Feasibility and Acceptability of a Telephone-based ACT Intervention for Caregivers (TACTICs) of Adults with Alzheimer’s Disease and Related Dementias (ADRD): A Randomized Pilot during the COVID-19 Pandemic

Abstract

Objectives

Examine the feasibility, acceptability, and preliminary effects of the Telephone Acceptance and Commitment Therapy Intervention for Caregivers (TACTICs) on dementia caregivers’ anxiety, depression, caregiver burden, suffering, and anticipatory grief.

Methods

TACTICs was a 2-arm pilot randomized trial with dementia caregivers ≥ 21 years old with clinically elevated anxiety or anxiety-related functional interference. Two cohorts were recruited at the beginning and end of the first year of the COVID-19 pandemic. Intervention participants received 6 telephone sessions delivered by a non-licensed interventionist and control participants received readings and a list of dementia caregiver support groups in their area. Outcomes were measured at baseline, post-intervention, and at 3 and 6 months. Feasibility was measured by enrollment rates, completion rates, and adherence to TACTICs by interventionists. Acceptability was measured with participant satisfaction surveys.

Results

TACTICs was feasible and acceptable with 96% enrollment, 98.6% adherence, 65.5% completion, and a mean satisfaction score of 9.35 out of 10 (SD 0.91). The TACTICs group showed clinically relevant reductions in anxiety post-intervention (SRM 4.1, 95% CI [2.4,5.8]), however reductions were not significantly different from the control group (p=0.98).

Conclusion

Implementation of TACTICs during the COVID-19 pandemic was feasible and acceptable. The preliminary outcomes were not as strong as expected. The results will inform the design of future trials with larger samples.

INTRODUCTION

Developing effective and scalable interventions to support family caregivers of people living with Alzheimer’s disease or a related dementia (ADRD) is a national priority (Alzheimer’s Association, 2024; Hebert et al., 2013). Among the 6.9 million individuals in the U.S. living with ADRD, 83% are cared for by nonpaid family caregivers (Schulz & Martire, 2004). Over 11 million people in the U.S. provide care to a family member or friend with ADRD (Alzheimer’s Association, 2024; National Alliance for caregiving, 2004). ADRD caregivers are often called the “silent second patient” and experience significantly more stress and burden than their non-caregiver peers (Pinquart & Sorensen, 2003). Caregiving for someone with ADRD is associated with negative physical and mental health consequences (Cooper, Balamurali, & Livingston, 2007; Pinquart & Sorensen, 2007; Schulz & Martire, 2004; Vitaliano et al., 2003), including symptoms of anxiety and depression (Cooper, Balamurali, & Livingston, 2007; Crespo et al., 2005; Cuijpers, 2005; Joling et al., 2015; Pinquart & Sorensen, 2007), burden (Cheng, 2017), and existential suffering (Schulz et al., 2010). Studies have shown that the COVID-19 pandemic exacerbated these negative psychological factors (Carbone et al., 2021; Hwang et al., 2021; Lucas et al., 2021; Masoud et al., 2022), with 96% of caregivers reporting the pandemic affected their anxiety(Masoud et al., 2022), and up to 44% of caregivers reporting increases in hours and duties caregiving as a result of the pandemic (Hwang et al., 2021; Masoud et al., 2022).

Over the past three decades, many interventions and programs have been developed and evaluated to support ADRD caregivers (Boyt et al., 2022; Dalton et al., 2018; Leng et al., 2020; Zhang et al., 2020). Many have focused on building caregivers’ skills for managing and coping with the symptoms of ADRD along with reducing their stress and burden. These skill-building protocols primarily focus on changeable aspects of the caregiving experience, such as how to reduce repetitive question asking or de-escalating behavioral outbursts of the person living with ADRD. However, not all caregivers equally benefit from these programs, and not all caregiver well-being outcomes, such as anxiety, are appropriate outcomes for the unchangeable aspects of caregiving such as existential suffering and anticipatory grief. More recently, focus has shifted to examining alternative approaches for intervening to support ADRD caregivers (Cheng et al., 2019; Molero Jurado et al., 2020).

Cognitive behavioral therapy (CBT) is among the class of psychotherapeutic interventions that are grounded in behaviorism (Chen et al., 2015; Gallagher-Thompson et al., 2000; Hepburn et al., 2007; Sepe-Monti et al., 2016; Tremont et al., 2015). CBT interventions emphasize changing or correcting one’s thoughts (e.g., cognitive restructuring) to alleviate suffering. Some CBT based interventions have proven to be moderately effective in reducing ADRD caregiver distress, broadly defined (Kishita et al., 2018; Walter & Pinquart, 2020; Gitlin et al., 2015; Pinquart & Sorensen, 2006; Walter & Pinquart, 2020). CBT is less effective at addressing ADRD caregiver anxiety or suffering (Cheng et al., 2019; Cooper, Balamurali, Selwood, et al., 2007; Gitlin et al., 2015; Olazarán & et al., 2010; Walter & Pinquart, 2020). Given that CBT aims to challenge and modify dysfunctional thoughts, interventions based on this approach may not be effective for anxious ADRD caregivers facing constant, unchangeable experiences during their caregiving journey.

Acceptance and Commitment Therapy (ACT) is part of the larger CBT tradition that includes approaches which share many features (Herbert et al., 2013) and focuses on noticing and accepting one’s thoughts (e.g., mindful awareness) without taking thoughts literally (e.g., cognitive defusion) so that these thoughts do not prevent the person from living a flexible, meaningful life (e.g., values-based action).

ACT is rooted in functional contextualism and (Hayes et al., 2011) incorporates mindfulness and acceptance-based processes along with commitment and behavior change processes to support adaptive coping with life’s realities (Hayes et al., 2012; Twohig & Levin, 2017). ACT is informed by Relational Frame Theory, a theory of language and cognition which explains how one’s thoughts and experiences can become interconnected and influence behavior (Hayes et al, 2012). ACT focuses on accepting one’s thoughts and feelings about an unchangeable situation without getting caught up in them and engaging in behavior aligned with one’s values (Baer, 2010; Wicksell et al., 2010). This psychological flexibility is achieved through six core processes: acceptance, cognitive defusion, mindful awareness of the present moment, self-as-context (perspective taking), values, and committed action (Hayes, et al., 2012). Psychological flexibility is theoretically linked to improvements in anxiety (Bluett et al., 2014), depressive symptoms, and wellbeing (Kohtala et al., 2015).

Mindfulness-based interventions like ACT can teach ADRD caregivers how to observe and accept their thoughts and feelings about their experience without expending energy trying to control or change them. A number of protocols have included mindfulness activities or meditation as all or part of the intervention (Paller et al., 2015; Waelde et al., 2017). A Cochrane Review regarding efficacy of mindfulness-based stress reduction for ADRD family caregivers found minimal and low quality evidence that mindfulness alone reduces short-term anxiety and depressive symptoms in this population (Liu et al., 2018). Meta-analyses and systematic reviews have highlighted the variability in intervention designs and measured outcomes, indicating a potential misalignment (Cooper, Balamurali, Selwood, et al., 2007; Olazarán & et al., 2010; Selwood et al., 2007). Therefore, mindfulness-based techniques combined with other behavioral approaches may facilitate greater improvement in caregiver outcomes, such as mood and quality of life with reductions in anxiety and suffering.

ACT may be ideal for ADRD caregivers with anxiety because unlike CBT or interventions that include only mindfulness practice, ACT aims to support the development of a healthier relationship with unchangeable personal experiences so caregivers can be more mentally flexible and have more meaningful interactions. ADRD caregivers experience excessive distress as a result of the challenging behavioral and psychological symptoms of dementia that the caregiver is often unable to change (Cheng, 2017). Up to 60% of ADRD caregivers without anxiety or depression develop one or both diagnoses within the subsequent two years of caregiving (Joling et al., 2015). This psychological morbidity is strongly associated with maladaptive and less acceptance-based coping (Cooper et al., 2008; Joling et al., 2015; Li et al., 2014a, 2014b). Two European studies have investigated ACT interventions for ADRD caregivers and found improved symptoms of anxiety and depression; however, perceived burden, suffering, and anticipatory grief were not measured (Atefi et al., 2023; Losada et al., 2015; Marquez-Gonzalez et al., 2010). Additionally, these interventions were delivered in-person by doctoral level therapists, making them difficult and expensive to replicate, disseminate, and scale.

In recent years, more studies have addressed the limitations associated with in-person healthcare by testing the implementation of alternative and mixed methods of care delivery. The COVID-19 pandemic highlighted the need to reach people remotely for health care and support. Most studies have shown positive results regarding efficacy and feasibility of the delivery method, whether the intervention was delivered via telephone, videoconference, or software on websites or apps (Atefi et al., 2024; Boyt et al., 2022; Han et al., 2021; Leng et al., 2020; Zhang et al., 2020). Prior to the COVID-19 pandemic, we conducted a single arm pilot of Telephone Acceptance and Commitment Therapy intervention for ADRD caregivers (TACTICs) and the results were favorable, demonstrating significant reductions in caregiver anxiety after the TACTICs intervention (Fowler et al., 2021). These results led us to develop and implement a pilot randomized controlled trial to assess the feasibility, acceptability, and safety of TACTICs and to measure the initial direction of changes in outcome scale scores comparing TACTICs to minimally enhanced usual care (mEUC). An additional goal of this project was also to provide estimates of effects for a larger, fully powered effectiveness trial of ADRD caregivers with anxiety and to determine if an ACT-based intervention could be delivered to ADRD caregivers by phone, and by non-licensed interventionists.

The primary outcome of this pilot randomized controlled trial was ADRD caregivers’ anxiety symptoms, and secondary outcomes included caregiver depressive symptoms, burden, suffering, and anticipatory grief.

MATERIALS AND METHODS

Participants

ADRD caregivers were eligible for TACTICs if they were ≥ 21 years of age, the primary caregiver for a family member with ADRD, able to provide informed consent, intended to continue caregiving for ≥ 12 months, able to communicate in English, willing to attend six weekly intervention sessions on the phone if randomized to the intervention, and had clinically elevated anxiety, defined as Generalized Anxiety Disorder Scale (GAD-7) (Spitzer et al., 2006) score ≥ 10 for the cohort recruited from March-July 2020 to or GAD-7 ≥ 10 or 5–9 with interference in daily activities for the cohort recruited from April-October 2021. This change in eligibility criteria for the second cohort was made to be more inclusive of caregivers with mild anxiety that interfered in their lives. Caregivers were excluded if they were a non-family member or had a diagnosis of ADRD or a serious mental illness (e.g., bipolar disorder or schizophrenia). Non-family members with power of attorney (POA) for the person with ADRD who met all other criteria were also considered eligible. Additionally, caregivers were not enrolled if their family member with ADRD was living in long-term, supportive housing such as assisted living, personal care, or a nursing home.

Procedures

Caregivers were actively recruited at two time points (i.e., two cohorts) during the COVID-19 pandemic from health care settings and the community in Indiana, USA. The first cohort was recruited from March 2020 to July 2020. Given that recruitment started two weeks into the COVID-19 pandemic and anxiety was the primary outcome, we recruited a second group of caregivers from April 2021 to October 2021 with the broadened eligibility criteria of anxiety noted above. In both cohorts, caregivers were recruited from local primary care practices, neurology, and specialty memory clinics from both a safety-net health system that includes federally qualified primary health care centers in Marion County, Indiana and from an academic health system that has specialty care and primary care across the state of Indiana. We also recruited from community sites in central Indiana such as Area Agencies on Aging. While the study did not do widespread, public-facing recruitment, caregivers who learned about the project from word of mouth were able to self-refer and were enrolled if they met eligibility criteria.

The Indiana University institutional review board approved all study procedures. For clinic recruitment, names of potential caregivers were identified through electronic medical record (EMR) review of patients with a diagnosis of ADRD. Letters of invitation were mailed to caregivers approved for contact. Trained research assistants called caregivers within 7–10 days of mailing the letter to screen for interest and eligibility and to obtain informed consent. We also advertised the study via the Indiana Clinical and Translational Science Institute research registry ALL IN and caregiver word of mouth.

Caregivers who expressed interest and were eligible underwent the informed consent process that included an assessment of decisional capacity with four open-ended study-specific teach back questions assessing, in their own words, key aspect of the information presented to them about the research study and the risks and benefits of their participation. Following the informed consent, a baseline assessment was completed, and caregivers were randomly assigned to TACTICs or to a mEUC control condition. See Figure 1 for the CONSORT flow diagram. [Figure 1]

Figure 1.

Study CONSORT

Caregivers completed three follow-up telephone assessments and received $25 gift cards for each assessment. Trained research assistants administering the assessments were blind to study condition. Caregivers randomized to TACTICs had their first intervention session scheduled for 1–3 weeks following the baseline assessment. Sessions were designed to be delivered weekly; however, caregivers had up to 12 weeks to complete all 6 sessions. Additionally, to measure acceptability, caregivers who completed all the TACTICs sessions were encouraged to share their impressions and satisfaction with the intervention qualitatively and quantitatively during the post-interview assessment.

Measures

Primary and secondary outcomes were assessed with validated caregiver-reported measures. All data were collected from caregivers via telephone by trained research assistants and recorded in a secure REDCap (Research Electronic Data Capture) database. Assessments of the TACTICs intervention group were performed at baseline (T1), within 7 days post-intervention (T2), 3 months post-intervention (T3), and 6 months post-intervention (T4). To mirror the same timing of assessments, the mEUC group assessments were performed at baseline (T1), 8 weeks +/− 7 days after baseline (T2), 3 months +/− 14 days after T2, and 6 months +/− 14 days after T2.

The primary outcome of anxiety was measured with the GAD-7 (Kroenke et al., 2007; Spitzer et al., 2006). The GAD-7 consists of seven items that assess the frequency of psychological and physical manifestations of anxiety, such as feeling nervous, not being able to stop or control worrying, becoming easily annoyed or irritable, or exhibiting restless behavior. Scores range from 0 to 21 with 0–4 indicating minimal, 5–9 indicating mild, 10–14 indicating moderate, and 15–21 indicating severe anxiety (Spitzer et al., 2006).

Secondary outcomes included depressive symptoms, caregiver burden, caregiver suffering, and anticipatory grief. For depressive symptoms, the Patient Health Questionnaire–8 (PHQ-8) was used to measure eight items that assess psychological and physical symptoms of depression (Kroenke et al., 2009). Caregiver burden was measured with the Zarit Burden Interview (ZBI), which includes 22 items that measure the objective and subjective burden experienced by family caregivers (Bedard et al., 2000; Hébert et al., 2000; Zarit et al., 1985). Caregiver suffering was assessed with The Experience of Suffering measure that contains 33 items across three subscale indices: physical (9 items), psychological (15 items), and existential (9 items) symptoms of suffering (Schulz et al., 2010). We also measured anticipatory grief, which is the psychological phenomenon wherein feelings of grief are experienced prior to the physical death of a loved one. Commonly experienced by caregivers of those with ADRD, anticipatory grief is independently related to higher levels of anxiety, depression, and burden (Garand et al., 2012; Gómez et al., 2023; Theut et al., 1991). Anticipatory grief was measured using the Anticipatory Grief Scale (AGS), a 27-item measure of grief-related feelings and attitudes of loved ones of those with ADRD using a Likert scale (Theut et al., 1991). All measures have been statistically validated in adult caregiving populations and have demonstrated good internal consistency in prior trials (Bond et al., 2011; Grégoire & et al., 2020; Hébert et al., 2000; Kroenke et al., 2001; Schulz et al., 2010; Spitzer et al., 2006; Theut et al., 1991).

At baseline, caregivers also reported their demographics, including relationship to the person living with ADRD, time spent caregiving, and use of other resources such as support groups and home and community-based services. We also administered the Dementia Severity Rating Scale (Clark & Ewbank, 1996) to assess dementia severity.

Intervention

TACTICs is a telephone-based behavioral intervention designed to reduce anxiety and enhance psychological flexibility through the practice of six core skills: acceptance, cognitive defusion, mindful awareness of the present moment, self-as-context (perspective taking), values clarification, and committed action.

TACTICs was innovatively designed to be delivered completely by telephone by a non-licensed, Bachelor’s-prepared interventionist to specifically address the issue of scalability (Alzheimer’s Association, 2024; Gitlin et al., 2015) of ADRD caregiver interventions. Use of non-licensed personnel could support implementation in a variety of settings.

Of the three study interventionists, one was a research staff member at Regenstrief Institute, and two were enrolled in a clinical psychology PhD program. Each had a bachelor’s degree in psychology. The first trained and primary interventionist (TG) was assigned 41 caregivers, the second interventionist was assigned six caregivers, and the third interventionist was assigned nine caregivers.

Interventionist training included completion of an online 8-hour ACT training course offered through Praxis Continuing Education and Training, assigned readings from an ACT text (Harris, R., 2019), in-depth study of the TACTICs study protocol, intervention manual, and participant notebook, listening to recordings of each of the six TACTICs sessions completed in the non-randomized pilot of TACTICs (Fowler et al., 2021), and role playing each of the six TACTICs sessions and debriefing with a doctoral level clinical health psychologist (SAJ). Throughout the study, interventionists attended weekly supervision sessions with a master’s level ACT clinician. Additionally, all TACTICs sessions were audio-recorded and assessed for fidelity and ACT skill level using a standardized fidelity checklist adopted from our previous ACT pilots (Fowler et al., 2021; Johns et al., 2020). Written feedback for each rated session was provided to interventionists in a timely fashion.

Prior to the first session, caregivers received a TACTICs binder including reading materials, worksheets, and handouts summarizing session topics. The 6-session intervention consisted of 1 hour telephone sessions that included the practice of one or more of the six ACT skills in each session. The interventionist guided caregivers in brief mindfulness meditation practices that encouraged non-judgmental awareness of thoughts, feelings, and bodily sensations in the present moment in each session. Participants were invited to practice mindfulness at home in between sessions using 10-min audio recordings available via a computer download or compact disc, according to each participant’s preference. Caregivers also identified deeply held values to serve as a guide when choosing how to spend limited time or energy and set values-based action goals each week.

Safety of the pilot trial was monitored throughout the study and ended at the conclusion of data collection. Since anxiety and depression were measured as a part of the study outcomes, all participants with PHQ-8 and/or GAD-7 scores ≥ 15 (suggesting moderate to severe levels of depression and anxiety) at any follow-up assessment were contacted by the clinical psychologist on the team (SAJ). During these contacts, SAJ informed participants of the meaning of their scores and offered referrals to and coordination of follow-up psychological care as needed.

Control Condition

Minimally enhanced usual care (mEUC) was chosen to reflect what may be commonly offered to caregivers of people who are diagnosed with ADRD in a clinic setting or what ADRD caregivers may find online if searching for information about caregiving support. It included a mailed packet of educational reading materials and information on how to access caregiver support groups in their area (National Institute on Aging, 2023a, 2023b).

Statistical Analyses

Descriptive statistics for caregivers’ social and demographic characteristics were summarized as frequency and percent for categorical variables, and as means and standard deviations for normal continuous variables, including their relationship to the person with ADRD and the severity of that individual’s ADRD. Feasibility measures included the percent of eligible caregivers who enrolled in the study and the completion rate of the six TACTICs sessions. Acceptability was assessed using satisfaction surveys with quantitative scales and open ended questions (Wagner et al., 2016). Inferential statistics included Student’s t-tests for two group comparisons of continuous outcomes at baseline and at follow-up time points. Paired t-tests were used to analyze the change from baseline to both follow-up time points. Factorial ANOVA models were then used to analyze the effect of both treatment and time, along with the interaction between the two variables.

The standardized response mean (SRM) effect size for the outcomes was calculated to assess the magnitude of intervention effects within-group at each timepoint. To determine SRM, mean change post intervention, 3-month post, and 6-month post relative to baseline was calculated and divided by the standard deviation (SD) of change. The 95% confidence intervals (CIs) were computed for each caregiver (caregiver’s mean change divided by the sample’s SD of change scores). The SAS MEANS procedure with the LCLM and UCLM options were used to compute the lower and upper 95% confidence limits for the SRM statistic.

The primary efficacy-related goal of this pilot was to estimate effect sizes, and the 2-sided paired t test was used to determine significant (P < 0.05) responsive over time. Given that this study was a pilot, marginal significance (0.05 < P < 0.10) was also measured and reported. Cohen’s d values for between-group analysis, were calculated using the difference in the SRMs of both groups divided by the standard deviation. Cohen’s d values of 0.2, 0.5, and 0.8 indicated small, medium, and large effect sizes, respectively (Cohen, 1998). Analyses were performed using SAS version 9.4 (SAS Institute, Inc., Cary, NC) and all analytical assumptions were verified.

RESULT

Social and Demographic Characteristics

Caregivers were 60.7 (SD 11.3) years old on average, with 40.2% providing care for their spouse, and 54.7% providing care for their parent or parent in-law with ADRD. 84% of caregivers were female, with 73.5% identifying as White, 22.2% identifying as Black, and 4.3% identifying as another race. The majority (97.4%) identified as non-Hispanic or Latino. More than half of the caregivers reported having a college degree or more education (59.8%), and their income level as comfortable (63.3%). Caregivers were caring for people with a range of ADRD severity, including mild (29.9%), moderate (51.3%), and severe (18.8%) ADRD. Caregivers reported their own health status as very good or excellent (35%), good (44.4%), and fair or poor (20.5%). Most caregivers (≥ 86%) reported not participating in any community support groups or individual counseling. There are no statistically significant differences between the TACTICs and mEUC groups at baseline. [Table 1.]

Table 1.

Baseline Characteristics of Caregivers by Randomization Group

	Overall N=117	TACTICs n=58	mEUC n=59	P-value
Age in years, mean (SD)	60.7 (11.3)	59.0 (11.0)	62.3 (11.5)	.1169
Self-identified sex, n (%)
Female	98 (83.8)	52 (89.7)	46 (78.0)	.0865
Self-identified race, n (%)
White	86 (73.5)	43 (74.1)	43 (72.9)	>.9999
Black	26 (22.2)	13 (22.4)	13 (22.0)
Other race	5 (4.3)	2 (3.5)	3 (5.1)
Self-identified ethnicity, n (%)
Hispanic or Latino/a	3 (2.6)	2 (3.5)	1 (1.7)	.6185
Non-Hispanic or Latino/a	114 (97.4)	56 (96.6)	58 (98.3)
Highest level of completed education, n (%)
Not a college graduate	47 (40.2)	23 (39.7)	24 (40.7)	.9102
College graduate	70 (59.8)	35 (60.3)	35 (59.3)
Self-reported income status, n (%)
Comfortable	74 (63.3)	36 (62.1)	38 (64.4)	.7974
Have just enough to make ends meet	35 (29.9)	17 (29.3)	18 (30.5)
Do not have enough to make ends meet	8 (6.8)	5 (8.6)	3 (5.1)
Caregiver relationship to care recipient with ADRD, n (%)
Adult child or child-in-law	64 (54.7)	30 (51.7)	34 (57.6)	.2779
Spouse	47 (40.2)	23 (39.7)	24 (40.7)
Other relationship	6 (5.1)	5 (5.6)	1 91.7)
Severity of ADRD of care recipient, * n (%)
Mild	35 (29.9)	14 (24.1)	21 (35.6)	.1126
Moderate	60 (51.3)	29 (50.0)	31 (52.5)
Severe	22 (18.8)	15 (25.9)	7 (11.9)
Caregiver self-reported health status, n (%)
Poor or Fair	24 (20.5)	12 (20.7)	12 (20.3)	.7716
Good	52 (44.4)	24 (41.4)	28 (47.5)
Very good or Excellent	41 (35.0)	22 (37.9)	19 (32.2)

Abbreviations: Alzheimer’s Disease and Related Dementias (ADRD); Telephone Acceptance and Commitment Therapy Intervention for Caregivers (TACTICs); Minimally Enhanced Usual Care (mEUC).

Intervention Effects and Caregiver Outcomes

At baseline, the average GAD-7 score for anxiety was 12.3 (SD 3.8), with 18%, 54.7%, and 27.4% falling in mild, moderate, and severe categories respectively. The average PHQ-8 score was 9.2 (SD 4.8), with 19.7%, 35.0%, 29.9%, 13.7%, and 1.7% falling into minimal, mild, moderate, moderately severe, and severe categories respectively. The average caregiver’s burden, measured by ZBI was 44.1 (SD 15.5), which indicates high levels of caregiver burden. The average anticipatory grief score was 80.9 (SD 14.9), which is higher than reported in other studies of ADRD caregivers (Fowler et al., 2013; Marwit & Meuser, 2005). There were no statistically significant differences between the TACTICs group and mEUC group for any outcome at baseline. Table 2 shows outcomes with mean scores each time point.

Table 2.

Caregiver Outcomes at each time point

Outcome	Time	Overall mean ± SD	TACTICs	mEUC	P-value
Anxiety (GAD-7 score)	T1	12.3 ± 3.8	12.8 ± 3.8	11.8 ± 3.9	.1616
	T2	8.2 ± 5.0	7.6 ± 4.9	8.6 ± 5.1	.3699
	T3	7.8 ± 4.9	7.5 ± 4.3	8.0 ± 5.3	.6360
	T4	7.8 ± 5.5	8.7 ± 5.6	7.1 ± 5.4	.2026
Depressive symptoms (PHQ-8 score)	T1	9.2 ± 4.8	9.3 ± 5.2	9.0 ± 4.4	.7280
	T2	7.4 ± 5.1	7.4 ± 5.4	7.4 ± 5.0	.9466
	T3	7.1 ± 4.6	7.3 ± 4.7	6.9 ± 4.6	.7296
	T4	6.7 ± 4.7	7.0 ± 4.9	6.4 ± 4.7	.5958
Caregiver burden (ZBI score)	T1	44.1 ± 15.5	44.3 ± 12.5	44.0 ± 18.1	.9284
	T2	38.9 ± 13.9	39.8 ± 14.7	38.3 ± 13.4	.6205
	T3	37.3 ± 14.5	37.8 ± 14.6	37.0 ± 14.6	.8242
	T4	36.3 ± 15.4	37.7 ± 16.0	35.2 ± 15.1	.4744
Caregiver suffering (Experience of Suffering measure)
Physical symptoms index sub score	T1	7.1 ± 3.5	6.7 ± 3.2	7.5 ± 3.8	.2111
	T2	6.9 ± 4.2	5.9 ± 3.6	7.5 ± 4.4	.0772
	T3	7.4 ± 4.5	6.8 ± 3.7	7.9 ± 5.0	.2872
	T4	6.5 ± 4.3	6.6 ± 3.7	6.5 ± 4.8	.9033
Psychological symptoms index sub score	T1	15.2 ± 6.1	15.7 ± 7.0	14.8 ± 5.1	.4404
	T2	13.7 ± 6.3	13.5 ± 6.2	13.9 ± 6.4	.7753
	T3	13.6 ± 6.6	14.5 ± 6.3	13.0 ± 6.8	.3090
	T4	13.0 ± 6.7	14.3 ± 6.8	12.1 ± 6.6	.1661
Existential symptoms index sub score	T1	15.7 ± 3.1	16.0 ± 2.8	15.4 ± 3.3	.2874
	T2	16.3 ± 3.2	16.5 ± 3.0	16.2 ± 3.3	.5955
	T3	16.0 ± 3.0	16.1 ± 2.8	16.0 ± 3.1	.8462
	T4	16.0 ± 3.0	16.6 ± 2.4	15.6 ± 3.4	.1985
Anticipatory grief (Anticipatory Grief Scale score)	T1	80.9 ± 14.9	79.8 ± 10.8	82.0 ± 18.1	.4205
	T2	76.3 ± 11.9	74.1 ± 10.3	78.0 ± 12.7	.1232
	T3	74.0 ± 12.0	73.2 ± 9.9	74.5 ± 13.2	.6367
	T4	73.8 ± 13.3	72.6 ± 12.5	74.6 ± 14.0	.5247

Abbreviations: Generalized Anxiety Disorder Scale-7 (GAD-7); Patient Health Questionnaire–8 (PHQ-8); Zarit Burden Interview (ZBI); Telephone Acceptance and Commitment Therapy Intervention for Caregivers (TACTICs); Minimally Enhanced Usual Care (mEUC)

Notes: T1= baseline assessment; T2= within 7 days of last intervention visit; T3= 3 months post intervention; T4= 6 months post intervention.

At T2, caregivers in the TACTICs group showed a moderate reduction in anxiety symptoms (SRM 4.7, 95% CI [3.2,6.2]), which was maintained to T3, and reduced slightly at T4 (SRM 4.1, 95% CI [2.3,5.8]). These reductions are not considered statistically significant because the mEUC group also demonstrated a reduction in anxiety symptoms at T2 (SRM 3.3, 95% CI [2.0,4.6]) and at T4 (SRM 4.1, 95% CI [2.4,5.8]). These reductions, however, are clinically significant because the TACTICs group baseline score (12.8 SD 3.8) categorizes the group as experiencing “moderate” anxiety, but their T4 score (8.7 SD 5.6) categorizes the group as experiencing “mild” anxiety. [Table 2]

Caregivers in the TACTICs group showed a mild reduction in depressive symptoms T2, T3, and T4 (SRM 2.3, 2.5, 3.1, 95% CI [0.8,3.8], [0.5,4.6], [1.3,5.0]). Caregivers in the TACTICs group showed a mild reduction in levels of caregiver burden at T2, T3, and T4 (SRM 5.3, 7.0, 6.3, 95% CI [2.2,8.4], [2.9,11.0], [2.7,9.9]). Caregivers in the TACTICs group showed a mild reduction of anticipatory grief at T2, T3, and T4 (SRM 1.3, 1.5, 3.9, 95% CI [−1.2,3.8], [−1.5,4.4], [1.2,6.7]). The mEUC group also demonstrated reductions in depressive symptoms, caregiver burden, and anticipatory grief. The reductions in the TACTICs group were not found to be statistically significant compared to the mEUC for depression (T4 p=0.4955), burden (T4 p=0.6169), or anticipatory grief (T4 p=0.9888).

The suffering subscales followed the same pattern, however, there was a significant difference for the psychological symptoms of suffering subscale between the TACTICs group (SRM 3.4, 95% CI [2.1,4.8]) and mEUC group (SRM 0.6, 95% CI [−0.9,2.2]) at T2. The d= 0.55, indicating a medium effect size, p= 0.0118. The significant difference diminished at T3 and T4. Table 3 shows within-group and between-group differences at each time point. [Table 3]

Table 3.

Within-Group and Between-Group Changes in Caregiver Outcomes from baseline

Outcome	Time	TACTICs SRM (95% CI)	mEUC SRM (95% CI)	TACTICs vs mEUC Cohen d	TACTICs vs mEUC P-value
Anxiety (GAD-7 score)	T1 - T2	4.7 (3.2, 6.2)	3.3 (2.0, 4.6)	0.31	.1465
	T1 - T3	4.7 (3.0, 6.5)	3.4 (1.8, 5.1)	0.23	.3087
	T1 - T4	4.1 (2.3, 5.8)	4.1 (2.4, 5.8)	−0.01	.9789
Depressive symptoms (PHQ-8 score)	T1 - T2	2.3 (0.8, 3.8)	1.5 (0.4, 2.5)	0.20	.3576
	T1 - T3	2.5 (0.5, 4.6)	2.1 (0.8, 3.3)	0.09	.6809
	T1 - T4	3.1 (1.3, 5.0)	2.3 (0.8, 3.9)	0.15	.4955
Caregiver burden (ZBI score)	T1 - T2	5.3 (2.2, 8.4)	4.8 (0.2, 9.3)	0.04	.8522
	T1 - T3	7.0 (2.9, 11.0)	6.8 (2.1, 11.4)	0.01	.9535
	T1 - T4	6.3 (2.7, 9.9)	7.8 (2.9, 12.8)	−0.11	.6169
Caregiver suffering (Experience of Suffering measure)
Physical symptoms index sub score	T1 - T2	0.7 (−0.4, 1.7)	−0.5 (−1.4, 0.4)	0.36	.0909
	T1 - T3	−0.1 (−1.2, 1.0)	−0.7 (−1.8, 0.4)	0.18	.4360
	T1 - T4	0.1 (−1.0, 1.1)	0.3 (−0.9, 1.5)	−0.06	.7978
Psychological symptoms index sub score	T1 - T2	3.4 (2.1, 4.8)	0.6 (−0.9, 2.2)	0.55	.0118
	T1 - T3	2.4 (0.7, 4.1)	1.7 (0.1,3.3)	0.14	.5473
	T1 - T4	2.2 (0.1, 4.3)	2.0 (0.2,3.8)	0.03	.8858
Existential symptoms index sub score	T1 - T2	−0.9 (−1.8, 0.1)	−0.7 (−1.5, 0.2)	−0.07	.7282
	T1 - T3	−0.6 (−1.7, 0.4)	−0.3 (−1.3, 0.7)	−0.10	.6694
	T1 - T4	−0.6 (−1.6, 0.4)	−0.1 (−1.0, 0.9)	−0.18	.4468
Anticipatory grief (Anticipatory Grief Scale score)	T1 - T2	5.8 (3.5, 8.2)	3.4 (−1.4, 8.2)	0.17	.3694
	T1 - T3	6.2 (2.6, 9.8)	7.1 (2.3, 11.9)	−0.06	.7673
	T1 - T4	6.1 (2.2,9.9)	6.1 (0.6, 11.6)	−0.01	.9888

Abbreviations: Generalized Anxiety Disorder Scale-7 (GAD-7); Patient Health Questionnaire–8 (PHQ-8); Zarit Burden Interview (ZBI); Telephone Acceptance and Commitment Therapy Intervention for Caregivers (TACTICs); Minimally Enhanced Usual Care (mEUC); Standard Response Mean (SRM); Confidence Interval (CI)

Notes; T1= baseline assessment; T2= within 7 days of last intervention visit; T3= 3 months post intervention; T4= 6 months post intervention.

Repeated measures analyses are depicted in Tables 4 and 5. Table 4 shows the results of an “intent to treat” analysis and Table 5 is a sensitivity analysis comparing outcomes from the only participants who completed all six TACTICs sessions compared to the mEUC control group. We did find a statistically significant effect of time on anxiety, depressive symptoms, and caregiver burden scores between the caregivers in the TACTICs group vs mEUC group, but not a time and intervention effect. [Table 4] Given that 65.5% (n=38) of the caregivers completed all six sessions of TACTICs, we conducted a sensitivity analysis to measure any differences in trends for those who received the full “dose” of TACTICs compared to the mEUC. Results show that there are significant trends over time for anxiety, caregiver burden, and anticipatory grief scores, but there is no difference in the trend over time as the interaction effect is not significant. [Table 4 and Table 5]

Table 4.

Intent to treat repeated measures analysis

Outcome	Effect	F	P-value
Anxiety (GAD-7 score)	Treatment (TACTICs vs. mEUC)	0.29	.5907
	Time	22.64	<.0001
	Treatment * time	1.40	.2411
Depressive symptoms (PHQ-8 score)	Treatment (TACTICs vs. mEUC)	0.32	.5710
	Time	5.27	.0014
	Treatment * time	0.06	.9784
Caregiver Burden (ZBI score)	Treatment (TACTICs vs. mEUC)	0.61	.4343
	Time	5.31	.0014
	Treatment * time	0.10	.9612
Caregiver suffering (Experience of Suffering measure)
Physical symptoms index sub score	Treatment (TACTICs vs. mEUC)	3.66	.0566
	Time	0.58	.6296
	Treatment * time	0.60	.6154
Psychological symptoms index sub score	Treatment (TACTICs vs. mEUC)	2.34	.1266
	Time	1.85	.1376
	Treatment * time	0.60	.6122
Existential symptoms index sub score	Treatment (TACTICs vs. mEUC)	2.32	.1282
	Time	0.84	.4736
	Treatment * time	0.23	.8730
Anticipatory grief (Anticipatory Grief Scale score)	Treatment (TACTICs vs. mEUC)	2.73	.0995
	Time	6.51	.0003
	Treatment * time	0.15	.9308

Abbreviations: Generalized Anxiety Disorder Scale-7 (GAD-7); Patient Health Questionnaire–8 (PHQ-8); Zarit Burden Interview (ZBI); Telephone Acceptance and Commitment Therapy Intervention for Caregivers (TACTICs); Minimally Enhanced Usual Care (mEUC)

Table 5.

Repeated measures analysis comparing TACTICs caregivers who completed the intervention (n=38) vs mEUC group

Outcome	Effect	F	P-value
Anxiety (GAD-7 score)	Treatment (completed TACTICs vs. mEUC)	0.29	.5907
	Time	22.64	<.0001
	Treatment * time	1.40	.241
Depressive symptoms (PHQ-8 score)	Treatment (completed TACTICs vs. mEUC)	0.32	.5710
	Time	5.27	.0014
	Treatment * time	0.06	.9784
Caregiver Burden (ZBI score)	Treatment (completed TACTICs vs. mEUC)	0.61	.4343
	Time	5.31	.0014
	Treatment * time	0.10	.9612
Caregiver Suffering
Physical symptoms index sub score	Treatment (completed TACTICs vs. mEUC)	3.66	.0566
	Time	0.58	.6296
	Treatment * time	0.60	.6154
Psychological symptoms index sub score	Treatment (completed TACTICs vs. mEUC)	2.34	.1266
	Time	1.85	.1376
	Treatment * time	0.60	.6122
Existential symptoms index sub score	Treatment (completed TACTICs vs. mEUC)	2.32	.1282
	Time	0.84	.4736
	Treatment * time	0.23	.8730
Anticipatory grief (Anticipatory Grief Scale score)	Treatment (completed TACTICs vs. mEUC)	2.73	.0995
	Time	6.51	.0003
	Treatment * time	0.15	.9308

Abbreviations: Generalized Anxiety Disorder Scale-7 (GAD-7); Patient Health Questionnaire–8 (PHQ-8); Zarit Burden Interview (ZBI); Telephone Acceptance and Commitment Therapy Intervention for Caregivers (TACTICs); Minimally Enhanced Usual Care (mEUC)

Feasibility

A priori feasibility benchmarks for TACTICs included at least 50% of eligible caregivers enrolling in the study, and among caregiver randomized into the TACTICs group, ≥ 70% would complete all six TACTICs sessions. We achieved a 96% recruitment rate, with 120 of 125 eligible caregivers were enrolled. Allocation was 1:1, with 58 enrolled in TACTICs and 59 enrolled in the mEUC group. Of the 58, 65.5% (n=38) completed all six TACTICs sessions. Figure 1. While the rate of enrollment exceeded our threshold during the COVID-19 pandemic, the rate of intervention completion was lower than anticipated. The participants who did not complete the intervention were given the opportunity to provide reasons for non-completion, with the most common being life events such as a death in the family and lack of time.

Fidelity ratings of TACTICs sessions yielded 98.9% (SD=3.10) and 98.4% (SD=3.75) adherence for the first and second cohorts, respectively, suggesting the interventionists delivered TACTICs with high adherence to the intervention manual. Results not shown.

Acceptability

The a priori benchmark for acceptability was that at least 70% of caregivers randomized to TACTICs would report a satisfaction of ≥ 8 via 10-point scale. The mean satisfaction score was 9.35 (SD 0.91) out of 10, indicating very high satisfaction with the intervention, as demonstrated in Table 4. On average, caregivers also indicated that they would be “quite a bit” or “very much” confident that they would recommend the program to others, and “likely” or “extremely likely” to continue using what they learned in the intervention. Results not shown.

In addition to the quantitative measures, participants completed a satisfaction questionnaire at post-intervention. Most comments on the questionnaire were positive, with one caregiver noting, “The most memorable and helpful part of the program was the opportunity to share how I was feeling as a caregiver, and as someone dealing with the changes the virus was making in our lives.” Another caregiver stated, “It gave me confidence and tools to have time to take care of myself. I enjoyed the meditation, which I continue doing. The workbook made it very easy and helped organize the material.” We also received valuable comments about improving the delivery of TACTICs, such as this quote from a caregiver: “Older adults need an easier way to access the audio relaxation exercises. I struggled each day trying to find it in my email, go to the site, and pull it up. It was a hassle. There may be an easier way to access, but don’t assume older adults know how to do those computer tech things.”

DISCUSSION

The pilot results demonstrate that the TACTICs intervention was feasible and acceptable according to the benchmarks outlined. Although pilot studies are primarily designed to examine feasibility and implementation, preliminary outcomes and effects were measured and were not as strong as expected based on the results of the initial pilot study (Fowler et al., 2021). No outcomes consistently showed statistical significance or lasting differences for TACTICs participants compared to the mEUC controls, however, the TACTICs group saw reductions in anxiety, depressive symptoms, caregiver burden, anticipatory grief, and suffering. The effects were simply not strong enough to defeat the placebo effect noted in the mEUC group, but that does not negate impact of the reductions themselves. It is additionally reassuring that the participants reported such a high satisfaction with the intervention and intend to use the skills they learned moving forward. The results, including the observed trends, will be helpful to consider when evaluating plans for future research with larger samples.

Given the mixed results, it is difficult to pinpoint exactly why the outcomes were not as strong as anticipated, as there are several potential factors. The first to consider is the content of the ACT intervention administered; the intervention manual may need more specialized applications or experiential exercises to be more effective in this population of ADRD caregivers. Additionally, the dosage of the intervention may need to be extended from a 6-week protocol to a longer more robust duration along with booster sessions (e.g., 8 to 12 weeks). Given the daily challenges caregivers experience, more time may be needed to engage and fully meet the needs of caregivers. For example, Losada and colleagues (2015) implemented an 8-week ACT protocol that found significant reductions in depressive symptoms and, more recently, implemented a series of 3 booster sessions that followed an 8-week protocol (Márquez-González et. Al., 2020). Similarly, Han and colleagues (2021) conducted a small pilot study that implemented a 10-session ACT protocol.

Another factor to consider is the use of Bachelor’s-level interventionists. Interventionists with higher levels of training may have improved the protocol’s effectiveness, however the fidelity data for the Bachelor’s-level interventionists was impressive and thus a less likely source. Specifically, this success may have stemmed from the rigorous training and oversight of interventionists implementing ACT (i.e., didactic training, readings, demonstrations, role playing, on-going supervision, and fidelity monitoring). It also is worth noting that the interventionists did not have prior training or experience in working with distressed caregivers of persons living with dementia and, anecdotally, interventionists were considered highly promising Bachelor’s level individuals who were viewed as kind and compassionate. Last, our sample size and inclusion of individuals with mild anxiety could have lessened effects, and the sample size and inclusion criteria could be addressed in future projects.

An important consideration when interpreting these study results is the potential effects of the COVID-19 pandemic on the mental state of participants. Both cohorts participated during a different phase of the pandemic, with the first trial running May 2020 to March 2021, and the second running June 2021 to June 2022. The NIH Coronavirus Health Impact Survey (CRISIS) was administered to participants in the second cohort (only) to test for pandemic effects, which found a significant relationship between COVID-19 anxiety and caregiver burden overall(Lucas et al., 2021). While COVID-19 should not be considered a confounding variable since it likely impacted the ACT and mEUC groups similarly, the environmental context of the pandemic is important to acknowledge as it may also have significantly contributed to the < 70% completed rate in the TACTICs arm.

Additionally, we must contemplate how the next wave of these types of caregiver interventions should be further developed and studied. As noted earlier, current evidence-based protocols typically address different elements of the caregiving experience and may not benefit all caregivers. As such, it is imperative that novel approaches to address the unchangeable aspects of ADRD caregiving are explored. Compared to other protocols, it may be that ACT- based interventions like TACTICs require caregivers to employ a different level of engagement and participation that necessitates greater caregiver readiness, and ACT-based protocols may serve as a complimentary approach to traditional programs.

Learning how to accept and skillfully navigate unchangeable aspects of providing dementia care may result in benefits not assessed in the current study. For example, it may be that ACT-based interventions target different caregiver outcomes, such as compassion, emotion regulation, caregiver stress and strains, quality of life, personal gain, and purpose which were not assessed in the current study. For example, Fauth and colleagues (2022) found decreased caregiver stress due to behavioral symptoms along with improvements in positive aspects of caregiving and quality of life. These types of measures may better reflect the caregiving experience in learning how to manage difficult yet unchangeable aspects of one’s caregiving roles and experiences. As such, future research should include a more diverse set of both negative and positive caregiver outcomes in order to determine the relevant targets of ACT.

It is important to consider the underlying mechanism of psychological flexibility and how this may inform future studies. Morimoto (2024) identified four separate caregiver profiles that were related to depressive symptoms, anxiety symptoms, and work-family conflict. Profiles included: high psychological flexibility; moderate psychological flexibility with high commitment; moderate psychological flexibility with low commitment, and low psychological commitment. Caregivers with high psychological flexibility or moderate psychological flexibility high commitment fared better. These findings suggest psychological flexibility may be an important mediator to include in subsequent analyses for more precisely examining the impact of ACT on outcomes. Caregivers also could be screened based on psychological flexibility to determine who might benefit the most from ACT and who may need additional support and resources. It also may be that one of the 6 underlying core components of psychological flexibility (i.e., acceptance, defusion, present-moment awareness, self-as-context, values clarity, and values-aligned committed action) are more relevant to a dementia caregiving population. Further research is needed to explore and delineate the underlying mechanism of psychological flexibility or one of the core components as the key mechanism.

Overall, research has found support for the acceptability and feasibility of ACT protocols, yet questions remain pertaining to the appropriate intervention does and length of time needed for implementation along with the appropriate measures for demonstrating efficacy. Future work also should continue to examine the underlying mechanism of ACT (i.e., psychological flexibility or a specific component). This inquiry will provide the needed theoretical information for guiding implementation and evaluation of ACT as a viable intervention approach for dementia caregivers.

Data Availability Statement

The data presented in this manuscript is available upon request from Dr. Nicole R. Fowler at fowlern@iu.edu. Detailed ACT session descriptions are available from Dr. Shelley A. Johns at sheljohn@iu.edu.