Psychosomatic Medicine Tips for Better Browsing
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Published online before print August 31, 2007, 10.1097/PSY.0b013e31814d4e6a
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by White, V. M.
Right arrow Articles by Giles, G. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by White, V. M.
Right arrow Articles by Giles, G. G.
Related Collections
Right arrow Personality
Right arrow Cancer
Psychosomatic Medicine 69:667-674 (2007)
© 2007 American Psychosomatic Society

ORIGINAL ARTICLES

Is Cancer Risk Associated With Anger Control and Negative Affect? Findings From a Prospective Cohort Study

Victoria M. White, PhD, Dallas R. English, PhD, Hamish Coates, PhD, Magdalena Lagerlund, PhD, Ron Borland, PhD and Graham G. Giles, PhD

From the Centre for Behavioural Research in Cancer (V.M.W.), The Cancer Council Victoria, Australia; Cancer Epidemiology Centre (D.R.E., G.G.G.), The Cancer Council Victoria, Australia; School of Population Health (D.R.E., G.G.G.), University of Melbourne, Australia; Department of Epidemiology and Preventive Medicine (D.R.E., G.G.G.), Monash University, Australia; Centre for the Study of Higher Education (H.C.), University of Melbourne, Australia; Department of Epidemiology and Biostatistics (M.L.), University of Western Ontario, Canada; VicHealth Centre for Tobacco Control (R.B.), The Cancer Council Victoria, Australia.

Address correspondence and reprint requests to Victoria M. White, Centre for Behavioural Research in Cancer, The Cancer Council Victoria, 1 Rathdowne Street, Carlton Vic 3053, Australia. E-mail: Vicki.White{at}cancervic.org.au


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Objective: To examine the associations between anger control and negative affect and the risk of five common cancers and total cancers. Possible associations between emotional states and the risk of cancer have long been postulated.

Methods: Prospective cohort study with average follow-up of 9 years. A total of 19,730 adults (99% aged between 40 and 69 years) answered questions on negative affect and anger control at baseline. A total of 1952 cancers were diagnosed, including 352 breast cancers, 318 prostate cancers, 88 lung cancers, 280 colorectal cancers, and 261 melanomas.

Results: After adjustment for potential confounders, there was no significant association between anger control or negative affect and risk of breast cancer, melanoma, or total cancers. Weak associations were identified between anger control and prostate cancer, hazards ratio (HR) 1.17 (1.04–1.30) for a 1-unit increase in the standardized scale, negative affect, and lung cancer, HR 1.24 (1.01–1.52) and colorectal cancer, HR 1.14 (1.01–1.28). There was no evidence of an interaction effect between anger control and negative affect.

Conclusions: Results suggest that anger control and negative affect are not associated with breast cancer, melanoma, or total cancer risk, although they may have a small role in risk of prostate, colorectal, and lung cancer. Although more research is needed to confirm these latter associations, the results suggest that if affective states are associated with cancer development, the association may differ for different cancers and argue against the use of total cancer as an outcome measure for studies in this area.

Key Words: cancer risk • cohort study • anger control • negative affect

Abbreviations: BMI = body mass index; CECS = Courtauld Emotional Control Scale; CI = confidence interval; HR = hazards ratio; HRT = hormone replacement therapy; MCCS = Melbourne Collaborative Cohort Study; MMPI = Minnesota Multiphasic Personality Inventory; PANAS = Positive And Negative Affect Scale.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Psychological factors, including stress and depression, are often cited as a cause of cancer by cancer patients (1–4). Although these beliefs may reflect a desire to introduce an element of control into cancer risk, they may also reflect research suggesting that negative emotional states can influence immunological responses and hormone levels, which in turn have been found to influence carcinogenic processes. Research examining the association between psychological factors and cancer risk has focused mostly on breast cancer (5–19) but some studies have examined associations with all cancers (20–30), colorectal cancer (31,32), and lung cancer (33). Reviews of this literature have generally concluded that personality dimensions such as extraversion and neuroticism and the experience of negative life events are not associated with cancer risk (34–38). These reviews have generally found a marginal association between emotional control and cancer risk, suggesting that this association warrants further investigation.

Emotional control refers to the tendency to minimize emotional upset and/or suppress the expression of negative emotions—most commonly anger, anxiety, and depression. In the literature examining the association between this construct and cancer risk, emotional control has been operationalized using a number of different scales including the Repression/Sensitization scale of the Minnesota Multiphasic Personality Inventory (MMPI) (10) and the Courtauld Emotional Control Scale (CECS) (39). Findings for the different scales have been mixed. Prospective studies using the MMPI's repression/sensitization scale have found no association with cancer risk (11, 25). Whereas one study using the CECS found positive associations between cancer risk and emotional control (8), other studies have found no association (12,13,27). Methodological weaknesses limit the strength of the conclusions that can be drawn from most studies in this area. For instance, one did not control for age (40), one of the major risk factors for cancer, whereas others (12,13) used a limited prospective design where women recalled for biopsy after routine mammography were interviewed before being told their biopsy results. One truly prospective study (27) examined the incidence of cancer over a 14-year period among men aged between 65 and 85 years at baseline and found no association between emotional control as measured by the CECS and cancer risk. Although this study controlled for several known risk factors for cancer including age and smoking, its relatively small sample size (n = 590) and the small number of cancers diagnosed (n = 114) limited its ability to find associations. This study used total cancers as the outcome measure rather than examine the association between emotional control and the onset of specific cancers (e.g., prostate cancer). Because cancer is a biologically diverse disease, with some cancer types being more responsive to the endocrine or immune systems, studies that use total cancer as the outcome measure may mask any association between emotional control and specific cancers.

A review of the literature examining the association between different psychosocial factors and breast cancer onset concluded that the evidence was most supportive for an association with the emotional control dimension of anger control (35). Although anger control is part of the emotional control construct, it is also considered an aspect of the "Type C" or "Cancer Prone Personality." An individual with this personality type is thought to be cooperative, unassertive, compliant with external authorities, and controls or suppresses negative emotions especially anger (41). It has been hypothesized that people with this personality type are at an increased risk for developing cancer (41). Although recent studies have not found an association between cancer onset and the anger control subscale of the CECS (12,27), given the limitations of many studies in this area, a role for anger control in cancer onset is still uncertain.

Should anger control be a risk factor for cancer, it might be expected to play a stronger role among those who experience more negative emotions. Research examining the role of negative emotions in cancer onset has focused mainly on depression (5,10,16–18,22,24,25,28–30,33,42) and stress (14,19). Findings for the association between depression and cancer have been mixed with some showing a positive association (25,30,33), some finding no association (16,22,33,42), and one finding a negative association (17). The majority of studies assessing depression have used instruments designed to screen for clinical depression or "caseness" (e.g., Centre for Epidemiological Studies – Depression Scale (24,29), Beck Depression Inventory (16), MMPI depression subscale (10,25). One prospective study found no association between breast cancer onset and either trait anger or trait depression (5). Few studies have looked at the association between cancer onset and more general trait measures of negative affect. Negative affect is one of the two main dimensions of mood and subsumes a variety of aversive mood states including anger, fear, distress, and sadness (43). Although negative affect can be a state dimension reflecting current mood, it is also considered to behave as a trait (43,44) and has been found to show moderate levels of stability over time (45,46). As negative affect is the dimension of mood being controlled in emotional control scales, it seems appropriate to determine whether the trait measure of this construct plays a role in cancer onset either independently or in interaction with anger control.

In this paper, we examine associations between cancer onset and anger control and trait negative affect in the Melbourne Collaborative Cohort Study (MCCS). This study commenced in 1990, and although designed primarily to examine the association between diet and cancer, it provided an opportunity to examine prospectively the associations between psychosocial factors and risk of cancer. At that time, the CECS (39), particularly its anger control subscale, was identified as the most promising measure. This scale directly assesses the conscious control or inhibition of feelings of anger. It has been suggested that using the onset of total cancer as the outcome measure may reduce the likelihood of finding an association between psychological factors and cancer risk (35). We examine this by using risk of total cancer and the five most common cancers (breast, colorectal, prostate, lung cancer, and melanoma) as the outcome measures. We hypothesize that, if anger control and/or negative affect are associated with cancer onset, this association will be seen when individual cancers are the outcome measures rather than when the outcome measure is total cancers. In addition, we examine whether there is an interaction between anger control and negative affect on cancer onset. We hypothesize that the association between anger control and cancer onset will be stronger when negative affect is high.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Study Cohort
The MCCS is a prospective cohort study of 41,528 people aged between 27 and 75 years at baseline (99.3% of whom were aged 40–69 years). Recruitment occurred between 1990 and 1994 in Melbourne. Details of the study have been published elsewhere (47). Participants were volunteers recruited through electoral roll mail outs (registration to vote is compulsory for people ≥18 years in Australia), advertisements, and community announcements in local media (such as television, radio, newspapers). Participants attended an assessment center, completed a questionnaire battery, and underwent a number of physical assessments including giving blood and having their height and weight measured. Subjects gave written consent to participate and release their medical records to the investigators. The Cancer Council Victoria's Human Research Ethics Committee approved the study protocol. To obtain wide variation in participants' diets, people with either an English-speaking background or a Southern European background were targeted for study recruitment. Only people from an English-speaking background completed the psychosocial questionnaire due to concerns about differences in understanding as a function of culture and/or English proficiency. This analysis is restricted to 19,730 participants who completed this questionnaire and did not have a diagnosis of cancer before study entry.

Demographic Variables and Confounders
The baseline questionnaire collected information on biological and demographic variables that were known to be associated with the risk of cancer, including age, gender, education, smoking status, number of pack years smoked (nonsmokers were given score of 0), frequency of red meat intake, alcohol consumption, frequency of walking and vigorous and moderate exercise in the last 6 months and, for women, number of children and use of hormone replacement therapy (HRT). Body mass index (BMI) was calculated from measured height and weight.

Psychosocial Variables
The psychosocial questionnaire included the 7-item anger control subscale from the CECS (39) and a shortened revised version of the Positive And Negative Affect Scale (PANAS) (43). The CECS anger-control subscale assesses how people respond when feeling angry or annoyed (express the anger or not). Responses are made on a 4-point scale ranging from never/almost never (score = 1) to always (score = 4). The seven items were summed with higher scores indicating greater anger control. The PANAS asks questions about the frequency people feel different negative emotions, and produces a measure of negative affect (six items). For this study, we used the trait version of this scale with subjects indicating the extent they generally experience a number of different negative emotions (43). Responses were made on a 5-point scale ranging from not at all/very slightly (score = 1) to extremely (score = 5). Responses were summed with higher scores indicating higher levels of negative affect. Both measures had high internal consistency (Cronbach's {alpha} values ≥0.80) and were essentially normally distributed. The correlation between the measures was low (0.12).

Cohort Follow-Up and Case Ascertainment
Addresses and vital status were determined by record linkage to electoral rolls, Victorian death records, the National Death Index, from electronic phone books, and from responses to mailed questionnaires and newsletters. By December 31, 2003, 416 (2%) participants were known to have left Victoria after baseline and 1119 (6%) participants had died.

It has been a legal requirement since 1982 that all cancers diagnosed in the Australian State of Victoria are reported to the Victorian Cancer Registry and case ascertainment for the registry is close to 100%. Cancers diagnosed during follow-up were identified by record linkage to this registry. Information on the date of diagnosis and tumor type was recorded.

Statistical Analysis
Cox regression, with age as the time metric, was used to estimate hazard ratios (HRs) and 95% confidence intervals (CI) for the measures of anger control and negative affect and for their composite variable. For these analyses, raw scores on the anger control and negative affect scale were standardized by conversion to Z scores to allow ready comparison of results for the two variables. Separate analyses were performed for breast, colorectal, prostate, lung cancer, melanoma, and total cancer. For each cancer outcome, tests based on Schoenfeld residuals found no evidence that proportional hazards assumptions were violated for the two psychosocial measures. For each analysis, follow-up began at baseline and ended at diagnosis of the cancer of interest, death, the date last known to be in Victoria or December 31, 2003, whichever came first.

The analyses were adjusted for measured biological and behavioral factors shown to be risk factors for the specific cancer in the epidemiological literature and found to be associated with cancer onset in this data set. Specifically, for breast cancer—we adjusted for parity, BMI, alcohol consumption, education, and use of HRT; for prostate cancer— education; for lung cancer—smoking status, pack years, education, and gender; for colorectal cancer—BMI, red meat consumption, alcohol consumption, education, smoking status, pack years, physical activity, and gender; for melanoma—education, physical activity, and gender; and for total cancers—BMI, alcohol consumption, education, smoking status, pack years, physical activity, and gender. All risk factors were assessed at study entry. All analyses were repeated excluding the first 2 years of follow-up in case health issues related to as-yet-undiagnosed cancers might have influenced the subjects' scores on the measures. During the first 2 years of follow-up, 405 cancers were diagnosed (including 72 breast, 69 prostate, 18 lung, 57 colorectal, and 42 melanoma cases).

To test the hypothesis that associations between anger control and cancer onset may be more pronounced among those who experience high negative affect, we repeated the above analyses and included anger control and negative affect and a product term of these two variables to test their interaction.

Statistical analyses were performed using Stata 8.2 (StataCorp, College Station, TX). All statistical tests were two-sided.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 NOTES
 REFERENCES
 
At baseline, 61% of participants were female, 55% were aged ≥55 years (range: 40–79 years), and 70% were married or living as married. In addition, 41% had been smokers, and 10% still smoked. Of women, 20% had not had any children, and of those who have children, 9% had their first pregnancy over the age of 30 years.

Table 1 shows the association between anger control and negative affect and major demographic and confounding factors. In general, there were only small differences between the different groups in scores on anger control or negative affect. The strongest association was found for education with higher scores on anger control and negative affect found among those with the least education.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Mean Scores and Standard Deviations of Anger Control and Negative Affect for Main Demographic Factors, Smoking Behaviors, Alcohol Consumption, BMI Group, Red Meat Consumption, Physical Activity Levels, and for Women use of HRT (Unstandardized Scores Reported)

 

Altogether, 1952 cancers were diagnosed during an average of 9 years of follow-up, including 352 breast cancers, 318 prostate cancers, 88 lung cancers, 280 colorectal cancers, and 261 melanomas. Table 2 gives descriptive information in terms of cancer diagnosis and demographics and confounding factors. Except for breast cancer, cases were generally older than all respondents. Respondents with the lowest level of education had the highest risks of lung, colorectal, and total cancer. Current smokers or former smokers had higher risks of lung, colorectal, and total cancer than never smokers. People with a BMI >30 were at increased risk of total cancer. Higher levels of alcohol consumption were associated with increased risk of colorectal cancer. Women who were current users of HRT at baseline were at increased risk of breast cancer.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Numbers of Total and Male and Female Respondents in Each of the Demographic Groups and Risk Factor Groups and Percent in Each Group Developing Each of the Five Specific Cancers and Total Cancer

 

Table 3 shows the age, and where appropriate, gender-adjusted mean scores for anger control and negative affect for cases and noncases for each cancer of interest and total cancers. In general, any differences in mean scores between those who developed the cancer and those who did not were small. The only statistically significant differences in mean scores were for anger control for prostate cancer, and negative affect for lung cancer. However, these differences were <0.2 or 0.3 of a standard deviation, indicating, at most, a small effect size. Table 4 shows for each of the specific cancers and total cancers, the HRs for anger control and negative affect and risk factors adjusted for in the different analyses. Standardized scores for anger control and negative affect were used in these analyses. The HRs represent the change in risk for developing the cancer associated with a 1-U increase in the standardized score. All HRs were consistent with weak associations at most, and the upper limits of all confidence intervals, including those that overlapped the null value, were also consistent with only weak associations. Of the two gender-specific cancers, only the HR for the association between prostate cancer and anger control was increased. For other cancers, there were weak positive associations between negative affect and both lung cancer and colorectal cancer after adjusting for known risk factors, including smoking. For lung cancer, the increases in risk associated with higher negative affect (HR = 1.24; 95% CI: 1.01–1.52) was small compared with that associated with being a current (HR = 13.10; 95% CI: 6.13–27.92) or former (HR = 4.90; 95% CI: 2.33–10.32) smoker. For colorectal cancer, the association with negative affect was weaker than the associations with alcohol and red meat consumption. Although it was not statistically significant, the HR for anger control and lung cancer was of a similar magnitude to that found for prostate cancer. Melanoma and total cancers were not associated with the two emotion measures, with both HRs close to unity. There was no significant interaction between anger control and negative affect for any of the cancer outcomes (breast: HR = 0.98, 95% CI: 0.89–1.08; prostate: HR = 1.00, 95% CI: 0.90–1.12; lung: HR = 1.15, 95% CI: 0.96–1.37; colorectal: HR = 1.05, 95% CI: 0.94–1.16; melanoma: HR = 1.04, 95% CI: 0.92–1.17; total cancer: HR = 0.99, 95% CI: 0.96–1.04). The upper limits of the confidence intervals associated with the HRs for the interactions were consistent with, at most, only a weak interaction effect.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Age and Gender (Where Appropriate) Adjusted Mean Scores for Anger Control and Negative Affect for Those With and Without Each of the Five Specific Cancers and Total Cancer

 

View this table:
[in this window]
[in a new window]

  		TABLE 4. Multivariate Associations (Hazard Ratios and 95% Confidence Intervals) Between Anger Control and Negative Affect and Each of the Five Specific Cancers and Total Cancers Adjusted for Risk Factors for Each Cancer Type

 

After excluding the first 2 years of follow-up, the association between anger control and colorectal cancer was slightly stronger (adjusted HR = 1.22; 95% CI: 0.98–1.52), but all other associations were similar to those seen in Table 4.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
NOTES
 REFERENCES
 
Overall, we found no evidence that anger control and negative affect were associated with risk of breast, melanoma, or total cancer and the confidence intervals ruled out the possibility of anything other than weak associations. The few positive associations we found, for anger control in relationship to risk of prostate cancer and for negative affect in relationship to risk of lung cancer and colorectal cancer, were small and substantially less than the associations between lifestyle factors and the respective cancers. We also found no evidence that negative affect and anger control interact to influence the risk of any of these cancers.

Our study has several strengths and limitations. Its strengths include the prospective nature of the study, the almost complete ascertainment of cancers, the large sample size and a reasonable period of follow-up, which resulted in precise estimates of associations as demonstrated by the narrow confidence intervals. We also measured and controlled for the effects of some important confounding variables, although we acknowledge that not all risk factors for each type of cancer (e.g., family history) were measured. Whereas our measure of anger control was suggested to show the most promise of being related to cancer risk (35), it is only one component of a larger instrument, and using it without the other components as an indicator of emotional control may limit the strength of our conclusions. In addition, we used only a single measurement of negative affect and anger control, and people's responses on these measures may have changed over time. For example, our measure of trait negative affect has been shown to have a moderate level of stability over 7 years (45). This variability may have weakened the associations between negative affect and cancer onset. On the other hand, because we conducted several analyses, it is possible that some of the associations we found may be due to chance.

Few studies examining the relationship between emotional factors and cancer risk have reported findings for more than one cancer. Most have examined either specific cancers such as breast cancer (5,12,13,15) or total cancer (24,27). Because the number of cases in the present study is one of the largest reported in this area, this is one of the few studies where it has been possible to conduct cancer-specific analyses. We hypothesized that there would be no association between anger control or negative affect and total cancers and this was supported. Our finding that associations between emotional factors and cancer risk may differ by cancer type supports the suggestion that it is inappropriate to use total cancer as an outcome measure (35). Whether the differential role emotional factors may play in the onset of the different cancers is due to the different biological basis of these cancers or due to different psychological factors influencing different behavioral risk factors (e.g., smoking) needs to be examined in future studies. We found that negative affect was higher among smokers than never smokers, suggesting that further research into the latter explanation may be fruitful. Efforts to understand how negative affect influences risk of lung cancer may examine how smoking behaviors differ between those with higher and lower levels of negative affect.

This is the only study we know to have examined the interaction between anger control and negative affect on cancer risk. We hypothesized that if anger control influences cancer risk, this association may be stronger among people who experience high negative affect. We found no interaction between negative affect and anger control on risk of the different cancers or total cancers. This finding suggests that any associations between anger control and negative affect on cancer risk are independent of each other and additionally that negative affect and anger control may be associated with risks for different cancers.

The lack of association we have observed between anger control and breast cancer risk is important as it provides convincing evidence that anger control is not a risk factor for breast cancer, especially when taken together with the null findings from several quasi-prospective studies (12,13), and two prospective studies (5,10). Reports of associations between emotional control (including anger) and breast cancer have come from small, retrospective studies that have failed to control for confounding factors such as age. With data from 12,039 women, including 352 who developed breast cancer in the follow-up period, our study is one of the largest of its kind, and our narrow CIs are compatible with extremely weak associations only. Although we could not control for family history of breast cancer in our analyses, we did control for other major risk factors for breast cancer, including age, parity, HRT, BMI, and education. Other studies that did not find an association between anger control and breast cancer risk have also controlled for age and several other known risk factors. These findings show the importance of controlling for established risk factors in analyses examining associations between emotional factors and cancer risk.

Our findings with respect to prostate cancer are not so clear-cut and there are few comparative data available. We found a small increase in risk associated with higher scores for anger control. This finding is novel. Most studies of emotional factors and the risk of cancer among men have lacked statistical power to examine specific cancers (25) or have used "total cancer" as the outcome (27). As little is known about biological or behavioral risk factors for prostate cancer, speculations as to the reasons behind this association may be premature. In addition, given the novelty of our findings and the relatively small association found, other studies showing the same association are needed before these findings can be taken as anything other than tentative.

For melanoma, we observed no associations with our two psychosocial measures. An early study (41) examined the association between the "Type C" personality and melanoma thickness. Although there was a significant positive association between their measure of Type C and melanoma thickness in that study, the association was not significant after adjustment for skin pigmentation.

One of the strongest associations we found was for lung cancer risk and high scores on negative affect. There was no significant interaction between anger control and negative affect and risk for lung cancer. This finding suggests that high levels of negative emotions per se increase the risk of lung cancer. As few studies in this area have examined the association between negative affect and cancer onset in general and lung cancer in particular, it is not possible to compare this finding with other work. This link may be due to factors associated with smoking that we have not been able to control in our analyses. For instance, people with high negative affect may be less likely to quit smoking and/or differ in their smoking behaviors (e.g., they may inhale their cigarettes more deeply). The relative risk for lung cancer associated with negative affect was considerably smaller than the relative risk associated with smoking, suggesting that efforts to reduce lung cancer risk should first and foremost focus on preventing smoking uptake and encouraging smokers to quit.

The risk of colorectal cancer was also weakly associated with higher negative affect scores, but was not associated with anger control. To our knowledge, only two other studies have looked at the association between emotional factors and colorectal cancer risk (31,32). A case control study that adjusted for risk factors such as diet, alcohol consumption, and age showed that people with colorectal cancer were more likely to exhibit a particular personality profile that included repressing or controlling negative emotions, including anger, and inhibiting reactions that may offend others (32). Our failure to find an association between anger control and colorectal cancer risk suggests that the experience of negative emotions, rather than the repression or control of these emotions, influences colorectal cancer risk. Because of the small amount of work examining the role of emotional factors in the onset of colorectal cancer, and our inability to control for risk factors such as family history of colorectal cancer, our findings should be treated with caution.

In conclusion, we found evidence that measures of anger control and negative affect may have a relatively weak positive association with risk of colorectal, lung, and prostate cancer, but are not associated with breast cancer, melanoma, or total cancers. These differential associations suggest that future studies in this area should use specific cancers rather than total cancer as the outcome variable. Our positive associations should be considered preliminary and require replication in other large prospective studies that control for known risk factors for cancer. The relationships between anger control and negative affect and cancer risk found in this study were weak when compared with the associations found for other risk factors, specifically smoking for lung cancer, and suggest that we need to ask whether these associations are of any practical importance. We believe that the associations we found are not strong enough to warrant interventions to reduce levels of anger control and negative affect among the population. In terms of cancer prevention, we consider it more important to focus public health and individual interventions on modifying the principal behavioral risk factors for cancer such as smoking, diet, and physical activity, as it may be that the associations that we have observed are mediated via these behaviors. Nevertheless, because few large prospective studies have investigated the association between anger control, negative affect, and cancer onset, and because many people believe that cancer risk is associated with emotional factors, the results of this study are important.

The authors thank the many thousands of Melbourne residents who participated in the study as well as the original investigators and those who worked to recruit study participants.


   NOTES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 NOTES
REFERENCES
 
Received for publication November 17, 2006; revision received May 22, 2007.

Cohort recruitment was funded by VicHealth and The Cancer Council Victoria. The study is also funded by Grant 209057 from the National Health and Medical Research Council and is supported by infrastructure provided by The Cancer Council Victoria. Magdalena Lagerlund was funded by a postdoctoral fellowship from the Swedish Council for Working Life and Social Research.

DOI:10.1097/PSY.0b013e31814d4e6a


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 

  1. Stewart DE, Cheung AM, Duff S, Wong F, McQuestion M, Cheng T, Purdy L, Bunston T. Attributions of cause and recurrence in long-term breast cancer survivors. Psychooncology 2001;10:179–83.[CrossRef][Medline]
  2. Stewart D, Duff S, Wong F, Melancon C, Cheung A. The views of ovarian cancer survivors on its cause, prevention, and recurrence. Medscape Womens Health 2001;6:5.[Medline]
  3. Costanzo ES, Lutgendorf SK, Bradley SL, Rose SL, Anderson B. Cancer attributions, distress, and health practices among gynecologic cancer survivors. Psychosom Med 2005;67:972–80.[Abstract/Free Full Text]
  4. Wold KS, Byers T, Crane LA, Ahnen D. What do cancer survivors believe causes cancer? (United States). Cancer Causes Control 2005;16:115–23.[CrossRef][Medline]
  5. Bleiker EM, van der Ploeg HM, Hendriks JH, Ader HJ. Personality factors and breast cancer development: a prospective longitudinal study. J Natl Cancer Inst 1996;88:1478–82.[Abstract/Free Full Text]
  6. Chen CC, David AS, Nunnerley H, Michell M, Dawson JL, Berry H, Dobbs J, Fahy T. Adverse life events and breast cancer: case-control study. BMJ 1995;311:1527–30.[Abstract/Free Full Text]
  7. Edwards JR, Cooper CL, Pearl SG, de Paredes ES, O'Leary T, Wilhelm MC. The relationship between psychosocial factors and breast cancer: some unexpected results. Behav Med 1990;16:5–14.[Medline]
  8. Fox CM, Harper AP, Hyner GC, Lyle RM. Loneliness, emotional repression, marital quality, and major life events in women who develop breast cancer. J Community Health 1994;19:467–82.[CrossRef][Medline]
  9. Greer S, Morris T. Psychological attributed of women who develop breast cancer: a controlled study. J Psychosom Res 1975;19:147–53.[CrossRef][Medline]
  10. Hahn R, Petitti D. Minnesota multiphasic personality inventory-rated depression and the incidence of breast cancer. Cancer 1988;61:845–8.[CrossRef][Medline]
  11. Lillberg K, Verkasalo PK, Kaprio J, Helenius H, Koskenvuo M. Personality characteristics and the risk of breast cancer: a prospective cohort study. Int J Cancer 2002;100:361–6.[CrossRef][Medline]
  12. O'Donnell M, Fisher R, Irvine K, Rickard M, McConaghy N. Emotional suppression: can it predict cancer outcome in women with suspicious screening mammograms? Psychol Med 2000;30:1079–88.[CrossRef][Medline]
  13. Price MA, Tennant CC, Butow PN, Smith RC, Kennedy SJ, Kossoff MB, Dunn SM. The role of psychosocial factors in the development of breast carcinoma: Part II. Life event stressors, social support, defense style, and emotional control and their interactions. Cancer 2001;91:686–97.[CrossRef][Medline]
  14. Roberts FD, Newcomb PA, Trentham-Dietz A, Storer BE. Self-reported stress and risk of breast cancer. Cancer 1996;77:1089–93.[CrossRef][Medline]
  15. Watson M, Pettingale K, Greer S. Emotional control and autonomic arousal in breast cancer patients. J Psychosom Res 1984;28:467–74.[CrossRef][Medline]
  16. Aro A, de Koning H, Screck M, Henriksson M, Anttila A, Pukkala E. Psychological risk factors of incidence of breast cancer: a prospective cohort study in Finland. Psychol Med 2005;35:1515–21.[CrossRef][Medline]
  17. Nylicek I, Louwman W, Van Nierop P, Wijnands C, Coebergh J, Pop W. Depression and the lower risk for breast cancer development in middle aged women: a prospective study. Psychol Med 2003;33:1111–7.[CrossRef][Medline]
  18. Ollonen P, Lehtonen J, Eskelinen M. Anxiety, depression, and the history of psychiatric symptoms in patients with breast disease: a prospective case-control study in Kuopio, Finland. Anticancer Res 2005;25:2527–33.[Medline]
  19. Jacobs JR, Bovasso GB. Early and chronic stress and their relation to breast cancer. Psychol Med 2000;30:669–78.[CrossRef][Medline]
  20. Dattore PJ, Shontz FC, Coyne L. Premorbid personality differentiation of cancer and noncancer groups: a test of the hypothesis of cancer proneness. J Consult Clin Psychol 1980;48:388–94.[CrossRef][Medline]
  21. Hansen P, Floderus B, Frederiksen K, Johnansen C. Personality traits, health behaviour and risk for cancer: a prospective study of a Swedish twin cohort. Cancer 2005;103:1082–91.[CrossRef][Medline]
  22. Kaplan GA, Reynolds P. Depression and cancer mortality and morbidity: prospective evidence from the Alameda county study. J Behav Med 1988;11:1–13.[CrossRef][Medline]
  23. Nakaya N, Tsubono Y, Hosokawa T, Nishino Y, Ohkubo T, Hozawa A, Shibuya D, Fukudo S, Fukao A, Tsuji I, Hisamichi S. Personality and the risk of cancer. J Natl Cancer Inst 2003;95:799–805.[Abstract/Free Full Text]
  24. Penninx BW, Guralnik JM, Pahor M, Ferrucci L, Cerhan JR, Wallace RB, Havlik RJ. Chronically depressed mood and cancer risk in older persons. J Natl Cancer Inst 1998;90:1888–93.[Abstract/Free Full Text]
  25. Persky V, Kempthorne-Rawson J, Shekelle R. Personality and risk of cancer: 20 year follow-up of the Western electric study. Psychosom Med 1987;49:435–49.[Abstract/Free Full Text]
  26. Schapiro I, Nielsen L, Jorgensen T, Boesen E, Johansen C. Psychic vulnerability and the associated risk for cancer. Cancer 2002;94:3299–306.[CrossRef][Medline]
  27. Tijhuis M, Elshout J, Feskens E, Janssen M, Kromhout D. Prospective investigation of emotional control and cancer risk in men (the Zutphen elderly study) (The Netherlands). Cancer Causes Control 2000;11:589–95.[CrossRef][Medline]
  28. Dalton SO, Mellemkjaer L, Olsen JH, Mortensen PB, Johansen C. Depression and cancer risk: a register-based study of patients hospitalized with affective disorders, Denmark, 1969–1993. Am J Epidemiol 2002;155:1088–95.[Abstract/Free Full Text]
  29. Zonderman AB, Costa PT Jr, McCrae RR. Depression as a risk for cancer morbidity and mortality in a nationally representative sample. JAMA 1989;262:1191–5.[Abstract/Free Full Text]
  30. Gallo JJ, Armenian HK, Ford DE, Eaton WW, Khachaturian AS. Major depression and cancer: the 13-year follow-up of the Baltimore epidemiologic catchment area sample (United States). Cancer Causes Control 2000;11:751–8.[CrossRef][Medline]
  31. Kavan M, Engdahl B, Kay S. Colon cancer: personality factors predictive of onset and state of presentation. J Psychosom Res 1995;39:1031–9.[CrossRef][Medline]
  32. Kune GA, Kune S, Watson LF, Bahnson CB. Personality as a risk factor in large bowel cancer: data from the Melbourne colorectal cancer study. Psychol Med 1991;21:43–60.[Medline]
  33. Knekt P, Raitasalo R, Heliovaara M, Lehtinen V, Pukkala E, Teppo L, Maatela J, Aromaa A. Elevated lung cancer risk among persons with depressed mood. Am J Epidemiol 1996;144:1104–6.[Free Full Text]
  34. McKenna M, Zevon M, Corn B, Rounds J. Psychological factors and the development of breast cancer: a meta-analysis. Health Psychol 1999;18:520–31.[CrossRef][Medline]
  35. Butow PN, Hiller JE, Price MA, Thackway SV, Kricker A, Tennant CC. Epidemiological evidence for a relationship between life events, coping style, and personality factors in the development of breast cancer. J Psychosom Res 2000;49:169–81.[CrossRef][Medline]
  36. Dalton SO, Boesen EH, Ross L, Schapiro IR, Johansen C. Mind and cancer: do psychological factors cause cancer? Eur J Cancer 2002;38:1313–23.[CrossRef][Medline]
  37. Garssen B, Goodkin K. On the role of immunological factors as mediators between psychosocial factors and cancer progression. Psychiatry Res 1999;85:51–61.[CrossRef][Medline]
  38. Garssen B. Psychological factors and cancer development: evidence after 30 years of research. Clin Psychol Rev 2004;24:315–38.[CrossRef][Medline]
  39. Watson M, Greer S. Development of a questionnaire measure of emotional control. J Psychosom Res 1983;27:299–305.[CrossRef][Medline]
  40. Grassi L, Cappellari L. State and trait psychological characteristics of breast cancer patients. New Trends Exp Clin Psychiatry 1988;IV:99–109.
  41. Temoshok L, Heller BW, Sagebiel RW, Blois MS, Sweet DM, DiClemente RJ, Gold ML. The relationship of psychosocial factors to prognostic indicators in cutaneous malignant melanoma. J Psychosom Res 1985;29:139–53.[CrossRef][Medline]
  42. Sturmer T, Hasselbach P, Amelang M. Personality, lifestyle and risk of cardiovascular disease and cancer: follow-up of a population based cohort. BMJ 2006;332:1359–65.[Abstract/Free Full Text]
  43. Watson D, Clark L, Tellegen A. Development and validation of brief measures of positive and negative affect: the PANAS scales. J Pers Soc Psych 1988;37:1418–20.[CrossRef]
  44. Watson D, Clark L. Negative affectivity: the disposition to experience aversive emotional states. Psych Bull 1984;96:465–90.[CrossRef][Medline]
  45. Watson D, Walker L. The long-term stability and predictive validity of trait measures of affect. J Pers Soc Psych 1996;70:367–77.
  46. Headey B, Wearing A. Understanding happiness: a theory of subjective well-being. Melbourne: Longman Cheshire; 1992.
  47. Giles G, English D. The Melbourne collaborative cohort study. IARC Sci Publ 2002;156:69–70.[Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by White, V. M.
Right arrow Articles by Giles, G. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by White, V. M.
Right arrow Articles by Giles, G. G.
Related Collections
Right arrow Personality
Right arrow Cancer

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS