The role of psychological functioning in the use of complementary and alternative methods among disease-free colorectal cancer survivors

A Report from the American Cancer Society's Studies of Cancer Survivors


  • Presented at the Fourth Biennial Cancer Survivorship Research Conference entitled “Cancer Survivorship Research: Mapping the New Challenges,” Atlanta, Georgia, June 18-20, 2008.



The medical and demographic correlates of complementary and alternative medicine (CAM) use among cancer survivors have been well documented. However, the role of psychological functioning in cancer survivors' CAM use and the degree to which such factors apply to survivors of colorectal cancer require additional study. In addition, sex differences in CAM use and its correlates among colorectal cancer survivors are not well understood.


By using data from a large-scale national population-based study of quality of life and health behaviors among cancer survivors, the authors examined the prevalence and psychological correlates of CAM use among 252 male and 277 female colorectal cancer survivors.


Use of CAM was more common among women, those with more education, and recipients of chemotherapy and radiation therapy. Several psychological factors predicted increased use of CAM among female colorectal cancer survivors, including anxiety, fear of cancer recurrence, fatigue, vigor, anger, mental confusion, and overall emotional distress. Depression was associated with decreased CAM use among female survivors, both for overall CAM use and across several standard CAM domains. In contrast, psychological functioning had little impact on male colorectal cancer survivors' CAM use. The only nonmedical/demographic variable associated with men's use of CAM was fatigue, which predicted use only of biologically based practices, such as diet and nutritional supplements.


Psychological functioning has a significant impact on CAM use among female colorectal cancer survivors. Decreased use of CAM among women with depressive symptoms was unexpected and warrants additional investigation. Cancer 2009;115(18 suppl):4397–408. © 2009 American Cancer Society.

Several recent studies with the general population have documented widespread use of nonconventional healthcare and wellness interventions, known collectively as complementary and alternative medicine (CAM).1-3 Consistent with this general trend, many people dealing with chronic illnesses, including cancer, use CAM during and after treatment.2-7 Reasons commonly cited by people with cancer for using CAM include attempting to improve one's physical and emotional health,8-10 the desire to exert a sense of personal control over one's illness,11 dissatisfaction with the medical care system or health providers,11, 12 and doubts concerning the effectiveness of conventional treatment.13

This research, however, has been limited in important ways. Typically, sample sizes are relatively small and focus on a particular population segment.14, 15 Most studies involve patients at urban, academic cancer centers,16-18 or are limited by the geographic location of the participants,7, 19, 20 and therefore are not representative of survivors across different socioeconomic or geographic groups. This limitation may influence access to some forms of CAM (eg, acupuncture/acupressure) that tend to be more available in urban areas.5 In addition, many studies examine CAM use only during active treatment,12, 21, 22 or less commonly, in advanced cancer patients,23 who may be receiving palliative care. Among studies involving post-treatment, disease-free cancer survivors, most have either focused on breast cancer survivors24-26 or have included mixed samples of individuals with different cancers.5, 27

One type of cancer that has received relatively little attention with respect to CAM use is colorectal cancer (CRC), which is the third most common cancer in the United States, with approximately 150,000 cases expected to be diagnosed in 2008.28 Because CRC affects men and women almost equally, CRC is well suited to investigate sex differences in CAM use. The treatments for CRC typically include surgery with adjuvant chemotherapy or radiation, and may result in side effects such as problems with bowel movements, pain, fatigue, nausea, and appetite changes,29, 30 which negatively impact quality of life (QOL).31, 32 Although many symptoms and side effects associated with CRC treatment may attenuate after the cessation of treatment, a sizable proportion of CRC survivors continue to experience troublesome long-term effects after treatment ends.33-36 Both the acute and the persistent symptoms may lead CRC survivors to seek CAM to manage such symptoms.

Despite the growing number of CRC survivors and the impact of the disease and its treatment on both physical and psychological functioning, to our knowledge relatively little research has explored the use of CAM in this population. Indeed, a literature search using Pubmed identified only 3 published papers that focus specifically on CAM use among CRC survivors.15, 37, 38 The prevalence rates for CAM use among CRC survivors in these studies varied widely, ranging from 32% to 75%. CAM use was generally associated with female sex, younger age, and receipt of conventional cancer therapy. The most common forms of CAM used were psychological and spiritual therapies,37 herbal medicines,38 and homeopathy.15

Results of these studies underscore the range of CAM use among CRC survivors and help identify potential sociodemographic correlates. However, only 1 investigated the psychosocial correlates of CAM use among CRC survivors.15 In this study of 191 CRC survivors, poorer social support, more intrusive thoughts, and poorer QOL were associated with CAM use, although none of these psychosocial variables remained significant in multivariate analyses.15 This study included only 12 CAM items, and did not categorize the items into the standard domains established by the National Center for Complementary and Alternative Medicine. In addition, the sample focused on CRC survivors treated at an urban medical center, limiting the generalizability of the findings to CRC survivors from rural areas, who likely have poorer availability and accessibility of certain types of CAM.

Although there is a dearth of data regarding psychosocial predictors of CAM use among CRC survivors, other studies with the general population or mixed samples of cancer survivors have included psychological variables. In general, individuals with poorer psychological functioning are more likely to use CAM.39-41 Depression is among the most common psychological correlates of CAM use, with most studies finding higher levels of depressive symptoms being associated with increased CAM use, although some report no association.42, 43 A national telephone survey of over 3000 women found that of 220 with depression, 54% reported using 1 of 9 CAMs in the past year.44 Other psychosocial variables associated with CAM use include fear of cancer recurrence,42 optimism,45 internal locus of control,46, 47 and poorer mental health.12 Although women are more likely to use CAM than men, the role of psychological processes in such sex differences has received little attention, particularly among CRC survivors. Some evidence suggests that men and women use CAMs for different reasons,48 which may reflect differing psychological needs and processes, but additional research is needed.

Thus, the aims of this study were to 1) investigate the prevalence of CAM use and the medical, demographic, and psychosocial correlates among a population-based sample of CRC survivors; and 2) assess differences in CAM use and psychosocial predictors as they vary with the sex of the study participants.



Data were collected through the American Cancer Society's Study of Cancer Survivors-I, a national, population-based prospective investigation of cancer survivors' QOL and health behaviors.49 Stratified random samples of 1-year survivors of 10 cancers were drawn from 11 statewide cancer registries. Oversight was provided by the institutional review board of Emory University, with additional approvals obtained from each state. Eligibility criteria included: 1) diagnosis with 1 of 10 high-incidence cancers (breast, prostate, bladder, uterine, skin melanoma, colorectal, kidney, non-Hodgkin lymphoma, ovarian, and lung); 2) localized, regional, or distant Surveillance, Epidemiology, and End Results summary stage (except inclusion of in situ cases for bladder cancer); 3) diagnosis between January 2000 and September 2003; 4) at least 18 years of age at diagnosis; and 5) fluent in English or Spanish. Younger survivors (<55 years old), minorities, and survivors of cancers with higher mortality rates (lung, colorectal, kidney, non-Hodgkin lymphoma, and ovarian) were oversampled. Depending on state registry requirements, survivors' physicians either received notification or provided consent before case recruitment. A mixed-mode recruitment method using mail and telephone was used for recruitment and data collection. The physician consent rate was 93%, the questionnaire return rate was 53%, and the overall study response rate was 36%. Compared with nonresponders, survivors who completed a study questionnaire were more likely to be women, younger, white, breast cancer survivors, and diagnosed with early stage disease (all P < .0001). A full description of the rationale and methodology for Study of Cancer Survivors-I is provided in a previous publication.49

For the present study, responses to the baseline survey from 529 colorectal cancer survivors (252 men and 277 women) were extracted from the sample. To create a sample of early stage, disease-free individuals with complete data on relevant variables, the analytic sample was limited to CRC survivors who 1) were diagnosed with localized or regional stage only and 2) did not report disease progression or additional primaries on the baseline survey.


Demographics and medical characteristics

Demographic information and types of cancer treatment received were collected on the baseline survey. Other medical variables such as date of diagnosis and stage of disease were obtained from the state cancer registries.

Cancer Problems in Living Scale

The Cancer Problems in Living Scale contains a list of 29 problems that cancer survivors may experience. The instrument was originally developed with bone marrow transplant recipients, and has been used previously with cancer survivors.33, 50, 51 Respondents indicate how much of a problem each item has been since cancer diagnosis: not a problem, somewhat of a problem, or a severe problem. A recent factor analysis of the Cancer Problems in Living Scale identified 4 underlying dimensions: 1) physical distress, 2) emotional distress, 3) employment/financial problems, and 4) fear of recurrence.52

Profile of Mood States-Short Form

The Profile of Mood States-Short Form (POMS-SF)53 is a 37-item short form of the original 65-item POMS,54 and has been used previously with cancer survivors.55, 56 For each item, respondents indicate on a 5-point scale how much they had been feeling that way during the past 2 weeks. Six factor-based subscale scores can be derived: anger-hostility, confusion-bewilderment, depression-dejection, fatigue-inertia, tension-anxiety, and vigor-activity.

Use of CAM

Survivors were asked whether they had used any of the 19 different CAM modalities shown in Table 1. These 19 items were categorized into domains defined by the National Center for Complementary and Alternative Medicine27: whole medical systems, biologically based practices, manipulative and body-based practices, energy medicine (EM), and mind-body methods (MBM). The mind-body methods domain was further categorized into nonreligious/spiritual mind-body methods and religious/spiritual mind-body methods, which allowed for the examination of differences in the prevalence and correlates of mind-body–based methods with and without religious/spiritual items included. In addition to the CAM domain variables, 2 additional summary outcome variables reflected use of any CAM when all 19 methods were considered together. The first, any CAM use with spiritual mind-body methods, included the spiritual-based practices; the second, any CAM use without spiritual mind-body methods, excluded spiritual-based practices. All outcomes, including each domain score and the 2 any CAM use summary scores, were dichotomized, with 1 indicating use of CAM and 0 indicating no use of CAM.

Table 1. Complementary and Alternative Methods by Domain
Whole medical systems
Mind-body medicine–religious/spiritual methods
 Faith/spiritual healing
 Prayer/spiritual practice
 Religious counseling
Mind-body medicine–nonreligious/spiritual methods
 Attended support groups
 Biofeedback therapy
Biologically based practices
 Herbal therapy
 Nutritional supplements/vitamins
 Special diet
Manipulative and body-based practices
Energy medicine
 Tai chi/yoga
 Therapeutic touch

Statistical Analysis

Univariate analyses were conducted to evaluate the association between the medical/demographic variables and the 2 any CAM use variables (any CAM use with and any CAM use without spiritual mind-body methods) as well as to compare CAM use by sex. The psychosocial correlates of CAM use for each outcome measure were evaluated after controlling for demographics (sex, race/ethnicity, education, and age) and medical characteristics (stage of disease, time since diagnosis, and receipt of chemotherapy and/or radiation). Because nearly all participants (96%) had received surgery, surgery was excluded from the model. Nested effects of the psychosocial correlates within sex were tested to evaluate their impact on the likelihood of CAM use for each sex. Significance criteria were set to alpha = .05 for all analyses. Three domains (whole medical systems, manipulative and body-based practices, and energy medicine) were excluded from further evaluation, because low prevalence of the use of these CAM domains in our sample led to large standard errors and unreliable estimates.


Table 2 summarizes the sample characteristics and prevalence of CAM use by demographic and medical variables. The sample was predominantly white (86%) and had a mean age of 60 years. Greater than half of the survivors had a high school education, and a majority had an annual income of at least $40,000. On average, survivors were approximately 15 months from diagnosis, and 55% had been diagnosed with regional disease. Ninety-six percent of the patients had received surgical treatment; 61% had received chemotherapy; and 33% had received radiation therapy. No differences between sexes were found for demographic and medical characteristics. In terms of the demographic and medical correlates of any CAM use with and any CAM use without spiritual mind-body methods, univariate analyses revealed that female CRC survivors and those who received chemotherapy or radiation were more likely to use CAM. When we included spiritual mind-body methods in the any CAM use category, those with higher education were also significantly more likely to use CAM, and younger CRC survivors tended to use more CAM, although this did not reach statistical significance (P = .065). When spiritual mind-body methods were excluded from the any CAM use category, younger CRC survivors were significantly more likely to use CAM (P = .009), as were those diagnosed with regional disease.

Table 2. Sample Characteristics and Any CAM Use
Sample DescriptionOverall Distribution, No. (Row %)*Any CAM Use With MBM-S, No. (Row %)*PAny CAM Use Without MBM-S, No. (Row %)*P
  • CAM indicates complementary and alternative medicine, MBM-S, mind-body medicine–religious/spiritual methods; SEER, Surveillance, Epidemiology, and End Results program.

  • Age at questionnaire completion: 59.8 ± 13.3 years (mean ± standard deviation).

  • Time since diagnosis: 15.0 ± 3.7 months (mean ± standard deviation).

  • *

    Percentage was generated based on number not missing.

  • Chi-square test P values were reported.

  • No sex difference was found across the medical demographic variables stated above.

Demographic characteristics
 Sex  <.001 .001
  Men252 (47.6)176 (69.8) 144 (57.1) 
  Women277 (52.4)238 (85.9) 197 (71.1) 
 Race  .366 .896
  White455 (86.0)359 (78.9) 294 (64.6) 
  Nonwhite74 (14.0)55 (74.3) 47 (63.5) 
 Age group, y  .065 .009
  <65329 (62.3)266 (80.9) 226 (68.7) 
  ≥65199 (37.7)147 (73.9) 114 (57.3) 
 Education  .042 .098
  <High school224 (43.0)165 (73.7) 135 (60.3) 
  ≥High school297 (57.0)242 (81.5) 200 (67.3) 
 Income  .442 .336
  <$20,00089 (19.6)65 (73.0) 51 (57.3) 
  $20,000-$39,999116 (25.6)90 (77.6) 72 (62.1) 
  $40,000-$74,999146 (32.2)119 (81.5) 97 (66.4) 
  ≥$75,000103 (22.7)83 (80.6) 71 (68.9) 
Medical characteristics
 SEER summary stage  .526 .006
  Localized238 (45.0)183 (76.9) 138 (58.0) 
  Regional291 (55.0)231 (79.4) 203 (69.8) 
 Surgery  1.000 1.000
  Yes502 (96.4)394 (78.5) 12 (63.2) 
  No19 (3.6)15 (78.9) 325 (64.7) 
 Chemotherapy  .029 <.001
  Yes317 (60.8)259 (81.7) 224 (70.7) 
  No204 (39.2)150 (73.5) 113 (55.4) 
 Radiation therapy  .006 .001
  Yes172 (33.0)147 (85.5) 128 (74.4) 
  No349 (67.0)262 (75.1) 209 (59.9) 

Table 3 shows use of CAM domains and individual CAM items for male and female CRC survivors. Overall, 86% of women and 70% of men reported using at least 1 CAM to cope with their cancer (P < .001) when spiritual mind-body methods modalities were included, whereas 71% of women and 57% of men reported using at least 1 CAM when spiritual mind-body methods were excluded. A higher proportion of women also reported use of CAM in each of the CAM domains. These differences were significant (P < .05) for all domains except nonspiritual mind-body methods (P = .068). For both sexes, mind-body methods were the most commonly used domain, whereas whole medical systems were the least commonly used.

Table 3. Self-Reported Use of CAM by Sex
CAM Domains and ItemsMen, n = 252, No. (%)Women, n = 277, No. (%)Total, N = 529, No. (%)P*
  • CAM indicates complementary and alternative medicine; MBM-S, mind-body medicine–religious/spiritual methods.

  • *

    Chi-square test P values were displayed for each CAM domain but not for individual items due to small sample sizes.

Any CAM use with MBM-S176 (69.8)238 (85.9)414 (78.3)<.001
Any CAM use without MBM-S144 (57.1)197 (71.1)341 (64.5).001
Whole medical system1 (0.4)9 (3.2)10 (1.9).022
 Acupuncture/acupressure1 (0.4)2 (0.7)3 (0.6) 
 Homeopathy1 (0.4)7 (2.6)8 (1.5) 
Mind-body medicine-overall163 (64.7)225 (81.2)388 (73.3)<.001
Mind-body medicine–nonspiritual117 (46.4)151 (54.5)268 (50.7).068
 Aromatherapy2 (0.8)17 (6.1)19 (3.6) 
 Art1 (0.4)6 (2.2)7 (1.3) 
 Attended support groups13 (5.2)15 (5.4)28 (5.3) 
 Biofeedback therapy0 (0.0)3 (1.1)3 (0.6) 
 Hypnosis0 (0.0)1 (0.4)1 (0.2) 
 Imagery/visualization8 (3.2)23 (8.3)31 (5.9) 
 Meditation19 (7.5)40 (14.4)59 (11.2) 
 Relaxation105 (42)130 (47.1)235 (44.7) 
Mind-body medicine–spiritual133 (52.8)207 (74.7)340 (64.3)<.001
 Faith/spiritual healing82 (32.7)152 (55.5)234 (44.6) 
 Prayer/spiritual practice127 (50.8)196 (71.3)323 (61.5) 
 Religious counseling23 (9.2)36 (13.3)59 (11.3) 
Biologically based practices85 (33.7)134 (48.4)219 (41.4).001
 Herbal therapy5 (2.0)19 (6.9)24 (4.6) 
 Nutritional supplements/vitamins75 (29.8)124 (45.4)199 (37.9) 
 Special diet27 (10.7)41 (14.9)68 (12.9) 
Manipulative and body-based practices6 (2.4)26 (9.5)32 (6.1).001
 Massage6 (2.4)26 (9.5)32 (6.1) 
Energy medicine5 (2.0)15 (5.4)20 (3.8).042
 Tai chi/yoga4 (1.6)10 (3.6)14 (2.7) 
 Therapeutic touch1 (0.4)7 (2.6)8 (1.5) 

Table 4 summarizes the results from the generalized linear models, which tested the relationships between the psychosocial variables and the likelihood (odd ratios [OR]) of CAM use for each sex, after accounting for CAM use associated with relevant medical and demographic covariates. In general, psychosocial functioning had a significant impact on the likelihood of CAM use among female CRC survivors. In addition, the associations differed by sex for different domains, and stronger associations were found among female cancer survivors. For any CAM use with spiritual mind-body methods, fear of recurrence (OR, 4.91), fatigue-inertia (OR, 1.19), tension-anxiety (OR, 1.25), and vigor-activity (OR, 1.11) were significantly and positively associated with the likelihood of CAM use, whereas depression-dejection (OR, 0.75) was associated with decreased use of any CAM. Each of these relationships was significant only among female CRC survivors. As for any CAM use without spiritual mind-body methods, only tension-anxiety (OR, 1.22) and depression (OR, 0.79) were significantly related to CAM use among female survivors. Consistent with the findings for any CAM use with and any CAM use without spiritual mind-body methods, it was also noted that for the mind-body methods, nonspiritual mind-body methods, and spiritual mind-body methods domains, the negative relation between depression and CAM use was detected among female cancer survivors only, with estimated ORs of 0.76, 0.85, and 0.83, respectively. In addition, fear of recurrence (OR, 3.95) and fatigue-inertia (OR, 1.23) significantly predicted use of mind-body methods, and anger-hostility (OR, 1.13) predicted nonspiritual mind-body methods use among female cancer survivors only. Physical distress (OR, 5.08) and confusion-bewilderment (OR, 1.21) predicted biologically based practices use among female survivors, whereas only fatigue-inertia (OR, 1.13) predicted biologically based practices use among male survivors. Consistent with the impact of depression-dejection on both any CAM use variables, emotional distress, as measured by the Cancer Problems in Living Scale-Emotional Distress, was negatively associated (OR, 0.19) with use of biologically based practices among female but not male survivors.

Table 4. Odds Ratios and 95% Confidence Intervals for Predictors of CAM Use
Covariates/ Predictors*Any CAM Use With MBM-SAny CAM Use Without MBM-SMBMMBM-SMBM-NSBBP
  • CAM indicates complementary and alternative medicine; MBM-S, mind-body medicine–religious/spiritual methods; MBM-NS, mind-body medicine– nonreligious/spiritual methods; BBP, biologically based practices; CPILS, Cancer Problems in Living Scale; PD, Physical Distress; ED, Emotional Distress; EF, Employment Financial; FR, Fear of Disease Recurrence; POMS, Profile of Mood States; AH, Anger-Hostility; CB, Confusion-Bewilderment; DP, Depression-Dejection; FI, Fatigue-Inertia; TA, Tension-Anxiety; and VA, Vigor-Activity.

  • *

    Results for whole medical systems, manipulative, and body-based practices, energy medicine was not shown due to nonexistence/instable maximum likelihood estimates for the outcome measures.

  • P < .01.

  • P < .05.

 Sex (referent: women)2.35 (0.27-20.21)0.78 (0.13-4.89)1.07 (0.15-7.45)0.37 (0.06-2.24)0.69 (0.12-3.86)0.77 (0.12-4.87)
 Race (referent: nonwhite)1.15 (0.53-2.46)1.09 (0.56-2.12)0.97 (0.47-1.98)1.01 (0.52-1.97)1.44 (0.78-2.66)1.03 (0.55-1.92)
 Education (referent: <high school)2.10 (1.22-3.59)1.83 (1.16-2.89)1.63 (1.00-2.68)1.89 (1.20-2.98)1.13 (0.74-1.74)1.92 (1.22-3.02)
 Age (continuous)1.01 (0.99-1.03)1.00 (0.99-1.02)1.01 (0.99-1.03)1.00 (0.98-1.02)1.02 (1.00-1.03)1.00 (0.98-1.02)
Medical variables
 Stage (referent: regional)1.22 (0.65-2.29)0.91 (0.53-1.54)1.27 (0.71-2.27)1.63 (0.96-2.77)0.75 (0.46-1.23)1.32 (0.78-2.22)
 Chemotherapy (referent: no)1.31 (0.66-2.61)1.43 (0.80-2.56)0.74 (0.39-1.41)0.88 (0.49-1.60)1.09 (0.63-1.89)2.11 (1.18-3.79)
 Radiation (referent: no)2.15 (1.11-4.17)1.80 (1.05-3.09)3.02 (1.64-5.55)2.20 (1.28-3.77)2.05 (1.26-3.34)0.93 (0.57-1.53)
 Time since diagnosis (continuous)1.03 (0.96-1.10)1.00 (0.95-1.06)1.01 (0.95-1.08)1.00 (0.94-1.06)1.00 (0.95-1.05)1.00 (0.94-1.06)
Psychosocial variables
 CPILS_PD (men)4.08 (0.66-25.13)2.56 (0.52-12.56)5.09 (0.90-28.68)1.91 (0.40-9.13)1.80 (0.40-8.16)3.48 (0.72-16.9)
 CPILS_PD (women)4.88 (0.45-53.03)2.63 (0.56-12.34)1.87 (0.29-12.09)3.61 (0.74-17.71)0.78 (0.22-2.79)5.08 (1.32-19.48)
 CPILS_ED (men)0.48 (0.06-3.81)1.25 (0.18-8.50)0.47 (0.07-3.40)0.18 (0.03-1.19)1.79 (0.29-11.16)0.56 (0.08-3.89)
 CPILS_ED (women)2.31 (0.12-43.02)1.71 (0.29-10.2)2.04 (0.24-17.61)0.78 (0.13-4.56)1.84 (0.45-7.57)0.19 (0.04-0.85)
 CPILS_EF (men)0.75 (0.22-2.54)1.21 (0.40-3.69)0.76 (0.24-2.38)0.75 (0.26-2.16)2.24 (0.76-6.59)0.93 (0.31-2.77)
 CPILS_EF (women)0.24 (0.05-1.07)0.75 (0.26-2.16)0.58 (0.16-2.17)0.63 (0.22-1.82)1.34 (0.53-3.37)1.14 (0.46-2.83)
 CPILS_FR (men)1.68 (0.64-4.40)1.47 (0.62-3.52)1.37 (0.55-3.40)2.05 (0.87-4.84)1.30 (0.56-3.02)2.09 (0.86-5.06)
 CPILS_FR (women)4.91 (1.32-18.21)2.22 (0.89-5.55)3.95 (1.33-11.73)2.10 (0.84-5.27)2.10 (0.96-4.56)1.52 (0.69-3.33)
 POMS_AH (men)0.89 (0.77-1.03)0.91 (0.80-1.05)0.94 (0.82-1.08)1.04 (0.92-1.19)0.95 (0.83-1.09)1.02 (0.89-1.18)
 POMS_AH (women)1.12 (0.92-1.37)1.08 (0.94-1.24)1.13 (0.96-1.33)1.09 (0.95-1.25)1.13 (1.01-1.27)1.02 (0.91-1.14)
 POMS_CB (men)1.04 (0.87-1.24)1.06 (0.90-1.24)1.07 (0.90-1.27)1.13 (0.96-1.33)1.09 (0.93-1.27)0.96 (0.81-1.14)
 POMS_CB (women)1.05 (0.83-1.32)1.06 (0.89-1.25)0.97 (0.80-1.17)1.00 (0.85-1.18)0.95 (0.83-1.09)1.21 (1.05-1.40)
 POMS_DP (men)0.99 (0.86-1.14)0.93 (0.81-1.06)0.95 (0.83-1.09)0.93 (0.81-1.07)0.92 (0.81-1.06)0.95 (0.82-1.09)
 POMS_DP (women)0.75 (0.61-0.92)0.79 (0.68-0.92)0.76 (0.64-0.90)0.83 (0.72-0.95)0.85 (0.75-0.96)0.91 (0.81-1.03)
 POMS_FI (men)1.05 (0.94-1.18)1.09 (0.98-1.21)1.00 (0.90-1.11)1.03 (0.93-1.14)0.98 (0.89-1.09)1.13 (1.01-1.26)
 POMS_FI (women)1.19 (1.01-1.40)1.03 (0.92-1.15)1.23 (1.06-1.42)1.07 (0.96-1.20)1.05 (0.96-1.15)0.94 (0.86-1.04)
 POMS_TA (men)1.06 (0.92-1.23)1.02 (0.90-1.16)1.09 (0.96-1.25)1.04 (0.92-1.18)1.05 (0.93-1.20)0.91 (0.79-1.03)
 POMS_TA (women)1.25 (1.01-1.54)1.22 (1.06-1.41)1.14 (0.97-1.33)1.12 (0.98-1.28)1.10 (0.99-1.23)1.09 (0.98-1.21)
 POMS_VA (men)1.07 (0.99-1.15)1.06 (0.99-1.14)1.07 (1.00-1.15)1.05 (0.98-1.12)1.06 (0.99-1.14)1.04 (0.96-1.12)
 POMS_VA (women)1.11 (1.01-1.22)1.05 (0.98-1.13)1.06 (0.98-1.15)1.02 (0.95-1.10)1.04 (0.98-1.11)1.05 (0.99-1.12)


This study investigated the prevalence and correlates of CAM use among CRC survivors who participated in a large, national, population-based study of cancer survivors' QOL and health behaviors. Overall, CAM use was higher than found in other studies of CAM use among CRC survivors, with nearly 86% of female survivors and nearly 70% of the male survivors in our sample reporting using at least 1 form of CAM, when spiritual mind-body methods modalities were included. The most common CAMs used were prayer/spiritual practices (61.5%) and faith/spiritual healing (44.6%). Among nonspiritual CAMs, relaxation (44.7%) and nutritional supplements/vitamins (37.9%) were used most often.

The distinction between prevalence when spiritual and nonspiritual CAMs are included is important, as opinions and practices differ regarding whether spiritual practices should be considered as CAMs. For this reason, some studies exclude spiritual-based practices,4, 7, 15, 26 whereas others report the results with and without spiritual-based practices,5, 7, 57 as we have done here. Indeed, when we excluded spiritual mind-body methods modalities, the overall prevalence rates were lowered to 71.1% for women and 57.1% for men.

CAM use among CRC survivors was associated with several medical and demographic characteristics, including female sex, higher education, and receipt of chemotherapy or radiation. Age approached but was not significantly related to CAM use (P = .065) when spiritual mind-body methods items were included, but was significant (P < .01) when spiritual mind-body methods items were excluded. Although these findings are of interest and corroborate previous research, the major contribution of this study is the direct assessment of the impact of CRC survivors' psychological functioning on self-reported CAM use across several domains. Our results suggest that emotional functioning, even after controlling for the roles of medical and demographic variables, is a powerful predictor of CAM use for female CRC survivors. Fears of cancer recurrence, anxiety, depressive symptoms, fatigue, anger, mental confusion, and vigor were all found to be significantly related to CAM use for women. For the most part, poorer emotional functioning predicted greater CAM use, but this was not always the case. Surprisingly, we found that female CRC survivors with depressive symptoms were less likely to use CAM, and this was found consistently across several CAM domains as well as for the 2 any CAM use variables. In contrast to the findings for female survivors, psychological functioning did not appear to predict CAM use among male survivors. Fatigue was the only nonmedical or demographic variable significantly associated with CAM use for men, and this relation was found for biologically based practices (eg, dietary and nutritional supplements) only, which may be perceived as addressing physical rather than psychological issues. Furthermore, although fatigue often has an affective or emotional component and may impact psychological functioning, it is often perceived as more physical than emotionally oriented.

Such findings may simply reflect a main effect of sex, with women being more likely to use CAM than men regardless of domain. However, it is interesting to consider the relative roles of psychological variables for CAM use among female CRC survivors. In particular, why would women with depressive symptoms be less, rather than more, likely to use CAM? This is inconsistent with what is generally found in the literature, where depression tends to be associated with increased use of CAM.44, 57, 58 Social cognitive learning theories59, 60 would hold that depression may result from a generalized expectation that 1 has little or no control over their personal situation, thus lowering motivation and leading to passivity and poorer self-efficacy. Applied to the findings of the current study, depressed female CRC survivors may therefore lack the motivation or self-efficacy to engage in CAM. Although this explanation is plausible, it appears to be challenged by the finding that the women in our study with higher scores on the fatigue and vigor subscales of the POMS-SF actually engaged in more CAM, compared with those with less fatigue and energy.

Another potential explanation for the negative relation noted between depression and CAM use among female survivors (but not men) is that women may be more inclined to express their emotions and admit vulnerability, and feel more comfortable in seeking conventional therapy for depression.61 Thus, the women in our sample may have sought psychotherapy or conventional psychopharmacological treatments for their depressive symptoms, rather than using CAM. In contrast, men have been found to be less likely to seek conventional treatment for depressive symptoms for several reasons, including concern about stigma, feeling that they should “tough it out,” or aversion to talking about emotional issues.62-64 Therefore, they may turn to alternative means of managing their depression, including some forms of CAM, which do not necessitate talking to a mental health professional or exploring one's emotions.

Finally, we considered whether women in our sample were simply more depressed than the male survivors, which may have reduced their overall motivation to engage in a wide range of behaviors, including CAMs. A recent review65 found that evidence is mixed regarding the rates of depression among male and female cancer survivors. Some studies have found higher rates of depression for female survivors, whereas others have found men and women with cancer to have similar levels of depressive symptoms. However, a comparison of the POMS depression-dejection subscale scores revealed no differences in depressive symptoms between the men and women in this study (data not shown).

In addition to assessing self-reported depressive symptoms, we also included 2 measures of anxiety. General anxiety was measured by the POMS-SF tension- anxiety subscale, and cancer-specific anxiety was measured with the Cancer Problems in Living Scale fear of recurrence subscale. In this study, both measures of anxiety were related to increased use of CAM in female CRC survivors. Given the finding with depression already noted, this result suggests that anxiety and depression may exert independent influences on CM use for women with a history of CRC. In both research and clinical settings, anxiety and depression are often grouped together under “emotional distress,” which may obscure the different influences these emotional states have on behavioral outcomes, such as CAM use. Specifically, anxiety may serve to motivate women to engage in practices (eg, CAM) that will either reduce their perceived risk of recurrence or help them manage the distress associated with the possibility of their disease recurring. At the same time, anxiety predicted overall CAM use but was not associated with any of the individual CAM domains.

Surprisingly, both fatigue and vigor (energy) were found to be positively associated with CAM use for women when spiritual mind-body methods modalities were included. This may seem paradoxical, as higher levels of fatigue are often associated with lower levels of energy. Thus, higher fatigue and lower energy would be expected to be associated with increased CAM use, as opposed to higher levels of both fatigue and energy predicting increased use of CAM. It may be that, as with depression and anxiety, fatigue and energy act independently and exert their own influence on CAM use. Fatigued women may use CAM to help manage this symptom even if they are experiencing low energy. At the same time, women with more vigor may have the energy to seek out and use CAM, whether or not they feel fatigued.


Although the somewhat provocative results of the current study provide new information regarding CAM use among male and female CRC survivors, our study had several limitations. First, <40% of the eligible survivors completed a study questionnaire, potentially biasing our results. Indeed, our study produced prevalence estimates that were somewhat higher than other studies have reported. Conversely, the sociodemographic and medical correlates of CAM use identified in this study are similar to other studies. In addition, when spiritually oriented CAMs were excluded, the prevalence estimates were more closely aligned with the existing literature.

Second, the measure of emotional functioning used in this analysis (the POMS-SF) provides only an indicator of self-reported emotional distress, and cannot be interpreted as measuring clinical depression or anxiety. Identifying individuals receiving care for clinically diagnosed mental health conditions (eg, depression) would more accurately indicate the role of emotional functioning in CAM use among cancer survivors. Nevertheless, these findings offer the ability to look at the relationship between emotional distress and the use of CAM. Furthermore, the intent of this study was not to compare CAM use among those with and without a mental health condition.

Finally, because these findings are based on survivors' rating of their cumulative use of CAM at a single point in time, no conclusions regarding the directionality or causality of the relationships can be drawn. As longitudinal data on CAM use among CRC survivors become available from this and other prospective studies, such issues can be explored further.


In conclusion, this study provides additional evidence that CAM use among CRC survivors is common and may be determined not only by disease and demographic factors, but also by emotional functioning. The data indicate that female CRC survivors are more likely than male CRC survivors to use CAM, and their use is more likely to be related to emotional functioning. We also found that the influence of psychological factors varied, depending on which factor was considered; most psychological factors predicted increased use of CAM, but depressive symptoms were inversely related to likelihood of CAM use only for the women in this study. Further research is needed to corroborate and further explicate this unexpected finding.


We acknowledge the cooperation and efforts of the cancer registries and public health departments from Alabama, Arizona, California (regions 2-6), Colorado, Connecticut, Delaware, Illinois, Iowa, Idaho, Maine, Massachusetts, Michigan, Minnesota, Nebraska, New Jersey, Ohio, Pennsylvania, Rhode Island, South Carolina, Washington, and Wyoming. We also thank the staff of the hospitals contributing cases to the participating cancer registries. Lastly, we salute the thousands of cancer survivors, their physicians, and their loved ones who contributed to the collection of these data. We assume full responsibility for analyses and interpretation of these data.

Conflict of Interest Disclosures

Cosponsored by the National Cancer Institute's Office of Cancer Survivorship, the Office of Cancer Survivorship of the Centers for Disease Control and Prevention, and the American Cancer Society's Behavioral Research Center.