Mindfulness and its efficacy for psychological and biological responses in women with breast cancer

1Although there have been enormous improvements in breast cancer diagnosis and advances in treatment, less attention has been paid to alleviating patients' breast cancer experience by preserving their physical, functional, and psychosocial well‐being 1. Women with breast cancer are challenged to cope over time with a high symptom burden and distress, which affects their well‐being and quality of life 2, 3, 4, 5, 6, 7. The prevalence of mood disorders is highest in the first year after breast cancer diagnosis and then decreases gradually over time 8.

2Still, individuals report persistent coexistent physical and psychological symptoms that contribute to interference with daily life after breast cancer treatment 9.

3An increasing body of research has established an association between distress and changes in immune function 10, 11. Distress seems to have a significant negative effect on immune function, such as lowered natural killer cells (NK cells) and T lymphocytes (T cells) 11, 12. T cells have been linked to breast cancer recurrence and survival 13, 14, 15. Other important parameters are cytokines, such as interleukin‐6 (IL‐6) and interleukin‐8 (IL‐8), which are independently correlated with breast cancer disease stage and progression 16, 17, 18, 19.

4Thus, previous research indicates a significant need for interventions to improve well‐being, alleviate distress and symptom burden, and to reinforce immunity in women during breast cancer diagnosis, treatment, and recovery.

5Originating from ancient Buddhist and yoga traditions, mindfulness‐based interventions have become increasingly popular in the Western world. Mindfulness is described as a “way of being” and defined as the capacity for awareness in each moment, by “paying attention in a particular way: on purpose, in the present moment, and nonjudgmentally”20. The use of mindfulness‐based interventions in oncology has proliferated over the past decade and research in the field has rapidly expanded 21, 22 While there appears to be evidence to support the use of mindfulness‐based interventions with cancer patients, the overall quality of existing trials varies considerably 23.

6Mindfulness‐based stress reduction (MBSR) is an 8‐week, standardized program combining mindfulness meditation, yoga and other techniques designed to reduce stress and improve well‐being and quality of life in patients with a wide range of chronic pain and stress disorders 20, 24, 25. MBSR and has also been shown to improve mood disorders 21, 26and reduce stress in cancer patients 27, 28. Furthermore, MBSR reduces fear of recurrence and improves physical functioning which in turn leads to reduced stress and anxiety in women with breast cancer 29. Evidence from nonrandomized, uncontrolled studies suggests that MBSR improves quality of life and coping, decreases stress and alters cortisol and immune patterns 30, 31, 32, 33.

7These results raise important questions as to whether MBSR is related to positive outcomes in mood disorders, symptom burden and health status, as well as strengthened immune system functioning in breast cancer survivors. Despite growing evidence that MBSR influences immune function, there is a need for studies to determine how biomarkers relate to changes in mindfulness and psychosocial outcomes, including symptom reduction and well‐being 34. While noting the effectiveness of MBSR, authors of several reviews have pointed out the inherent methodological problems in the published studies 35, 36. There is also a need for randomized controlled studies with long‐term follow‐up 36.

8The primary purpose of our study was to determine the efficacy of MBSR intervention for mood disorder symptom improvements in women with breast cancer. Secondary goals were to evaluate their symptom experience, distress, health status, coping capacity, mindfulness, posttraumatic growth, and immune status.

9In this 3‐month follow‐up study, we present the first results of a 5‐year longitudinal, randomized, controlled trial (RTC). Details of this trial have been described elsewhere 37.

10The trial was designed in accord with Consort recommendation 38, 39, 40. In an unblinded RTC, participants' expectations about the intervention may lead to a placebo effect in the intervention group and/or a negative response among controls. In order to minimize a potential placebo effect in the active intervention group and a “frustrebo response”41 in controls, a three‐armed design was chosen.

11Patients diagnosed with breast cancer were consecutively recruited to participate after completion of adjuvant chemotherapy and/or radiation therapy, with or without endocrine therapy. Patients were excluded on the basis of having another advanced illness at diagnosis that might interfere with the ability to participate, ongoing major depression, ongoing Herceptin therapy, or who had previously, as well as during the intervention, used MBSR and other mind‐body programs (including yoga).

12This trial was approved by the Regional Ethical Review Board, University of Gothenburg, and informed consent was obtained before enrolment.

13Eligible patients were contacted by research nurses at the first follow‐up appointment for patients receiving hormonal therapy or at the last treatment for patients undergoing chemotherapy. After oral and written information, interested patients provided written consent to participate in the study. Participants were first invited to a baseline health check‐up appointment, which included blood sample collection and questionnaire completion. Randomization was computerized and conducted in blocks of 9, 12, and 15, varied randomly. Assignment codes were kept in sequentially numbered, opaque, sealed envelopes, prepared by the research coordinator.

14Participants were randomized into one of three groups:

15MBSR (8 weeks self‐instructing MBSR program + instructor and weekly group sessions), active controls (8 weeks self‐instructing MBSR program) or non‐MBSR (no intervention).

16Participants in the MBSR group attended a standardized, group‐based, 8‐week course once a week for an average of 2 h each week with homework assignments consisting of 20 min sessions, 6 days/week. Participants were provided with information material, including a 20‐page introduction to mindfulness training, a compacted disk (CD) with meditation exercises, the training program and a diary in order to report the time allotted to mindfulness training including patients′ reflections about the meditation exercises. Led by a certified MBSR instructor, these weekly group sessions focused on the participants' experiences of mindfulness, and including gentle meditation and yoga training. Active controls received information material, a CD, 8 weeks of self‐instructing training program and a diary. The only difference between MBSR group and active controls was the weekly group sessions. Participants in both MBSR group and active controls were provided with written and oral instructions how to use information material, CD and diary. All participants received standard care for follow‐up for breast cancer according to the national and local guidelines 42.

17Socio‐demographic data were collected through chart review and interviews. Clinical characteristics, patient self‐reported outcomes and biomarkers were collected at health checks both pre and postintervention. Follow‐ups for MBSR group and active controls were conducted 1 month after the intervention, and at similar time points. The same procedures, at similar time points of 3 months were conducted for those in the non‐MBSR group.

18Mood disorder was measured using the Hospital Anxiety and Depression scale (HAD), which is one of most widely used instruments to screen for anxiety and depression in cancer patients 43, 44, 45. The HAD is a 14‐item questionnaire consisting of two subscales: anxiety and depression. Subscale scores range from 0 to 21; scores for each subscale are defined as: 0–7 (normal), 8–10 (possible cases), and 11–21 (cases of psychological morbidity) 46. The internal consistency of reliability for both subscales are satisfactory, with Cronbach's alpha 0.72–0.89, respectively, 0.78–0.93 45

19Symptom experience was evaluated using the Memorial Symptom Assessment Scale (MSAS), a questionnaire consisting of 32 symptoms and symptom frequency, severity, and distress 47. The MSAS generates two subscales including physical and psychological symptoms, and two global indicators: Total Symptom Burden Scale (TMSAS) and the Global Symptom Distress Index (GDI). The MSAS is a reliable and valid multidimensional measure of symptom experience in cancer populations 47, 48 including the Swedish version of the MSAS 49.

20Health status was measured using the 36‐item Short Form Health Survey (SF‐36), which consists of eight scaled scores: vitality, physical functioning, bodily pain, general health perceptions, physical, emotional and social role functioning, and mental health. The maximum score is 100 points. Reliability measurements of the SF‐36 are consistently good 50, 51, 52, 53.

21Coping capacity was evaluated using the Sense of Coherence scale (SOC) 54, 55, which consists of a 7‐point Likert scale evaluating perceived comprehensibility (5 items), manageability (4 items) and meaningfulness (4 items). Higher scores represent stronger sense of coherence. Reliability and validity of the SOC scale have been demonstrated, with Cronbach's α ranging from 0.74 to 0.93 56, 57, 58.

22Mindfulness was measured using the 29‐item short form Five Facets of Mindfulness Questionnaire (FFMQ–Swedish version), consisting of five key facets of mindfulness: observing, describing, acting with awareness, nonjudging, and nonreactivity to inner experience 59, 60.

23Personal growth was evaluated using the Posttraumatic Growth Inventory (PTGI), which measures positive life changes after traumatic events. The PTGI yields a total score based on five dimensions: relating to others, new possibilities, personal strength, spiritual change, and appreciation of life 61. The PTGI has shown good reliability in previous research with a total score Cronbach's α of 0.96 62.

24Lymphocyte distribution in peripheral blood was analyzed by flow cytometry using a FACSCanto II flow cytometer and the FACSDiva software. The absolute number of blood lymphocytes was determined with Trucount reference beads using the method recommended by the manufacturer. The following subpopulations were reported CD3+, CD3+4+ and CD3+8+T cells, CD19+B cells, and CD3‐16+56+NK cells. The results for each subpopulation were expressed as the percentage of lymphocytes and as the number of cells × 109/L. Antibodies to the antigens above. Trucount beads, the FACSCanto II flow cytometer and the FACSDiva software were all from BD Biosciences, Mountain View, CA.

25NK‐cell activity was measured using a Flow‐cytometric Assay of Natural Killer cell Immune response in Activated whole blood (FANKIA) a modified version of a previously published method using flow cytometry and stained K562 cells as target cells 63. Whole blood was mixed with a defined number of target cells transfected with the gene for green fluorescent protein (GFP) 64. After incubation the same volume was collected from tubes with: blood and target cells, target cells and medium; and blood and medium and analyses were performed as described above. The lytic activity was defined as the reduction in the number of target cells after mixing with the blood, expressed in percentage of target cells.

26Determination of cytokine concentrations were determined in sera using commercial high sensitivity IL‐6 and IL‐8 ELISA kits (R&D Systems, Inc., Abingdon, UK) according to the instructions from the manufacturer.

27Sample size calculation was based on the primary outcome: breast cancer patient's mood disorder symptoms. A one‐unit change on the HAD‐subscales from baseline to 3‐month follow‐up was regarded as clinically relevant. The detection of such a difference would require 50 participants per group (a total of 150 participants) to achieve a statistical power of 80%. Descriptive statistics were used to summarize socio‐demographic and clinical characteristics. Spearman's correlation coefficients were calculated to determine the strength of relationships between selected variables. As most of the variables we explored were of ordinal data type and most of the continuous variables deal with skewed distributions deviating from normal‐distribution, we used nonparametric tests (Wilcoxon's test for comparison within groups and Mann–Whitney's test for comparison between groups).

28P < 0.05 was considered as statistically significant result.

29A total of 177 women consented participation and were randomly assigned to one of three groups. There were 11 drop‐outs after randomization, that is, two patients were excluded as they did not complete the intervention, two patients withdraw their participation due to rapid breast cancer disease progression, and seven patients did not visited first follow‐up (MBSR = 4; active controls = 5; non‐MBSR = 2).

30The final groups were MBSR (n = 62), active controls (n = 52) and non‐MBSR (n = 52). Postintervention data were missing for one active control participant. A participation flowchart is depicted in Figure 1.

31Participants' ranged from 34 to 80 years (mean = 57.2, SD = 10.2). No statistical differences were found between groups on demographic or clinical characteristics. Participant descriptions are listed in Table 1.

32Study results revealed significant changes in psychological and biological responses to the MBSR intervention, summarized in Tables 2, 3, 4, 5, 6, 7, 8.

33On the primary outcome measures, MBSR participants reported significant improvements on depression symptoms both within group (mean = 4.3; SD = 3.7 to mean = 3.3; SD = 3.3; P = 0.001) as well as compared to non‐MBSR (P = 0.015), but not on anxiety symptoms (Table 2).

34The number of reduced cases of depression were 7 (11%) in MBSR participants compared to 4 (8%) reduced cases of depression in active controls and non‐MBSR, respectively.

35Pre and postintervention HAD scores for depression and anxiety are presented in Figure 2.

36There were significant improvements in physical symptoms (mean = 0.7; SD = 0.5 to mean = 0.6; SD = 0.4; P = 0.007), psychological symptoms (mean = 1.4; SD = 0.8; to mean = 1.2; SD = 0.9; P = 0.008), and total symptom burden (mean = 0.8; SD = 0.5 to mean = 0.7; SD = 0.5; P = 0.004) within the MBSR. There were also significant improvements between MBSR versus non‐MBSR regarding psychological symptoms (P = 0.019), as well as global distress (P = 0.013) (Table 3).

37Within the MBSR group, changes in health status were seen in improved vitality (mean = 49.5; SD = 27.5 to mean = 60.9; SD = 20.1; P < 0.001), physical functioning (mean = 77.4; SD = 16.7 to mean = 83.1; SD = 15.4; P < 0.001), general health perceptions (mean = 61.2; SD = 21.6 to mean = 67.6; SD = 19.1; P = 0.003), physical role functioning (mean = 45.9; SD = 42.9 to mean = 63.3; SD = 39.7; P = 0.003), and mental health (mean = 67.9; SD = 19.0 to mean = 74.1; SD = 17.1; P < 0.001). Changes were reported in physical functioning within the non‐MBSR group (mean = 78.5; SD = 19.6 to mean = 82.8; SD = 19.0; P = 0.007), physical role functioning with active controls (mean = 42.2; SD = 42.6 to mean = 67.6; SD = 35.8; P = 0.001), non‐MBSR (mean = 40.4; SD = 41.5 to mean = 59.6; SD = 42.3; P = 0.006), social functioning for active controls (mean = 78.4; SD = 25.9 to mean = 84.8; SD = 22.8; P = 0.017), non‐MBSR (mean = 72.1; SD = 26.0 to mean = 84.1; SD = 25.3; P = 0.001). A significant improvement between groups was reported in mental health for the MBSR group (P = 0.001) and active controls (P = 0.038) compared to non‐MBSR (Table 4).

38Additional secondary outcome measures showed that MBSR participants reported improved coping capacity (P = 0.028) versus non‐MBSR (Table 5).

39Enhanced elements of mindfulness were shown regarding Nonreactivity within MBSR (mean = 2.9; SD = 0.7 to mean = 3.3; SD = 0.5; P < 0.001), and between groups compared to non‐MBSR (P = 0.010), as well as on Observe both within MBSR group (mean = 3.3; SD = 0.7 to mean = 3.6; SD = 0.5; P < 0.001) and compared to non‐MBSR (P = 0.006) (Table 6). Enriched posttraumatic growth was reported within MBSR (mean = 59.78; SD = 19.5 to mean = 64.65; SD = 17.7; P = 0.005), and between groups for active controls versus non‐MBSR (P = 0.049) (Table 7).

40Mean baseline NK‐cell activity increased significantly (mean = 19.1; SD = 8.2 to mean = 22.0 SD = 7.7; P = 0.015) within the MBSR group. The absolute number of CD19+B‐lymphocytes increased (mean = 0.15; SD = 0.16 to mean = 0.17; SD = 0.15; P = 0.004). The proportion of CD3+T‐lymphocytes decreased (mean = 72.8; SD = 11.1 to mean = 71.4; SD = 10.5; P = 0.027) as did that of CD3+8+T‐lymphocytes (mean = 29.6; SD = 11.3 to mean = 28.4; SD = 10.5; P = 0.035), whereas the proportion of CD19+B‐lymphocytes increased (mean = 9.5; SD = 8.4 to mean = 11.6; SD = 7.6; P = 0.001). Analyses also demonstrated significant changes (P = 0.041) between MBSR participants and non‐MBSR regarding decrease in the absolute number of NK cells (CD3‐16+56+NKx10e9/l).

41There were also some significant changes pre and postintervention in active controls and in the non‐MBSR group suggesting a decrease in the absolute number of NK cells (CD3‐16‐56+NKx10e9/I) for active controls (P = 0.011) compared to non‐MBSR. There were also significant changes for active controls regarding CD3‐16+56+NK% versus non‐MBSR (P = 0.025) (Table 8). There were no significant differences in serum concentrations of IL‐6 or IL‐8 between any of the study groups.

42Our trial provides evidence in support of the efficacy of MBSR for psychological and biological response among women with breast cancer. The primary purpose of this study was to determine the efficacy of MBSR intervention on mood disorder, that is, depression and anxiety. Our finding of improvements in depression is consistent with other RTCs that have evaluated MBSR and mood disorders in breast cancer patients 21, 26. Unlike those studies, our findings revealed no significant changes in anxiety. However, consistent with our trial, a meta‐analysis of mindfulness‐based interventions that included participants who met the diagnostic criteria for a current episode of anxiety or depressive disorder show that MBSR is effective for reducing symptoms of depression, but not anxiety 65.

43MBSR participants reported significantly greater improvements in symptoms, especially psychological symptoms. In addition, their symptom burden and distress significantly decreased. MBSR participants also improved in functional status; in line with previous research showing significant intergroup improvements in mental health 66, our findings indicate significant improved mental health between groups.

44A common assumption is that mindfulness increases the individual's ability to cope, but few RCTs have examined the effect of MBSR on coping capacity. The MBSR intervention appears to improve coping effectiveness in breast cancer patients 32, and behavioral and cognitive coping 67. Results from our trial show that women who participated in the MBSR experienced improved coping capacity, here measured as sense of coherence (SOC). Previous research has identified SOC as a significant predictor of distress, number and type of coping strategies in women with breast cancer 68, suggesting the lower the SOC, the higher the levels of symptom burden 3. While Antonovsky 54 believed that SOC is a relatively stable personality state, our findings show evidence that MBSR may influence SOC (i.e., to improve patients' ability to manage, comprehend, and finding meaning living with breast cancer).

45Enhanced elements of mindfulness were shown for non‐reactivity and observing in the MBSR group. Future research is needed to explore the complexity and relations among the different dimensions of mindfulness, and to gain a deeper understanding about which factors facilitate the cultivation of mindfulness 69.

46Our trial indicates that the benefits of MBSR may also extend to posttraumatic growth. Relatively little research has investigated the relationship between posttraumatic growth and immunity. A study of patients with hepatoma suggests that higher PTGI scores are associated with higher peripheral blood leukocytes and longer survival 70. Further research addressing the interrelationship of MBSR with posttraumatic growth and immune response is warranted.

47In terms of biological response, changes in NK‐cell activity and numbers of both NK cells and B cells within the MBSR group as well as between groups were seen. Of note was the finding that there were no changes in numbers of IL‐6 or IL‐8. The clinical relevance of these discrete findings is difficult to estimate and more research is needed to fully explain the clinical meaning of these biological parameters. However, consistent with our findings, there is intriguing evidence suggesting that finding meaning and personal growth is associated with T‐cell levels 71 and NK‐cell activity 72. Furthermore, there have been prior reports that distress is associated with immune downregulation, including reduced NK‐cell activity 12, 32. In addition, participation in mindfulness training leads to a shift from proinflammatory response in cancer patients 33 and a pilot study has suggested that improvements in well‐being following MBSR was associated with increased NK activity and decreased CRP levels 73.

48Given the beneficial efficacy of MBSR on both psychological and biological outcomes, future longitudinal studies may be needed to investigate the effect of these outcomes on disease progression and survival. There were several tendencies in favor for MBSR group, although not statistically significant, and for some outcome variables statistically significant changes could be detected only within MBSR group but not between the groups. A larger sample size might have resulted in more significant differences between groups. Several improvements were also seen in active controls (i.e., physical and social role functioning, and observing) as well as between groups (i.e., global distress, mental health, and posttraumatic growth). Future research is needed to explore who benefits from participating in an MBSR intervention with weekly group sessions versus using a self‐instructing training program.

49This study is characterized by several strengths, including use of active and non‐MBSR controls, random assignment, inclusion of patient‐reported outcomes and immune response among a homogenous group of women diagnosed with breast cancer. To our knowledge, this is the first MBSR intervention study to include a comparison between standardized MBSR and both active controls using a self‐instructing program and passive controls. The notable study limitation was that all women who fulfilled the inclusion criteria were invited without undergoing screened for mood disorder before study invitation. Furthermore, despite randomization, there were differences in distribution regarding disease stage (tumor size and type of breast cancer) which might have affected study results.

50In conclusion, results from this RCT suggest that MBSR is beneficial and leads to psychological and biological improvements. MBSR may hold potential for alleviating depression, distress and symptom experience, and to strengthen coping capacity, which may improve breast cancer survivorship. Since there were also positive changes in the active control group, it is important to provide the self‐training program to patients who prefer to practice themselves, without weekly group exercises. Finally, longitudinal studies are required to investigate whether these positive psychological and biological responses remain constant, increases or decreases over time.

51The author(s) indicated no potential conflicts of interest.