Effects of self-acupressure on the fatigue-sleep disturbance-depression symptom cluster in breast cancer survivors: a phase III randomised controlled trial.

Effects of self-acupressure on the fatigue-sleep disturbance-depression symptom cluster in breast cancer survivors: a phase III randomised controlled trial

Abstract

Purpose

To explore the effectiveness and safety of self-acupressure intervention on the fatigue-sleep disturbance-depression symptom cluster (FSDSC) among breast cancer (BC) survivors and to explore within-trial costs associated with the intervention.

Methods

Eligible participants were randomly allocated to a true self-acupressure group, a sham self-acupressure group or a usual care group (36 in each). The primary outcomes were the FSDSC at both the individual symptom level and cluster symptom level. Individual symptoms were measured by the Brief Fatigue Inventory, the Pittsburgh Sleep Quality Index, and the Hospital Anxiety and Depression Scale-Depression, respectively. The cluster level was calculated as the mean of three separate 0-10 numeric rating scale scores for fatigue, depression, and sleep disturbance. The secondary clinical outcome was quality of life (QoL) measured by Functional Assessment of Cancer Therapy-Breast (FACT-B). Outcomes were assessed at baseline (T0, week 0), immediately after intervention (T1, week 7) and after follow-up (T2, week 19). Safety and related costs within trial were also reported.

Results

Ninety-three participants completed the study. Significant group-by-time interactions were observed for all clinical outcomes and QoL. Subsequent pairwise comparisons showed that the true self-acupressure group had significantly better FSDSC than the other groups at T1 and T2 (all p < 0.01). At the individual symptom level, the true self-acupressure group also had significantly better PSQI and HADS-D scores than both comparison groups at T1 and T2 (all p < 0.05). For BFI, the true self-acupressure group had significantly lower scores than the usual care group at both T1 and T2 (both p < 0.0001), whereas the difference between the true and sham self-acupressure groups was only statistically significant at T2 (p = 0.0076). For QoL, the true self-acupressure group had significantly higher FACT-B total scores than the usual care group at T1 and T2 (both p < 0.0001). The difference in FACT-B total score between the true and sham self-acupressure groups was not statistically significant at T1 (p = 0.069) but was statistically significant at T2 (p = 0.0002). No adverse events were reported. During the trial, only one participant in each of the true self-acupressure and usual care groups reported healthcare resource use (medications plus clinical visits for symptom relief), with a sum cost of ¥931.06 and  ¥2247.44, respectively. The one-off cost of video production for the true self-acupressure group was ¥2879.

Conclusions

The self-acupressure is a safe intervention that demonstrated positive effects in managing the FSDSC in BC survivors. While self-acupressure is recommended to support symptom management in cancer survivorship, its cost-effectiveness remains uncertain and warrants further investigation.

Trial registration

ClinicalTrials.gov (NCT06412107), Date: 02/26/2024. The study protocol has been published.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1007/s00520-026-11155-2.

Keywords: Self-acupressure, Breast neoplasm, Fatigue, Sleep disorder, Depression, Symptom cluster, Randomised controlled trial

Introduction

With the advanced technologies in early detection and treatments, there has been considerable increase in the number of cancer survivors globally. There were over 53 million cancer survivors with 5-year survivorship worldwide in 2024 [1]. Among all cancer types, breast cancer (BC) in women has a relatively high 5-year survival rate of over 90% in high-income countries compared to over 80% in China [2, 3]. However, BC survivors often experience unpleasant symptoms during or after their cancer treatment which negatively affect their quality of life (QoL), particularly when these symptoms co-occur or interact biologically as a symptom cluster [4–6]. The fatigue-sleep disturbance-depression symptom cluster (FSDSC) is a common one among BC survivors [4, 7]. The occurrence of the FSDSC was mainly explained by the cytokine-induced sickness behaviour, of which pro-inflammatory cytokines trigger the fatigue, depression, and sleep disturbance [8, 9].

Effective management for the FSDSC in BC survivors is challenging. Guidelines from the Oncology Nursing Society, the National Comprehensive Cancer Network, and the American Society of Clinical Oncology emphasise the priority of non-pharmacological interventions in managing symptom clusters [10]. This is due to safety concerns (e.g. potential drug-drug interactions) and the absence of consistent evidence supporting pharmacological approaches [10]. Current evidence suggests that non-pharmacological interventions could be beneficial for managing the FSDSC in BC survivors [11, 12]. Cognitive behavioural therapy (CBT), acupressure, acupuncture, and physical exercise are proposed as effective and practical choices, since they may affect the physiological pathways associated with the FSDSC, significantly reducing the levels of pro-inflammatory cytokines [13–15] and thus potentially improving the FSDSC. However, the use of acupuncture may be constrained by potential risks such as soft tissue infection and subcutaneous ecchymosis given its invasive nature [16, 17]. Physical exercises such as dance or resistance exercise programmes are often energy-demanding, which may reduce participants’ interests and adherence, particularly among BC survivors experiencing persistent fatigue [18]. CBT typically requires multiple sessions with trained therapists, which may limit its accessibility in clinical settings [19]. Compared with those interventions, acupressure may offer a relatively safer and lower-cost option for cancer survivors, as it is non-invasive and can be self-administered following a short training without the need for ongoing clinical supervision [20].

Recent evidence has demonstrated potential benefits of acupressure in alleviating individual symptoms, such as fatigue, sleep issues, and depressive mood in cancer survivors (e.g. BC survivors) [20–23]. Benefits of acupressure have also been reported for other symptom clusters, such as the fatigue-pain-sleep disturbance-depression-anxiety symptom cluster and the pain-fatigue-sleep disturbance symptom cluster [11]. However, whether acupressure can effectively improve the FSDSC in BC survivors remains uncertain. Guided by the Medical Research Council Framework for Developing and Evaluating Complex Interventions [24, 25], a three-phase research programme was therefore proposed and conducted by our research team. In phase I, an evidence-based acupressure intervention protocol was developed and validated [26]. Somatic self-acupressure was chosen as the preferred modality as it allows cancer survivors to practise independently anytime and anywhere after minimal training [20]. In phase II, a pilot randomised controlled trial (RCT) was conducted to assess the feasibility and acceptability of the developed self-acupressure intervention protocol [27, 28]. A sham control group was also included to distinguish the specific and non-specific effects of acupressure [27, 29]. The phase II RCT demonstrated that self-acupressure is a feasible and safe approach with promising specific effects on the FSDSC in BC survivors [27, 28]. Based on these findings and participants’ feedback, several refinements were made to further optimise the study protocol to facilitate the study implementation of phase III RCT [27]. The Brief Fatigue Inventory (BFI) was adopted instead of the Multidimensional Fatigue Inventory (MFI) for fatigue assessment, as the MFI had a high proportion of missing values and was perceived by participants as burdensome to complete. WeChat (the most popular social media platform in China) combined with telephone contacts, was chosen as a follow-up strategy to improve participants’ adherence to home-based self-acupressure [30]. A 12-week follow-up period in the phase III RCT was implemented to capture the sustainability of self-acupressure’s beneficial effects. Moreover, a within-trial economic evaluation was also incorporated to identify the scope of costs and benefits, which could contribute to answering questions that matter most to decision makers. Given that self-acupressure showed significant positive effects on the FSDSC but not on QoL in the phase II RCT, the current study was designed as a subsequent fully powered RCT to further validate these findings. The study hypothesised that self-acupressure would alleviate the severity of the FSDSC at both cluster and single symptom levels and improve the QOL in BC survivors.

Methods

Study design and study settings

This phase III trial was a partially blinded, three-arm, sham-controlled RCT, consistent with the design of the phase II RCT. Participants were recruited from two tertiary hospitals affiliated with Zunyi Medical University. This trial has been registered at Clinicaltrials.gov (NCT06412107). A detailed study protocol was published separately [31], and the study methods were briefly outlined as follows.

Participants, randomisation, and masking

Eligible participants were those who (1) were diagnosed with early-stage female BC without distant metastases (from stages I to IIIa); (2) had completed chemotherapy for at least 1 month and up to 3 years (to capture persistent symptoms); (3) had experienced at least moderate symptom cluster with a score of 4 or above on an NRS scale (from 0 = “no” to 10 = “worst”) for fatigue, sleep disturbance, and depression in the past month; (4) had no scheduled chemotherapy or radiotherapy during the study period; and (5) consented to participate in the study [31]. Participants were ineligible if they (1) were currently taking any medications to treat fatigue, sleep disturbance, and depression; (2) had received acupressure treatment within the past 6 months; (3) had severe weakness and/or cognitive impairment that prevented them from following study procedures; and (4) involved in other studies that might potentially affect this study’s outcomes [31].

The adherence of participants to the intervention protocol in phase II RCT was suboptimal, which has confounded the observed effect size [27]. Therefore, a prior sample size for the current study was calculated based on effect sizes reported in similar studies (ranging from 0.32 to 0.96) [32–34]. An effect size of 0.32 was adopted in the G*Power calculation based on a conservative consideration. Considering an anticipated 25% attrition rate during the intervention and follow-up periods [31], a final sample size of 108 patients (36 per group) was determined.

The randomisation sequences were generated by an independent randomisation person who was involved only in the randomisation procedure, using Sealed Envelope (https://www.sealedenvelope.com/). After the eligibility was established and the baseline assessment was completed, the study investigators contacted the person to identify group allocation. Participants were randomly assigned to one of the three groups with a block size of nine (1:1:1 allocation). Due to the nature of self-acupressure intervention, blinding for the participants in the usual care group was deemed impossible, while participants in the true and sham groups would not know the self-acupressure intervention was a true or sham method. A partially blinded outcome assessment for the true and sham self-acupressure groups was also maintained as the measures in this study were self-rated. Data were analysed by an independent statistician who was blinded to group allocation. A total of 93 participants completed the study: true self-acupressure group (n = 31), sham self-acupressure group (n = 31), and usual care group (n = 31).

Intervention and procedures

Study intervention and procedures were consistent with the phase II RCT and were detailed in the published protocol [31]. All participants received an evidence-based and updated education booklet [35–37] with general strategies on the management of the FSDSC and usual care. Participants in the true self-acupressure group were required to perform acupressure at eleven acupoints (seven of which were bilateral) after training. The 11 acupoints were selected based on their relevance to alleviating fatigue, sleep disturbance, and depressive mood, including Zusanli (ST36), Sanyinjiao (SP6), Taixi (KI3), Hegu (LI4), Neiguan (PC6), Shenmen (HT7), Baihui (GV20), Qihai (CV6), Guanyuan (CV4), Yintang (EX-HN3), and Taichong (LR3). Participants in the sham self-acupressure group performed light pressure at eleven non-acupoints (seven of which were bilateral) after training. The non-acupoints were located 1–3 cm away from the acupoints used in the true self-acupressure group and were not on the meridians [26]. In addition, illustrations of the acupoint locations for the true self-acupressure and sham self-acupressure groups are provided in the supplementary figures. Both groups followed the same regimen: a 7-week period consisting of daily 36-min sessions at their convenient time, with each point stimulated for 2 min [26].

Clinical nurses at the study sites introduced this study to participants attending the hospitals for regular follow-up visits. Participants who were interested in study participation were referred to the study investigators for further eligibility assessment. Eligible participants who agreed to participate were asked to sign an informed consent and baseline assessment before group allocation. Participants assigned to either the true or sham self-acupressure group were required to attend face-to-face acupressure training sessions provided by the investigators [31]. Following the training, participants were asked to complete a return demonstration to ensure that they could perform self-acupressure correctly. The training session took about 20 min. Additional adhesive-coloured dots were applied to the acupoints to reinforce the correct locations on the body, when necessary. To further support participants’ self-acupressure at home, a home learning package with demonstration videos and charts was provided. During the 7-week intervention, daily WeChat reminders and weekly follow-ups were provided by an investigator to collect information about self-practice and any potential adverse events and to address any concerns related to self-practice at home [31].

Outcome measures

The primary endpoints were changes in the FSDSC at both the single symptom level and cluster symptom level. The secondary outcomes were QoL, safety, and relevant economic outcomes within the trial. Clinical outcomes were collected using paper-based patient-reported questionnaires at baseline (T0), the end of intervention (T1, week 7), and the follow-up (T2, week 19). The primary and secondary outcomes were specified in the published study protocol paper [31] and briefly outlined below.

Demographic and clinical information questionnaire

A questionnaire was used to collect participants’ socio-demographic data (e.g. age, occupation, educational background) and medical information (e.g. cancer stage, medical history).

The FSDSC composite score: three separated numerical rating scale (NRS)

The 0–10 NRS is a widely used tool for measuring symptom intensity and has been successfully used to assess the severity and impact of various symptoms [38]. Its straightforward format makes it easily accessible to population with different educational backgrounds. Thus, symptom severity at the cluster level (the FSDSC composite score) was calculated as the average score of three single symptom ratings on “0 = no present” and “10 = as bad as you can imagine” [39, 40].

Brief Fatigue Inventory (BFI; 9 items)

The single fatigue symptom was assessed using the BFI Chinese version [41]. Each item is scored from 0 to 10. The score of fatigue is the mean value of the nine items. A high score represents a high level of fatigue. The Cronbach’s alpha value was found to be 0.92 for the severity section and 0.90 for the interference section [41].

The Pittsburgh Sleep Quality Index (PSQI; 19 items)

The single sleep disturbance symptom was assessed using the PSQI Chinese version [42]. The total score is from 0 to 21, with higher scores indicating a higher level of sleeping difficulty (Cronbach’s alpha = 0.79) [42].

Hospital Anxiety and Depression Scale-Depression (HADS-D; 7 items)

The single depression symptom was assessed using the HADS-D Chinese version [43]. Each item is scored from 0 to 3. A total score was calculated by summing seven items, yielding a range of 0 to 21. A high score represents a high level of depression (Cronbach’s alpha = 0.93) [43].

The Functional Assessment of Cancer Therapy-Breast (FACT-B; 37 items)

The Chinese version of the FACT-B with five subscales was used to assess QoL [44]. The Cronbach’s alpha values of the subscales ranged between 0.59 and 0.85, while the test-retest reliability values for all subscales ranged from 0.82 to 0.91 [44]. Each item is scored from 0 to 4. The total score (0–148) was obtained by summing all subscale scores. A higher score indicates a better QoL [44].

Safety

Any adverse events related to self-acupressure were collected through the regular contact between the participants and study investigators across the intervention period. Any discomfort or incidents were assessed and managed by the on-site professional acupuncture practitioner.

Measurement of costs

The costs for healthcare use and intervention-related expenses were collected as outlined in the protocol [31]. Costs related to healthcare use were collected prospectively and retrospectively from patient self-reports and medical records. A self-designed form was used to record healthcare utilisation costs related to fatigue, sleep disturbance and depression (e.g. general practice visit, medicine use). The intervention-related costs were tracked and recorded in detail throughout the study.

Data analysis

Data analysis was performed generally in accordance with the published study protocol [31] with minor deviations from the original procedures. A brief description and clarification of the variations are provided in this section. Data cleaning, missing data handling, and baseline comparison were conducted in IBM SPSS Statistics 29.0.1.0. Baseline characteristics were summarised using descriptive statistics. Comparisons across the three study groups were conducted using one-way ANOVA or the Kruskal-Wallis H test for continuous variables, and the chi-square test or Fisher’s exact test for categorical variables, as appropriate [31]. Further analyses were conducted in R (version 4.3.3) using established packages for data wrangling, penalised regression, longitudinal modelling, and marginal-means estimation. Least absolute shrinkage and selection operator (LASSO) was used to select the potential covariates. The generalised estimating equations (GEE) model was conducted to identify the effects of group, time, group-by-time, and covariates on clinical outcomes. Estimated marginal means (EMMs) were obtained for each study group at each time point with covariate-adjusted outcomes. Pairwise contrasts among groups were computed within time and adjusted using Holm’s method.

The adjusted and unadjusted models yielded qualitatively similar estimates; however, the adjusted model provided more robust effect estimates by including potential covariates. This paper, therefore, primarily reported results from the adjusted model in the main analyses. Effect sizes were expressed as Cohen’s d (standardised mean differences) and interpreted using conventional small, medium, and large guidelines [45]. All statistical tests were two-sided with a significance threshold of 0.05. T2 outcome assessments were unavailable for 25 participants (7 in the true self-acupressure group, 9 in the sham self-acupressure group and 9 in the usual-care group) because data collection was temporarily interrupted during the PhD student’s transfer between universities. The GEE analyses used the available repeated outcome measurements for participants according to their originally assigned groups. For economic data, no discounting was necessary as the study period was less than 1 year. A cost-effectiveness and cost-utility analysis was not conducted due to the very limited data on healthcare resource use within trial. Moreover, clinical nurses assisting with participants’ recruitment and study investigators were provided as in-kind support in the study sites, and the only direct intervention-related cost was a one-off video production. The limited data collected did not allow for a statistically robust analysis; therefore, only descriptive reporting of intervention and healthcare resource costs is presented.

Some discrepancies between the study protocol and the actual implementation should be acknowledged. While the published protocol specified the use of SPSS for data analysis [31], both SPSS and R software were used to ensure more flexible and comprehensive statistical analysis. Least absolute shrinkage and selection operator (LASSO) with repeated cross-validation was used to estimate out-of-sample prediction error and to tally variable selection frequencies under the one-standard-error rule. The purpose was descriptive, mapping covariate importance to guide sensitivity analyses, rather than to hard-select predictors for the primary models. In addition, although the repeated-measures ANOVA was also initially planned, only the GEE was adopted to avoid over-analysis of the data. This variation was further justified by the superior robustness of GEE in handling missing data and its capacity to model within-subject correlations across repeated measurements.

Results

Participant recruitment and baseline characteristics

Participant recruitment was completed during an 8-month period (Fig. 1). A total of 108 BC survivors participated in the study, with 93 completing the study. The eligibility rate for the screened patients was 11.17% (125/1119), and the recruitment rate was 86.4% (108/125). The retention rate of the participants was 86.11% (93/108), and the attrition rate was 13.89% (15/108). The baseline characteristics of the participants (n = 108) are presented in Table 1. The enrolled participants were 48.63 ± 8.73 (mean ± SD) years old. Most of the participants (n = 84, 77.8%) were diagnosed with stage II BC. There were no statistically significant differences in socio-demographic or clinical characteristics between groups (p > 0.05).

Table 1Participant characteristics at baseline

Variables True SA Sham SA Usual care Total Statistics value p

N (%) N (%) N (%)

Education status (n = 108)

   Primary school 10 (9.3) 16 (14.8) 16 (14.8) 42 (38.9) 5.872a 0.441

   Middle school 12 (11.1) 12 (11.1) 14 (13.0) 38 (35.2)

   High school 7 (6.5) 4 (3.7) 2 (1.9) 13 (12.0)

   Tertiary education or above 7 (6.5) 4 (3.7) 4 (3.7) 15 (13.9)

Religious background (n = 108)

   No 34 (31.5) 35 (32.4) 36 (33.3) 105 (97.2) 1.874a 0.771

   Yes 2 (1.9) 1 (0.9) 0 (0.0) 3 (2.8)

Ethnicity (n = 108)

   Han 33 (30.6) 34 (31.5) 33 (30.6) 100 (92.6) 8.255a 0.366

   Miao 3 (2.8) 2 (1.9) 0 (0.0) 5 (4.6)

   Tujia 0 (0.0) 0 (0.0) 1 (0.9) 1 (0.9)

   Mongolian 0 (0.0) 0 (0.0) 1 (0.9) 1 (0.9)

   Others 0 (0.0) 0 (0.0) 1 (0.9) 1 (0.9)

Marriage status (n = 108)

   Married 34 (31.5) 30 (27.8) 35 (32.4) 99 (91.7) 4.367a 0.141

   Single 2 (1.9) 6 (5.6) 1 (0.9) 9 (8.3)

Family income (monthly) (n = 107)

   ≤ 3000 CNY 9 (8.4) 10 (9.3) 12 (11.2) 31 (29.0) 3.810a 0.717

   3001–6000 CNY 15 (14.0) 16 (15.0) 11 (10.3) 42 (39.3)

   6001–10,000 CNY 7 (6.5) 9 (8.4) 10 (9.3) 26 (24.3)

   ≥ 10,001 CNY 4 (3.7) 1 (0.9) 3 (2.8) 8 (7.5)

Residence (n = 107)

   Rural/town 31 (29.0) 32 (29.9) 29 (27.1) 92 (86.0) 1.707a 0.450

   City 5 (4.7) 3 (2.8) 7 (6.5) 15 (14.0)

Living situation (n = 108)

   Living alone 1 (0.9) 11 (10.2) 6 (5.6) 18 (16.7) 12.588b 0.050

   Living with children only 10 (9.3) 7 (6.5) 4 (3.7) 21 (19.4)

   Living with spouse only 17 (15.7) 12 (11.1) 18 (16.7) 47 (43.5)

   Living with whole family 8 (7.4) 6 (5.6) 8 (7.4) 22 (20.4)

Primary caregiver (n = 108)

   None 9 (8.3) 15 (13.9) 11 (10.2) 35 (32.4) 5.733a 0.183

   Family members 27 (25.0) 19 (17.6) 21 (20.4) 68 (63.0)

   Others 0 (0.0) 2 (1.9) 3 (2.8) 5 (4.6)

Employment status (n = 108)

   Retired 5 (4.6) 5 (4.6) 1 (0.9) 11 (10.2) 3.628a 0.467

   Employed 9 (8.3) 10 (9.3) 10 (9.3) 29 (26.9)

   Unemployed 22 (20.4) 21 (19.4) 25 (23.1) 68 (63.0)

Medical insurance (n = 108)

   No 1 (0.9) 0 (0.0) 0 (0.0) 1 (0.9) 1.836a 1.000

   Yes 35 (32.4) 36 (33.3) 36 (33.3) 107 (99.1)

Medicare payment (n = 108)

   Fully self-paid (100%) 1 (0.9) 0 (0.0) 0 (0.0) 1 (0.9) 2.559a 0.738

   Mostly self-paid (> 50% out-of-pocket) 20 (18.5) 19 (17.6) 17 (15.7) 56 (51.9)

   Partially self-paid (< 50% out-of-pocket) 15 (13.9) 17 (15.7) 19 (17.6) 51 (47.2)

   Fully reimbursed 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)

Cancer stage (n = 108)

   Stage I 5 (4.6) 3 (2.8) 6 (5.6) 14 (13.0) 4.845a 0.572

   Stage IIA 17 (15.7) 24 (22.2) 16 (14.8) 57 (52.8)

   Stage IIB 10 (9.3) 6 (5.6) 11 (10.2) 27 (25.0)

   Stage IIIA 4 (3.7) 3 (2.8) 3 (2.8) 10 (9.3)

Surgery type (n = 108)

   Modified radical mastectomy 17 (15.7) 17 (15.7) 17 (15.7) 51 (47.2) 2.191a 0.995

   Breast cancer simple excision 10 (9.3) 9 (8.3) 11 (10.2) 30 (27.8)

   Breast-conserving surgery 9 (8.3) 9 (8.3) 8 (7.4) 26 (24.1)

   Others 0 (0.0) 1 (0.9) 0 (0.0) 1 (0.9)

Chemotherapy regimen received (n = 108)

   AC/ACT 14 (13.0) 14 (13.0) 16 (14.8) 44 (40.7) 6.219a 0.634

   EC/TEC 13 (12.0) 13 (12.0) 16 (14.8) 42 (38.9)

   TC/TAC 4 (3.7) 7 (6.5) 2 (1.9) 13 (12.0)

   Others 1 (0.9) 1 (0.9) 1 (0.9) 3 (2.8)

   Unclear 4 (3.7) 1 (0.9) 1 (0.9) 6 (5.6)

Allergy history (n = 108)

   No 32 (29.6) 31 (28.7) 34 (31.5) 97 (89.8) 1.449a 0.617

   Yes 4 (3.7) 5 (4.6) 2 (1.9) 11 (10.2)

Completed chemotherapy cycles (n = 105)

   2–4 cycles 13 (12.4) 15 (14.3) 17 (16.2) 45 (42.9) 3.059b 0.555

   5–6 cycles 10 (9.5) 9 (8.6) 12 (11.4) 31 (29.5)

   7–8 cycles 11 (10.5) 12 (11.4) 6 (5.7) 29 (27.6)

Received targeted therapy (n = 106)

   No 29 (27.4) 27 (25.5) 31 (29.2) 87 (82.1) 0.992b 0.598

   Yes 6 (5.7) 8 (7.5) 5 (4.7) 19 (17.9)

Received radiotherapy (n = 105)

   No 20 (19.0) 25 (23.8) 15 (14.3) 60 (57.1) 5.833b 0.065

   Yes 15 (14.3) 10 (9.5) 20 (19.0) 45 (42.9)

Received hormone therapy (n = 106)

   No 4 (3.8) 3 (2.8) 1 (0.9) 8 (7.5) 2.011a 0.351

   Yes 31 (29.2) 32 (30.2) 35 (33.0) 98 (92.5)

Family history of BC (n = 107)

   No 34 (31.8) 34 (31.8) 35 (32.7) 103 (96.3) 0.668a 1.000

   Yes 1 (0.9) 2 (1.9) 1 (0.9) 4 (3.7)

Diabetes (n = 108)

   No 35 (32.4) 36 (33.3) 35 (32.4) 106 (98.1) 1.261a 1.000

   Yes 1 (0.9) 0 (0.0) 1 (0.9) 2 (1.9)

Hypertension (n = 108)

   No 34 (31.5) 33 (30.6) 32 (29.6) 99 (91.7) 0.784a 0.907

   Yes 2 (1.9) 3 (2.8) 4 (3.7) 9 (8.3)

FSDSC composite (mean/SD) 5.79 (1.04) 5.51 (0.87) 5.58 (0.93) 5.63 (0.95) 0.834c 0.437

BFI total (mean/SD) 4.89 (1.48) 4.56 (1.30) 4.61 (1.49) 4.69 (1.42) 0.559c 0.573

PSQI total (median/IQR) 12.00 (3) 13.00 (3) 12.50 (3) 12.50 (3) 0.487d 0.784

HADS-D (median/IQR) 10.00 (4) 10.00 (3) 9.00 (3) 10.00 (3) 1.058d 0.589

FACT-B total (mean/SD) 78.29 (16.37) 82.00 (15.34) 82.20 (15.78) 80.83 (15.79) 0.698c 0.500

Age (mean/SD) 48.25 (7.49) 47.61 (9.58) 50.03 (9.04) 48.63 (8.73) 0.738c 0.481

aFisher’s exact test, bchi-square test, cANOVA, dKruskal-Wallis H test, SA self-acupressure; chemotherapy regimen, A adriamycin, E epirubicin, C cytoxan, T taxol or taxotere, FACT-B functional assessment of cancer therapy-breast, PSQI pittsburgh sleep quality index, HADS-D hospital anxiety and depression scale-depression, BFI brief fatigue inventory, HADS-A hospital anxiety and depression scale-anxiety, BC breast cancer

Participants’ adherence to the self-acupressure intervention

Most participants in the true (91.7%) and sham self-acupressure groups (75%) performed more than 30 min daily. During the entire study period, approximately one-third of participants in the true self-acupressure group (27.8%) and less than one-sixth in the sham self-acupressure group (13.9%) completed the practice for 49 days. No significant difference in daily session (p = 0.151) and total duration (p = 0.143) was found between groups.

Effects of self-acupressure on the FSDSC

The adjusted GEE analysis showed significant group-by-time interactions for the FSDSC scores, indicating greater reductions in the true self-acupressure group compared with the sham and usual care groups over time. Compared with the usual care group and T0, the true self-acupressure group demonstrated significantly greater reductions in FSDSC scores at T1 (B = 1.87, p < 0.001) and T2 (B = 2.49, p < 0.001) (Table 3). Similarly, significant reductions were observed at T1 (B = 1.05, p < 0.001) and T2 (B = 1.38, p < 0.001) when compared with the sham self-acupressure group and T0 in the adjusted model (Table 3). A confounding effect was observed with “Medical payment (Partially self-paid)” (p = 0.037) in the model (Table 3). Between-group comparisons also indicated that the FSDSC composite score in the true self-acupressure was significantly lower than those in the usual care group and sham self-acupressure group with large effect sizes at T1 (all p < 0.001) and T2 (all p < 0.001) (Table 2 and Fig. 2). The sham self-acupressure group also significantly outperformed the usual care group at T1 (p < 0.001) and T2 (p < 0.001). The true self-acupressure group showed a large effect compared with the usual care group at T1 (− 1.68) and a stronger effect at T2 (− 2.33), while effect sizes between the sham self-acupressure group and the usual care group at T1 and T2 were − 0.84 and −1.15. Moreover, the effective sizes between true self-acupressure and sham self-acupressure groups at T1 and T2 were −0.84 and −1.19 (Table 2).

Table 3Adjusted GEE model for BFI, PSQI, HADS-D, and the FSDSC composite scores (using a first-order autoregressive correlation structure)

Predictor term Coefficient estimate Standard error Wald statistic p-value

Dependent variable: BFI total score

   Intercept 6.08 0.12 2374.81  < 0.0001

   Group = sham SA (vs true SA) −0.26 0.31 0.72 0.3951

   Group = usual care (vs true SA) −0.17 0.32 0.28 0.5943

   Time = T1 (vs T0) −1.94 0.23 73.96  < 0.0001

   Time = T2 (vs T0) −2.34 0.23 104.65  < 0.0001

   Medical payment = mostly self-paid (> 50%) −0.89 0.21 18.96  < 0.0001

   Medical payment = partially self-paid (< 50%) −1.68 0.23 54.32  < 0.0001

   Interaction: sham SA × T1 0.81 0.29 8.08 0.0045

   Interaction: usual care × T1 1.77 0.25 48.33  < 0.0001

   Interaction: sham SA × T2 1.00 0.29 12.15 0.0005

   Interaction: usual care × T2 2.16 0.26 69.22  < 0.0001

Dependent variable: PSQI total score

   Intercept 12.31 0.37 1104.76  < 0.0001

   Group = sham SA (vs true SA) 0.28 0.50 0.31 0.5752

   Group = usual care (vs true SA) −0.33 0.55 0.37 0.5417

   Time = T1 (vs T0) −2.94 0.50 34.27  < 0.0001

   Time = T2 (vs T0) −3.92 0.46 73.98  < 0.0001

   Interaction: sham SA × T1 1.47 0.60 6.04 0.0140

   Interaction: usual care × T1 2.92 0.57 26.45  < 0.0001

   Interaction: sham SA × T2 1.83 0.56 10.74 0.0010

   Interaction: usual care × T2 4.14 0.53 61.50  < 0.0001

Dependent variable: HADS-D score

   Intercept 10.17 0.41 628.90  < 0.0001

   Group = sham SA (vs true SA) −0.61 0.50 1.49 0.2215

   Group = usual care (vs true SA) −0.86 0.54 2.52 0.1121

   Time = T1 (vs T0) −3.36 0.38 78.59  < 0.0001

   Time = T2 (vs T0) −3.81 0.38 100.03  < 0.0001

   Interaction: sham SA × T1 2.25 0.49 20.90  < 0.0001

   Interaction: usual care × T1 4.00 0.50 65.08  < 0.0001

   Interaction: sham SA × T2 2.89 0.49 34.40  < 0.0001

   Interaction: usual care × T2 5.00 0.52 92.23  < 0.0001

Dependent variable: FSDSC composite score

   Intercept 6.13 0.27 522.46  < 0.0001

   Group = sham SA (vs true SA) −0.25 0.21 1.47 0.2251

   Group = usual care (vs true SA) −0.28 0.20 1.96 0.1611

   Time = T1 (vs T0) −1.95 0.16 142.43  < 0.0001

   Time = T2 (vs T0) −2.54 0.21 143.53  < 0.0001

   Income = 3001–6000 RMB −0.43 0.22 3.83 0.0503

   Income = 6001–10,000 RMB 0.16 0.24 0.43 0.5096

   Income ≥ 10,001 RMB −0.16 0.38 0.17 0.6808

   Living situation = with children only −0.23 0.30 0.60 0.4373

   Living situation = with spouse only 0.26 0.26 1.02 0.3132

   Living situation = with family (extended) 0.10 0.27 0.14 0.7099

   Medical payment = mostly self-paid (> 50%) −0.08 0.18 0.18 0.6688

   Medical payment = partially self-paid (< 50%) −0.48 0.23 4.37 0.0365

   Interaction: sham SA × T1 1.05 0.19 31.20  < 0.0001

   Interaction: usual care × T1 1.87 0.19 100.84  < 0.0001

   Interaction: sham SA × T2 1.38 0.24 31.57  < 0.0001

   Interaction: usual care × T2 2.49 0.24 111.66  < 0.0001

T0 baseline, T1 post-intervention, T2 post-follow-up, SA self-acupressure, GEE generalised estimating equations, BFI brief fatigue inventory, PSQI the pittsburgh sleep quality index, HADS-D hospital anxiety and depression scale – depression, FSDSC fatigue-sleep disturbance-depression symptom cluster, RMB Chinese yuan; interaction means, Bold entries indicate statistical significance at p < 0.05

Table 2Estimated marginal means from adjusted GEE and standardised effect sizes for pairwise contrasts (using a first-order autoregressive correlation structure)

Outcomes Time point True SA group (T) Sham SA group (S) Usual care group (U) T vs S T vs U S vs U

EMM (95% CI) EMM (95% CI) EMM (95% CI) p Effect size p Effect size p Effect size

FSDSC T0 5.87 (5.54, 6.20) 5.61 (5.28, 5.95) 5.59 (5.25, 5.93) 0.4863 0.27 0.486 0.29 0.8977 0.03

T1 3.91 (3.53, 4.30) 4.71 (4.35, 5.08) 5.50 (5.16, 5.85) 0.0015 −0.84  < 0.0001 −1.68 0.0005 −0.84

T2 3.32 (2.89, 3.76) 4.45 (4.06, 4.83) 5.53 (5.21,5.85)  < 0.0001 −1.19  < 0.0001 −2.33  < 0.0001 −1.15

BFI T0 5.23 (4.88,5.58) 4.97 (4.55,5.39) 5.06 (4.59, 5.52) 1.0000 0.18 1.0000 0.12 1.0000 −0.06

T1 3.29 (2.97,3.61) 3.84 (3.43, 4.26) 4.89 (4.46, 5.31) 0.0634 −0.39  < 0.0001 −1.12 0.0007 −0.73

T2 2.88 (2.56, 3.20) 3.62 (3.26, 3.98) 4.87 (4.50, 5.25) 0.0076 −0.52  < 0.0001 −1.40  < 0.0001 −0.88

PSQI T0 12.31 (11.58, 13.03) 12.58 (11.93,13.23) 11.97 (11.18, 12.76) 1.0000 −0.12 1.0000 0.15 0.7210 0.27

T1 9.36 (8.30, 10.42) 11.11 (10.26, 11.96) 11.94 (11.16, 12.72) 0.0238 −0.78 0.0004 −1.16 0.1562 −0.37

T2 8.39 (7.40, 9.38) 10.50 (9.74, 11.26) 12.19 (11.52, 12.87) 0.0020 −0.94  < 0.0001 −1.70 0.0020 −0.76

HADS-D T0 10.17 (9.37, 10.96) 9.56 (8.98, 10.13) 9.31 (8.60, 10.01) 0.4449 0.28 0.3392 0.40 0.5901 0.12

T1 6.81 (5.94, 7.67) 8.44 (7.85, 9.04) 9.94 (9.16, 10.73) 0.0046 −0.75  < 0.0001 −1.44 0.0046 −0.69

T2 6.36 (5.66, 7.06) 8.64 (8.17, 9.11) 10.50 (9.82, 11.18)  < 0.0001 −1.05  < 0.0001 −1.90  < 0.0001 −0.86

FACT-B total T0 78.29 (72.99, 83.58) 82.00 (77.04, 86.96) 82.20 (77.10, 87.31) 0.8860 −0.24 0.8860 −0.25 0.9556 −0.01

T1 96.45 (92.53, 100.38) 90.52 (85.47, 95.57) 80.93 (76.47, 85.39) 0.0689 0.38  < 0.0001 0.98 0.0109 0.61

T2 103.12 (98.99, 107.25) 90.91 (86.13, 95.69) 76.59 (72.61, 80.56) 0.0002 0.77  < 0.0001 1.68  < 0.0001 0.91

PWB subscale T0 19.31 (18.12, 20.49) 19.73 (18.34, 21.12) 20.19 (19.15, 21.24) 1.0000 −0.11 0.8054 −0.24 1.0000 −0.12

T1 23.03 (22.03, 24.03) 21.39 (20.09, 22.69) 20.25 (19.18, 21.31) 0.1011 0.44 0.0007 0.74 0.1828 0.31

T2 23.83 (22.89, 24.78) 21.58 (20.50, 22.67) 18.69 (17.77, 19.62) 0.0022 0.60  < 0.0001 1.37 0.0002 0.77

SWB subscale T0 13.12 (11.11, 15.13) 13.69 (12.04, 15.33) 13.60 (11.51, 15.68) 1.0000 −0.10 1.0000 −0.08 1.0000 0.02

T1 14.76 (13.39, 16.13) 14.46 (12.96, 15.96) 13.20 (11.43, 14.98) 0.7741 0.05 0.5183 0.26 0.5731 0.21

T2 16.56 (15.38, 17.75) 14.50 (12.99, 16.00) 13.11 (11.25, 14.96) 0.0680 0.35 0.0065 0.59 0.2530 0.24

EWB subscale T0 14.98 (13.60, 16.36) 14.99 (13.30, 16.69) 15.17 (13.66, 16.69) 1.0000 0.00 1.0000 −0.04 1.0000 −0.04

T1 18.70 (17.42,19.98) 17.27 (15.71, 18.83) 14.62 (13.31, 15.92) 0.1198 0.31  < 0.0001 0.90 0.0136 0.58

T2 19.67 (18.44, 20.90) 16.94 (15.52, 18.36) 13.54 (12.29, 14.78) 0.0012 0.60  < 0.0001 1.35 0.0002 0.75

FWB subscale T0 8.81 (7.48, 10.13) 10.22 (8.73, 11.71) 10.73 (9.29, 12.16) 0.3265 −0.32 0.1622 −0.43 0.6311 −0.11

T1 13.14 (11.82, 14.46) 12.19 (10.72, 13.67) 9.67 (8.58, 10.76) 0.3490 0.21 0.0003 0.78 0.0141 0.57

T2 14.81 (13.46, 16.15) 12.25 (10.83,13.67) 9.08 (8.14, 10.02) 0.0104 0.58  < 0.0001 1.29 0.0006 0.72

BCS subscale T0 23.64 (22.07, 25.21) 24.53 (23.25, 25.81) 23.88 (22.66, 25.10) 1.0000 −0.21 1.0000 −0.06 1.0000 0.16

T1 28.39 (27.38, 29.39) 26.36 (25.00, 27.72) 24.57 (23.32, 25.81) 0.0375 0.48  < 0.0001 0.91 0.0563 0.43

T2 29.81 (29.01, 30.60) 26.81 (25.47, 28.14) 23.54 (22.43, 24.64) 0.0003 0.71  < 0.0001 1.49 0.0003 0.78

EMM estimated mean, SA self-acupressure, 95% CI 95% confidence interval, GEE generalised estimating equations, T0 baseline, T1 post-intervention, T2 post-follow-up, FACT-B the functional assessment of cancer therapy – breast, PSQI the pittsburgh sleep quality index, HADS-D hospital anxiety and depression scale – depression, BFI brief fatigue inventory, FSDSC fatigue-sleep disturbance-depression symptom cluster, PWB physical well-being, SWB social well-being, EWB emotional well-being, FWB functional well-being, BCS breast cancer subscale, Bold entries indicate statistical significance at p < 0.05

Model specification: FSDSC composite score: group, time, group by time, income, living situation, medical payment

BFI score: group, time, group by time, medical payment

PSQI total score: group, time, group by time

HADS-D Score: group, time, group by time

FACT total score: group, time, group by time

PWB subscale: group, time, group by time

SWB subscale: group, time, group by time

EWB subscale: group, time, group by time, medical payment, allergy history, primary caregiver

FWB subscale: group, time, group by time

BCS subscale: group, time, group by time

Effects of self-acupressure on fatigue

The adjusted GEE model indicated that the true self-acupressure group achieved significant improvements in fatigue (assessed by BFI). Compared with the usual care group and T0, the BFI total scores were significantly lower in the true self-acupressure group at T1 (B = 1.77, p < 0.001) and T2 (B = 2.16, p < 0.001) in the adjusted GEE model (Table 3). Similar significant improvements were also indicated at T1 (B = 0.81, p < 0.005) and T2 (B = 1.0, p < 0.001) when compared with the sham self-acupressure group and T0 in the adjusted model (Table 3). Between-group comparisons at T1 and T2 found significant differences in the BFI scores between the true self-acupressure and usual care group, with large effect sizes (Table 2 and Fig. 2). The sham self-acupressure group also outperformed significantly usual care group at both T1 and T2 (all p < 0.001, Table 2). However, between-group comparisons only indicated a significantly lower BFI total score at T2 (effect size = −0.52) between true self-acupressure group and sham self-acupressure group.

Effects of self-acupressure on sleep disturbance

The results of the adjusted GEE model identified that self-acupressure resulted in a greater relief of sleep disturbance (assessed by the PSQI). The PSQI scores were significantly lower in the true self-acupressure at T1 (B = 2.92, p < 0.001) and T2 (B = 4.14, p < 0.001) when compared with the usual care group and T0. When compared with the sham self-acupressure and T0, the true self-acupressure also performed better at T1 (B = 1.47, p = 0.014) and T2 (B = 1.83, p = 0.001). At T1 and T2, between-group comparisons indicated that the PSQI scores in the true self-acupressure group were significantly lower than those in sham self-acupressure and usual care groups (all p < 0.05), see Table 2 and Fig. 2. However, sham self-acupressure group outperformed usual care only at T2 (p = 0.002, Table 2). The true self-acupressure group showed large effects compared with usual care at both T1 (− 1.16) and T2 (− 1.70), while effect sizes between sham self-acupressure and usual care were smaller at T2 (− 0.76). In addition, the effect sizes between true self-acupressure and sham self-acupressure were − 0.78 at T1 and − 0.94 at T2, confirming stronger and sustained benefits of true self-acupressure over time (Table 2).

Effects of self-acupressure on depression

The adjusted GEE results indicated that self-acupressure contributed to significant relief of depression (assessed by the HADS-D). The HADS-D scores were significantly lower in the true self-acupressure group at T1 (B = 4.00, p < 0.001) and T2 (B = 5.00, p < 0.001) when compared with the usual care group and T0 (Table 3). Compared with the sham self-acupressure and T0, the true self-acupressure outperformed significantly at both T1 (B = 2.25, p < 0.001) and T2 (B = 2.89, p < 0.001) (Table 2). At T1 and T2, between-group comparisons also found significant differences in the HADS-D scores among study groups (all p < 0.01), see Table 2 and Fig. 2. The true self-acupressure group showed a large effect compared with usual care at T1 (− 1.44) and a stronger effect at T2 (− 1.90), while effect sizes between the sham self-acupressure group and the usual care group at T1 and T2 were − 0.69 and − 0.86. Furthermore, the effective sizes between true self-acupressure and sham self-acupressure groups at T1 and T2 were −0.75 and −1.05 (Table 2).

Effects of self-acupressure on quality of life

The adjusted GEE model showed that the true self-acupressure group had significant improvements in overall QoL (assessed by the FACT-B). Specifically, the FACT-B total and all subscale scores were significantly higher at T1 (all p < 0.05) and T2 (all p < 0.001) than the usual care and T0 (Table 4).

Table 4Adjusted GEE model for FACT-B and subscales scores (using a first-order autoregressive correlation structure)

Predictor term Coefficient estimate Standard error Wald statistic p-value

Dependent variable: FACT-PWB

   Intercept 19.31 0.60 1032.90  < 0.0001

   Group = sham SA (vs true SA) 0.43 0.93 0.21 0.6462

   Group = usual care (vs true SA) 0.89 0.80 1.23 0.2676

   Time = T1 (vs T0) 3.72 0.50 55.21  < 0.0001

   Time = T2 (vs T0) 4.53 0.51 78.85  < 0.0001

   Interaction: sham SA × T1 −2.06 0.71 8.45 0.0036

   Interaction: usual care × T1 −3.67 0.79 21.69  < 0.0001

   Interaction: sham SA × T2 −2.68 0.75 12.71 0.0004

   Interaction: usual care × T2 −6.03 0.74 66.14  < 0.0001

Dependent variable: FACT-SWB

   Intercept 13.12 1.02 165.01  < 0.0001

   Group = sham SA (vs true SA) 0.56 1.32 0.18 0.6691

   Group = usual care (vs true SA) 0.47 1.47 0.10 0.7468

   Time = T1 (vs T0) 1.64 0.79 4.32 0.0378

   Time = T2 (vs T0) 3.44 0.92 13.87 0.0002

   Interaction: sham SA × T1 −0.86 0.97 0.79 0.3738

   Interaction: usual care × T1 −2.03 1.02 3.95 0.0468

   Interaction: sham SA × T2 −2.63 1.08 5.90 0.0151

   Interaction: usual care × T2 −3.93 1.13 12.14 0.0005

Dependent variable: FACT-EWB

   Intercept 15.78 1.36 135.49  < 0.0001

   Group = sham SA (vs true SA) 0.02 0.98 0.00 0.9875

   Group = usual care (vs true SA) 0.20 0.88 0.05 0.8227

   Time = T1 (vs T0) 3.72 0.66 31.77  < 0.0001

   Time = T2 (vs T0) 4.69 0.64 53.72  < 0.0001

   Medical payment = mostly self-paid (> 50%) −2.34 1.08 4.69 0.0304

   Medical payment = partially self-paid (< 50%) −0.17 1.18 0.02 0.8872

   Allergy history = yes 1.94 1.07 3.32 0.0683

   Primary caregiver = family member −1.61 0.73 4.93 0.0264

   Primary caregiver = other −1.20 1.35 0.80 0.3721

   Interaction: sham SA × T1 −1.44 0.91 2.55 0.1106

   Interaction: usual care × T1 −4.28 0.89 23.16  < 0.0001

   Interaction: sham SA × T2 −2.75 0.87 9.94 0.0016

   Interaction: usual care × T2 −6.33 0.86 53.71  < 0.0001

Dependent variable: FACT-FWB

   Intercept 8.81 0.67 170.42  < 0.0001

   Group = sham SA (vs true SA) 1.42 1.01 1.95 0.1623

   Group = usual care (vs true SA) 1.92 0.99 3.74 0.0532

   Time = T1 (vs T0) 4.33 0.59 53.14  < 0.0001

   Time = T2 (vs T0) 6.00 0.71 72.00  < 0.0001

   Interaction: sham SA × T1 −2.36 0.90 6.86 0.0088

   Interaction: usual care × T1 −5.39 0.81 44.21  < 0.0001

   Interaction: sham SA × T2 −3.97 0.94 17.86  < 0.0001

   Interaction: usual care × T2 −7.64 0.91 70.17  < 0.0001

Dependent variable: FACT-BCS

   Intercept 23.64 0.80 878.56  < 0.0001

   Group = sham SA (vs true SA) 0.89 1.03 0.75 0.3867

   Group = usual care (vs true SA) 0.24 1.01 0.06 0.8114

   Time = T1 (vs T0) 4.75 0.71 44.80  < 0.0001

   Time = T2 (vs T0) 6.17 0.74 68.93  < 0.0001

   Interaction: sham SA × T1 −2.92 0.95 9.38 0.0022

   Interaction: usual care × T1 −4.06 0.94 18.57  < 0.0001

   Interaction: sham SA × T2 −3.89 0.97 16.04 0.0001

   Interaction: usual care × T2 −6.51 0.98 44.48  < 0.0001

Dependent variable: FACT total score

   Intercept 78.29 2.69 846.81  < 0.0001

   Group = sham SA (vs true SA) 3.72 3.69 1.02 0.3136

   Group = usual care (vs true SA) 3.92 3.74 1.10 0.2945

   Time = T1 (vs T0) 18.17 2.27 64.23  < 0.0001

   Time = T2 (vs T0) 24.83 2.62 90.01  < 0.0001

   Interaction: sham SA × T1 −9.65 2.95 10.71 0.0011

   Interaction: usual care × T1 −19.44 2.88 45.52  < 0.0001

   Interaction: sham SA × T2 −15.92 3.20 24.69  < 0.0001

   Interaction: usual care × T2 −30.45 3.16 93.07  < 0.0001

T0 baseline, T1 post-intervention, T2 post-follow-up, SA self-acupressure, GEE generalised estimating equations, FACT-B the functional assessment of cancer therapy – breast, PWB physical well-being, SWB social well-being, EWB emotional well-being, FWB functional well-being, BCS breast cancer subscale, Bold entries indicate statistical significance at p < 0.05

Significant group-by-time interaction effects were observed for the FACT-B total score and the PWB, FWB, and BCS subscales at both T1 and T2 (all p < 0.05), indicating that changes over time differed across the groups (Table 4). The pattern of group-by-time interactions for the FACT-B subscales varied according to subscale, comparator and assessment point; detailed estimates and p values are presented in Table 4. Between-group comparisons at T1 and T2 showed that FACT-B total scores were significantly higher in both the true and sham self-acupressure groups compared with the usual care group (all p < 0.05); see Table 2 and Fig. 2. The difference in FACT-B total score between the true and sham self-acupressure groups was not statistically significant at T1 (p = 0.069) but was statistically significant at T2 (p = 0.0002). A confounding effect was observed with “primary caregiver (family member)” in the model (p = 0.026) (Table 4). At T1, true self-acupressure showed a large advantage over usual care (effect size = 0.98) and a moderate, though nonsignificant, trend compared with sham (effect size = 0.38). By T2, effects were more pronounced: true self-acupressure exceeded sham (effect size = 0.77) and usual care (effect size = 1.68). Sham also outperformed usual care (effect size = 0.91). These results confirm meaningful, sustained QoL benefits (Table 2).

Safety of the self-acupressure

No adverse event was reported in this study.

Costs within trial

During the trial, only one participant in each of the true self-acupressure and usual care groups reported healthcare resource use (medications plus clinical visits for symptom relief), with a sum cost of ¥931.06 and ¥2247.44, respectively. The one-off cost of video production for the true self-acupressure group was ¥2879. Given the very limited number of participants reporting resource use, the cost-effectiveness of self-acupressure intervention is not conclusive.

Discussion

This study is a fully powered RCT to provide evidence, to the best of our knowledge, on the sustained effects of self-acupressure in managing the FSDSC and QoL in BC survivors. The results revealed statistically significant reductions in the FSDSC and improvements in QoL. These results highlight the benefits of incorporating self-acupressure into existing self-management programmes for BC survivorship care.

The encouraging results from the phase II study provided a foundation for this current full-scale trial. With a larger sample size and improved follow-up strategies, this current study confirmed and strengthened the previously observed beneficial effects of self-acupressure on FSDSC management among BC survivors and demonstrated that these effects were sustained over time through 12 weeks’ follow-up. The beneficial effects observed on individual symptoms in this study align with recent evidence confirming the effectiveness of acupressure intervention in improving the cancer-related fatigue, depressive distress and sleep issues [22, 46, 47]. This study also demonstrated significant positive effects at a cluster level, and the effect sizes of self-acupressure on the cluster-level were stronger than single symptom level. Due to the limited studies on this specific topic, direct comparisons to studies with the same participants and outcomes are not possible. However, our study findings are comparable to those from similar studies of applying self-acupressure in other types of cancer survivors and with those using other non-pharmacological interventions targeting BC survivors. For example, a pilot RCT with sham-control (38 participants in each group) examined the effects of self-acupressure and reported positive outcomes in insomnia-depression-anxiety symptom cluster from cancer survivors [40]. Although another study targeting BC survivors demonstrated the preliminary positive effects of complex multimodal intervention in alleviating the FSDSC, it required substantial professional involvement [48]. Self-acupressure in our study produced comparable benefits while being simpler and may offer a more accessible and practical alternative for symptom management for BC survivors. By addressing the symptoms of fatigue, sleep disturbance, and depression as an integrated cluster, self-acupressure showed beneficial effects at both the cluster level and on individual level. These findings also suggest potential influences on interrelationships among the three symptoms, although the underlying mechanisms require further investigation. The results of this study may imply that a cluster-based management approach could provide comprehensive and sustained benefits for cancer survivors compared with interventions targeting single symptoms alone.

This study demonstrated significant improvements in QoL, with the true self-acupressure group showing greater improvement than the other groups. This finding is consistent with previous study showing that self-acupressure is an effective method to improve QoL among BC survivors [32, 49]. Given the confirmed association between the FSDSC and QoL in the phase II RCT [27], the positive effect of self-acupressure on QoL may be attributed to improvements in physical, emotional, and functional well-being related to reduced fatigue, sleep disturbance, and depression. However, no significant between-group difference was observed in social well-being between the true self-acupressure and usual care groups after the intervention. A possible explanation is that Chinese culture places a strong emphasis on family bonds, with family members often serving as the main source of social support [50]. Previous study found that patients in China commonly seek support from family members and peer patients [51]. Therefore, participants in the intervention and control groups may have received similar levels of family support after diagnosis which may further contribute to the limited impact of the intervention on social well-being.

Consistent with the previous pilot study [27], the study results indicated that both true self-acupressure and sham self-acupressure were superior to usual care in alleviating the FSDSC at both the cluster and individual symptom levels. Of which, the true self-acupressure group was more effective in alleviating the FSDSC than other groups. Moreover, the sham self-acupressure group also outperformed the usual care group in managing the FSDSC. This finding can be explained by the potential positive changes from psychosocial factors, including participants’ strong desire for symptom relief or participants’ expectations of the acupressure intervention or active self-engagement in modulating symptom perception [52]. The reduction in the FSDSC in this current study may not be entirely attributed to specific therapeutic effects of self-acupressure but could also involve placebo effects. Even when applied to non-acupoints, it may trigger beneficial psychological, physiological, and behavioural responses [29, 53]. Such placebo effects highlight the importance of using these psychological factors in designing symptom management interventions for BC survivors. The findings regarding the placebo effects of acupressure are consistent with a previous systematic review on sham acupressure [29].

Medical payment showed a significant association with fatigue and the FSDSC in the adjusted GEE model. Specifically, partial self-payment (less than 50%) of medical payment was associated with a lower FSDSC composite score (p = 0.037). Such findings indicated that financial burden may contribute to symptom burden in BC survivors. However, self-acupressure remained significantly effective after accounting for these covariates. Financial burden may increase psychological distress, limit access to supportive care resources, and reduce adherence to self-management activities, thereby exacerbating the overall symptom experience [54, 55]. Therefore, future studies could consider incorporating stratified randomisation based on medical payment type to achieve better group balance and improve internal validity.

In terms of the cost-effectiveness of self-acupressure in this study, making robust conclusions is challenging due to the limited data within the trial. It is also possible some over-the-counter medicines may have been used but not captured as not all information on medicine use was available in the medical records. In addition, relying on participants’ self-report and recall may have led to incomplete data. A more comprehensive economic evaluation, incorporating costs from the healthcare provider perspective, the societal perspective, and the patient perspective, is warranted to further assess the value of acupressure in both clinical practice and future trials [56].

Implications

Findings from this phase III RCT, together with previous related studies [26–28], supported that self-acupressure is a feasible, safe, and effective intervention for symptom management in BC survivors. Self-acupressure requires minimal pre-training and can be performed independently at any time in any setting once the technique has been learned. Such features of self-acupressure make it highly appropriate for primary healthcare where accessible, self-managed, low-burden strategies are essential. Moreover, the evidence-based self-acupressure intervention protocol [26] provides a standardised approach that can be translated into routine practice to support symptom management for BC survivors. Although participants generally adhered to the daily 36-min practice in this study, sustaining continuous practice for the full 49 days was challenging. In real-world implementation, individualised adjustments in sessions could be made based on participants’ functional status, symptom burden, and personal preferences, thereby supporting adherence to the overall intervention regimen.

Limitations

This study has several limitations. The assessment of clinical outcomes relied on participants’ self-reported measures, which could be quite subjective sometimes despite the fact that the measures used in this study are well-validated. Future studies should include objective measures, such as biomarkers or actigraphy, to supplement self-reported assessments of symptoms. Due to the lack of a validated instrument for assessing the FSDSC at the cluster level, the composite score was calculated as the average of three separate NRS ratings. Despite this method was used in previous studies [39, 40], this approach implicitly assumes equal weighting of each symptom and may not reflect potential interactions among symptoms. While potential mechanisms of self-acupressure (e.g. decreased pro-inflammatory cytokines) can be theoretically substantiated, these cytokines were not directly assessed within the scope of the study.

Conclusion

This clinical trial was theory-driven and practice-based, following an evidence-based intervention protocol with promising findings and subsequent protocol refinements. It provides robust evidence on the positive effects of self-acupressure in improving the FSDSC and QoL in BC survivors. Self-acupressure could be safely integrated into clinical practice for BC survivors with symptom burden.

Supplementary information

Below is the link to the electronic supplementary material.

Acknowledgements

The authors are grateful to participants in this study. We also sincerely thank to the on-site clinical staff—Ms. Xiaoting Gan, Ms. Sisi Gu, Ms. Xiaoming Cheng, Mr. Taolang Li, Ms. Suhong Sun, and Mr. Kang Hu—for their support in this study.

Author contribution

Mengyuan Li: writing-original draft, conceptualization, methodology, data interpretation, quality assurance. Wai Hang Kwok: writing-review & editing, conceptualization, methodology, data interpretation, quality assurance. Tao Wang: writing – review & editing, conceptualization, methodology, supervision. Renli Deng: writing-review & editing, methodology, resources, study coordination. Haiying Wang: writing-review & editing, methodology, supervision. Li-Qun Yao: writing-review & editing, methodology, supervision. Linda Deravin: Writing -review & editing, supervision. Li Yang: resources, study coordination. Xiaojing Tian: resources, study coordination. Jing-Yu (Benjamin) Tan: Writing- review & editing, conceptualization, methodology, supervision.

Funding

Open Access funding enabled and organized by CAUL and its Member Institutions This study is funded by the Charles Darwin International PhD Scholarship (CDIPS) and the University of Southern Queensland International Fees Research Scholarship.

Data availability

Individual participant data are not available for sharing in accordance with the ethical requirements of the study.

Declarations

Ethics approval

This study was performed in line with the principles of the Declaration of Helsinki. This study was reviewed and approved by the Human Research Ethics Committees at the Charles Darwin University (H22110) and the University of Southern Queensland (ETH2025-0294, ETH2025-0584*, ETH2026-0469 (Milestone)), as well as the Clinical Trial Ethics Committees at the Affiliated Hospital of Zunyi Medical University (KLL-2023-594) and the Second Affiliated Hospital of Zunyi Medical University (KYLL-2023-058).

Consent to participate

Informed consent was obtained from all individual participants included in the study.

Consent for publication

Consent for publication was obtained from all individual participants included in the study without any identifiable personal information.

Competing interests

The authors declare no competing interests.

Footnotes

Mengyuan Li and Wai Hang Kwok are both first authors.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1.World Health Organization (2024) Global cancer burden growing, amidst mounting need for services. Available at: https://www.who.int/news/item/01-02-2024-global-cancer-burden-growing%2D%2Damidst-mounting-need-for-services. Accessed 15 Oct 2025

2.Kim J, Harper A, McCormack V, Sung H, Houssami N, Morgan E, Mutebi M, Garvey G, Soerjomataram I, Fidler-Benaoudia MM (2025) Global patterns and trends in breast cancer incidence and mortality across 185 countries. Nat Med 31:1154–1162. 10.1038/s41591-025-03502-3

3.Zeng H, Zheng R, Sun K, Zhou M, Wang S, Li L, Chen R, Han B, Liu M, Zhou J, Xu M, Wang L, Yin P, Wang B, You J, Wu J, Wei W, He J (2024) Cancer survival statistics in China 2019-2021: a multicenter, population-based study. J Natl Cancer Cent 4:203–213. 10.1016/j.jncc.2024.06.005

4.He X, Ng M, Choi K, Li L, Zhao W, Zhang M, So W (2022) Synergistic interactions among fatigue, sleep disturbance, and depression in women with breast cancer: a cross-sectional study. Oncol Nurs Forum 49:243–254. 10.1188/22.ONF.243-254

5.Jing F, Zhu Z, Qiu J, Tang L, Xu L, Xing W, Hu Y (2023) Symptom clusters and quality of life in patients with breast cancer receiving endocrine therapy in China. Semin Oncol Nurs 39:151443. 10.1016/j.soncn.2023.151443

6.So WK, Law BM, Ng MS, He X, Chan DN, Chan CW, McCarthy AL (2021) Symptom clusters experienced by breast cancer patients at various treatment stages: a systematic review. Cancer Med 10:2531–2565. 10.1002/cam4.3794

7.Kim YB, Lee IJ, Byun HK, Choi YY, Hong B, Lee J (2024) Symptom network and quality of life of breast cancer patients receiving multimodal cancer treatment: cross-sectional study. Eur J Oncol Nurs 71:102661. 10.1016/j.ejon.2024.102661

8.Barsevick A (2016) Defining the symptom cluster: how far have we come? Semin Oncol Nurs 32:334–350. 10.1016/j.soncn.2016.08.001

9.Kwekkeboom KL, Tostrud L, Costanzo E, Coe CL, Serlin RC, Ward SE, Zhang Y (2018) The role of inflammation in the pain, fatigue, and sleep disturbance symptom cluster in advanced cancer. J Pain Symptom Manage 55:1286–1295. 10.1016/j.jpainsymman.2018.01.008

10.Kwekkeboom K (2020) Guideline-recommended symptom management strategies that cross over two or more cancer symptoms. Oncol Nurs Forum 47(5):498–511. 10.1188/20.ONF.498-511

11.Li MY, Yao LQ, Liu XL, Tan JYB, Wang T (2024) Effects of nonpharmacological interventions on symptom clusters in breast cancer survivors: a systematic review of randomized controlled trials. Asia-Pac J Oncol Nurs 11:100380. 10.1016/j.apjon.2024.100380

12.Wong WM, Chan DN, He X, So WK (2023) Effectiveness of pharmacological and nonpharmacological interventions for managing the fatigue–sleep disturbance–depression symptom cluster in breast cancer patients undergoing chemotherapy: a systematic review. Cancer Nurs 46:E70–E80. 10.1097/NCC.0000000000001048

13.Khosravi N, Stoner L, Farajivafa V, Hanson ED (2019) Exercise training, circulating cytokine levels and immune function in cancer survivors: a meta-analysis. Brain Behav Immun 81:92–104. 10.1016/j.bbi.2019.08.187

14.Liu Z, Jiao Y, Yu T, Wang H, Zhang Y, Liu D, Xu Y, Guan Q, Lu M (2022) A review on the immunomodulatory mechanism of acupuncture in the treatment of inflammatory bowel disease. Evidence-based complementary and alternative medicine : eCAM 2022:8528938. 10.1155/2022/8528938

15.Xu Y, Hong S, Zhao X, Wang S, Xu Z, Ding S, Zhang K, Zhang Y, Xu L, Yu N, Zhao T, Yan Y, Yang F, Liu Y, Yu K, Liu B, Guo Y, Pang G (2018) Acupuncture alleviates rheumatoid arthritis by immune-network modulation. Am J Chin Med 46:997–1019. 10.1142/S0192415X18500520

16.Huang CC, Kotha P, Tu CH, Huang MC, Chen YH, Lin JG (2024) Acupuncture: a review of the safety and adverse events and the strategy of potential risk prevention. Am J Chin Med 52:1555–1587. 10.1142/S0192415X24500617

17.Kim SA, Lee JS, Kim TH, Lee S, Lee JD, Kang JW (2024) The reporting quality of acupuncture-related traumatic adverse events: a systematic review of case studies in Korea. BMC Complement Med Ther 24:120. 10.1186/s12906-024-04421-5

18.Bower JE (2014) Cancer-related fatigue--mechanisms, risk factors, and treatments. Nat Rev Clin Oncol 11:597–609. 10.1038/nrclinonc.2014.127

19.Semenenko E, Banerjee S, Olver I, Ashinze P (2023) Review of psychological interventions in patients with cancer. Support Care Cancer 31:210. 10.1007/s00520-023-07675-w

20.Cheng HL, Yeung WF, Wong HF, Lo HT, Molassiotis A (2023) Self-acupressure for symptom management in cancer patients: a systematic review. J Pain Symptom Manage 66:e109–e128. 10.1016/j.jpainsymman.2023.03.002

21.Doğan MD (2024) The effect of acupressure on fatigue in cancer patients: a meta-analysis study. Indian. J Palliat Care 30. 10.25259/IJPC_95_2023

22.Yang C, Huang Y, Ling W, Cheung DST, Lee JJ (2025) The effectiveness of acupressure on sleep quality in cancer patients: a systematic review and meta-analysis of randomised controlled trials. J Clin Nurs 34:4560–4577. 10.1111/jocn.17707

23.Zhang C, Chen J, Zhang Y, Yan C (2025) The effect of acupressure on chemotherapy-induced anxiety and depression: a systematic review and meta-analysis. Wirkung von Akupressur auf chemotherapiebedingte Angst und Deprimiertheit: Eine systematische Übersicht und Metaanalyse. Complement Med Res 32:55–66. 10.1159/000542022

24.Skivington K, Matthews L, Simpson SA, Craig P, Baird J, Blazeby JM, Boyd KA, Craig N, French DP, McIntosh E, Petticrew M, Rycroft-Malone J, White M, Moore L (2024) A new framework for developing and evaluating complex interventions: update of Medical Research Council guidance. Int J Nurs Stud 154:104705. 10.1016/j.ijnurstu.2024.104705

25.Skivington K, Matthews L, Simpson SA, Craig P, Baird J, Blazeby JM, Boyd KA, Craig N, French DP, McIntosh E, Petticrew M, Rycroft-Malone J, White M, Moore L (2021) A new framework for developing and evaluating complex interventions: update of Medical Research Council guidance. BMJ (Clinical Research ed) 374:n2061. 10.1136/bmj.n2061

26.Tan JY, Wang T, Zhao I, Polotan MJ, Eliseeva S (2022) An evidence-based somatic acupressure intervention protocol for managing the breast cancer fatigue-sleep disturbance-depression symptom cluster: development and validation following the Medical Research Council framework. Int J Environ Res Public Health 19:11934. 10.3390/ijerph191911934

27.Li MY, Kwok SWH, Tan JYB, Bressington D, Liu XL, Wang T, Chen SL (2023) Somatic acupressure for the fatigue-sleep disturbance-depression symptom cluster in breast cancer survivors: a phase II randomized controlled trial. Eur J Oncol Nurs 66:102380. 10.1016/j.ejon.2023.102380

28.Wang T, Tan JB, Liu XL, Zheng SL, Zhao I, Eliseeva S, Polotan MJ, Cheng HL, Huang HQ (2022) Implementing an evidence-based somatic acupressure intervention in breast cancer survivors with the symptom cluster of fatigue, sleep disturbance and depression: study protocol of a phase II randomised controlled trial. BMJ Open 12:e054597. 10.1136/bmjopen-2021-054597

29.Tan JY, Suen LK, Wang T, Molassiotis A (2015) Sham acupressure controls used in randomized controlled trials: a systematic review and critique. PLoS ONE 10:e0132989. 10.1371/journal.pone.0132989

30.Sui Y, Wang T, Wang X (2020) The impact of WeChat app-based education and rehabilitation program on anxiety, depression, quality of life, loss of follow-up and survival in non-small cell lung cancer patients who underwent surgical resection. Eur J Oncol Nurs 45:101707. 10.1016/j.ejon.2019.101707

31.Li M, Wang T, Liu XL, Deng RL, Kwok WH, Yao LQ, Tan JYB (2025) Somatic acupressure for the management of the fatigue–sleep disturbance–depression symptom cluster in breast cancer survivors: a study protocol for a phase III randomised controlled trial. BMJ Open 15:e089515. 10.1136/bmjopen-2024-089515

32.Zick SM, Sen A, Wyatt GK, Murphy SL, Arnedt JT, Harris RE (2016) Investigation of 2 types of self-administered acupressure for persistent cancer-related fatigue in breast cancer survivors: a randomized clinical trial. JAMA Oncol 2:1470–1476. 10.1001/jamaoncol.2016.1867

33.Cheung DST, Yeung WF, Chau PH, Lam TC, Yang M, Lai K, Ip CY, Lao L, Lin CC (2022) Patient-centred, self-administered acupressure for Chinese advanced cancer patients experiencing fatigue and co-occurring symptoms: a pilot randomised controlled trial. Eur J Cancer Care 31:e13314. 10.1111/ecc.13314

34.Khanghah AG, Rizi MS, Nabi BN, Adib M, Leili EKN (2019) Effects of acupressure on fatigue in patients with cancer who underwent chemotherapy. J Acupunct Meridian Stud 12:103–110. 10.1016/j.jams.2019.07.003

35.Andersen BL, Lacchetti C, Ashing K, Berek JS, Berman BS, Bolte S, Dizon DS, Given B, Nekhlyudov L, Pirl W, Stanton AL, Rowland JH (2023) Management of anxiety and depression in adult survivors of cancer: ASCO guideline update. J Clin Oncol 41:3426–3453. 10.1200/JCO.23.00293

36.Fabi A, Bhargava R, Fatigoni S, Guglielmo M, Horneber M, Roila F, Weis J, Jordan K, Ripamonti CI, Guidelines Committee ESMO (2020) Cancer-related fatigue: ESMO Clinical Practice Guidelines for diagnosis and treatment. Ann Oncol 31:713–723. 10.1016/j.annonc.2020.02.016

37.Grassi L, Zachariae R, Caruso R, Palagini L, Campos-Ródenas R, Riba MB, Lloyd-Williams M, Kissane D, Rodin G, McFarland D, Ripamonti CI, Santini D, Guidelines Committee ESMO (2023) Insomnia in adult patients with cancer: ESMO Clinical Practice Guideline. ESMO Open 8:102047. 10.1016/j.esmoop.2023.102047

38.Oldenmenger WH, Pleun J, de Klerk C, van der Rijt CC (2013) Cut points on 0–10 numeric rating scales for symptoms included in the Edmonton Symptom Assessment Scale in cancer patients: a systematic review. J Pain Symptom Manage 45:1083–1093. 10.1016/j.jpainsymman.2012.06.007

39.Nguyen LT, Alexander K, Yates P (2018) Psychoeducational intervention for symptom management of fatigue, pain, and sleep disturbance cluster among cancer patients: a pilot quasi-experimental study. J Pain Symptom Manage 55:1459–1472

40.Hoang HTX, Molassiotis A, Chan CW, Vu AH, Bui PT (2022) Pilot randomized sham-controlled trial of self-acupressure to manage the symptom cluster of insomnia, depression, and anxiety in cancer patients undergoing chemotherapy. Sleep Breath 26:445–456. 10.1007/s11325-021-02370-8

41.Wang XS, Hao XS, Wang Y, Guo H, Jiang YQ, Mendoza TR, Cleeland CS (2004) Validation study of the Chinese version of the Brief Fatigue Inventory (BFI-C). J Pain Symptom Manage 27:322–332. 10.1016/j.jpainsymman.2003.09.008

42.Ho RT, Fong TC (2014) Factor structure of the Chinese version of the Pittsburgh sleep quality index in breast cancer patients. Sleep Med 15:565–569. 10.1016/j.sleep.2013.10.019

43.Li Q, Lin Y, Hu C, Xu Y, Zhou H, Yang L, Xu Y (2016) The Chinese version of hospital anxiety and depression scale: psychometric properties in Chinese cancer patients and their family caregivers. Eur J Oncol Nurs 25:16–23. 10.1016/j.ejon.2016.09.004

44.Wan C, Zhang D, Yang Z, Tu X, Tang W, Feng C, Wang H, Tang X (2007) Validation of the simplified Chinese version of the FACT-B for measuring quality of life for patients with breast cancer. Breast Cancer Res Treat 106:413–418. 10.1007/s10549-007-9511-1

45.Cohen J (1992) Statistical power analysis. Curr Dir Psychol Sci 1:98–101. 10.1111/1467-8721.ep10768783

46.Hsieh SH, Wu CR, Romadlon DS, Hasan F, Chen PY, Chiu HY (2021) The effect of acupressure on relieving cancer-related fatigue: a systematic review and meta-analysis of randomized controlled trials. Cancer Nurs 44:E578–E588. 10.1097/NCC.0000000000000997

47.Yang LH, Li SY, Yang Y, Yang SN, Guo HL, Hu MM, Duan PB, Xia C (2025) Effectiveness of acupressure in managing the pain-fatigue-sleep disturbance-depression symptom cluster in patients with cancer: a systematic review and meta-analysis of randomized controlled trials. J Integr Complement Med 31:860–875. 10.1089/jicm.2024.0914

48.Wong WM, Chan DN, Choi KC, Choy YP, So WK (2023) A multi-modal intervention for managing the fatigue–sleep disturbance–depressed mood symptom cluster in breast cancer patients undergoing chemotherapy: a pilot study. Asia Pac J Oncol Nurs 10:100269. 10.1016/j.apjon.2023.100269

49.Kim SY, Park JS (2021) The effect of self-acupressure on peripheral neuropathy, disturbance in daily activity, and quality of life in breast cancer patients undergoing chemotherapy. Asian Oncol Nurs 21:129–139. 10.5388/aon.2021.21.3.129

50.So WK, Leung DY, Ho SS, Lai ET, Sit JW, Chan CW (2013) Associations between social support, prevalent symptoms and health-related quality of life in Chinese women undergoing treatment for breast cancer: a cross-sectional study using structural equation modelling. Eur J Oncol Nurs 17:442–448. 10.1016/j.ejon.2012.11.001

51.Cheng H, Sit JW, Chan CW, So WK, Choi KC, Cheng KK (2013) Social support and quality of life among Chinese breast cancer survivors: findings from a mixed methods study. Eur J Oncol Nurs 17:788–796. 10.1016/j.ejon.2013.03.007

52.Williams CM, Kamper SJ (2012) Non-specific effects of acupuncture - does the 'placebo' effect play an important role? Br J Sports Med 46:578–579. 10.1136/bjsports-2012-09122

53.Miller FG, Colloca L (2010) Semiotics and the placebo effect. Perspect Biol Med 53(4):509–516

54.Benedict C, Fisher S, Schapira L, Chao S, Sackeyfio S, Sullivan T, Pollom E, Berek JS, Kurian AW, Palesh O (2022) Greater financial toxicity relates to greater distress and worse quality of life among breast and gynecologic cancer survivors. Psychooncology 31:9–20. 10.1002/pon.576

55.Perry LM, Hoerger M, Seibert K, Gerhart JI, O’Mahony S, Duberstein PR (2019) Financial strain and physical and emotional quality of life in breast cancer. J Pain Symptom Manage 58:454–459. 10.1016/j.jpainsymman.2019.05.011

56.Molassiotis A, Russell W, Hughes J, Breckons M, Lloyd-Williams M, Richardson J, Hulme C, Brearley S, Campbell M, Garrow A, Ryder W, Ryder WD (2013) The effectiveness and cost-effectiveness of acupressure for the control and management of chemotherapy-related acute and delayed nausea: assessment of nausea in chemotherapy research (ANCHoR), a randomised controlled trial. Health Technology Assessment (Winchester, England) 17(26):1

Associated Data

Supplementary Materials

Data Availability Statement

Individual participant data are not available for sharing in accordance with the ethical requirements of the study.