Fax: (011) 61 7 3257 1306
Describing and predicting psychological distress after colorectal cancer
Article first published online: 18 JAN 2008
Copyright © 2008 American Cancer Society
Volume 112, Issue 6, pages 1363–1370, 15 March 2008
How to Cite
Lynch, B. M., Steginga, S. K., Hawkes, A. L., Pakenham, K. I. and Dunn, J. (2008), Describing and predicting psychological distress after colorectal cancer. Cancer, 112: 1363–1370. doi: 10.1002/cncr.23300
- Issue published online: 3 MAR 2008
- Article first published online: 18 JAN 2008
- Manuscript Accepted: 1 OCT 2007
- Manuscript Revised: 25 SEP 2007
- Manuscript Received: 17 JUL 2007
- Cancer Council Queensland
- colorectal neoplasms;
- psychological distress;
- cancer survivorship
Psychological distress in cancer survivors can be detrimental to treatment adherence and self-care tasks and is associated with poor health behaviors and decreased overall quality of life. The prevalence, course, and predictors of psychological distress after the diagnosis of colorectal cancer are to date not well described.
A prospective survey of 1822 colorectal cancer patients was undertaken assessing psychological distress and hypothesized predictors including optimism, cancer threat appraisal, social support, and physical activity at 6 and 12 months postdiagnosis. Logistic regression identified correlates of psychological distress at 12 months postdiagnosis.
The prevalence of global psychological distress was low: 8.3% and 6.7% at 6 and 12 months postdiagnosis, respectively. When baseline measures of independent variables were included in a logistic regression model, distress at 6 months postdiagnosis (odds ratio [OR] = 10.84), comorbidities (OR = 1.64), optimism (OR = 0.93), cancer threat appraisal (OR = 0.92), and social support (OR = 0.94) were significantly associated with distress at 12 months postdiagnosis. A second logistic regression model that included concurrent measures of cancer threat appraisal, social support, and physical activity found that distress at 6 months postdiagnosis (OR = 12.49), comorbidities (OR = 1.64), optimism (OR = 0.94), and concurrent cancer threat appraisal (OR = 0.85) were significantly associated with distress at 12 months postdiagnosis.
Distress screening at regular intervals is needed to efficiently detect colorectal cancer patients who require in-depth psychological intervention. Threat appraisal is a modifiable variable that should be included in interventions for colorectal cancer survivors. Further research is needed to investigate the potential for physical activity to reduce distress after cancer. Cancer 2008. © 2008 American Cancer Society.
Colorectal cancer is 1 of the leading causes of morbidity and mortality worldwide, particularly in developed countries. In Europe, the US, and most of the industrialized world, colorectal cancer is the third most common cancer and the second most common cause of cancer deaths,1, 2 and the prevalence of colorectal cancer is expected to increase as worldwide trends in population aging continue. Treatment for colorectal cancer includes surgical, radiation, and systemic therapies, all of which may have debilitating side effects such as fatigue and nausea. For patients whose diagnosis warrants adjuvant therapy, the treatment trajectory can be up to 12 months or more.3 In addition, as with other cancers, colorectal cancer survivors can be expected to experience psychological distress that for a significant proportion of survivors will endure over time.4 Elevated psychological distress is a key clinical issue in such individuals, as it can be detrimental to treatment adherence and self-care tasks and decrease overall quality of life.5, 6 Furthermore, such distress is associated with increased likelihood of poor lifestyle behaviors and health outcomes, including smoking, physical inactivity, and obesity,7 which in turn have negative sequelae.
Few studies have assessed the prevalence of psychological distress specifically for colorectal cancer survivors, with prevalence estimates ranging from 12% to 37%. In a cross-sectional study of 494 mostly newly diagnosed colorectal cancer patients, 32% were found to reach caseness for psychological distress.6 A survey of 128 colorectal cancer survivors, who were on average 40 months postsurgery, found that 8% of participants met caseness for anxiety and 37% for depression.8 Nordin et al.4 interviewed 104 colorectal cancer patients at the time of diagnosis and found that, for those with localized disease, 18% met caseness for anxiety and 12% for depression, with incident cases higher among those with advanced disease (37% for anxiety and 21% for depression). A study of 154 colorectal cancer survivors aged 65 years or more, assessed 4 to 6 weeks postsurgery, found that 18% met the cutoff for depressed mood.10 Finally, Ramsey et al.9 surveyed 227 colorectal cancer patients who were at least 5 years postdiagnosis and found that 14% were depressed. Limitations of these studies include the use of cross-sectional designs4, 6, 8; nonrepresentative samples,6, 8 absence of response rate,10 or disease stage data6; and limited sample sizes for the prospective studies.9 Thus, the prevalence of psychological distress among colorectal cancer patients, and the trajectory of this distress over time, is not yet well described.
A number of variables have been found to predict poorer adjustment after cancer, including sociodemographic, medical, psychological, and lifestyle factors. Sociodemographic and medical factors predictive of poorer psychological outcomes include younger age at diagnosis, being female, having dependent children in the home, economic adversity, lower levels of education, later-stage disease, and high physical burden treatment approaches.10–14 Psychological variables include predisposing factors such as dispositional optimism; adaptive approaches such as positive threat appraisals; and coping resources such as social support.11, 15–17 Dispositional optimism is proposed as a stable personality characteristic whereby a person has generalized outcome expectancies that lead them to expect either positive or negative outcomes to result from events.18 For people with breast and prostate cancer, optimism has been consistently linked to positive outcomes such as better psychological and social adjustment, less anxiety, depression, and mood disturbance, lower decision-related distress, and more active coping.16, 19, 20 Cancer threat appraisal involves the cognitive assessment of the impact that cancer has had on personal identity and on the future. A less negative cancer threat appraisal has been found to predict lower decision-related distress and better psychological adjustment in prostate and breast cancer patients.16, 21, 22 Social support that encompasses emotional, informational, and practical support may mitigate the adverse psychological effects of a cancer diagnosis by acting as a coping resource and is associated with lower distress after breast and colorectal cancer.15, 17, 23 In combination these psychosocial variables have been found to concurrently predict up to 38% of baseline mood in breast cancer patients; however, such effects typically diminish at subsequent prospective assessments where baseline distress is accounted for.23
Finally, physical activity has been found to predict improved mental health in the general population24 and better quality of life in cancer survivors.25, 26 The mechanisms underlying this effect are not clear; however, it is proposed that physical activity may act on multiple physiological and psychological systems. First, physical activity may induce changes in neuroreceptor levels of monoamines or endogenous opiates, but these effects tend to be limited to participation in vigorous activity.27 Second, psychological aspects of physical activity, such as increased social interaction, feelings of self-mastery and self-efficacy, and improved ability to deal with stress may also improve mental health.24, 28
Carver et al.29 recently demonstrated that for breast cancer patients different aspects of quality of life have different antecedents. Specifically, medical variables were related to financial problems and appearance worries; social support was related to quality of life at the bivariate but not multivariate level; optimism by contrast consistently predicted quality of life across multiple domains. To date, the relative contributions of sociodemographic and medical variables, psychological influences, and physical activity on psychological distress have not been compared. Identifying variables that predict positive psychological outcomes for people diagnosed with colorectal cancer is important for supportive care interventions. First, variables that have strong predictive power may have efficacy as screening targets to identify patients at higher risk of poorer long-term outcomes so that they can be triaged to more intensive psychosocial intervention.30 Second, significant predictor variables that are potentially modifiable can be targeted in psychosocial interventions. In this regard, given the recent increase in focus on physical activity interventions for cancer patients,31, 32 there is a need to assess the strength of physical activity as a predictor of adjustment, relative to sociodemographic, medical, and psychological variables.
Accordingly, the present study aimed to describe the prevalence of psychological distress among a population-based sample of colorectal cancer survivors, and second to identify sociodemographic, medical, psychological, and lifestyle predictors of this distress.
MATERIALS AND METHODS
Participants and Procedures
All persons resident in Queensland, Australia, with a histologically confirmed diagnosis of a first, primary colorectal cancer, notified to the state-based cancer registry between 1 January 2003 and 31 December 2004, were eligible for the study. Eligibility criteria also included speaking English, having no hearing, speech, or cognitive disabilities that would prevent completing a telephone interview, and being aged between 20 and 80 years. Ethical approval was obtained from the University of Queensland.
A flow diagram describing recruitment to and participation in the study is presented in Figure 1. The treating doctors of 3626 eligible patients were approached in writing for permission to contact their patients regarding the study. The 3182 patients for whom consent from the physician was obtained were written to about the study. All patients who agreed to take part and provided their written, informed consent were posted a questionnaire for self-administration and were telephoned by a trained interviewer at approximately 6 (Time 1) and 12 months (Time 2) postdiagnosis. Data were collected by these 2 methods so that neither the telephone interview nor questionnaire was too burdensome for the participants. At each assessment participants who had not returned their self-administered questionnaire within 6 weeks were followed-up with 2 reminder telephone calls, 2 weeks apart. Optimism and sociodemographic and medical variables were only measured at the Time 1 assessment. All other measures were administered at both timepoints.
Telephone interview measures
Current marital status, educational attainment, and health insurance status were self-reported during the interview. Sex and age had been ascertained previously from the cancer registry records.
Participants also reported the modes of treatment they had received, whether a stoma had been formed, their current comorbid conditions, and height and weight (to determine body mass index [BMI]) at the time of interview. Site and stage of disease were collected from pathology reports held within the cancer registry.
Physical activity was assessed using questions based on the Active Australia Survey, a validated instrument used to monitor leisure-time physical activity participation in the Australian adult population.33 Based on national physical activity guidelines that recommend the accumulation of at least 150 minutes of moderate-intensity activity per week,34 participants were categorized as being either inactive (0 minutes per week), insufficiently active (1–149 minutes per week), or sufficiently active (150 minutes or more per week).
Self-administered questionnaire measures
The Brief Symptom Inventory-18 (BSI-18)35 was used to assess psychological distress. Respondents were asked to report the degree of distress caused by each symptom during the previous week on 5-point scales ranging from 0 (not at all) to 4 (extremely). The items of the BSI-18 can be grouped evenly into 3 subscales (anxiety, depression, somatization), and the sum of all responses forms a global severity index (GSI). Used with a heterogeneous cancer sample, internal reliability was found to be 0.89, and the BSI-18 total score was significantly correlated with the GSI of the full BSI (r = 0.84, P < .001).35
Dispositional optimism was measured using the Revised Life Orientation Test.36 The 10 items (4 are fillers) are scored on a 5-point scale where participants rate their agreement with each item from 0 (strongly disagree) to 4 (strongly agree). Negatively worded target items are reverse-scored, then responses to target items are added for an overall optimism score, with a higher score indicating greater dispositional optimism.
Cancer threat appraisal
The Constructed Meaning Scale measures cognitive assessment of the implications that colorectal cancer has had on participants' identity, relations, and perceived future.37 Participants are asked to indicate their agreement with 8 items using a 4-point scale from 1 (strongly disagree) to 4 (strongly agree). Negatively worded items are reverse-scored, with responses summed to give an overall score where a lower score indicates a more negative cognitive appraisal. The measure has good internal consistency (α 0.81).
Social support was assessed using the 6 items from the Brief Social Support Questionnaire.38 Participants are asked to rate how satisfied they are with various aspects of social support on a 5-point scale ranging from 1 (not at all satisfied) to 5 (very satisfied). This scale demonstrates high internal reliability (0.92–0.93).38
Raw scores from both timepoints for the BSI-18 symptom subscales and GSI were converted to standardized T-scores. A T-score of 63 or more indicates a positive score for each specific scale. Overall caseness for the BSI-18 was defined as a GSI T-score of 63 or higher, or if any 2 symptom subscale T-scores were 63 or higher.35 Hence, participants were categorized as positive for depression, anxiety, somatization, GSI, and BSI-18-defined caseness.
We used logistic regression to identify independent variables significantly associated with meeting BSI-18-defined caseness at Time 2. The first model was created to determine the contribution that baseline measures of predictor variables made to the BSI-18 GSI score at Time 2. The second logistic regression assessed the extent to which cancer threat appraisal, social support and physical activity concurrently predicted psychological distress at Time 2.
In all, 1822 participants completed both the telephone interview and self-administered questionnaire, giving an overall response rate of 53.2% (only those found to be ineligible were removed from the denominator). The characteristics of the study participants are described in Table 1. To examine the representativeness of our sample, we compared participants with those who were eligible, but not recruited to the study, across sex, age category, site, and stage of disease (the variables available to us through the cancer registry for nonparticipants). There was no difference in sex distribution between participants and nonparticipants. However, our sample underrepresents older (70–80 years) colorectal cancer survivors, those with rectal cancer, and those with more advanced disease (χ2 test, P < .05 for each).
|Age categories, y|
|Never married||4%||Not available|
|Less than 8 y||14%||Not available|
|12 y/Technical college||33%||—|
|Private insurance||60%||Not available|
|No private insurance||40%||—|
The proportions of participants who scored above a T-score of 63 on any of the BSI-18 scales are reported in Table 2. This table also presents the proportion of participants who met caseness for distress, defined as a GSI T-score of 63 or higher, or if any 2 symptom subscale T-scores are 63 or higher.35 Of the 143 participants who met caseness for distress at Time 1, 55 (38%) remained highly distressed at Time 2. These individuals accounted for just under half (47%) of the 116 who were distressed at Time 2.
|Participants in BSI-18 clinical range|
|6 Months postdiagnosis||12 Months postdiagnosis|
|Men n = 1089||Women n = 733||Total n = 1822||Men no. (%)||Women no. (%)||Total no. (%)|
|Anxiety dimension*||91 (8.5%)||40 (5.6%)||131 (7.4%)||49 (5.8)||46 (7.7)||95 (6.7)|
|Depression dimension†||95 (8.9%)||38 (5.3%)||133 (7.5%)||62 (7.4)||39 (6.7)||101 (7.1)|
|Somatization dimension‡||161 (15.3%)||106 (14.9%)||267 (15.1%)||110 (13.2)||76 (13.0)||186 (13.1)|
|Global Severity Index (GSI)§||92 (8.9%)||46 (6.6%)||138 (8.0%)||66 (7.9)||47 (8.0)||113 (8.0)|
|Caseness (TGSI ≥ T63 or T2DIM ≥ T63)∥||96 (9.3%)||47 (6.7%)||143 (8.3%)||69 (8.3)||47 (8.0)||115 (8.2)|
We initially entered baseline physical activity and all baseline sociodemographic, medical, and psychological variables into a logistic regression model to determine which were independently associated with meeting caseness for distress at 12 months postdiagnosis. Variables were then removed in a stepwise fashion, until only those that had statistically significant associations (overall or within variable) with distress remained. The final model is presented in Table 3.
|BSI-18 case at T1|
|No. of comorbidities at T1|
|4 or more||10.60||3.19, 35.24|
|Optimism at T1 continuous variable||0.93||0.84, 0.99||<.001|
|Cancer threat appraisal at T1 continuous variable||0.92||0.86, 0.98||<.001|
|Social support at T1 continuous variable||0.94||0.91, 0.97||<.001|
A second regression model was then created to determine the contribution that concurrent physical activity, cancer threat appraisal, and social support made to global distress score at 12 months postdiagnosis (Table 4). The strongest predictor of distress at Time 2 was again distress at Time 1. Comorbid conditions and optimism (assessed at Time 1) and Time 2 cancer threat appraisal were also associated with distress at Time 2.
|BSI-18 case at T1|
|No. of comorbidities at T1|
|4 or more||11.59||3.19, 35.24|
|Optimism at T1 continuous variable||0.94||0.88, 0.99||<.001|
|Cancer threat appraisal at T2 continuous variable||0.85||0.80, 0.90||<.001|
The present study is the first to document prospectively the prevalence of psychological distress in a population-based sample of colorectal cancer survivors. By comparison to previous studies6, 8–10 we found the prevalence of clinically significant levels of distress to be low at 6 and 12 months after diagnosis. The reason for this disparity is unclear, but may relate to different sampling frames: our participants were recruited from a population-based cancer registry rather than cancer treatment centers, and so may more realistically represent the distress burden in the community. As well, participants in our study were several months postdiagnosis at recruitment, and so would be expected to have moved past the high distress typically experienced at diagnosis. Consistent with this, other researchers have suggested that for colorectal cancer patients psychological distress is greater preoperatively in anticipation of treatment, with significant improvements in the 3 to 8 months after treatment.39, 40 The study response rate of 53.2% may also have impacted on distress prevalence, both in terms of underrepresenting those with more advanced cancers and the possibility that more distressed patients declined to participate and so are underrepresented. However, the data presented come from a large, population-based sample of colorectal cancer survivors collected using a well-validated and recommended screening tool.41 Accordingly, we propose that, whereas our results may underestimate the distress likely to be found in this population group, the findings highlight the need for oncology service providers to identify the high-distress minority and more efficiently target them for intensive psychological intervention.30, 41, 42
Psychological distress can be detrimental to decisions about treatment, symptoms, adherence, and outcomes.43 In addition, ongoing distress may significantly increase direct and indirect healthcare costs to the individual, as well as the health system more broadly.6, 44 Clinical practice guidelines for psychosocial care in oncology in North America and Australia currently advise that cancer patients be routinely screened for distress to facilitate triage to appropriate support.45–47 This approach directs care to those who most need it and is proposed as more cost-effective.41 As in previous research,9, 16, 42, 48 the most powerful prospective predictor of psychological distress 12 months after diagnosis was previous distress. However, using distress alone at the initial 6 months assessment to identify patients at risk of ongoing distress would have missed 49% of longer-term distressed participants. This finding highlights the need for regular distress screening for cancer patients throughout the illness continuum.
When considering the relative importance that sociodemographic, medical and psychological variables, and physical activity made to psychological distress, comorbidities and the psychological variables of optimism and threat appraisal were predictive, but the effects were not strong. It has been suggested that the predictive effects of optimism on longer-term adjustment are partly mediated by the association of optimism with initial distress15 and this may explain these results. Cancer threat appraisal emerged as a significant and potentially modifiable variable, providing further support for the inclusion of social cognitive components, such as outcome expectations, in psychosocial interventions for cancer patients.49 Outcome expectations encompass a person's perceptions of their current and potential future outcomes, both physical and psychosocial, that may result from a behavior or action. Nezu et al.50 propose that a cancer patients' subjective appraisal of their cancer strongly influences their responses, such that magnifying the problems to be faced may hinder effective coping. Thus, it is likely that interventions that include cognitive challenging of highly negative threat appraisals will be more effective.
In the present study physical activity was not significantly associated with distress. This may relate to exercise intensity: specifically, levels of vigorous activity reported by our sample were low, with only 3.7% of participants participating in any vigorous-intensity activity on a weekly basis.51 There is evidence to suggest a dose-response relation between levels of physical activity and psychological distress.24 Problematically, achieving sufficient intensity and amount of physical activity may be difficult for cancer survivors who are still experiencing negative sequelae of the disease. As well, cancer risk increases with age such that many cancer patients will have comorbidities that make regular exercise difficult to achieve. However, cancer survivors who do successfully modify their lifestyle to include regular physical activity may reduce their risk for other chronic disease and cancer recurrence.52–54 Hence, there remains a strong imperative to develop ways to promote regular physical activity for this population.
This study is the first to assess the prevalence of psychological distress in colorectal cancer patients using a population-based approach and is unique in its consideration of sociodemographic, medical, psychological, and lifestyle-related variables as predictors of psychological distress in colorectal cancer survivors. The data confirm the need for regular distress screening for cancer patients, but also highlights the limitations of distress screening if only applied at a single timepoint. The data also support the need for further indicators to be developed, adjunctive to current distress, to more accurately identify cancer patients who need in-depth psychosocial intervention and those will adjust effectively with their own personal resources and standard healthcare services.
- 1GLOBOCAN 2002: Cancer Incidence, Mortality and Prevalence Worldwide. IARC CancerBase No. 5 v. 2.0. Lyon, France: IARC Press; 2004., , , .
- 4Quality of life in long term survivors of colorectal cancer. Am J Gastroenterol. 2002; 97: 1228–1234., , , , .Direct Link:
- 24United States Department of Health and Human Services, physical activity and health: a report of the Surgeon General. Atlanta, Georgia: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, 1996.
- 33Australian Institute of Health and Welfare. The Active Australia Survey: a guide and manual for implementation, analysis and reporting. Canberra: AIHW; 2003.
- 34Department of Health and Aged Care. National Physical Activity Guidelines for Australians. Canberra: Australian Government; 1999.
- 35Brief Symptom Inventory 18: Administration, Scoring and Procedures Manual. Minneapolis: National Computer Systems; 2000., .
- 45Canadian Association of Psychosocial Oncology. Standards: Canadian Association of Psychosocial Oncology. Montreal: Canadian Association of Psychosocial Oncology; 1999.
- 46National Breast Cancer Centre, National Cancer Control Initiative. Clinical Practice Guidelines for the Psychosocial Care of Adults with Cancer. Canberra: National Health and Medical Research Council; 2003.
- 47National Comprehensive Cancer Network, Distress Management. Rockledge, PA: National Comprehensive Cancer Network; 2007.
- 50A problem-solving approach helping cancer patients cope. Washington, DC: American Psychological Association; 1998., , , , .
- 52The 2006 Nutrition, Physical Activity and Cancer Survivorship Advisory Committee; American Cancer Society. Nutrition and physical activity during and after cancer treatment: an American Cancer Society guide for informed choices. CA Cancer J Clin. 2006; 56: 323–353., , , et al.;