DEPRESSION AND QUALITY OF LIFE: A SYSTEMATIC REVIEW OF THE CANCER LITERATURE,
3) What is the association between illness cognitions and emotions on psychosocial outcome for cancer patients?
2.5 LITERATURE SEARCHING .1 Study Location
2.5.7 Amendment of review objectives
Despite highlighting a large number of papers in the early stages, question one (that of association between personality and appraisal) was removed as a study objective at this stage. The objectives were initially reviewed because it was felt that the review had become too large to feasibly answer all research questions with sufficient clarity and quality. In exploring each question individually it became clear that question one was the weakest. There were two primary reasons. First, of the three review questions, the least amount of research had been conducted on question one and, therefore, it was felt more salient to review those that had been the focus of greater research effort. Reviews of this question will be important in due course. However, with the heterogeneity observed in this
53
field of research, the studies identified were not expected to be similar enough to be synthesised in an informative way. Instead, (as with many other excluded papers) the studies were used in the narrative review (Chapter One). Second, it was expected that in extraction of data relating to the remaining two questions, a partial (albeit not strictly systematic) answer to the association between
personality and appraisal may be possible.
Although this modification to the study objectives did not dramatically reduce the literature size (ultimately, only six studies were removed as many of the other papers highlighted by this specific search were relevant to other research questions), by reducing the scope of the review it made data synthesis more straightforward and the findings more comprehensible.
54
Table 2.3. Publication details and variables measured fo r the 68 included studies.
Relevant measures used1
ID Authors Date Title
Personality
2005 Mood, anxiety and sense of humour in head and neck
• Higher anxiety/lower depression than controls.
• Some clinical variables predict outcome.
■ Sense of humour at diagnosis predicts QOL and depression at six year follow up.
Ahmed,
Depression ■ Anxiety, depression and neuroticism higher in (ZDI) patients than controls.
■ Concurrent anxiety and neuroticism predict depression in patients.
1996 Quality of life in oncological patients.
A study of 105 cases.
Locus of Control (MHLC)
Quality of ■ Co-morbidity and locus of control were strongly Life (GRQLI) associated with QOL
Allison, 2000 A prospective Guichard & investigation of
Gilain. dispositional
Quality of • Pre- and post- treatment, optimists reported Life (EORTC) better QOL
55
Table 2.3. Publication details and variables measured fo r the 68 included studies ( continued).
ID Authors Date Title Relevant measures used' Summarised Main Findings
5 Andrykowski 1994 Health locus of ■ Locus of Control "Distress • Disease severity and treatment moderate the
& Brady control and (MHLC) (PAIS) relationship between LOC and distress (LOC has
psychological ■ Depression most effect on distress at low levels of perceived
distress in cancer (POMS) severity),
patients: interactive effects of context
6 Andrykowski 1996 Psychosocial ' ■ Positive Affect "Depression • Patients reported poorer physical health,
, Curran, adjustment and (PANAS) (POMS/CES- functioning but greater psychosocial adaptation
Studts, quality of life in ■ Negative Affect D) than controls (benign breast problem patients).
Cunningham, women with breast (PANAS) ■ Quality of ■ No differences in distress between patients and
Carpenter, cancer and benign • Personal Change Life (PHQ) controls.
McGrath, breast problems: a
Sloan & controlled
Kenady._______________ c o m p a r i s o n . _________ _________________ _______________________________________
7 Badger, 2004 Depression burden, ■ Negative Affect ■ Well Being ■ Mood at diagnosis is predictive of mood at
Braden, psychological (NAS) (IW B) longitudinal follow-up.
Mishel & adjustment, and ■ Greatest adjustment observed in those reporting
Longman. quality of life in perceived high depression burden in an
women with breast intervention group,
cancer: Patterns
____________________ over time.______________________________
8 Baider, 2003 Effects of age on ■ Fatalism (MAC) ■ Distress ■ Cross-cultural differences in distress observed.
Andritsch, coping and ■ Hopeless/Helples (PAIS) • Across all subsamples, age and distress were
Uziely, psychological sness (MAC) significantly associated (young women, more
Goldzweig, distress in women ■ Anxious distress).
Ever-Hadani, diagnosed with Preoccupation
Hofman, breast cancer: (MAC)
Krenn& review of literature ■ Fighting Spirit
Samonigg. and analysis of two ■ Thought
different settings. Intrusion/
Avoidance (IES)
56
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Title Relevant measures used1 Summarised Main Findings
Bleiker, van der Ploeg &
Ader.
1995 Personality traits of • Rationality women with breast
Depression ■ Rationality and emotional expression differed (SAQ-N) between patients and controls.
• No significant depression differences.
10 Bleiker, 2000 Psychological Pouwer, van distress two years der Ploeg, after diagnosis of Leer & Ader. breast cancer:
frequency and
Depression • 16% still distressed, two years after diagnosis.
(SAQ-N) ■ Intrustive thoughts, anxiety and health complaints were best predictors of two year distress levels.
11 Boer, Elving 1998 Psychosocial factors
& Seydel and mental health in cancer patients:
opportunities for __________________ health promotion.
Self-Efficacy (GSES)
Quality of Life (RAND)
Mental health in cancer survivors is lower than general population controls.
Self-efficacy and loneliness were strong predictors of mental health.
57
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Title Relevant measures used1 Summarised Main Findings
12 Brown, King, 2000 Patterns over time ■ Isolation (PAC) • Quality of • QOL fluctuates highly over time; mood remained
Butow, Dunn in quality of life, • Minimisation Life stable.
& Coates. coping and (PAC) (LASA/GLQ) ■ Well-being deteriorated over time
psychological ■ Anger (PAC) " Between patient variability accounted for 60%,
adjustment in late ■ Perceived aim of 45%, and 44% of the total variance in effort to
stage melanoma treatment cope, mood and well-being respectively.
patients: An ■ Perceived ability • Perceived treatment aims, minimization, anger,
application of to cope with marital status and better QOL were significant
multilevel models. illness (PACIS)
■ Coping (COPE) distress: a study of women with early
Optimism was significantly associated with distress (prior levels controlled for) at each time point
Several coping reactions (esp humour, denial and disengagement) are longitudinally predictive of lower distress levels.
14 Carver,
1994 Optimism versus pessimism predicts
Pessimism was significantly predictive of longitudinal well-being even when previous levels were controlled for.
No effect of pessimism on quality of sex life or thought intrusions.
Optimism-pessimism are adequately measured using single item scales.
58
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
2000a How important is the perception of
Perceived control over remaining cancer free was not associated with distress.
Expectancy of remaining cancer free was associated with distress levels.
2000b How important is the perception of
17 Cohen, Moor 2000 The association
& Amato between treatment-specific optimism and depressive symptomatology in patients enrolled in a phase I cancer clinical trial.
Treatment Specific Optimism
Depression ■ Treatment specific optimism was significantly (CES-D) associated with fewer depressive symptoms, Distress lower mood disturbance, and fewer symptoms (POMS) of distress (after demographic and clinical
variables were controlled for).
■ Clinically depressed patients reported
significantly lower treatment specific optimism at diagnosis.
59
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Title Relevant measures used< Summarised Main Findings
18
• Perceived Stress (PSS)
■ Coping mediated the relationship between health status and QOL
• Self-accusation and perceived stress mediated the relationship between trait anxiety and QOL
• Perceived stress mediated the relationship between tumour status and QOL
19 De Valck &
Vinck
1996 Health locus of control and quality of life in lung cancer patients.
■ Locus of Control (MHLC)
• Quality of Life (DCL)
■ Patients reported higher internally oriented LOC beliefs than controls.
• Fluctuations in QOL best explained by increasing physical complaints.
■ No significant association between LOC and QOL
20 Dropkin 2001 Anxiety, coping strategies, and
■ Self-care was the only significant predictor of anxiety.
■ Optimism ■ Monitoring
■ Thought Intrusion (IES)
■ At diagnosis, age and distress relationships were mediated by intrusive thoughts. Optimism - distress relationships were mediated by coping.
■ At three month follow up, distress was only predicted by intrusive thoughts.
■ Six month distress was predicted by optimism (mediated by coping).
60
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Title Relevant measures used' Summarised Mala Findings
22 Evans,
being of adults with acute leukaemia in
• Some differences in well being found to be associated with variability across a number of personality characteristics (endurance, affiliation, cognitive structure, autonomy, and nurtu ranee).
23 Faller, Bulzebruck, Drings &
Lang.
1999 Coping, distress, and survival among
■ Depressive coping and emotional distress were both significant predictors of distress.
24 Gallagher,
• Psychological functioning was significantly related to psychiatric history, tumour grade and cognitive appraisals (esp. threat and self- efficacy).
• Behavioural self-blame was predictive of concurrent distress levels only;
characterological self-blame was predictive of changes in distress over time.
26 Golden-
■ Perceived Stress (PSS)
• Thought Intrusion (IES)
■ Thought Avoidance (IES)
■ Depression ■ Perceived stress and neuroticism both significantly predict depressive symptoms.
61
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Tide
Relevant measures used1 Summarised Main Findings
27 Goodwin, quality of life and psychosocial status
Psychosocial status and quality of life were not associated with medical outcome.
1988 Pattern of emotional control and
Some associations between emotional control, hostility and psychosocial outcome were found.
29 Green, Pakenham, Headly &
Gardiner.
2002 Coping and health- related quality of
Self Efficacy Threat appraisal Coping (COPE) Satisfaction with life
Quality of • Patient groups did not differ on measures of Life (EORTC) distress, life satisfaction, cognitive function, Depression physical symptoms, or social/role functioning.
(DASS) • Hormonal treatments were associated with Anxiety poorer psychosexual functioning, but improved (DASS) physical symptoms.
• Threat appraisals and coping were predictive of _________ longitudinal QOL__________________ ________
30 Hack& 2004 Coping responses
Degner. following breast
Baseline depression and cognitive avoidance and minimal use of approach based coping strategies were associated with poorer adjustment at three years.
62
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Title Relevant measures used1 Summarised Main Findings
31 Hassanein, 2001 Functional status of » Fatalism (MAC) • Quality of • 25% of patients had anxiety/or depression.
Musgrove & patients with oral ■ Fighting Spirit Life • Poorer functioning was associated with many
Bradbury cancer and its (MAC) (UWQOL/EO clinical and demographic variables.
relation to style of ■ Anxious RTC) " Functionality was also associated with anxiety.
coping, social Preoccupation • Anxiety depression, coping style (support and fighting
support and (MAC) (HADS) spirit were only very weakly associated).
psychological status. • Hopelessness/Hel Depression ■ Anxiety and depression were only weakly
Hassanein, 2005 Psychological plessness (MAC) (HADS) correlated with socio-demographics and
Musgrove & outcome of patients medical variables.
Bradbury. following treatment of oral cancer and its relation with functional status and coping mechanisms.
• Anxiety and depression were highly correlated with cancer specific QOL
• MAC scores were strongly associated with psychosocial outcome.
32 Hee, Kim, 2005 Quality of life in ■ Extroversion • Quality of • FACT C is a valid tool for measuring QOL in
Eremenco & colorectal cancer (EPI) Life cancer patients.
Han patients with
33 Helgeson, 2004 Psychological and Snyder & physical adjustment
Seltman to breast cancer
over 4 years:
Quality of • Age, self-image, perceived control, and social Life (SF-36) resources all distinguished between different
trajectories of adjustment.
63
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Title Relevant measures used1 Summarised Main Findings
34 Helgeson &
Lepore
2004 Quality of life following prostate cancer: the role of agency and
• Agency was associated with indicators of good QOL
• Unmitigated agency was associated with intrusive thoughts, depressive symptoms and poorer mental functioning.
• Role of agency was related to self-efficacy.
35 Iwamitsu,
■ Emotional Control (CECS)
■ Anxiety (MAS)
■ Distress (POMS)
” Pre surgery anxiety and emotional suppression were associated with distress pre- and post- surgery.
36 Kessler 2002 Contextual
variables, emotional state, and current and expected quality of life in breast cancer survivors.
■ Affects (PANAS)
• Satisfaction with life
■ Quality of Life
■ Affect was associated with both life satisfaction and QOL
■ Survivors reported greater comparative life satisfaction and life satisfaction future expectancies.
2001 Distress, coping and social support
■ Distress and hopelessness/helplessness were both reported at very high levels.
64
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Title Relevant measures used1 Summarised Main Findings
38 Laubmeier, 2004 The role of ■ Perceived Life • Quality of ■ Spirituality was associated with less distress
Zakowski & spirituality in the Threat Life (FACT) and better QOL regardless of levels of perceived
Bair psychological • Well being life th reat
adjustment to (SWBS) ■ Existential, but not religious well being
cancer: a test of the accounted for most variance in QOL
transactional model of stress and coping.
39 Lehto, 2005 Predictors of quality ■ Coping (WoC) • Distress • Psychosocial factors were the strongest
Ojanen & of life in newly ■ Anger Expression (RSCL) predictors of QOL (when compared with cancer
Kellokrumpu diagnosed ■ Depression and treatment types).
-Lehtinen melanoma and (DEPS) ■ Breast cancer patients received far more social
breast cancer ■ Quality of support than melanoma patients.
patients. Life (EORTC) • Adjuvant treatment may comprimise the
beneficial role of psychosocial factors that enhance QOL
40 Lewis 1982 Experienced
personal control and quality of life in late-stage cancer patients.
Self Esteem (RSE)
Purpose in Life (PLT)
External Locus of Control (HLCS)
Anxiety (ZAS)
Perceived meaning (but not control) of the illness was associated with anxiety and self*
esteem.
Lewis 1989 Attributions, experienced meaning, and psychosocial w ell
being in patients with advanced cancer.
65
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Tide Relevant measures used' Summarised Main Findings
41 Lloyd, Parker, Ludlam &
Maguire
1984 Emotional impact of diagnosis and early
■ No age or treatment associations with outcome.
■ Female patients were more likely to meet psychiatric disorder criteria.
" Neuroticism was strongly associated with dissatisfaction with communication with medical staff.
• Affect was associated with anxiety and depression, but not quality of life.
43 Lowery,
■ Thought Intrusion /Avoidance(IES)
■ Locus of Control (MHLC)
■ Distress (PAIS)
■ 'Why me?' attributions, perceived loss of control, and higher impact of event scores were associated with both poorer adjustment and prediction of psychological distress.
1995 Cognitive factors in adjustment to
■ Self blame was associated with longitudinal distress (four months), but not concurrent distress levels.
• Perceptions of control had no direct or mediating effects for distress.
66
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Title Relevant measures used< Summarised Main Findings
45 Marks, Richardson, Graham &
Levine
1986 Role of health locus of control beliefs
" Perceived control mediated the relationship between perceptions of disease severity and depression. cancer: the effects of optimism and
■ Optimism and coping were both associated with adjustment, even after controlling for functional status and prior adjustment measures.
■ Optimism was positively associated with w ell
being, and negative associated with distress.
47 Morris,
" Anger expression was higher in patients than controls.
• Of all personality traits, only neuroticism differed between patients (higher) and controls (lower neuroticism).
1996 Quality of life and anxiety before and locus of control on depression.
■ Demonstrated an interactive effect between some locus of control scores and anxiety in the prediction of depression.
• Suggest that those loci of control assumed to be adaptive in the general population may, in fact, be maladaptive in early stage cancer survivors.
67
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Title Relevant measures used' Summarised Main Findings
50 Nelson,
• Of four patient profile clusters, only one demonstrated high distress and global maladjustment
• The remainder all had normal affect levels, but differed on various QOL/functioning subscales.
• Suggests that even non-emotionally distressed patients have varying adjustment trajectories and may all benefit from intervention.
51 Nordin &
being over time in patients with
■ Fighting spirit was associated with emotional well-being; hopeless/helplessness and anxious preoccupation was not
• MAC scores were generally stable over time.
• Marked differences observed between HADS clinical/non clinical subsamples. beliefs and coping in adult cancer
patients.
■ Perception of Control
■ Coping (CS1)
■ Thought Intrusion (IES)
• Problem focussed coping was associated with low anxiety/depression; emotion focussed coping with more symptoms.
■ Interactions between PF coping and control were predictive of lower anxiety and depression at baseline, but not four months later.
53 Osoweicki &
Compas
1999 A prospective study of coping, perceived
■ PF coping associated with lower distress near diagnosis.
■ EF disengagement coping associated to distress at 6 months controlling for earlier levels.
■ No main effects for perceived control.
■ Interaction between coping and control significant at baseline only, not longitudinally.
68
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Title Relevant measures used' Summarised Main Findings
54 Padilla,
■ Mastery • Illness uncertainty (MUIS)
• Mood, illness ambiguity, mastery and danger focussed appraisal are all key predictors of QOL.
explaining far more variance than clinical factors.
• Coping was not predictive of QOL 55 Perczek,
Burke, Carver, Krongrad &
Terris
2002 Facing a prostate cancer diagnosis:
• Avoidance was predictive of increased distress;
optimism, cancer status, and all other coping strategies weren’t
■ Interactions between optimism and coping weren't significant predictors either.
■ High neuroticism was associated with high distress in short and long term analyses.
■ Higher social support was associated with higher distress.
• L-Scores (reflecting the cancer prone personality) were high at diagnosis and associated with risk of death at 5 years.
■ Depression was also an independent 5 year survival risk factor.
58 Rondorf- Klym &
Colling
2003 Quality of life after radical
■ Anger Expression (AEI)
■ Perceived support, self-esteem, and health locus of control were all significant predictors of QOL.
■ Urinary function appraisal was mediated by locus of control.
• Anger suppression and depression were not significant predictors of QOL.
• Higher social support, optimism, high perceived meaning in life and lower avoidant coping were associated with better adjustment
69
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Tide
Relevant measures used% Summarised Main Findings
60 Schou 2004 Pessimism as a ■ Optimism ■ Positive ■ Anxiety • Prevalence of emotional morbidity was
Ekeberg, predictor of (LOT) expectations (HADS) significantly higher among pessimists.
Ruland, emotional morbidity • Fatalism (MAC) • Depression • Pessimism was the strongest predictor of one
Sandvik & one year following • Fighting Spirit (HADS) year anxiety and depression.
Karesen. breast cancer (MAC) • Quality of ■ Optimists and pessimists have different coping
surgery. ■ Hopelessness/Hel Life (EORTC) approaches which may interact to predict
plessness (MAC) outcome.
Schou, 2005a The mediating role ■ Anxious Optimism had both a direct and a mediated (via
Ekeberg & of appraisal and Preoccupation fighting spirit and hopeless/helplessness) effect
Ruland. coping in the (MAC) on QOL
relationshp between ■ Treatment ■ Relationships between predictors and outcomes
opdmism- decision making were also mediated by threat appraisals at the
pessimism and ■ Primary time of diagnosis.
quality of life Appraisals (VASA) Compared to general population controls, at
■ Treatment diagnosis patients reported lower emotional,
Schou, 2005b Multiple predictors decision making cognitive and social functioning.
Ekeberg, of health-related ■ Cognitive and social functioning disparity
Sandvik, quality of life in continued until one year post surgery.
Hjermstad & early stage breast ■ Breast conservation surgery a chemotherapy
Ruland. cancer. Data from a were predictive of functioning one year after
year follow-up study surgery.
compared with the ■ Throughout, optimism was associated with
general population. better QOL and fewer symptoms.
61 Stanton, 2000 Emotionally ■ Hope ■ Quality of ■ Emotionally expressive coping was associated
Danoff-Burg, expressive coping ■ Coping (COPE) Life (FACT) with fewer cancer related medical
Cameron, predicts ■ Perceived Health ■ Distress appointments, enhanced health and vigour, and
Bishop, psychological and (POMS) decreased distress at three month (age, coping
Collins, Kirk, physical adjustment and initial levels controlled for).
Sworowski & to breast cancer. • Relationships between emotionally expressive
Twillman coping and QOL were mediated by social
receptivity.
70
Table 2.3. Publication details and variables measured fo r the 68 included studies (continued).
ID Authors Date Tide Relevant measures used' Summarised Mala Findings
62 Taylor,
1999 The role of coping in the psychological adjustment of African American women with early- stage breast cancer.
■ Thought Intrusion (IES)
• Avoidant coping was associated with
psychological distress, particularly in younger patients.
• Relationships between illness attributions and adjustment were mediated by perceived
• Relationships between illness attributions and adjustment were mediated by perceived