), 1332–1339. [DOI] [PubMed] [Google Scholar]
  6. Buck HG, Cairns P, NE BP, Hernandez DF, Mason TM, Bell J, Kip KE, Barrison P, & Tofthagen C (2020). Accelerated Resolution Therapy: Randomized controlled trial of a complicated grief intervention. American Journal of Hospice & Palliative Medicine, 1049909119900641. 10.1177/1049909119900641 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Center for Complicated Grief of the Columbia University School of Social Work. (2015). Official Diagnostic Criteria. Retrieved May 27 from https://complicatedgrief.columbia.edu/professionals/complicated-grief-professionals/diagnosis/
  8. Dhama K, Latheef SK, Dadar M, Samad HA, Munjal A, Khandia R, Karthik K, Tiwari R, Yatoo MI, Bhatt P, Chakraborty S, Singh KP, Iqbal HMN, Chaicumpa W, & Joshi SK (2019). Biomarkers in stress related diseases/disorders: Diagnostic, prognostic, and therapeutic values. Frontiers in Molecular Biosciences, 6, 91. 10.3389/fmolb.2019.00091 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dohrenwend BP (2006). Inventorying stressful life events as risk factors for psychopathology: Toward resolution of the problem of intracategory variability. Psychological Bulletin, 132(3), 477–495. 10.1037/0033-2909.132.3.477 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Granger DA, Hibel LC, Fortunato CK, & Kapelewski CH (2009). Medication effects on salivary cortisol: Tactics and strategy to minimize impact in behavioral and developmental science. Psychoneuroendocrinology, 34(10), 1437–1448. 10.1016/j.psyneuen.2009.06.017 [DOI] [PubMed] [Google Scholar]
  11. Herrera AP, Snipes SA, King DW, Torres-Vigil I, Goldberg DS, & Weinberg AD (2010). Disparate inclusion of older adults in clinical trials: Priorities and opportunities for policy and practice change. American Journal of Public Health, 100(Suppl 1), S105–S112. 10.2105/AJPH.2009.162982 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Holland JM, Futterman A, Thompson LW, Moran C, & Gallagher-Thompson D (2013). Difficulties accepting the loss of a spouse: A precursor for intensified grieving among widowed older adults. Death Studies, 37(2), 126–144. 10.1080/07481187.2011.617489 [DOI] [PubMed] [Google Scholar]
  13. Kip KE, Elk CA, Sullivan KL, Kadel R, Lengacher CA, Long CJ, Rosenzweig L, Shuman A, Hernandez DF, Street JD, Girling SA, & Diamond DM (2012). Brief treatment of symptoms of post-traumatic stress disorder (PTSD) by use of Accelerated Resolution Therapy (ART®). Behavioral Sciences (Basel), 2(2), 115–134. 10.3390/bs2020115 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kudielka BM, Hellhammer DH, & Wüst S (2009). Why do we respond so differently? Reviewing determinants of human salivary cortisol responses to challenge. Psychoneuroendocrinology, 34(1), 2–18. 10.1016/j.psyneuen.2008.10.004 [DOI] [PubMed] [Google Scholar]
  15. Manthey L, Giltay EJ, van Veen T, Neven AK, Vreeburg SA, Penninx BW, & Zitman FG (2010). Long-term benzodiazepine use and salivary cortisol: the Netherlands Study of Depression and Anxiety (NESDA). Journal of Clinical Psychopharmacology, 30(2), 160–168. 10.1097/JCP.0b013e3181d41f41 [DOI] [PubMed] [Google Scholar]
  16. Marques L, Bui E, LeBlanc N, Porter E, Robinaugh D, Dryman MT, Nadal-Vicens M, Worthington J, & Simon N (2013). Complicated grief symptoms in anxiety disorders: Prevalence and associated impairment. Depression and Anxiety, 30(12), 1211–1216. 10.1002/da.22093 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Mason TM, & Duffy AR (2019). Complicated grief and cortisol response: An integrative review of the literature. Journal of the American Psychiatric Nurses Association, 25(3), 181–188. 10.1177/1078390318807966 [DOI] [PubMed] [Google Scholar]
  18. Mason TM, Tofthagen CS, & Buck HG (2020). Complicated grief: Risk factors, protective factors, and interventions. Journal of Social Work in End-of-Life & Palliative Care, 1–24. 10.1080/15524256.2020.1745726 [DOI] [PubMed] [Google Scholar]
  19. Nater UM, & Rohleder N (2009). Salivary alpha-amylase as a non-invasive biomarker for the sympathetic nervous system: Current state of research. Psychoneuroendocrinology, 34(4), 486–496. 10.1016/j.psyneuen.2009.01.014 [DOI] [PubMed] [Google Scholar]
  20. Prigerson HG, Horowitz MJ, Jacobs SC, Parkes CM, Aslan M, Goodkin K, Raphael B, Marwit SJ, Wortman C, Neimeyer RA, Bonanno GA, Block SD, Kissane D, Boelen P, Maercker A, Litz BT, Johnson JG, First MB, & Maciejewski PK (2009). Prolonged grief disorder: Psychometric validation of criteria proposed for DSM-V and ICD-11. PLoS Medicine, 6(8), e1000121. 10.1371/journal.pmed.1000121 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Prigerson HG, & Maciejewski PK (2017). Rebuilding consensus on valid criteria for disordered grief. JAMA Psychiatry, 74(5), 435–436. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Prigerson HG, Maciejewski PK, Reynolds CF 3rd, Bierhals AJ, Newsom JT, Fasiczka A, Frank E, Doman J, & Miller M (1995). Inventory of Complicated Grief: A scale to measure maladaptive symptoms of loss. Psychiatry Research, 59(1-2), 65–79. [DOI] [PubMed] [Google Scholar]
  23. Richardson VE, Bennett KM, Carr D, Gallagher S, Kim J, & Fields N (2015). How Does bereavement get under the skin? The effects of late-life spousal loss on cortisol levels. The Journals of Gerontology: Series B, 70(3), 341–347. 10.1093/geronb/gbt116 [DOI] [PubMed] [Google Scholar]
  24. Riegel B, Dickson VV, Lee CS, Daus M, Hill J, Irani E, Lee S, Wald JW, Moelter ST, Rathman L, Streur M, Baah FO, Ruppert L, Schwartz DR, & Bove A (2018). A mixed methods study of symptom perception in patients with chronic heart failure. Heart & Lung, 47(2), 107–114. 10.1016/j.hrtlng.2017.11.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Ruhé HG, Khoenkhoen SJ, Ottenhof KW, Koeter MW, Mocking RJ, & Schene AH (2015). Longitudinal effects of the SSRI paroxetine on salivary cortisol in Major Depressive Disorder. Psychoneuroendocrinology, 52, 261–271. 10.1016/j.psyneuen.2014.10.024 [DOI] [PubMed] [Google Scholar]
  26. Sahu GK, Upadhyay S, & Panna SM (2014). Salivary alpha amylase activity in human beings of different age groups subjected to psychological stress. Indian Journal of Clinical Biochemistry, 29(4), 485–490. 10.1007/s12291-013-0388-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Sarah HY, Jernigan JA, & McDonald LC (2016). Prevalence of probiotic use among inpatients: A descriptive study of 145 US hospitals. American Journal of Infection Control, 44(5), 548–553. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Serra-Blasco M, Torres IJ, Vicent-Gil M, Goldberg X, Navarra-Ventura G, Aguilar E, Via E, Portella MJ, Figuereo I, & Palao D (2019). Discrepancy between objective and subjective cognition in major depressive disorder. European Neuropsychopharmacology, 29(1), 46–56. [DOI] [PubMed] [Google Scholar]
  29. Shear MK (2015). Complicated grief. New England Journal of Medicine, 372(2), 153–160. Doi: 10.1056/NEJMcp1315618 [DOI] [PubMed] [Google Scholar]
  30. Shear MK, Ghesquiere A, & Glickman K (2013). Bereavement and complicated grief. Current Psychiatry Reports, 15(11), 406. 10.1007/s11920-013-0406-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Shear MK, Reynolds CF 3rd, Simon NM, Zisook S, Wang Y, Mauro C, Duan N, Lebowitz B, & Skritskaya N (2016). Optimizing treatment of complicated grief: A randomized clinical trial. JAMA Psychiatry, 73(7), 685–694. 10.1001/jamapsychiatry.2016.0892 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Shear MK, Wang Y, Skritskaya N, Duan N, Mauro C, & Ghesquiere A (2014). Treatment of complicated grief in elderly persons: A randomized clinical trial. JAMA Psychiatry, 71(11), 1287–1295. 10.1001/jamapsychiatry.2014.1242 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Simon NM, Shear MK, Reynolds CF, Cozza SJ, Mauro C, Zisook S, Skritskaya N, Robinaugh DJ, Malgaroli M, Spandorfer J, & Lebowitz B (2020). Commentary on evidence in support of a grief-related condition as a DSM diagnosis. 37(1), 9–16. 10.1002/da.22985 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Steckl AJ, & Ray P (2018). Stress biomarkers in biological fluids and their point-of-use detection. ACS Senorss, 3(10), 2025–2044. 10.1021/acssensors.8b00726 [DOI] [PubMed] [Google Scholar]
  35. Strahler J, Skoluda N, Kappert MB, & Nater UM (2017). Simultaneous measurement of salivary cortisol and alpha-amylase: Application and recommendations. Neuroscience & Biobehavioral Review, 83, 657–677. 10.1016/j.neubiorev.2017.08.015 [DOI] [PubMed] [Google Scholar]
  36. Sung SC, Dryman MT, Marks E, Shear MK, Ghesquiere A, Fava M, & Simon NM (2011). Complicated grief among individuals with major depression: Prevalence, comorbidity, and associated features. Journal of Affective Disorders, 134(1-3), 453–458. 10.1016/j.jad.2011.05.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Tanasiewicz M, Hildebrandt T, & Obersztyn I (2016). Xerostomia of various etiologies: A review of the literature. Advances in Clinical and Experimental Medicine: Official Organ Wroclaw Medical University, 25(1), 199–206. [DOI] [PubMed] [Google Scholar]
  38. Thoma MV, Joksimovic L, Kirschbaum C, Wolf JM, & Rohleder N (2012). Altered salivary alpha-amylase awakening response in Bosnian War refugees with posttraumatic stress disorder. Psychoneuroendocrinology, 37(6), 810–817. 10.1016/j.psyneuen.2011.09.013 [DOI] [PubMed] [Google Scholar]
  39. Utz RL, Caserta M, & Lund D (2012). Grief, depressive symptoms, and physical health among recently bereaved spouses. The Gerontologist, 52(4), 460–471. 10.1093/geront/gnr110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Wiener RC, Wu B, Crout R, Wiener M, Plassman B, Kao E, & McNeil D (2010). Hyposalivation and xerostomia in dentate older adults. The Journal of the American dental association, 141(3), 279–284. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES