| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLE |
From the Department of Epidemiology and Public Health, University College London, UK.
Address correspondence and reprint requests to Jane Wardle, PhD, Cancer Research UK Health Behavior Unit, Department of Epidemiology and Public Health, University College London, Gower Street, London, WC1E 6BT, UK. E-mail: j.wardle{at}ucl.ac.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Methods: Sigmoidoscopy screening for colorectal cancer was identified as a potential stressor. A subset of participants (N = 3535) from the U.K. Flexible Sigmoidoscopy Trial completed pre- and postscreening questionnaires regarding psychologic well-being. All trial participants were sent a postscreening questionnaire after 3 months (post-flexible sigmoidoscopy [FS] sample, N = 29,804), including measures of distress (the General Health Questionnaire), anxiety (State-Trait Anxiety Inventory), a single-item measure of bowel cancer worry, and a 6-item measure of positive consequences of screening. Socioeconomic status (SES) was coded from postcodes with the Townsend Index. SES differences in changes in emotional well-being over the course of screening were evaluated in the longitudinal sample. SES differences in postscreening well-being in relation to screening outcome were evaluated in the post-FS sample.
Results: Bowel cancer worry and anxiety were higher in lower SES groups before screening. Both reduced after screening, but there were no SES differences in the change. In the post-FS sample, there was an SES gradient in anxiety but not in distress. Lower SES groups indicated more positive reactions. There were no interactions between SES and screening outcome for any indicator of well-being.
Conclusions: Lower SES was associated with worse psychologic well-being before and after screening, but lower SES participants did not show any differentially greater adverse reactions compared with higher SES participants. Moderately stressful experiences in everyday life do not necessarily more unfavorably affect those with fewer educational and economic resources.
Key Words: SES • bowel cancer • colorectal cancer • screening • psychologic well-being • stress • distress
Abbreviations: SES = socioeconomic status; FS = flexible sigmoidoscopy.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Few studies have examined SES differences in emotional reactions to everyday stresses and strains, although a number have investigated exposure to severe trauma. Lower SES appears to be a risk factor for developing posttraumatic stress disorder (9), with evidence coming from responses to a variety of traumatic situations, including assault (10), air disaster (11), and treatment for breast cancer (12). A limitation of these studies is that traumatic exposures are not always uniform for all participants; people are exposed to different amounts of stress depending on the impact that the incident has on, for example, their financial circumstances, which may compound the effects of the stressor. People from less-privileged social backgrounds are likely to have fewer economic or social resources to deal with a major traumatic event, so the total adverse exposure is greater, making it difficult to interpret any differential emotional responses between SES groups.
The effect of lower psychosocial resources is an impaired capacity to respond to new difficulties (13). A weakened ability to respond adequately to novel stressors has been referred to as "reactive responding" (14), a concept characterized by a state of chronic vigilance, a lack of options, lack of opportunity for learning and skill development, short-term goals, and an emotionally charged response. The present study explores one part of this concept in relation to social background, testing the idea that lower SES groups will have a more emotionally charged response than higher SES groups to a stressful experience. The study uses emotional reactions to a moderately stressful medical examination to test this hypothesis.
The stressor investigated in this study is a screening examination for early detection of colorectal cancer. Colorectal screening is not currently included in the national screening program in the United Kingdom, but a randomized, controlled trial of the efficacy of flexible sigmoidoscopy (FS) is in progress (15). This article uses data from this trial. FS screening involves endoscopic examination of the distal colon to identify precancerous changes (polyps) and early, asymptomatic lesions. FS is followed up with colonoscopy if there are indications of disease in the proximal colon. The bowel preparation for the FS examination (a self-administered enema) is a novel procedure for most adults, and the endoscopy itself can be embarrassing and uncomfortable. In our study, people who were found to have polyps waited another 2 weeks before being given the "all-clear," whereas those needing colonoscopy waited an average of 6 weeks for the appointment, and then endured another more invasive and uncomfortable procedure. The worry associated with the possible detection of cancer also added to the potential stress. Previous research into the psychologic impact of cancer screening suggests there is increased distress both in anticipation of attending screening and while waiting for follow-up tests (16,17). Longer-term adverse consequences have been identified among women who are recalled for further investigations after mammography screening (18,19). We judged that the FS screening procedure would be mildly stressful and that the experience of people recalled for colonoscopy would be moderately stressful, thus providing an appropriate setting to look for SES differences.
We had data on anxiety and cancer worry before and after the screening examination(s) in a subset of trial participants (the longitudinal sample) and were able to compare changes in emotional well-being over time between SES groups. We expected the lower SES groups to be more anxious beforehand and to find the examination procedure more threatening, resulting in more adverse emotional outcomes in these groups. Data on psychologic well-being 3 months after screening were available for the whole screening group (the post-FS sample). This allowed us to examine SES differences in the well-being in people who received negative (clear) results compared with those who had polyps detected and removed during FS and those who were referred for colonoscopy. We hypothesized that the screening experience would be most stressful for those who were referred for colonoscopy, and predicted that there would be an interaction between SES and screening outcome in postscreening well-being.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Design and Procedure
Background questionnaires had been sent to a randomly selected subset of participants (N = 5942 of 57,254, approximately 10% of the randomized sample) approximately 5 to 6 months before they came for the screening examination in 3 centers: Portsmouth, Oxford, and Swansea. This group comprised the smaller sample used to examine changes in emotional well-being from before to after the examination (the longitudinal sample). Follow-up questionnaires were sent to all screening participants, except those diagnosed with cancer (N = 140), 3 months after the first examination (N = 40,534, the post-FS sample).
The Screening Examination
The screening examinations were carried out in dedicated units in local hospitals by medical endoscopists. The procedure took approximately 5 minutes, and the participants could watch the progress of the scope and any polyp removal on a screen. The design of the study is shown in Figure 1. Seventy-two percent (N = 29,406) had no polyps detected and were given a letter confirming the negative result at the end of the examination. A total of 9607 (23.6%) had small polyps detected and removed at FS. A total of 2131 (5.3%) were referred for colonoscopy either at the time of the FS screen or because the polyps that were removed at FS showed high-risk pathology. The average time from FS to completion of colonoscopy was 6 weeks. All screened participants, except those diagnosed with cancer, were mailed a follow-up questionnaire 3 months after the FS examination to assess the psychologic impact.
|
Measures
Socioeconomic Status
SES was indexed using an area-level measure of socioeconomic deprivation, which could be derived from postcode information within England and Wales. The Townsend Index (21) is based on levels of car ownership, housing tenure, unemployment, and overcrowded living conditions as recorded in the 1991 census across enumeration districts (on average, approximately 150 households). Each enumeration district receives a score based on its relative levels of deprivation. A score of zero represents the national average with negative values representing less deprivation (higher SES) and positive values representing more deprivation. Postcode information was collected for each individual and linked to enumeration districts; the Townsend score for the district was assigned to the individual. For the purposes of analyses, the scores were grouped into quintiles to form 5 categories with "1" indicating lower socioeconomic deprivation and "5" signifying higher deprivation.
Screening Outcome
The 3 possible screening outcomes were: 1) flexible sigmoidoscopy with a negative (clear) result (FSN), 2) flexible sigmoidoscopy in which small polyps were found and removed (FSP), and 3) flexible sigmoidoscopy followed by colonoscopy to examine the proximal bowel and remove large and numerous polyps (FSC). The small number of participants (0.34%, N = 140) who received a cancer diagnosis were not included in these analyses.
Bowel Cancer Worry
Worry about bowel cancer was assessed with the single item: "How worried are you about getting bowel cancer?" (22). Response options were "not at all worried," "a bit worried," "quite worried," and "very worried."
Psychologic Well-Being
The shortened (6-item) state version (23) of the State-Trait Anxiety Inventory (STAI (24)) was included in both the prescreening and postscreening questionnaire. The short form has good concurrent validity compared with the full form and good internal reliability (
= 0.82) (23). Item scores were summed to produce a scale from 6 to 24 with higher scores indicating greater anxiety.
At follow up, participants were additionally sent the 12-item General Health Questionnaire (GHQ-12 (25)). The GHQ is designed as a self-administered test to assess psychologic distress in primary care patients (26). It assesses present state in relation to usual state eg, "in the last three months have you been able to face up to your problems?" The items have 4 response options: "better/more than usual," "same as usual," "less than usual," and "worse/much less than usual." The GHQ-12, which has good reliability and validity (25), produces scores from 0 to 24, with higher scores indicating greater distress.
Positive consequences of screening were also measured at follow up using 6 items from the Psychological Consequences of screening Questionnaire (PCQ (27)). All items started with the stem: "Do you think that your experience of having the Flexi-Scope test has... ?" followed by: "given you a sense of reassurance that you do not have bowel cancer," "made you feel more able to do the things that you normally do," "made you feel more hopeful about the future," "made you feel less anxious about bowel cancer," "made you get on better with those around you," or "given you a greater sense of well-being." Response options were "not at all," "a little bit," "quite a bit," or "a great deal." Items were scored 1 to 4 and scores were summed to produce a scale from 4 to 30 with higher scores indicating more positive consequences. The scale had good internal reliability (Cronbach's = 0.86) in this sample.
Analyses
In the longitudinal sample, repeated-measures analysis of variance with deprivation quintiles as the between-subjects factor, time (pre- to postscreening) as the within-subjects variable, and screening outcome included as a control variable, was used to test for the hypothesized SES by time interactions for state anxiety and bowel cancer worry. For the analyses of SES differences in response to the different screening outcomes, we used the larger post-FS sample. Results were analyzed using analysis of variance with SES and screening outcome as the independent variables and age and gender included as control variables. Family history was not used in the analyses as a control variable because it was unrelated to SES and did not alter the results when included in the analyses. If, as we hypothesized, a positive FS test followed by colonoscopy is experienced as more stressful for lower SES participants, then we would expect to find an outcome by SES interaction for anxiety, psychologic distress (GHQ), and PCQ.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
The overall mean level of anxiety at baseline was 9.83 (SD = 3.49). There was a significant main effect of SES for both anxiety (F [4,3387] = 14.76, p <.001) and bowel cancer worry (F [4,3373] = 4.11, p <.01), with more deprived groups being more anxious and the more worried about developing bowel cancer both before and after screening (see Figs. 2 and 3). Anxiety was reduced (F [1,3389] = 27.85, p <.001) from before (STAI mean = 9.83; standard error [SE], 0.06) to after screening (STAI mean = 9.49; SE, 0.06), but there was no evidence for a differential effect across levels of SES. Bowel cancer worry also decreased significantly over the period of screening (F [1,3383] = 55.96, p <.001) (prescreening worry mean = 1.97; SE, 0.01, postscreening worry mean = 1.87; SE, 0.01), but again the interaction with deprivation was not significant.
|
|
The average age of participants was 60.5 years, with 14,876 women (49.9%) and 14,928 men (50.1%). Men were more likely than women to have a positive result (chi-square = 697.95, df = 2, p <.001) as expected from their higher incidence of bowel cancer. Participants who had polyp removal (FSP) or colonoscopy (FSC) were slightly older (60.60 vs. 60.42 years) than those with an initially clear result (F [2,32214] = 16.71, p <.001). SES was also significantly related to screening outcome (chi-square = 15.85, df = 1, p <.01), although the effect was small. A total of 21.8% of the most deprived group were referred for colonoscopy compared with only 19.6% in the most affluent group.
Scores on the STAI, GHQ, and PCQ by SES and screening outcome are shown in Table 1. Overall, the levels of distress (GHQ) and anxiety (STAI) in this sample were low compared with other community and population samples (23,28,29). Distress was not associated with SES, but there was a graded association between SES and anxiety (F [4,28951] = 68.40, p <.001). Ratings of positive consequences of screening (PCQ) were also linked to deprivation (F [4,28997] = 195.91, p <.001); but lower SES groups reported more positive consequences of screening (Table 1).
|
Screening outcome was related to psychologic well-being, but contrary to expectation, psychologic well-being was highest in those who had the colonoscopy. This group had the lowest scores for anxiety (F [2,28239] = 10.38, p <.001) and distress (F [2,28239] = 10.38, p <.001). There were no differences in positive consequences scores (F [2,29311] = 2.39, NS) between the outcome groups.
Interactions between SES and outcome group were not significant for any of the psychologic variables: STAI (F [8,28952] = 1.35, p = .21), PCQ (F [8,28998] = 1.50, p = .15), or GHQ (F [8,27940] = 1.42, p = .18; Table 1).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
We anticipated that the subset of participants referred for colonoscopic examination would experience greater stress because of the additional medical investigation and the greater delay before receiving a clear result, and we expected that this would create relatively more distress in the lower SES groups. The post-FS sample provided an adequate sample size to explore SES differences by screening outcome. Overall, the levels of distress and anxiety were low postscreening. More importantly, they were lower for the 2 outcome groups who had more pathology detected (FSP and FSC outcomes). People referred for colonoscopy also reported the most positive consequences of taking the test. One possible explanation is that people who had a colonoscopy experienced the greatest "relief" when the results were received. If relief is positively valued, then it might, at least transiently, give the best outcome. Results from other screening studies have also shown that on receipt of a negative test, there can be an immediate decrease in distress, again defined in terms of relief, because the threatening situation is positively resolved (31–34). Contrary to our prediction, lower SES groups were not any more adversely affected by being referred for colonoscopy. This gives no support to the idea that low SES groups have an impaired capacity to respond to new difficulties (13), at least in terms of their psychologic health. It is possible that lower SES participants benefited to a greater extent from the additional attention and health information provided at the time of colonoscopy than did others.
Across the groups as a whole, lower SES was associated with identifying more positive psychologic consequences of taking the test despite having higher levels of anxiety and worry about cancer. Previous research has shown that individuals may simultaneously feel distress and well-being (35,36), but the extent to which demographic and health status outcomes relate to levels of positive experiences has not been fully explored. One recent study found that "benefit-finding" increased with more severe disease stage, lower SES, and higher negative affect in women with breast cancer (37). The authors argued that the more severe the threat, the greater the need to mobilize resources to minimize its impact, and finding benefit in the experience is one way of doing this. If lower SES groups typically experience higher levels of life stress, they may have more experience of using this strategy to minimize impact. Supporting evidence for this is that people who experience more traumatic life events report more benefits than those who experience fewer traumas (38,39).
The results do not support our original hypotheses regarding psychologic reactions to stress in low SES groups, raising the possibility that the conditions in which we tested the hypotheses were not satisfactory. A limitation of the study design was that only people who had expressed an interest in the test were subsequently invited to take part in the trial. This introduced a bias in the selection of the participants according to SES. Higher SES participants were more likely to express an interest in taking the FS test and so were more likely to be invited to take part (40) and also more likely to attend (41). This same bias is seen in all national screening programs as well as in countries that use opportunistic screening (42,43). However, it is not clear that this SES gradient in participation should explain why the lower SES groups had a more positive reaction to the test. A second limitation is that the timing of the post-FS survey may have been too long after the screening took place, and any transient differences in reactions to the test may have subsided during the 3 months after the test. However, in the colonoscopy group, follow up was only 4 to 6 weeks after the result, on average, and less in some groups. In addition, all participants in the study had low levels of distress at follow up. This could be because trial situations introduce a "Hawthorne effect" whereby clinic staff provide greater reassurance and information than would be usual outside of a trial setting. A review of this effect found only limited evidence in support of this idea (44). The low levels of distress also suggest that the stress experienced by people was not severe enough to allow any differences by SES to be displayed. There is increasing evidence that a clear screening test, even if more than 1 investigation is required before the all-clear, is a positive experience for some people, perhaps as a reprieve from persistent worry or in terms of relief after the stress of the procedure, as previously discussed. Future studies should address these limitations by choosing a situation that is both nonvoluntary and highly stressful.
This research is important despite these limitations. It shows that, of the people who do come to take these tests, low SES participants experience a reduction in distress similar in magnitude to high SES participants and may even perceive more positive consequences of participating in screening. There is no reason to think that the same would not be true for similar routine screening situations. This study demonstrates that despite our expectations, lower SES groups were able to gain as much, if not more, psychologic benefit from undergoing screening as more privileged groups. Few studies have examined SES differences in psychologic reactions to stressors. These results suggest that low SES groups may be more resilient to novel stressors than previously expected (14). In the future, it will be useful to pursue this idea and discover whether these findings are consistent across different situations.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Received for publication May 12, 2004; revision received October 6, 2004.
DOI:10.1097/01.psy.0000155665.55439.53
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
= 1; * = 2; 
= 3; X = 4, 
= 5.