Breast Cancer Research and Treatment

, Volume 141, Issue 2, pp 277–285

Use of complementary and alternative medicine (CAM) and changes in depressive symptoms from 3 to 15 months after surgery for primary breast cancer: results from a nationwide cohort study

Authors

    • Unit for Psychooncology and Health PsychologyAarhus University Hospital & Aarhus University
  • Søren Christensen
    • Unit for Psychooncology and Health PsychologyAarhus University Hospital & Aarhus University
  • Anders Bonde Jensen
    • Department of OncologyAarhus University Hospital
  • Robert Zachariae
    • Unit for Psychooncology and Health PsychologyAarhus University Hospital & Aarhus University
Epidemiology

DOI: 10.1007/s10549-013-2680-1

Cite this article as:
Pedersen, C.G., Christensen, S., Jensen, A.B. et al. Breast Cancer Res Treat (2013) 141: 277. doi:10.1007/s10549-013-2680-1

Abstract

Background This large population-based prospective study explored the associations between use of complementary and alternative medicine (CAM) and prevalence and changes in depressive symptoms from 3 to 15 months after surgery for primary breast cancer. Methods In an ongoing nationwide cohort study, depressive symptoms and the use of ten different types of CAM since time of diagnosis were assessed by questionnaire at 3 (N = 3233) and 15 months post-surgery (N = 2833). Clinical and socio-demopraphic variables were obtained from the Danish Breast Cancer Cooperative Group (DBCG) and national longitudinal registries. Results At 3 months post-surgery, 40.2 % reported having used CAM since the diagnosis, and in the time from 3 to 15 months after surgery, 49.9 % had used CAM. Compared with non-users, CAM users experienced more depressive symptoms at both 3 and 15 months follow-up when adjusting for several possible confounders. In a fully adjusted model, including all CAM variables and adjusting for depressive symptoms at 3 months, use of dietary/vitamin supplements was the only independent, statistically significant risk factor for experiencing more depressive symptoms at 15 months (RM = 1.10; 95 % CI 1.02–1.19, P = 0.014). Conclusion Cross-sectionally, CAM use was associated with higher levels of depressive symptoms at both time-points. Prospectively, users of dietary or vitamin supplements at 3 months after surgery reported higher levels of depressive symptoms than non-users 1 year later. Possible explanations could be an increased vulnerability for depressive symptoms among women using CAM or adverse effects of dietary or vitamin supplements.

Keywords

Breast cancerDepressive symptomsComplementary and alternative medicine (CAM)Cohort studiesHealth behavior

Introduction

Compared to women in the general population, breast cancer patients have been found to exhibit elevated levels of depressive symptoms 3 months after primary surgery [1]. In addition to negative consequences for the women’s psychological well-being, depressive symptoms have also been associated with poorer physical function, increased mortality, and increased use of health care services [2].

Although the findings reported in the literature are inconsistent [3, 4], cancer patients using Complementary and alternative medicine (CAM) have generally been found to be more distressed and to worry more about recurrence than non-users [5]. This could indicate that the internationally wide-spread and growing use of CAM among cancer patients reflects their attempts to cope with feelings of helplessness and depression. This interpretation is supported by studies showing that CAM use is associated with a more active coping style [4] and a higher degree of perceived control over the cancer disease [6], indicating that decisions to use CAM in relation to cancer could be interpreted as attempts to regain control and play an active role by taking responsibility of one’s own health [7]. CAM use may thus be viewed as a way of coping, which could have possible beneficial influences on emotional well-being and quality-of-life (QoL) [7]. Furthermore, some types of CAM have been suggested to have more direct effects on QoL. For instance, some herbal medicines and nutritional supplements, e.g., St. John’s wort and omega-3 fatty acids, have been reported to have some beneficial effects on mild to moderate depression [8].

Although the available results are inconsistent [9] and of varying methodological quality [5], beneficial effects of CAM have been published, with CAM-users reporting reductions in distress [10, 11] and improvements in emotional well-being [12]. On the other hand, there are also indications that some types of herbal medicines and dietary or nutritional supplements may lead to potentially life-threatening adverse events in combination with conventional anti-cancer treatments [8]. Furthermore, when used as the primary treatment, CAM has been related to disease progression and higher risk of recurrence and death, with more severe outcomes in those delaying or omitting surgery [13]. Possible effects of interactions between CAM and conventional anti-cancer treatment on emotional well-being, QoL, or depressive symptoms have so far not been studied.

Taken together, the available research appears to suggest that use of CAM may have beneficial influence on emotional well-being and overall QoL in cancer patients. The existing studies, however, generally suffer from being based on relatively small non-representative samples, and the majority have failed to differentiate between different types of cancer diseases, to control for the timing of assessment in relation to treatment and stage of disease, and to differentiate between specific types of CAM. Furthermore, most studies have used non-standardized or multi-dimensional QoL—or coping measures, and so far, only one study [9] have examined the possible influence of CAM on depressive symptoms measured prospectively using validated instruments.

We addressed these limitations by using prospective data from a large national cohort of women treated for primary breast cancer to explore the possible associations between the use of specific types of CAM and depressive symptoms assessed at 3 and 15 months after surgery with a well validated instrument, while adjusting for a number of possible confounders.

Patients and methods

Study design and materials

The present study utilized data from a nationwide prospective inception cohort of 4917 Danish women who underwent surgery for primary breast cancer between October 2001 and March 2004 and were treated in accordance with standardized guidelines provided by the Danish Breast Cancer Cooperative Group (DBCG). Details concerning the cohort have previously been published [1, 14]. Eligible women were informed about the study at the surgical departments, and the Charlson Comorbidity Index [15] was completed for each patient. At the first assessment, 3 months post-surgery, all eligible women were mailed additional information, an informed consent form, a questionnaire package, and a prepaid return envelope, and were invited to participate. A single reminder was mailed after 2 weeks. Approximately 15 months post-surgery, a follow-up questionnaire package was mailed to the participants. The participants received no compensation. The study was approved by the Regional Research Ethics Committees and The Danish Data Protection Agency.

Eligibility

Eligible patients were 18–70 years old resident Danish women with histologically confirmed breast cancer T1-3, N0-3, and M0 according to the TNM classification [16] with no history of other cancers, except non-melanoma skin cancer or carcinoma in situ of the cervix uteri. Additional requirements were: Ability to read Danish and complete a questionnaire.

Data collection

Addresses, CPR-numbers (personal identification numbers), and data concerning eligibility, comorbidity, disease status, and treatment-related variables, including tumor size, nodal status, tumor grade, ER/PR receptor status, type of surgery, chemotherapy, radiotherapy, and hormone therapy, were obtained from the surgical departments and the DBCG-registry. Validation studies suggest that at least 97 % of all eligible women receiving surgery for primary breast cancer in Denmark during the inclusion period were identified for the study [17]. Demographic data and information on socio-economic status and psychiatric history were collected from six nationwide Danish longitudinal registries through a linkage serviced by Statistics Denmark [18]. Additional information on health behaviors, health status, and psychosocial variables was obtained through questionnaires.

Assessment of depressive symptoms

The 21-item Beck Depression Inventory, Second Edition (BDI-II), is a widely used questionnaire developed for the assessment of symptoms corresponding to criteria for diagnosing depressive disorder listed in the Diagnostic and Statistical Manual of Mental Disorders—Fourth Edition (DSM-IV) [19]. According to the BDI manual, a cut-off score of 17 has yielded a 93 % true-positive rate and a 18 % false-positive rate for the presence of major depression [20]. Its two factor-analytically derived subscales measuring somatic- and cognitive/affective depressive symptoms makes it particularly useful in populations with somatic diseases. Internal consistencies (Cronbach’s Alpha) were satisfactory (Total: 0.90; Somatic symptoms: 0.75; Cognitive/Affective symptoms: 0.89) [1].

Assessment of CAM use

CAM use between the time of the cancer diagnosis and 3 months after surgery and between 3 and 15 months after surgery was assessed with an adapted version of the Danish health and morbidity survey [21], asking participants if they had used one or more of the following types of CAM during the time since their cancer diagnosis: (1) Herbal medicine, (2) Reflexology; (3) Relaxation or yoga; (4) Kinesiology; (5) Nutrition/exercise counseling; (6) Healing, laying on hands or similar; (7) Massage/manipulation; (8) Meditation; (9) Needle acupuncture; (10) Dietary or vitamin supplements beyond ordinary vitamin pills, or (11) Other types (specified). When possible, the last variable was re-coded independently and negotiated by two investigators into one of the 10 listed types of CAM. “No CAM use” was assessed with the control item: “No, I did not use any type of complementary or alternative treatment after I was diagnosed with breast cancer.”

Covariates

Covariates included the socio-demographic variables of marital status, number of children, educational level, occupational status, personal income, mean household net-wealth, ethnicity, urbanicity and psychiatric history, using relevant national registries (see [1] for further details). The socio-demographic variables refer to pre-cancer conditions either in the year prior to surgery minus 1 month or, when appropriate, at the date of surgery minus 1 month. Additional covariates included the weighted index score of the Charlson Comorbidity Index (CCI) [22], The physical activity scale for the elderly (PASE) [23], the 10-item physical functioning subscale (PF) of the MOS Short Form (SF-36) [24], body mass index (BMI) [25], smoking habits, and alcohol consumption presented in units per day (beers, glasses of wine, or drinks).

Missing values

Following the manual, missing values on the SF-36 PF subscale were substituted with the mean of the remaining completed items [24]. Likewise, missing values on the BDI were substituted with the mean of the remaining completed items of the relevant subscale. Subscale totals with >50 % missing values were coded as missing and no total score was calculated. When internal consistency (Cronbac’s alpha) >0.7, this method is considered preferable to other procedures, e.g., list-wise deleting or scale mean substitution of scale scores [26].

Statistical analysis

Negative binomial logistic regression analyses (NBREG) were used to evaluate the cross-sectional associations between CAM use and depressive symptoms at 3 and 15 months follow-up. The results are presented as ratios of means (RM), presenting the expected BDI score in a category relative to the expected BDI score in the reference category. First, unadjusted associations were calculated. Next, covariates were added to the analyses in 4 successive steps according to the phases of the women’s cancer history. Pre-cancer variables (age, socio-demographic, and socio-economic variables) were entered at step 1, health status at step 2, disease -and treatment-related factors at step 3, and physical function, BMI and health behaviors at step 4. The same analytical approach was used at 15 months post-surgery, but adjusting for depressive symptoms at 3 months. The NBREG analyses were conducted with STATA 12.1 and all other analyses with SPSS 19 for Windows.

Role of the funding sources

The funding sources had no involvement in the study design, collection, analysis, and interpretation of data, or writing of the paper.

Results

As previously reported, a total of 3343 women (68.0 %) returned a questionnaire 3 month after surgery [1]. The majority (77 %) was married or cohabiting and 30 % had a tertiary education. Fifty-one percent had a tumor positive axillary lymph node status, 54 % were treated with mastectomy, and 44 % received chemotherapy. Responders were younger than non-responders (median 55.7 vs. 58.0 years; range 23–70 years). Age-adjusted analyses revealed that participation was not influenced by tumor—and treatment related factors, and the sample can therefore be considered as nationally representative with respect to disease—and treatment-related variables [1]. At the follow-up, approx. 15 months post-surgery, 166 women, who had died or suffered a relapse, and 47 women with unknown disease status were excluded together with 2 women who had not met the original inclusion criteria [14]. A total of 2931 (94 %) of 3128 eligible women completed additional questions concerning CAM use and depressive symptoms.

Use of CAM

At 3 months post-surgery, complete data on BDI score and CAM use were available for 3233 women. A total of 1301 (40.2 %) women had used CAM since their diagnosis. At follow-up, 15 months post-surgery, complete data on the BDI and CAM use were available for 2833 recurrence-free women. A total of 1413 (49.9 %) had used one or more types of CAM since the first assessment. The use of specific types of CAM is shown in Tables 1 and 2.
Table 1

Associations between CAM use since diagnosis and depressive symptoms at 3 months post-surgery

CAMa

N

Depressive symptoms (BDI-II total)

BDI-II total

Unadjusted

+ Age and socio-demographicsb

+ Pre-cancer health statusc

+ Treatment and histopathologyd

+ Physical function, BMI, and health behaviorse

Mean

RM

(95 % CI)

RM

(95 % CI)

RM

(95 % CI)

RM

(95 % CI)

RM

(95 % CI)

CAM use

   

P < 0.001

 

P < 0.001

 

P < 0.001

 

P < 0.001

 

P < 0.001

 Yes

1301

9.70

1.17

1.11–1.24

1.16

1.09–1.23

1.17

1.10–1.24

1.17

1.10–1.24

1.20

1.13–1.27

 No

1932

8.27

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Dietary/vitamin supplements

   

P = 0.001

 

P = 0.012

 

P = 0.005

 

P = 0.011

 

P = 0.001

 Yes

891

9.58

1.12

1.05–1.19

1.09

1.02–1.16

1.10

1.03–1.17

1.09

1.02–1.16

1.11

1.05–1.19

 No

2342

8.56

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Massage

   

P = 0.016

 

P = 0.013

 

P = 0.011

 

P = 0.004

 

P = 0.004

 Yes

202

10.12

1.16

1.03–1.30

1.16

1.03–1.31

1.17

1.04–1.31

1.19

1.06–1.35

1.18

1.05–1.33

 No

3031

8.76

1.00

Referent

        

Herbal medicine

   

P = 0.015

 

P = 0.138

 

P = 0.059

 

P = 0.089

 

P = 0.047

 Yes

313

9.84

1.13

1.02–1.24

1.08

0.98–1.19

1.10

1.00–1.21

1.09

0.99–1.20

1.10

1.00–1.21

 No

2920

8.74

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Nutrition/exercise counseling

   

P = 0.032

 

P = 0.010

 

P = 0.008

 

P = 0.011

 

P = 0.001

 Yes

314

9.73

1.11

1.01–1.23

1.14

1.03–1.25

1.14

1.03–1.25

1.14

1.03–1.25

1.18

1.07–1.30

 No

2919

8.75

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Needle acupuncture

   

P = 0.006

 

P = 0.010

 

P = 0.007

 

P = 0.011

 

P = 0.026

 Yes

229

10.21

1.17

1.05–1.31

1.16

1.04–1.29

1.17

1.04–1.30

1.16

1.03–1.30

1.13

1.01–1.26

 No

3004

8.74

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Reflexology

   

P = 0.376

 

P = 0.354

 

P = 0.206

 

P = 0.186

 

P = 0.018

 Yes

210

9.29

1.05

0.94–1.18

1.06

0.94–1.19

1.08

0.96–1.21

1.08

0.96–1.22

1.15

1.02–1.28

 No

3023

8.81

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Relaxation, yoga, or similar

   

P = 0.024

 

P = 0.010

 

P = 0.006

 

P = 0.007

 

P = 0.003

 Yes

193

10.08

1.15

1.02–1.30

1.17

1.04–1.32

1.18

1.05–1.33

1.18

1.05–1.33

1.19

1.06–1.34

 No

3040

8.77

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Healing

   

P = 0.135

 

P = 0.255

 

P = 0.209

 

P = 0.217

 

P = 0.059

 Yes

172

9.70

1.10

0.97–1.25

1.08

0.95–1.22

1.08

0.96–1.23

1.08

0.95–1.23

1.12

1.00–1.27

 No

3061

8.80

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Meditation

   

P = 0.950

 

P = 0.894

 

P = 0.894

 

P = 0.863

 

P = 0.466

 Yes

108

8.89

1.01

0.86–1.18

1.01

0.86–1.19

1.01

0.86–1.18

1.01

0.86–1.19

1.06

0.91–1.24

 No

3125

8.84

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Kinesiology

   

P = 0.728

 

P = 0.675

 

P = 0.964

 

P = 0.865

 

P = 0.800

 Yes

77

8.56

0.97

0.80–1.17

0.96

0.80–1.59

1.00

0.83–1.20

0.98

0.81–1.19

1.02

0.85–1.23

 No

3156

8.85

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

RM ratio of means. RMs in bold differ significantly (95 % CI) from the reference group (RM = 1.00). CI confidence interval

aNo. of users = Prevalence of CAM use (No. of users within variable category)

bAdjusted for all socio-demographic variables (age, marital status, children, education, urbanicity, occupational status, personal income, household net-wealth, and ethnicity)

cAdjusted for all socio-demographic variables and pre-cancer health status (psychiatric history and comorbidity)

dAdjusted for all socio-demographic variables, pre-cancer health status, and clinical factors (tumor size, nodal status, tumor grade, ER/PR receptor status, surgery, chemotherapy, radiotherapy, and hormone therapy)

eAdjusted for all socio-demographic variables, pre-cancer health status, clinical factors, and physical function (SF-36), BMI, and health behaviors (physical activity (PASE), smoking, and alcohol consumption)

Table 2

Associations between depressive symptoms at 15 months post-surgery and CAM use between 3 and 15 month’s post-surgery

CAMa

N

Depressive symptoms (BDI-II total)

BDI-II total

Unadjusted

+ Age and socio-demographicsb

+ Pre-cancer health statusc

+ Treatment and histopathologyd

+ Physical function, BMI, and health behaviorse

Mean

RM

(95 % CI)

RM

(95 % CI)

RM

(95 % CI)

RM

(95 % CI)

RM

(95 % CI)

CAM use

   

P < 0.001

 

P < 0.001

 

P < 0.001

 

P < 0.001

 

P < 0.001

 Yes

1413

8.42

1.33

1.23–1.43

1.32

1.23–1.43

1.32

1.22–1.43

1.33

1.23–1.44

1.33

1.23–1.43

 No

1420

6.34

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Dietary/vitamin supplements

   

P < 0.001

 

P < 0.001

 

P < 0.001

 

P < 0.001

 

P < 0.001

 Yes

882

8.58

1.26

1.16–1.36

1.26

1.16–1.37

1.26

1.16–1.37

1.26

1.16–1.36

1.26

1.16–1.36

 No

1951

6.83

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Massage

   

P < 0.001

 

P = 0.002

 

P = 0.003

 

P = 0.002

 

P = 0.002

 Yes

402

8.71

1.22

1.09–1.36

1.19

1.07–1.32

1.18

1.06–1.32

1.19

1.07–1.33

1.18

1.06–1.31

 No

2431

7.16

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Herbal medicine

   

P < 0.001

 

P < 0.001

 

P < 0.001

 

P < 0.001

 

P = 0.001

 Yes

321

9.30

1.30

1.16–1.47

1.28

1.14–1.44

1.28

1.13–1.43

1.27

1.12–1.43

1.21

1.08–1.36

 No

2512

7.13

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Nutrition/exercise counseling

   

P = 0.002

 

P = 0.002

 

P = 0.002

 

P = 0.003

 

P = 0.005

 Yes

287

8.78

1.22

1.07–1.38

1.22

1.08–1.39

1.22

1.08–1.39

1.22

1.07–1.38

1.19

1.05–1.35

 No

2546

7.22

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Needle acupuncture

   

P = 0.018

 

P = 0.010

 

P = 0.009

 

P = 0.013

 

P = 0.016

 Yes

267

8.49

1.17

1.03–1.33

1.19

1.04–1.35

1.19

1.04–1.35

1.18

1.04–1.35

1.17

1.03–1.32

 No

2566

7.26

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Reflexology

   

P = 0.391

 

P = 0.314

 

P = 0.322

 

P = 0.318

 

P = 0.261

 Yes

251

7.78

1.06

0.93–1.21

1.07

0.94–1.23

1.07

0.94–1.22

1.07

0.94–1.23

1.10

0.96–1.25

 No

2582

7.34

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Relaxation, yoga, or similar

   

P = 0.047

 

P = 0.006

 

P = 0.009

 

P = 0.008

 

P = 0.001

 Yes

247

8.35

1.15

1.00–1.31

1.21

1.05–1.38

1.20

1.05–1.37

1.20

1.05–1.38

1.25

1.09–1.42

 No

2586

7.28

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Healing

   

P = 0.010

 

P = 0.048

 

P = 0.063

 

P = 0.049

 

P = 0.008

 Yes

162

9.01

1.24

1.05–1.46

1.18

1.00–1.38

1.16

0.99–1.37

1.18

1.00–1.38

1.23

1.06–1.44

 No

2671

7.28

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Meditation

   

P = 0.653

 

P = 0.359

 

P = 0.423

 

P = 0.478

 

P = 0.123

 Yes

109

7.71

1.05

0.86–1.28

1.10

0.90–1.34

1.08

0.89–1.32

1.07

0.88–1.31

1.16

0.96–1.40

 No

2724

7.36

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Kinesiology

   

P = 0.182

 

P = 0.306

 

P = 0.287

 

P = 0.295

 

P = 0.145

 Yes

54

8.87

1.21

0.92–1.59

1.15

0.88–1.51

1.16

0.88–1.52

1.16

0.88–1.52

1.22

0.93–1.59

 No

2779

7.35

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

RM ratio of means. RM in bold differs significantly (95 % CI) from the reference group (RM = 1.00). CI confidence interval

aNo. of users = Prevalence of CAM use (No. of users within variable category)

bAdjusted for all socio-demographic variables (age, marital status, children, education, urbanicity, occupational status, personal income, household net-wealth and ethnicity)

cAdjusted for all socio-demographic variables and pre-cancer health status (psychiatric history and comorbidity)

dAdjusted for all socio-demographic variables, pre-cancer health status, and clinical factors (tumor size, nodal status, tumor grade, ER/PR receptor status, surgery, chemotherapy, radiotherapy, and hormone therapy)

eAdjusted for all socio-demographic variables, pre-cancer health status, clinical factors, and physical function (SF-36), BMI, and health behaviors (physical activity (PASE), smoking, and alcohol consumption)

Associations between type of CAM use and depressive symptoms at 3 and 15 months post-surgery

The cross-sectional, unadjusted, and adjusted associations between CAM use and depressive symptoms at 3 and 15 months post-surgery are shown in Tables 1 and 2.

At 3 months post-surgery, CAM users exhibited more depressive symptoms than non-users, and the RM were only marginally affected when adjusting for the covariates in the analytical steps (range 1.16–1.20). In the fully adjusted analyses, users of dietary/vitamin supplements; massage; herbal medicine; nutrition/exercise counseling; needle acupuncture; reflexology; and relaxation, yoga, or similar (relaxation/yoga) were all found to report more depressive symptoms than non-users. Differences in depressive symptoms between non-users and users of healing, meditation, and kinesiology did not reach statistical significance.

At 15 months post-surgery, overall CAM use within the past year was also associated with more depressive symptoms. RM were again only marginally affected when including the covariates (range 1.32–1.33). Higher levels of depressive symptoms were associated with use of dietary/vitamin supplements, massage, herbal medicine, nutrition/exercise counseling, needle acupuncture, relaxation/yoga, and healing. The associations with reflexology, meditation, and kinesiology did not reach statistical significance.

CAM use and change in depressive symptoms from 3 to 15 months post-surgery

Table 3 shows CAM use since diagnosis measured 3 months post-surgery as a predictor of depressive symptoms at 15 months post-surgery, adjusting for depressive symptoms at 3 months. Both overall use of CAM and specific use of dietary/vitamin supplements and relaxation/yoga emerged as predictors of higher levels of depressive symptoms at 15 months, independent of depressive symptoms at 3 months and the included covariates of age, socio-demographic factors, pre-cancer health status, disease- and treatment related factors, physical function, BMI, and health behaviors. In a fully adjusted model including all CAM variables, use of dietary/vitamin supplements emerged as the only independent, statistically significant risk factor for higher levels of depressive symptoms at 15 months (RM = 1.10; 95 % CI 1.02–1.19, P = 0.014).
Table 3

CAM use at 3 months post-surgery as predictor of depressive symptoms at 15 months post-surgery, adjusted for depressive symptoms at 3 months post-surgery (baseline-adjusted)

CAMa

N

Depressive symptoms (BDI-II total)

BDI-II total

Baseline adjusted

+ Age and socio-demographicsb

+ Pre-cancer health statusc

+ Treatment and histopathologyd

+ Physical function, BMI, and health behaviorse

Mean

RM

(95 % CI)

RM

(95 % CI)

RM

(95 % CI)

RM

(95 % CI)

RM

(95 % CI)

CAM use

   

P = 0.002

 

P = 0.007

 

P = 0.012

 

P = 0.018

 

P = 0.002

 Yes

1210

9.59

1.10

1.03–1.17

1.09

1.02–1.16

1.09

1.02–1.16

1.08

1.01–1.15

1.11

1.04–1.18

 No

1819

8.23

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Dietary/vitamin supplements

   

P < 0.001

 

P = 0.002

 

P = 0.003

 

P = 0.006

 

P = 0.002

 Yes

827

9.47

1.13

1.06–1.21

1.12

1.04–1.20

1.11

1.03–1.19

1.10

1.03–1.18

1.11

1.04–1.20

 No

2202

8.51

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Massage

   

P = 0.326

 

P = 0.448

 

P = 0.403

 

P = 0.242

 

P = 0.200

 Yes

192

9.99

1.06

0.94–1.20

1.05

0.93–1.19

1.06

0.93–1.20

1.08

0.95–1.23

1.09

0.96–1.24

 No

2837

8.69

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Herbal medicine

   

P = 0.085

 

P = 0.186

 

P = 0.178

 

P = 0.227

 

P = 0.127

 Yes

287

9.74

1.10

0.99–1.21

1.07

0.97–1.19

1.07

0.97–1.19

1.07

0.96–1.19

1.09

0.98–1.21

 No

2742

8.67

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Nutrition/exercise counseling

   

P = 0.248

 

P = 0.302

 

P = 0.301

 

P = 0.428

 

P = 0.336

 Yes

295

9.61

1.06

0.96–1.18

1.06

0.95–1.17

1.06

0.95–1.17

1.04

0.94–1.16

1.05

0.95–1.17

 No

2734

8.68

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Needle acupuncture

   

P = 0.879

 

P = 0.828

 

P = 0.859

 

P = 0.681

 

P = 0.604

 Yes

216

10.14

0.99

0.88–1.11

0.99

0.88–1.11

0.99

0.88–1.11

0.98

0.86–1.10

0.97

0.86–1.09

 No

2813

8.67

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Reflexology

   

P = 0.292

 

P = 0.259

 

P = 0.253

 

P = 0.355

 

P = 0.385

 Yes

197

9.42

1.07

0.95–1.21

1.07

0.95–1.22

1.08

0.95–1.22

1.06

0.94–1.20

1.06

0.93–1.20

 No

2832

8.73

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Relaxation, yoga, or similar

   

P = 0.049

 

P = 0.051

 

P = 0.051

 

P = 0.045

 

P = 0.022

 Yes

183

10.06

1.13

1.00–1.29

1.13

1.00–1.29

1.13

1.00–1.29

1.14

1.00–1.29

1.16

1.02–1.32

 No

2846

8.69

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Healing

   

P = 0.196

 

P = 0.322

 

P = 0.322

 

P = 0.279

 

P = 0.334

 Yes

158

9.51

1.09

0.95–1.26

1.07

0.93–1.23

1.07

0.93–1.23

1.08

0.94–1.24

1.07

0.93–1.23

 No

2871

8.73

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Meditation

   

P = 0.344

 

P = 0.316

 

P = 0.317

 

P = 0.346

 

P = 0.374

 Yes

103

8.87

1.09

0.92–1.29

1.09

0.92–1.29

1.09

0.92–1.29

1.09

0.91–1.29

1.08

0.91–1.28

 No

2926

8.77

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

Kinesiology

   

P = 0.158

 

P = 0.141

 

P = 0.128

 

P = 0.261

 

P = 0.280

 Yes

67

8.09

1.16

0.94–1.43

1.17

0.95–1.44

1.17

0.95–1.44

1.13

0.91–1.39

1.12

0.91–1.38

 No

2962

8.79

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

1.00

Referent

RM ratio of means. RM in bold differs significantly (95 % CI) from the reference group (RM = 1.00). CI confidence interval

aNo. of users = Prevalence of CAM use (No. of users within variable category for women completing the BDI at both follow-up’s. BDI means at 3 months follow-up for women having completed the BDI at both follow-up’s

bAdjusted for depressive symptoms at 3 months follow-up and all socio-demographic variables (age, marital status, children, education, urbanicity, occupational status, personal income, household net-wealth, and ethnicity)

cAdjusted for depressive symptoms at 3 months follow-up and all socio-demographic variables and pre-cancer health status (psychiatric history and comorbidity)

dAdjusted for depressive symptoms at 3 months follow-up and all socio-demographic variables, pre-cancer health status, and clinical factors (tumor size, nodal status, tumor grade, ER/PR receptor status, surgery, chemotherapy, radiotherapy, and hormone therapy)

eAdjusted for depressive symptoms at 3 months follow-up and all socio-demographic variables, pre-cancer health status, clinical factors, and physical function (SF-36), BMI, and health behaviors (physical activity (PASE), smoking and alcohol consumption)

Finally, we explored possible interactions between CAM use and cancer treatment (chemotherapy and endocrine therapy). A statistically significant interaction was found between CAM use and chemotherapy in the fully adjusted analysis (RM = 0.80; 95 % CI 0.71–0.91, P = 0.001). A stratified analysis revealed that only CAM users (vs non-users) who had not received chemotherapy had more depressive symptoms from 3 to 15 months post-surgery (RM = 1.22; 95 % CI 1.11–1.34, p < 0.001). This was not the case for women who had received chemotherapy (RM = 0.99; 95 % CI 0.91–1.09, P = 0·88). For specific types of CAM, similar statistically significant interactions were found for dietary/vitamin supplements, healing, massage, and relaxation/yoga. No significant interaction was found for CAM use and endocrine therapy (data not shown).

Discussion

CAM use is prevalent among breast cancer patients and appears predominantly to be used with the intent of improving QoL [5]. In the present study of a large nationwide cohort of women treated for early breast cancer, we examined—both cross-sectionally and prospectively—the associations between CAM use and prevalence and changes of depressive symptoms at 3 and 15 months after surgery.

In concordance with previous cross-sectional results, we found the use of CAM from time of diagnosis to 3 months after surgery to be associated with more depressive symptoms [12]. Our study, however, is the first to show that this relationship exists not only for overall CAM use, but also for individual types of CAM, including dietary and vitamin supplements, massage, nutrition- and exercise counseling, needle acupuncture, and relaxation, yoga or similar. While this could be interpreted as if the use of these types of CAM may represent an attempts to cope with depressive symptoms, women having used these types of CAM, together with those using herbal medicine and healing within the past year, were also found to report more depressive symptoms 15 months after surgery when compared to non-users.

Our prospective results revealed, in contrast to most of the findings of previous studies of varying methodological quality which have suggested beneficial influences of CAM on adjustment, emotional well-being, or QoL [10, 11, 27, 28], that CAM users experienced 8–11 % more depressive symptoms from 3 to 15 months post-surgery when compared with non-users. This appeared to be particularly true for those using dietary/vitamin supplements, relaxation, and yoga or similar interventions. When we adjusted for the use of multiple types of CAM, the higher level of depressive symptoms at 15 months was found to be restricted to users of dietary/vitamin supplements only. Further explorative analyses showed that it was CAM users not treated with chemotherapy, rather than women who had received chemotherapy, who were at increased risk of more depressive symptoms at 15 months when compared to non-users.

The finding that CAM users fared worse than non-users was unexpected, and our data provide no clear explanation. One could speculate that some unknown confounding characteristics of CAM users, especially those using dietary and vitamin supplements, could make them both more vulnerable to depression and more attracted to CAM use. It is also possible that the patterns of CAM use vary according to different needs among women who receive chemotherapy and those who do not. We have, for example, previously shown that CAM use is more prevalent in the chemotherapy group when adjusting for socio-demographic- and socio-economic factors [29]. Finally, attention has recently been drawn to so-called “holistic sickening,” which refers to a commonly held view among many CAM practitioners that cancer is merely a symptom of deeper-rooted problems beyond the cancer itself, and that patients therefore—to some extent—are responsible for their own cancer [30]. One could speculate that if patients subscribe to this view, this could have a negative influence on their psychological well-being. Finally, it is also a possibility that some types of dietary and vitamin supplements could have more direct effects on the patients’ emotional states acting through biological mechanisms. In line with this thinking, a recent meta-analysis of non-clinical samples found effects of vitamin and mineral supplements on stress, mood and mild psychiatric symptoms but not on depression [31]. In any instance, considering that the present findings suggest that use of some forms of CAM—especially oral CAM supplements—could be indicative of more depression, it may be important for clinicians to discuss use and motives for use of these with their patients, as a means of promoting evidence-based and patient-centered care as well as informed decision-making regarding CAM.

This first, large population-based, prospective study of CAM use and depressive symptoms has several strengths, including our assessment of the use of specific types of CAMs, controlling for use of multiple CAMs as well as a number of potential socio-demographic and clinical confounders, and adjusting for depressive symptoms at baseline. One limitation of the present study is that we were unable to identify individual types of dietary or vitamin supplements. Another is that some participants may have been unable to distinguish between herbal medicines and dietary or vitamin supplements. Finally, it should be noted that participants were younger than non-participants and that women older than 70 years were not included in the study.

Conclusion

Several types of CAM were associated with more depressive symptoms at both 3 and 15 months post-surgery, suggesting that many types of CAM may be used by patients in attempts to cope with psychological distress. However, in contrast to our expectations, we found no evidence that CAM use was associated with fewer depressive symptoms over time. To the contrary, CAM users who had not received chemotherapy, in particular users of dietary/vitamin supplements between diagnosis and 3 months after surgery, experienced more depressive symptoms 1 year later compared with non-users. Our findings call for further studies of possible adverse effects of specific types CAM on depressive symptoms and QoL in general.

Acknowledgments

We thank the participating women, the staff at the participating surgical departments, Susanne Møller, the Danish Breast Cancer Cooperative Group (DBCG), Rigshospitalet, Copenhagen University Hospital and Michael Væth, Department of Biostatistics, Aarhus University. We also thank the Danish Cancer Society and the Knowledge and Research Centre for Complementary and Alternative Medicine (ViFAB) for funding the study.

Conflict of interest

None to declare.

Copyright information

© Springer Science+Business Media New York 2013