Skip to content
ALL Metrics
-
Views
Get PDF
Get XML
Cite
Export
Track
Research Article

Accepting the unacceptable? Exploring how acceptance relates to quality of life and death anxiety in a cancer population

[version 1; peer review: awaiting peer review]
Lucinda Brabbins, Nima Moghaddam
https://orcid.org/0000-0002-8657-4341
David Dawson
Lucinda Brabbins, Nima Moghaddam
https://orcid.org/0000-0002-8657-4341
David Dawson
PUBLISHED 14 Apr 2020
Author details Author details
OPEN PEER REVIEW
REVIEWER STATUS AWAITING PEER REVIEW

This article is included in the Healthy Lives gateway.

Abstract

Background: Quality of life is a core concern for cancer patients, which can be negatively affected by illness-related death anxiety; yet understanding of how to appropriately target psycho-oncological interventions remains lacking. We aimed to explore experiential acceptance in cancer patients, and whether acceptance – as an alternative to avoidant coping – was related to and predictive of better quality of life and death anxiety outcomes.
Methods: We used a longitudinal, quantitative design with a follow-up after three months. Seventy-two participants completed a questionnaire-battery measuring illness appraisals, acceptance and non-acceptance coping-styles, quality of life, and death anxiety; 31 participants repeated the battery after three months.
Results: Acceptance was an independent explanatory and predictive variable for quality of life and death anxiety, in the direction of psychological health. Acceptance had greater explanatory power for outcomes than either cancer appraisals or avoidant response styles. Avoidant response styles were associated with greater death anxiety and poorer quality of life.
Conclusions: The findings support the role of an accepting response-style in favourable psychological outcomes, identifying a possible target for future psychological intervention. Response styles that might be encouraged in other therapies, such as active coping, planning, and positive reframing, were not associated with beneficial outcomes.

Keywords

ACT, Acceptance, Death, Cancer, Quality of Life, Death anxiety

Cancer and experiential distress

In the UK, an estimated 50% of people born after 1960 will experience cancer in their lifetime (Cancer Research UK, 2014). During phases of diagnosis, treatment, and remission, up to 75% of cancer patients experience anxiety (Cardy et al., 2006), reduced quality of life (QoL) (Ciarrochi et al., 2011), and heightened levels of grief, pain, fatigue, and depression (Barraclough, 1999). Many cancer patients concomitantly report intrusive thoughts and worry in relation to their survival, and preoccupation with existential thoughts and death anxiety (Adelbratt & Strang, 2000).

Death anxiety is described as “an emotional reaction involving subjective feelings of unpleasantness and concern”, evoked by “the anticipation of a state in which the self does not exist” (Hoelter, 1979; Tomer & Eliason, 1996), and correlates with general anxiety, depression, and perceived shortened life-expectancy. Advanced cancer patients may experience different appraisal and response styles (Rinaldis et al., 2009), and higher rates of death anxiety and negative wellbeing (Adelbratt & Strang, 2000; Neel et al., 2015; Vodermaier et al., 2011). Patients with aggressive or late-stage pancreatic, lung, and prostate cancers report greater levels of existential distress and suicidal risk (Bill-Axelson et al., 2010; Zabora et al., 2001), as the subjective and objective threat to life increases. It is logical to suggest that death anxiety and negative quality of life may therefore augment in line with worsening prognosis and perceived severity of the disease (Gao et al., 2010; Rinaldis et al., 2009), and that increasingly-proximal threats to life may affect appraisal, response styles, and wellbeing. It is also clear that not everyone with cancer experiences acute distress, suggesting there are probable mechanisms in action, which warrant further investigation.

Terror Management Theory, acceptance, and response styles in cancer

Terror Management Theory (TMT) (Greenberg et al., 1986) suggests that death anxiety is innate and universal, with the degree of death awareness varying between individuals; potentially being higher in those whose lives are more threatened by disease. TMT suggests we must stave off the conscious ‘terror of death’ through proximal defensive mechanisms of thought suppression, denial, and distraction (Greenberg et al., 1986), and that such avoidant defences must be strengthened to alleviate anxiety and promote quality of life (Mosher & Danoff-Burg, 2007). TMT posits that, when confronted by an overt mortality-threat (as when diagnosed and living with cancer), individuals who can avoid or control unwanted internal experiences (threatening thoughts, feelings, and sensations) will have more favourable outcomes.

However, emerging evidence for the role of acceptance in physical health populations supports an alternative model of psychological functioning and intervention. Acceptance and Commitment Therapy (ACT) (Hayes et al., 2011) suggests that efforts to avoid and control inner experiences come at a high cost to the individual, and do not provide a successful or protective long-term solution. For example, attempts to suppress thoughts have been shown to heighten their frequency and emotional salience (Hayes et al., 2011), such that problematic avoidance behaviours inadvertently increase in an effort to escape increasing numbers of anxiety-eliciting related stimuli (Solomon et al., 2015).

Although empirical evidence is mixed, avoidant responses – including denial, disengagement, self-blame, and emotion-focused control – have been found to be associated with anxiety, depression, lower treatment compliance, and poorer QoL in cancer patients (Carver et al., 1993; Hulbert-Williams et al., 2015; Nipp et al., 2016; Šoštarič & Šprah, 2004; Stanton et al., 2000). Conversely, open responses, which facilitate the expression of affect, have been linked to better adjustment and QoL in cancer (Stanton et al., 2000). Furthermore, research has implicated the role of illness appraisals (i.e., cognitive appraisals of how illness will affect the patient’s life) in determining quality of life, yet attempts to directly modify these appraisals, as might be applied in traditional Cognitive Behaviour Therapy (CBT), encounter the same limitations (and ironic effects) as thought-suppression (Vilardaga et al., 2013). This provides scope to investigate whether cancer- and death-related thoughts and anxieties can be safely accepted and experienced, rather than avoided.

Accepting unwanted experiences in cancer

ACT seeks to change relationships to psychological events, rather than to directly lessen, control, or alter the events themselves (Hayes et al., 2006; Hayes et al., 2011). In contrast to TMT, ACT would therefore suggest that attempts to avoid and control painful cancer-related inner experiences inadvertently create a state of suffering and unhelpful ‘experiential avoidance’. Acceptance here is a willingness to allow all thoughts and feelings to occur without judgement or avoidance, and is therefore the opposite of the experiential avoidance theorised to underpin poor psychological health, anxiety, and depression (Kashdan & Rottenberg, 2010), and is also diametrically opposed to the response styles advocated by TMT. ACT is one of several contemporary cognitive and behavioural intervention-models that promote open (versus avoidant) responding to difficult experiences (such as death anxiety) – positing that such responses may be more adaptive (Hayes et al., 2011).

Present study

ACT is effective at improving QoL in populations with health conditions such as chronic pain, diabetes and HIV (e.g. A-Tjak et al., 2014). Although research into ACT processes and cancer is inchoate, initial findings are promising (Hulbert-Williams et al., 2015; Low et al., 2016; Tauber et al., 2019) and warrant further analytical study of the putative relationship between experiential acceptance and better psychological outcomes. Given that contrasting predictions may be made from a TMT perspective (which would arguably favour avoidance- over acceptance-based coping) empirical evidence is needed to resolve theoretical differences (with important implications for practice). A need for longitudinal and processual studies on acceptance/ACT processes in cancer has been identified (Dunne et al., 2017; Rand et al., 2012). This study therefore aimed to investigate concurrent and temporal relationships between acceptance, QoL, and death anxiety in an inclusive sample of cancer patients.

Methods

Aims and design

The primary aim of this study was (1) to examine whether acceptance was concurrently related to better quality of life and death anxiety outcomes in cancer patients. Secondary aims were to examine: (2) whether/how illness attributions and alternative (non-acceptance) response-styles concurrently related to QoL and death anxiety outcomes, and (3) any temporal relationships between acceptance and outcomes, over a three-month follow-up period.

The study used a longitudinal, quantitative design. At two timepoints, participants completed several demographic and clinical questions, and a battery of standardised questionnaires, which measured cancer appraisal, response styles, and both death anxiety and QoL.

Ethics and consents

The study was granted ethical approval by the National Health Service (NHS) East Midlands Research Ethics Committee (REC reference 14/EM/1224), with governance approved locally by four participating NHS trusts.

Participants and procedure

A priori power calculation in G*Power software (3.1) estimated that (for 80% power at an alpha of 0.05) at least 32 participants were needed to detect the expected medium-sized correlation between acceptance and QoL (based on a correlation of r = 0.47, observed in previous research; Hulbert-Williams et al., 2015) – i.e., an n of 32 was required to address the primary study aim. Moreover, an n ≥ 51 would provide enough (80%) power to detect similar (or larger) effect sizes in regression analyses incorporating up to 5 potential explanatory variables. Over-recruiting was desirable, to allow for attrition and to power any later multivariate analyses. At the first timepoint (time one of two) 74 participants were recruited, but two participants’ data sets were excluded due to having missing data on over 20% of the questionnaire; a threshold utilised elsewhere in research (e.g. Gillanders et al., 2015). Pairwise deletion was used appropriately for the remaining items with missing data (Tabachnick & Fidell, 2007), as the relatively few missing data (7.4%) was shown to be concentrated across a few variables and missing at random, as established by Little’s test (x2 = 321.54, df = 283, p = 0.057).

Participants with cancer were invited to the study via one of two recruitment streams. Recruitment for the study took place between September 2015 and February 2016. Participants were either identified and approached by NHS clinical gatekeepers, or self-selected via advertisements placed online, using social media platforms and websites, such as Facebook, Twitter, and Cancer Chat. Clinical gatekeepers were recruited from hospitals and hospices, from where they identified potential participants and distributed questionnaire packs. Packs contained participant information sheets, consent forms, questionnaire batteries, and return envelopes. Online participants accessed the same information via an online survey programme. Participants were eligible if they were over 18, had experience of cancer, and lived in the UK. 46% of total participants completed paper questionnaires, and 54% completed online questionnaires. All participants were invited to take part in a follow-up questionnaire, conducted online, which took place three months after initial participation. The overall sample size was n = 72 for time one, and n = 31 for the follow-up.

Measures

A questionnaire battery was administered to measure three domains: background variables (demographics, cancer characteristics, and illness appraisals); response style variables (responses to appraisals); and outcome variables (QoL and death anxiety) (Table 1). The following standardised measures of appraisal, response styles, and outcomes were issued:

Table 1. Theoretically informed measurement framework and the measures used in each category.

123
Theoretical categoryBackground variablesResponse style variablesOutcome variables
Conceptual targetsIndividual and clinical characteristics;
cancer-related appraisals
Response-focused measures; ways
of responding to cancer-related
appraisals
Psychological
outcome measures
Measurement targetsDemographic and clinical variables;
cancer characteristics and beliefs
Acceptance and alternative coping/
response styles
Quality of life and
death anxiety
Measures usedBrief IPQ
Age
Education
Religion/spirituality
Psychological support
Bereavement
Cancer site
Cancer stage
No. previous cancers
AAQ-II
Brief COPE
FACT-G
DAS

Note. AAQ-II = Acceptance and Action Questionnaire II; DAS = Death Anxiety Scale; FACT-G = Functional Assessment of Cancer Therapy – General; Brief IPQ = Brief Illness Perceptions Questionnaire; Brief COPE = Brief Coping Orientation to Problems Experienced

Appraisal measure

Brief Illness Perception Questionnaire (Brief IPQ). The Brief IPQ (Broadbent et al., 2006) is a nine-item questionnaire assessing the cognitive and emotional representations of illness. It is quick to administer and therefore suitable for populations who may feel unwell (Ng, 2012), and has good test-retest and discriminant reliability, and predictive and discriminant validity (Van Oort et al., 2011).

Scale items were made specific to a cancer population, e.g. by substituting the wording of ‘illness’ for ‘cancer,’ and to improve the comprehensibility of items 3, 7, 8, and 9, as recommended (Broadbent et al., 2015). The anchoring system was also adapted to five-point Likert scale from the IPQ-Revised (using response anchors of ‘strongly disagree’ and ‘strongly agree’), as the original 0–11 scaling has been shown to be unsuitable for those with short-term prognoses (Price et al., 2012). Higher Brief IPQ scores reflect a more threatening view of the illness.

Response style measures

Acceptance and Action Questionnaire II (AAQ-II). The AAQ-II (Bond et al., 2011) is a seven-item measure of experiential acceptance, and in this study was scored in the direction of higher scores indicating greater experiential acceptance (versus experiential avoidance). Use of the AAQ-II is well-established in the ACT literature, and the measure has demonstrated satisfactory reliability and validity in general research and with cancer populations (Bond et al., 2011; Feros et al., 2013).

Brief COPE. The Brief COPE (Carver, 1997) measures coping (response) styles across 28 items on a 4-point Likert scale; instructions were adapted to focus on coping in the context of cancer (“the ways you’ve been coping with the stress in your life since you found out you had cancer”). Factor analysis indicates that the various response styles reflect two core factors (Eisenberg et al., 2012): (1) avoidant coping (comprising self-distraction, denial, substance use, behavioural disengagement, venting, and self-blaming) and (2) approach coping (comprising active coping, use of emotional and instrumental support, positive reframing, planning, and [passive, resigned] acceptance [distinct from the active, willing acceptance targeted by the AAQ-II]). In the current study, the Brief COPE was scored accordingly, deriving two summary scores ([1] avoidant coping; [2] approach coping), with higher scores reflecting greater use of the respective class of responses. The Brief COPE has been used in cancer populations of varying cancer sites and stages, and has adequate validity and reliability for cancer populations (Yusoff et al., 2010).

Outcome measures

Functional Assessment of Cancer Therapy - General (FACT-G). The FACT-G (Cella et al., 1993) is a widely-used measure of cancer-related QoL (Ciarrochi et al., 2011) across four domains on a Likert scale: emotional, functional, physical, and social. It has total score good reliability and validity; established across cancer subtypes (Webster et al., 2003), and through correlations with mood, anxiety, and other health-related QoL measures (Luckett et al., 2011). Higher scores suggest better cancer-related QoL.

Death Anxiety Scale (DAS) The DAS (Templer, 1970) uses a fixed choice, true/false format to assess attitudes towards death on 15 items. The DAS is a brief measure, which has internal validity, test-retest reliability (Templer, 1970), and remains the most widely used measure of death anxiety. The DAS has been used in both palliative and non-palliative cancer populations (e.g. Gonen et al., 2012; Royal & Elahi, 2011), and has also been validated internationally in non-health populations (Sharif Nia et al., 2014). Higher scores indicate increased death anxiety, with a cut-off score of 7 out of a possible 15.

Data analyses

Time One (T1). Preliminary analyses allowed for exploration and assumption checks to be carried out on the data, using IBM SPSS Statistics (version 22.0). Correlation analyses (Pearson’s r) were carried out to examine any zero-order relations among cancer stage, illness appraisal, response styles, and focal outcome variables (QoL and death anxiety).

Research aims 1 and 2 were met using hierarchical multiple regression analyses. Five models were run – one for each outcome/dependent variable of interest ([1] social, [2] emotional, [3] functional, and [4] physical QoL; [5] death anxiety). Acceptance (AAQ-II) and appraisal (IPQ) scores were included as a priori predictors of interest in all regression models, alongside response-style variables (avoidant coping and/or approach coping) that demonstrated significant zero-order correlations with the modelled outcome variable (as recommended by Tabachnick & Fidell, 2007). Explanatory variables were entered into the model in two blocks: Block one without the AAQ-II, and block two with the AAQ-II. This allowed R2-change scores to be calculated for the incremental contribution of acceptance (AAQ-II) to each model.

Time Two (T2). Preliminary t-tests were carried out to establish whether any significant changes occurred between T1 and T2 (three-month follow-up). Aim 3 was met using partial correlations and hierarchical multiple regression analyses, with T1 outcomes controlled for in each analysis of T2 outcomes.

Results

Participant characteristics

At T1, 72 participants completed a battery of questionnaires. The sample comprised 41 females, and 31 males (see Table 2 and underlying data (Moghaddam, 2020)). Of these participants, 58% were over 50 years old, with a modal age range of 65–74 years. There was a wide range of cancer sites reported, the most common of which were: breast (26.4%), prostate (25%), bowel (12.5%), lung (9.7%) and ‘other’ (16.7%). Of participants, 75% knew the stage of their cancer, and of these, 33.3% reported stage I or II cancers, and 41.6% reported stage III or IV. Secondary cancers were reported by 24% of participants, and 24% of participants had also had cancer at least once before. Clinically significant levels of distress were reported by 60% of the sample, as measured by HADS scores reaching 8 or above (Zigmond & Snaith, 1983), and 44% experienced high death anxiety, as measured on the Death Anxiety Scale, equalling or exceeding a cut-off score of 7.

Table 2. Characteristics of the overall sample at time one.

Sample demographicsSample (N = 72)Percentage %
Gender
  Male3143.1
  Female4156.9
Age range
  18–240-
  25–3479.7
  35–4468.3
  45–541723.6
  55–641216.7
  65–741926.4
  75+1115.3
Cancer site
Primary:
  Unknown34.2
  Breast1926.4
  Prostate1825.0
  Other1216.7
  Bowel912.5
  Lung79.7
  Blood45.6
Secondaries1723.6
Cancer stage
  I34.2
  II2129.2
  III1622.2
  IV1419.4
  Unknown1825.0
Highest level of
education
Missing22.8
None1216.7
Level 1 or below22.8
GCSE1723.6
A-Level45.6
Higher Education or
above
3548.6
Religion/spirituality
Yes2534.7
No4765.3
Number of previous
cancers
  05576.4
  11622.2
  2 +11.4

At T2, a follow-up rate of 53% (n = 31) was obtained from participants who had consented to the follow-up and then completed it. There were no significant differences in demographic or outcome variables between T2 participants who did and did not respond. Three participants had died between T1 and T2. Similar to the make-up of the T1 sample, the most common cancers reported were breast (26%), prostate (16%), and blood (10%), with 23% of participants reporting secondary cancers, and 13% having experienced more than one episode of cancer.

T1: Correlational analyses

Correlations between cancer stage, appraisal, response style, and outcome variables of interest are presented in Table 3. Cancer appraisal scores, measured on the Brief IPQ, demonstrated significant positive correlations with avoidance coping and death anxiety (rs = 0.34 and 0.35). Cancer appraisal was also negatively related to acceptance (AAQ-II score) and emotional QoL (rs = -0.36 and -0.26, ps < 0.05), and positively correlated with stage of cancer (r = 28, p < 0.05). Earlier stage of cancer was associated with worse physical and functional QoL (rs = -0.37 and -0.27, ps < 0.05).

Table 3. Time one bivariate correlation matrix between background, response style, and outcome variables.

12345678910
1 Stage of cancer.283*-.007.039.016.072-.141-.268*-.372**-.013
2 Brief IPQ-.361**.338**.190.204-.257*-.161-.165.353**
3 Acceptance (AAQ-II)-.709**-.039.298*.729**.496**.256*-.571**
4 Avoidance coping.266*-.217-.637**-.435**-.222.498**
5 Approach coping.210-.049.059-.088.097
6 FACT-G Social-.326**.456**.292*.000
7 FACT-G Emotional.621**.327**-.461**
8 FACT-G Functional.695**-.225
9 FACT-G Physical-.003
10 DAS

*p < 0.05, **p < 0.01

Note 1: Rows 1–2 = background variables, 3–5 = response style variables, and 6–10 = outcome variables

Note 2: Background variables demonstrating ≤1 significant association with focal response-style and outcome variables were suppressed

Note 3: AAQ-II = Acceptance and Action Questionnaire II; DAS = Death Anxiety Scale; FACT-G = Functional Assessment of Cancer Therapy – General; Brief IPQ = Brief Illness Perceptions Questionnaire

Avoidance coping (derived from the Brief COPE) demonstrated significant, moderate-to-large associations with poorer emotional and functional QoL (rs = -0.44 to -0.64) and greater death anxiety (r = 0.50) – and was thus entered as a response-style variable (alongside experiential acceptance) in subsequent regression analyses of these outcome variables. Avoidant response-style and experiential acceptance demonstrated a (conceptually consistent) inverse relationship of large magnitude (r = -0.71). Higher levels of acceptance (as measured by the AAQ-II) were significantly associated with more desirable outcomes on all outcome measures: Demonstrating small-to-moderate positive associations with social and physical QoL (rs = 0.26 to 0.30), large positive associations with emotional and functional QoL (rs = 0.50 to 0.73), and a large negative association with death anxiety (r = -0.57).

T1: Regression analyses

Alongside a priori explanatory variables of interest (acceptance [AAQ-II] and illness appraisal [IPQ]), avoidant coping was entered into regressions for functional, emotional, and death anxiety outcomes (i.e., those outcomes demonstrating significant zero-order correlations with avoidant coping).

Hierarchical regression results are displayed in Table 4. The addition of acceptance (AAQ-II, in block two) led to statistically significant increases in R2 and F values for four of the five models. Indeed, the social QoL model only reached significance following the addition of acceptance (increase in R2 = 0.11, F(1,69) = 8.68, p = 0.004). Acceptance demonstrated the strongest associations with outcome (based on absolute values of standardised coefficients) in four (of five) models and explained between 9% and 18% of unique model variance in these models (as indicated by incremental R2). Across models, acceptance had a significant positive relationship with emotional, functional, and social QoL, and a significant negative relationship with death anxiety. Cancer appraisals achieved significant explanatory power for emotional and physical outcomes – more negative appraisals were associated with poorer QoL in these domains, and these associations remained significant when modelled alongside acceptance. Notably, physical QoL was not associated with acceptance: Cancer appraisal was an informative variable in this domain (accounting for 12% of the variance in physical QoL) and acceptance was not incrementally informative (only explaining an additional 2% of the variance).

Table 4. Two-block hierarchical multiple regression results for time one variables.

Outcome variable (n)BlockPredictors
entered
St. βtpΔR2dfFp
FACT-G Social (72)1IPQ.0340.28.780.0011,700.00.780
2IPQ.1421.19.238.112**1,698.68.004
AAQ-II.351**2.95.004
FACT-G Emotional (65)1IPQ-.236*-2.25.028.398**2,6220.51<.001
Avoidance-.510**-4.86<.001
2IPQ-.219*-2.46.017.176**1,6125.21<.001
Avoidance-.104-0.87.389
AAQ-II.588**5.02<.001
FACT-G Functional (65)1IPQ-.189-1.53.132.162**2,625.99.004
Avoidance-.297*-2.40.020
2IPQ-.177-1.50.139.090*1,617.38.009
Avoidance-.006-0.04.971
AAQ-II.421**2.72.009
FACT-G Physical (72)1IPQ-.346**-3.081.003.119**1,709.49.003
2IPQ-.295*-2.519.014.0241,691.95.167
AAQ-II.1641.40.167
DAS (64)1IPQ.082.687.495.238**2,619.55<.001
Avoidance.454**3.82<.001
2IPQ.0670.61.546.110**1,6010.11.002
Avoidance.1340.90.375
AAQ-II.464**3.18.002

*p < 0.05, **p < 0.01

Note 1: AAQ-II = Acceptance and Action Questionnaire II; DAS = Death Anxiety Scale; FACT-G = Functional Assessment of Cancer Therapy – General; IPQ = Illness Perceptions Questionnaire (assessed via Brief IPQ)

Note 2: Only response style variables which significantly correlated with each outcome were entered in the models. Brief IPQ and AAQ-II were entered into all models.

Avoidant coping demonstrated significant associations with emotional and functional QoL and death anxiety, but these associations did not survive entry of acceptance in the second block of respective models. This indicated that the acceptance measure (AAQ-II) overlapped with avoidant coping whilst accounting for additional unique variance in outcomes.

T2 analyses

Repeated-measures t-tests (Table 5) revealed that outcome scores significantly changed over time in all four QoL subdomains: emotional, social, physical, and functional. The direction of change showed that emotional and physical QoL scores significantly increased over time, whereas social and functional QoL decreased. There was no significant change in reported death anxiety. Moreover, scores were stable over time for illness appraisals and all response style variables (ps > 0.05). No shifts in participants’ demographic or clinical details were identified between T1 and T2.

Table 5. Dependent t-test, means, and standard deviations for time one and two outcome variables.

Time OneTime Twot-test
Outcome VariableMeanSDMeanSDp
FACT-G Social16.735.9110.177.61<.001**
FACT-G Emotional16.324.7119.302.81.001**
FACT-G Functional16.126.2010.688.65.004**
FACT-G Physical21.304.9226.431.92<.001**
DAS6.473.215.802.93.235

Note: DAS = Death Anxiety Scale; FACT-G = Functional Assessment of Cancer Therapy – General

*p < 0.05, **p < 0.01

T2: Partial correlations and regression analyses

Partial correlations were carried out between the T1 response style that had previously emerged as significant (avoidant coping), and the outcome variables showing significant change at T2 (QoL sub-domain scores; Table 6). Avoidant coping at T1 was not significantly correlated with T2 outcomes and was consequently excluded from subsequent regression analyses.

Table 6. Time Two partial correlation matrix (controlling for Time One outcome variables).

Established predictor & response variables: Time One
AvoidanceAAQ-IIIPQ
Outcomes: Time Two
FACT-G Physical T2-.056.121-.149
FACT-G Social T2-.207.332.252
FACT-G Emotional T2-.339.110-.320
FACT-G Functional T2-.241.442*.363

*p < 0.05, **p < 0.01. T2 = Time Two.

Note 1: AAQ-II = Acceptance and Action Questionnaire II; FACT-G = Functional Assessment of Cancer Therapy – General; IPQ = Illness Perceptions Questionnaire (assessed via Brief IPQ)

Note 2: Partial correlation analyses were limited to (1) response style variables that were significantly correlated with outcomes at Time One and (2) outcome variables demonstrating significant change over time (Time One to Time Two)

A two-block hierarchical regression model was tested for each T2 outcome variable, regressing the T2 QoL outcomes onto T1 values for cancer appraisal (IPQ) and acceptance (AAQ-II; as a priori explanatory variables of interest) whilst controlling for QoL at T1, such that T2 outcomes represent change-scores (Table 7).

Table 7. Two block hierarchical regression results for predicting outcome variables at Time Two.

Outcome variable T2 (n)BlockPredictors enteredSt. βtpΔR2dfFp
FACT-G Social T2 (30)1IPQ.2021.28.210.341**2,276.98.004
Social T1.524**3.33.003
2IPQ.2621.75.092.105**1,264.90.036
Social T1.3231.87.073
AAQ-II.381*2.21.036
FACT-G Emotional T2 (31)1IPQ-.312-1.79.084.264*2,285.03.014
Emotional T1.3111.79.084
2IPQ-.334-1.89.069.0211,270.80.380
Emotional T1.2020.95.351
AAQ-II.1780.89.380
FACT-G Functional T2 (31)1IPQ.373*2.08.047.1672,282.80.078
Functional T1.3011.67.105
2IPQ.388*2.42.023.194**1,278.19.008
Functional T1.1981.21.239
AAQ-II.454**2.86.008
FACT-G Physical T2 (30)1IPQ-.146-0.74.463.0992,271.49.243
Physical T1.2311.17.250
2IPQ-.138-0.69.496.0111,260.33.570
Physical T2.2261.14.266
AAQ-II.1070.58.570

*p < 0.05, **p < 0.01. T2 = Time Two.

Note 1: AAQ-II = Acceptance and Action Questionnaire II; FACT-G = Functional Assessment of Cancer Therapy – General; IPQ = Illness Perceptions Questionnaire (assessed via Brief IPQ)

Note 2: Only response style variables which significantly correlated with each outcome were entered in the models. Brief IPQ and AAQ-II were entered into all models.

T1 acceptance was a significant explanatory predictor of future functional QoL (βstandardised = 0.45, p = 0.008) and social QoL (βstandardised = 0.38, p = 0.036). The direction of these coefficients was consistent with T1 regressions, with acceptance predictive of higher future QoL. Cancer appraisal was also a significant explanatory predictor of functional QoL (βstandardised = 0.38., p = 0.023), but acceptance accounted for 19% of the unique variance in this model, over and above variance explained by cancer appraisal and functional QoL at T1. In regression models for other T2 outcome variables, none of the entered predictor variables reached significance – although the T2 analyses only had sufficient power to detect large effects.

Discussion

This longitudinal quantitative study aimed to explore whether acceptance was related to QoL and death anxiety in cancer patients, both concurrently and prospectively. Acceptance and death anxiety were significantly and negatively associated, in line with previous literature on anxiety in cancer patients. Avoidant response styles (such as denial, behavioural disengagement, and self-distraction) showed a pattern of moderate to strong negative correlations with QoL, and significant positive correlations with death anxiety in the cancer population. Cancer appraisals demonstrated a greater number of significant correlations with response styles and outcomes than measures of physical disease characteristics. In line with previous research, this supports the role of more threatening cognitive appraisals in poorer outcomes; relationships that were not demonstrated by or contingent upon disease characteristics alone. Contrastingly, acceptance was an independent explanatory variable for (concurrent) QoL and anxiety outcomes, in directions consistent with psychological health. Acceptance also demonstrated predictive power for these outcomes over time, over and above the influence of cancer appraisals.

In contrast to relations observed for acceptance, ‘approach coping’ demonstrated negligible-to-small, non-significant associations with QoL and death anxiety. As defined by the COPE, approach coping includes strategies that are promoted in traditional CBT (such as positive reframing and efforts to actively control and problem-solve stressors). It has been argued that such (problem-focussed) strategies are less apt for relatively uncontrollable stressors (such as cancer; Hulbert-Williams et al., 2015; Park et al., 2004); moreover, a recent meta-analysis of interventions for anxiety in cancer (specifically around fear of recurrence; Tauber et al., 2019) found that traditional CBT was outperformed by contemporary cognitive-behavioural interventions (including ACT) – which place greater emphasis on relating to difficult experiences with active openness and acceptance (Hayes et al., 2011). Thus, the present finding – that coping strategies aligned with traditional CBT demonstrated little association with wellbeing or death anxiety, whereas active acceptance was associated with greater wellbeing and lower death anxiety – is congruent with conceptual arguments and recent empirical evidence underpinning ACT and other contemporary cognitive-behavioural intervention-models (Hayes et al., 2011).

Theoretical implications

In contrast to the ‘protective’ and defensive role of avoidance posited by TMT, avoidance was significantly associated with lower QoL, and higher death anxiety. TMT suggests that avoidance (denial, suppression, and distraction) should be utilised to protect against the terror of mortality and its associated impact upon QoL; yet here the findings that avoidance was significantly associated with death anxiety and poorer QoL (in the presence of the proximal mortality-threat of cancer) suggest the opposite.

QoL declined over time on social and functional domains, and deteriorations in these domains were predicted by lower levels of acceptance (i.e., a response-style characterised by experiential avoidance). Thus, contrary to TMT, and consistent with ACT theory, openness to experience appeared to dispose participants to better psychological outcomes over time. Of further relevance to ACT theory, it was notable that (within the present sample) functional and social QoL both declined over time, despite physical and emotional QoL improving, lending support to the ACT notion that valued living (functional and social QoL) is independent of change in physical and distress symptoms (Hulbert-Williams et al., 2015).

Given that avoidant response styles did not retain significant explanatory power following the addition of the acceptance measure, it is likely that avoidant responses and experiential avoidance (as measured by low AAQ-II scores) share common variance, and that this prevented avoidant response styles from achieving significance. Experiential avoidance may therefore be a generalised avoidance phenomenon, rather than a unique concept, given its large correlations with other measures of avoidant responding (Chawla & Ostafin, 2007). Acceptance, however, as measured by the AAQ-II, made a unique and statistically independent contribution as an explanatory variable: accounting for additional model variance, over and above cancer appraisals or avoidant response styles. Although these are modest percentages, fluctuations in the amount of variance explained may reflect the relevance of the variables entered in explaining outcome. For example, a large proportion of variance was explained for concurrent emotional QoL (40%, plus an additional 18% with the addition of AAQ-II in the final model), but relatively little was explained for concurrent social QoL (11% in the final model). We may therefore theorise that the way we respond to stressors – e.g. via acceptance or avoidance – can contribute to emotional outcomes, but the response styles captured here are less likely to be direct determinants of the external influencers of social QoL, such as support from family and friends. Nevertheless, response styles may still indirectly affect social support over time, e.g. by influencing whether family and friends make themselves available, or perhaps by influencing subjective appraisals of whether friends and family are available, hence the smaller degree of variance explained.

Limitations

Limitations of this study include its observational research design, and the cross-sectional nature of the data collected at time-one, which was subject to the limitations of correlational, multivariate data analyses and the non-causal conclusions that are drawn from them. Inclusion of data at a second timepoint partially addressed these limitations but was limited by attrition (only large effects could be detected at follow-up).

Time since diagnosis was not accurately captured, as it has been elsewhere in cancer literature (e.g. Hulbert-Williams et al., 2015). A further limitation may be posed by the use of the AAQ-II, as items may be conflated with distress outcomes (Wolgast, 2014). Although the AAQ-II was used as a process measure indicative of acceptance, with implications for targeted assessment and intervention in this process, the observed relationships with outcomes may be artificially inflated by overlapping content and weak discriminant construct validity. However, use of the AAQ-II is well-established in the ACT literature, and has been deemed to be of satisfactory reliability and validity in research with cancer populations (Feros et al., 2013).

Conclusions

Acceptance is supported as a helpful response style, and a potential target for intervention in psychological work with cancer patients. Psychological interventions targeting acceptance are also supported over and above those directed towards illness appraisals in cancer patients, which were somewhat correlated with distress outcomes but had less predictive and explanatory power. Results also showed that many response styles that might be targeted in more traditional, often appraisal-focused, work such as CBT or expressive therapies, are unlikely to be associated with positive outcomes in cancer patients. Such strategies include active coping, emotional support, planning, positive reframing, and venting. As experiential avoidance and avoidant response styles had greater predictive power than cancer appraisals in QoL and death anxiety outcomes, increasing acceptance and shifting relationships to distressing cognitions, rather than attempting to modify their content, gives a clear direction for future intervention studies with this population.

Data availability

Underlying data

Open Science Framework: Accepting the unacceptable? https://doi.org/10.17605/OSF.IO/XSTZF (Moghaddam, 2020)

This project contains the following underlying data:

  • Data - exploring acceptance in a cancer population.sav (Participants level data on questionnaire battery responses)

Data are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).

Looking for the Open Peer Review Reports?

They can now be found at the top of the panel on the right, linked from the box entitled Open Peer Review. Choose the reviewer report you wish to read and click the 'read' link. You can also read all the peer review reports by downloading the PDF.

Comments on this article Comments (0)

Version 1
VERSION 1 PUBLISHED 14 Apr 2020
Comment
Author details Author details
Copyright
Download
 
Export To
metrics
VIEWS
$counts.viewCount
 
downloads
Citations
CITE
how to cite this article
Brabbins L, Moghaddam N and Dawson D. Accepting the unacceptable? Exploring how acceptance relates to quality of life and death anxiety in a cancer population [version 1; peer review: awaiting peer review] Emerald Open Res 2020, 2:13 (https://doi.org/10.35241/emeraldopenres.13524.1)
NOTE: it is important to ensure the information in square brackets after the title is included in all citations of this article.
track
receive updates on this article
Track an article to receive email alerts on any updates to this article.

Open Peer Review

Current Reviewer Status:
AWAITING PEER REVIEW
AWAITING PEER REVIEW
?
Key to Reviewer Statuses VIEW
ApprovedThe paper is scientifically sound in its current form and only minor, if any, improvements are suggested
Approved with reservationsKey revisions are required to address specific details and make the paper fully scientifically sound
Not approvedFundamental flaws in the paper seriously undermine the findings and conclusions

Comments on this article Comments (0)

Version 1
VERSION 1 PUBLISHED 14 Apr 2020
Comment

Do you have research related to the global impact of coronavirus?

We're funding £20,000 of Emerald Open Research APCs (article processing charges) for research relating to public health outbreaks.

Find out more
Alongside their report, reviewers assign a status to the article:
Approved - the paper is scientifically sound in its current form and only minor, if any, improvements are suggested
Approved with reservations - key revisions are required to address specific details and make the paper fully scientifically sound
Not approved - fundamental flaws in the paper seriously undermine the findings and conclusions
Sign In
If you've forgotten your password, please enter your email address below and we'll send you instructions on how to reset your password.

The email address should be the one you originally registered with F1000.

Email address not valid, please try again

You registered with F1000 via Google, so we cannot reset your password.

To sign in, please click here.

If you still need help with your Google account password, please click here.

You registered with F1000 via Facebook, so we cannot reset your password.

To sign in, please click here.

If you still need help with your Facebook account password, please click here.

Code not correct, please try again
Email us for further assistance.
Server error, please try again.