Effects of an exercise and hypocaloric healthy eating intervention on indices of psychological health status, hypothalamic-pituitary-adrenal axis regulation and immune function after early-stage breast cancer: a randomised controlled trial

Many women experience emotional distress, depression and anxiety after a diagnosis of breast cancer, which can persist for prolonged periods, irrespective of the clinical treatment outcome [1, 2]. Studies show that a quarter of breast cancer patients have clinically important levels of emotional distress up to 12 months after treatment [3] and almost 50% of women experience depression and/or anxiety during this period [1]. Psychological stress and depression have been associated with hypothalamic-pituitary-adrenal (HPA) axis dysregulation in breast cancer survivors, including aberrations in diurnal cortisol rhythm [4, 5], and elevated levels of the inflammatory cytokine IL-6 [6] that may impair immune system functioning [7] and adversely impact upon survival [8, 9].

The importance of healthy lifestyle behaviours to prevent the adverse effects of weight gain after a breast cancer diagnosis is widely acknowledged and sustainable lifestyle change was recently identified as one of the key areas for future research [10]. Lifestyle interventions that incorporate exercise and healthy eating advice could also have an important role in modulating indices of psychological health status during the early recovery phase after breast cancer treatment. Physical inactivity and poor dietary habits are independently associated with depressive symptoms in breast cancer survivors [11, 12], whereas evidence suggests that regular exercise [13] and positive dietary change [14] can improve depressive symptoms and psychological wellbeing. By improving psychological health status, lifestyle interventions have the potential to impact upon the HPA axis regulation and enhance immune function in women recovering from breast cancer treatment. This could improve long-term outcome, as restoration of immune function after treatment is predictive of survival [15, 16].

We recently reported the results of a randomised controlled trial that investigated the effects of an exercise and hypocaloric healthy eating intervention on body weight and metabolic biomarkers associated with disease recurrence in women recovering from early-stage breast cancer treatment [17]. Women in the intervention group experienced a wide range of health benefits, including reductions in central adiposity, dietary fat intake, total cholesterol and resting diastolic blood pressure and an improvement in cardiopulmonary fitness [17]. Here, in the same patient cohort, we report the effects of the lifestyle intervention on indices of psychological health status, HPA axis regulation and immune function.

The two groups were generally well-matched at baseline for key demographic and anthropometric variables (Table 1). In addition, levels of depression and perceived stress were well-balanced across the two groups. Nearly all the women were white Caucasian and on average, were 7 to 9 months post-treatment. Compliance to the supervised exercise sessions was excellent, with women attending on average 84% (42 to 98%) of the sessions that were offered to them.

Depressive symptoms were reduced in the intervention group in comparison with controls at the 6-month follow-up (P = 0.004). This was due to a greater reduction in BDI-II scores from baseline in the intervention group (P <0.001) in comparison with the controls (P = 0.02; Table 2). A total of 15 (34%) of participants in the intervention group were at least mildly depressed at baseline (BDI-II ≥14 [19]), in comparison with 2 participants (4.5%) at the 6-month follow-up, whereas in the control group the number of at least mildly depressed participants remained constant (22%) throughout the study period (χ² = 5.93; P = 0.02). In contrast, the reduction in PSS score in the intervention group versus controls at the 6-month follow-up was not significant (P = 0.16), despite a significant reduction from baseline PSS score (P <0.001, Table 2).

A pronounced diurnal cortisol rhythm was evident at baseline in both groups, with highest levels being recorded in the morning sample and lowest values in the evening (Figure 1). Although a similar diurnal cortisol rhythm was apparent in both groups at the 6-month follow-up, an increase in diurnal salivary cortisol rhythm AUC was observed in the lifestyle intervention group in comparison with the controls (P = 0.03; Figure 2). This was attributable to a higher post-intervention morning salivary cortisol concentration (P = 0.03; Figure 1). In contrast, the lifestyle intervention had no effect on circulating levels of the inflammatory cytokines, IL-6 (P = 0.93) or TNF-α (P = 0.61; Table 2).

An increase in total leukocyte count in the controls in comparison with the intervention group was observed at the 6-month follow-up (P = 0.04; Table 3), attributable to an increase in lymphocyte count (P = 0.04) and neutrophil count (P = 0.05). The CD3⁺CD4⁺ and CD3⁺CD8⁺ T cell counts showed a different pattern of response between the two groups (P ≤0.05; Table 3). However, the lifestyle intervention had no effect on CD4⁺:CD8⁺ T cell ratio (P = 0.87), NK cell count (P = 0.46), NK cell cytotoxicity (P = 0.85) or lymphocyte proliferation in response to PHA stimulation (P = 0.11; Table 3 and Figure 2).

Although pragmatic lifestyle interventions could be important for helping to prevent the adverse effects of weight gain after breast cancer treatment [10], few studies have assessed the impact of combined exercise and healthy eating programmes on psychological wellbeing and associated prognostic biomarkers in breast cancer survivors [23, 24]. This is the first study to report the effects of a combined exercise and healthy eating intervention on psychological health status, indices of HPA axis regulation and immune function in overweight women during the early recovery phase (3 to 18 months) after stage-I to -III breast cancer treatment. There was excellent compliance with the exercise sessions (84%) and low attrition, with only three participants from each group being lost to follow up in this analysis of psychological and neuro-immunological variables. An improvement in depressive symptoms in the intervention group was not accompanied by a significant reduction in perceived stress. However, the lifestyle intervention also had a significant impact on diurnal salivary cortisol rhythm, as evidenced by an increase in morning salivary cortisol at the 6-month follow-up.

At baseline, BDI-II depression scores in our patient cohort were consistent with previous studies of breast cancer survivors [2, 11, 25]. Furthermore, the reduction in mean BDI-II depression score from baseline in the intervention group was similar to that previously reported following exercise interventions [25]. Improvements in depression scores have also been observed after lifestyle interventions involving a dietary component to promote healthy eating and healthy body weight maintenance [26, 27] but this has not been a consistent finding [23]. Using the criterion of 0.5 SD for our sample population at baseline [28], a reduction in BDI-II score >3 points is clinically meaningful. Hence, the 6.1-point reduction from baseline and between-group difference of >3 points at the 6-month follow-up suggests that women in the intervention group experienced a clinically important change in depressive symptoms. Depressive symptoms in women recovering from breast cancer treatment have been associated with fatigue, pain, sleep disturbances, low libido, poor financial status and social support, and a range of other psychological factors [2]. Physical inactivity and poor dietary habits have also been independently associated with depressive symptoms in breast cancer survivors [11, 12].

Evidence suggests that lifestyle interventions, which include support for exercise and dietary behaviour change, may improve depressive symptoms by improving mood, enhancing self-efficacy expectations, reducing intrusive thoughts about breast cancer and future morbidity/mortality and buffering self-concept perceptions [27, 29]. Reductions in body weight and/or body fat and improvements in cardiopulmonary fitness (as observed in the intervention participants [17]) could also have a positive impact on depression scores [30, 31]. In addition, biological mechanisms, including the release of monoamine neurotransmitters (that is, serotonin, dopamine and norepinephrine) and endorphins during exercise have been proposed to account for improvements in depressive symptoms in physically active individuals [32]. However, it is also important to be mindful of attention effects on depression symptoms, as an improvement in BDI-II score was previously reported in the attention control arm of a short-term exercise trial in a similar cohort of early-stage breast cancer patients [13]. The increased level of personal contact between the intervention group and researchers is a potential limitation of the present study, which needs to be taken into consideration. In summary, continued research is necessary to understand the extent to which such factors influenced the improvement in depressive symptoms in the present study and the impact of changes on long-term prognosis.

Although the intervention group also exhibited a significant reduction in PSS scores from baseline levels, the smaller difference between groups at the 6-month follow-up was not significant. PSS scores in our patient cohort at baseline were comparable to those reported previously for breast cancer survivors [33‐37]. However, the lack of effect on this outcome measure contrasts with previously reported reductions (in the range of 4 to 9 points) following stress management and yoga interventions in early-stage breast cancer patients [33‐36]. A large proportion of breast cancer survivors are reported to use complementary therapies (including exercise) to manage stress [38], but only one previous study has reported reductions in PSS scale scores following exercise training [37]. The reduction in PSS scores after a 10-week programme of aerobic and resistance exercise was only marginally greater in this study than the non-significant difference between groups in our patient cohort at the 6-month follow-up. Hence, the lack of a significant effect on perceived stress in the present study may be a reflection of low statistical power for this outcome.

The intervention had a significant impact on HPA axis regulation, as evidenced by an increase in diurnal salivary cortisol rhythm at the 6-month follow-up. This was attributable to a higher morning salivary cortisol concentration. Cortisol levels follow a strong circadian rhythm in healthy adults, with the highest levels being evident in the morning after awakening and the nadir occurring at around midnight [39]. A pronounced diurnal cortisol rhythm was apparent at baseline in our patient cohort, and this is consistent with previous studies of non-metastatic breast cancer patients and survivors [4, 5]. However, this is the first study to report a steepening of diurnal cortisol rhythm following a lifestyle intervention in breast cancer patients in the early recovery phase after treatment. The implications of this are difficult to define but flatter diurnal cortisol rhythms (blunted morning and/or elevated evening levels) have been associated with depressive symptoms, poor sleep quality and persistent fatigue after breast cancer treatment [5, 40]. Hence, the intervention may have influenced morning cortisol levels via its positive impact on one or all of these factors, or via its impact on other factors not monitored in this study. Interestingly, an increase in morning cortisol level, accompanied by improvements in depressive symptoms and psychological stress, was reported after an 8-week stress reduction programme involving yoga, meditation and group discussions in breast cancer patients [33]. This was interpreted as evidence that the intervention resulted in a normalisation of the cortisol response. In a similar way, our results may indicate a modest improvement in HPA axis regulation in the intervention group but further research is needed to confirm this and delineate the precise determinants of HPA axis modulation.

Circulating levels of the inflammatory cytokines IL-6 and TNF-α have also been used as evidence of aberrant HPA axis regulation in breast cancer patients and other populations [6, 7, 9]. Depression and psychological stress are characterised by elevated levels of IL-6 and TNF-α, [6, 41] and IL-6 stimulates cortisol secretion by adrenocortical cells [42]. However, in the present study, levels of these inflammatory cytokines were unchanged at the 6-month follow-up, and this is consistent with previous lifestyle intervention studies involving breast cancer survivors [23, 24]. Elevated levels of these inflammatory cytokines have more generally been associated with severe and/or chronic depression and more marked HPA axis dysregulation in previous studies [6, 9]. The pronounced diurnal cortisol rhythm at baseline in our patient cohort, coupled with evidence of mild depressive symptoms, suggests a more modest disruption of HPA axis regulation, which probably explains why the intervention had no effect on these inflammatory cytokines.

This is also consistent with our results for immune function. Previous studies show that high levels of stress and depression, coupled with the persistent activation of the HPA axis, impair immune responses and potentially contribute to the development and progression of cancer (reviewed by Reiche et al. [7]). However, indices of immune function (lymphocyte proliferation and NK cell cytotoxicity) were unchanged from baseline levels in the present study. Furthermore, the different pattern of response in CD3⁺CD4⁺ and CD3⁺CD8⁺ T cell counts is difficult to interpret, as baseline values were higher in the intervention group. Nevertheless, an increase in total leukocyte count, attributable to an increase in lymphocyte and neutrophil counts, was observed in the controls. The clinical implications of the latter are unknown but surgically treated early-stage breast cancer patients who experienced disease recurrence exhibited higher lymphocyte and neutrophil counts relative to disease-free patients in the 17 months prior to recurrence detection [8]. This contrasts with earlier evidence of an association between low lymphocyte counts and breast cancer recurrence [43]. Although changes in circulating leukocyte counts during recovery from cancer treatment may be indicative of specific and/or non-specific immune reactions to sub-clinical disease [8, 43], the clinical significance of such changes in our patient cohort is difficult to interpret without the longer-term follow-up of clinical endpoints such as disease-free interval and mortality.

In conclusion, this study shows for the first time that an improvement in depressive symptoms resulting from supervised exercise and healthy eating advice in the early recovery phase, 3 to 18 months, after breast cancer treatment is accompanied by a significant change in diurnal salivary cortisol rhythm. The increase in morning salivary cortisol at the 6-month follow-up suggests a normalisation of HPA axis regulation, which may be linked to the improvement in psychological health status. The precise determinants and clinical significance of these changes warrants further investigation.